Medical Malpractice

Medical Malpractice

Frank A. Sloan and Lindsey M. Chepke

The MIT Press

Cambridge, Massachusetts
London, England
© 2008 Massachusetts Institute of Technology

All rights reserved. No part of this book may be reproduced in any form by any
electronic or mechanical means (including photocopying, recording, or informa-
tion storage and retrieval) without permission in writing from the publisher.

For information about special quantity discounts, please email special_sales

@mitpress.mit.edu

This book was set in Sabon by SNP Best-set Typesetter Ltd., Hong Kong.
Printed and bound in the United States of America.

Library of Congress Cataloging-in-Publication Data

Sloan, Frank A.
Medical malpractice / Frank A. Sloan and Lindsey M. Chepke.
p. ; cm.
Includes bibliographical references and index.
ISBN 978-0-262-19572-0 (hardcover : alk. paper) 1. Medical personnel–
Malpractice–United States. 2. Insurance, Liability–Malpractice–United
States. I. Chepke, Lindsey M. II. Title.
[DNLM: 1. Insurance, Liability–legislation & jurisprudence–United States.
2. Malpractice–legislation & jurisprudence–United States. 3. Insurance,
Liability–economics–United States. 4. Malpractice–economics–United States.
W 33 AA1 S6m 2007]
KF2905.3.S66 2007
344.7304′11–dc22
2007019119

10 9 8 7 6 5 4 3 2 1
Contents

Preface vii
1 Introduction 1
2 Why the Crises in Medical Malpractice? 27
3 An Increased Threat of Lawsuits and Higher Premiums:
The Consequences 51
4 Governments’ Responses to Medical Malpractice Crises—and
Their Effects 85
5 Ceilings on Nonmonetary and Total Losses 107
6 Compensating Plaintiffs’ Attorneys 135
7 Juries and Health Courts 163
8 Patient Safety and Medical Malpractice 189
9 Medical Malpractice Insurance and Insurance Regulation 217
10 Reinsurance 247
11 No-Fault for Medical Injuries 277
12 Reforms: What Can Be Done 309
Notes 337
References 401
Index 441
Preface

For more than three decades, medical malpractice has been at the top of
the U.S. policy agenda from time to time. Medical malpractice crises
begin when medical malpractice premiums spike and some insurers with-
draw from the market. Major changes in the price and availability of
coverage in turn lead to pressures on public policymakers for relief.
Policymakers are pressured to implement quick solutions, typically on
the basis of little objective information. After a few years, premiums
stabilize and insurers return to the market. The disruptions, both real
and imagined, are clearly unfortunate. But there is also a benefit. Although
the controversies and many of the solutions do not seem to vary much,
with each crisis we learn more about how the system really operates.
The “system” reflects the actions of citizens as patients (more frequently)
and plaintiffs (much less frequently), physicians, hospitals, and other
health care providers, lawyers for plaintiffs and defendants, judges,
insurers, and governments.
This book is about what we do and do not know about the ways in
which this complex system operates, the important lessons to be learned
from the past, and recommended policies for the future. We have written
the book for a broad readership, ranging from individuals (as patients
and voters), physicians, hospital administrators, attorneys, and insurers
to state legislators, as well as the advocates for the various policy posi-
tions on medical malpractice. Scholars in the various disciplines will find
a lot of material on medical malpractice, with documentation, in the
endnotes. Although the authors are an economist and a lawyer, respec-
tively, we have tried to write this book in nontechnical language. Some
issues, such as those pertaining to insurance, are inherently technical.
viii Preface

Major funding for this research is from an Investigator Award from

the Robert Wood Johnson Foundation to Frank Sloan. The Investigator
Award program is a very valuable source of research support. Even
investigator-initiated research grants do not generally give scholars
the scope to explore paths totally unanticipated at the time the grant
proposal is written, as the Investigator Awards do.
A purpose of these awards is to give scholars the time to think broadly
about health care issues. In this spirit, we have ventured into methods
of paying professionals in chapter 6, and no-fault insurance as applied
to areas outside of personal health care services, such as for motor vehicle
liability and workers’ compensation, in chapter 11. The material on
insurance cycles and reinsurance in chapters 2 and 10 applies much more
broadly than only to medical malpractice. We cite experience with spe-
cialized courts from other fields, such as family and bankruptcy courts,
in our analysis of the desirability and feasibility of health courts in
chapter 7. This analysis also draws on evidence from product liability.
Our research also benefited from a grant from the Pew Charitable
Trust and help from the principal investigator of the Pew grant, William
Sage, who was on the faculty of Columbia Law School when the funding
was active. Parts of the chapters on reinsurance and medical no-fault
programs come from work funded by the Pew grant.
Kenneth Morris, senior vice president, chief financial officer, and trea-
surer of the Duke University Health System, provided helpful perspec-
tives on Duke’s approach to and experiences with its hospitals and
physicians, as well as important details on how reinsurance markets
function in practice.
Medical Malpractice
1
Introduction

Focus of the Book

In the first years of this century, for the third time in three decades, the
United States faced a crisis over medical malpractice. Each crisis has had
its unique features, but each crisis shares important attributes with its
predecessors as well. In large part because there are always new persons
in the policy arena, old discussions and experiences tend to be forgotten,
and the same old questions are asked anew. Since each crisis stimulates
new research, we know increasingly more about medical malpractice and
medical malpractice insurance. In fact, we know much more than is
commonly acknowledged—not that we have all the answers. For example,
we do not know precisely why insurance crises reoccur, except that it is
inevitable that they will. Also, evidence on the effects of past reforms,
other than the ones that have been implemented repeatedly by state leg-
islatures since the mid-1970s, is limited.
Interestingly, there is little fundamental disagreement among
most researchers active in this field about the nature and extent of the
“medical malpractice problem” and even about many of the solutions.
However, there are some important differences in emphasis in policy
recommendations. Virtually all experts would agree that the current
system is “ill-suited” for deterring injuries and compensating injury
victims efficiently. These same experts would agree that the issues
in medical malpractice are at a crossroads, in that the policy responses
to previous crises have not altered the fundamental incentives of the
participants in the care, claiming, claims resolution, and insurance
processes.
2 Chapter 1

By contrast, there are substantial differences in views among the public

at large, health professionals who are on the front line but seldom
conduct quantitative analysis, medical malpractice insurers, large busi-
nesses in general, trial lawyers, and consumer groups. In large part, these
differences in perceptions and recommendations for reform reflect the
adage “Where you stand depends on where you sit.” In addition, to a
considerable extent these differences reflect reliance on personal experi-
ences versus reliance on information assembled and analyzed by methods
generally accepted by the scientific community.
While we identify deficiencies of the existing system, we do not propose
sweeping reforms. Rather, our emphasis is on achievable reforms,
although achieving any meaningful reform is admittedly not an easy task.
In deciding what is and what is not achievable, it is essential to consider
the institutional and political context which has guided how health care
is financed and delivered in the United States. The experiences of other
parts of the U.S. economy or of other countries are important to con-
sider, but it is doubtful that these experiences would be replicated by the
U.S. health care system, at least without substantial modifications.
An objective of this book is to make empirical evidence available, in
a balanced fashion, to readers who do not typically read scholarly jour-
nals or attend academic conferences. Several books have been written on
medical malpractice. Most are now over one or even two decades old.
In addition, there has been much empirical research and writing since
about 2000 which is reflected in this book for the first time.
Our evaluation of current practices is based on a substantial amount
of empirical evidence and several premises. First, considerable research
has been conducted on many aspects of medical malpractice, such as the
claims resolution process, juries, defensive medicine, and the effects of
tort reforms, especially those termed “first-generation tort reforms.”1
The research consistently indicates that rhetoric of the stakeholders on
both sides of the debate has been highly exaggerated.
Second, the approach taken here is centrist. Some of the findings and
our proposals will not be readily embraced by either organized medicine
or the organized trial bar. For example, we are highly critical of flat caps
on nonmonetary or total loss, which has been advocated by organized
medicine. On the other hand, we find flaws with the existing method of
 Introduction 3

assessing damages and recommend that alternative approaches for sched-

uling damages be considered. The organized trial bar has consistently
opposed legislation that would constrain the latitude that juries now
have to make such determinations.
Third, for any program of reform to succeed, the financial interests of
health care providers must be aligned with the social objective of injury
prevention and claims prevention. The financial incentives that providers
have to engage in error reduction and quality improvement currently are
insufficient. Under the best of circumstances, medical malpractice alone
cannot succeed in encouraging implementation of all desirable quality
improvements. Yet, rather than be an ineffective or even a negative force,
medical malpractice can have a productive role in deterring iatrogenic
injuries and in quality improvement.2
Fourth, the states have been laboratories for tort reform, whereas the
U.S. government has been inactive.3 Most of the recommendations in
this book can be implemented by individual states.
Fifth, especially when empirical evidence on effects of some promising
reforms is lacking or inconclusive, the law should permit health care
organizations to implement reforms. That is, in general, voluntary
approaches are to be favored over legislating broad “one size fits all”
packages.

Medical Malpractice and Tort Law

Tort law has several important goals, among the most important of
which are deterring misconduct (and hence injury) and compensating
injury victims.4 Other goals of tort include meting out justice and provid-
ing a safety valve for airing victims’ grievances. These latter objectives
are important to maintaining a civil society. While concerns arise more
frequently during times of rising insurance premiums, there are standing
concerns among scholars of tort law in general and of medical malprac-
tice in particular. Specifically, how successful is tort in attaining the
above objectives?
To date, there is no evidence that the threat of tort deters medical
injuries, although such evidence exists for other applications of tort law,
such as dram shop for alcohol sellers5 and automobile liability.6 Why the
4 Chapter 1

threat of a civil lawsuit is effective under some circumstances and not

under others is not entirely clear. A possible explanation is that the
underlying technology of injury prevention is easier in some areas than
it is in others. Having a bartender call a taxicab for a patron who has
consumed too many beers involves a simpler technology than is necessary
for preventing a mishap in transplanting an organ.

Indicators of a Broken System

A logical first step in the public policy process is to assess “what works”
and “what is broken” and needing change. Diagnosis is an essential first
step in the policy process, and much of our attention is devoted to
diagnosis, especially on the empirical evidence required for a good
diagnosis.
The above rationale for tort, as embodied in such goals as injury deter-
rence and compensation, assumes that the legal system is efficient and
accurate in adjudicating claims. At some point, inefficiencies and inac-
curacies tip the balance against use of tort liability. Critics often cite the
inefficiencies and inaccuracies of tort liability in general and of medical
malpractice in particular. For example, legal disputes, especially those
involving medical malpractice, can take years to resolve, requiring sub-
stantial use of legal resources on both sides of the dispute. Overall, the
medical malpractice process is a slow and expensive approach for
compensating persons who are injured as a result of receiving (or not
receiving) medical care.
Compensation under tort is indeed very expensive—when measured
in terms of the legal fees incurred by plaintiffs and defendants, court
costs, and insurer overhead.7 For tort liability as a whole (not just for
medical malpractice), between 40 and 50 cents on the dollar is returned
to plaintiffs as compensation for their injuries, which is much lower than
the share of the premium dollar returned to insured individuals in other
contexts, such as private health insurance, Medicare, and Social
Security.
Even though the current system has these very negative attributes, it
also has some important positive ones. For one, being able to sue in
combination with the contingent fee method for paying plaintiffs’ attor-
 Introduction 5

neys gives patients who are unsatisfied with outcomes of care a mecha-
nism for addressing their grievances that would not be possible through
other channels. The regulatory apparatus, in principle designed to serve
the public interest, is sometimes controlled or substantially influenced
by the parties it is designed to regulate. Because of health care provider
influence or for some other bureaucratic reason, regulatory agencies
may be unresponsive to patients’ complaints. Ironically, one reason that
medical malpractice is so aggravating to the health care provider com-
munity is that patients’ ability to file a tort claim empowers patients to
obtain justice and compensation when other systems, likely to be more
subject to providers’ influence, fail. Not all patients’ complaints prove
to be meritorious in the end, but some do.
Advocates for tort see it as an effective private mechanism for meting
out justice, especially when other systems, such as public regulation and
self-regulation by hospitals and physicians, fail to achieve their stated
purposes. Moreover, defenders of the current system argue that individu-
alized justice is costly to achieve and hence is inherently expensive.
Fortunately, there is now a large body of empirical evidence on
the performance of tort in general and of medical malpractice in par-
ticular. Unfortunately, much of the public discourse continues to be
based on anecdotes, which may be valid in isolated cases but do not
generalize.
The rising cost of medical malpractice insurance premiums is cited in
public discourse as an indicator that the medical malpractice system is
“broken.” While complaining about spending more money for insurance
is certainly understandable—no one feels good about spending more for
insurance—rising premiums, numbers of medical malpractice claims, and
payments per paid claim are not in themselves valid social indicators that
the system is broken. In a dynamic economy, expenditures on some
goods and services rise and expenditures on others fall. There have been
attempts to link trends in medical malpractice to increased spending on
personal health services and reduced patient access to care.8 But the
empirical evidence indicates that medical malpractice is a factor of
second- or even of third-order importance among determinants of
increases in health care spending and of decreases in patient access
to care.
6 Chapter 1

If one views secular trends in medical malpractice payments and pre-

miums, adjusted for general price changes, increases in payments and
premiums have actually been quite modest, only slightly higher than the
changes in prices overall. It is the substantial increase in premiums and
withdrawal of medical malpractice insurance at the onset of “hard
markets” that have attracted widespread attention. These premium
shocks as well as shocks to availability of insurance have indeed been
substantial, but sharply contrast with secular trends which have been
much more moderate.
Levels, trends, and cycles in premiums, claims, and payments are
invalid indicators that the system is functioning or is failing and badly
in need of repair. But, as documented below, one reason that the threat
of medical malpractice litigation fails to deter medical injuries is that
there are too few rather than too many lawsuits, at least meritorious
ones. If the state police went on strike, it would seem unlikely that the
accident rate on highways would fall. In fact, having fewer troopers
patroling the roads would reduce the probability of being apprehended
for speeding and drunk driving. Thus, drivers might be inclined to be
less cautious. Most commentators tend to be unwilling to make a parallel
inference about too few medical malpractice suits—that more suits might
make operating rooms safer places for patients under the knife.
Further, if medical malpractice litigation were serving a useful role,
one would expect to observe situations in which more beneficial care is
given and greater precautions are taken in providing such care than
would occur in the absence of the threat of tort. The lack of empirical
demonstration that the threat of medical malpractice has improved the
quality of care, and hence has deterred iatrogenic injuries, is much more
troubling. Yet some elected officials gauge the success of statutory
changes, called “tort reform,” in terms of their effects on claims fre-
quency (box 1.1).
Not only is the number of claims no indication of success, but such
reforms have often not had lasting effects, even on claims frequency.
State courts have found some of the changes to be unconstitutional. Most
of the changes may have appeared to be likely to reduce claims, payments
per claim, and premiums at the time they were enacted but, as docu-
mented in chapter 4, for various reasons they did not have this effect.
 Introduction 7

Box 1.1
Court Statistics as an Indicator of Success of State Policy

“Citing data released from the state Supreme Court as “proof that our
reforms are making a difference,” Governor Edward G. Rendell proposed
a series of immediate and long-term proposals to address Pennsylvania’s
medical malpractice situation. The proposals supplement those offered by
the Governor in June 2003.
“Court statistics released last week showing a 30 percent decrease in the
number of malpractice suits filed in 2004 are great news and a tribute to
the reforms enacted by Governor Schweiker, the General Assembly and
Supreme Court prior to my tenure as Governor,” Governor Rendell said.
“They confirm what I’ve said for more than a year—the reforms are good
ones and will have an impact on the medical malpractice situation in
Pennsylvania.”
Pennsylvania has experienced much higher rates of claims, payouts, and
premiums than other states, especially in Philadelphia and Pittsburgh
(Bovbjerg and Bartow 2003).

Source: “Governor Rendell Announces Medical Malpractice Liability Pro-

posals,” PAPowerPort (March 2004). (http://www.state.pa.us/papower/
cwp/view.asp?A=11&Q=436418 (accessed Dec. 30, 2004).

History Repeats Itself

During each crisis, as the medical malpractice issue gains steam, experts
receive the same types of inquiries from the media. Questions include:
Has this happened before? Why are malpractice premiums and losses
skyrocketing? Where is the primary source of the problem—insurance
markets, dispute resolution, medical errors, failure of government over-
sight, some other factor, or a combination of these factors? Which solu-
tions have been tried and “work”? One’s definition of “work” depends,
of course, on one’s perspective. Those who believe that premiums are
too high define “work” as a reduction in insurers’ losses and in premi-
ums. For the victim of a medical injury, the concept of “work” is quite
different.
There are also questions of fact relating to specific consequences of a
medical malpractice crisis. For example, are physicians really relocating
because of rising premiums? Is statistical evidence from past years still
relevant? The growth of managed care and fixed physician fee schedules,
8 Chapter 1

which make it much more difficult for physicians to pass higher premi-
ums forward to patients and health insurers in the form of higher fees,
inform the responses to these questions. The responses from the early
twenty-first century may differ considerably from those in the 1970s and
1980s.

Four Markets

Medical malpractice is a complex topic in large part because its effects

are so far-reaching. It is not only about physicians and patients, not only
about lawyers, not only about insurers, and not only about governments
which are cajoled into responding in times of medical malpractice insur-
ance crisis. It is really about them all—each with different objectives,
constraints, and cultures. Outcomes are determined in four conceptually
distinct markets—legal, medical malpractice insurance, medical care, and
government activity (not stated in any particular order). The simultane-
ous operation of these four markets is largely what makes the issue of
medical malpractice as complex as it is. Though each market is discrete
in theory, in practice they interact. Most previous studies have focused
on one aspect (or one market). In this book, we analyze them all.

The Legal Market

In the legal market, injury victims and physicians as defendants demand
legal services, supplied by lawyers and the courts. In theory, lawyers are
to be perfect agents for their clients. But in practice, this may not occur
for financial and nonfinancial reasons. For instance, choice of compensa-
tion method—paying lawyers on a contingent fee or hourly basis—may
make a difference in attorney willingness to represent a client or on
lawyer effort on behalf of clients, as well as certain decisions, such as
whether and on what terms the dispute is resolved.9 In discussing lawyer
compensation, a report from a state governor alleged lawyer fee-splitting,
noting, “It is common practice for lawyers who do not handle medical
malpractice cases to refer these cases to lawyers who do specialize in
this area. For simply referring a case a lawyer will negotiate a fee agree-
ment with the other lawyer, which range[s] from 25 to 50 percent of any
fee for a settlement or judgment. The referring attorney is not obligated
 Introduction 9

to perform any actual work on the case.”10 Courts have expressed con-
cerns that juries’ decisions are unduly swayed by the severity and cir-
cumstances of the plaintiff’s injury, but this is disputed by other studies.11
In spite of some limitations in the judicial process, the American
jury gives ordinary citizens, in their roles as jurors, a part in the dispute
resolution process.
While insurers who engage lawyers for the defense are obligated to act
in the interest of the physicians they insure, they also have their own
interests, such as seeking to minimize total expense per case, which do
not take account of unquantifiable losses, such as the loss of a physician
defendant’s reputation. Market failures would occur if, for example,
(1) claimants consistently file nonmeritorious claims and obtain settle-
ments, (2) payments to claimants systematically exceed injury cost, or
(3) courts often make legal errors in determining liability and damages.

The Medical Malpractice Insurance Market

Another market is for medical malpractice insurance.12 Medical malprac-
tice insurance is part of a larger market—the market for property-
casualty insurance. This market is competitive but subject to a considerable
amount of state regulation. In the medical malpractice insurance
market, physicians and other health professionals are the consumers
and medical malpractice insurers are the suppliers. Insurers decide
which physicians to insure (whom to underwrite) and at what premium.
Physicians have an underlying suspicion that they may not be well
represented by commercial insurers. This is due, in part, to the fact that
an individual physician, whose business is not insurance, may be no
match for a large insurer whose business is insurance. For this reason,
as well as the exit of many commercial insurers from the medical mal-
practice insurance line, physician-sponsored medical malpractice insurers
were formed in many states in the mid-1970s. Many of these insurers
survive today.
Embedded in the premium-setting decision is the issue of risk classifica-
tion,13 the process of placing insured individuals into separate groups for
purposes of assigning a premium to each group. At first glance, this seems
like a purely technical decision, of interest only to the experts. In general,
there is an important trade-off between the goal of risk-sharing which
10 Chapter 1

insurance provides and maintaining an incentive for policyholders to

exercise due care.
However, hospital liability insurance is another matter. Many hospi-
tals self-insure for much of their coverage and buy excess insurance to
cover catastrophic risk. Excess insurance is highly experienced-rated.
If a careless policyholder, or one who works in an area in which inju-
ries are more likely to occur, pays the same premium as one who exer-
cises due care or works in an area in which injuries are less likely to
occur, there are two potentially adverse consequences. First, the incentive
for physicians to exercise due care may be reduced. Second, the insured
physician with a low claims risk subsidizes the physician with a high
claims risk. Thus, risk classification has important practical implications,
not only for the premiums that insured individuals pay, but also for any
effects medical malpractice insurance may have on deterring injuries.
As already noted, there are no documented deterrent effects of medical
malpractice. This may partly reflect the broad risk classes commonly
employed in the medical malpractice insurance field. Experience rating
is used far less in medical malpractice than in other lines of property-
casualty insurance, such as automobile liability insurance.14
For example, drivers with prior accidents or traffic violations routinely
pay higher premiums. The reason for the lack of experience rating in
medical malpractice may be a common perception that outcomes of the
claims are random, and hence physicians should not be made to pay
higher premiums after payments are made on their behalf. Charging
higher premiums for physicians who engage in risky practices, such as
delivering babies, may cause physicians to stop such practices—an
adverse reaction from a societal point of view.15 Rather than charge
higher premiums, insurers may refuse to underwrite high-risks, leaving
high-risk physicians without coverage or having to purchase coverage
from “surplus line carriers,” insurers which specialize in hard-to-insure
risks.16
Insurers also make decisions about which markets to enter, how
aggressively to defend claims, and about the amounts of reserves to set
aside for future losses incurred during a particular year in which they
insure risks (policy year). Physicians decide how much coverage to pur-
chase and from whom. Rising premiums and exits of insurers are the
 Introduction 11

immediate causes of crisis, but these decisions plausibly reflect more

complex underlying factors.

The Medical Care Market

The third market is for medical care. Here, consumers are patients, and
physicians, hospitals, and others are suppliers. Higher premiums may
lead to increased medical fees. In theory, medical malpractice would lead
to provision of optimal levels of care,17 but for various reasons, there
may be under- or overdeterrence (the latter called “defensive medicine”).
Underdeterrence may arise from the asymmetric relationship between
patients and physicians. For example, patients may fail to request certain
types of care because they do not have advance knowledge that such care
would potentially benefit them. In principle, imposing liability on physi-
cians is a method for preventing underdeterrence. However, overdeter-
rence could arise if the liability threat is excessive and/or imposed
arbitrarily. As a result of overdeterrence, physicians may overprescribe
diagnostic tests or therapeutic procedures, or avoid certain types of pro-
cedures or practice locations associated with higher probabilities of
lawsuits.
Critics often use such presumed effects as a pretext for public interven-
tion. However, it is difficult for physicians to argue for government
assistance because higher medical malpractice premiums have depressed
physicians’ incomes. It is much more persuasive for physicians adversely
affected by premium increases to argue that this will reduce provision of
highly beneficial medical services, such as obstetrical and emergency
room care. This is not to deny the validity of such arguments, only to
admit to the possibility that such arguments may be self-serving.

The Market for Government Activity

Finally, there is the market for government activity. The law-as-market
view asserts that legislation (and government activity more generally) is
a good demanded and supplied much like other goods.18 Legislative
protection flows to groups obtaining the greatest value from public sector
decisions, irrespective of their impact on social welfare. Citizens are on
both sides of the market, benefiting and paying taxes. Those benefiting
more from a type of decision form special interest groups to advocate
12 Chapter 1

for it (in this context, physician, lawyer, and insurer groups). The cost
of the policy is much more widely shared, mainly by taxpayers who
do not enjoy a direct benefit and who have an insufficient incentive
to inform themselves about policies in advance and act in their
self-interest.
Political officials maximize the aggregate political support they receive
from all interest groups.19 In the context of medical malpractice, these
officials are associated with executive or administrative, legislative, and
judicial branches of state government. They regulate the solvency, pre-
miums, and marketing practices of insurers. Legislatures enact laws
affecting claims resolution. They create special organizational forms
(e.g., mutual insurers). In addition, in some states, the state is a medical
malpractice insurer, providing no-fault coverage and public reinsurance.
Operated on an actuarially sound basis, publicly supplied reinsurance
has much to recommend it,20 but even so, such programs have been
subject to manipulation (see e.g., box 1.2). Empirical evidence on prior
effects of government intervention is often cited in debates, albeit
selectively.

What Is Known and Unknown About the Four Markets?

At one end of the continuum, there is the view that nothing is known
about medical malpractice, so it is necessary to resort to anecdotes. At
the other extreme is the view that everything is known, so public policy
remedies are obvious. Neither of these views is accurate. From an opti-
mistic or “cup half full” perspective, scholars have presented an enor-
mous amount of information about medical malpractice and tort liability,
and insurance more generally.
On the legal market, there are research findings, both theoretical and
empirical, on why injury victims file claims.21 There are studies about
the universe of injuries relative to claims frequency,22 determinants of
award sizes,23 and comparisons of injury cost with compensation.24 In
addition, scholars have researched the relative awards when claimants
use a specialist lawyer versus when they do not,25 the choice made by
injury victims between tort and no-fault when they have a choice,26 and
the outcomes in medical no-fault versus tort and in other contexts (auto
 Introduction 13

Box 1.2
“Raiding” the Medical Malpractice Patient Compensation Fund

Physicians in Wisconsin feel that their low medical liability insurance rates
are being jeopardized by the governor’s plan to deal with the state budget
deficit.
Wisconsin Gov. Jim Doyle has proposed taking $200 million from the
state’s Patient Compensation Fund to help offset a $454 million budget
deficit. Physicians, hospitals and other health care professionals have paid
into the fund annually for more than 25 years. The money is used to pay
damages that exceed the coverage of medical liability insurance policies.
The governor says the fund has money to spare, but physicians and some
lawmakers say it doesn’t and that it would be irresponsible to use the
money for other budget expenses.
“They are trying to plug a hole in one situation, but are creating a per-
manent problem in another area,” said Mark L. Adams, general counsel
for the Wisconsin Medical Society.
In addition to worrying that injured patients won’t be fairly compen-
sated, physicians and some lawmakers say that the fund, as part of exten-
sive tort reform efforts in Wisconsin, has helped keep medical liability
insurance rates low, even as they have soared in much of the rest of the
country. Wisconsin is one of only six states that the AMA says is not
showing signs of being in the midst of a medical liability crisis.

Source: Tanya Albert, AMEDNews.com. (Apr. 7, 2003). http:www

.ama-assn.org/amednews/2003/04/07/prsd0407.htm (accessed Dec. 30,
2004).

no-fault).27 Finally, scholars have studied the effects of contingent fees

on legal outcomes, and on jury behavior in tort litigation in general.28
Although much evidence suggests well-functioning legal markets, there
is recent evidence, not specifically in regard to medical malpractice, that
courts located in areas with high proportions of minorities and low-
income households award higher amounts of compensation than in other
areas.29 Also, medical malpractice appears to pay higher compensation
to injury victims than in other contexts (e.g., auto liability).30
In addition to research findings, there is much literature on markets
for medical malpractice insurance and other relevant work on property-
liability insurance. There is general literature on insurance cycles and
their causes.31 To preserve a deterrent incentive in the presence of com-
plete insurance, experience rating is desirable.32 Although experience
14 Chapter 1

rating is feasible in medical malpractice,33 it is rare.34 Even so, surplus

line carriers insure some physicians not able to obtain coverage from
standard insurers.35 There is some empirical evidence on exit/entry, loss
reserving, investment, premium-setting practices, and profitability of
medical malpractice insurers,36 and general findings on loss prevention
practices.37 Research exists on malpractice insurers’ decisions to rein-
sure,38 but we know relatively little about how reinsurer premium
increases and stricter underwriting practices affect primary insurers. The
literature misses some of the more recent developments.
Less is known about effects of medical malpractice on the market for
medical care. Some important theoretical work on the effects of alterna-
tive liability regimes on physician decisions has been conducted.39 We
know that in the past, premium increases have been shifted forward to
consumers in the form of higher fees,40 but this may not be possible now,
given the growth of fixed fee schedules (e.g., under Medicare Part B) and
the growth of managed care.
The effect of the threat of tort liability on physician care levels has
been studied.41 Medical errors remain frequent, even with the threat of
tort claims.42 Also, one study43 found that prior claims experience is
unrelated to subsequent technical quality of care. Although often asserted,
there is no conclusive evidence that physicians are leaving practice on
account of high medical malpractice premiums or altering product mixes
(e.g., dropping obstetrics). On the contrary, in one study based on data
from the 1970s and 1980s,44 physicians who had experienced high fre-
quency of claims were less likely to change their practices (e.g., leave the
state or retire). The potential deterrent benefit of liability may be greater
under capitation than under fee-for-service,45 but there is no empirical
evidence on this.
Conceptual research in economics, political science, and law has
described the government market in terms quite similar to any private
market. In this context, with strong, well-organized stakeholders advo-
cating for and against change, this analogy appears to be particularly
apt. Empirically, we know a lot about effects of tort reform on medical
malpractice premiums, claims frequency and severity, and total loss.46
Much less is known about why states implement specific statutory
changes in medical malpractice or property-casualty more generally.47
 Introduction 15

Crises in premiums and availability of insurance appear to be leading

precipitators of change. There is no quantitative evidence on the effect
of lobbies on legislative decisions affecting medical malpractice. Regula-
tory practices of state departments of insurance have been described.48
Government decisions ideally would be guided by both empirical evi-
dence and well-articulated social objectives. Economic analysis of tort
liability has been excellent in conceptually describing how alternative
legal rules may produce socially optimal levels of care.49 The level of care
that is optimal from society’s vantage point is one at which the marginal
social benefit of care equals the marginal social cost of providing it. For
example, if the total cost of conducting and interpreting a diagnostic test
is $1,000 per test, society’s welfare is maximized if tests are conducted
only on persons for whom the benefit of the test is also at least $1,000.
If the test is conducted on persons for whom the benefit is only $10, this
is a waste of society’s scarce resources.
Such benefits reflect both the effectiveness of care in producing better
health outcomes and the society’s willingness to pay for such better
outcomes. Such cost reflects resource outlays for care borne by all payers
as well as nonmonetary costs. For example, the price of a colonoscopy
is not the only out-of-pocket cost for the procedure. There are also other
costs, including transportation, the opportunity cost of the patient and
a family member taking time from work or other pursuits, and the pain
and suffering from preparing for the procedure and possible rare adverse
outcomes from the procedure itself.
Discussing the socially optimum level of care conceptually is one thing.
Determining what the marginal social benefit and marginal social cost
curves look like in practice is quite another matter. A major reason that
the calculation is so complicated is that benefits and costs differ substan-
tially among individuals.50 “Defensive medicine” presumably occurs to
the extent that the legal system causes too much care to be provided—
that is, care for which the marginal social cost exceeds the marginal
social benefit. What this means, for example, is that if there is a one-in-
a-hundred chance of a person having a disease and the loss incurred if
the disease goes undetected is $100,000, then the test should be con-
ducted if its cost is $1,000 or less. Otherwise, the test should not
be conducted. Defensive medicine occurs in the latter case. Although
16 Chapter 1

defensive medicine is said to be a major driver of health care cost growth,

there really is no evidence of how much it is.51 And health care cost
growth is the wrong metric in any case. The correct method applies the
principle of cost versus benefit.
The effects of various statutory changes on the well-being of injury
victims, individuals as patients, physician defendants, and attorneys ulti-
mately always involve equity issues. These issues, although considered
in the political process and by the popular media, have received little
attention from scholars in any discipline, with a very few notable excep-
tions.52 Statutory changes in tort liability inevitably involve trade-offs.
For example, implementing limits on contingent fees has potential effects
on the levels of care doctors provide to patients, since the threat of being
sued may decline after the fee limits are imposed,53 but also may raise
barriers for injury victims in obtaining legal representation. Lawyers, for
both the plaintiff and the defense, are made worse off, the latter because
legal effort is roughly equivalent on both sides of the dispute.54 A priori,
the net effect of this statutory change on societal well-being is not clear,
although the well-being of particular types of individuals is clearly
affected, either positively or negatively.
There are trade-offs among categories of injured persons. For example,
is a person who is the victim of medical negligence more worthy of
compensation than a person injured in a natural disaster, such as a hur-
ricane or an earthquake? Currently, in a comparison of victims of medical
malpractice and hurricanes, the former would generally receive higher
compensation than the latter would. In the end, when there are two
groups that stand to gain or lose from a public policy decision, political
officials must make the trade-off decision in the well-being between the
two groups. The best an analyst can do is identify the nature of the
trade-off.

Five Myths of Medical Malpractice

Any discussion of malpractice would be incomplete without acknowledg-

ing the myths associated with medical malpractice. In some lectures and
in the first author’s undergraduate class on health economics, the first
author states and discusses five myths of medical malpractice. The
 Introduction 17

purpose of the discussion of the myths is not only to present empirical

evidence demonstrating that the statements are myths, but also to be
provocative and elicit interest in the topic (though hopefully not too
provocative). The five myths are the following:
Myth 1. There are too many medical malpractice claims.
Myth 2. Only “good” doctors are sued.
Myth 3. Dispute resolution in medical malpractice is a lottery.
Myth 4. Medical malpractice claimants are overcompensated for their
losses.
Myth 5. Medical care is costly because of medical malpractice.
Myths 2–5 will be discussed at greater length in subsequent chapters.
Thus, only a very brief preview is provided here.
Doctors who are sued for medical malpractice are neither of higher
nor of lower quality, on average, than those who are not sued (myth 2).
Some critics of medical malpractice contend that being at the cutting
edge technologically makes a physician more vulnerable to being sued.
There is no empirical evidence that being sued is an indicator of superior
performance, as is sometimes alleged. There is some empirical evidence
about differences in physician–patient relationships between those physi-
cians with adverse medical malpractice claims histories and physicians
with no claims, the latter rated by their patients as being, or at least
appearing to be, more understanding, more caring, more available, and
a good communicator. A more patient-oriented practice style is good
defensive medicine, a point rarely mentioned in public discourse on
medical malpractice.
There is a definite relationship, albeit an imperfect one, between inde-
pendent assessments of liability and of injury cost and outcomes of legal
disputes alleging medical malpractice (myth 3). Assertions that medical
malpractice outcomes as a general matter are random or a lottery are
not supported by the empirical evidence. Based on available evidence
comparing the costs of injuries of medical malpractice claimants and
compensation actually received, such claimants are, on average, under-
compensated, not overcompensated (myth 4). In the one study that has
compared injury cost versus compensation, compensation exceeded cost
18 Chapter 1

by 22 percent for claimants who received compensation at verdict. The

22 percent included payment for pain and suffering.
Practice decisions attributable to the threat of a medical malpractice
suit may have raised spending on personal health services above what it
would have been absent the threat, but threats of a medical malpractice
suit are not a major cause of rising expenditures on personal health ser-
vices (myth 5). In addition, any reduction in spending that would occur
if the threat were eliminated or appreciably reduced would inevitably
lead to a decline in physician and hospital revenue. For example, if 15–20
percent of personal health expenditures were eliminated due to “effec-
tive” tort reform, the health care sector would fall into a deep recession
or even a depression, and depending on how funds released from the
health care sector were reallocated, Gross Domestic Product for the
economy as a whole might suffer more than a blip as well. Physicians
would leave practice in large numbers, and many hospitals would close.
Although expenditures on personal health services would probably fall,
patient access to physicians and hospitals would be much worse than it
is now.
Myth 1 deals with the excess number of medical malpractice claims.
As background for discussing this myth, two pathbreaking studies are
especially pertinent. The first was conducted in California in 1974.55 The
second was conducted in New York in 1984.56 In both studies, surveys
of medical records of hospitalized patients were conducted to ascertain
(1) rates of injury (“adverse events”) attributable to provision of medical
care to these patients and (2) rates of adverse events due to provider
negligence, termed “negligent adverse events.” Follow-up studies to the
New York study were conducted in Colorado and Utah.
In the California study, records of 20,864 randomly selected patients
at twenty-three hospitals were reviewed. Records were abstracted accord-
ing to a protocol and then analyzed by medical and legal experts to
determine whether the injury occurred while the patient was receiving
medical care and whether or not the injury could be attributed to pro-
vider negligence.
In the New York study, 31,429 records of patients were reviewed. The
data collection, abstraction process, and medical and legal expert review
were patterned after the process used in California a decade earlier. The
 Introduction 19

California study revealed that 5 percent of patients experienced an

adverse health event while in the hospital, and 17 percent of these
patients suffered a negligent adverse event. In New York, the correspond-
ing rate was 4 percent for adverse events, and of those, 28 percent were
classified as due to negligence.
The authors of the New York study and the Institute of Medicine of
the U.S. National Academy of Sciences used estimates from the New
York study to compute the annual number of deaths due to iatrogenic
injuries in the United States and the number of such injuries for which
providers were at fault. These estimates have been criticized on grounds
that the implied number of deaths due to iatrogenic injuries is too high.57
Like any research, these studies have strengths and weaknesses; they will
be discussed in detail in a later chapter. But the exact number of deaths
is really unimportant. The number is large, and this should suffice for
private and public decision-making. It is a bit like the critiques of the
estimates that six million Jews died in the Holocaust—not that we wish
to compare those who deny that there are numerous medical errors with
those who deny that the Holocaust occurred, but our example does make
an important point. Suppose the true number of Holocaust deaths were
four million. Would this change the analysis in any important way?
The New York researchers obtained data on medical malpractice
claims filed on behalf of the injury victims identified in their study. The
authors found “invalid” claims—those not matching the study’s deter-
mination of liability from raters’ evaluations of the medical records—
outnumbered valid claims by a ratio of three to one—providing empirical
support to myth 1.58 However, it is not appropriate to stop here. They
also found that only 2 percent of negligent adverse events resulted in
medical malpractice claims. There were 7.6 times as many negligent
injuries as there were claims. Thus, there were errors in both directions:
individuals filed too many invalid claims and not enough valid claims.
Do these results serve to strike down myth 1? In two important senses,
the answer is no. First, many valid claims are not filed. What is filed is
the tip of the iceberg of potentially valid claims. Second, the raters viewed
only medical records for the patients when they were hospitalized.59 In
addition, the litigation process involves accumulating much more infor-
mation than is contained in a single medical record. It is quite possible
20 Chapter 1

that with the additional information, the expert findings could have been
reversed. Of course, the reversal could be in both directions. Yet it seems
highly likely that the true percentage of negligent adverse events resulting
in claims is somewhere between 0 and 4 percent. The percentage of
negligent adverse events resulting in claims is very, very unlikely to be
in, say, the 50 to 80 percent range of negligent adverse events.
Then how does the large number of invalid claims affect the validity
of myth 1? At least at first glance, there appears to be support for the
allegation that there are too many lawsuits. The measurement error in
estimating the number of negligent adverse events that result in claims
is plausibly relevant here as well. Applying the same logic, the true ratio
of invalid claims to valid claims could be much higher than three to one.
Further, medical malpractice plaintiffs lose the overwhelming majority
of suits for which a verdict is reached.60 Thus, the system does weed out
many invalid claims.
From another perspective, litigation can be viewed as an information-
gathering process; many claims that are filed are dropped by claimants
after initial investigation reveals that the claim has little or no legal
merit.61 Would we want to say that the large number of negative test
results provides conclusive evidence of overtesting, even when the tests
are justifiable ex ante on a clinical basis? The fact that the person in the
above example had a negative test result even though the ex ante chance
of having the disease was one in one hundred does not imply that the
test was unnecessary. This is not to assert that there are no frivolous
claims, but rather that not every claim that turns out to have been
“invalid” is frivolous. The term “frivolous lawsuits” has been used much
too loosely in public discourse about medical malpractice.
At a superficial level, the New York study offers “good news” for
advocates on both sides of the medical malpractice debate. In the end,
myth 1 is partially valid; there are both too many invalid claims and too
few valid claims. The system, in sum, is imperfect.

Chapter Roadmap

The intended audience for this book is nonspecialists. With this in mind,
the text unites several areas of academic research in medical malpractice.
 Introduction 21

It is difficult to be well versed in all of the topics presented, and for that
reason we aim to provide a text useful for anyone interested in medical
malpractice. A reader who is familiar with the legal issues will be made
aware of the political issues. In the same way, a reader who is familiar
with the political issues may not have previously understood how insur-
ance works.
This chapter and the next three chapters describe what is known about
the functioning of medical malpractice insurance, defensive medicine,
and the effects of tort reforms implemented to date. Chapter 2 describes
why insurance cycles arise. “Hard markets” characterized by sharply
rising medical malpractice premiums and withdrawal of insurers from
the market have led to much political pressure for policy changes. Further,
advocates for particular statutory changes base their arguments for
change on their theories of the origins of cycles.
Chapter 3 analyzes effects of rising medical malpractice premiums and
the threat of lawsuits. Arguments for change are often linked to the
concepts of positive and negative defensive medicine. Positive defensive
medicine involves the use of diagnostic and therapeutic procedures in
excess of levels that physicians would recommend solely on the basis
of their professional clinical judgments. Negative defensive medicine
involves withdrawal of care because of high premiums and/or the threat
of lawsuits.
Chapter 4 describes tort reforms and their effects. Reflecting the goals
of the stakeholders who promote tort reforms and reflected in the term
“tort reform,” success has been gauged in terms of whether or not, and
the extent to which, specific tort reforms reduce medical malpractice
claims frequency, insurance payments, and premiums. The “prizewin-
ners” to date are caps on nonmonetary loss and on total loss. A large
body of evidence is generally consistent with the view that caps reduce
payments of insurers, and also decrease the variance in anticipated loss,
and lead to lower medical malpractice premiums.
Some tort reforms in effect transfer money from injury victims and
their attorneys to health care providers. Flat caps on damages, the over-
whelming favorite of the lobbies for provider organizations, parti-
cularly since 2000, fall in this category. Placing a cap on damages
has no potential for improving patient safety. In addition, this policy
22 Chapter 1

disproportionately makes plaintiffs with severe injuries worse off. A

stronger argument for limits on awards is that there is considerable varia-
tion in awards, limiting, among other things, the injury deterrent signal.
As an alternative to flat caps, chapter 5 examines proposals to fully
schedule damages for nonmonetary loss rather than just place limits on
the high payments. Another alternative to fixed lump-sum payments,
periodic payments, or even a complement to scheduled damages, is
service benefit insurance contracts to cover medical, custodial, educa-
tional, and rehabilitative services. Chapter 5 also describes a proposal
for service benefit contracts.
Damage caps do reduce medical malpractice premiums and are favored
for this reason, but again, they disproportionately disadvantage claim-
ants with severe injuries. We propose more equitable methods for
limiting damages and providing more consistent compensation, both
horizontally (for plaintiffs with similar injuries) and vertically (for plain-
tiffs with more severe or less severe injuries and loss).
Much of this book is about avoiding misguided reforms. Limitations
on lawyers’ contingent fees are a case in point, for reasons described in
chapter 6. Such limitations represent a partial incomes policy, one which
redistributes income from plaintiffs’ lawyers to physicians and perhaps
to others as well. The data that have been presented to show that plain-
tiffs’ attorneys have excessive earnings are seriously flawed. Moreover,
if there is concern about earnings inequality, this should be addressed
by a broad policy change, not by concentrating on a single occupation,
such as plaintiffs’ lawyers. Even if the data on plaintiff lawyers’ earnings
were accurate, such lawyers are not the only persons to enjoy high earn-
ings. Defendants’ lawyers are likely to have similar earnings, on average.
Moreover, some insurance executives are very well compensated, as are
physicians in some specialties. As much as fee limits bar access to legal
representation, they do this indiscriminately, affecting access to legal
representation of potential claimants with and without meritorious
claims.
Chapter 6, which describes empirical evidence on this issue, as well as
some prominent policy proposals, concludes that no change is warranted
in this area. Aside from the lack of empirical support for arguments
supporting change, in terms of horizontal equity (equal treatment of
 Introduction 23

equals) it seems imprudent to impose constraints on fees of attorneys

who represent claimants without likewise imposing constraints on indi-
viduals in other highly compensated professions.
Chapter 7 examines empirical evidence on jury behavior, finding that
“runaway juries” perhaps appear in a few highly publicized cases and
more generally make good fiction, but do not apply to the typical jury.
One solution to a “runaway jury” problem, thought in part to stem from
jurors’ inability to process relevant scientific information and in part due
to jurors’ alleged sympathy with injury victims, is to shift medical mal-
practice cases to specialized health courts. Regulatory agencies and even
judges may not be equally sensitive to consumer interests, although, after
considerable discussion in this book, we leave as an open question
whether or not health courts should be implemented. The pros and cons
of this policy option are evaluated in chapter 7.
Existing tort reforms do not make care safer. There has been much
discussion of patient safety, especially since around 2000, including the
high rates of medical errors that occur. As chapter 8 indicates, in spite
of all the national attention the issue of medical errors has received,
surprisingly little progress has been made in implementing error reduc-
tion systems. This is partly because meaningful financial incentives for
health care providers to adopt patient safety measures have been
lacking.
In principle, tort liability could provide such incentives, but not as it
is currently structured in the medical field. And under the best of circum-
stances, tort liability is only one of several policy instruments that can
be applied to make medical care safer than it is now. The threat of a
lawsuit would presumably deter such errors, but (1) there is empirical
evidence that such errors are widespread (although there is some dis-
agreement about how widespread they are), and (2) some experts even
argue that medical malpractice has been counterproductive in achieving
the objective of reducing medical error rates.
One impediment appears to be the lack of a financial incentive to
implement such systems. Not only is there a general lack of a market
incentive, but medical liability is not playing a positive role. Physicians
practicing alone or in small groups may not be well positioned to make
major investments in error reduction approaches. This suggests that
24 Chapter 1

hospitals may be in a relatively good position to do this; there is a good

case for enterprise liability at the hospital level. Physicians would be
included for the care they provide within the walls of the hospital, in
both inpatient and outpatient settings.
Following chapter 2’s discussion of insurance cycles, chapters 9 through
11 discuss medical malpractice insurance. Chapter 9 focuses on the lack
of experience rating of medical malpractice, insurance regulation, the
various alternative ownership forms that insurance companies can and
do take, and government interventions to assure health care providers
access to medical malpractice insurance coverage and to cover losses in
the event of insurer bankruptcy.
Chapter 10 discusses private markets for reinsurance and the rationale
for and experience with government provision of such insurance. Volatil-
ity in reinsurance markets is one cause of cycles in markets for primary
medical malpractice insurance. One approach for reducing the ampli-
tudes of cycles, which can be quite disruptive, especially in some geo-
graphic locations and physician specialties, is to substitute publicly
provided for privately provided reinsurance, an option considered in
chapters 10 and 12.
Chapter 11 focuses on provision of medical no-fault insurance as a
substitute for standard third-party medical malpractice (or professional
liability) insurance. A voluntary no-fault plan in which hospitals and
patients can elect no-fault has attractive features. However, no-fault
insurance for iatrogenic injuries, such as exists in Sweden and New
Zealand, is not an achievable alternative in the United States. The experi-
ences with medical no-fault in Florida and Virginia, and of the federal
Vaccine Injury Compensation Program, demonstrate limitations of
medical no-fault, U.S.-style. While there are some advocates of substan-
tial expansions in no-fault coverage among academic experts, the proof
of the pudding is in actual implementation. The successes of no-fault in
some other countries have been impossible to replicate in the United
States. Therefore, the recommendation in chapter 11 is to expand
voluntary, hospital-based, no-fault insurance as a substitute for tort
liability/insurance and to do this by contract.
In the concluding chapter, chapter 12, the focus is on public policy
reforms that have promise in terms of improving quality of care and
 Introduction 25

determining injuries in health care settings. The chapter recommends

locating insurance coverage with the hospital as the insuring unit for all
care delivered within its walls. Much of the loss is incurred in hospital
settings. Certainly the high premiums in such specialties as obstetrics/
gynecology and neurosurgery reflect the medical malpractice risk from
care delivered at hospitals. Nevertheless, even though physicians would
be covered for care they deliver in hospital inpatient and outpatient
facilities, they would still purchase medical malpractice coverage for
services delivered in their own practices. Enterprise liability with the
hospital as the enterprise is also an attractive option, but it is a somewhat
more radical change that may face greater resistance. Under enterprise
liability, only the enterprise (the hospital) would be sued.
In sum, in terms of the four markets, chapters 6 and 7 are most directly
about the legal market; chapters 2 and 9–11, about the medical malprac-
tice insurance market; chapters 3 and 8, mostly about the medical care
market; and chapters 4, 5, and 12 mostly about the government market
and public policy.
2
Why the Crises in Medical Malpractice?

Most industries experience business cycles, yet it is not clear why the
property-casualty insurance industry, of which medical malpractice
insurance is part, has such prominent cycles. The property-casualty
industry’s cycle, termed more generally “insurance cycle” or “underwrit-
ing cycle,” is characterized by periods of “soft” and “hard” market
conditions. Cycles in medical malpractice insurance are important
because they can be at least temporarily disruptive to health care delivery
(at a minimum, this is a widespread perception), and they elicit strong
demands for change in the political arena.
This chapter has three purposes: (1) to examine how deeply entangled
medical malpractice is with the insurance cycle; (2) to provide insight
into the functioning of the medical malpractice insurance business;
(3) to explore the dynamics of the cycle by looking at both internal
and exogenous causes. The description of the insurance industry
requires a look into the frequency and severity of claims; reserves
and premium-setting; the insurer’s income statement and balance
sheet; and their relationship with each other. Discussion of internal
and external, largely exogenous causes reflects the true complexity
of the insurance industry and requires attention to several factors:
inflation shocks; capacity constraints; the oligopolistic (few firms
in a market with each firm taking into account the actions of its
competitors—in pricing, for example) structure of the property-
casualty insurance industry; and price and availability of reinsurance.
This chapter provides context essential for understanding the nature of
the crisis.
28 Chapter 2

The Nature of Insurance Cycles

Before beginning a scrutiny of the insurance cycle, some background

is necessary. To start with, a major feature of insurance cycles is
the intensity of competition among insurers during one phase and the
seeming lack of sellers during another. After the hard market sub-
sides, premiums decline as insurers compete to increase their individual
market shares. As the soft phase of the market matures, and losses
on previously sold, underpriced insurance policies mount, profits
decline, often to the point of insurers experiencing losses. Competition
among insurers becomes far less intense, and underwriting standards1
become more stringent. The supply of insurance decreases because
of a decrease in capital available to underwrite insurance, and premiums
rise. Then, improved profitability attracts more capital to the industry
from external sources, and a new soft phase with increased competition
begins, eventually leading to another hard market. Interestingly, the
periods during which the property-casualty insurance industry has
experienced hard markets since 1970—1975–1978, 1984–1987, and
2001–2004—coincide with the periods of crisis in medical malpractice
insurance.2
Widespread media and public policy interest mainly exists only in
times of crisis.3 Insurance cycles are not an important issue on many
politicians’ agenda, absent pressure from stakeholders. However, politi-
cal pressures for legislative change rise markedly during hard markets.
Sharply rising premiums and nonavailability of insurance coverage
provide the immediate impetus.
Even the frequency of published research papers on medical malprac-
tice per year varies with the cycle, more articles appearing after the onset
of hard markets. During periods of soft markets (e.g., 1988–2000),
analogies to Rip van Winkle and the title (but not the plot) of the 1946
Humphrey Bogart movie The Big Sleep are apt. This is not to say that
individuals are unconcerned about medical malpractice issues during the
quiescent periods. Rather, political activity is much lower, because either
the stakeholders’ organizations are focused on other issues, or individu-
als are not sufficiently riled up about medical malpractice to compel their
leaderships to act. During a quiescent period for medical malpractice
 Why the Crises in Medical Malpractice? 29

insurance during the 1990s, for example, organized medicine was preoc-
cupied with mounting a campaign for patient protection laws to combat
managed care’s alleged excesses.
The dynamics of the insurance cycle are not fully understood. Two
features of cycles are prominent. Premium increases are in excess of
increases in costs incurred by insurers, and there is a lack of supply of
insurance at the beginning of hard markets.
Absent some external interference, such as government controls over
prices, lack of supply is an anomaly in markets. Even though Picasso
paintings are no longer produced, there is no shortage of such paintings.
Rather, if there is increased demand, the price of such artwork rises. At
higher prices, some potential purchasers of Picasso paintings drop out
of the market. Nonavailability of such paintings is at most a very short-
term phenomenon. In the longer run, the paintings are available but at
higher prices than before.
In the context of medical malpractice insurance—in sharp contrast to
the market for Picasso artwork, not being able to obtain insurance at
any price, even for weeks or months, can be highly disruptive for physi-
cians. In addition, the premium increases lead to a great deal of com-
plaining in public forums, especially when the victims are as politically
influential as physicians.

Medical Malpractice Insurance and the Cycle

The cycle is inextricably linked to medical malpractice insurance for

several reasons. First, “crisis” is a by-product of cycles, not of secular
trends. The trend in losses from medical malpractice claims does not
exceed the trend in the medical Consumer Price Index (CPI), at least
when viewed over several decades. Certainly the trend in the medical CPI
has not produced a “crisis.” In other contexts, price spurts are seen as
predictors of adverse future trends, as was the case with the increase in
oil prices following the oil embargo of the 1970s and the “stagflation”
that ensued. A similar example is the concerns about $100 or even $300
per barrel oil prices that followed the substantial increase in oil prices
in 2005–2006. Clearly, medical malpractice is not the only subject area
in which cycles and trends are confused.
30 Chapter 2

Second, advocates on both sides of the medical malpractice issue make

their cases for change or preserving the status quo based on whom they
identify as the “culprits”—those who are allegedly responsible for the
crisis. On the one side there are those who link the crisis to “runaway
juries”4 and “greedy lawyers.”5 On the other side are those who blame
interest rates, and possibly insurer pricing practices, for the crisis. If one
attributes the crisis to falling interest rates and bad investments in the
stock market, the policy implications are markedly different than if
softhearted and cognitively limited juries and ambulance-chasing lawyers
are to blame.
Third, cycles are here to stay. In the mid-1960s, some advocates of
Keynesian macroeconomic policies6 thought that business cycles had
been eliminated because governments knew precisely when and how to
apply fiscal policies. History proved them wrong. The same goes for
insurance cycles.
Inept juries and greedy lawyers cannot explain cycles, unless one
argues that there are cycles in softheartedness, ineptness, and ambulance
chasing. Poor investments in the stock market are not a plausible reason.
Property-casualty insurers have the bulk of their investments in interest-
bearing securities, not in equity. Furthermore, insurers could not recoup
losses from bad investments by charging more to current customers.
Another insurer, which had been a prudent investor, could sell insurance
at a lower price, and the insurer with imprudent past investments would
lose market share. Or an insurer with neither a good nor a bad invest-
ment history could enter the market and attract business from the impru-
dent insurer. There are cycles in interest rates, but only anticipated,
not prior interest rates, are relevant for insurers’ premium-setting
decisions.
Another argument is that increased premiums are due to insurers’
exploiting their market power in setting premiums. This argument can
be rejected for three reasons. First, one would have to argue that there
are cycles in insurers’ pursuit of their profit-maximizing objectives, which
seems implausible. Second, many of the medical malpractice insurers are
sponsored by medical societies. It is doubtful that such organizations
would exploit their own sponsors. Third, there is empirical evidence for
 Why the Crises in Medical Malpractice? 31

the period including the first two crises that, at least in the long run,
medical malpractice premiums are actuarially fair.7 Thus, the above
arguments can be rejected as causes of insurance cycles; to find likely
causes, it is necessary to look elsewhere.

Background Pertinent to Analysis of Cycles

Frequency and Severity

Unlike most insurance, medical malpractice is a low claims frequency
and high severity line of insurance. Health and automobile liability
claims are much more frequent than medical malpractice insurance
claims; and earthquake and life insurance claims are much less frequent.
Severity refers to the size of paid claims. Malpractice claims are relatively
high in mean payment, and the distribution of such payments is highly
skewed, with a few very large payments, reflecting a few very severe
injuries, at the high end of the frequency distribution. Low claims fre-
quency makes actuarial predictions more difficult; high-dollar severity,
especially when coupled with a few very large paid claims, increases the
importance of reinsurance.8

Tail of an Insurance Policy

Premiums for all types of insurance are received before claims are paid.
The mean lag between the date of the occurrence of an insurable event,
the date the claim is filed, and the date the claim is paid differs appre-
ciably among lines of insurance. This lag is called the “tail” of an insur-
ance policy. Thus, the “claims tail” is the length of time from filing to
resolution of a claim.
Claims on health insurance are paid relatively quickly.9 The fact that
a covered expense occurred is easily documented, and improvements in
electronic transmission have reduced the delay even further.10 Since the
insurance is taken out on oneself, there is no opposing party to dispute
the validity of the claim, which applies to all first-party insurance in
general.
Demonstrating negligence in a line of third-party insurance, such
as medical malpractice insurance, can be a lengthy process. For some
32 Chapter 2

injuries, such as from automobile accidents, the injury is documented at

the site of the accident and rapidly confirmed—say, when the injured
person is transported to a hospital emergency room. In contrast, some
medical malpractice injuries take years to be identified (e.g., delayed
development in a child attributed to an injury the child suffered at birth),
leading to a long delay between the injury date and the date the claim
is eventually resolved. To shorten this time period, many medical mal-
practice insurers switched from occurrence to claims-made insurance
during the 1970s, but some occurrence policies continue to be written
three decades later.
Occurrence policies cover liability from the date of occurrence. Claims-
made policies cover liability from the date the claim is filed.11 Neverthe-
less, in spite of this change in much of insurance practice, the claims tail
remains unusually long for medical malpractice insurance.
Based on a conversation its personnel had with a national association
of insurers, the U.S. General Accounting Office (June 2003, p. 8) reports
that many medical malpractice claims take five years to resolve, including
discovering the malpractice, filing a claim, determining payment respon-
sibilities, and paying the claim. Some claims can take as long as eight to
ten years to resolve.12

Insurer’s Income Statement

A company’s income statement records its revenue, expenses, and profits
(losses) for a particular period, its fiscal year. In the context of insurance,
there are essentially two revenue streams—revenue from the sale of insur-
ance and revenue from investments. Expenses consist of payments to
policyholders, other expenses attributable to individual claims, “adjusted
loss expense” (ALE, which includes payments for defending cases, such
as attorneys’ fees), and expenses for marketing, managing investments,
and general overhead. The difference between total revenue and total
expenses is total profit. The core of the insurance business is assuming
(or underwriting) the risk of uncertain future events in exchange for a
premium. Underwriting profits are the difference between premiums and
paid losses and ALE in a fiscal year. Premium income is invested from
the time it is received to the time that payments for loss and ALE are
made.
 Why the Crises in Medical Malpractice? 33

Insurer’s Balance Sheet

A firm’s balance sheet lists assets, liabilities, and equity, the latter being
the difference between assets and liabilities. Because of government regu-
lation of insurer solvency and their own decisions to avoid bankruptcy,
insurers are generally very conservative investors. With some excep-
tions,13 the major part of their portfolios is invested in interest-bearing
securities.14 These investments were estimated by the U.S. General
Accounting Office (June 2003) to account for 79 percent of medical
malpractice insurers’ portfolios of the fifteen largest medical malpractice
insurers nationally, which consisted of both commercial stock and
medical society-sponsored mutuals and reciprocals. In other words,
insurance companies have relatively little common stock in their portfo-
lios, and thus are not greatly affected by the performance of stock
markets—to a far lesser extent than for most personal investors.
More important to an insurer’s bottom line is its liabilities, which
consist of its reserves on the insurance policies it issues. When an insurer
issues an insurance policy, it incurs a dollar obligation that must be
backed by assets of corresponding size. Losses are incurred throughout
the policy year, not on the date the premium is received (or day 1 of the
policy year). With some exceptions, insurers do not immediately know
that a potential loss has been incurred. Further, there is a distinction
between unearned and earned premiums. Premiums switch from unearned
to earned as the policy year evolves. For earned premiums, there are two
types of loss, reserves—incurred but not reported (IBNR)—and case
reserves.
As the name suggests, IBNR reserves are set to cover losses that have
occurred but have not yet been reported to the insurer. Once the report
is made and the insurer has determined the most likely loss from the
specific claim, some part of IBNR reserves converts to a case reserve on
that claim. The case reserve reflects the insurer’s best estimate of the loss
that will ultimately result from that claim. The case reserve or, equiva-
lently, the anticipated loss on the claim, is not set in stone; rather, it is
adjusted periodically as new information pertinent to the claim is revealed
to the insurer. New information could include a previously unseen tran-
script of a deposition, a communication by an attorney of the opposing
party, or a jury verdict in a similar case.
34 Chapter 2

When an insurer sells a policy before the policy year begins, the entire
reserve is unearned. The anticipated loss, especially in a long-tail line,
such as medical malpractice insurance, is best seen as an educated guess.
It reflects the insurer’s view of trends in claims frequency and severity,
which in turn reflects anticipated inflation, trends in legal decisions, and
so on. Some premiums may be ceded to a reinsurer.15 In this transaction,
a primary insurer exchanges money (pays a premium) for a promise by
the reinsurer to cover a specific expense, should the criteria for payment
warrant this. Reinsurance is insurance for the primary insurer. By rein-
suring, the primary insurer reduces its liabilities and its risk of insolvency,
but it sacrifices some revenue in return.
The term used in the insurance industry for the difference between
total assets and total liabilities is the surplus. The surplus for some single-
line medical malpractice insurers is often quite small relative to the
potential loss on a handful of claims, exposing these insurers to substan-
tial bankruptcy risk, absent some action, such as reinsuring to reduce
such risk.

Setting Premiums
Premiums are set on the basis of forecasts of future losses from insurance
written in a particular policy year, an adjustment for risk, the tail, and
the anticipated returns to be earned on investments from the premium
funds collected for the policy year, as well as market factors. Because it
can earn investment return on premiums, the insurer can charge a
premium that is below the anticipated loss on the insurance policy.
Higher risk tends to result in higher premiums. A longer tail and higher
anticipated returns tend to result in lower premiums. When an insurer
computes a premium for a future period, it takes account of expected
returns from investing money it collects as premiums that it will retain
until payments on losses are made. The potential return is positively
related to the length of the time period that elapses from the date the
claim is filed until payment is actually made.
When considering returns on investments, insurance companies look
forward, not backward. As explained earlier, in a competitive insurance
market, no one should be willing to pay a higher premium to a company
just to allow the company to recover losses from errors in its past invest-
 Why the Crises in Medical Malpractice? 35

ments. Forecasting future returns is fraught with uncertainty. Interest

rates reflect many factors, including inflationary expectations and mon-
etary policies of central banks. If insurers forecast higher real returns on
investments, premiums can be lowered, and conversely.
Adjusting for risk is both a complex and a not fully resolved issue (at
least to scholars). At the most basic level, insurers diversify the risk of
loss by pooling uncorrelated risks of individual insureds. By pooling,
what is a risk of loss to an individual is a much smaller risk to the
members of the pool. In any period, most insureds are fortunate not to
incur losses. Others are not so fortunate, and incur losses. But as long
as losses are uncorrelated, risk pooling works.
However, losses may be correlated. Outcomes at verdict/appeal may
be correlated. A particular judicial decision, for example, may be prece-
dent-setting. Juries and others may hear about a large award, and a string
of experts may be unduly influenced by a scientific theory that carries
over to a number of legal disputes. Or a state supreme court may find a
particular statute to be unconstitutional, and its decision in turn may
affect a number of pending cases. For any of these reasons for correlated
loss on the underwriting side of the insurer’s business, risk may not be
diversified away by insuring additional individuals or organizations.
In principle, diversification of assets on the investment side of the
insurer’s business should eliminate much market risk, and holding a large
number of assets should reduce credit risk.16 Insurers can hold shorter-
term securities and engage in various types of hedging activities to reduce
some risk. However, for various reasons, diversification will not elimi-
nate all such risk, and higher “nondiversifiable” risk is expected to be
reflected in higher premiums as well.
Another nondiversifiable risk relates to the relationship between the
variability of underwriting results and the variability of returns on a
well-diversified portfolio of securities. If returns on the sale of insurance
and on investments are negatively correlated (as in personal portfolios,
bonds tend to increase in price during recessions as interest rates fall
while stock prices tend to decrease), then selling insurance reduces the
overall nondiversifiable risk, and conversely if there is a positive relation-
ship. If losses fall when the economy is depressed, insurers’ underwriting
losses and performance of the general economy are negatively correlated.
36 Chapter 2

This implies that adding policyholders with uncorrelated risks reduces

overall nondiversifiable risk. Adding uncorrelated risks would in turn
increase the insurance firm’s share value and decrease the cost of raising
equity capital from external sources.
On the other hand, if underwriting losses rise when the general
economy is depressed, losses on the underwriting and investment sides
of the business are positively correlated. Then, following the same rea-
soning, adding policyholders increases the insurer’s cost of external
equity capital (the return investors demand for investing in the
company).
Some types of insurance are likely to be correlated with the business
cycle—for example, disability claims rise with the unemployment rate.
Medical malpractice payments do not seem to vary systematically with
or counter to the business cycle.17 Thus, although underwriting risks may
be correlated, overall underwriting losses and investment losses appear
to be uncorrelated. This suggests why a prudent insurer would be cau-
tious of putting all of its eggs in the medical malpractice insurance
underwriting basket, although many insurers, by specializing in a single
line—medical malpractice insurance—largely do this.

Dynamics of the Cycle

Before delving into explanations of cycles, it is important to have a

picture of changes in losses from underwriting, as well as in profits, as
the cycle progresses. This discussion and the accompanying figure (figure
2.1) are adapted from T. Baker (2005a), who presents a very clear
account of the dynamics of the cycle.
There are three lines in figure 2.1: operating profit, which is akin to
underwriting profit; initial incurred losses; and developed losses for
medical malpractice insurers (1980–2003). The period covered by figure
2.1 (1) follows the hard market that ended in 1978, (2) includes the
entire hard market of 1984–1987, and (3) includes part of the hard
market of 2001–2005.
Operating profit (from the income statement), reflecting underwriting
losses, becomes negative before the onset of the hard market, but reaches
its low around the beginning of the hard market, reflecting outlays on
 Why the Crises in Medical Malpractice? 37

50 $8,000,000
Operating Profit (in percent)

40 $7,000,000
30 $6,000,000

Losses (1000s)
20
$5,000,000
10
$4,000,000
0
$3,000,000
–50
–20 $2,000,000
–30 $1,000,000
–40 $0
1980 1985 1990 1995 2000

Operating Profit Developed Losses Initial Incurred Loss

Figure 2.1
U.S. Medical Malpractive Insurers’ Operating Profit and Loss Experience, 1980–
2003. Source: Baker (2005a).

previously underpriced insurance policies. Poor profit performance is

expected to be reflected in declines in the insurance companies’ share
prices.
Initial incurred losses are losses projected at the time the insurance
policy is sold. As figure 2.1 shows, initial losses (in nominal terms)
increase throughout the period, except briefly during the late 1990s,
when there is a decrease. There are substantial increases in such losses
during hard markets. What this means is that insurers substantially
increased their initial loss projections at these times. These are antici-
pated—not actual or realized losses—losses reported at the end of the
year during which the policies for the policy year were sold.18
The third line is for developed losses. These are calculated at ten years
after the policies were sold or as of the end of 2003, whichever is earlier,
and are assigned to the year in which the policies were sold. Developed
losses represent revisions based on “development” of losses for the policy
year, as well as other information revealed after the policy year, such as
changes in inflation rates and judicial decisions most relevant to the
insurance sold during a policy year. The beginning and the end points
of the two loss series are nearly identical. Prior to the hard market of
the 1980s, developed losses exceeded initial incurred losses, implying
that insurers were increasing their estimates of loss on policies they had
38 Chapter 2

previously sold. From 1986 to 1998, initial incurred losses exceeded

developed losses. Insurers then believed that they had been too pessimis-
tic about losses on these policies when the policies were sold, and they
subsequently revised their projections downward. In 1998, developed
losses again exceeded initial incurred losses, implying that insurers
increased anticipated losses after the policies were sold. This is not sur-
prising in view of the increase in the real value of paid claims that were
occurring then.19 The pattern from the 1980s shown in figure 2.1 is not
exactly repeated during the 2001–2004 hard market, but this could
reflect the lack of a ten-year follow-up for computing developed losses
for this period.
When insurers are revising their loss projections upward, the funds to
cover their additional losses have to come from somewhere. More specifi-
cally, the funds have to come from revenue from other lines of insurance
sold by the company for which the initial incurred loss proved to be too
pessimistic—from surplus and/or from additional cash flows obtained
from higher insurance premiums on policies from subsequent policy
years. The data in figure 2.1 are for the medical malpractice insurance
industry. All, or virtually all, firms were being similarly affected. Deterio-
rating financial performance is not an isolated case of mismanagement
by a single insurer, but rather of simultaneous, similar decisions by many
insurers.

Explanations of Insurance Cycles: Factors Internal to the Companies

and the Industry versus Factors External or Exogenous to the
Companies and the Industry

Overview
The above description of changes in losses in the balance sheet and in
profit as reflected in the companies’ income statements reveals nothing
about the causes of these changes. The trends in losses say nothing about
why sharp increases in premiums occurred, although higher premiums
are reflected in higher operating profits. For purposes of this discussion,
it is useful to distinguish internal factors from external factors accounting
for insurance cycles, the latter including changes in interest rates, changes
in the relationship between investment returns and underwriting returns,
 Why the Crises in Medical Malpractice? 39

depletion of capacity to write insurance, and changes in the relationship

between return.20 The external factors are typically exogenous (not influ-
enced by) to the insurance industry. These exogenous factors reflect
developments in the national or global economy as a whole, weather,
war, and so on.

Competition Among Oligopolists: An Internal Factor

The recurrent crises in medical malpractice insurance have been attrib-
uted to an oligopolistic structure of the medical malpractice insurance
market.21 In oligopoly22 there is a very small number of firms competing
in the same product market, with each firm taking into account the
actions and reactions of its competitors in setting its price. Each firm’s
demand curve is kinked at the current price, with demand being elastic
above and inelastic below the current price. Demand is elastic for price
increases because each firm assumes that competitors will not follow a
price increase; thus, raising price will result in large decreases in the firm’s
market share. Decreases in price, by contrast, will be matched by com-
petitors, with the consequence that the firm will not gain market share
by reducing price. Because of these demand conditions, oligopolists often
engage in nonprice competition rather than compete on price. Nonprice
competition is designed to establish consumer loyalty. In the medical
malpractice market, there might be competition on service following the
filing of a claim. Or the medical society-sponsored medical malpractice
insurers might hypothetically be able to convince some physicians that
they are more loyal to the medical profession than their commercial
competitors.
In their account of oligopolistic behavior as a reason for medical mal-
practice insurance cycles, Nye and Hofflander (1987) begin with the
observation that physician demand for malpractice insurance across the
entire industry is not very sensitive to its price (i.e., is highly inelastic).
However, the authors’ demand for an individual insurer’s product may
be very sensitive to price changes. Some doctors may be very quick to
switch to a different company if their current insurer raises premiums.
In such a situation, an insurer cannot charge a substantially higher
premium than its competitors. Further, the authors assert that insurers
face constant returns to scale, meaning that much larger firms do not
40 Chapter 2

inherently have a proportionally greater competitive (cost) advantage

than smaller firms. Although existing firms in a market may be in a price
equilibrium, a new entrant can rapidly build its market share by under-
pricing existing firms in the market. Relative ease of entry means that it
would be relatively easy for any single firm to dominate a given market
with a low-price strategy unless other firms follow its lead in pricing.
With these conditions established, the authors make the argument that
any price cut in premiums by one insurer, such as a new market entrant,
would result in price decreases by its competitors in an effort to avoid
losing customers and market share. The result would be a reduction in
premium levels but no change in the number of physicians insured.
However, eventually the insurers in the market would realize they are
verging on bankruptcy and raise premiums nearly simultaneously.
One would think that the consequences of a price reduction would be
anticipated in advance, and no single company would begin the price
reduction process. But a new entrant into the market needs to have a
way of enticing physicians in the market to switch insurers, and having
the new entrant brag about its great service or loyalty to the medical
profession may not be enough to induce physicians to switch. A price
incentive may be just what is needed.
Even without entry, an existing insurer in a market may become
myopic (see box 2.1). Under one scenario, a combination of entry and
myopic behavior occurs. A firm with poor management enters a market,
underestimates future losses, and sets premiums below the actuarial
value of the losses. Some of this does occur, as the PHICO experience,
described more fully in Chapter 8, illustrates. PHICO entered new insur-
ance markets and underpriced its medical malpractice insurance product,
which led to its demise. St. Paul, the largest seller of national medical
malpractice insurance, did not go bankrupt but may have seen itself
as a victim of price-cutting, and as a result dropped the medical mal-
practice insurance line at the beginning of the 2001–2005 hard market
(box 2.2).
As described in box 2.1, myopia may reflect internal conflict within
insurance companies.23 Fitzpatrick (2003–2004) discusses the interplay
of forces within insurance companies during various phases of the cycle.
During downturns in the cycle, when the companies are losing money
 Why the Crises in Medical Malpractice? 41

Box 2.1
The Price War of 1979–1981

Nye, Gifford, Webb, et al. (1988) describe a price war in the medical
malpractice insurance market during 1979–1981. At the time, nominal
returns on Treasury bills were very high (15.5 percent in 1981). With
interest rates on riskless securities this high, insurers aggressively reduced
premiums to gain market share. They expected to offset any underwriting
losses by increased investment income. According to the authors, this
practice, called cash-flow underwriting, is not necessarily unsound or
imprudent business practice. It represents a legitimate and desirable
response from insurers that benefits insurance buyers in the form of reduced
rates.
During 1979–1981, however, excesses of cash-flow underwriting
occurred, and virtually all insurers reduced premiums to preserve their
market shares. The result, in many instances, was cutthroat price competi-
tion. Insurance companies’ chief executives were faced with pleas from
heads of their marketing departments that unless prices were reduced, the
company would lose market share. Such claims, in most instances, pre-
vailed over the protests of actuaries that the premiums in 1981 were not
actuarially sound. This intense price competition resulted in premiums well
below actuarially sound premiums. The inevitable upward pressure on
insurance rates as a result of increasing paid claims was delayed and
obscured. In other words, in hindsight, premium rates for physicians and
others in 1981 were lower than they should have been (p. 1526).

and are under pressure to restore profitability, actuaries, who tend to be

conservative in premium-setting, become relatively influential in company
decision-making, and their advice, particularly in setting premiums, is
followed. By contrast, during prosperous times, underwriters, who may
be compensated on the basis of the volume of business they attract to
the organization, are relatively more influential.24 Underwriters often
lobby for lower premiums with the objective of expanding market share.
Their arguments are more persuasive in times of good financial perfor-
mance. Following this type of reasoning, we might observe substantial
increases in premiums as companies pay less attention to the opinions
of their underwriters and more attention to their actuaries, who tend to
be more concerned about the adequacy of insurance premiums as the
companies move in the direction of restoring profitability from the
bottom of the cycle.
42 Chapter 2

Box 2.2
St. Paul Stops Selling Medical Malpractice Insurance in 2001

During the hard market of the 1980s, St. Paul increased its reserves. But
by decade’s end, claims frequency and severity had leveled, and the
company concluded that it had overreserved. The company released $1.1
billion in reserves during 1992–1997. According to a newspaper report
(Zimmerman and Oster 2002), “The money flowed through its income
statement and boosted its bottom line. St. Paul stated clearly in its annual
reports that excess reserves had enlarged its net income. But that part of
the message didn’t get through to some insurers—especially bedpan
mutuals—dazzled by St. Paul’s bottom line, according to industry officials.
In the 1990s, some bedpan mutuals began competing for business beyond
their original territories. . . . With St. Paul seeming to offer a model for
big, quick profits, ‘no one wanted to sit still in their own backyard, says
Scipie’s [a California-based physician-sponsored medical malpractice
mutual] Mr. Zuk. The boards of directors said, ‘We’ve got to grow.’ Scipie
expanded into Connecticut, Florida, and Texas, among other states, start-
ing in 1997. . . . The newer competitors soon discovered, however, that
‘the so-called profitability of the ’90s was the result of those years in the
mid-’80s when the actuaries were predicting the terrible trend, says Donald
J. Fager, president of Medical Liability Mutual Insurance Co. [New
York]. . . . The competition intensified, even though some insurers ‘knew
rates were inadequate from 1995 to 2000’ to cover malpractice claims,
says Bob Sanders, an actuary. . . . In the late 1990s, the size of payouts for
malpractice awards increased, carriers say. . . . St. Paul’s malpractice busi-
ness sank into the red. The new Chief Executive announced that the
company would drop the coverage line. St. Paul reported a $980 million
loss on the [malpractice] business for 2001” (p. A1).

Some studies have focused on errors in loss reserving as a cause of

insurance cycles. If companies act rationally, but face chance variation
in losses for reasons beyond their control, there will be errors in loss
reserving; however, these errors should also be random. Using the loss
reserve errors shown in figure 2.1, the differences between initial incurred
losses and developed losses seem serially correlated, not random.
Such nonrandom variation could reflect several types of behavior.
First, loss forecasts could be based on past losses rather than on a full
consideration of factors likely to occur in the future.25 Any changes in
initial incurred loss, which has increased sharply at the beginning of hard
markets in medical malpractice, should have much greater effects on
 Why the Crises in Medical Malpractice? 43

premiums in a long-tail line, such as medical malpractice, than in lines

with shorter tails. An increase in anticipated inflation from 2 percent to
3 percent will have a larger effect on future anticipated losses if the
payout period is seven years than if it is two years due to
compounding.
Second, because stockholders desire stability in earnings, firms could
manipulate losses to reduce the volatility in earnings.26 However, this
second explanation encounters two problems of its own. First, many
malpractice insurers, organized as mutuals, reciprocals, or some alterna-
tive to the stock company form, have no stockholders. In recent years
there has been substantial growth of self-insured plans run by large
hospitals and large groups of physicians.27 These organizations have no
incentive to manipulate their losses either. Second, profits are volatile,
and hardly seem to have been substantially smoothed. A succession of
inflation shocks or adverse (to companies) judicial decisions could also
generate serial correlation in loss reserve errors.
In sum, competitive behavior among companies in the industry can
account for some aspects of the cycle but not for others. It potentially
describes pricing behavior. Nominal prices of medical malpractice insur-
ance fall in some areas during soft markets.28 Whether or not nominal
prices decrease, insurers interested in gaining market share offer insur-
ance at premiums below the actuarial value (expected value) of the loss.
Price wars and below actuarially fair pricing may result as a response to
an entrant’s low price and/or to high profitability, as was the case with
St. Paul before it exited (box 2.2).
Such price wars could in principle be prevented by state insurance
regulation. Although insurance regulation is found in some form in all
U.S. states, not all states regulate premiums. And even when they do, a
small line of insurance, such as medical malpractice insurance, may be
below state regulators’ radar in states where the focus is on much larger
lines of insurance. Moreover, state regulation of premiums may itself be
a cause of insurance cycles. Insurance regulators may be too slow to
approve justifiable premium increases.29
The notion that there are internal conflicts about pricing within firms,
and that the relative influence of marketing personnel and actuaries
within companies in premium decisions is cyclical, is plausible, especially
44 Chapter 2

since the same story comes from several different sources.30 This is a hard
topic to study quantitatively since discussions within firms are private
information. It also seems plausible that publicly traded companies are
sensitive to the effects of company earnings reports on short-run fluctua-
tions in the share price, a type of thinking that is not unique to insurance.
However, many medical malpractice companies are not publicly traded.
Thus, the importance to be attached to this latter explanation of company
behavior is correspondingly smaller.
Industry pricing dynamics do not explain withdrawal of coverage at
the beginning of hard markets. In addition, it seems a bit of a stretch
that the large changes in loss reserves associated with cycles merely reflect
efforts to reduce the volatility in reported earnings. Under the rational
expectations hypothesis, an insurer’s predictions of future losses should
be unbiased.31 Thus, although the ex ante forecasts would not coincide
with the realized losses, taken over a number of years, the difference
between realized and anticipated losses should be small. This is true
especially for the industry as a whole, since decision makers use all avail-
able information in making decisions, and errors made by individuals
are offset by errors by other individuals in the opposite direction, which
is the “wisdom of crowds.”32
Finally, increases in payments to policyholders and for ALE do not
explain short-run fluctuations. While increases in such payments should
be fully reflected in higher premiums in the long run, the cycle is about
the short run.

Exogenous Factors: Changes in Returns on Interest-Bearing Securities

Part of the premium-setting process involves anticipating the pattern of
future interest rates over time. Even if insurers’ expectations of future
interest rates are unbiased, they may make major mistakes in estimating
future interest rates after the fact. For example, Harrington and Litan
(1988) contend that in the early 1980s, insurers did not anticipate the
substantial decrease in interest rates that occurred later in the decade.
As a result, premiums were set too low, and were raised after the sub-
stantial decrease in interest rates became apparent.
Real interest rates were declining before each of the three hard markets
since 1970. However, nominal interest rates33 increased before the hard
 Why the Crises in Medical Malpractice? 45

market of the 1970s, the difference between trends in nominal and real
rates reflecting high inflation during this period.34 Nominal rates decreased
before the onset of the other two hard markets. It seems unlikely that
insurers overpriced medical malpractice insurance on the basis of falling
interest rates, since, with history as a guide, they almost certainly would
have increased within a few years. Interest rates at best explain only part
of cyclical underpricing of insurance premiums.

Exogenous Factors: Unanticipated Inflation Shocks

In principle, anticipated losses would be revised upward in response to
anticipated increases in prices, particularly in medical prices. Here again,
there is no consistent pattern. The late 1990s and early 2000s, for
example, were years of relative price stability, including medical
prices.35

Exogenous Factors: Capacity Constraints

Financial models of insurance pricing based on various standard frame-
works36 link premiums to the present value of expected claims expense
plus a loading factor.37 In these models, the cost of capital to the insurer
does not vary with the amount of insurance the company sells.
A separate group of models embodies capital constraints.38 In these,
the cost of raising a dollar of capital (marginal cost of capital) rises with
the amount of insurance sold. To the extent that the cost rises steeply or
is vertical at a certain threshold of insurance sales (i.e., at a specific level
of underwriting volume, no extra capital is forthcoming at any price)
this places an effective limit on the amount of insurance that will be sold.
Capacity constraints may go at least partway to explaining a puzzle
about cycles, namely, the refusal of insurers to offer coverage at any
premium.39
Insurance contracts represent liabilities of insurers. As a matter of
prudence of the companies and a requirement of state insurance regula-
tors, companies need assets to match the liabilities. A problem arises if
the insurer lacks assets to back the sale of insurance policies. Because
companies have made upward revisions on policies sold in prior years
or for some other reason, they may lack internally generated funds to
back the sale of insurance policies.
46 Chapter 2

The capacity constraint explanation begins with the notion that a

dollar of externally generated capital may be more expensive than a
dollar of internally generated capital. For insurers that are not organized
as stock companies, such as mutuals and reciprocals, and certainly for
self-insured plans, there is practically no recourse at all to equity supplied
by investors. These organizations must fully rely on internally generated
capital, such as profits derived from premium income.
To see how the capacity constraint works, suppose that the firm is
initially at the equilibrium P0 and q0 (figure 2.2). In the short run, the
firm would issue more insurance policies, but along the supply curve of
insurance S0. While in the long run the firm can expect growth in inter-
nally generated capital at a constant (opportunity) cost—hence the curve
SLR—in the short run, extra capital could be obtained only at relatively
unfavorable rates, which is reflected in S0. Now suppose the firm experi-
ences a loss shock due, for example, to a string of adverse judicial deci-
sions,40 reflected in increased developed losses. Then, even if the demand
for insurance curve (D) is stable, the equilibrium price of insurance will
increase in the short run. For other types of insurance, such shocks may
be the result of natural disasters or terrorism.

P
S1 S0
D

P1

SLR
P0

q0 Q

Figure 2.2
Insurance Market Equilibrium.
 Why the Crises in Medical Malpractice? 47

In lines of insurance in which losses from underwriting are relatively

volatile, underwriting must be covered by relatively large amounts
of surplus. When insurers’ capacity is high, relatively volatile lines are
allocated comparatively more surplus in relation to premium volume.
However, when capacity is low, the uncertain or capacity-using lines are
allocated the least capacity. At the same time, one observes withdrawal
from those lines or less withdrawal coupled with steep premium increases.
In lines with more predictable losses, such as automobile liability, avail-
ability of insurance is not curtailed as much and premiums remain
relatively stable.
Using aggregate time series data on stock property-casualty insurers
for the years 1935–1997, Choi, Hardigree, and Thistle (2002) report
evidence that an increase in surplus or surplus growth leads to lower
premiums in the short run. Doherty and Posey (1987) likewise find
support for an effect of surplus on insurance availability, using a sample
of stock general liability insurers from A. M. Best for 1979–1989.
However, applying these results to medical malpractice insurance should
be done with some caution, since stock companies are declining in
market share in the medical malpractice line. Also, although the capacity
constraint explanation seems plausible, and some evidence supports the
capacity constraint hypothesis, other evidence does not, or provides
weak support.41

Exogenous Factors: Role of Reinsurance in Cycles

Another factor exogenous to primary medical malpractice insurers and
self-insuring organizations in the United States is the price and avail-
ability of reinsurance. As explained more fully in Chapter 10, insurers
purchase reinsurance, termed excess loss insurance when purchased by
a self-insured organization, such as a hospital, to protect itself against
large, unpredictable losses. Medical malpractice insurers, especially small
single-line insurers, are highly dependent on reinsurance as a hedge
against risk.42 According to interviews conducted with reinsurers
described by the U.S. General Accounting Office (June 2003), reinsur-
ance premiums increased at the beginning of the hard market of 2001–
2005 for two reasons. First, there was a general increase in reinsurance
premiums following the terrorist acts of September 11, 2001. This may
48 Chapter 2

have reflected a combination of reinsurers’ reassessment of their expo-

sure to serially correlated losses following this event and the loss of
surplus. Second, for medical malpractice insurance in particular, reinsur-
ers had experienced increased loss payments since about 1998, and for
this reason raised reinsurance premiums to customers in this line of
insurance substantially, with increases ranging from 50 to 100 percent.
Higher reinsurance premiums were in turn passed forward to medical
malpractice insurers and then to their customers. To hold premium
increases down, some insurers raised the dollar threshold of loss before
reinsurance coverage applied.
Many of the reinsurers are not domiciled in the United States, but
rather in Caribbean countries, where they are not subject to corporate
taxation and much insurance regulation, or in Europe (e.g., Swiss Re,
Münchener Rückversicherung). Insurers and self-insured groups contract
with several reinsurers which cover various levels of loss (“layers”).
Insurance brokers often serve as intermediaries in developing reinsurance
packages for clients—primary insurers and self-insuring organizations.
Unlike many primary medical malpractice insurers, reinsurers are
involved in multiple lines of insurance. They are also subject to the
shocks (losses attributable to natural disasters, terrorism) described
above. However, primarily because of state regulation of insurance, the
market area for medical malpractice insurance in the U.S. is limited to
individual states, whereas the market for reinsurance is global.
A reinsurer/excess insurer similarly faces risks from insuring against
large claims. By definition, large claims are rare. When claims are infre-
quent, it is more difficult to judge whether an individual claim is a
random event or represents an underlying change in claiming; the latter,
but not the former, would require a change in the premium.
The fact that claims are resolved in the far distant future is problematic
for the reinsurer/excess insurer because it increases the likelihood that
events affecting claim resolution will intervene. For example, a high jury
verdict in a jurisdiction may be precedent-setting, affecting how juries
decide other claims as well as settlements which mirror jury verdicts. To
the extent that the precedent affects all large claims in the jurisdiction
(i.e., the correlation in claims outcomes increases), the price of coverage
is likely to increase to reflect this risk. In fact, in markets for coverage
 Why the Crises in Medical Malpractice? 49

of catastrophic risk, the premium may be several times the expected value
of the loss for the policy year.43 For these reasons, a few large payouts
may lead to substantial increases in reinsurance/excess insurance premi-
ums. These, in turn, are paid directly by self-insured entities or are shifted
forward by primary insurers in the premiums they charge physicians and
other primary insurance premiums.
Particularly because many reinsurers are not subject to the same extent
of public regulation as other insurers, and many reinsurers are not public
corporations, much less is known about reinsurers than about primary
insurers. There is virtually no academic literature on the subject of rein-
surance. Froot (2001) provides the most in-depth analysis of reinsurance
markets in general. His article does not specifically deal with reinsurance
for medical malpractice, however. Nor does it consider events that have
occurred since 2000. Froot examined several nonmutually exclusive
hypotheses to explain why reinsurance premiums were a multiple of
expected loss, even prior to 2000. As discussed in Chapter 10, Froot
does not reach a definitive conclusion as to why premiums are so high.
Although he does not discuss reinsurance after the year 2000, a combina-
tion of shocks to capacity from natural disasters and 9/11, and some
unanticipated payments on medical malpractice claims, seems like the
most plausible reason for the withdrawal of coverage and substantial
reinsurance premiums charged to malpractice insurers and the self-
insured for medical malpractice in hospitals that occurred after 2000.

Conclusion

Insurance cycles reflect a number of factors. Several conclusions can be

reached with near certainty. Others are much more uncertain.
First, insurance cycles, like business cycles and cycles in weather,
clearly are here to stay. There may be changes in amplitude, but we can
expect cycles to reoccur with certainty. Second, it is far easier to say what
are not causes of cycles than to say definitively what causes them.
Prominent among the noncauses are (1) attempts of insurers to recoup
losses on poorly performing investments by raising premiums and (2)
exercise of market power by insurers to raise premiums well above actu-
arially fair values. On the former, the market for medical malpractice
50 Chapter 2

insurance is too competitive to allow for such actions. Even in markets

with a single seller, entry of other sellers is always a possibility. On the
latter, in many states there is more than one insurer competing for the
business of physicians and hospitals which do not self-insure for medical
malpractice. There is no evidence that insurers collude. In fact, the notion
that a medical society-sponsored mutual medical malpractice insurer
would collude with a for-profit insurer, such as St. Paul when it was in
the medical malpractice insurance market, is implausible on its face.
Having eliminated the noncauses, we are left with considering possible
causes. Underpricing of insurance is a common phenomenon leading up
to crises. Shocks are also factors, even though the shocks pertinent to
each cycle differ. Substantial changes in interest rates and prices that
characterized the 1970s and 1980s did not reoccur after 2000. In con-
trast, the massive shocks from natural disasters and 9/11 were more
likely factors after 2000.
The relative importance of precipitating causes of cycles may well
differ among cycles. Similar to motor vehicle fatalities, we can be sure
that they will occur in the future. In some seasons, they may be predomi-
nantly due to slick roads. In others, they may be due to heavy drinking
and driving on holidays, reflecting fewer police patrols.
If there is nothing to be done about preventing cycles, what can be
done to reduce their amplitude and disruptive effects? Among the possi-
bilities discussed in later chapters are public provision of reinsurers—
public reinsurers being much less likely to withdraw coverage during
hard markets—and hospital provision of medical malpractice insurance
for all care provided within the walls of the hospital. Hospitals would
cover much primary insurance on a self-insured basis with reinsurance
for the large losses. Here, too, the hospital would be far less likely to
withdraw coverage during hard markets. However, it should be clear
that there is no free lunch. Each proposal has its own pluses and
minuses.
3
An Increased Threat of Lawsuits and Higher
Premiums: The Consequences

Is there a doctor in the house? Increasingly, in Florida and around the country,
the answer is no—not in the house, not in the doctor’s office, and not in the
hospital. Many physicians are choosing to retire early or to practice in other
states because medical malpractice insurance in Florida has become unaffordable
and, in some, cases, unavailable.

Thus begins the Executive Summary of the report of the Governor’s

Select Task Force on Healthcare Professional Liability Insurance, State
of Florida, submitted to Governor Jeb Bush in early 2003.1 The same
report finds that “The concern over litigation and the cost and lack of
medical malpractice insurance have caused doctors to discontinue high-
risk procedures, turn away high-risk patients, close practices, and move
out of the state. In some communities, doctors have ceased delivering
babies and discontinued hospital care.”2 The report concludes that previ-
ous tort reforms have failed and that limitation of damages, the only
provision proven to be effective in reducing the severity of judgments,
was struck down by the Florida Supreme Court.3
In the political debate about tort reform during 2000–2005, the
high cost of personal health services in the United States was frequently
attributed to litigation and the high cost of medical malpractice insur-
ance. For example, President George W. Bush explained in a speech
delivered on January 5, 2005, “Many of the costs we are talking about
don’t start in an examining room or an operating room, they start in a
courtroom.”4
A decade and a half earlier, Danzon, Pauly, and Kingston5 made the
following observation in response to similar assertions: “Although mal-
practice insurance is still less than 2 percent of total health care expen-
ditures, many observers argue that medical malpractice is a major factor
52 Chapter 3

contributing to rising health care costs. One implication of such argu-

ments is that virtually all the cost is borne by patients. This seems at
odds with arguments, made by others, that a significant fraction of physi-
cians are giving up practice, or at least high-risk procedures, because of
liability.”
This chapter presents evidence on two of the five myths of medical
malpractice listed in chapter 1. First, myth 2—Only “good” doctors are
sued. We will see that both high- and low-quality physicians are sued
for medical malpractice. Second, myth 5—Medical care is costly because
of medical malpractice. At most, only a small part of the growth in real
expenditures on personal health services in the United States can be
attributed to medical malpractice. A focus of this chapter is whether
there is empirical evidence that medical malpractice is to blame for creat-
ing access barriers to health care services, and has increased the cost of
such services. Much of the conventional wisdom is that there is clearly
a link of sufficient importance to require statutory changes. The effects
of damage caps will be discussed in the next two chapters.
In a democracy such as ours, each of us has a right to pursue our
private interest in a public forum. Throughout history, farmers have
lobbied for price supports; the steel industry has argued for tariffs on
foreign steel. Domestic automobile manufacturers have sought to restrain
others who would impose more stringent gasoline mileage standards on
the automobiles that they manufacture. Pursuit of private interest is not
necessarily contrary to the public interest, but the two frequently do not
coincide. Lower medical malpractice premiums and lower probabilities
of being sued are certainly in the private interest of those who advocate
tort reform. However, these policies may or may not serve the public
well.
Lobbying on the basis that physicians have suffered reductions in their
net income because premiums have risen is not likely to be popular with
voters since physicians earn relatively high incomes. Therefore, the ratio-
nale for policy intervention is often cast as a need to help the public;
assertions that patients have been disadvantaged by medical malpractice
are common.
A major question this chapter addresses is whether the empirical
evidence establishes a link between rising premiums and claim severity,
 An Increased Threat of Lawsuits and Higher Premiums 53

on the one hand, and patient access to medical services and spending on
personal health care services, on the other. The issue is not whether there
is ever a link. For example, some physicians may have left practice for
reasons at least partly related to medical malpractice. Rather, it is impor-
tant to know whether there is evidence of widespread access barriers the
existence of which can be attributed to rising premiums and threat of
litigation. Further, the issue is not whether there are access barriers; the
existence of access barriers has been amply documented. The issue is,
compared to other causes of access barriers to patient care, how impor-
tant medical malpractice is as a cause. All of the discussion assumes that
premiums and outlays for awards have risen appreciably. The data pre-
sented below, however, do not show appreciable increases over long time
periods.
Another potential impact is on spending for personal health care ser-
vices. Such expenditures have clearly increased, but how much of the
increase can be attributed to medical malpractice, including the threat
of being sued?
The public discussion of the issue of defensive medicine has been par-
ticularly confused. Any increase in spending on personal health care
services attributable to the threat of litigation is often considered waste-
ful. However, a well-functioning tort liability system would encourage
physicians to order procedures for which benefit exceeds cost. As effec-
tive diagnostic tests and procedures are developed, their use should be
encouraged, not discouraged. A malfunctioning tort system would
encourage the use of procedures for which cost falls short of benefit.
Thus, it is not enough to demonstrate that expenditures have risen as a
consequence of medical malpractice; it is also necessary to document that
such expenditures were wasteful, measured in cost/benefit terms. Such
documentation is very rarely provided.
Referring back to an example provided in chapter 1, if the cost of a
diagnostic test (to all parties, not just the out-of-pocket cost to the
patient) is $1,000 per test, this test is appropriately provided to all
patients for whom the benefit is $1,000 or more (again considering all
benefits when benefits accrue to persons other than the patient, as would
be the case, for example, in testing for AIDS). In the case in which the
benefit is $1 and the cost is $1,000, the test should not be conducted.
54 Chapter 3

This is an admittedly economic viewpoint. In what some describe

as a medical model, a test should be provided if the benefit is positive.
Thus, the test would be provided in cases in which the benefit is $1
and the cost is $1,000. When one describes this “medical model”
in abstract terms, it seems to makes sense. Few, however, would
recommend testing when the benefit versus cost is as stark as our
$1 versus $1,000 example. Perhaps a fully insured patient would
demand this, but few would support such decision criteria as sound
public policy.
Benefit includes nontangible benefits, such as relief of pain and suffer-
ing, and of anxiety, which are very real and sometimes avertable costs.
For example, people spend a great deal of money for painkillers and for
anxiety relief. These costs are very real.
Often physicians are quoted as stating that they order tests and pro-
cedures in excess of what they would order in the absence of a threat of
being sued. This could be so in a well-functioning system as well as a
dysfunctional one, since an objective of tort is to provide an incentive to
use tests and procedures for which benefit exceeds cost in an economist’s
sense. There is undoubtedly a lot of waste (care for which cost exceeds
benefit or even care for which the benefit is zero) in the U.S. health care
system. Nevertheless, allegations that health care spending is high because
of medical malpractice generally do not include an accompanying analy-
sis of the effects of other factors on such spending. Waste would almost
surely exist even if there were no threat of lawsuits.
A distinction is made between “positive” and “negative defensive
medicine.” Positive defensive medicine refers to increases in the cost of
personal health care services attributable to the threat of being sued.
Negative defensive medicine applies to withdrawal of care due to retire-
ments, changes in physician location, and dropping of particular proce-
dures which often lead to lawsuits, such as obstetrical care. A rigorous
definition of positive defensive medicine should be limited to provision
of services for which the added cost is less than the anticipated benefit,
not to increased cost per se.
If only premiums are considered, it is difficult to make a case that tort
liability has had a major impact on the cost of medical care.6 As docu-
mented below, U.S. health expenditures exceed medical malpractice pre-
 An Increased Threat of Lawsuits and Higher Premiums 55

miums by a factor of 50 or more. The trend in the rate of growth in

premiums relative to inflation has been modest. In addition, about half
of a premium goes to pay injury victims and involves a transfer of
income, not the use of scarce resources. For this reason, to obtain a large
effect of medical malpractice on health care spending, one must incor-
porate the cost of positive defensive medicine in addition to higher pre-
miums. By creating access barriers, negative defensive medicine may
actually reduce health care spending.
The next section discusses what economic theory has to say about
effects of increased medical malpractice premiums on physicians’ fees
and incomes; this is followed by a description of trends in medical mal-
practice premiums, claims frequency, payment per paid claim, and an
empirical evaluation of the effects of increased medical malpractice pre-
miums on physicians’ fees and incomes. The next three sections are
organized according to the trilogy often used to assess health care: access,
cost, and quality or injury deterrence. We then review evidence on the
association between a physician’s claims history and indicators of quality.
The final section summarizes the empirical evidence and implications
based on the findings.

Effects of Premium Increases on Physicians’ Fees and Incomes: Theory

and Empirical Evidence

Economic theory offers predictions about price and output responses to

an increase in the price the firm pays for an input used to produce its
output, but the predictions depend on the underlying assumptions. As a
practice expense, medical malpractice premiums are an input, as are
expenses on aides and medical supplies, in the production of practice
outputs.

Premiums as a Fixed Cost to a Physician’s Practice

Premiums do not typically vary with physician hours of work or the
volume of services delivered, and hence are plausibly viewed as a fixed
cost to the physician’s practice. The obstetrician who delivers five babies
a year pays the same premium as a colleague in the same geographic
location who delivers 150 annually.
56 Chapter 3

A lesson from the microeconomics7 of the firm, at least at its most

basic level, is that an increase in the price of fixed input reduces the firm’s
profit, but does not affect the firm’s product price or its output (sales).
For physicians, a reduction in profit is typically equivalent to a reduction
in net physician income.8
Although premiums do not reflect physician hours of work or practice
volume, they do reflect the types of services a physician provides. For
example, medical malpractice premiums can be reduced appreciably if a
physician drops obstetrical deliveries from his or her practice. Thus, one
would expect that raising premiums for physicians who deliver obstetri-
cal care would reduce the number of physicians who provide such care.
The decrease in supply should be particularly concentrated among those
physicians with low obstetrical caseloads prior to the increase in the
premiums among physicians who deliver such care.

Premium Changes as Physician Supply in a Market Changes

Immediately following an increase in the premium, physicians may make
some alterations in their practices, but they are unlikely to move or quit
practice. Such changes become more likely over time. To the extent that
increased premiums reduce net incomes of physicians, the supply of
physicians in the area experiencing the premium increase may fall in the
long run. A drop in net income can induce older physicians to retire, and
younger physicians may be less prone to enter practice in locations with
high premiums. Other physicians may migrate to areas with relatively
smaller premium increases. Physicians’ fees may increase as the number
of physicians in an area decreases. For example, a managed care orga-
nization (MCO) is likely to be in a poor bargaining position vis-á-vis
a physician who is the only physician performing obstetrical deliveries
in a geographic area. Women do not want to drive long distances
when in labor, and MCOs unable to include an obstetrician in their local
networks can expect to lose market share.
To the extent that the amount the insurer pays physicians eventually
rises to reflect increased medical malpractice premiums, the increased
premiums would lead to higher fees actually paid to physicians. This
seems rather unlikely in markets in which there are many physicians and
relatively few MCOs, but may occur in areas in which physicians have
 An Increased Threat of Lawsuits and Higher Premiums 57

more bargaining power due, for example, to their relative scarcity. In

addition, some insurers, most notably Medicare, base fee schedules in
part on medical malpractice premiums in the geographic area and the
specialty. Medicare fee schedules change slowly in response to changes
in medical malpractice premiums.9

Premiums as a Variable Cost to a Physician’s Practice

Physicians, however, may not view increased premiums simply as an
increase in fixed cost. Rather, they may view the premium increase as a
signal that the probability of being sued, or being sued for a lot of money,
has risen. Each additional patient seen represents another chance to be
sued. And when premiums grow, that chance of being sued increases. In
contrast, if premiums are seen as a fixed cost, seeing another patient does
not increase the malpractice cost. But if each additional patient adds to
the overall risk of being sued, then seeing additional patients does add
to the malpractice cost, and doctors will reduce practice volume and raise
price, if they have the market power to do this.
With the growth of managed care and changes in physician reimburse-
ment, such as Medicare’s change in its method of paying physicians in
the late 1980s, individual physicians have lost a great deal of power to
set prices for their services. To the extent that this has occurred, physi-
cians may wish to raise their fees in response to a medical malpractice
premium increase, but be unable to do so. Rather, they must wait for
offers from insurers to raise their fees in response to changes in practice
expense.

Effects of Premium Changes on Positive Defensive Medicine

The above simple framework can explain fee changes that mirror changes
in premiums. The model is equally appropriate for analyzing visits to the
barber and barbers’ fees.
With positive defensive medicine, the physician orders more tests,
more surgical procedures, and more follow-up care than the patient
would if he or she was fully knowledgeable about the efficacy of the
tests, procedures, and follow-up care in the same situation. Models that
allow for such physician-induced demand are necessarily more complex,
and a detailed discussion of these models would take us far afield.10
58 Chapter 3

While less specific models seem more realistic, the cost of adding general-
ity is that even some of the above predictions no longer hold. For
example, an increase in a fixed practice expense leading to decrease in
practice income could cause physicians to supply more rather than less
output.

Summary of Theoretical Results

If the discussion of theory to this point seems like another product of a
two-armed economist, weighing options on the one hand as well as the
other hand, this is not a false impression. Theories are useful for provid-
ing a conceptual framework for analysis, but predictions are sensitive to
the assumptions underlying the analysis. Using one set of assumptions,
an increase in medical malpractice premiums will have no effect on physi-
cians’ fees or on positive defensive medicine, but will increase negative
defensive medicine. Under alternative assumptions, a rise in premiums
will increase fees and positive defensive medicine. To the extent that
these increase, there would be less of a tendency to leave practice. In
other words, negative defensive medicine would decrease.
There is a limit to how far logic, economic or noneconomic, can take
us. Logic does, however, suggest there are trade-offs between positive
and negative defensive medicine. A physician who orders many tests
because the threat of lawsuits has increased, and for which he or she is
compensated, is not as likely to experience a major loss of income and
leave practice. Economic theory is often good for generating hypotheses
about relationships between variables, but the proof of the pudding is
in the empirical evidence.

Empirical Evidence: Trends in Premiums, Claims Frequency, and

Payment per Paid Claim

Actual trends in medical malpractice premiums have been more moder-

ate than much of the rhetoric asserts. Using data from national surveys
of physician practice conducted by the American Medical Association
during 1970–2000, Rodwin et al. (2006) document trends in self-
employed physicians’ mean malpractice premiums, total practice
expenses, and net income. Mean medical malpractice premiums were
 An Increased Threat of Lawsuits and Higher Premiums 59

$18,400 in 2000, which amounted to 7.5 percent of total practice

expenses and 3.8 percent of gross physician revenue. Three decades
earlier, before any of the major crises in medical malpractice, medical
malpractice premiums as percentages of total practice expenses and gross
physician revenue were 5.5 percent and 2.0 percent, respectively.
In 1975, liability costs were 0.91 percent of U.S. health care spending.
In 2002, liability cost as a share of health care spending had risen to
1.58 percent. In 1975, physicians paid 50.7 percent of liability cost in
premiums, and hospitals paid 41.0 percent. By 2002, the shares were
57.7 and 30.2 percent, respectively.11
These changes hardly seem dramatic, raising the question “Where’s
the beef?” Following spikes in premiums, physicians, hospitals, and other
health care providers have sought relief, pointing fingers at others, the
alleged culprits, while making widespread appeals to politicians.
Admittedly, as Rodwin and his coauthors acknowledge, these esti-
mates obscure areas of medical practice and specialties in which premium
growth was more substantial. On the other hand, there are areas and
specialties for which premium growth was less than average. Further,
because the AMA discontinued its surveys of physician practice after
2000, these estimates exclude the premium growth that occurred after
2000. Between 2000 and 2004, according to data from the Medical
Liability Monitor, mean premiums paid by obstetrician/gynecologists
increased by about 70 percent in real terms (2004 dollars). General
surgeons’ mean premiums about doubled. However, the increase for
specialists in internal medicine was much lower, about 50 percent
(figure 3.1).
Nevertheless, as already noted, it seems likely the “beef” is really about
more than the increase in premiums. Even though medical malpractice
insurance covers the dollar loss of the claim and the associated legal
expense, there are other costs to physicians and hospitals, such as stress
and loss of time in defending the cases.
Medical malpractice claims frequency spiked in 1975 and 1985, both
crisis years, but claims frequency has increased slowly since the mid-
1990s. The crisis that occurred in the early 2000s was associated with
a modest increase in claims frequency at most. In some states, claims
frequency declined.12
60 Chapter 3

$90,000
$80,000 Internal Medicine OB/GYN
General Surgery
$70,000
$60,000
$50,000
$40,000
$30,000
$20,000
$10,000
0
1991 1993 1995 1997 1999 2001 2003

Figure 3.1
Mean Insurance Premiums in Three Specialties, 1991–2004 (2004 $). Source:
Medical Liability Monitor surveys (1991–2004).

However, payment per claim has increased substantially since the mid-
1990s (figure 3.2). Relationships between medical malpractice premi-
ums, claims frequency, mean payment size, and total payments are
complex. In particular, payments in a particular year come from insur-
ance policies sold in the past, not from policies sold in the same year.
The more fundamental relationship is between the premiums set in a
particular year and anticipated losses in future years on claims filed in
the policy year.

Empirical Evidence on Effects of Changes in Premiums

Danzon et al. (1990) report results of cross-sectional and time series

analyses on the effects of changes in premiums on physicians’ fees. Using
cross-sectional data from the 1970s, they find that premium increases
boosted fees by more than can be explained solely by the effect of the
increase on overall practice expense. In other words, there was more than
100 percent forward shifting of the burden of the premium increase onto
patients and insurers. If more than a full pass-through of the additional
cost occurred, physician incomes would have risen as a consequence of
premium increases. However, the authors’ time series analysis of effects
of premium changes during 1976–1983 reveals that the increased pre-
miums had no impact on physicians’ incomes.
 An Increased Threat of Lawsuits and Higher Premiums 61

$400,000
$350,000 Mean
Median
$300,000
$250,000
$200,000
$150,000
$100,000
$50,000
$0
1988 1990 1992 1994 1996 1998 2000

Figure 3.2
Claims Payouts, 1988–2002 (2004 $). Source: PIAA Data Sharing Project, in
Smarr (2003).

Danzon et al. explore several possible reasons for the lack of change
in incomes, the most plausible of which is that insurers increased reim-
bursements to physicians by the amount of the premium increase. More-
over, physicians could have increased fees for reasons besides the
reimbursement increase, which led to a compounding effect. But given
the important changes in insurer payment practices that have occurred
since the early 1980s, it is not at all certain their results would generalize
to the post-2000 period.13
Only one study has updated Danzon et al.’s (1990) analysis. This
study, described in Pauly (2006), covering the years 1994, 1998, and
2002, reached essentially the same conclusion as Danzon et al. (1990),
a rather surprising finding, given the changes in insurer payment prac-
tices that have occurred since the early 1980s. The bottom line is that
the empirical evidence indicates that increases in medical malpractice
premiums lead to physician fee increases. Patients, taxpayers, and
premium payers—and not physicians—ultimately bear the burden of
malpractice premium increases.
Pauly adds a cautionary note, however, that on average, the conclusion
that physicians do not bear the financial burden of increased premiums
appears to hold. It is quite possible that when premiums increase, some
physicians lose, and others gain, income. Perhaps those physicians who
lose financially are the most vocal in demanding relief from the burden
of medical malpractice.
62 Chapter 3

Data collected by the American Medical Association indicate that

physicians’ incomes rose in real terms over the period 1970–1996, but
declined between 1996 and 2000, just before the most recent sharp rise
in medical malpractice premiums occurred.14 The decline in real income
between 1996 and 2000 did not result from a rise in premiums, as pre-
miums remained constant in real terms during these years. As of 2000,
medical malpractice premiums were 7.5 percent of total practice expenses,
up from 5.5 percent in 1970.15

Does Medical Malpractice Reduce Patient Access to Care? Physician

Supply

Aggregate Trends and Disparities by Physician Specialty and Geographic

Location
Starting about the late 1960s, the number of physicians relative to the
to U.S. population increased appreciably. Between 1980 and 2003, for
example, the number of physicians in the United States nearly doubled.16
Supply has grown even faster in some states, such as Florida, which
seems inconsistent with the quotations at the beginning of this chapter.
In 1980, the U.S. Graduate Medical Education National Committee
released a report concluding that the United States faced a massive phy-
sician surplus by the year 2000. Although by various indicators this
surplus did not materialize,17 there were no widespread policy concerns
that the supply in the aggregate was insufficient. Concerns have been
expressed about excess supply in some physician fields, especially surgery.18
Disparities in the geographic distribution of physicians have been long-
standing,19 predating public awareness of a medical malpractice issue.

Studies Linking Medical Malpractice to Physician and Hospital Supply:

Anecdotes and Evidence from Two Surveys of Physicians
There is considerable anecdotal evidence that physicians and hospitals
have reduced availability of medical care in response to higher medical
malpractice premiums. The anecdotes based on newspaper stories in box
3.1 illustrate this point. Not only do they add color; the anecdotes also
serve to personalize the problems. They say nothing about the effect of
a rise in premiums on availability of care for the population as a whole
Box 3.1
Anecdotal Evidence on the Impact of Higher Medical Malpractice Premiums and
Availability of Medical Malpractice Insurance on the Availability of Medical
Care

• Arizona The city of Bisbee, along the Mexican border, lost the mater-
nity ward at its local hospital when malpractice rate increases led to four
of the city’s six obstetricians to stop delivering babies.
• Florida The number of insurers offering medical malpractice coverage
dropped in half (from forty to twenty) over the past decade, pushing pre-
miums up and reducing the availability of coverage. Malpractice insurance
premiums in 2002 averaged $201,376 for ob/gyns, while the average was
$174,268 for general surgeons. The Orlando Regional Medical Center is
currently at risk of closing its trauma center due to the lack of neurosur-
geons willing to work the emergency room.
• Georgia A recent study of Georgia physicians projected that 2,800
doctors in the state (or about one in five) would stop providing higher-risk
procedures in order to reduce their liability exposure. One in three ob/gyns
said they would limit their services (including delivering babies), and 11
percent would stop working in emergency rooms. Four percent of the
state’s doctors reported that high malpractice premiums have led them to
retire early or leave the state. Overall, the study reported that malpractice
premiums increased between 11 percent and 30 percent in the state.
• Nevada It has been reported that dozens of doctors have stopped
practicing in the state due to the medical malpractice crisis. The decision
by St. Paul Companies to cease writing malpractice insurance left 60
percent of Las Vegas doctors seeking new insurers, and 10 percent of the
city’s doctors are expected to quit or relocate as a result. The crisis in
Nevada was made particularly clear when the state’s only Level 1 trauma
center closed for ten days in July 2002, during which time the hospital’s
CEO warned the public to “Drive home carefully.”
• New Jersey Medical liability premiums have been increasing 20 percent
to 25 percent annually, and the Medical Society of New Jersey estimates
that 3,000 physicians in the state are at risk of losing coverage due to
reduced coverage by insurers. Over a period of less than a year, three
insurers—the MIXX Group, PHICO, and St. Paul Companies—covering
55 percent of the state’s doctors stopped writing coverage for malpractice,
leaving doctors rushing to find new sources of insurance.
• Pennsylvania The state’s largest malpractice insurer, the PHICO Group,
has been placed in liquidation, and the MIXX Group and Princeton Insur-
ance have ceased writing new policies. Rising malpractice costs have
induced doctors to leave the state, retire early, or stop performing certain
procedures. Difficulty obtaining malpractice coverage caused Abington
Memorial Hospital outside Philadelphia to close its trauma center for
almost two weeks. Among doctors hit the hardest, according to Pennsyl-
vania Hospital, are radiologists specializing in mammography. The loss of
radiologists in the state has resulted in waiting periods for routine mam-
mograms of up to eight months.
64 Chapter 3

Box 3.1
(continued)

• Texas Doctors along the Rio Grande River have experienced significant
increases in malpractice premiums, with neurosurgeons paying up to
$120,000 a year and ob/gyns paying up to $100,000 for coverage. Numer-
ous surgeons, internists, and the only pediatric surgeon in El Paso left
the city. According to one physician, “The physicians along the Mexican
border have a lower percentage of patients who are privately insured, and
to have a line item like medical liability insurance go up 100 percent to
300 percent in a year’s time is a lot for some practices to swallow.”
• West Virginia Higher malpractice rates have contributed to about 5
percent of the state’s doctors either retiring early or leaving the state. The
Charleston Area Medical Center had to pay $2,000 daily in malpractice
premium subsidies in order to retain the doctors necessary to keep its
trauma center open. After the last emergency room neurosurgeon left
Wheeling, the local hospital had to transport trauma patients by helicopter
to other emergency rooms. The departure of St. Paul Companies from the
malpractice insurance market has forced two-thirds of the state’s doctors
to seek coverage from other sources.
• Washington Increased losses forced Washington Casualty Co., the
state’s largest provider of malpractice coverage to rural hospitals, into
receivership. The firm provided coverage to forty-six hospitals and twenty
community health clinics in the state, and covered 75 percent of the state’s
rural hospitals. PedMac, which provides health care services to the poor,
reported that its annual malpractice insurance costs increased by 150
percent, and the average cost for malpractice coverage for hospitals
increased 60 percent statewide. A survey by the state medical association
found that obstetricians have been hit hard, with 19 percent reporting that
they have already stopped practicing obstetrics and 8 percent saying they
plan to stop in the near future.

Source: Saxton, U.S. Congress, Joint Economic Committee (2003),

pp. 15–17.
 An Increased Threat of Lawsuits and Higher Premiums 65

or even for segments of the population. Many of the accounts speak in

terms of “up to.” Minimum, mean, and median values, however, are not
presented.
The brief examples in box 3.1 may be oversimplifications of reality.
For example, when Bisbee, Arizona, lost its maternity ward at the local
hospital, where were did residents of Bisbee go for obstetrical services?
Did the ward reopen later? Did the two obstetricians remaining in Bisbee
increase their capacity? Were new obstetricians recruited to the commu-
nity? If there was no entry and if residents of the community had no
nearby alternative, the situation would be far more serious than if the
example describes only a very temporary event.
Surveys of physicians in Pennsylvania and in Florida were conducted
in 200320 and 2004,21 respectively. Both surveys were conducted by mail,
with response rates of 65 percent and 40 percent. Seven percent of spe-
cialist respondents to the Pennsylvania survey said that they would defi-
nitely retire in the next two years because of liability costs, and another
32 percent said that they would be very likely or somewhat likely to
retire for this reason. Surgeons were more inclined to state that they
would retire early than were other specialists. Four percent said that they
would definitely relocate to another state for this reason in the next two
years, and another 29 percent said that they were very likely or some-
what likely to do so. Surgeons (general surgeons, neurosurgeons, and
orthopedic surgeons) were more likely to say that they were apt to relo-
cate in the future than were other specialists. Twelve percent of special-
ists had reduced or eliminated “high-risk aspects of their practices,”
which included delivering babies, performing back surgery, and avoiding
high-risk patients, such as the obese and high-risk pregnancies, because
of liability costs, and another 38 percent stated that they were very likely
or somewhat likely to do so.
The format of the questions about future retirement and practice loca-
tion intentions in the Florida survey differed from that of its Pennsylvania
counterpart. When asked about how long the physician respondents
intended to stay in their current practice, 11.2 percent said “less than
two years.” These responses included retirements, relocations to other
geographic areas, and relocations within the same area. Retirements
accounted for about two-fifths of the decisions regarding planned
66 Chapter 3

changes. Among those planning to leave their current practices within

two years, 61.6 percent said that “inability to find medical liability insur-
ance played a role” in their plans. An even larger share of respondents,
77.8 percent, said that “inability to pay for medical liability insurance
played a role.”
Much of the Florida survey focused on changes in availability of spe-
cific services attributable to the medical liability insurance market. The
most commonly eliminated services were nursing home coverage (42.1
percent of respondents), vaginal deliveries (29.1 percent), cesarean deliv-
eries (26.0 percent), emergency department coverage (22.8 percent), and
mental health services (21.2 percent). There were no statistically signifi-
cant differences between rural and suburban/urban physicians in these
assessments.
These types of survey findings can be influential in the legislative
decision-making process. Not only are physicians in a financial position
to contribute to political campaigns, but they are also on the front line
and, therefore, presumably are “in the know.” Nevertheless, such survey
findings are subject to several important limitations.
First, response rates to physician surveys tend to be low, especially to
mail questionnaires, as was the case with these two surveys, especially
the Florida survey. This raises the possibility that those physicians who
were most upset about recent events in the market for medical malprac-
tice insurance would be overrepresented among respondents.
Second, the surveys measured planned exits, but not planned entry. If
physicians stay in practice, they have to move somewhere. No state, to
the authors’ knowledge, has acknowledged that it has benefited from
receiving many expatriates from other states because of high medical
malpractice insurance costs in these other states. If, following a decrease
in supply, payments from insurers eventually rise to cover at least part
of the physician income initially lost due to the increase in premiums,
this will provide opportunities for physicians to locate to these states.
An immediate response to a substantial increase in premiums may be
to contemplate retirement or moving to another state. But after two years
during which premiums do not increase further, with increases in insurer
payments to partly offset the premium increase, and when malpractice
insurance again becomes widely available, many physicians are likely to
 An Increased Threat of Lawsuits and Higher Premiums 67

change their plans. Added to this are uncertainties associated with retire-
ment and the appreciable cost and uncertainty of moving to another state
and establishing practice there. For these latter reasons as well, many
physicians who planned to leave are likely to decide to stay put.
Also, moving to another state for reasons of medical malpractice is a
risky undertaking. Such states as West Virginia were “crisis states” for
the first time in the post-2000 crisis. A physician could have moved there
from Florida, only to find that West Virginia was just as bad or even
worse. Actual changes may differ from planned ones. In one study based
on data from the 1970s and 1980s,22 physicians who had previously
experienced high frequency of claims were less likely to actually change
their practice status (e.g., leave the state or retire).
Key informant interviews used to help design the mail questionnaire
for the Pennsylvania survey revealed that “Physician migration particu-
larly affected hospitals located near Pennsylvania’s borders, because
physicians could easily commute across state lines.”23 However, at about
the same time, the New Jersey Hospital Association reported impacts of
the medical malpractice crisis that threatened New Jerseyans’ access to
health care services. This included a group of seven New Jersey retinal
specialists who stopped providing care to premature infants, a commu-
nity that lost its only neurosurgeon to New York, and reductions in
access to obstetrical care.24 It might be advisable for the tort reform
advocates in both states to coordinate their public relations activities.
Third, the decreases in services may seem alarming, but no information
was provided on the actual volume of the reductions in services. For
example, for obstetrical deliveries, the physician may have performed
very little obstetrical care before quitting the performance of deliveries.
Cutbacks in service provision may sometimes be desirable. Low-volume
providers tend to provide lower-quality care, holding other factors con-
stant.25 Another concern has been that rates of cesarean sections are too
high. Some declines in these rates may be a welcome outcome.
Fourth, some physicians may view surveys of impacts of medical mal-
practice on their practices as political tools, and the responses may have
been strategic to elicit legislative changes favorable to physicians. Fortu-
nately, some researchers have studied the relationship between medical
malpractice and physician supply, using multivariate analysis. Medical
68 Chapter 3

malpractice is certainly not the only determinant of physician supply.

Using multivariate analysis, researchers can ascertain the contribution of
changes in medical malpractice variables to changes in physician supply,
while accounting for other determinants of physician supply.

Studies Linking Medical Malpractice to Physician Supply: Results of

Multivariate Analysis
Baicker and Chandra (2004) assess changes in physician supply by state
between 1992 and 2001, using several alternative measures of the threat
of medical malpractice, premiums, frequency of paid claims, severity of
paid claims, and a measure of the loading factor on medical malpractice
insurance,26 and holding several other potential determinants of physi-
cian supply constant in their analysis. They conclude that medical mal-
practice has no effect on physician location.
Hellinger and Encinosa (2003) study physician location patterns in
selected years from 1985 to 2000. Rather than use a measure of medical
malpractice cost or threat, the authors assessed the impact of dollar limits
on damages in medical malpractice cases as the key variable explaining
the geographic distribution of physicians. Among all statutory changes
involving tort liability of physicians, limits on damages have been shown
to be the most effective in terms of reducing medical malpractice pay-
ments and premiums.27 Hellinger and Encinosa conclude that (1) states
with caps on nonmonetary damages had about 12 percent more physi-
cians than states without caps and (2) states with relatively high caps
were less likely to experience an increase in numbers of physicians during
the observational period than were states with lower caps.
The Hellinger-Encinosa study differs from Baicker and Chandra’s in
several respects, including the definition of the physician supply, the
dependent variable, the other explanatory variables included in the anal-
ysis, and the measure of medical malpractice threat. Perhaps some other
factor, not included, such as a favorable political climate for physicians,
was responsible both for the passage of damage caps and for physician
entry into the state. Baicker and Chandra report no effect of medical
malpractice on supply, and another study, by Kessler et al. (2005),
described more fully in a later chapter, finds a much smaller effect (2.9
percent) than did Hellinger and Encinosa.
 An Increased Threat of Lawsuits and Higher Premiums 69

The most appropriate conclusion seems to be that there may be a link

between a state’s medical malpractice climate and its supply of physi-
cians. The effect is probably not large, especially when compared with
the difference in supply among geographic areas that persists, and rela-
tive to the growth in aggregate physician supply that occurred in the U.S.
during the latter part of the twentieth century.

Patient Access to Medical Care: Qualitative Evidence of Access

Problems
Like the quantitative evidence, qualitative evidence on the relationship
between the medical malpractice environment and physician location is
ambiguous. The U.S. General Accounting Office (GAO) conducted case
studies in five states (Florida, Mississippi, Nevada, Pennsylvania, and
West Virginia).28 Medical malpractice crises were reported in all of these
states during this period.29 The GAO reports some examples of reduced
access to services, such as lack of full orthopedic surgeon coverage in an
emergency room. However, it also finds that some reports of physicians
relocating to other states, retiring, or closing practices were either inac-
curate or involved relatively few physicians.30
The American Medical Association was asked to comment on a pre-
liminary report of GAO’s research. The point-counterpoint (as described
by the GAO) is interesting since it mirrors the general tenor of the debate
between groups advocating tort reform and evidence assembled by inde-
pendent organizations with no particular stake in the outcome.
The AMA had several criticisms of the GAO’s analysis of access. First,
the AMA criticized the limited scope of the study, in particular its cover-
ing only five states. The GAO responded that reports from other sources
had indicated that the five states were experiencing medical malpractice
crises. Second, the AMA criticized the GAO for failing to cover two
specialties in which problems with medical malpractice were most acute,
obstetrics-gynecology and emergency room medicine. These are indeed
specialties which have experienced problems with litigation from time to
time over decades. Yet they are not the only specialties to have been
adversely affected in this way.
Third, the AMA said that the GAO’s aggregate analysis obscured
access problems in particular specialties and localities. Among its
70 Chapter 3

responses to this criticism, the GAO noted that it had conducted some
specialty and state-specific analysis. It cited cases in which results from
its detailed investigation contradicted prior reports of access problems
that had been attributed to medical malpractice.
Fourth, the AMA commented that some of the GAO’s analysis was
not sufficiently recent. The GAO responded that its analysis of 2000–
2002 data included a period during which medical malpractice premiums
rose appreciably.

Patient Access to Medical Care: The Relationship between Medical

Malpractice and Patient Travel Time
To the extent that hospitals are closing units at their facilities, one would
expect to observe increases in patient travel time to the facilities that
remain open. Using hospital inpatient utilization data from Florida for
the years 1997, 2000, and 2003, Dranove and Gron (2005) assess utili-
zation of inpatient procedures that would normally be performed by
obstetricians and neurosurgeons, two high-risk specialties in a high-risk
state. From the data, they could determine the distance between the
patient’s place of residence and the hospital at which the procedure was
performed. The most notable finding on travel time for craniotomies is
an increase in travel time from home to hospital of from thirty-seven to
forty-two minutes between 2000 and 2003—before and during the
medical malpractice crisis in that state. Although the travel times were
more than twenty minutes longer for rural craniotomy patients than for
similar patients statewide, there was no differential change in travel times
for such patients between 2000 and 2003. For high-risk obstetrical
deliveries, travel time increased by less than half a minute, on average,
between 2000 and 2003.

Access to Care: Implications

Overall, the empirical evidence suggests that reports of physicians leaving
practice because of the professional liability crisis are overstated, as are
statements that the crisis is causing problems of patient access. Lack of
health insurance, which has not received nearly as much political support
from the advocates of tort reform as has tort reform, has been clearly
established as a barrier to care. “Negative defensive medicine” undoubt-
 An Increased Threat of Lawsuits and Higher Premiums 71

edly exists at some times and in some places. However, if the policy intent
is to improve access to care, a much broader strategy is warranted.

The Rising Cost of Medical Care: Is Medical Malpractice the Culprit?

Positive Defensive Medicine

Lack of Adequate Definition

Given that medical malpractice premiums are a small fraction of total
personal health care spending, rising premiums per se cannot be a major
driver of the increased cost of personal health services. Those who see
medical malpractice as the culprit point to the effects of the threat of
lawsuits on decisions of physicians and other providers. Providers prac-
tice “defensively” to avoid litigation.
The vast majority of providers carry medical malpractice insurance,
which tends to cover all payments to claimants as well as the legal expense
of defending such cases. Since medical malpractice insurance premiums
are not generally experience-rated,31 insured providers face very little
direct financial risk from medical malpractice litigation. However, there
is a substantial uninsured expense in terms of the value of time and the
emotional burden on parties accused of medical malpractice, as well as
possible loss of reputation. Thus, even insured physicians and other pro-
viders may have an incentive to practice defensively.
The lack of an adequate definition for defensive medicine has led to
much confusion; the vast majority of assertions have not been based on
a precise definition of defensive medicine, nor has quantification of the
extent of this practice been attempted. As explained above, to the extent
that the threat of medical malpractice litigation increases provision of
care for which marginal benefit exceeds marginal cost, then such litiga-
tion serves its desired purpose, and conversely. Thus, not all additional
services attributable to the threat are “defensive” or unproductive. Any
effort to measure the effect of defensive medicine should start with an
operational definition of the term.
The U.S. Office of Technology Assessment (OTA) defined defensive
medicine in the following way: “Defensive medicine occurs when doctors
order tests, procedures, or visits, or avoid high-risk patients or proce-
dures, primarily (but not necessarily solely) to reduce their exposure to
72 Chapter 3

malpractice liability. When physicians do extra tests or procedures

primarily to reduce malpractice liability, they are practicing positive
defensive medicine. When they avoid certain patients or procedures, they
are practicing negative defensive medicine.”32 The OTA definition, based
on a view shared by the vast majority of health professionals, starts from
a very different premise. Medical liability has little or nothing to do with
optimal care. Rather, the threat of being sued is a disruption, and any
changes in resource allocation attributable to the threat are wasteful.
In contrast to the OTA definition, the law and economics tradition
starts with the concept of optimal care, the level that would maximize
consumer well-being, given available scarce resources. Health profession-
als are called upon to serve as the patient’s agent in deciding on which
care is expected to yield a benefit in excess of cost. According to this
view, only care for which expected cost exceeds expected benefits is
excessive and defensive. There may be uncertainty about the anticipated
benefit, but the expectation is that the parties act on the best information
they have to achieve the objective.

Quantifying the Cost of Medical Malpractice: The Role of Positive

Defensive Medicine
Reynolds et al. (1987) estimate the net impact of medical malpractice on
the cost of physicians’ services in 1985. These costs include malpractice
insurance premiums, the costs of defending claims, and the costs of
practice changes made in response to increasing medical liability risk.
The authors used two methods. Their first approach used statistical
techniques to assess the relationship between medical malpractice
premiums and the price and volume of services provided, an approach
followed in most subsequent studies of defensive medicine.
Their second approach utilized a survey of physicians. The authors
questioned physicians about how they have responded to the threat of
medical liability. The physicians stated that they gave more detailed
documentation of patient encounters in medical records and took more
time to discuss care with patients. To the extent that good documenta-
tion leads to better care, this is something that patients are likely to value.
At least some patients are likely to want to discuss their diagnoses and
therapies with physicians. As some of the physician responses to the
 An Increased Threat of Lawsuits and Higher Premiums 73

threat of medical malpractice litigation are desirable, those desirable

responses should not contribute to the cost of medical liability unless
there is a separate calculation for the benefit as well.
Payments to claimants as reflected in premiums are not a social cost
of medical malpractice. Social costs reflect use of scarce resources. The
resources are used up and are not available for another purpose. Instead,
such payments represent transfer payments from defendants to claim-
ants. Such payments are a private cost to defendants and private revenue
to claimants.
For these reasons, Reynolds and his coauthors plausibly overestimate
the cost of medical malpractice. They report that in 1985 in the U.S.,
physicians paid $3.7 billion in premiums. The value of physicians’ time
in litigation was $100 million. The cost of defensive medicine in that
year was $11.7 billion, for a total cost of medical malpractice of $15.4
billion. This was relative to an expenditure of $82.8 billion on physi-
cians’ services. Thus, their conclusion is that premiums, defensive medi-
cine, and the total cost of professional liability were, respectively 4.5,
14.1, and 18.6 percent of spending on physicians’ services in that year.
Because the $11.7 billion figure includes additional resources associ-
ated with desirable responses by physicians (a benefit of the threat of
lawsuits, not a cost), and the $3.7 billion figure includes transfer pay-
ments to injury victims, the estimated total cost of professional liability is
probably too high. One component, the $100 million estimate for time in
litigation, is too low. Physicians (like the vast majority of defendants in
litigation) plausibly would place a value on time spent in litigation at far
more than their hourly wage. (We admit that we would!) And the number
of hours in litigation excludes all the time defendants are likely to spend
fretting about the case. But even if the $100 million figure were multiplied
by a factor of 5 or even 10, it seems unlikely that a more accurate estimate
of the total cost of professional liability would be anywhere near 18.6
percent of spending on physicians’ services in a year.

Quantifying the Cost of Medical Malpractice from Clinical Scenarios

Other approaches used more recently are clinical scenario surveys and
case studies describing the impact of malpractice liability concerns on
the use of specific medical technologies. In clinical scenario surveys,
74 Chapter 3

physicians were asked by the OTA to say how they would respond in
various situations and to rate the relative importance of various reasons
for the physicians’ stated choices.33 The list of reasons included the threat
of medical malpractice. A positive aspect of this method is that the sce-
narios represent clinically relevant situations and choices that physicians
often make in practice. There is a risk, however, that respondents might
tailor responses to help achieve a political outcome that they desire.
Based on a review of existing studies, the OTA reports no convincing
previous empirical evidence on defensive medicine at the time it con-
ducted its own research on the topic. The OTA administered scenarios
to physicians practicing in states with different liability climates. Based
on its findings, the OTA report concludes that defensive medicine raises
the use of diagnostic procedures by less than 8 percent, and the effect
varies substantially by clinical situation.

Quantifying the Cost of Medical Malpractice: Kessler and McClellan’s

Analysis of Positive Defensive Medicine in Cardiac Care
The most cited scholarly paper34 on the topic of defensive medicine and
one that is often used to support the case for tort reform in recent years
is by Kessler and McClellan (KM).35 KM used longitudinal data on all
elderly (aged sixty-five and over). Medicare beneficiaries who were hos-
pitalized for a new acute myocardial infarction (AMI) or a new episode
of ischemic heart disease (IHD) in 1984, 1987, and 1990. Like subse-
quent studies by Hellinger-Encinosa (2003) and Kessler, Sage, and Becker
(2005), KM’s measures of the medical malpractice threat are variables
reflecting tort reforms implemented in the state in which the beneficiary
was admitted for treatment. KM assessed the effect of the statutory
changes on total hospital Medicare payments made during the year fol-
lowing admission for the AMI or IHD. These data were used to measure
the effect of the statutory changes on intensity of treatment. KM also
studied the impact of the tort reforms on patient outcomes.
According to KM’s methodology, defensive medicine is reduced if the
reforms (1) reduced treatment intensity and (2) did not adversely affect
patient outcomes. Their outcome measures are mortality within one year
of admission for the index event (admission to a hospital for AMI or
IHD) and whether or not the patient experienced a subsequent AMI or
 An Increased Threat of Lawsuits and Higher Premiums 75

heart failure, measured by admission for either condition in the year

following the index event. Rather than analyze effects of tort reforms
individually, KM combine reforms into two variables: “direct” and
“indirect.” The direct reforms are caps on damage awards,36 abolition
of punitive damages,37 no mandatory prejudgment interest,38 and collat-
eral source rule reform.39 KM define indirect reforms as other reforms
that may affect pressure from tort on care provision, but affect awards
only indirectly. An example of an indirect reform is limitations on the
plaintiff’s attorney’s contingency fees, because it makes it more difficult
for injury victims to file medical malpractice claims. Indirect reforms
include limits on contingent fees,40 mandatory periodic payments,41 joint
and several liability reform,42 and the availability of a patient compensa-
tion fund.43 The study controlled for the effects of other factors’ hetero-
geneity by including explanatory variables for state and year.
KM find that in states adopting direct reforms (relative to states
without direct reforms), Medicare payments for hospital care during the
first year following the index admission declined by from 5 percent to 9
percent; for the indirect reforms, the decline was 1.8 percent. Mortality
was virtually the same in reform and nonreform states. Since the reforms
reduced cost of care, while not adversely affecting outcomes, KM con-
clude that liability reforms can reduce defensive medicine practices.
The KM study has several important strengths. It assesses both cost
and outcomes. It is national in scope and uses a large sample (200,000+
hospital admissions). KM acknowledge that elderly persons are less likely
to file medical malpractice claims than others.44 However, as KM argue,
results from analysis of a group that is not suit-prone provides a conser-
vative estimate of the extent of defensive medicine.
The study also has some weaknesses. The most important deficiency
is its possible lack of generalizability. While some proponents of tort
reform have used the paper to make generalizations about the effects of
tort reform, in fact, admissions for AMI and IHD (although an important
category of admissions) constitute only a small part of Medicare hospital
admissions. Implementation of tort reforms may reduce the cost of care
without adversely affecting mortality for AMI and IHD, but on average
it does not reduce the cost of care when elderly persons are admitted to
the hospital for other reasons.
76 Chapter 3

This deficiency is remedied at least in part by a later study by Kessler

and McClellan (1997), which uses national data to study effects of direct
and indirect reforms with a national survey of physician data from the
American Medical Association. That study focuses on effects of reforms
on claims frequency and premiums, finding some evidence that direct
reforms affected both, but with a lag.45 Further, they find that a variable
for direct and indirect reforms reduced referrals for consultation and
time spent with patients.46

Quantifying the Cost of Medical Malpractice: Studies of Positive

Defensive Medicine in Obstetrics
Several studies have assessed the effect of the threat of medical malprac-
tice lawsuits on the probability that a cesarean section rather than a
vaginal delivery would be performed. Birth injuries are a major allegation
in medical malpractice suits, far more frequent than suits involving
failure to perform a cardiac procedure.47
Sloan, Entman, Reilly, et al. (1997), using data from Florida, find no
effect of malpractice pressures on the method of obstetrical delivery
(cesarean versus vaginal delivery). An earlier study of the effect of the
threat of tort on the probability of cesarean section, using data from
New York State, also reports no effect.48
More recently, Dubay et al. (1999) have used national data from
birth certificates from 1990 thorough 1992 to assess the impact of
medical malpractice risk on cesarean rates and infant health. They
find that a $10,000 reduction in malpractice premiums would result
in a 1.4–2.4 percent decline in the cesarean section rate for some
mothers, but not for women with the highest socioeconomic status.
The authors conclude that a total cap on damages would reduce the
number of cesarean sections by 3 percent and total obstetrical charges
by 0.27 percent.

Bottom Line on Positive Defensive Medicine

In the end, this review of studies goes to show how hard it is to find
evidence that the threat of tort alters practice patterns in an important
way. In spite of its limitations, the KM study is well executed. And with
 An Increased Threat of Lawsuits and Higher Premiums 77

the limitations noted, it does provide empirical evidence for positive

defensive medicine in a much more rigorous fashion than the anecdotal
accounts and the studies based on surveys of physician opinion. On the
other hand, the evidence on cesarean sections is mixed. Placed in context,
even if direct reforms reduce the cost of care by 5–9 percent, real spend-
ing on personal health services increases by about this much in a two-
to-three-year period. These reforms are not a panacea for reducing the
growth in such expenditures.
The potential effect of the threat of liability may interact with
how the physician is paid. For example, any deterrent benefit of
liability may be greater under capitation, since providers have a greater
incentive to reduce services for financial reasons independent of the
threat of lawsuits. Perhaps facing the threat of lawsuits, physicians would
be more reluctant to cut services. On the other hand, under fee-for-
service, physicians may have a strong incentive to provide tests and
perform surgery. The threat of being sued may reinforce this incentive
to provide services. However, there is no reliable empirical evidence on
this issue.49

Good “Bedside Manner” as a Defense Against Lawsuits

Finally, one method of fending off claims has received some, but insuf-
ficient, attention. Having a good relationship with patients might be a
productive defense against being sued.
In 1992 Hickson et al. (1994) surveyed 963 women who had given
birth in Florida in 1987. The identities of the patients’ obstetricians were
obtained from the vital statistics files. Information on claims and loss
experience of these obstetricians was obtained from closed medical mal-
practice claims information filed with the Florida Department of Insur-
ance. Although mothers with adverse birth outcomes were oversampled,
none of the 963 women filed a medical malpractice claim, though twenty-
four had spoken with an attorney about this. Women seeing physicians
who had the greatest numbers of claims were more likely to complain
that they felt rushed, never received explanations for tests, and were
ignored. In response to the open-ended question “What part of your care
78 Chapter 3

were you least satisfied with?” women seeing obstetricians with a high
frequency of medical malpractice claims offered twice as many com-
plaints as those seeing obstetricians who had never been sued. Problems
with physician-patient communication were the most commonly offered
complaints.
An issue with any one survey conducted at one point in time in one
state and for one specialty is its generalizability. In a follow-up study,
Hickson et al. (2002) report an association between physicians’ patient
complaint records and their risk management records for the period
January 1992–March 1998. The risk management records, which came
from a large multispecialty group, included incidents with and without
associated expense. No attempt was made to ascertain whether the risk
management records reflected valid or invalid allegations of wrongdoing.
Data came from a state other than Florida, but neither the physician
group nor the state was specified.
A relatively small number of physicians generated a disproportionate
share of complaints. Physicians’ complaint frequency was positively cor-
related with risk management outcomes, ranging from opening medical
malpractice files to multiple lawsuits. The authors offer practical sugges-
tions about how management might intervene to improve patient satis-
faction and hence reduce the probability of lawsuits.
Other research is consistent with this finding. Levinson et al. (1997)
find that physicians without medical malpractice claims provide patients
with more orienting and facilitating comments, and use more humor,
than those with medical malpractice claims.
Being open and understanding with, and being accessible to, patients
may well be a good defense against lawsuits before adverse health out-
comes occur. After such outcomes occur, it is probably too late to insti-
tute many of these actions. However, in recent years, several commentators
have suggested that a lawsuit may be averted if the physician apologizes
for the outcome and, if appropriate, his or her role in causing it. Lawyers
representing the defense and insurers caution physicians that such apolo-
gies may compromise the defense if a lawsuit is filed in spite of the
apology. “I’m sorry” legislation has been proposed to protect physicians
who apologize. However, if enacted, the statutes are likely to be chal-
lenged. Given this uncertainty, we feel much more confident in recom-
 An Increased Threat of Lawsuits and Higher Premiums 79

mending the ex ante approach just described than apologizing after an

adverse outcome.

Implications
As with “negative defensive medicine,” some “positive defensive
medicine” undoubtedly exists. Tests are ordered and procedures
performed for which the marginal benefit falls short of marginal cost,
just to improve the defense in the event that one is sued for medical
malpractice. Statements that physicians order more tests because of the
threat of being sued seem persuasive, but estimates of the effect size are
never provided.50 Further, physicians have a financial incentive to order
tests and perform procedures. They own labs, CT scanners, and MRIs.
Thus, another way of viewing such statements is that physicians use
medical malpractice to justify practices that are in their financial
interest.
Compared with the substantial rise in the cost of medical care that has
occurred, with the exception of the Reynolds et al. (1987) estimate, even
high-end estimates of the cost of defensive medicine fall very short as
cost containment measures. Reform of medical malpractice, as argued in
later chapters, is justified, but not to reduce the amount of positive
defensive medicine.

Does the Threat of Lawsuits Deter Medical Injuries?

Medical errors remain frequent even with the threat of tort claims
(Institute of Medicine 2000), in spite of the fact that tort liability
has existed for years. The question posed here is whether or not the
threat of tort liability deters injuries and hence improves quality of
care.
The Harvard Medical Practice Study provides the best-known attempt
to determine whether the threat of tort indeed deters injuries. Using data
from forty-nine hospitals, the authors specified and estimated a two-
equation model (Weiler, Hiatt, Newhouse, et al. 1993). One equation
measured the effect of the threat of tort on the hospital’s injury rate, and
a second equation measured the relationship between the threat of tort
and characteristics of the area in which the hospital was located that
80 Chapter 3

might affect the threat. Most important, the second equation contained
exogenous variables that had no theoretical role in the first equation:
urbanization and population density. There is no apparent reason that
urbanization and population density would directly affect a hospital’s
injury rate. However, people may be more prone to sue in urban and
densely populated areas because social ties among residents are weaker
than in a small town or in a rural community. The threat of a malpractice
claim was measured as the fraction of negligent injuries (as determined
by the researchers’ assessments of medical records at the hospital) that
actually resulted in a medical malpractice claim. Dependent variables for
the main equation were the fraction of hospitalizations that resulted in
injuries, and the fraction of all injuries that were attributable to
negligence.
Weiler and colleagues (1993, p. 132) are not as cautious about inter-
preting these findings as they might have been. They state:
Our econometric analysis provides some evidence, though not scientific demon-
stration, that the higher the number of malpractice claims, the lower the number
of negligent injuries experienced by the patient population as a whole (patient
population at the hospital). That result emerged from our data even though the
host of constraints on the data set combined to reduce rather than enhance the
likelihood that such a causal connection would manifest itself. Indeed, some
might suggest that our point estimate of the impact of tort on injuries is probably
understated.

And they proceed to suggest a reason. Also, Danzon (2000) infers from
an insignificant coefficient in the Weiler et al. analysis and her own cal-
culations that the threat of medical malpractice claims provides a non-
trivial deterrent effect.51
Using the same database, Mello and Brennan (2002) report results of
their reanalysis of this issue. The interested reader is urged to consult
their forthright description of their investigation. They find inconsistent
results, and are unable to replicate the Weiler et al. findings; in the end,
they abandon their investigation.52

Bottom Line
There is no convincing empirical evidence to indicate that the threat of
a medical malpractice claim makes health care providers more careful.
 An Increased Threat of Lawsuits and Higher Premiums 81

This lack of empirical support represents a serious indictment of medical

malpractice as it currently exists and has existed.

Are Doctors Who Are Sued More Frequently Bad Doctors?

If it is not possible to show that the threat of a medical malpractice suit

deters medical injuries, is it a least possible to show that physicians with
adverse claims experience provide lower-quality care, as the term
“quality” is typically understood? The best indication is that an adverse
claims record in the past is a reasonable predictor of a future adverse
claims experience, but physicians with adverse claims experience are not
necessarily worse physicians.
Using data from a survey of 963 women who had delivered in
Florida in 1987, merged with data from medical records on some of the
same women,53 Entman et al. (1994) report results of subjective and
objective evaluations of the medical records of 484 of the same
women.
An independent expert panel of obstetricians and pediatricians identi-
fied 166 items that conformed to expected practice in 1987. Perinatal
nurses abstracted in detail all medical records according to a specific
study protocol. The study team assessed compliance with defined param-
eters in the areas of documentation, appropriate use of ancillary tests,
and events with marginal or inadequate care. Further, the reviewers
decided whether the error in care was irrelevant because the outcome
was good, or, if the outcome was adverse, whether the error was attrib-
utable to some underlying action or inaction.54 At the end of the review,
the team was asked, “Would you refer a member of your family to this
obstetrician?”
Although the reviewers found a number of cases in which care was
marginal or inadequate, in only 3 percent of cases did they conclude that
quality of care was substandard. There was no relationship between
prior claims experience of the obstetrician and the technical quality of
the practice in 1987 as indicated by the charts on the 484 women. Fur-
thermore, the reviewers concluded that attempting to identify physicians
at risk for future clinical errors by using data on prior malpractice claims,
82 Chapter 3

such as the National Practitioner Data Bank, may be misjudging the

likelihood that substandard care is provided by physicians with prior
claims.
In another study based on the survey of the 963 women and the linked
birth-death vital statistics file obtained for use in the same study, Sloan,
Whetten-Goldstein, Githens, et al. (1995) examine whether birth out-
comes were better in areas of Florida in which obstetricians were at
higher risk of a lawsuit than in areas in the risk was lower. They con-
sidered various types of birth outcomes: fetal death, low Apgar score,
death within five days of birth; infant death (death within a year of birth),
and death or permanent impairment of the child at five years of age. The
authors conclude that no systematic improvement in birth outcomes in
response to an increased threat of medical malpractice litigation could
be demonstrated.

Discussion and Conclusions

This chapter presents some bad news for both advocates and critics of
the current tort system as it applies to medical injuries. For the supporters
of the present system, the bad news is that evidence that the threat of
tort deters medical injuries is lacking and an adverse claims record is not
an indicator that a physician provides poor-quality care. While it is not
true that only good doctors are sued (myth 2), being sued is not a marker
of being a bad doctor either.
There is also bad news for the critics of medical malpractice as it exists
today. Statements asserting that tort law accounts for much of the
increase in spending on personal health care services in the U.S. (myth
5) are unfounded, as are statements that this is a major cause of barriers
to such services. At most, the threat of being sued accounts for a minor
part of the increase in spending. If negative defensive medicine exists,
and undoubtedly it does under specific circumstances, more important
access barriers are lack of health insurance and geographic inaccessibil-
ity. Some argue that the increased cost of care has led to increases in the
numbers of uninsured persons in the U.S. and that exit of providers has
led to geographic inaccessibility. If that is so, the magnitudes of the
changes are small.
 An Increased Threat of Lawsuits and Higher Premiums 83

Some important theoretical work has been conducted on the effects of

alternative liability regimes on physician decisions linking the threat of
a lawsuit to more careful behavior on the part of potential tort-feasors.55
Although there is some empirical support for the notion that the threat
of tort induces potential tort feasors to be more careful, thereby deterring
injuries in other contexts, the evidence in the context of medical mal-
practice provides little or no support. In principle, the system should be
reformed to provide a much greater incentive for deterrence; as a practi-
cal matter, the case would be even stronger if such reform could be sup-
ported by empirical evidence.
4
Governments’ Responses to Medical
Malpractice Crises—and Their Effects

Motivations for Reform

Since the first medical malpractice crisis, physicians’ organizations have

spearheaded what is widely known as “tort reform.”1 For convenience,
this terminology is used here. Ordinarily, one thinks of “reform” as an
improvement. Most of the tort reforms that have been implemented to
date are based on the underlying assumptions that the amount of litiga-
tion is excessive, as are payments to plaintiffs. Any statutory change that
results in fewer suits and lower payments is viewed as a “success” and,
conversely, as a failure when the statutory changes have no such effect.
The yardsticks against which statutory changes have been measured are
effects on medical malpractice premiums, the number of medical mal-
practice claims, and payments per paid claim. These end points reflect
the narrow agenda of premium and litigation control. By contrast, a
more general view of the social objective is minimization of the sum of
social costs: harm from injury to victims, costs of precautions, and costs
associated with the use of the legal system, which include legal costs
incurred by victims, defendants, and the government in resolving
disputes.
There is no consensus that tort reform has been motivated by
financial gain; indeed, there is an alternative interpretation. To many
physicians, medical malpractice claims are assaults on their sense
of capability and distort what they see as competent clinical practice.
Such claims seem counter to the concept of the physician as a profes-
sional in whom control over the meaning of competence in clinical
work is vested.2 Fielding (1990), taking a sociological perspective,
86 Chapter 4

argues that the medical malpractice crisis has been socially construed in
economic rather than autonomy terms because traditionally economics
has provided a more acceptable basis for labor unrest in the United
States.3
In the end, whether tort reform, especially tort reform of the conven-
tional variety, is about money, autonomy, or some other nonfinancial
goal, it seems unlikely that it is fundamentally about the goals espoused
by its proponents, which are said to benefit people in their roles as
patients and tax and premium payers. Of course, tort reform is by no
means unique in this regard. Irrespective of whether the plea benefits
society more generally, private stakeholders routinely seek help from the
government. In this sense, “What else is new?”

Past Policy Responses: First- and Second-Generation Reforms

Tort reforms have been classified in various ways. One method divides
them into (1) traditional or first-generation reforms and (2) second-
generation reforms. First-generation reforms are relatively minor modi-
fications to the existing tort liability system as it applies to medical
malpractice or, in some cases, to personal injuries more generally. The
ideas underlying second-generation reforms are more recent and involve
more fundamental change. Perhaps because they are more novel, but
more likely because they lack strong advocates outside of the community
of scholars, second-generation reforms have been enacted only rarely.
Scholars are often brought in as experts by activist groups advocating a
particular policy position, but as T. Baker (2005b) explains, scholars
lack both the financial resources and, more important, the incentives to
mount sustained efforts in terms of a particular policy position. For
scholars, the incentives are to publish and to teach; public advocacy is
often frowned upon.
This chapter addresses the states’ responses to the three crises that
have occurred since 1970. We begin by clarifying the distinction between
first-generation and second-generation reforms. Using this distinction,
the extent to which legislators are informed by research and the politics
surrounding adoption of tort reforms is explored. Next, we discuss
how state judiciaries affect the application and impact of tort reforms.
 Governments’ Responses to Medical Malpractice Crises 87

Following this, the remainder of the chapter discusses empirical evalua-

tions of the effects of reforms and the difficulties in evaluating them, as
well as their unintended consequences.
Some first- and second-generation reforms will be discussed in far
greater detail in later chapters, as will some major insurance reforms. An
overview is provided here.

First-Generation Reforms

First-generation reforms include tort and insurance reforms (box 4.1).

Tort Reforms
The direct intended effects of the tort reforms may be subdivided into
those that are aimed at size of recoveries, the number of suits filed,
plaintiffs’ difficulty of winning, and functioning and cost of the judicial
process. In practice, all reforms affecting size of recovery also affect the
number of suits. If there is less potential reward from suing, there will
be fewer suits. Among reforms aimed at recovery size, states have enacted
dollar limits on nonmonetary, total compensatory, or punitive damages;
permitted payments to be made to plaintiffs periodically rather than as
a lump sum; have eliminated joint and several liability; have modified
the common-law collateral source offset rule; and have restricted the use
of ad damnum clauses.
The rationale for periodic payments is to reduce windfalls to plaintiffs
and their families, which would occur, for example, in the event that the
injury victim died before the date anticipated at the time the verdict was
reached. Also, some have argued that juries may assign more modest
dollar amounts when the payment is made on a regular basis than as a
lump sum.4 A suggested approach to paying damages described in chapter
5 uses the periodic payments concept.
Under the traditional common-law rule, tort awards are not reduced
by the amount of compensation the injury victim receives from sources
other than the tort claim, such as from public and private insurance. The
rationale for disregarding compensation from collateral sources is that
defendants should not benefit from the fact that the injury victim obtained
first-party insurance coverage. Also, if amounts obtained from collateral
88 Chapter 4

Box 4.1
First-Generation Malpractice Reforms

Aimed at size of recoveries’ difficulty (or severity)

Caps on awards
Periodic payments of damages
Collateral source offset
Joint and several liability changes
Punitive damage limits
Ad damnum clauses restricted promises
Aimed at the number of suits
Pretrial screening panels
Arbitration
Statutes of limitations
Attorney fee controls
Certificate of merit
Costs awardable
Aimed at increasing plaintiffs’ difficulty (or costs) of winning
Expert witness requirements
Informed consent limits
Professional standard of care asserted
Res ipsa loquitur restrictions
Statute of frauds for medical guarantees
Aimed at functioning/cost of judicial process
Mediation
Notice of intent to sue
Precalendar conference required
Preferred scheduling
Insurance Reforms
Patient compensation funds
Joint underwriting associations
Limits on insurance cancellation
Mandates for liability coverage
Reporting requirements

Sources: Bovbjerg (1989); Kinney (1995).

 Governments’ Responses to Medical Malpractice Crises 89

sources are deducted from the award, potential tort-feasors will under-
estimate the loss to be incurred should they commit a tort. Offsetting
collateral sources would reduce the deterrent effect of tort liability. This
rule has been modified as part of tort reform either to allow juries to
consider payments from the other sources or to require them to do so
when they decide on compensatory damages.
States have changed rules for joint and several liability so that one
defendant would not be liable for the awards against other defendants.
For example, under the traditional rule, a hospital defendant may be
obligated to pay the amount that a defendant physician was required to
pay but did not have the funds to pay. States have also restricted the use
of ad damnum clauses.5 Sometimes plaintiffs’ attorneys have claimed
huge dollar amounts when the suit is filed. Such large claims have public-
ity value and may force the defendant to settle quickly.
Reforms aimed directly at suit frequency include panels and certificates
of merit to screen out nonmeritorious claims, arbitration as an alterna-
tive to tort, and eliminating the discovery rule. The discovery rule tolls,
delays or suspends the statute of limitations for injuries that could not
have been discovered with reasonable effort. The rationale is that many
medical injuries are not discoverable at the time of the occurrence—for
example, when a sponge is left in a patient by a surgeon.
To address the inability of some patients to file timely suits for their
injuries, many states have specified that the statute of limitations does not
begin until the person has at least a reasonable chance of discovering
the injury. However, this has resulted in an extremely long lag between
the injury occurrence and the date the claim was filed, which added to the
uncertainty of pricing medical malpractice insurance.6 Other reforms
directed at suit frequency involve implementing maximum percentages on
plaintiff attorneys’ contingent fees, and requiring that the loser in litiga-
tion pay the winning party’s legal expense (“costs awardable”).7
A third category of reforms intends to make it more difficult for plain-
tiffs to prevail in court. These statutory changes include setting minimum
qualifications for experts who testify at trial, setting and defining the
types of misleading information that would constitute fraud, and specify-
ing the requirements for consent forms. Some reform of consent forms
merely outlines the information that needs to be included on written
90 Chapter 4

consent forms, while other reforms define the amount of information

needed for appropriate disclosure of the risks and benefits of various
interventions.
Another reform that makes it more challenging for plaintiffs to prevail
is setting standards of care that constitute adequate medical care. Several
states have enacted legislation requiring plaintiffs to prove that the stan-
dard of care was not met. Previously, leaving a surgical implement in the
patient’s body constituted sufficient evidence that the surgeon was neg-
ligent (res ipsa loquitur, which is Latin for “the thing speaks for itself”).
Placing limits on the res ipsa loquitur doctrine made it more difficult
for plaintiffs to recover for injuries; proving an implement was left in
the patient’s body is easier than proving the occurrence constitutes
negligence.
Reforms aimed at improving the functioning and/or boosting the effi-
ciency of the judicial process include offering mediation, requiring a
notice of intent to sue, precalendar conferences, and preferred schedul-
ing. Overall, these are relatively minor reforms.
None of the above reforms has addressed the very high administrative
costs of the liability system. Also, first-generation reforms retain the same
traditional legal system, including the adversarial process, and do not
really constrain juries, features that physicians find problematic.8

Insurance Reforms
Insurance reforms aim to improve the availability of medical malpractice
insurance to health care providers. The two most important of these are
patient compensation funds (PCFs) and joint underwriting associations
(JUAs).9 PCFs are state-operated risk pools designed to supplement
primary coverage for large losses and thereby stabilize the insurance
market. JUAs are designed to be risk pools that provide coverage to
providers who are unable to obtain coverage from the private market.
The other insurance reforms listed in Box 4.1 are relatively minor.

Second-Generation Reforms

Second-generation reforms make more basic changes in dispute resolu-

tion (box 4.2).10
 Governments’ Responses to Medical Malpractice Crises 91

Box 4.2
Second-Generation Malpractice Reforms

Use of medical practice guidelines to set the standard of care

Scheduling of damages
Mandated use of alternative dispute resolution methods in lieu of tort
No-fault approaches
Limited no-fault early compensation
Neo-no-fault early compensation
Pure no-fault approaches
Enterprise liability
Private contract to implement malpractice reform

Source: Kinney (1995).

Medical practice guidelines are produced and disseminated by various

medical societies and other organizations to provide guidance to physi-
cians on best medical practice, but may also be used as a defense in liti-
gation. The rationale for using medical practice guidelines is to provide
clearer guidance to providers, and ultimately to courts, about the bound-
ary between nonnegligent and negligent care. Production and dissemina-
tion of guidelines have the potential of improving quality of care in
general, and thus have implications well beyond medical malpractice. In
principle, use of guidelines should reduce defensive medicine, since all
the physician has to do is follow them without feeling a need to exceed
them. Use of guidelines in the context of professional liability replace the
traditional approach, which has been to rely on physician expert testi-
mony about the usual standard of care in the jurisdiction in which the
case is being tried.
As of January 1, 1992, Maine began a five-year demonstration project
using standards of care that were defined by guidelines for four special-
ties (obstetrics-gynecology, radiology, anesthesiology, and emergency
medicine). Physicians choosing to participate in the demonstration project
could use the guidelines as a legal or affirmative defense in a medical
malpractice lawsuit. Most of the physicians in the four specialties signed
up to participate. In 1993, the U.S. General Accounting Office predicted
that this demonstration would be subject to litigation, which in fact it
92 Chapter 4

was.11 Medical malpractice insurers were concerned at the time that if

guidelines as an affirmative defense were found to be unconstitutional,
they might be held liable retrospectively.
Next, scheduling damages is an alternative to damage caps.12 Rather
than establish a ceiling on awards, as described in detail in chapter 5,
scheduling damages involves a more comprehensive methodology for
setting such limits.
Although first-generation reforms included use of alternative dispute
resolution (ADR)13 as a substitute for tort, some proposals make this
substitution a requirement. The main advantage to ADR, whether in the
form of arbitration14 or of mediation,15 is that the process tends to be
speedier than a trial. In addition, with binding arbitration, the decision
reached is comparable to a jury verdict and can be overturned only if
there is evidence of malfeasance in the process of reaching a particular
decision.
No-fault approaches are also designed to be substitutes for tort, provid-
ing compensation without regard to fault. The no-fault concept has been
widely used as a substitute for tort in auto liability and is used in workers’
compensation. However, medical no-fault has been implemented in only
two states, Florida and Virginia, and then on a very limited basis.16
An approach that had garnered some attention in the recent past is the
early offer approach, which was proposed by Jeffrey O’Connell over two
decades ago (O’Connell 1982). In this approach, a variant on no-fault,
there is no compensation for nonmonetary loss, and attorneys’ fees are
lower; there is a provision for offset of collateral sources; and the defen-
dant can offer a settlement for net economic loss. If the offer is accepted,
the defendant pays the plaintiff’s attorney fees, which would be on an
hourly rather than a contingency basis. If the offer is accepted, the case is
terminated. The defendant would presumably make an early offer if the
expected loss from doing so was less than the expected loss of proceeding
to trial. Thus, when a lawsuit is frivolous, an early offer would not be
made. Advocates for this type of proposal argue that it would result in
faster compensation, an upshot of which is lower legal expenses.17
Enterprise liability shifts liability from the individual health care pro-
vider to an enterprise, such as a hospital or an insurer.18 Advocates for
enterprise liability maintain that it would both improve deterrence and
 Governments’ Responses to Medical Malpractice Crises 93

reduce litigation cost.19 To the extent that errors are system errors rather
than mistakes on the part of individual providers, enterprises may be
more effective in quality assurance than individuals.
Finally, private contracts are offered as an alternative to the traditional
tort approach. The rationale for private contracts is that tort liability
determines compensation on the basis of standards of care that may
differ from the standards that patients might prefer. Private contracts
might set out specific circumstances in which providers might be liable,
schedule damages, and specify alternative resolution mechanisms when
disputes arise.20
The strength of private contracts is that they can reflect preferences of
individuals. Individuals with higher willingness to pay for safety pay
more for such care. However, individual choice opens the door to adverse
selection. That is, persons who are more likely to suffer an injury because
their health is more fragile may be more willing to pay for contracts
offering extra precautions.21
Opponents of private contracting as a substitute for tort liability point
out that the relationship between the patient and the provider is not one
of equal power. A hospitalized patient or even an outpatient may not be
well positioned to negotiate with a physician. Courts have overturned
contracts reached at the point of service for this reason. But this is not
when contracting would occur. Rather, contracts could be options
offered to persons at the time they enroll in a health plan. A lower stan-
dard of care or a less generous schedule of damages would command a
lower premium.

The Politics of Tort Reform

Adoption of First-Generation and Nonadoption of Second-Generation

Reforms
By 1990, all states had adopted some form of first-generation reforms.
Yet another medical malpractice crisis arose in the early 2000s. By con-
trast, implementation of second-generation reforms has been quite rare.
Litigation delayed implementation of Maine’s guideline demonstration.22
The two states with medical no-fault were slow to enroll injured
claimants.
94 Chapter 4

Kinney (1995) reports that physicians and other health care providers
oppose most second-generation reforms. Even though there is some
support for alternative dispute resolution mechanisms, to the extent that
they would streamline the dispute resolution process, she observes that
these groups have not supported, or even opposed, using medical practice
guidelines to establish the standard of care, especially if plaintiffs can
also use the same guidelines offensively. Support for medical no-fault is
also lacking.23 Although they are supported by legal and policy scholars,
there is no widespread political support for scheduling damages, enter-
prise liability, or for private contracting.24

Limited Input from the Public and from Researchers in the States’
Decision-Making Process
Tort reform is essentially a state issue.25 Numerous summaries of tort
reforms and empirical evidence of their effects exist. Empirical evidence
has been generally unimportant in determining legislative outcomes of
the tort reform debate.
Prior research indicates that state legislators on the floor rely on con-
sultation with committee members and experts rather than on their own
reading of the data.26 For example, Songer assesses the use of empirical
research in legislative committee and floor decision-making in South
Carolina, and his work is particularly pertinent because it deals with tort
reform.27 He finds that although the majority of House members on the
floor had not read the report prepared for their deliberation and relied
on opinions of committees for background to their votes, committee
members did review the data prior to making their recommendations.
Legislators are asked to vote on many issues and, especially given their
part-time status, may have difficulty becoming expert in a number of
substantive areas.
Often input from members of the public is lacking in legislative delib-
erations about tort reform. D’Arcy (1986) provides a case study of
legislative decisions about tort reform in the mid-1980s in Illinois. The
governor convened a task force consisting of eighteen members from the
business, academic, legal, and political communities. The Trial Lawyers’
Association, the Illinois Medical Society, and state malpractice insurers
were not represented on the task force, but they did make a substantial
 Governments’ Responses to Medical Malpractice Crises 95

contribution to the task force’s fact-finding process. Although they had

an opportunity to do so, no group representing the public interest
appeared before the task force.
In the end, many of the task force’s recommendations were enacted
into law. In general, the task force proved to be a filter for various pro-
posals, even though a few proposals supported by the medical society,
but not the task force, were enacted into law. The task force process
provided a mechanism for resolving disputes among stakeholders, but
with little or no input from the public more generally.

Conflicts between the Legislative and Judicial Branches of State

Government
While reformers have been suspicious of judges and juries, there has been
little inclination among legislators to allocate more decision-making
power to judges. Rather, as Rabin (1988, p. 40) notes, “The more attrac-
tive option has been a set of cutoff rules: caps on awards, a bar to
joint and several liability, or the elimination of collateral source
recovery . . . [but] the implications are rarely articulated in full.”

The Role of the State Judiciary in Challenges to Legislatively Enacted

Tort Reforms

Many of the legislatively enacted tort reforms have been challenged in

state courts. Decision-making in state courts differs from decision-making
in federal courts in several important ways. Much of this is due to state
judges being elected for fixed terms rather than appointed for life. Further,
amendments to state constitutions either by referenda or by popular ini-
tiatives are rather common.28 This is in contrast to the U.S. Constitution,
where the process of amending is arduous. Because of the difficulty in
amending, the U.S. Constitution has remained relatively stable over the
course of its existence.29
State constitutions vary considerably from the U.S. Constitution. On
average, they are three times the length, average around 120 amendments
per constitution, and, in the twentieth Century alone, eighteen states
ratified entirely new constitutions.30 Judicial selection, though not
as important as state constitutions, plays an important role in the
96 Chapter 4

development of tort law. The role of the judiciary is to interpret and

apply legislative and constitutional authority.31 This gives judges con-
siderable power to interpret laws and the constitution, as well as to fill
in any gaps in between with “judge-made” law, also known as common
law.32
Judges are often not accountable to the public for their decisions, even
though judges at the state level are mostly elected.33 This is due in part
to the fact that judicial decisions receive little publicity aside from the
rare case that catches the media’s attention. This creates concern about
judicial policymaking. In a common-law system, judicial policymaking
is inevitable and frequently necessary. Nevertheless, the judiciary’s power
to do so has been challenged, questioned, and criticized for at least a
century. As Thomas Jefferson stated in a letter to Judge Spencer Roane
in September 1819, “The constitution . . . is a mere thing of wax in the
hands of the judiciary, which they may twist and shape into any form
they please.”34
The state courts are split as to the constitutionality of tort reform
statutes. States look to each other for guidance, but the holdings vary
wildly from state to state and from reform to reform. A quick summary
of a small sample of cases demonstrates the variance across the country,
court, and reform. In 1996, Ohio passed a comprehensive tort reform
law. The law placed limits on noneconomic and punitive damages;
created statutes of repose35 for medical claims and offsets for collateral
benefits; and reformed the state’s joint and several liability legislation.36
Close to a year after the reform law came into effect, the Ohio Academy
of Trial Lawyers and the AFL-CIO filed suit, challenging its constitution-
ality.37 The statute was challenged on the grounds it violated provisions
in the state constitution, including separation of powers, the one-subject
rule, right to trial by jury, damages for wrongful death, right to remedy,
equal protection, prohibition of special privileges, and prohibition of
retroactive laws.38
The Ohio Supreme Court ruled the statute unconstitutional in toto.39
The court addressed the other concerns raised in the challenge individu-
ally, but did not include the analysis as part of the holding. One issue
the court addressed, an issue that appears in other courts as well, is that
 Governments’ Responses to Medical Malpractice Crises 97

of separation of powers.40 Because there was existing case law striking

down previous provisions of the bill, the court held the enactment of the
bill an intrusion on the exclusive authority of the judiciary.41 In their
opinion, the justices underscored the power struggle among the branches
of government, stating that the court and the legislature had “endeavored
to comport with the principle of separation of powers and respect the
integrity and independence of the other, that is, until now.” The court’s
authority even to hear such a case has been reexamined by the Ohio
Supreme Court since 2001.42
The Missouri Supreme Court challenged the statute of limitations law
in that state.43 The statute posed a two-year statute of limitations for
malpractice actions, but required minors under ten to bring their claim
by the age of twelve.44 A nineteen-year-old challenged this statute, alleg-
ing it violated the Missouri Constitution, which guarantees to every
citizen “that the courts of justice shall be open to every person, certain
remedy afforded for every injury to person. . . .”45 Ultimately, the court
struck down the law as violating a minor’s right to access the courts.
However, in a well-researched dissent, Judge Welliver lists twenty-three
different state cases in which courts have found similar laws constitu-
tional under equal protection grounds.46 Judge Welliver further suggests
that this was a case of inappropriate judicial policymaking, “reminiscent
of the Lochner era. . . .” when state and federal courts acted like super-
legislatures in striking down legislation not consistent with their own
views.47
Contrary to what many tort reform advocates would have you believe,
tort reform does not begin and end with the enactment of legislation.
State courts, guided by such legislation as well as their state constitution,
play an important role in the application and evolution of the law. As a
result, it is often difficult to determine the status and impact of state
laws. The uncertainty about whether a particular law will be found to
violate the state constitution can cause a lag in the effect on medical
malpractice insurance premiums. For this reason, insurers may be slow
to adjust premiums in response to a cap on damages.48 Some studies
assess changes from court decisions as well as changes enacted by legis-
latures. Others do just the latter.
98 Chapter 4

Empirical Evaluations of Effects of Tort and Insurance Reforms

Difficulties in Evaluating Reforms

A substantial number of studies have evaluated impacts of tort reforms.49
As indicated above, the outcome measure of interest is generally some
measure of claims activity, payments, or premiums. Even if these “end
points” are correct, several difficulties in evaluating state-level tort
reforms remain.
There are many differences in tort reform statutes. For example, in
states with damage caps, limits are set at very different levels. In addition,
tort reforms are often not enacted singly but in groups, making it more
difficult to isolate the effects of particular laws. In spite of these limita-
tions, evaluations of the impact of reforms are critically important. Leg-
islation is often enacted on the basis of specific assumptions about effects,
but the various participants in this “market”50 are clever, and can succeed
in circumventing the legislation’s original intent. Medical malpractice
review panels are a case in point. About half of the U.S. states have
enacted statutes authorizing the use of such panels.51
The exact details of the panels vary by state, but in general there are
three to seven members. One may be an attorney, a health care provider,
and/or a judge. Some panels include laypersons. In general, the panels
are more informal than a trial, but they do have the power to subpoena
witnesses and documents, including medical records, and to hear testi-
mony of expert witnesses, but not at the parties’ expense. Arizona is one
of these states. An evaluation of this program used insurance company
claim files data from before and after implementation of panels.52
Although the authors find that the panel system did not affect frequency
of claims and mean payments of paid claims, there was an increase in
the number of disputes in which the parties sought formal adjudication,
an increase in the cost of the adjudication process, and an increase in
the time within which disputes were resolved.
Functionally the panels simply added another step in the litigation
process. For plaintiffs, panels are a convenient method for gathering
information and determining the strength of the case for trial. They
provide a new, relatively low-cost, method for determining the outcome
of the suit. Other unintended consequences are possible (box 4.3). For
 Governments’ Responses to Medical Malpractice Crises 99

Box 4.3
Unintended Consequences of First-Generation Tort Reforms

Sloan and Bovbjerg (1989, pp. 16–17) listed examples of several possible
unintended consequences of first-generation tort reforms. Some of these
consequences have not been documented as having occurred empirically,
yet they remain conceptual possibilities and demonstrate the risks of enact-
ing statutory changes without in-depth analysis of their possible effects.

Example 1
The restriction formally imposed by the legislative change may not in fact
be binding.
Legislatures may not realize, for instance, that a statutory limitation on
recovery may be higher than the vast majority of recoveries actually paid
in recent years. Similarly, a contingent fee limit may be above most con-
tingent fees in the state. [The latter has occurred; see chapter 6.]

Example 2
Changes may be very slow to take effect, given the backlog of cases that
can be brought under prereform law.

Example 3
Pressing down the “balloon” at one place may cause a bulge in the balloon
elsewhere. For instance, limiting payment for nonmonetary loss may cause
juries to be more lenient in their assessments of monetary loss. For this
reason, some statutes seek to keep the juries from knowing the limits,
which might work before juries hear about them generally, and claimants’
lawyers adjust their tactics.

Example 4
A limitation on awards may be seen by various parties as targets or even
floors rather than as ceilings. A limit on payments for nonmonetary loss
(see chapter 5) could thus seem to become the right value to award in most
or even all cases, including the smaller ones. Claimants’ attorneys with
contingent fees below the statutory limit may use the legislation to justify
a fee increase; or the fee schedule may be used to facilitate price-fixing
among lawyers. Reforms also may change the patterns of filings. Reduc-
tions in the statute of limitations, for instance, may speed up filings or
pending cases to beat the new deadlines. Such increases in litigation,
however, are likely to be short-lived.

Example 5
Proposals that decrease litigation costs to society, and thus to claimants,
may encourage them to bring formerly unremunerative claims and increase
the total cost to the system.
To the extent that arbitration is a low-cost alternative to litigation, for
example, it may encourage claims. There is some evidence that this has
occurred.
100 Chapter 4

Box 4.3
(continued)

Example 6
Even if reforms generate savings in the long run, they may affect premiums
much later.
Because of uncertainties, in particular whether specific changes will survive
constitutional challenge and how they will be interpreted, medical mal-
practice insurers may be unwilling to reduce premiums until many years
of claims data reassure actuaries that reforms have had their intended
effects. There is some empirical evidence to support this, too.

example, a limit on payment of nonmonetary loss could be a target for

plaintiffs more often than a binding constraint, thus raising payments.
Tightening the statute of limitations could have the effect of shifting
claims filing to fit the tighter limit. This would, in effect, decrease time
to dispute resolution, which is advantageous to the plaintiff, who, unlike
the insurer, does not collect investment returns on reserves during the
time the case is pending.

Empirical Evidence on First-Generation Reforms on Claims Frequency,

Payments per Paid Claim, and Premiums
There is widespread interest in the effects of first-generation tort reforms,
and there are a number of excellent reviews of the literature.53 That the
impetus of the reforms was to reduce claims’ frequency and severity, and
premiums’ effectiveness, has been evaluated in these terms. The studies
have used regression analysis with reforms and other determinants of
frequency, severity, and premiums as explanatory variables (table 4.1).
Among the reforms, the evidence is clearest that caps on damages
reduce the frequency and severity of medical malpractice claims. The
more recent evidence54 indicates that caps reduce premiums as well. Even
though premiums reflect underlying losses, they also reflect prospective
investment returns and rate regulation.55
Evidence is uniformly negative on the effects of collateral source offsets
on claims severity and claims frequency, and there is no evidence that
they reduce premiums. Example 4 in Box 4.3 provides an explanation.
Among the reforms aimed most directly at claims frequency, the most
Table 4.1
Study Findings on First-Generation Tort Reforms Based on Regression Analysis

Significant Decrease in Claim Significant Decrease in Claim Significant Decrease in

Payments? Frequency? Premiums?

Reform Yes No Yes No Yes No

Governments’ Responses to Medical Malpractice Crises

Damages cap Sloan 2 Zuckerman Zuckerman Zuckerman Sloan 1
Danzon 1, 2 Thorpe
Collateral source Danzon 1, 2 Sloan 2 Danzon 2 Zuckerman Sloan 1
offset Zuckerman Zuckerman
Thorpe
Pretrial screening Danzon 1, 2 Danzon 1, 2
panels Sloan 2 Zuckerman
Zuckerman
Shorter statute of Danzon 1 Sloan 2 Danzon 2 Danzon 1 Zuckerman Sloan 1
limitations Zuckerman Zuckerman
Binding arbitration Danzon 2 Sloan 2 Danzon 1, 2 Sloan 1
Zuckerman Zuckerman Zuckerman
Contingent fee Danzon 1, 2 Zuckerman Sloan 1
limits Sloan 2 Zuckerman
Zuckerman Thorpe

Key: Danzon 1 = Danzon (1984); Danzon 2 = Danzon (1986); Sloan 1 = Sloan (1985); Sloan 2 = Sloan, Mergenhagen, and
Bovbjerg (1989); Thorpe = Thorpe (2004); Zuckerman = Zuckerman, Bovbjerg, and Sloan (1990). This table reproduces a table
in Studdert, Mello, and Brennan (2004), with the Thorpe reference added.

101
102 Chapter 4

analyzed are pretrial screening, binding arbitration, attorney fee limits,

and, to a lesser extent, costs awardable. The evidence on binding arbitra-
tion, although mixed, is more positive than for the other reforms. For
the latter, more often than not, measures of these policies have been
statistically insignificant in studies of claims frequency, severity, and
premiums. Effects of shorter statutes of limitation are mixed.
Effects of periodic payments of damages have not been assessed.56
There is no evidence on effects of joint and several liability changes or
on punitive damage limits, although finding much of an effect for the
latter seems unlikely since punitive damages have rarely been awarded
in medical malpractice cases, at least until very recently.
The other tort reforms listed in box 4.1 are much more limited in
purpose and scope, and their effects have not been evaluated statistically.
The insurance reforms seem to have been instrumental in restoring avail-
ability of coverage during the 1970s.57
Some studies have assessed dependent variables other than claim fre-
quency and severity, and premiums. Born, Viscusi, and Carleton (1998),
Barker (1992), and Thorpe (2004) find that damage caps improve insurer
profitability. However, Viscusi, Zeckhauser, Born, et al. (1993) do not
find a statistically significant relationship between damage caps and
profitability.58 That damage caps would increase profitability is plausible,
since insurers have been reluctant to reduce premiums immediately fol-
lowing enactment of reforms, in particular since the reforms are subject
to constitutional challenges.59 Thus, it would not at all be surprising that
a reform that decreases losses, such as damage caps, improves insurer
profitability.
Using data on jury verdicts in California that were subject to the state’s
$250,000 limit on payments for nonmonetary loss, Studdert, Yang, and
Mello (2004) document that the limit resulted in much greater (seven
times greater) absolute dollar reductions in awards for persons incurring
a grave injury than a minor injury. For this reason, the authors conclude
that such limits are regressive.

Effects of Damage Caps on Physician Supply

The first studies on the effects of damage caps on physician supply found
mixed results. Klick and Stratmann (2003) report that states that adopted
 Governments’ Responses to Medical Malpractice Crises 103

caps experienced a 3 percent increase in physician supply relative to those

without caps, holding other factors constant. By contrast, Matsa (2007)
fails to find an effect. More recently, two studies have found that caps
increase physician supply.
Encinosa and Hellinger (2005) find that damage caps increased the
supply of physicians over the period 1985–2000.60 The caps had a larger
impact on physician supply in rural counties than on supply in all areas.
Caps on nonmonetary loss set at or below $250,000 had more of an effect
on the supply of surgeons and obstetricians-gynecologists than caps set
above this dollar level, presumably because they are more likely to have
been finding in the former case. While the effects of damage caps on physi-
cian supply are statistically significant at conventional levels, the magni-
tude of effect is to raise physician supply by about 2 to 3 percent, on
average, and by up to between 5 and 6 percent for $250,000 caps in rural
counties. Thus, while the effects are statistically significant, they are insuf-
ficiently large to improve patient access to care in important ways.
Kessler, Sage, and Becker (2005) use the methodology developed by
Kessler and McClellan (1996) to assess the effects of “direct” and “indi-
rect” reforms on physician supply. Recall that the direct reforms include
caps on damage awards, abolition of punitive damages, no mandatory
prejudgment interest, and collateral source offset reform. The indirect
reforms include caps on contingent fees, mandatory periodic payments,
joint and several liability reform, patient compensation fund, and statute
of limitations reform. The authors measure physician supply relative to
state population of physicians, physicians with less than twenty years’
experience, physicians with greater than twenty years’ experience, and
physicians in nongroup practice. There was a separate analysis of physi-
cians in emergency medicine, anesthesiology, radiology, and general
surgery. The rationale for considering doctors separately by years of
experience is that more experienced or older doctors may be more likely
to retire in response to an increased threat of medical malpractice and/or
higher premiums. The hassle factor may be greater for nongroup physi-
cians than physicians practicing in large groups.
Controlling for fixed differences (differences that do not vary over
time) among states, market factors and various political characteristics
of the state, and implementation of the direct reforms (which include
104 Chapter 4

caps), physician supply increased by 2 to 3 percent, on average. The

effects were greater more than three years after adopting than before
this. Direct reforms had a greater effect on retirements and deterring
entry into the state than on the propensity of physicians to move between
states.
The two studies provide empirical support for the assertions that
adopting such reforms as damage caps can increase physician supply over
and above what it would be otherwise. Yet the effects, on average, are
not large, and this policy should be compared with alternative methods
for increasing physician supply. The implication is that damage caps
alone are unlikely to be a panacea for communities with few or no physi-
cians. And given the effect sizes, it is unlikely that the threat of lawsuits
has led physicians to flee specific states.

Empirical Evidence on Second-Generation Reforms

Second-generation reforms seek more fundamental change, yet most of
the reforms have not been implemented. Among those that have, the
most empirical evidence exists on no-fault as a substitute for tort. When
no-fault has been applied to the medical field, it has been applied in a
very limited way.61 For what it covers, no-fault has greatly reduced the
overhead of administration below what it is in tort. Yet there is no indi-
cation that no-fault has substituted for tort. Tort claims frequency in
areas covered by no-fault remain high.
The Institute of Medicine (2002) and the Governor’s Select Task Force
on Healthcare Professional Liability (2003) in Florida have recommended
pilot projects of second-generation or system reforms.62 In contrast, the
overwhelming majority of states saw no reason for study prior to wide-
spread implementation of first-generation or conventional tort reforms.
And, when evaluated, most of these reforms have shown no effect, even
on the goals they were implemented to achieve.

Conclusions and Implications

After three decades of experience with state tort reform and evaluations
covering a period almost as long, the key finding is that only damage
caps have consistently affected various outcomes of interest, including
 Governments’ Responses to Medical Malpractice Crises 105

claim frequency and severity, medical malpractice premiums, and physi-

cian supply. Profitability of insurers has also increased in states with
caps, at least in the short run. However, Studdert, Yang, and Mello
(2004) conclude from their empirical analysis that California’s single
$250,000 cap is unfair in the sense that it imposes a much greater reduc-
tion in payments to plaintiffs with grave injuries than to those with minor
injuries. The authors recommend scheduled damages or maximum pay-
ments within each severity category.
Such evidence should be sufficient for private stakeholders to argue
for caps, at both the state and federal levels. However, except for the
modest impact on physician supply, caps offer nothing for patients,
health insurance premium payers, and taxpayers.
One is left with an unsatisfying feeling that there must be a better way
than caps. Perhaps some second-generation reforms can be recast in ways
that will have more popular appeal while retaining their ability to im-
prove system performance. The following chapters address such reforms,
except for the chapter on contingent fees, which argues that limiting such
fees is a step in the wrong direction.
In discussing his finding that only caps have statistically significant
effects, Thorpe (2004) states:
At issue is whether we should have short-term, stop-gap solutions to slow the
growth in premiums or use the recent experience to more fundamentally evaluate
and perhaps reform the liability system. The recent spike in medical malpractice
insurance premiums allows us an opportunity to reexamine whether the tort
system is achieving its goals. If it isn’t, what changes in the system would improve
the dual goals of deterrence and compensation? The results suggest that capping
awards may improve the profitability of malpractice carriers and reduce premi-
ums. Whether this is socially desirable or improves the goals of deterrence and
compensation remains an open question.”63

Finally, the discussion of second-generation reforms and the goals of

tort in general are part of a larger discussion that the United States and
perhaps some other industrialized societies should have as well. Issues
of social justice raised by a broad taxpayer-supported no-fault program
for iatrogenic injuries are pertinent to several major private and social
insurance programs as well. While addressed in response to a crisis in
medical malpractice, the efficiency and equity issues of no-fault are
common to these other programs as well.64
5
Ceilings on Nonmonetary and Total Losses

Compensation for Personal Injuries: The Alphabet Soup

As of 2005, more than half of the U.S. states had some form of limit on
awards, mostly on nonmonetary awards, but some on total awards.1 The
attraction of such limits is that they do reduce mean payment per claim
(claim “severity”) by as much as 40 percent.2 Some studies show reduc-
tions in premiums as well, but by a smaller amount than claim severity,3
while others show no effect on premiums.4 Yet even in states with limits
on damages, premiums have risen appreciably in absolute terms.5 Findings
from research on limits on damages and other statutory changes in medical
malpractice are similar to those for other types of tort claims, such as
automobile liability.6 One reason for the smaller effect on premiums is
that laws limiting awards have been subject to constitutional challenges
in the states, which have resulted in the laws being overturned in a minor-
ity of instances.7 Insurers have been cautious about granting premium
reductions, given the uncertainties of outcomes of legal challenges.8
Given that these laws have been more effective than any other type of
tort reform in reducing losses from medical malpractice claims, they have
substantial political support among health care providers, medical mal-
practice insurers, and some attorneys who represent defendants in
medical malpractice litigation. The advocates for limits on awards fre-
quently cite the success of California’s 1975 medical malpractice law.9
The major criterion for success is the relatively low growth in medical
malpractice premiums since the 1975 law was enacted.10 The most
notable feature of this law was a ceiling on awards for nonmonetary loss
(a cap on pain and suffering.)
108 Chapter 5

Awards are meant to make the injury victim whole. This is the ratio-
nale for compensatory damages. Compensatory damages is a general
category which includes payment for monetary losses, such as from
medical care, lost earnings, and other services that are directly attribut-
able to the injury. Payment for nonmonetary loss is also part of com-
pensatory damages; however, nonmonetary loss can go under several
headings, depending on the circumstances of the case, such as payment
for pain and suffering, emotional distress, loss of consortium, loss of
enjoyment of or a chance at life, and so on. The object of compensatory
damages is to make an individual as well off as before the injury occurred.
When there is permanent damage to health, this cannot be accomplished
by restoring the person to his or her original health. However, assuming
that people get utility out of money as well as health, money is used to
compensate for the loss in health.
The rationale for payment for nonmonetary loss is that not all of the
loss is monetary. To exclude payment for nonmonetary loss would
reduce the potential deterrent effects of tort.11 Injuries can be painful.
They can result in loss of a lifelong partner. They can limit a person’s
opportunities for enjoyment of life, such as participation in nonprofes-
sional sports.
However, theory is one thing; in practice, full compensation may be
infeasible when the injury is seriously disabling with the result that loss
is irreplaceable.12 Another concern about payment for nonmonetary loss
is much more widespread, namely, that there is no objective yardstick
for ascertaining the extent of the nonmonetary loss even if compensation
could, in principle, be made in full.
Many injuries are temporary. That is, the person was injured and after
some time, the injury is self-correcting or, frequently, there is a medical
or surgical intervention to correct the injury. Then payment is made to
compensate for the period during which health was impaired.
Many injuries are permanent. In such cases, accuracy in payment for
future monetary losses is difficult since computing such loss involves a
prediction, which by its nature involves uncertainty, such as the injury
victim’s longevity; the level of care, which will be indicated at various
points of time; the prices of such care; and the future ability of the indi-
vidual to engage in remunerative activities. There is uncertainty about
 Ceilings on Nonmonetary and Total Losses 109

future rates of general inflation, and thus which discount rate to use.
Computing future nonmonetary loss also involves many assumptions.
In several European countries, the process of computing future loss is
simplified by having tables giving values of loss depending on such
factors as the injury victim’s age and degree of impairment. There is,
however, a trade-off between adherence to table values, which offers
simplicity and precision, and the facts of the particular case.13
Punitive damages are the final component of a medical malpractice
award. They are intended to punish the defendant for wrongful conduct
and to deter others from engaging in such behavior. They are not based
on the severity of the plaintiff’s injury, but rather on the culpability of
the defendant’s conduct and on an assessment of the amount required
to get the defendant to notice the award. Thus, a large corporation may
be assessed a larger punitive damage than an individual, such as a health
professional, for the same injury. In contrast to product liability, punitive
damages are rarely awarded in medical malpractice cases.14 Studdert and
Brennan (2000) report that punitive damages are awarded in less than
1.5 percent of medical malpractice verdicts. Nevertheless, many states
have enacted laws to limit such damages in medical malpractice as well
as other kinds of personal injury cases.15
Given that limits on awards curb losses paid by defendants and, to a
lesser and more uncertain extent, medical malpractice premiums, they
make good private policy. As one commentator observes, “Caps are a
political success. First, they are popular and widely enacted. Second, caps
in general work as intended.”16 For this reason, the American Medical
Association and various other organizations representing health care
providers have placed enacting caps at both federal and state levels at
the top of their political agendas. There is some question, however,
whether or not, at least in their present form, caps make good social or
public policy.

Is There a Compelling Rationale for Paying for Nonmonetary Loss?

The Controversies
However controversial ascertaining monetary loss might be, payment for
nonmonetary loss is all the more so. Not only has nonmonetary loss been
110 Chapter 5

limited by many more states than has total loss, but certain proposals
for reform, such as medical no-fault,17 either severely restrict payment
for such loss or eliminate it entirely.
Paying for nonmonetary loss is controversial for several reasons. Prob-
ably foremost, there is no objective standard by which to measure an
injury victim’s physical pain or mental anguish. Such concepts as “pain
and suffering,” “loss of consortium,” and “loss of enjoyment of life,”
all elements of nonmonetary loss, are somehow unreal and cannot be
quantified. That these losses are unreal is clearly contradicted by the fact
that a substantial number of individuals seek medical care for these sup-
posedly “unreal” conditions. Also, people frequently purchase prescrip-
tion and over-the-counter drugs to reduce pain and discomfort. That
these conditions cannot be quantified is further contradicted by the fact
they have been quantified, and the results of such quantification have
been published in refereed medical journals.18 Specialists in the field of
decision analysis have developed techniques for quantifying nonmone-
tary loss and are constantly refining them.
In addition to the difficulties in quantifying, some scholars have noted
that there is no voluntary market for insurance for nonmonetary losses.
The argument is that if there is no voluntary market for such loss, there
should be no payment for such loss under tort.19
In a comprehensive review of the medical malpractice literature aimed
at an audience of economists, Danzon (2000, p. 1373) argues:
that consumers do not voluntarily buy coverage for non-monetary loss in any
other private or social insurance program suggests that such coverage may not
be worth its cost. The lack of a voluntary market for insurance for non-monetary
losses may reflect severe ex post moral hazard of exaggeration of such losses,
which cannot be objectively measured. Assuming that this moral hazard of loss
exaggeration is at least as severe in the tort system, the evidence from private
choices supports the case for limits on compensation for nonmonetary loss
through the tort system.

This quotation contains three statements. The first is the observation

that people do not purchase insurance for nonmonetary loss. This state-
ment is clearly valid. The second is that such insurance may not be pur-
chased because people would have a tendency to overstate their pain,
which cannot be directly observed by others. This form of moral hazard
may be a reason that a voluntary market for insurance for nonmonetary
 Ceilings on Nonmonetary and Total Losses 111

loss does not exist. This point also is not controversial. The third state-
ment, that exaggeration of nonmonetary loss means that there should be
limits on compensation for such loss from tort, is very controversial and
likely to be wrong.20 The legal process of ascertaining nonmonetary loss
can be expensive. But the adversarial process of a jury trial permits
examination and cross-examination of assertions of loss.
Croley and Hanson (1995, pp. 1842–1844) make a further argument
for compensation for nonmonetary loss. Even if there is no private
market for nonmonetary loss insurance, the Consumers Union, which
purports to represent the interests of consumers, has consistently sup-
ported compensation for such loss.
Analytical techniques for quantifying such phenomena as pain and
suffering have been applied in research contexts, but they are otherwise
not in widespread use. Individuals’ willingness to pay to avoid specific
injuries, and pain and suffering from these injuries, is elicited from
surveys in which respondents have nothing to gain by suggesting that
they place a high value on the underlying loss. The goal of this research
is to determine the average willingness of survey respondents, as a group,
to pay. By contrast, in the context of insurance, if the payment were
made to depend on the injury victim’s depiction of his or her own pain
and suffering, surely there would be an incentive to exaggerate. In litiga-
tion, experts for plaintiffs have an incentive to compute the highest
possible values of damages incurred. On the other hand, experts for the
defense, in criticizing the estimates of the plaintiffs’ experts, have the
opposite incentive. Every defendant has a right to have these counterar-
guments presented at trial on his or her behalf. Therefore, it is not at all
clear that outlandish estimates presented on behalf of plaintiffs would
prevail.
The lack of a practical, objective standard which can be applied to
individual cases is likely to be the major reason that there is no private
market for insuring against nonmonetary loss. It is easy to understand
why an insurer would not offer health or disability insurance for non-
monetary loss. Any insurance company offering this kind of coverage in
addition to insurance for financial loss would potentially fall prey to
“adverse selection.” That is, people who are particularly pain-prone
would demand such coverage disproportionately. And, without a
112 Chapter 5

benchmark, insurers would be at the mercy of the injury victim’s asser-

tion of the extent of such loss, and such claimants would indeed have
an incentive to exaggerate the loss. As explained in greater detail below,
lack of an objective benchmark in the United States is also a problem
for juries’ estimation of nonmonetary loss.
An underlying presumption is that the law should provide the level of
damages, and only those damages, which a fully informed consumer, in
advance of the injury, would choose to be paid if injured. This makes
the decision of the amount of nonmonetary damages to pay depend on
what a sovereign consumer’s preferences are.21 People are willing to pay
for additional health and safety, for example, in their cars and to forgo
higher wages for greater safety in employment.22 The willingness to pay
for air bags and to forgo wages need not reflect only monetary loss
associated with an injury. One reason people want air bags is that motor
vehicle crashes hurt, not just because they may lead to wage loss and
medical expense.
There are more practical reasons for paying for nonmonetary loss.
Plaintiffs must pay a considerable part of the award—generally 33 to 40
percent—to their lawyers.23 These fees can be paid at least in part out
of the compensation for nonmonetary loss.24 Although seemingly plau-
sible at first glance, this argument is a weak rationale for paying non-
monetary losses. Taken to the extreme, if contingent fees were 67 percent
of compensation, payment for nonmonetary loss would even be higher
than it is today.25 Plaintiffs’ attorneys and consumer advocates have
argued that caps on damages, with zero damages being a special case of
a cap, have made it difficult for some injury victims to obtain legal rep-
resentation, even when their cases are valid, because of the limited
payouts.26 Some analysts, however, have correctly noted that the extent
to which caps have created a barrier to legal representation is
unknown.27
A more convincing argument can be made on equity grounds. For
example, in a well-publicized case from the authors’ own university,
Duke University, a Mexican teenager died after a second heart-lung
transplant because, for the first transplanted organs, the blood types of
the organ donor and recipient did not match.28 Future monetary loss was
relatively low since the woman had not acquired skills that would have
 Ceilings on Nonmonetary and Total Losses 113

commanded high compensation in the labor market. Yet clearly the

family was indeed made worse off by the pain and the suffering of the
experience and the subsequent loss of a daughter. In this type of case,
paying the family more than if their daughter were a successful business
executive would probably violate social norms of equity. Many citizens
would plausibly feel that an executive’s family should not receive con-
siderably more compensation than this low-income family, particularly
given the nature of the error.

Legal Challenges to Damage Caps

Several constitutional theories have been used in legal challenges to
damage caps in litigation at the state level.29 Limits have been challenged
using the open-courts guarantee included in most state constitutions. An
example of an argument employed in these challenges is that a dollar
limit denies catastrophically injured patients the right to collect their full
damages, and that the law creating caps contains no offsetting provision
that would be of benefit to plaintiffs. A second theory is that limits
violate the right to trial by jury, the premise in one state being that juries
must be able to determine the appropriate level of damages, and limits
are inconsistent with this right. Third, these laws have been subjected to
claims of lack of equal protection. Laws that create special categories or
classifications, such as distinguishing between medical malpractice and
other personal injury victims, may be subject to this type of legal
challenge.30
Due process challenges have been brought against nonmonetary
damages caps. In one variant, the argument is that the standard damage
cap deprives plaintiffs of the full amount of compensation without giving
them an opportunity to present evidence as to why full compensation is
appropriate. The separation of powers theory argues that many state
constitutions vest judicial powers exclusively in the judicial system. Thus,
in passing limits on damages, the legislature infringes on a judicial
prerogative.
More often than not, limits on nonmonetary losses (comprising the
majority of legal challenges) and on total losses have survived constitu-
tional challenges, particularly the challenges of the former. Rather than
rely on theoretical arguments or arguments about economic justice, a
114 Chapter 5

major and successful defense of such limits is a very practical one,

namely, that they have the effect of stabilizing medical malpractice insur-
ance premiums. In any case, payment for nonmonetary loss is most likely
here to stay.
The public policy issue is less whether or not to pay it, and more
about the process involved in setting damages in particular cases.
Seen from this perspective, there is clearly considerable room for
improvement.
In malpractice trials, large awards command a lot of public attention,
but they often are not actually paid in full, given subsequent reductions
of awards on appeal.31 The complaints voiced in the public arena,
however, tend to be about the magnitude of total awards, not specifically
about the nonmonetary component.
Even for jury awards, the allocation between awards for monetary and
nonmonetary losses is unknown. In many states, jurors are not instructed
to divide the award in this way. In some states, for example, where there
is a cap on nonmonetary loss, jurors are instructed to make such an
allocation for purposes of complying with the statutory cap, but data
have not been collected systematically, at least until recently.32 It is often
difficult, and generally impossible, to disentangle payments for each
purpose from extant databases.33 It is often possible to know the plain-
tiff’s initial demand for payment, but not to separate payment for mon-
etary from payment for nonmonetary loss from payments made in
settlements.
Few medical malpractice claims reach the trial stage and relatively few
verdicts are decided in favor of plaintiffs.34 The vast majority of medical
malpractice claims are either dropped by plaintiffs or settled. Then there
is no compensation at all.

Imperfections in the Current System of Ascertaining Payment for

Nonmonetary Loss

The relevant public policy issue is much less whether or not to compen-
sate for nonmonetary loss, and more about the process involved in
determining damages in particular cases. Seen from this perspective,
there is clearly considerable room for improvement. Whether viewed
 Ceilings on Nonmonetary and Total Losses 115

at close range or at a distance, the current methods for determining

nonmonetary loss leave much to be desired.
Even though the objective of compensatory damages is to make the
injury victim “whole,” the law in the U.S. gives very little guidance as
to how damage awards are to be calculated. In much the same way, the
process whereby such losses are computed by juries is not well under-
stood.35 For financial losses, there are prices and units of loss. For pain
and suffering, by contrast, there are no direct measures. Values must be
imputed, but the law gives no guidelines for valuation. Whereas prece-
dent influences legal decisions, there is no role for precedent in comput-
ing compensatory loss. Each decision is independent and presumably
highly individualized to fit the particular circumstances of the case.
Not surprisingly, variability in compensatory damages is the conse-
quence—in large part by design, since circumstances of individual cases
are thought to be highly variable. Not only is there variability, but there
is empirical evidence that patterns of awards reflect the income, racial,
and ethnic composition of the area in which cases are tried, and hence
the composition of the jury,36 as well as the type of case being tried—
factors that should have no bearing on the individual plaintiff’s loss.
Awards for the same injury appear to be higher in medical malpractice
than in automobile liability cases,37 suggesting that juries take into
account “deep pockets” of defendants in medical malpractice cases. An
alternative interpretation is that auto injury victims are even more under-
compensated than those with medical injuries.38
High payment variability—especially at the upper end of award sizes—
coupled with the prospect of even higher variability over time, may be
expected to have several untoward consequences. First, high awards may
be precedent-setting, particularly in determining future settlement
amounts.39 Given unpredictability of the outcome of the claim, health
care providers are unable to predict how their behavior will be judged
if that behavior causes harm, which reduces any deterrent effect the
threat of lawsuits might otherwise have.40
Second, once harm has occurred and a claim has been filed, the parties
to the litigation are less likely to settle to the extent that their predictions
of the outcome at verdict diverge.41 Third, since insurers face greater risk,
the availability of liability insurance is likely to be decreased and, when
116 Chapter 5

insurance is available, premiums are likely to be higher to reflect this

added nondiversifiable risk.42
Decisions at verdict may influence terms of subsequent settlements,
and the law provides no objective benchmarks for valuing nonmonetary
loss. Juries are generally given broad instructions for valuing loss, such
as that payment should providing “fair compensation” or “reasonable
compensation,” rather than providing specific criteria for valuing loss.43
Juries do not have the benefit of knowing how loss has been determined
in the past in roughly comparable situations. Such data are generally
unavailable, and there is no provision for providing it even if such infor-
mation were available. There is no requirement that the rationale or
methods used for arriving at a specific value be justified or even explained.
An appeals process exists, but appellate judges can also lack objective
standards for assessing loss. Postverdict reductions in awards often occur,
but most often in a settlement arrived at by the parties.44

Flawed Public Policy Responses

The public policy response to what is perceived as capricious and

excessive awards in medical malpractice has been to impose limits
on payment for nonmonetary damages (e.g., California), and in some
states payment for total loss (e.g., Indiana). Not all limits on nonmone-
tary and total are fixed in nominal dollar terms as is California’s, but
most are.
Conceptually, the assumption underlying flat caps is that injury victims
with large losses are relatively overcompensated, but there is no empirical
evidence to support this assumption. In fact, the only systematic com-
parison of injury cost versus compensation in medical malpractice indi-
cates that such injury victims are undercompensated on average.45 Many
plaintiffs receive no compensation for their injuries, either because their
cases are dropped or because they lose at trial. On average, even those
plaintiffs who won at trial received only 22 percent more than their
monetary loss. The 22 percent was presumably for nonmonetary loss.
This hardly implies that the system is out of control on average. Further,
as explained above, imposing caps on payment for nonmonetary loss
 Ceilings on Nonmonetary and Total Losses 117

may have a disproportionate effect on low wage earners and on persons

with severe injuries.46
In addition, it seems highly implausible that some awards are “too
high” while no award is “too low.” If juries make errors at the high end,
presumably they make errors at the low end as well—for example, as a
reaction to testimony by an expert for the plaintiff whom the jurors
dislike.
Imposition of a cap not indexed to the rate of inflation for the economy
overall or for medical inflation in the limit is both arbitrary and unfair.
Considering inflation that has occurred since 1975, California’s cap was
effectively about $70,000 around 2005, and with inflation, its value was
falling fast.47 Pace et al. (2004) quantified the effect of not allowing the
California cap to rise with increase in overall prices. Rather than reduce
payments by 30 percent, which is what their empirical analysis revealed
the effect of California’s cap to have had, indexing the cap would have
reduced payments by 21 percent (through 1999).
In a conversation the first author had with an attorney-physician who
has advocated actively for flat caps, the latter argued that allowing the
limits to increase with increases in the Consumer Price Index would
introduce too much complexity. Yet it seems ironic that policies requir-
ing indexing of Medicare fees for physicians are promoted by the same
organizations that promote flat, unindexed caps. If an unindexed limit
were the solution to an excessive number of medical malpractice claims
and payments, then it could become a precedent for paying for particular
medical or surgical procedures judged by a few experts to be in “excess
supply.” Social Security benefits are indexed annually by the Social Secu-
rity Administration to account for inflation, as are damage values from
tables developed for providing guidance as to levels of compensation for
personal injury in some large European countries. Complexity is not an
issue for such benefits.
In California, the Supreme Court upheld caps on nonmonetary
damages.48 However, in other states, damage caps have been held to
violate federal equal protection and due process guarantees, and to
violate various state constitutional provisions that give individuals the
right to trial by jury.49 They are especially unconstitutional given that no
118 Chapter 5

corresponding societal quid pro quo existed to replace the victim’s right
to full recovery.50 This argument stems from a focus on the rights of
individuals who were injured.51
From another perspective, caps in particular, and reforms in general,
are meant to benefit a larger class of people, nonclaimants. By virtue of
these policies, nonclaimants enjoy greater accessibility to medical care at
lower fees and premiums.52 However, as discussed in chapter 3, caps may
adversely affect access to care and have a minor effect in reducing health
care costs.
In sum, a flat dollar limit on damages, expressed in nominal dollars,
does have the advantage of simplicity. However, it is too simple and is
a bad precedent.

Better Ways to Reform Payment for Nonmonetary Loss in Medical

Malpractice Registries

Lack of Guidance to Juries

Juries are asked to set damages in individual cases on a one-time
basis, without any prior experience in performing this task. Nor do they
have the benefit of prior experience of other juries. With exceptions of
data collection efforts, which are not part of the judicial system, there
are no systematic, official records of deliberations or findings to inform
future jury decisions. Some institutional memory exists in trial judges’
memory of prior legal decisions and in reported appellate decisions.
Nevertheless, there is in general little to guide jurors in valuing loss to
the plaintiff.
A few states (e.g., Florida and Texas) collect information on closed
medical malpractice claims. At the federal level, the National Practitioner
Data Bank collects information on all closed medical malpractice claims
resulting in payment. Although useful for some purposes, such as a rough
guide for monitoring quality and for research, this information is insuf-
ficient to serve as a guide for juries in ascertaining the appropriate size
of an award.
For one thing, there is insufficiently detailed information on the sever-
ity of the injury and on the circumstances under which the injury
occurred.53 Moreover, for any particular case, it is not currently possible
 Ceilings on Nonmonetary and Total Losses 119

to know about the existence of similar cases and how they have been
decided.54

Scheduling Damages
Rather than set a limit on the maximum size of an award, it would be
preferable to set payment criteria for all awards, not only the large ones.55
Scheduling damages involves such an approach. Because scheduling
affects the whole distribution, not just the upper tail, it is conceptually
superior to flat caps on grounds of equity of payment to claimants with
very severe injuries relative to those with less severe injuries (“vertical
equity”), and has been advocated by several commentators and by the
Institute of Medicine.56 The trial bar opposes this approach on grounds
that it would limit the ability of plaintiffs to make a case for their special
circumstances and for their attorneys to exercise skill in arguing the merits
of these circumstances. Such flexibility, however, must be measured
against the horizontal (equal or unequal treatment of equals in terms of
injury severity) and vertical inequities of the current system.
Bovbjerg, Sloan, and Blumstein (BSB; 1989) propose three approaches
worthy of consideration for reforming payment for nonpecuniary loss in
medical malpractice as replacements for flat caps as well as for use in
states without any limits on awards. Since they were proposed in 1989,
these ideas have elicited interest in the academic community, but they
have not received serious consideration by any state, probably for several
reasons. First, organized medicine finds flat caps more attractive. Second,
the trial bar does not wish to introduce any limitations on determinations
of awards. And third, the proposals have no proactive supporters. The
road from the university seminar room to the political marketplace is
often not well trodden.
The first approach is an award matrix for nonmonetary loss. The
matrix would display awards for nonmonetary loss for injuries with
specific characteristics. The underlying presumption is that payment for
nonmonetary loss should be based on objective factors, such as the per-
son’s age or life expectancy, severity of injury, and type of body part
affected. More permanent and serious injuries are likely to have higher
underlying nonmonetary loss, although the relationship between severity
rank and loss may not be linear.
120 Chapter 5

Existing injury severity scales may be used to classify injuries by sever-

ity levels, such as a nine-point injury severity scale in widespread use,
which varies from “emotional only” to “permanent grave” injuries and
death (table 5.1). The nine categories of injury severity are broad, rather
objective, and mutually exclusive, thus permitting relatively little upward
“creep” in category assignment by advocates of the plaintiff. Assignment
to a category would be open to challenge in any event.
For temporary injuries, a younger person may recover more quickly
than would an older person. However, for permanent injuries, a younger
person would be expected to have to endure the nonmonetary loss for
a more extended time. Thus, it would be appropriate to cross-classify by
age as well as severity of injury. The values in the matrix could be based
on prior awards, past research valuing life and quality of life, and/or set
by a state legislature.
The approach of basing matrix values on prior awards for nonmone-
tary loss by past juries, adjusted by trial and appellate courts, is straight-

Table 5.1
Severity of Injury Scale

Severity of Injury Examples

1. Emotional only Fright, no physical damage

2. Temporary insignificant Lacerations, contusions, minor scars, rash. No
delay in recovery
3. Temporary minor Infections, mis-set fractures, fall in hospital.
Recovery delayed
4. Temporary major Burns, surgical material left, drug side effect,
brain damage. Recovery delayed
5. Permanent minor Loss of fingers, loss of or damage to organs.
Includes nondisabling injuries
6. Permanent significant Deafness, loss of limb, loss of eye, loss of one
kidney or lung
7. Permanent major Paraplegia, blindness, loss of two limbs, brain
damage
8. Permanent grave Quadriplegia, severe brain damage, lifelong
care or fatal prognosis
9. Death

Source: National Association of Insurance Commissioners (1980).

 Ceilings on Nonmonetary and Total Losses 121

forward in principle. In practice, states would have to set up a registry

to assemble the data. The matrix would be subject to judicial review and
might be modified accordingly.
There is a large literature on valuing life,57 but it is not sufficiently
specific to fill in cells of a matrix with a nine-point severity scale, several
age categories, and data splits by body parts. This would require more
research.58 There is much more evidence on willingness to pay to avoid
a death than for outcome measures set according to a dimension of
quality of life. Values are based on inferences from market data or from
surveys eliciting individuals’ willingness to pay to avoid certain adverse
outcomes.59
Also, in using estimates of value from value of life or quality of life
studies, one would be making the implicit assumption that the context
by which life or its quality is lost does not matter. That is, losing a life
in a war, which has popular support, in a hit-and-run motor accident,
or at the hands of an incompetent surgeon in the operating room have
the same nonmonetary loss. Yet the context in which the injury occurred
may matter. On the other hand, it may be desirable that awards are made
context-free and solely on more objective factors. Values would be
inflated by a price index, such as the Consumer Price Index.
Previous research has documented that (1) payments rise monotoni-
cally through the nine injury severity categories, with payments for loss
of life being lower on average than for persons with grave, permanent
impairments and (2) despite evidence of a plausible relationship between
mean payments and injury severity, there is considerable variation in
payments within each severity category.60 This suggests that the legal
system is taking more into account than severity per se in assessing
damages. Even though paying on a nine-point scale would improve verti-
cal equity over flat caps, merely adjusting for severity may not be an
improvement over the current approach in states with flat caps, which
is not to reduce payments in amounts below the statutory ceilings.
In principle, this method is simple, but it would require several
steps to implement. In its defense, however, it is conceptually more
appropriate than a flat cap. Payments based on an award matrix
for nonmonetary loss would probably not at all resemble those based on
a flat cap.
122 Chapter 5

A simpler variant of the award matrix, which encompasses all awards,

would be to use flexible ranges in lieu of caps. This system would regulate
the lowest and the highest payments for nonpecuniary loss. The state
would designate payments in the lowest quartile and the top deciles of
payments (assuming that the data were made available through a registry
or in some other way).61 Payments for nonmonetary loss would be
limited at both extremes, raised at the bottom and lowered at the top.
Flexible ranges are less likely to be opposed by the trial bar since they
leave some room for the application of lawyers’ skill in obtaining
payment for clients. However, such ranges may violate the principle of
vertical equity since payments for lower and higher severity injuries may
overlap.
Under a third approach, a common set of injury scenarios would be
developed and assembled to assist juries in valuing nonmonetary loss.
Juries would be given a notebook containing descriptions of particular
injuries and descriptions of the pain and suffering associated with each.
Dollar values would be attached to each scenario, but jurors would not
be told the values. Values associated with each scenario could be set by
the state legislatures, perhaps with the guidance of a study commission.
Dollar amounts would be assigned to the injury after the choice of sce-
nario was made by the jury. Jurors might be shown up to ten different
scenarios from a larger number made available to the court. Juries would
be told that none of the scenarios fit the case under consideration but
would be asked to indicate which of the scenarios in the notebook most
nearly resembles the case under consideration. Jurors might respond, for
example, that pain and suffering seem less than scenario 3 but worse
than scenario 5. In such a case, the midpoint of the two might be selected
by the judge when setting compensation for nonmonetary loss. The judge
would thus have some discretion in picking the final value. This is analo-
gous to stating in undergraduate microeconomics that indifference curve
3 lies below indifference curve 5, and a point is selected in between the
two curves.62 There are many points between the curves, as with the
scenarios.
A neutral scenario might read: “Permanent minor injury (level 5 sever-
ity). Life expectancy 25 years. Mild persistent pain, usually controllable
with aspirin. Unable to engage in more than light housework.”
 Ceilings on Nonmonetary and Total Losses 123

A more colorful scenario might read: “Plaintiff Peters has completely

and permanently lost the use of her right arm. Her life expectancy is 25
years, according to standard life insurance tables. Her arm throbs with
pain most of the time, but the pain usually can be controlled with aspirin.
She cannot do more than light housework.”
The scenarios could be more detailed than this. Each scenario might
be delivered to jurors by video. The scenarios might characterize actual
cases tried in the state in recent years. The process of selecting cases in
the notebook that are closest to those in the case being tried could be
mandatory or voluntary. If voluntary, the scenarios could provide guide-
lines for jury deliberations. In that case, juries would be told the associ-
ated dollar values after they had identified the closest scenarios, and they
would be asked to reconsider their determination of a dollar value of
nonpecuniary loss arrived at independently. Even if the system were not
mandatory, large discrepancies between the jury award and the scenar-
ios’ values may be subject to review on appeal.
An objection to BSB’s suggestion that juries be provided information
on past awards for nonmonetary loss or scheduling based on a matrix
that is based on averages of past losses is that if juries have made mis-
takes in the past, it is inadvisable to use past decisions as guides for
future ones.63
Geistfeld (1995) proposes an alternative method of promoting consis-
tency of jury awards—in our list of suggestions, the fourth approach.
His valuation method does not rely on past jury awards. Rather, his jury
survey approach proposes that the jury be asked to indicate how much
a reasonable person would have been willing to pay to avoid the injury
in question.64 The approach would allow the jury to consider the
nuances of the case being tried, an element that scheduled damages
can only approximate. The approach Geistfeld suggests is feasible to
implement; it has been widely applied to study valuation in various
contexts, and researchers now have considerable experience in question
design (much more so than in 1995). Most frequently, the surveys are
computer-administered.65
Geisfeld has juries asking the appropriate question—What would pre-
venting the injury have been worth before the injury actually occurred?
However, the valuation studies have employed hundreds of persons in
124 Chapter 5

the valuation task, not just six to twelve jurors (the usual size of juries).
Having few persons valuing the loss will almost certainly lead to very
imprecise estimates. One outlier response could greatly distort the esti-
mates, but there are several possible ways to deal with this problem.66
Avraham (2006) is critical of all of the above approaches. He states
that:
All of the proposed solutions are administratively complicated and prohibitively
complicated. Who decides the schedules, matrices, scenarios, or guidelines? What
criteria do they use? The more detailed the scenarios or guidelines are, the more
costly it is to design them, and the less discretion the jury has. How do we know
that the jury will not be overburdened with these new tasks? Even a simple
matrix (conventionally used for evaluating malpractice insurance cases into
which all injuries, including death, are collapsed) introduces a wide range of
awards within each category. If this “simple matrix” results in such a massive
variation within injury-severity type, it is clearly unhelpful, and one can reason-
ably expect that an even wider range of awards would result if states adopt
Bovbjerg, Sloan, Blumstein’s suggestion that a jury apply different scenarios to
the case at hand. This problem is further complicated when it is extended to
the determination of what the criteria for the judicial review of jury verdicts
would be.67

Outcomes of any method of scheduling depend on details of how the

method is actually applied. If juries are given less flexibility in applying
the schedule and there are fewer categories, there will be correspondingly
less variability. Flat caps on damages give no discretion to juries for
awards above a given dollar value. Simplicity is gained at the expense of
precision. In addition, this approach reduces variability in payment for
severe injuries, but it introduces horizontal inequities since there is het-
erogeneity in injuries above the flat cap as well as vertical inequities,
because compensation for severe injuries is disproportionately reduced.
Avraham (2006) proposes a simple approach that is sixth in our list,
basing compensation for nonmonetary loss on compensation for medical
loss. Medical loss would be multiplied by a “multiplier” to yield the
award for nonmonetary loss. To illustrate, he suggests that the multiplier
for medical losses of up to $100,000 could be 0.5, yielding nonmonetary
damages in the range of $0 to $50,000. For losses in the $100,001–
$500,000 range, the multiplier might be 0.75. And for losses of $500,001–
$1,000,000 and above $1,000,000, the multipliers might be 1.0 and
1.25, respectively.
 Ceilings on Nonmonetary and Total Losses 125

The idea clearly is simple, but it, too, has deficiencies. For example, a
person who had a leg amputated would receive less compensation for
pain and suffering than would a person who received several complex
surgical procedures to reattach nerves in the leg and who retained use
of the leg. The proposal could increase use of personal health care ser-
vices, such as home care and rehabilitation services, in order to capture
a higher payment for noneconomic loss. The nine-point injury severity
scale shown in table 5.1 is less easily gamed.
In sum, each of the above six proposals for scheduling damages has
both strengths and weaknesses, and there are undoubtedly other propos-
als we have not mentioned. Some form of scheduling makes sense. In
deciding how best to schedule, states will need to weigh the pros and
cons of the alternatives.

Better Ways to Reform Payment for Nonmonetary or Total Loss in

Medical Malpractice Insurance Contracts for Future Services: Payment
in the Form of Provision of Service Benefits

Some plaintiffs who seek compensation through tort have suffered per-
manent, serious injuries requiring lifelong care. It is such cases, rather
than temporary injuries, that often lead to volatility of compensation.
Currently, awards are not well synchronized with the outlays that such
injury victims incur. Imprudent injury victims may squander their awards
and later require public subsidies or rely on uncompensated care subsi-
dies. Furthermore, even if awards are not excessive on average, they are
likely to be either insufficient or excessive in many cases. Substantial
lawyers’ fees appreciably reduce any payment that plaintiffs might
receive, leading to the likelihood that net compensation after fees is
insufficient to cover injury cost. Grave injuries, such as quadriplegia and
severe brain damage, may require substantial sums, often exceeding
$100,000 per year. Even thirty annual payments of $40,000, discounted
at 5 percent, have a present value of $615,000.
At present, to compensate plaintiffs for the cost of care, experts are
employed to compute the sum of past cost, often including interest on
such cost, and the anticipated cost of future care, often discounted to
present value. Once awarded a lump sum, plaintiffs or the persons des-
126 Chapter 5

ignated to manage their care have an incentive to economize on care.

However, such persons may not be adequately informed about care
options, and as individuals they would not typically have the bargaining
power to obtain the best deals for their money.
The suggestions just described focus on determination of nonmonetary
loss. There is reason for concern about the accuracy and adequacy of
payment for monetary (and total) loss as well.
Even under the best of circumstances, forecasts of future economic loss
will inevitably have been too high or too low ex post. Future costs of
medical care and other personal care services are difficult to project. In
addition, the course of disease, disability, and longevity is highly indi-
vidualized and extremely difficult to predict ex ante. Life expectancy, for
example, refers to the midpoint in the frequency distribution of future
dates of death. There is considerable dispersion around this midpoint.
To the extent that forecasts and payments are too high, there is a
potential windfall to plaintiffs. If the forecasts are too low, there is the
risk that the plaintiff will not receive adequate financial support. Payment
of cash at the settlement or verdict does not eliminate the financial risk
that the recipient faces. In principle, the recipient could purchase private
health and disability insurance with funds from the lump-sum payment.
However, in practice, individual health and disability insurance policies
tend to be very costly relative to group insurance policies. Also, with a
preexisting condition, it may be difficult for some persons to obtain
coverage, especially full coverage, at any price anywhere near actuarially
fair rates. Finally, the recipient and/or his or her agents may be insuffi-
ciently prudent to plan for the future.
Periodic payments, as a “tort reform,” are an annuity-like contract to
provide payments on a regular basis in the future. They have two advan-
tages over lump-sum payments. First, they deal with uncertainty about
the plaintiff’s longevity. Although empirical evidence is lacking, provi-
sions for periodic payments have been enacted by state legislatures under
the presumption that they will reduce overpayment to the extent that
juries, in attempting to the minimize the risk of a shortfall in compensa-
tion to the plaintiff, set damages too high on average.
Periodic payments address the problem of a plaintiff being overpaid
to the extent that plaintiffs die prematurely as well as the problem of
 Ceilings on Nonmonetary and Total Losses 127

the imprudent plaintiff who spends the lump sum soon after receiving
it. Yet there are several potential problems not addressed by periodic
payments.
First, periodic payments are often provided in nominal terms and
therefore do not shield the recipient against the risk of unanticipated
inflation. Second, they do not provide protection against unanticipated
deterioration in health, although there may be a windfall if the person’s
health improves more than was anticipated. Third, to the extent that
plaintiffs would want to insure against future risk, periodic payments do
not take account of the high price and/or unavailability of insurance to
such individuals. Fourth, if the plaintiff-injury victim receives cash rather
than insurance as an in-kind benefit, and thus becomes a cash-paying
“self-pay” patient, he or she is likely to be paying top dollar because of
an asymmetry in negotiating power between the individual and provid-
ers. Fifth, plaintiffs and/or their agent may mismanage cash payments
provided on a periodic basis. There may well be an advantage to paying
injury victims in a noncash form. Sixth, particularly since the award is
reduced by the amount of the lawyer’s fee, it seems unlikely that the
plaintiff could purchase an insurance policy that would provide financial
protection against unforeseen changes in health.
As an alternative to periodic payments or a lump-sum amount, a
potentially attractive reform would be for the defendants found liable to
fund an insurance contract for future services that would provide for or
pay for future services as the needs arise. In essence, rather than being
paid money, the plaintiff would be guaranteed that a set of services
would be provided, conditional on patient need—hence the term “service
contract proposal.”68
Under the service contract proposal, rather than award damages, juries
would specify features of an insurance contract appropriate for the care
of the injured person. Such features would include duration of coverage
and services to be covered, typically specified in considerable detail, but
would exclude services not causally connected to the injury and could
provide for collateral source offsets.
Proposals from prospective contractors would specify details of ser-
vices they would provide and their associated prices, which would reflect
the expected cost of the contract in terms of service benefits plus a
128 Chapter 5

loading factor, which includes administrative costs to the insurer as well

as profit.
The court would then select a contractor. Potential contractors would
include academic medical centers, hospitals, and larger physician group
practices, as well as HMOs and other insurers. This would be a business
opportunity for providers who could specialize in particular types of
injuries, such as care for the severely and permanently impaired, neuro-
logically impaired children, or persons with cancer that becomes more
advanced because of a misdiagnosis.
This alternative is more complex than the above proposals, and resolu-
tion of various details would be critical to its success. Under a service-
contracting system, services to help persons cope with nonmonetary loss,
such as counseling, could be part of the contract and could substitute
for much or all of the payment for such loss. Since injury victims would
receive more comprehensive, specialized care tailored to each individual’s
circumstances, including psychiatric needs, they would receive something
in return for eliminating or substantially reducing compensation for non-
monetary loss.
As with the other proposals, even though the proposal may seem
attractive conceptually, the devil is in the practical details of implementa-
tion. First, since prices and conditions of the contract would be expected
to differ, plaintiffs and defendants may not agree on the choice of
contractor.
Moreover, there may be details about future care that are not specified
by the court, as well as differences of opinion on where care is to be
provided. Should a brain-damaged infant be housed near the family resi-
dence or in a state facility an hour or two away? Should the injury victim
receive individual or group psychotherapy? At what point is care to be
discontinued because the patient is considered to be terminal? In many
cases, disputes would be settled informally or by private negotiation,
with results being finally accepted by the court.
However, if the parties cannot agree within a prespecified time period,
a court-appointed arbitrator could resolve the dispute by final-offer
arbitration.69 In this type of arbitration, each side presents its best offer.
The arbitrator has his or her own idea of the appropriate settlement
 Ceilings on Nonmonetary and Total Losses 129

terms and selects the proposal which most closely resembles his or her
estimate of fairness. Given that both parties know that the arbitrator will
select the contract that is most “reasonable,” each has an incentive to
make offers toward the middle of the range of possible values. In effect,
final-offer arbitration aligns the incentives of the parties. Although this
approach may seem cumbersome, under the present circumstances, injury
victims may be insufficiently empowered to gain resources that they will
eventually need.
Another problem may arise when the parties are constantly dead-
locked and/or lose trust of the insurer-provider, and the plaintiff wishes
to terminate the relationship. Or the insurer may file for bankruptcy.
Thus, there must be a provision for cash-out of the insurance contract.
However, this is not without complicating factors as well. For example,
a plaintiff who believes that he or she will use fewer services than the
average plaintiff with this contract type may wish to cash out. Thus,
there would have to be a financial penalty for cashing out which is part
of the initial contract provisions. On the other hand, providers-insurers
may wish to cash out in the event of unanticipated high rates of inflation
or unanticipated regulatory (e.g., new quality of care safeguards), or
technological changes (e.g., a new procedure for treating the disease).
The provider-insurer may not have a formal escape clause, but it could
offer an attractive settlement to encourage the plaintiff to exercise the
buyout option. It is indeed possible that an unhappy provider-insurer
could let service deteriorate to the point that the plaintiff voluntarily
seeks a buyout on unfavorable terms. To guard against this risk, it would
be important to provide safeguards in implementing the proposal. For
example, the plaintiff-injury victim could be allowed to seek redress in
a contempt proceeding in the court of original jurisdiction; the plaintiff
could file another lawsuit; and/or complaints could be kept on file in a
clearinghouse to provide a Better Business Bureau type of service to
courts and future litigants.
This plan calls for a new market in insurance service-benefit contracts
to cover future losses of persons who have been awarded damages by
courts in personal injury cases, and specifically in medical malpractice.
Since the vast majority of cases are settled, there should have been ample
130 Chapter 5

opportunity for this type of market to emerge by now. The next question
is why such a market has not emerged, and what, more specifically, the
objections to the proposal are likely to be.
We can anticipate one objection. During preparation of Blumstein,
Bovbjerg, and Sloan (1991), which contained this proposal, the third
author presented the proposal at the home office of a major multiline
insurer. The proposal drew a skeptical response, primarily because there
are no actuarial data on which to base offer prices. One possible initial
source of data for projecting future losses may come from workers’
compensation insurers. These insurers have the obligation to cover life-
time costs of care for certain claimants. Furthermore, the same issue
arises for no-fault compensation plans, but this objection has not arisen
as a major shortcoming of such plans.
Second, particularly if the market were restricted to verdicts in medical
malpractice cases and there were barriers to entry from insurance com-
panies domiciled in other states, the market may not be of a sufficient
size to attract insurers willing to supply long-term service benefit con-
tracts. Without being able to exploit scale economies, the load on the
insurance policies is likely to be high. However, rather than being a
structural flaw, this concern suggests that the plan not be restricted to
medical malpractice cases, but to personal injury cases more generally.
In addition, federal as well as state courts could implement this idea.
A third possible concern involves adverse selection. Perhaps the only
plaintiffs who will consent to insurance contracts are those with private
information that they will be high users of services. One way to deal
with adverse selection is to make participation in the contract plan man-
datory; this might encourage settlements by those who do not expect to
be high users on a risk-adjusted basis. Another approach would be to
implement an outlier policy so that unusually high-cost cases could be
covered by a reinsurance pool.
Fourth, while service benefits may make the victim whole in pecuniary
terms, the proposal does not deal with nonmonetary loss. However, there
should be a gain in well-being to injury victims by not having to deal
with the substantial expenditure risk in the current system. A service-
contract proposal could be combined with scheduling for nonmonetary
loss.
 Ceilings on Nonmonetary and Total Losses 131

These are generally valid concerns. However, the current system masks
many of the problems that the insurance contract option makes explicit,
such as asymmetric negotiating power between individuals and insurer-
providers, the failure for cash payments to compensate for expenditure
risk, and, quite simply, the fact that after paying lawyers’ fees, injured
persons may not have the ability to pay for their health care.
Finally, Peter Schuck (1991, pp. 219–220) concludes that “The virtues
of the [service benefit contracting] proposal are less apparent than those
of damages scheduling [specifically Bovbjerg, Sloan, and Blumstein’s
scheduling proposals].” Schuck contends that the service benefit insur-
ance proposal depends on a paternalistic assumption that contract rights
to future services are superior to an immediate lump-sum payment because
the latter might be squandered. He further argues that it is unclear why
paternalism is any more justified in the case of personal injury victims
deciding how to dispose of their resources than for other adults.
In fact, the proposal is somewhat paternalistic, and for good reason.
Many families have to cope with serious, ongoing permanent injuries for
years. Budgeting over the long term can be very difficult. In the case of
a permanently injured child, the parents may predecease the child. The
individual family is not well positioned to bargain with suppliers of care.
To advocate for a paternalistic solution under these special circumstances
is not at all advocating for paternalism more generally.

Private Contracting

An option that has a few strong advocates is private contracting as an

alternative to tort. Private contracts might explicitly limit the circum-
stances under which tort liability would still apply. For example, tort
liability might continue only under gross negligence. Contract provisions
could also specify the method by which disputes are resolved (e.g., by
arbitration), guidelines under which care is delivered that, if adhered to,
would not result in payment of damages (e.g., specific guidelines or
organizations promulgating guidelines, as well as schedules for paying
for nonmonetary loss.
A major objection to private contracting is that patients are not as
well positioned as are parties to a contract, at least relative to health care
132 Chapter 5

providers. In specific instances, such as in medical emergencies, this argu-

ment has even greater force.
The counterargument is that contracts can be made when patients are
well, not sick. Moreover, agents for patients, such as employers, should
be in a better position to deal with provider groups. Employers deal with
provider groups on many matters, including fees and use of services.
Danzon (2000, p. 1382) argues that if health plans can lock in patients
to providers who have adopted cost-reducing contractual changes, some
of which substitute for tort, there can be benefits to patients in the form
of lower contributions to health insurance plans and lower wage offsets
for that part saved by employers on their contributions to health
plans.
Although the contracts would encompass many provisions other than
scheduled damages, they could be used for scheduling. In fact, the legal
environment permitting, this may be an effective mechanism for imple-
menting scheduling and demonstrating its feasibility.

Discussion and Conclusions

In chapter 1, we described myth 4 as “medical malpractice claimants are

overcompensated for their injuries.” There are surely cases in which
myth 4 is valid, but based on available empirical evidence, this myth is
invalid on average. If there is an argument for limits on payments, it is
that payments are highly variable, not that they are upward biased.
Ironically, limits on payment for nonmonetary loss and total loss may
have “worked” too well. They have reduced outlays of insurers and
medical malpractice premiums to a lesser extent. In accomplishing these
goals, they have served the interests of defendants, and obtaining passage
of such limits in states without them and at the federal level has become
a political priority for these stakeholders. In the first decade of the
twenty-first Century, proposals for flat caps have garnered more political
support than ever, although they have insufficient support to be enacted
at the federal level.
This chapter proposes specific alternatives to flat caps. The concept of
scheduling damages is not at all new. For example, in a chapter titled
“Visual Economics,” there is a detailed description of how pensions for
 Ceilings on Nonmonetary and Total Losses 133

disability, in this case for disability on account of vision loss, should be

determined.70 Although the task of the person applying principles of
visual economics was to compensate for monetary loss (i.e., loss in earn-
ings potential), aspects of the payment scheme pertained to nonmonetary
aspects as well. According to the payment schedule in 1905, loss of sight
in one eye with the eye in place was to be compensated at a rate of $12
per month. However, if the eye was missing, the rate was $17 per month.
The justification given was that potential employers would be willing to
pay less if they saw that the eye was actually in place. While this may
be considered monetary loss, from another perspective this could indeed
have been a way to circumvent the statute and pay for nonmonetary
loss.
Groups representing providers seek to limit outlays for medical mal-
practice. The above proposals could well lead to greater accuracy in
determining damages and to lower risk, especially to plaintiff-injury
victims, but the proposals would not automatically result in savings to
insurers and health care providers. Organized medicine understandably
wants policy solutions that would affect savings to its constituents.
Scheduled damages are an anathema to the trial bar, as are flat caps.
Both flat caps and scheduled damages limit opportunities for a skillful
and industrious attorney to obtain high compensation for clients.
In contrast to these special interests, the public in general, and injury
victims in particular, are not well organized politically. The public rarely
gets directly involved in a medical malpractice case as a juror, as an
injury victim, or as a relative of one. Thus, citizens are prone to follow
the finger-pointing of the well-organized groups.
This entire agenda will not be adopted at once, and considerable
compromise will be needed to adopt any of the options. The first step is
recognition that the facts are not as simple as they tend to be portrayed
and that existing data are insufficient to implement sound policy options.
The most realistic first steps may be implementation of registries to
provide the requisite data and demonstrations to test the feasibility of
the proposals before full-scale implementation takes place.
Given these political impediments, then what are the next steps? It
seems doubtful that insurers will implement the proposal for long-term
contracts for permanently injured individuals. Yet this is potentially an
134 Chapter 5

excellent opportunity for a health care provider, such as an academic

health center. These organizations have a wealth of talent on their facul-
ties and staffs, and should be well positioned to manage care for perma-
nently injured persons over the life course. This would provide teaching,
research, and income opportunities for academic health centers. The
advantages for teaching seem immediately apparent. For research, this
would provide opportunities to study individuals over the life course as
well as to study the effects of interventions of various types on the well-
being of these individuals. Only some of the interventions would be
medical. Others would involve special education, rehabilitation, housing,
and other aspects of these individuals’ lives.
Scheduled damages are part of some compensation systems, such as
workers’ compensation. They could be implemented as part of medical
no-fault. In addition, they may possibly be part of private contracts that
would provide an alternative to tort. It will be necessary to demonstrate
the practical feasibility of scheduled damages before this approach has
any prospect for success in a state legislature, especially given the lack
of support from both sides of the political debate about tort reform.
Chapter 12 will describe feasible options for reform. Because of the
political challenges in getting these proposals implemented, scheduling
damages and contracts for permanently impaired injury victims should
be part of the reform agenda but, with the political impediments to
implementation—perhaps not at the very top.
6
Compensating Plaintiffs’ Attorneys

Trial Lawyers, Contingent Fees, and Medical Malpractice Litigation:

Opposing Views

The Critique
Critics of medical malpractice quickly turn to the role of the trial lawyer
as a cause of the high cost of medical malpractice. In some accounts,
such lawyers are seen as “ambulance chasers,” arriving at the scene of
an injury and readily offering their services to injury victims who, without
these persuasive lawyers, would not have sued for medical malpractice.
The image of greedy lawyers stirring up lawsuits is linked to payment of
lawyers on a contingent fee basis because the contingent fee system pre-
sumably gives the lawyer an added incentive to pursue injury victims.1
In medical malpractice, the injury victim may not be found literally lying
on the side of the highway, but rather may be identified in some other
way—or attracted by a lawyer’s advertisement or other form of self-
promotion. The perceived motive for the lawyer is to make a quick and
big buck. There is a widespread perception in other countries that the
high rates of litigation in the United States are largely attributable to the
contingent fee system (box 6.1).2
While the victim may receive compensation from tort, whatever the
victim receives is substantially reduced by the lawyer’s high fee. Given
the high rewards to suing, the argument goes, physicians and other
potential health care defendants are sheep waiting to be fleeced. While
in normal markets, entry would reduce fees to a competitive level, trial
lawyers’ fees are kept far above competitive levels by anti-competitive
practices, typically not precisely specified by the proponents of this view.
136 Chapter 6

Box 6.1
Public Opinion About Lawsuits in the United States: Unfavorable to Trial
Lawyers

Michael Saks (1992, pp. 1162–1163) summarized results of a public

opinion poll conducted in the mid-1980s.
• By a 69–24 percent majority, people were convinced that it is too easy
for people to sue for damages when they think they have been injured or
wronged.
• 63 percent believed that the size of most cash settlements is excessive.
• A 77 percent–15 percent majority blamed the liability crisis on persons
who think they can make a lot of money from such suits—namely, trial
lawyers.

The near uniformity of contingent fee percentages, according to one

view3 in the 31–33 percent range or higher, is taken as evidence of lack
of competition. They may be fixed by custom if not by collusion.4 Some
criticisms focus on the high earnings of trial lawyers, especially on earn-
ings as calculated (but not paid) on an hourly basis, with a rationale for
reform seeming to reflect a concern about inequities in the distribution
of earned income. But underlying concerns about inequity, even if not
expressed, are probably that the prospects of lawyers striking it rich has
led to too many medical malpractice suits being filed.5 In light of these
allegations, states have passed limits on attorneys’ contingent fees in
efforts to reduce litigation rates and payments, and perhaps to address
an imbalance in earnings from the practice of law, at least on behalf of
plaintiffs.

Another Perspective
There are sharp differences in opinions about trial lawyers and how they
are paid (box 6.2). From the perspective of perhaps a minority of citizens,
including many injury victims, contingent fees are about access to justice
through the mechanism of civil litigation or the threat of it.
Much more forceful support for the contingent fee system comes
from consumer advocates. For them, the contingent fee system for
compensating plaintiffs’ attorneys in personal injury cases is a godsend
 Compensating Plaintiffs’ Attorneys 137

Box 6.2
Public Opinion About Lawsuits in the United States: Counterpoint

In concluding his book on contingent fee lawyer compensation and reputa-

tion, Kritzer (2004, pp. 253–254) said, “Supporters of contingency fees
describe them as the average person’s ‘key to the courthouse.’ In contrast,
critics of contingency fees see such fees as at least partly responsible for
many of the evils and excesses of the American legal system (Barry and
Rein 1999; Brickman 2003; Kagan 2001; O’Connell 1979) and as unjustly
enriching and empowering members of the legal profession (Brickman
1989, 1996; Olson 2003). Nonetheless, from the perspective of the average
citizen, contingency fees are about access to justice through the mechanism
of civil litigation, or the threat of civil litigation.
Does the availability of contingency fees increase the resort to litigation?
It seems that the obvious answer must be Yes, of course. However, that
answer is too simplistic, because one must go on to ask, “Compared to
what?”

for the injury victim. Absent the contingent fee system, trial lawyers
would adopt other payment approaches, such as charging by the
hour.6
However, unlike lawyers who can diversify away the risk of losses in
individual lawsuits by accepting many cases (analogous to assets in an
investment portfolio), the individual injury victim has no means of diver-
sification. He or she has only one claim. If it succeeds, the claimant may
get substantial compensation. If it fails, the claimant gets nothing; and
under an hourly compensation system or a loser-pay-all-legal-costs
system (the English Rule), losers would be stuck with a substantial legal
expense in addition to receiving no compensation. A risk-averse person
may be unlikely to file a claim for which the lawyer is paid on an hourly
basis because of the high risk of losing. The risk of losing when an
attorney is paid hourly means being forced to pay legal expenses without
the cushion of revenue from suing.
In contrast to the view that placing an upper limit on contingent
fee percentages is good public policy since it limits the ability of
people to file nonmeritorious claims and trial lawyers’ allegedly excessive
earnings, opponents of such limits see them as reducing individuals’
access to needed legal services that otherwise would not be available.
138 Chapter 6

More relevant to a prospective client is the expected recovery net of the

fee, which is likely to vary considerably among lawyers with differing
know-how and effort levels. To the plaintiff lawyer deciding to take a
case, the relevant datum is the expected revenue from the case, not the
fee percentage per se. If trial lawyers are doing good work obtaining
compensation for deserving injury victims and make good income in the
process, the income distribution suffers no harm and may even be
improved. Both the claimant, who may have incurred a substantial
income loss as a result of the injury, and the attorney are made financially
better off.
Sloan, Githens, Clayton, et al. (1993) report that plaintiffs in medical
malpractice cases who retained an attorney specializing in medical
malpractice litigation obtained higher recoveries than did plaintiffs
who used attorneys not specializing in this field. The specialized
attorneys were plausibly much more affluent than general attorneys, but
the plaintiffs who used them were much better off in the end. However,
just because the recoveries were higher does not necessarily mean that
specialized lawyers are more productive than others, although they
may be. The higher recoveries likely reflect better case selection on
their part.
Specialization is not limited to the trial bar. Specialization and skills
obtained by being a repeat player are more common among attorneys
for the defense in medical malpractice cases than they are among those
representing plaintiffs.7
In contrast to much public commentary, which is often critical of
contingent fees charged by attorneys, the American Bar Association’s
Standing Committee on Ethics and Professional Responsibility has issued
a formal opinion in support of contingent fees. As long as the fee is “both
appropriate and reasonable,” and the client has been informed of alter-
native billing options, the committee considers the fee agreement to be
ethical. The committee further states that it is ethical to charge a contin-
gency fee when liability is clear and recovery is expected. Also, the lawyer
is not obligated to solicit an early settlement offer when compensated on
a contingent fee basis.8
In sum, the issue may not be so much that trial lawyers generate too
many lawsuits. Rather, statutory limits on such fees, by limiting access
 Compensating Plaintiffs’ Attorneys 139

of injury victims to compensation through tort, may result in even fewer

meritorious claims being filed than in states in which such fee limits do
not exist.
While this line of argumentation is helpful to opponents of limits on
plaintiff attorneys’ contingent fees, it also has some weaknesses. For one,
among those who advocate for the existing contingent fee system on
grounds that it provides broad access to legal services, there are alterna-
tives worth considering, such as first-party insurance for legal expense,
as exists for personal health services, homeowners, and automobile col-
lision insurance.

Our View
Although all proposals have both pluses and minuses, the case for con-
tingent fee reform, at least as implemented to date, is weak on balance.
If reform is needed because the courts are overburdened, why focus on
medical malpractice, which accounts for less than 1 percent of lawsuits?
If the issue is too many nonmeritorious lawsuits, do limitations on con-
tingent fees winnow out nonmeritorious lawsuits or meritorious ones as
well? If the issue is an inequitable income distribution, why focus the
public policy discussion on a few trial lawyers with high earnings rather
than on highly paid professionals involved in other activities, such as
lawyers who handle mergers and acquisitions and patent disputes, health
professionals, and CEOs in general?
Statutes limiting contingent fees generally set the lowest fee percent-
ages for the largest payment awards. This implies that the large cases
disproportionately result in excess profits to plaintiffs’ attorneys. At
some threshold of awards, this is plausibly so, but empirical evidence on
this point is lacking, including the threshold above which litigation on
a contingent fee basis is “excessively” profitable.
By contrast, more recent proposals focus on limiting fees in cases that
are resolved quickly, but leave fees for disputes contested over a longer
time period unaffected. This implies that the cases that are quickly
resolved, often for lesser amounts, are excessively profitable. Empirical
evidence for this position is also lacking.
The only communality the proposals share is that they seek to limit
compensation of plaintiffs’ attorneys in medical malpractice litigation.
140 Chapter 6

But, they leave out attorney compensation in high-profile merger and

acquisition, patent, and class action suits, for example. In contrast to
lawsuits against pharmaceutical and device manufacturers, which fall
under the category of product liability, personal injury cases in which
physicians, hospitals, or other health care providers are named as defen-
dants do not achieve class action status.

Types of Compensation Arrangements

There are several methods of paying lawyers. There is no consensus

among theoretical studies about effects of alternative methods of payment
on lawyer decision-making and legal outcomes. The conclusions from
the theoretical studies about effects of payment mechanism on lawyer
decision-making and legal outcomes are highly dependent on the under-
lying assumptions made in modeling.
Kritzer (2002) classifies fee arrangements into six types:
1. Fixed fees specified in advance for routine work on tasks that are well
defined and predictable
2. Time-based fees in which an hourly fee is multiplied by the number
of hours worked
3. Task-based fees in which the lawyer charges fixed amounts for specific
subtasks, such as writing a letter, making a telephone call, and so on
(parallel to task-based physician fees)
4. Statutory, or other law-based, fee schedules based on the value of the
transaction or the amount in controversy
5. Commission-based fee arrangements in which the lawyer’s fee is
based on some percentage of the amount recovered or the value of the
matter being handled (e.g., in probate work, on the value of the estate
being handled)
6. Value-based fee systems, in which the lawyer assesses the value of the
work to the client and sets the fee on this basis
The most common approaches are the second, time-based compensa-
tion, and the fifth, standard contingent fee arrangements.
 Compensating Plaintiffs’ Attorneys 141

The Contingent Fee System

Historical Context The contingent fee system has had a long and some-
what checkered history in the United States. Until the late 1800s, there
was a common-law prohibition against contingent fees in many U.S.
jurisdictions; contingency fees were considered maintenance or cham-
perty, and were forbidden by law.9 This formally ended in 1884 when
the U.S. Supreme Court issued an opinion approving a contingent fee
contract in which the lawyer was to receive 50 percent of the award.10
Nevertheless, the adoption of contingent fees developed gradually in
the common law. State courts began recognizing contingent fees and the
state legislatures would take no action preventing this, resulting in the
integration of contingent fees into the legal system by individual states.
By the mid-1900s, most of the statutes preventing contingent fees had
been repealed. However, it was not until the 1960s that contingent fees
were allowed in all states; Maine was the last state to allow such fee
arrangements.11

Advantages of Contingent Fees There are several arguments for con-

tingent fees. First, they offer a financing arrangement for plaintiffs who
are too liquidity-constrained to pursue their cases because they lack the
internal funds and/or because banks are reluctant to provide loans for a
legal bill of a very uncertain size. Second, they may provide an efficient
mechanism for risk sharing, especially when other financing mechanisms,
such as personal insurance to cover legal expense, are unavailable.12
Contingent fee arrangements are not used by the defense. Unlike plain-
tiffs, the organizations bearing the financial risk for defendants are likely
to be insurers or large-self insured enterprises such as hospitals. These
organizations often have considerable ability to diversify away risk. In
this sense, contingent fees level the playing field.13 On grounds of being
able to bear risk, the financial risk bearers for the defendants may have
a greater self-interest in prolonging litigation than do most plaintiffs. But
for various other reasons, such as reputation loss, defendants may be
eager to settle.
Since attorney effort may be difficult for clients to observe, contingent
fees provide a method for coping with moral hazard when attorneys bill
142 Chapter 6

but expend little effort.14 Under the contingent fee system, attorneys are
paid for results, not just their inputs (that is, time expended on the case).
In fact, clients may achieve a larger recovery from a bifurcated (two-step
schedule) rather than single fee percentage in which the attorney gets a
larger fraction of the recovery if the case goes to trial.15 Since preparing
for trial generally involves substantial effort, the plaintiff’s attorney may
need an added incentive to be willing to make this investment.
This bifurcated approach contrasts with state-legislated fee caps which
place greater limits on fees for large recoveries. The contingent fee system
could encourage persons to file nonmeritorious cases since plaintiffs
incur no out-of-pocket expense if their case is lost. However, the attorney
lacks an incentive to accept such cases when the fee is paid only when
there is success in obtaining recovery.16

Disadvantages of Contingent Fees Contingent fees, as well as payment

by the hour, can lead to situations in which the welfare of plaintiffs and
of their attorneys may conflict.17 Two potential problems are particularly
noteworthy.
The first is for lawyers to settle cases too early. Lawyers may want to
settle rather than expend extra effort to obtain a larger amount. In fact,
some cases may yield (mostly modest) settlement offers with rather little
effort expended by attorneys. Although plaintiffs must consent to the
settlement of their claims, they often may not be able to gauge the extra
gain in their recoveries to be expected from extra time expended by the
lawyer on their claims.18 In fact, attorneys are retained in large part for
the expertise provided to clients. If the clients did not need the expertise,
they could pursue their legal grievances without the help of
attorneys.19
A countervailing influence is that attorneys who settle early will gain
a reputation for settling early for easy cash. Not only does reputation in
the community at large affect attorney demand, but so does reputation
in the community of attorneys—which would affect the number of refer-
rals an attorney receives from other attorneys. One would expect a trial
lawyer who consistently settles for limited amounts to eventually receive
ever smaller settlement offers. At least in some cases, plaintiffs’ attorneys
need to resist the temptation to settle early and hold out for higher settle-
 Compensating Plaintiffs’ Attorneys 143

ments and/or litigate some cases to verdict.20 In fact, the need for plain-
tiffs’ attorneys to appear tough is one reason that plaintiffs lose the vast
majority of cases litigated to verdict.21 On the other hand, critics of
contingent fees have argued that frivolous lawsuits can be profitable if
defendants prefer to settle such cases rather than risk a trial.22
By contrast, payment on an hourly rate may give attorneys an incentive
to prolong the case. Since lawyer compensation is not tied to the dollar
amount obtained for the client, lawyers paid on an hourly basis may
tend to exaggerate to their clients the chances of a large recovery at
verdict. Under hourly compensation, there may be an inclination to pad
bills and to allocate legal resources to cases at margins at which use of
such resources is unproductive.23 The trade-off seems to be between the
financial incentive to provide a realistic appraisal of the benefits of stop-
ping the case under the contingent fee versus an incentive to provide a
realistic appraisal of continuing litigation under hourly fee payment.24

Do Attorneys Paid a Contingent Fee Really Earn More than Other

Attorneys?
One strategy employed by advocates for contingent fee limits is to assert
that lawyers paid a contingent fee have inordinately high earnings. Most
often, such statements are made without reference to empirical evidence,
or the evidence presented is of low quality. Differences in compensation
by type of payment arrangement have often been exaggerated.
A case in point is the assertions advanced by Professor Lester Brickman
in his criticism of contingent fees. He asserts that since 1960, the effective
hourly rates of tort lawyers have increased 1000%–1400% in real dollars
while the overall risk of nonrecovery has remained unchanged.25 He
bases this assessment partly on average jury verdicts from the years 1960
and 2001: $132,000 and $1,454,800 (in 2004 dollars), respectively.
Unfortunately, Brickman relies on data from jury verdict reporters. The
vast majority of medical malpractice claims are settled before verdict. He
reports that changes in selection of cases that go to trial could explain
the dramatic increase in payments at verdict. Further, mean jury awards
may be substantially affected by a few high outlier awards. Thus, a better
comparison would be based on median rather than on mean jury
awards.
144 Chapter 6

These fees are much larger than those reported by Kritzer (2004),
which are based on data from a survey Professor Kritzer conducted in
Wisconsin. Results of the survey revealed a median hourly rate of $156
and a mean of $154 (2004 dollars; Kritzer 2004, p. 187). Taking his
data in combination with other data he reviewed, Kritzer concludes that
a useful range to consider is $156 to $174 per hour.26 He cautions,
however, that in computing hourly rates, one should deduct expenses
which would be billed separately by attorneys charging on an hourly
basis. Also, many lawyers do not maintain time records for their contin-
gent fee cases.27 Kritzer concludes that contingency fee lawyers, on
average, are paid an amount similar to defense attorneys with commen-
surate levels of experience.
Based on a 1988 survey of law firms, Aranson (1992) reports that
compensation of partners representing plaintiffs on a contingent fee basis
was 26 percent higher than for partners working for insurance defense
firms. Plaintiff attorneys’ mean hours of work were 12 percent lower
than those of defense lawyers. Thus, on an hours-adjusted basis, contin-
gent fee attorneys earned almost 30 percent more than those working
on an hourly basis for the defense. This may be an equilibrium differen-
tial generated by the market to compensate contingent fee lawyers for
assuming a risk-bearing role that lawyers paid per hour do not assume.
We do not know whether a 30 percent differential represents over- or
undercompensation for being risk bearers.
Daniels and Martin (2002) report results from a survey of plaintiffs’
lawyers in Texas. The mean value earned in contingent fee cases
was over $1 million. However, the median value was under $50,000.28
A team of Rand Corporation researchers conducted an indepen-
dent evaluation of effects of case management on dispute resolution
as part of the study, in which a survey was conducted of judges,
lawyers, and litigants.29 Most relevant to an analysis of fee arrange-
ments, the Rand authors conducted statistical analysis to determine
whether or not payment on an hourly versus a contingent fee arrange-
ment affected median days to case disposition, lawyer satisfaction with
management of the case, lawyers’ view that management of the case was
fair, and total work hours per litigant. No statistically significant differ-
 Compensating Plaintiffs’ Attorneys 145

ences in responses of hourly fee and contingent fee lawyers were

obtained.
A survey of 511 lawyers compensated on a contingent fee basis in
Wisconsin revealed that most cases that were not accepted were declined
because of lack of liability or lack of liability in combination with inad-
equate damages.30 Respondents to that survey indicated that they turned
away about half of the cases presented to them. This figure rose to 80–90
percent among attorneys who obtained a large share of their clients
though media advertising.31 Other studies reviewed by Kritzer32 suggest
that over half of inquiries are rejected by contingent fee attorneys. Brick-
man sets the figure at over two thirds.33 Brickman concludes that as a
result of this screening process, contingent fee attorneys are able to
recover nearly 100 percent of their out-of-pocket advances of litigation
costs, including the costs for unsuccessful cases.34
The same Wisconsin survey revealed that most lawyers in the study
(60 percent), charged a fixed 33 percent contingent fee; however, 31
percent employed a variable fee. For variable fees, the most common
pattern was a fee of 25 percent if there was no substantial trial prepara-
tion, rising to 40 percent if the case resulted in an appeal. There were
sometimes reductions in fees below the rate specified in the retention
agreement. Kritzer concludes that “At least in Wisconsin, the assertion
by contingency fee critics that there is a uniform contingency fee is clearly
false.”35 But as Brickman notes, relative to many if not most states,
Wisconsin is not a litigious state; as a result, data from this state may
be unrepresentative of contingent fee practices in more litigious
states.36
Overall, the quality of the evidence on whether or not contingent fee
lawyers do, in fact, earn more is not as good as we would like. At a
minimum, however, the evidence suggests that claims that attorneys as
a group make huge profits from contingent fees when representing indi-
vidual plaintiffs are exaggerated. Nevertheless, while the Wisconsin
survey results are probably representative of Wisconsin, there is consider-
able variability among states, and one should thus be careful not to
generalize to the United States as a whole from evidence from a single
state, even if it is “average.”
146 Chapter 6

State Policy and its Effects

Contingent Fee Limits and the Evolution of State Policy Following the
American Revolution, the opposition to federalism was strong, and led
to the repeal of caps on attorneys’ fees. Only in the mid-1970s, with the
first malpractice crisis, did political support for attorney fee statutes
reemerge in the legislatures of U.S. states, with some states enacting laws
and subsequently repealing them. For example, since 1976, Hawaii
enacted and repealed legislation regulating fee limits three times over the
course of a decade before settling on a much more lenient statute requir-
ing court approval of both parties’ counsel fees.37 Pennsylvania enacted
a statute in 1975,38 but in 1984 the Pennsylvania Supreme Court deter-
mined the statute to have been nullified by an earlier court decision.39
Oregon enacted a limit on all contingent fees in 197540 and repealed it
in 1987, replacing it with another statute which limits only attorney fees
stemming from punitive damages.41 Idaho enacted a cap on contingency
fees in 1975 with a sunset in 1981.42
New Hampshire’s Supreme Court ruled the state’s sliding-scale con-
tingent fee statute unconstitutional, and the legislature subsequently
repealed it.43 The Court held that the statute interfered with the freedom
of contract between a single class of plaintiffs and their attorneys: “It
does not regulate contingent fees generally; nor does it apply to defense
counsel in medical malpractice cases, whose fees consume approximately
the same percentage of the insurer premium dollar as do those of the
plaintiff bar.”44 The Court recognized that there were systemic problems
created by the contingent fee statute; the restrictions on contingent fees
unfairly burdened the malpractice plaintiffs and their attorneys. By regu-
lating only attorney fees in the medical malpractice arena, it made these
types of cases unappealing to the plaintiff bar.45 In its place, the legisla-
ture enacted a more lenient statute that requires court review of fees over
$200,000.
In 1980, Florida enacted a statute, repealed five years later, allowing
courts to award a reasonable attorney fee to the prevailing party in
medical malpractice cases.46 Attempting to curb rising costs associated
with a crisis in medical malpractice, the Florida legislature proposed an
amendment to the state constitution, filed with the secretary of state on
 Compensating Plaintiffs’ Attorneys 147

September 8, 2003. The amendment, titled “Claimant’s right to fair

compensation,” became effective after approval by voters in November
2004.47 Under the new amendment, when a contingent fee is involved,
a claimant in a medical malpractice case is entitled to 70 percent of the
first $250,000 recovered, and 90 percent of all damages in excess of
$250,000 recovered.48 This statute has simply reversed the wording of a
contingent fee limit—30 percent of the first $250,000, and 10 percent
of all the excess.
Florida’s approach to containing contingent fees was soon sidestepped
by plaintiffs’ attorneys. To avoid the limit, clients were asked to sign
waivers allowing their lawyers to take a higher fee. In 2006, the Florida
Supreme Court decided that any constitutional right can be waived, and
thus, the use of waivers was acceptable.49
In sum, states have been involved in limiting attorney fees for over
three decades. This legislation has largely arisen as a response to the
medical malpractice insurance crises. This implies that stakeholders rep-
resenting health care providers have been particularly active and effective
in assuring passage of statutes to limit attorney fees in litigation that
affects them most directly.

The Current Situation Existing rules governing fees generally treat

contingent fees as more worthy of regulation than other payment
methods—for example, the requirement that the terms of contingent fee
arrangements be in writing.50 The underlying assumption appears to be
that the state needs to provide more protection for clients when the
lawyer is paid on a contingency fee basis.51
Three types of regulations are currently used to limit attorney fees:
sliding scale,52 court review,53 and maximum percentage.54 Less than half
of the states use one of these restrictions, but almost all of the states that
do use restrictions have had their statutes in place since the mid-1980s
or earlier.
The sliding-scale limitation is employed by ten of the twenty-three
states with fee limits. The formulas limiting fees specify that contingent
fee percentages decrease as the award increases, implicitly assuming that
absent government intervention, lawyer compensation under contingent
fee arrangements is relatively excessive for large awards. Nevertheless,
148 Chapter 6

the formulas do not account for especially large jury verdicts. Thus, past
some threshold, the assumption appears to be that compensation does
not become more excessive.
To address this issue, eight states have enacted a process of court
review of attorney fees, which approves what is deemed to be a reason-
able fee.55 In states where judicial review is not mandatory, the right
must be affirmatively invoked; however, it is questionable whether a
typical plaintiff would contest a fee after months or possibly years of
litigation. As enacted, fee-capping statutes are not likely to be binding.
Most states with these statutes cap fees at the standard contingency fee
of one third of the recovery, but one state allows up to 50 percent of the
award to go to attorneys’ fees, a percentage that is well above the usual
contingency fee agreement.
As mentioned before, in Florida not only did limits recently win legisla-
tive approval, but they were voted upon by the electorate in 2004 as
well. The voters in Florida presumably believed the arguments of the
advocates for these tort reforms. Finally, as with Florida’s no-fault
program for neurologically impaired infants,56 lawyers are able to pursue
their financial incentive to sidestep the intent of the reform.

Empirical Evidence of Contingent Fee Regulation

An early study by Danzon and Lillard (1983) estimates the effect of
restricting contingent fees on settlement rates. They find fee restrictions
were associated with a 1.5 percentage point increase in the settlement
rate. Studies by Sloan (1985) and Danzon (1986) find no effect of attor-
ney fee regulation on medical malpractice premiums, claims frequency,
or mean payment per paid claim.
More recently Helland and Tabarrok (2003) used two databases,
one cross-sectional57 and the other time-varying,58 to investigate con-
tingent fee statutes. Their study aimed to find the effects of contingency
fee statutes, both those requiring judicial review and those imposing
limits, on the probability that a case is dropped, as well as the time to
case resolution. The underlying theory is that measures that make
contingent fee payment less attractive to attorneys will lead to greater
attorney reliance on pay by the hour. The probability that a case was
dropped is taken as an indicator of the legal quality of the case. Whether
 Compensating Plaintiffs’ Attorneys 149

or not a lawyer was willing to accept a case on a contingent fee basis is

interpreted as a signal of whether the lawyer thinks the case is meritori-
ous. In contrast, payment on an hourly basis provides an incentive for
lawyers to take cases, irrespective of their legal merit. Further, lawyers
paid on an hourly basis have a greater incentive to delay case
resolution.
Results from both databases indicate that the probability of a case
being dropped increased following imposition of statutory constraints
on contingent fees; the time to resolution also increased. The authors
conclude that “Limits on contingency fees were supposed to prevent
people with poor information from paying too much for legal representa-
tion. The results of this study cast doubt on whether this goal was
achieved. Our results indicate that, as in other areas, restrictions on the
freedom to contract have unintended consequences.”59
The fact that dropping of cases increases following enactment of con-
tingent fee limits may not have been anticipated by proponents of statu-
tory fee limits. But the result makes perfect sense. Once an attorney
receives the bad news about the prospects of winning a case, possibly as
a result of exchange of documents during the discovery process or from
depositions, the lawyer compensated on a contingent fee basis has an
incentive to recommend that the case be discontinued.60 Under an hourly
fee arrangement, lawyers do not have this incentive. Overall, limits on
contingent fees charged by attorneys appear to have none of the effects
sought by the proponents of these statutes—namely on the cost of
medical malpractice insurance to physicians—most likely because the
limits were not binding constraints on such fees.

Fee-Shifting Rules: The Loser Pays All Attorneys’ Fees

Historical Background
In the United States, the common-law rule, and the usual practice, is that
each side assumes its own attorneys’ fees irrespective of the outcome of
the legal dispute. However, the United States has some experience with
fee shifting wherein the loser is responsible for all litigation cost, also
known as the English Rule.61 As a new colony, the United States followed
the English Rule of fee shifting. The English Rule was in practice until
150 Chapter 6

1789, when Congress enacted legislation authorizing federal courts to

follow state law concerning fee awards.62 Despite this legislation, the use
of the English Rule by the states continued until the U.S. Supreme Court
specifically addressed fee shifting in a 1796 opinion. This case involved
an award of attorneys’ fees which were challenged; the Supreme Court
ultimately struck the award of attorneys’ fees. Because there were no
federal statutes on the issue during the period 1800–1853, and all the
statutes that allowed the recovery of attorneys’ fees had expired by 1800,
federal courts used state statutes on attorneys’ fees. As a result, the
English Rule was frequently applied. An 1830 decision by the Supreme
Court,63 despite precedent and relevant statutory authority, held that the
law did not limit the power to award attorney fees; this power was in
the Court’s discretion.64 Subsequently, Congress enacted the 1853 fee
bill that prevents collection of attorneys’ fees from a losing party. The
rule that developed from the bill was named the American Rule. This
new rule called for each litigant to pay its own attorneys’ fees, regardless
of the success of the parties.

Exceptions to the Use of the American Rule in the United States

The most comprehensive fee-shifting statute in the United States is in the
state of Alaska. This state has had a fee-shifting statute since the mid-
1800s, before Alaska even became a territory of the United States.65
Alaska’s legislative and judicial development partly explains its contin-
ued use of the English rule. After the United States purchased Alaska in
1867, seventeen years passed before Congress gave Alaska instructions
for a civil government. In 1884, Congress declared Alaska to be a civil
and judicial district, and the general laws of Oregon would apply.
Oregon’s statutes allowed one-way fee shifting for the plaintiff, including
witness fees, court fees, deposition expenses, and costs relating to the
preparation of documents used as evidence at trial. This law was in effect
until 1900, when Congress passed a code of civil procedure for Alaska.
This new code had a provision for attorneys’ fees similar to the Oregon
statute. In 1949, Congress amended the statute designating the civil
procedure code so that the Federal Rules of Civil Procedure would apply
to the U.S. District Court in Alaska. However, as Congress had not
 Compensating Plaintiffs’ Attorneys 151

repealed the statute granting jurisdiction over attorneys’ fees to Alaska,

and the Federal Rules were silent on the matter, Alaska kept the fee-
shifting practice.
Alaska’s rules evolved over the next several years to establish a more
reasonable fee schedule and grant courts the discretion to award fees
when cases fell outside the parameters of the fee schedule. After 1959,
when Alaska became a state, the former territorial laws and court rules
were codified in the newly created Alaska code. Rule 82, which still
applies, was adopted into the Alaska Rules of Civil Procedure.66 This
rule set forth schedules for trial courts to award attorney fees to the
prevailing party. Over the next several decades, the rules were repealed,
rewritten, and amended, and the Alaskan Supreme Court has heard
many challenges to Rule 82. A Supreme Court subcommittee has eva-
luated the rule twice; on both occasions, the Court has chosen to leave
the rule in effect. Substantial revisions were made the last time Rule 82
was reviewed (1992).
Rule 82 gives fixed percentages of actual reasonable attorneys’ fees
based on length, degree of difficulty, bad conduct, attorney conduct, and
a few other considerations. The prevailing party always receives attorney
fees as long as a motion is filed within ten days from the date shown on
the certificate of distribution on the judgment.67 However, the parties
can agree to waive the fee requirement. Attorneys in Alaska have said
that the motion practice for Rule 82 is routine, and simply another part
of the litigation process.68

Empirical Evidence on Effects of Switching from the American Rule to

the English Rule
There is some empirical research on the effects of changing to the English
Rule. This rule should reduce claiming since it places a presumably risk-
averse plaintiff under substantial financial risk. Hughes and Snyder69
investigate the effects of changing to the English Rule, using data on
Florida closed medical malpractice claims. They find claims were more
likely to be dropped under the English Rule, suggesting that weak claims
are less likely to be pursued. Those claims that were pursued were
more likely to be litigated. In general, they found that the English Rule
152 Chapter 6

encouraged plaintiffs with strong cases to pursue their claims but dis-
couraged those with weak cases from doing so.70 Interestingly, the Florida
Medical Association, which had favored the English Rule initially, sought
its repeal because, in practice, plaintiffs often did not have the funds to
pay the winner.71
Alaska’s experience with the English Rule has been evaluated by Di
Pietro and Carns.72 A deficiency with that study is that there was no
control group, although some comparisons were made.
The study reaches three conclusions. First, the English Rule discour-
ages some middle-class persons from suing. Second, it discourages pursuit
of cases of questionable merit on both plaintiff and defense sides. Third,
it encourages litigation in strong cases that would otherwise settle. While
wealthy persons presumably can afford to pay the winner’s fees, collec-
tion from the poor is infeasible, so the English Rule would have no effect
in such cases.
The authors also report general satisfaction with the rule among attor-
neys in Alaska. However, in the vast majority of cases, the amounts of
fees assessed the losing party were small (well under $10,000) and,
interviews of knowledgeable persons in the state suggested that the
English Rule did not affect plaintiff claiming.
There is a 12 percent difference between Alaska’s civil court filings
and those for states without fee shifting. In 1992, the national median
number of court filings was 6,610 per 100,000 population, compared
with Alaska’s 5,793 per 100,000, a difference of 810 lawsuits per
100,000 persons.73 In one sense, a 12 percent differene is meaningful.
Yet, we do not know anything about what types of cases were reduced
(e.g., whether or not cases involving low-income injury victims were
disproportionately affected).
Overall, although the English Rule has the potential to benefit premium
payers, it places injury victims at such a substantial disadvantage that
unless it is implemented as part of a reform package which gives some
important benefits to victims, it has virtually no chance of passage except
on a very limited scale. And if it is enacted, the Florida experience is
likely to be repeated. On balance, fee shifting seems worse than the
disease it is trying to cure.
 Compensating Plaintiffs’ Attorneys 153

Another Type of Reform of the Contingent Fee System: Early Offers

An additional approach is to limit the fees in cases in which the plaintiff

accepts an early offer from the defendant(s). Jeffrey O’Connell and
Patrick Bryan74 develop and advocate such a proposal they label “early
offers.” Under this plan, the defendant has up to 120 days after the filing
of a suit to offer a no-fault-like periodic payment of a claimant’s net
economic loss.75 In later variants of the proposal, the period is shortened
to sixty days.
Acceptance of the offer forecloses the plaintiff’s right to pursue a sepa-
rate tort claim for noneconomic losses. The plan specifies that an early
offer will compensate the plaintiff for all costs, including medical and
rehabilitation expenses and wage losses not covered by collateral sources.
The plan also encourages defendants to apologize for errors, something
that does not currently occur because such statements may be used
against the defendant in court.
In its 2004 version the plan required that contingent fees not be
charged against settlement offers made prior to plaintiff’s retention of
counsel. If the offer is accepted by the plaintiff, the attorney’s fees are
limited to hourly rate charges and are capped at 10 percent for the first
$100,000 of the offer and at 5 percent for any amounts in excess. There
is a provision requiring notices of claim submitted by plaintiff’s counsel
to include basic, routinely discoverable information designed to assist
defendants in evaluating plaintiff’s claim, and conversely for defendants
to provide such information to plaintiffs. When plaintiffs reject defen-
dants’ early offers, contingent fees may be applied against net recoveries
only in excess of these offers. Finally, if no offer is made within sixty
days, the early offer plan does not affect the attorney’s contingent
fee.76
Early offers have several potential advantages. The plan seeks to cut
time elapsed from the filing of a suit to the time it is resolved. By accep-
ting an early offer, the plaintiff can benefit from speedy compensation
at the cost of being foreclosed from suing in tort for nonmonetary loss.
Defendants save time and the aggravation of a lengthy legal process, and
third-party insurers save legal expense and payments for nonmonetary
loss as well as the unpredictability of some claims. The early offer
154 Chapter 6

includes the plaintiff’s attorney fees, based on a reasonable hourly rate,

which is substantially less than the customary range of contingent fees
described above. Some medical malpractice plaintiffs state that learning
what happened is a major motivation for suing,77 and thus plaintiffs may
be better off from receiving both an explanation and an apology. At the
same time, by not permitting contingent fees during settlement offers
made prior to plaintiff’s retention of counsel, the possibility of offers
soon after providers discover an error may be increased. The provision,
which disallows contingent fees on the early offer if the offer is rejected,
is meant to dissuade attorneys from convincing their clients to reject early
offers.
The plan is ostensibly designed to help the injury victims who are said
to be victimized by trial lawyers under the contingent fee system. That
the plan increases the well-being of injury victims is debatable. The basic
argument for the plan is that it would cut losses, in terms of both legal
expense and compensation, although there would be benefits in terms of
speed of claim resolution and the lower amount of the early offer taken
by the attorneys’ fees.
The plan is likely to make it harder for people to obtain legal repre-
sentation even if they have valid grievances. Proponents of the plan speak
of the windfall trial lawyers currently enjoy. However, evidence of client
dissatisfaction is at best ambiguous; if clients truly are disappointed with
their attorney and fee levels, they do have recourse to appeal; the judge
can overturn or reduce the fee. Further, a survey of medical malpractice
claimants in Florida revealed that the vast majority of claimants, whether
or not they received compensation, were satisfied with the system.78 This
evidence is from one state and only for medical malpractice claims involv-
ing birth injuries and injuries in emergency rooms. Thus, generalizability
is an issue. But evidence based on systematically collected data suggests
that clients are generally satisfied with the dispute resolution process.
The notion of widespread windfalls assumes that the cost incurred by
lawyers in making initial evaluation of the legal merits of a claim is
negligible. This may be true in some cases, but definitely not in others.
Advocates for the proposal can argue that defendants are precluded
under the current contingent fee system from making early offers of
settlement because this would signal a weakness in their case and would
 Compensating Plaintiffs’ Attorneys 155

certainly be rejected. If no early offer is made, defendants are in the same

position as before.
Litigation cost currently absorbs a very high share of the medical
malpractice premium dollar. Thus, any plan that would reduce such cost
would, other factors being equal, reduce the overhead of receiving com-
pensation. This is a plus. Another argument advanced by advocates of
early offers is that it will clear the court dockets by encouraging early
settlements. In turn, this would reduce the “excessive” costs of the
current tort system.79 However, it is not clear that the court dockets are
clogged with personal injury claims, and certainly not with medical
malpractice claims, which constitute only a very small proportion of total
personal injury claims.
In the end, this proposal leaves us with more questions than answers.
What should be done about poorer victims who incur insubstantial
monetary loss when injured? What about victims with early offers that
do not adequately compensate their damages, but whose case would not
survive the higher burden of proof required for trial?

Replacing the Contingent Fee System: Legal Expense Insurance

Overview
A more promising alternative to the current contingent fee system may
be legal expense insurance. If people have insurance coverage for legal
expenses, a large part of the rationale for the contingent fee system would
no longer apply. People would have access to legal representation without
having to grant the plaintiff’s attorney a large share of any recovery. On
the other hand, the experience with other forms of first-party insurance
has not been that good. For one, health insurance distorts individuals’
incentives to use medical care wisely.

Legal Expense Insurance in Europe

Legal expense insurance has been used for decades in Europe.80 Thus,
the experiences in these countries provide a basis for evaluating the
shortcomings and strengths of prepaid legal service plans. Currently in
Europe there are four types of legal expense insurance (LEI): after the
event (ATE); before the event (BTE); stand-alone; and add-on policies.
156 Chapter 6

While it seems that ATE would save a claimant the expense of insurance
without losing the benefits, insurers offer ATE only if the claimant’s
chance of success is high.
In England, most policyholders use BTE, with a small but growing
segment of the market using ATE. Most of the policies in England are
add-on policies; toward the end of the 1990s, the Association of British
Insurers was aware of only one personal stand-alone product on the U.K.
market.81 By contrast, in Germany, stand-alone BTE policies are the
dominant form. These policies allow their holders to mix and match the
areas of law they prefer to have covered, based on their needs.82 England
also has commercial policies, a form of insurance that does not exist in
Germany. These policies are sold on a stand-alone basis and account for
one third of the LEI market’s gross premiums.83
About 42 percent of households in Germany are covered by legal
expense insurance.84 Such insurance constitutes only a small share of the
total insurance market; in Germany, where legal expense insurance is
prevalent, it accounts for only 6 percent of insurance premiums, com-
pared with 46 percent for auto.85 Nevertheless, German LEI provides
coverage for 3.6 million cases annually and over *1.5 billion in lawyers’
fees, making up a quarter of all the fees earned by German lawyers.86
Mandatory use of fee schedules for attorneys’ fees is only one element
of public regulation. Germany and Switzerland require that LEI be
offered only by specialized insurance companies. As of 1986, there were
thirty-two such companies in Germany and twelve in Switzerland. Europe
as a whole, not including Scandinavia, had a total of ninety-five specialty
companies; and, over 600 other insurance companies provided some
form of LEI.87 In Germany, in-house lawyers are not allowed, ceding
private attorneys control over all private causes of action. Also, Germany
does not allow contingent and conditional fees; any fee that is outcome-
determined is not permitted.
Proponents of LEI argue that a well-developed LEI market removes
pressure for government funding of programs such as legal aid or other
organizations that focus their efforts on indigent clients. In Germany,
about eight times as much was spent on LEI as all the federal German
states combined spent on legal aid. Since such insurance covers tort liti-
gation, access to legal representation may be enhanced in this sense.
 Compensating Plaintiffs’ Attorneys 157

However, less than half of German households have legal expense

insurance. This could mean that the other half can afford legal assistance
at their own cost, or it could mean that a vast majority of German citi-
zens do not have access to the legal system. Of the small amount spent
on legal aid by the government in Germany, 80 percent is spent on family
law cases, an area not covered by legal expense insurance. By contrast,
in England and Wales, where LEI is relatively new and not as heavily
relied upon, governments spent twenty-eight times as much per capita
on legal aid than the population spent on LEI premiums.88
Legal expense insurance has achieved widespread acceptance in Europe.
Such insurance has the advantage of providing risk protection against
high legal losses without resorting to contingent fees. However, there are
important differences between the United States and Europe, especially
in acceptance of government regulation and in particular in government’s
role in setting fees. Some of the issues that have plagued health insurance,
discussed above, may also plague its legal counterpart. However, the
technological change that underlies much of the growth of personal
health spending is not present in the legal context.

Legal Expense Insurance in the United States

Prepaid legal service plans (LSPs) have had a challenging beginning in
the United States. At the beginning of the twentieth Century, both the
American Bar Association and many state bar associations had strict
ethical rules which prohibited a lawyer’s involvement with group legal
services, parallel to the objections against prepaid medical service plans
by the medical establishment.89 However, this opposition did not totally
preclude entry of some group legal service plans. In 1930, the Brothe-
rhood of Railroad Trainmen established one of the very first group legal
service plans.90 Its activities gave rise to a series of cases which ended in
1964 with Brotherhood of Railroad Trainmen v. Virginia ex rel. Virginia
State Bar.91 In this case, the U.S. Supreme Court found that Virginia
statutes prohibiting group legal services were a violation of the First and
Fourteenth Amendments. One year prior to this decision, the Supreme
Court held that the National Association for the Advancement of Colored
People (NAACP) had a constitutionally protected right of political asso-
ciation, which gave its attorneys the right to bring civil rights suits on
158 Chapter 6

behalf of the members, setting the precedent for group legal service
plans.92
Nevertheless, criticisms from the legal community continued after the
U.S. Supreme Court decisions. The concerns centered on one main
issue—conflict of interest.93 The conflict of interest is twofold, encom-
passing both confidentiality issues and interference with independent
professional judgment. Both of these issues stem from the involvement
of a third party, the legal service organization. Confidentiality issues are
raised because legal service organizations want to know the extent of
clients’ issues and to be involved in deciding what the course of action
should be. Concern about independent judgment arises because, with
insurer involvement, attorneys may not act in the best interests of their
clients. These concerns were addressed by a series of revisions to the
Code of Professional Responsibility, American Bar Association opinions,
and local bar opinions. Since 1983, the ABA has officially encouraged
development of legal service plans.
Interestingly, the same confidentiality and conflict of interest issues
have arisen in the context of health insurance in general and in managed
care in particular. These issues provide much of the reason for the
managed care backlash. Without some sort of insurer intervention,
however, moral hazard becomes a major problem in contexts of legal
and health insurance. The same issues arise on the defense side. Third-
party insurance covers litigation expense as well as payments to claimants.
The interest of the insurer and the defendant may often be in conflict.
Contemporary LSPs function much like preferred provider organiza-
tions (PPOs). Monthly premiums are paid and plan benefits limit out-of-
pocket expense, particularly if the covered person uses a lawyer in the
network.94 Legal service plans can be barebones or comprehensive in
their coverage.
Legal insurance does not provide adequate coverage for most personal
injury cases. For medical malpractice actions, few, if any, legal service
plans cover the amount of preparation and trial time required for litiga-
ting a claim. Some more comprehensive plans cover this type of litiga-
tion, but even those put a cap on the amount of time spent on the
case—past this limit, the plaintiff is responsible for the charges, albeit at
a discounted fee.
 Compensating Plaintiffs’ Attorneys 159

Differences Between Legal Expense Insurance in Europe and in the

United States
There are several important differences between the European legal
expense insurance and the legal service plans in the United States. These
differences affect the functionality and accessibility of group legal ser-
vices in the United States. First, LSPs do not insure against monetary
loss; rather, they allocate a preset number of lawyer’s hours to the
insured. Even though risk is curbed by limits on the number of hours,
attorney fees in the United States vary markedly, which complicates
LSPs’ risk calculations. Public regulation of fees in Europe simplifies this
aspect of the risk calculation.95 LSPs typically fund the expenses of liti-
gation only for a defendant. An ordinary plan allows a member to have
a limited amount of funding for consultation with an attorney, usually
over the phone, and basic legal services, such as will drafting. Neverthe-
less, even though the most comprehensive plans limit and/or exclude
coverage, most comprehensive plans are structured so that a typical
middle-class family is likely to be covered for 80–90 percent of personal
legal services they may need to use.96 In general, there is no requirement
that a specific event occur (such as a claim filed against the insured) for
the policy to take effect. Members may take advantage of routine services
and consultation immediately, in contrast to LEI plans.
Perhaps the most important difference between LSPs and LEIs, and
the largest obstacle to their growth, is the fact that LSPs are usually set
up as an employee benefit trust or as a union benefit.97 Given the growth
of the cost of other fringe benefits, expanding reimbursements to cover
still another type of benefit has been highly problematic. In contrast, LEIs
are directly financed by households.

Is Legal Insurance a Desirable Alternative to the Contingent Fee

System?
First-party legal insurance coupled with payment of plaintiff attorneys
could indeed be an alternative to contingent fees. If it is provided on a
group basis, adverse selection could be largely avoided. But employers
are likely to resist a mandate that they provide such coverage. Indivi-
dually purchased legal insurance would be potentially subject to adverse
selection.
160 Chapter 6

The other major problem plaguing insurances of all types is moral

hazard. With nearly complete or complete insurance, insured individuals
would use legal services at greater rates. To combat this, insurers would
need to impose limits on utilization of such services or otherwise impose
some type of active case management. This would in turn result in
conflict of interest issues discussed above; moreover, the limits may be
too restrictive to allow lawyers to actively pursue a complex medical
malpractice claim.
It matters from which point one starts. If the United States had started
with European-style LEI in the modern era, it would be relatively easy
to continue along this path. Yet contingent fees are so entrenched in the
U.S. culture, especially in the realm of personal injury, that changing
course at this stage would be hard to do.

Discussion and Conclusions

Some who would reform the contingent fee system fall into the same
trap as previous reformers. They presume to know how the system
works, but that presumption is based largely on a single case or even
atypical cases that capture media and public attention. Not surprisingly,
the direct beneficiaries of proposed changes in methods of compensating
lawyers rush to support the statutory changes that they like. We agree
with Herbert Kritzer,98 who concludes, “I do not presume that there are
never situations in which lawyers take advantage of clients to obtain fees
that raise equity and ethical issues. However, abuses of this type are
confined neither to lawyers working on a contingency fee basis99 nor to
members of the legal profession.” And Haltom and McCann note:
A public image that includes financial self-interest hardly separates tort
lawyers from other business persons, stock traders, and almost any powerful
figure in American society. . . . In short, the public assessment is contradictory;
lawyers are respected and distrusted for the same aggressive client
representation.100

This is not an excuse for overlooking injustices and inefficiencies when

they occur. However, to the extent that principal/agent problems are rife
when consumers deal with professionals, this is worthy of a more general
examination than a limited focus on personal injury attorneys and
 Compensating Plaintiffs’ Attorneys 161

attorneys who represent claimants in medical malpractice cases in

particular.
For those who would see imposing limits on attorney compensation
as a method for reducing the high cost of medical malpractice, empirical
evidence has provided no support for the claim that limits in fact lead
to savings. The evidence suggests that application of the English Rule
could accomplish this goal, but Florida’s experience with the English
Rule is instructive.
The system currently in place in the United States is a blend of the
English and American rules. It provides access to the courts to people who
must rely on contingent fees, and it also allows parties who desire fee
shifting a statutory basis for asking the court for attorneys’ fees, especially
in cases where there is abuse of the legal system. With over 1,200 statutes
(to date) allowing fee shifting in this country, the American Rule in effect
allows fee shifting. There is a strong argument that virtually all of the
plaintiffs and defendants we want to protect qualify for one exception or
another. Those who do not qualify for an exception are those whom most
people think of as typical litigants; they have a potentially legitimate
problem or issue, and they do not abuse the system in order to obtain
compensation. These are not the persons who should be deterred from
filing. What the proponents of the English Rule should be arguing for is
more frequent or expanded use of existing exceptions.
In the United States, the people—through their elected officials—decide
which cases will permit fee shifting, not the courts. Furthermore, the
English Rule is applied in an entirely different legal context in England.
In that country, judges, not juries, hear personal injury cases. Changing
this factor alone would satisfy those who claim there are “outrageous
jury verdicts.” But, Americans see jury trials in civil cases as their
right.
As in England, when the defendant in the United States is an organiza-
tion rather than an individual, the defendant tends to have considerably
greater resources and knowledge than the plaintiff. The defendant has
the resources to delay settlement and force a case to trial.—Under the
English Rule and combined with the high rate of success at verdict for
the defense in medical malpractice disputes, the result for an unsuccessful
plaintiff would be crushing attorney fees.
162 Chapter 6

In practice, most legal disputes in both the United States and England
are settled before trial. However, faced with the possibility of paying the
attorney fees of an opponent, many potential victims with legitimate
claims will never bring their claims; it is too large a risk for a plaintiff
with anything but a “sure-thing” case.
In regards to first-party insurance for legal expense, the concept is
intriguing, but is unlikely to be implemented as a substitute for contin-
gent fees in the United States. As with no-fault, discussed in chapter 11,
historical context can be critical to the success of a change. While legal
expense insurance has been successful in Europe, it faces special chal-
lenges in the United States. In particular, coping with moral hazard and
adverse selection would lead to many controls that both lawyers and
individual policyholders are likely to resist. Many of the issues facing
health insurance would likely be repeated, including how to provide
financial support for claims filed by persons without legal insurance.
Another issue that arises is the various cost containment issues when
personal out-of-pocket expense falls far short of total expense.
In the end, it seems that the current system for paying attorneys may
be as good as any of the other alternatives. As the saying goes, “If it ain’t
broke, don’t fix it.” If there are to be changes in medical malpractice
law, it is best to look elsewhere.
Nevertheless, one aspect that could be “broke” is the lack of a com-
petitive market for plaintiffs’ attorney services. However, rather than
implement regulatory interventions, or such proposals as early offers,
which would at best address part of the issue of lack of competition in
the market for attorneys’ services, a better course would be for antitrust
authorities to investigate whether or not there are impediments to price
competition in this sector. Such an investigation should not be limited
to contingent fee arrangements, but should investigate price fixing in
hourly fee arrangements and referral agreements as well.
7
Juries and Health Courts

Introduction and Overview

Several articles on juries start “The American jury is on trial” or an

equivalent statement.1 While this statement applies much more generally
than only to medical malpractice, one criticism of juries is that they are
ill-equipped to decide complicated issues of causation and duty of care.
Not only is it said that juries lack expertise to deal with the technical
issues in many medical malpractice cases, but in participating in a single
trial, individual jurors lack the experience that they would accumulate
by participating in many trials involving medical malpractice. Jurors’
ability to process the technical information presented at trial is certainly
not helped by the lawyers and attorneys representing both sides who
cannot agree on the applicable scientific principles themselves.2 Even if
they are not limited in technical expertise and experience, jurors are often
said to be sympathetic to plaintiffs. Their interpretation of the facts in
this light3 may cause them to view litigation as a vehicle for redistributing
wealth from wealthier defendants to less affluent plaintiffs. Some have
also alleged race and gender bias.4
In the United States, tort cases proceed through several stages. Most
simply, the stages are determination of liability, compensation (especially
nonmonetary compensation), and punitive damages. The first stage is
the trial stage, where the plaintiff must show proof of injury and proof
that defendant’s negligence was the cause in fact of the injury. After
successful completion of this stage, the trial moves to the compensa-
tory stage, in which the court decides the amount of compensation
necessary to make the plaintiff whole. In a third stage, the court may
164 Chapter 7

ask whether punitive damages are warranted, that is, whether or not
the defendant’s conduct was sufficiently malicious, reckless, or careless
to warrant an additional penalty designed to deter future behavior of
the sort presented at the trial, and if it is, the appropriate amount of
punitive damages.5
Criticism of juries is not new. Ever since juries or jurylike organiza-
tions have existed—in ancient Rome, in eighteenth-century England, and
in the early twentieth-century United States, they have been attacked for
their incompetence.6 American juries have been the subject of criticism
at all three stages. Jury sympathy for the plaintiff may spill over into
sympathy in setting compensatory and punitive damages.
Much of the literature focuses on punitive damages in product liability
cases, where such damages are fairly common, especially in the high-
profile, highly publicized cases. Punitive damages are now very uncom-
mon in medical malpractice.7 However, they may become more common
in the future, especially if caps on nonmonetary compensatory loss
become binding constraints on payment to medical malpractice
claimants.

“Runaway Juries”

One often hears about “runaway juries.” Quotations, such as the fol-
lowing, from the front page of the Wall Street Journal, are common:
“The real sickness is people who sue at the drop of the hat, judgments are
going up and up and up, and people getting rich out of this are the plaintiffs’
attorneys,” says David Golden of the National Association of Independent
Insurers, a trade group. The American Medical Association says Florida,
Nevada, New York, and Pennsylvania and eight other states face a “crisis”
because the “legal system produces multimillion-dollar jury awards on a regular
basis.”8

The media provide a biased sample of jury verdicts. A disproportionate

number of extremely high verdicts are discussed by the media.9 Although
high verdicts receive a substantial amount of publicity, many high awards
at verdict are subsequently reduced, either by the judge, on appeal, or
by a postverdict settlement.10 These reductions are much less likely to be
highly publicized. A study by the Rand Corporation finds that about 70
 Juries and Health Courts 165

percent of amounts awarded at verdict are actually paid. For punitive

damages, slightly less than 60 percent of the award is paid.11
Although concerned not with medical malpractice but tobacco litiga-
tion, John Grisham’s fictional account summarizes what many believe to
be nonfiction.12 (Parenthetically, some of the major plaintiffs’ attorneys
in tobacco litigation specialized in medical malpractice litigation before
becoming involved in lawsuits against tobacco companies.)
In medical malpractice cases, the defendant is typically a physician,
not a tobacco manufacturer. The alleged plot does not typically involve
a particular juror, but rather jurors more generally, who may be well-
intentioned, but whose decision-making is allegedly not governed by
medical knowledge or legal principles, but rather by sympathy for the
injury victim.
Compared to the large number of anecdotes, there are a small number
of in-depth empirical studies of juror behavior, most of which do not
specifically deal with medical malpractice. The empirical evidence is
mixed, suggesting some support for perhaps a partial runaway jury.

Empirical Evidence on Juror Behavior in Determination of Liability

The Percentage of Cases Won by Plaintiffs at Trial Is Low

A review of thirteen studies of verdicts in medical malpractice cases
revealed that the median plaintiff win rate for the studies taken as a group
was 29.6 percent.13 In a recent review of civil trials disposed of in state
courts in the seventy-five largest counties in the United States in 2001, the
plaintiff win rate in medical malpractice cases was nearly 27 percent.14
These low win rates do not eliminate the possibility of jury bias in favor
of plaintiffs, but they tend to place an upper bound on such bias.

Nontechnical Nature of Many Medical Malpractice Disputes

While the conventional wisdom is that medical malpractice claims involve
much technical material beyond the comprehension of juries, many
medical malpractice disputes are in fact not technical. Vidmar presents
brief summaries of the issues in a sample of verdicts from North Carolina
in the mid-1980s in which the verdict favored the plaintiff.15 These issues
included surgery on the wrong foot, permanent paralysis and brain
166 Chapter 7

damage following drug overdose administered to the patient after heart

surgery, and a patient receiving ten times the prescribed dosage of che-
motherapy, which resulted in nerve damage, paralysis, organ damage,
and loss of sexual function.
One can also find examples in which technical details are central to
the legal dispute, such as West v. Johnson & Johnson,16 a medical
product liability, not a medical malpractice, case. The medical issue was
whether Johnson and Johnson’s tampon was a defective product. Sugar-
man uses this case to argue that jurors selected at least in part for their
ignorance about the topic are asked to decide extremely difficult scientific
issues.17 He would presumably have the same opinion of choice of jurors
in complex medical malpractice litigation involving scientific issues.

Empirical Evidence on Myth 3: Dispute Resolution Is a Lottery

The above thinking about limitations of jurors leads to the inference that
outcomes of litigation are often random. There is a body of empirical
evidence, however, that leads to the opposite conclusion: that outcomes
are not random. Even so, this does not imply that juries never make
mistakes.
Taragin et al. compare outcomes of claims, whether or not the dispute
ended with payment to the claimant, with assessments of physicians
employed by an insurance company, the New Jersey Medical Inter-
Insurance Exchange.18 The physicians were asked to review the medical
records and provide a neutral evaluation of whether or not they believed
that the named defendants in specific cases were negligent. The raters’
decisions were confidential and could not be presented as evidence at
trial.
Most claims were “defensible,” meaning that the raters found no
negligence. The vast majority of defensible claims were resolved before
discovery. Of those defensible claims that reached the trial stage, the vast
majority resulted in no payment for the plaintiff. By contrast, among
indefensible cases that reached trial, of those that did not settle, virtually
all resulted in payment. About half of these cases resulted in payment
when the case was decided on appeal.
In sum, there is substantial consistency between assessments of physi-
cians working for the insurer and outcomes of the claims, both those
 Juries and Health Courts 167

settled prior to verdict and those decided at verdict and on appeal.

Moreover, judgments for the plaintiff and the severity of the plaintiff’s
injury were unrelated, contradicting the view that jurors are unduly
sympathetic to plaintiffs who sustain serious injuries. The authors’ con-
clusion is that “unjustified payments are probably uncommon.”19
Farber and White report similar results. Their data consisted of medical
malpractice claims filed against a single hospital, which the authors did
not identify.20 As in the Taragin et al. study (1992), physicians assessed
case-specific quality of care. Overall, the authors find strong evidence
that care quality affects the outcomes of medical malpractice disputes.
The hospital’s expected liability for damages was twenty-five times higher
when negligence occurred than when “good care” had been rendered.
The data covered all claims against the hospital, not just jury verdicts.
Sloan, Githens, Clayton, et al. compare independent physicians’ assess-
ments with outcomes from medical malpractice disputes.21 In 1986 and
1987, the authors surveyed 187 plaintiffs in Florida who had filed
medical malpractice claims stemming from permanent birth-related inju-
ries or injuries in a hospital emergency room. The survey obtained
information on the respondents’ clinical histories prior to the injury that
led to the claim, events that led the plaintiff to sue, information on uti-
lization of personal health services and other data used to estimate past
and future monetary loss for each injury, and plaintiff satisfaction with
the legal process. In addition, permission was obtained to give the authors
access to the respondents’ hospital records.
Medical evaluations of liability were conducted by a panel of physi-
cians. The ratings were systematically and significantly related to the
outcomes of the cases. Among cases dropped by plaintiffs without
payment, the physician raters were three times more likely to have found
no liability than to have found the defendant to be liable. Among cases
settled with payment, the ratio of not liable and liable cases was reversed.
For cases decided at verdict, the liability ratings were related to the trial
outcome. When the plaintiff lost at trial, the independent raters were
more likely to have found no liability, and the converse was true for
those cases won by plaintiffs. The relationship between the raters’ find-
ings and the case outcomes was not perfect. However, case outcomes
were systematically related to the independent findings of the raters. In
168 Chapter 7

a large number of cases, the raters indicated that they lacked the infor-
mation from the medical records and interviews needed to determine
defendant liability. If the authors had attempted to force an answer and
not allow for uncertainty about liability, there would have been more
discrepancies between ratings of liability and case outcomes.
A fourth and more recent study provides the most conclusive evidence
on myth 3. Studdert, Mello, Gawande, et al. report results of assessments
of a random sample of 1,452 medical malpractice claims extracted from
the files of five liability insurers to determine whether or not an injury
occurred, and if it had, whether or not the result was due to medical
error.22 Unlike the previous studies, the five companies were located in
all regions of the United States. The authors focused on four clinical
areas: obstetrics, surgery, missed or delayed diagnosis, and medication.
The physician reviewers were trained in one-day sessions at each study
site. Reviews lasted 1.6 hours per file on average; each file was handled
by one reviewer.
Reviewers scored severity of injury on a widely used index ranging
from emotional injuries to death.23 If no injury was apparent from the
record, the assessment of the file was terminated. This occurred in 3
percent of claims. More important, the physicians recorded their judg-
ments of whether or not an error occurred. This scale ranged from little
or no evidence that an adverse outcome resulted from one or more errors
(scored 1 out of 6) to virtually certain evidence that an error had
occurred. The full scale is shown in figure 7.1, which is reproduced from
the study. Claims receiving a score of 4 or more (with 4 being “more
than 50–50 chance of error”) were classified as claims involving medical
error. Claims with scores of 3 (“close call, but <50–50 chance of error”)
were classified as not involving error. Physician reviewers were not
blinded as to whether or not payment was actually made in the case.24
As in the three other studies, there was a clear pattern of physician
assessor agreement with the actual outcome of the claim (figure 7.1). For
those claims for which there was “little or no evidence” of medical error,
payment was made in only 19 percent of claims. By contrast, when the
evidence of error was “virtually certain,” payment was made in 84
percent of claims. The claims for which error was ambiguous fell between
the two extremes in terms of the percent of claims paid, with the percent
 Juries and Health Courts 169

350 Classified as not involving Classified as involving

error error
300
No payment
No. of Claims (N = 1404)

250 Payment

200

150

100

50
19% 32% 52% 61% 72% 84%
0
1, Little 2, Slight-to- 3, Close 4, Close 5, Moderate- 6, Virtually
or no modest call, but call, but to-strong certain
evidence evidence <50–50 >50–50 evidence evidence
Confidence in Judgment Regarding Error

Figure 7.1
Relationship Between Physician Rate Confidence in Judgment Regarding Error
and the Probability That Payment Was Made to Claimant. Source: Studdert,
Mello, Gawande, et al. (2006).

paid increasingly monotonically with the degree of certainty that an error

had been committed.
Interestingly, the authors drew a line between “close call, but <50–50”
and “close call, but >50–50.” Thus, a case for which the evidence was
49–51 was considered not to have involved an error, while a case for
which the evidence was 51–49 was considered a case in which an error
had occurred. The evidence clearly suggests that the probability that any
payment was made is systematically related to whether or not an error
had occurred, but the system is not perfect. Of course, there was only
one physician reviewer per case, and such reviewers may make mistakes
from time to time.
The article indicates that claims not involving errors accounted for 13
to 16 percent of the system’s total monetary cost. However one views
this percentage (substantial or small), and whether or not a case involves
an error, depends substantially on where in figure 7.1 one draws the line
between error and no error. This is a fundamental point to keep in mind
in assessing the accuracy of dispute resolution in medical malpractice.
170 Chapter 7

In most claims rated for this study, physician reviewers were somewhat
uncertain as to whether or not a medical error had been committed.
Unfortunately, the study conclusions do not stress or even mention that
the estimates of error are subject to a very high degree of uncertainty.
In particular, results from the Harvard Medical Practice study suggest
the opposite conclusion, namely, that most medical malpractice claims
are “invalid”.25 At first glance, this view seems to support myth 3 and
contradict the evidence from the four studies just described. Even if a
review of an entire file revealed that there is “little or no evidence of
error,” this may not have been obvious at the time the claim was filed.
As more evidence is accumulated, many claims are dropped without
payment.26 Filing a claim gives the claimant the option of pursuing a
claim to trial. The Sloan, Githens, Clayton, et al. study surveyed plaintiffs
many years after the injury occurred. Assessments were based on survey
evidence in addition to medical records. Also, the ratio of invalid to valid
claims strongly depends on whether or not the study allows for an
“uncertain about liability” category. Just as in a trial, there will be dis-
agreements among the experts about liability. It is not appropriate to
treat the views of raters as an ironclad “gold standard.”

Empirical Evidence Suggesting That Jurors Make Errors in Evaluating

Liability
All of the four studies just described focus on comparisons between case
outcomes and physician assessments of liability. There is a substantial
body of literature, not specific to medical malpractice, focusing on juries’
decision-making processes.
Literature from cognitive and social psychology documents various
biases that apply to individuals of all types, including jurors and judges.
People tend to underestimate the frequency of low frequency and to
overestimate that of higher frequency.27 People find an anchor or refer-
ence event for assessing probabilities, termed the availability heuristic.28
That is, the probability of the event in question is compared to some
other event. Well-publicized hazards tend to be assigned a higher risk
than less publicized ones, especially in the case of phenomena that were
previously unfamiliar, such as terrorist attacks or SARS.29
 Juries and Health Courts 171

Some research, as well as anecdotes, suggests that jurors are suscepti-

ble to relying on such legally irrelevant factors as the physical character-
istics of litigants, the presence and content of pretrial publicity, and
inadmissible evidence.30 Since there are no minimum educational require-
ments for individuals to serve on juries, and persons with high demands
on their time may be more likely to try to avoid jury duty, many jurors
have low levels of educational attainment.31
Much of this research is based on evidence from mock juries,32 that
is, individuals organized into fake juries in order to research how groups
of individuals make various assessments of the types that juries often
make. Results of research based on data obtained from mock juries criti-
cally depend on how well the questions are phrased.33
There is some empirical evidence that jurors have difficulty making
such benefit versus cost trade-offs. That is, jurors are swayed by the cost
associated with an iatrogenic injury while ignoring the benefits that could
have been anticipated at the time clinical decisions were made that
caused the injury.34
Viscusi compares assessments of ninety-four judges from all parts of
the United States versus those of jurors.35 Compared to the jurors, judges
tended to be better educated, had more exposure to risky decisions in
their roles as judges, and had experience with conditions under which
judicial errors are overturned on appeal. Furthermore, the judges were
better able to assess objective probabilities of risk, although they, like
the citizens, tended to underestimate large and overestimate small prob-
abilities of risk. The judges were more likely than the jurors to reach
decisions based on a benefit versus cost calculus. They were less likely
to confuse ex ante and ex post risks than were the jurors. While such
comparisons are interesting, they are not very surprising, and, as dis-
cussed more fully below, we are unclear what the implications of the
findings really are.
In sum, the myth that jury verdicts are random on liability is rejected
by the empirical evidence. However, viewed from another perspective,
there is no perfect consistency with the judgments of independent expert
raters either. In some cases, perhaps the raters may have been in error
because, for example, they lacked some of the information presented to
172 Chapter 7

juries. In many other cases, the raters may have been right. This is truly
a situation of “cup half empty” versus “cup half full.”
Rather than focus on deficiencies of juries, since the American jury is
here to stay, it seems advisable to focus on how legal procedures can be
improved. Courts are currently provided insufficient guidance about (1)
recent scientific findings and the weight given to those findings by inde-
pendent experts, and (2) the thinking that led to prior judicial decisions
about liability.

Empirical Evidence on Juror Behavior: Determination of

Compensatory Damages

Do Jurors Make Major Errors in Assessing Damages?

There is also a body of empirical evidence on jury assessments of com-
pensatory damages. Some studies have focused on the accuracy of assess-
ments of nonmonetary loss. Other research has compared total
compensation with monetary loss. Payment for nonmonetary loss is
derived as the difference between total compensation and monetary loss.
Rather than focus on outcomes and make indirect inferences about juror
behavior, some research examines juror behavior directly.
Vidmar and Rice compare assessments of nonmonetary loss between
persons waiting to serve on juries and twenty-one arbitrators for various
personal injury, contract, and labor disputes. They find that the assess-
ments of the laypersons and arbitrators were positively correlated and
that the laypersons did not assign higher values to such loss than did the
arbitrators.36 Another study found no difference in “pain and suffering”
awards when the defendant was an individual, two individuals, or a
hospital or corporation. This casts doubt on the “deep pockets”
hypothesis.37
Some other studies are more critical of juries. Optimal deterrence
requires that punishments increase when the probability of detection
decreases. In research using mock jurors, there was no difference in rec-
ommended penalties depending on the probability of detection they were
asked to assume.38 In theory, there should be a negative relationship
between the probability of detection and the size of the award. For
example, if the number of highway patrols per highway mile is reduced,
 Juries and Health Courts 173

the fines for speeding should be increased. Sunstein, Schkade, and

Kahneman report that even law students trained in deterrence theory
rejected making judicial decisions on the basis of such economic thinking
about damage awards.39 This suggests that the problem does not apply
only to jurors but also to those trained in the law.
Experience seems to help. Judges seem more likely to think in
terms of the risk-cost trade-offs inherent in the concept of optimal
deterrence than do jurors.40 This has led some experts to conclude
that a shift from jury to judge determination of damages might be
worthwhile.41
Vidmar criticizes this line of research, arguing that the personal injury
scenarios presented to the subjects were too short to allow them to make
an informed judgment.42 The longest of the scenarios presented to the
subjects contained fourteen sentences, much less information than would
be conveyed in a trial.
There is evidence from some studies of individual decision-making that
individuals have difficulty making the kinds of decisions about damages
that juries are asked to make. Yet even if assessments of plaintiff loss
are inaccurate by some standard yet to be defined, they are not upward
biased on average.

Empirical Evidence on Myth 4: Medical Malpractice Claimants Are

Overcompensated for their Losses
The survey of the 187 plaintiffs in Florida obtained information on the
cost of injuries.43 Using these survey data as well as information from
other primary and secondary sources, the authors performed an inde-
pendent analysis of monetary loss in each of the 187 cases for both past
and future loss. These estimates were then compared with the compensa-
tion the plaintiffs ultimately received.
The costs resulting from major injuries were considerable. Twenty-six
percent of the plaintiffs received no compensation. On average, including
those cases for which no compensation was received, compensation
amounted to about half of monetary loss. Even including compensation
for nonmonetary as well as monetary loss, compensation fell far short
of injury cost. In cases in which the plaintiff won at trial, compensation
exceeded monetary loss by 22 percent on average—hardly an indication
174 Chapter 7

of a “runaway jury.” Some or all of the 22 percent was probably

payment for nonmonetary loss.
The authors report a systematic relationship between injury cost and
compensation, although variations in injury cost explained only part of
the variation in compensation. A sensitivity analysis did not reverse the
findings; the analysis adjusted for lawyers’ fees, compensation from col-
lateral sources, and favorable tax treatment of funds received from tort.
None of the plaintiffs received punitive damages.
The authors conclude that the empirical evidence does not suggest that
compensation is excessive on average. This does not eliminate the possi-
bility that compensation was excessive in selected cases. Further, the
results do not support imposition of caps on damages.
Some reforms, such as those proposed in Chapter 5, may have merit
because payment for nonmonetary loss is inconsistent among cases;
moreover, juries are given little or no guidance regarding the criteria to
be used for setting such compensation.

Determination of Punitive Damages

Why is the Subject of Punitive Damages a Public Policy Concern?

Polinsky and Shavell begin with the statement “One of the most contro-
versial features of the American legal system is the imposition of punitive
damages.”44 Punitive damages are designed to fulfill the social functions
of punishment, deterrence of injuries, and general deterrence of the
defendant’s conduct that violates social mores.45 For punitive damages
to be awarded, the plaintiff must show that the defendant intentionally
harmed the plaintiff, or recklessly harmed the plaintiff by actions or
inactions undertaken with knowledge that harm would occur, or with
gross deviation from ordinary standards of care. Awards for punitive
damages are often highly publicized, but are rare in medical malpractice
cases.46
The sheer size of the dollar awards in some cases has raised
concerns.47 Other concerns are the substantial variability of awards
for punitive damages and discrepant punishments by different juries
for essentially the same act, reflecting in part the vague instructions
to juries for assessing the appropriate level of damages, leaving it to
 Juries and Health Courts 175

individual juries to develop their own criteria. The situation in civil

law, particularly in regard to punitive damages, stands in sharp contrast
to criminal law, where sentencing guidelines have become far more
common.48
One dispute in the academic literature stems from results of compari-
sons of patterns of awards for punitive damages between juries and
judges. In Hersch and Viscusi’s (2004) sample of jury and bench verdicts,
juries, which decided 95 percent of cases, were more likely to award
punitive damages than were judges, and award levels of punitive damages
were higher. They found little relationship between compensatory and
punitive awards, a finding in contrast to earlier studies by Eisenberg,
Goerdt, et al. (2002), Eisenberg, La Fountain, et al. (1997), and Karpoff
and Lott (1999), who found that the size of punitive damage awards is
positively related to the size of compensatory awards. Hersch and
Viscusi’s (2004) analysis shows that study findings are quite sensitive to
the explanatory variables included in the analysis. The unstated implica-
tion of Hersch and Viscusi’s analysis is that relying on juries to set puni-
tive damages is bad public policy.
Comparisons between the performance of judges and of juries raise
some more fundamental questions. First, when opinions differ, are the
opinions of the judges always the correct ones? Here it is useful to
examine other empirical evidence on the performance of judges. Tabar-
rok and Helland investigate why trial awards differ among U.S. states.49
The key hypothesis is that elected judges will be particularly sensitive to
the state of residence of plaintiffs and defendants. Plaintiffs almost always
live in the state in which the trial takes place. By contrast, for many
personal injury cases, not including medical malpractice, the defendant’s
headquarters is out of state. In medical malpractice, defendants tend to
be in state.
The authors present evidence from one state (Florida) that at least 80
percent of campaign contributions to judges in that state were made by
lawyers. Contributions from trial lawyers could lead to bias in favor of
granting high awards to plaintiffs. Tabarrok and Helland find that the
expected total award was about $240,000 higher in states with elected
judges when the defendant was out of state.50 Thus, judges, too, can be
biased.
176 Chapter 7

It is not surprising that decisions of judges and juries sometimes differ.

If we were to limit the voting population to persons with a graduate
degree and/or household incomes over $200,000, election outcomes
would probably differ appreciably. However, such restrictions hearken
back to the literacy tests enacted by states under Jim Crow.

The Bottom Line: Rationale for Reform of Courts

The American jury is under attack from a much broader group of stake-
holders than physicians, hospitals, and medical malpractice insurers. It
is also true of some criticisms of awards for compensatory damages and
of attorneys’ fees, the subjects of the previous two chapters.
Two criticisms of the existing system seem more fundamentally appli-
cable to the legal system and to resolution of medical malpractice dis-
putes in particular. The system’s outcomes may be inequitable in both
horizontal (equal treatment of equals) and vertical (equal treatment
according to harm done) dimensions. That is, while the current system
is highly individualized to the circumstances of a particular case, indi-
vidualization can go too far, especially if decision-makers are subject to
substantial errors in judgment and even biases. Random outcomes may
also weaken any deterrent signal that a judicial outcome may otherwise
have. Greater consistency in decisions seems like a worthwhile goal,
providing a possible rationale for health courts as well as for scheduling
damages (discussed in Chapter 5). Juries currently receive little guidance
about past judicial decisions, and their rationale, to inform their delibera-
tions. Scientific information is presented anew. Such replays in the court-
room seem inefficient at best.

Health Courts

Rationale for Health Courts

Proponents of health courts contend that these special courts would offer
several potential benefits: reducing medical malpractice insurance premi-
ums; increasing access for all who have been injured; and encouraging
patient safety measures.51 In addition to providing victims with consis-
tent, fast, and relatively easily obtained compensation when this is war-
 Juries and Health Courts 177

ranted, health courts are also intended to reduce cost by streamlining the
process, maintaining consistent medical standards, and capping or sched-
uling damages.
Several factors have led to proposals for specialized health courts. For
instance, there is a general distrust of juries, especially by groups likely
to be adversely affected by their decisions. In addition, there are concerns
about the inadequacies of juries to decide technical matters common in
medical malpractice litigation, and the inexperience of judges in the
mainstream judiciary in medical matters. Nevertheless, the concept of
specialized courts is not new. There are specialized courts for family
matters, juveniles, taxes, bankruptcy, admiralty, mental health, drug,
and even the homeless.
Administrative approaches are another alternative to the status quo.
In this chapter, we limit the discussion to health courts. (Administrative
approaches are discussed in detail in chapter 11.)

Health Court Proposals

Alternative specialty courts for medical malpractice are receiving a great
deal of public attention in the first decade of the twenty-first century.
The Common Good and the Public Policy Institute are driving the cam-
paign in both public and private spheres, garnering support along the
way from health policy scholars, physicians, attorneys, leaders of uni-
versities and medical schools, and directors and heads of consumer
organizations, among many others.52 A decade and a half earlier, the
American Medical Association (AMA) introduced a proposal utilizing a
specialized administrative health court system in 1988.53 The Public
Policy Institute and the Common Good’s proposals share essentially the
same features as the AMA proposal, albeit with a simplified design.54
However, unlike the current health court proposal, the AMA proposal
received little publicity or support.
Health courts eliminate jury trials for medical malpractice actions and
establish damage schedules with caps on nonmonetary damages. Although
they differ in some details, the major proposals for alternative courts for
medical malpractice legal disputes all are fundamentally the same. Juries
are removed and decision-making power is placed with a special judi-
ciary that has scientific training. In addition, to preserve the neutrality
178 Chapter 7

of expert opinions, experts are selected and compensated by the court,

not the parties. Also, decisions about standards of care are made as a
matter of law by the judge. Compensation schedules are set by an inde-
pendent body, and reviewed periodically to account for inflation and
changing costs specific to medical malpractice injuries.
The process is designed to be accessible and fast for claimants. Physi-
cians are required to inform the patient immediately after an adverse
event. Then, to initiate a claim, an injured patient must only complete a
simple claim form and send it to a review board. These forms would be
available throughout the state, and could be completed without legal
assistance. An initial review board would then evaluate the circumstances
under which the injury occurred.
After the review board receives the initial complaint, it would obtain
information from medical records and interviews with patients, health
care providers, and others with primary knowledge. Similar to some
variants of a medical no-fault, cases arising from avoidable events or
clearly due to negligence or malpractice would receive immediate payment
from the health care provider according to the schedule of damages.55
Cases that clearly do not constitute medical malpractice or are too minor
to merit an award would be dismissed with a limited right to appeal to
the health court.56 If it is not clear how the injury occurred, the claim
would be sent to the health court for a full trial. If, in the plaintiff’s
opinion, the judge fails to apply the law correctly, plaintiffs would have
the right to appeal the decision to a dedicated court of medical
appeals.
Full-time judges are a major feature of the health court proposals. The
judge would deal only with malpractice cases, entirely replacing the func-
tion of a jury. In one proposal (Common Good’s), specialized judges
would shape legal standards for medical malpractice, creating a body of
science-based common law that health care providers could rely on when
making treatment decisions.
In theory, a body of science-based common law seems valid and useful,
but raises issues of its own. In the context of health courts, standards
for medical practice would develop under state law. Yet, without federal
regulation each state would be free to develop medical standards by way
of the common law, allowing the possibility of variable legal and medical
 Juries and Health Courts 179

practice standards. Another proposal (the Public Policy Institute’s) sug-

gests creating legislation that would allow the U.S. government to regu-
late health courts under the power of the interstate commerce clause.57
Health courts face several implementation barriers, including chal-
lenges to the constitutionality of scheduled damages, the limited right to
appeal, removal of juries, the questionable skill of appointed judges, and
access to the system. In the past, tort reform statutes have been over-
turned on grounds of unconstitutionality. Opponents of such courts have
argued that health courts would not survive judicial scrutiny. Propo-
nents, however, have countered that the U.S. Supreme Court has not
extended the Seventh Amendment right to a jury trial to state courts.58
While many states have granted this right in their own constitutions,
advocates for health courts have suggested that the U.S. government
could easily overcome this barrier at the state level by enacting federal
legislation preempting existing state guarantees.
To the extent that access to the courts is improved, medical malprac-
tice payments and premiums could conceivably increase, not decrease.
Since currently so few adverse medical events actually result in claims,59
increased payments and premiums may be in the public interest. But this
is not the type of outcome that proponents of health courts are likely to
have in mind. This was the experience with state statutes facilitating
alternative dispute resolution in medical malpractice, which presumably
increased claimant access by lowering the cost of dispute resolution.60
Even if many larger awards were eliminated, partly due to fewer findings
for plaintiffs and partly due to scheduling of damages, these savings may
be offset not only by additional claims, but also by the additional cost
of the public subsidies for a new kind of court.

The Experience of Other Specialty Courts

The concept of specialty courts has been implemented in other contexts.

While there are differences as well as similarities, specialty courts provide
insight into the challenges of creating a separate system for health courts.
Two specialty courts with long histories are bankruptcy and tax courts—
the first American bankruptcy court emerged in 1841, and tax courts
began in 1924.61 Unlike proposed health courts, they are both federal
180 Chapter 7

administrative courts. More recently, drug courts were created in 1989

in Dade County, Florida, with positive reception; since then, virtually
every state has implemented a drug court.62 Mental health courts began
in 1997 in Broward County, Florida. Their development was accelerated
by subsequent federal legislation.63
The current trend toward specialization of courts has been well
accepted, but not without some controversy. Both these courts were
developed in order to provide defendants an opportunity to receive
financial assistance for their recovery in lieu of fines and jail time.
However, these courts often require a guilty plea before an alternative
judgment is issued. In some cases, only those who plead guilty have
access to the drug or mental health court; those entering a not-guilty plea
must remain in the general court system.
Because entry into the system does not require a guilty plea, unified
family courts may offer a better paradigm. A unified family court is a
system in which family cases, both criminal and civil, are tried in one
separate family court. Juvenile and family courts had parallel develop-
ments.64 The theory on which both courts were established is that state
intervention can prevent social problems (e.g., juvenile delinquency) if
courts are able to deal with the source: neglectful and abusive
families.65
These issues created a drive to develop an alternative court system.
For family courts, this meant centralizing resources, unifying jurisdiction
of all domestic relations cases, training judges, and resolving a family’s
legal problems in as few court appearances as necessary. Three national
organizations worked together to draft the Standard Family Court Act,66
with the stated purpose “to protect and safeguard family life in general
and family units in particular by affording to family members all possible
help in resolving their justiciable problems and conflicts . . . in a single
court, with one specially qualified staff under one leadership . . . with one
set of family records, all in one place under the direction of one or more
specially qualified judges.”67 This model is sometimes referred to as “one
judge, one family.”
When the Standard Family Court Act became law in 1959, several
states fashioned family courts completely separate from the general
docket. The family court concept was received positively; over the next
 Juries and Health Courts 181

few decades many more states would enact their own statewide family
court systems, some with exclusive jurisdiction and others permissive.
Some states expanded their existing juvenile courts to include family
issues.68
Mental health courts, which work with prosecutors’ offices, are a
much more recent development than family courts. Mental health courts
are designed to offer the mentally ill an opportunity to avoid extensive
prison time, provided they adhere to treatment protocols established by
the courts on a case-specific basis. Defendants must plead guilty and
submit to a detailed psychiatric evaluation before becoming subject to
the mental health court’s jurisdiction. Maintaining contact with mental
health providers and adhering to medication protocols are monitored by
the mental health court.69
Specialized courts offer many advantages over general courts. First,
duplication in judicial attention has been wasteful and ineffective. One
advantage of family courts is that they retain jurisdiction over defen-
dants, promoting continuity of supervision and accountability.70
Second, specialized courts are particularly appropriate when coordina-
tion between the judiciary and various other public and private organiza-
tions is likely to be productive. Such coordination takes place over time.
In the context of medical malpractice, a specialized court would be able
to monitor the service benefits described in Chapter 5.
Third, forum shopping or judge shopping is a common occurrence
among litigants.71 This problem is by no means limited to family law;
medical malpractice suits are also subject to this problem, perhaps even
more so. With giving just one court jurisdiction, and assigning the case
to just one judge, forum shopping is prevented. However, state law
determines medical malpractice and family law statutes, so it is likely
that when parties are in different states, some forum shopping will still
occur. Nevertheless, single jurisdiction family courts help reduce forum
shopping by limiting the number of available forums. Unifying the
authority of a court to hear certain types of cases eases the caseload from
the general court. With at least 25 percent of civil filings being family
cases, it eases the burden considerably.72
Fourth, specialized courts such as family courts have the potential to
save money in the long run. Initially, however, as would be the case for
182 Chapter 7

health courts as well, a unified family court system is an expensive

venture requiring much investment for its start-up.73
Despite the start-up costs, recent studies show that family courts have
realized some long-term savings through judicial efficiency, economy,
competency, quality of adjudication, and coordination among judicial
officers, police, and treatment personnel.74 Unified family courts also save
money for the state due to the need for fewer justices.75 Moreover, if the
court functions so as to resolve cases more quickly, nonmonetary costs
from a lengthy and drawn-out process will be saved immediately. In
addition, and perhaps more important, children are spared multiple trips
to the witness stand, and families will receive the aid they need quickly
and have the support services to obtain it. Fast compensation and reliable
support services are goals of health courts as well.
Domitrovich documents savings of about 20 percent in New Jersey.76
Speedier resolution of disputes can reduce the nonmonetary cost, as well
as legal cost, of medical malpractice litigation; however, the overall
savings to the system would be minimal, given that so few cases ever
reach the trial stage. Ironically, savings in per-case litigation costs could
lead to more litigation. Trial lawyers may be more willing to accept cases
on a contingent fee basis if their litigation cost is reduced. If there is a
case for health courts based on the family court experience, the case
should be advanced on another basis than reduced litigation cost.
Critics, and those interested in reforming existing unified family court
systems, point out that many courts fail to fulfill their purpose. To begin
with, in order to avoid judicial burnout,77 many jurisdictions require that
judges rotate through the court system as often as every eighteen months.78
Aside from preventing judges from developing expertise, rotating judges
may cause families to see more than one judge, defeating the fundamental
purpose of family courts. Even judges who do not rotate through the
system may have short terms on the bench; judicial appointments can
range from nine months to a lifetime.79
Another potential problem is hiring and retaining skilled justices.
Family and juvenile courts have a reputation as the training ground for
new justices before they move to a general court.80 Lower pay and fewer
opportunities for career advancement compound the problem. The nega-
tive effect of the short terms and high turnover of judges is exacerbated
 Juries and Health Courts 183

by the fact that unified family courts have wide jurisdiction over several
areas of law, both civil and criminal.81 Medical malpractice cases are
sometimes high profile. This disadvantage may not apply to health
courts.
Although instructive, the experience of other specialty courts is not
entirely comparable. Bankruptcy and tax courts are federal administra-
tive courts, and their jurisdiction is fundamentally different from that of
proposed health courts. Federal courts are governed by federal rules of
procedure and federal substantive law, which has produced a large body
of case law followed by all federal bankruptcy and tax courts. Without
federal intervention, states have jurisdiction over medical malpractice
claims and may create legislation as they choose.
Health courts would be created by state legislation, and there would
be considerable variation among states. In contrast, bankruptcy courts
have original and exclusive jurisdiction over bankruptcy cases. This
means states are prohibited from enacting bankruptcy laws or asserting
control of the field.82
Tax courts provide experience in one aspect that can instruct the
development of health courts: unrepresented litigants. In tax court, 43
percent of litigants are pro se.83 Taxpayers often have cursory knowledge
of the Tax Court Rules, the Internal Revenue Code, and the Federal
Rules of Evidence, whereas their adversary, a government agent, is well
versed in the rules, procedures, and techniques needed for trial. As a
result, the majority of cases result in verdicts for the government.84
Without contingent fees, many malpractice litigants in health court
would be faced with two options: represent themselves or not file a
claim.
As mentioned before, these plaintiffs may find it difficult to obtain
representation with scheduled damages and a 20 percent cap on attor-
neys’ fees. These factors indicate there may be many pro se litigants. This
is especially true if the process is as easily initiated, as health court pro-
posals suggest: by filling out a readily available user- friendly form. As
in tax court, health court plaintiffs would have only a cursory knowledge
of procedural rules and of tort law and its vast and ever-changing prec-
edent. Defendants, on the other hand, would have resources to hire
skilled counsel with expertise in procedural rules and the nuances of the
184 Chapter 7

substantive law. As with tax courts, there is a clear risk that health courts
could develop a defendant bias.

The Bottom Line

It is unrealistic to expect health courts to solve all, or even most, of the
problems with the existing system, as the advocates of such courts would
have it. However, with the more recent success of and demand for spe-
cialty courts in other areas of the law, this may be an idea worth
pursuing.85
Health courts have some potential benefits. Probably the major advan-
tage is the courts’ ability to coordinate with other agencies. For example,
rather than pay a lump sum to a family with a birth-injured child, care
can be monitored and coordinated over time.
Another positive feature is the possibility of broadening the standard
for individual negligence. Much like no-fault systems,86 this would have
the effect of increasing the number of injury victims compensated, allow-
ing clearly avoidable mistakes to be compensated in the first hearing,
without a trial, thereby reducing the burden on the court system and the
time it takes for injury victims to be compensated. Increasing the number
of compensated injuries and holding more adverse events as liable would
also create an incentive for health care providers to adopt patient safety
measures, but only if coupled with provisions requiring experience rating
of medical malpractice insurance premiums.
Currently, the mean time it takes to resolve a malpractice case is over
two years, with many of the more complex cases taking much longer to
resolve.87 For a patient needing compensation to pay health care and
other immediate expenses, this delay can be extremely burdensome. By
eliminating the need for trial for cases in which adverse events are clearly
avoidable, a greater number of patients would receive compensation
more nearly synchronized with the time it is most needed.
Very few cases in state courts currently involve medical malpractice.
Thus, judges cannot be expected to be highly knowledgeable about the
technical issues in these cases, making it more difficult, for example, to
know when inaccurate scientific information is being introduced by an
expert. By specializing in these cases, judges could become more gener-
ally knowledgeable about medical issues.
 Juries and Health Courts 185

Health courts also have some important weaknesses. For one, there is
concern about specialty courts becoming “ghettoized.” In addition, it
may be hard to find high-quality candidates for judicial positions who
are sufficiently neutral. Underlying biases may be a reason that judges
are currently appointed or elected. A judge with strong views on medical
malpractice may be elected or appointed for this reason. Consequently,
plaintiffs may prevail too frequently or not frequently enough. Addition-
ally, if appointment of experts is in the judge’s discretion, it is not entirely
clear that the experts chosen will be neutral. Allowing the parties to use
their own experts would assure that all viewpoints are heard and are on
the record.
To deal with the appointment process and its consequences, a state
could establish a special board whose sole responsibility is appointing
judges to health courts. Terms on the board could be rotating and of
sufficient length that a governor would not be able to appoint most of
the board members even if he or she serves two consecutive terms.
Another concern is that health courts would be overwhelmed by the
influx of cases. This has happened in other specialty courts; family court
dockets have been steadily increasing, and the burden is borne by the
individual justice.88 Over time, malpractice cases’ frequency is likely to
decrease but, at least initially, health court judges may be as over-
whelmed as family court judges are.
Opponents have raised the question of whether the development of
law in a specialty court reflects the development of other principles of
law in the general court system. They argue that an area of law can
become isolated from the development of law. Generally, the develop-
ment of law influences courts in their decisions across specialties. Stan-
dards for admitting evidence or expert testimony and the application of
constitutional rights have evolved over time, as have most other areas of
law, affecting development across subject matter. Bankruptcy and tax
courts have been accused of ignoring those developments and creating
an entirely distinct body of law. The common-law system functions on
precedent. It is important for specialty courts to be involved in incorpo-
rating and building on the precedents of other courts.
Perhaps the most common criticism is the removal of juries. The
Seventh Amendment to the U.S. Constitution guarantees a citizen’s right
186 Chapter 7

to a jury trial in civil cases.89 Channeling a medical malpractice plaintiff

to a health court is not necessarily denial of the right to trial by jury. It
depends on how such courts would operate in practice. Undoubtedly, if
they are instituted, health court constitutional challenges would eventu-
ally be considered by the U.S. Supreme Court.
In the end, even considering all the benefits of a specialty court, there
are limits to what specialty courts can accomplish. A family law expert
from the District of Columbia Bar Association acknowledged, “Reform-
ing the court will not solve the underlying societal problems that lead to
the abuse and neglect of our city’s children. . . . A judge can only do so
much.”90 Though he was discussing family courts in particular, his words
are entirely relevant to all specialty courts. Health courts are by no means
a panacea, as is sometimes asserted.

Conclusion

This chapter has both discussed performance of the existing system, as

in our refutation of myths 3 and 4, and assessed the advantages and
disadvantages of specialized health courts. Along the way, we have pre-
sented evidence on specialized courts in other contexts, with a view
toward whether and how the lessons learned from these other experi-
ences apply to proposals for health courts.
The criticisms of jurors raise much larger issues currently plaguing the
social sciences which have broad implications for public policy. Although
the critics of juries focus on the limitations of individuals in these set-
tings, many others have questioned the ability of individuals to make
fully informed rational choices in many other contexts. Perhaps because
of the “wisdom of crowds” (albeit small ones, six to twelve persons),
collective decision-making of juries may be much better than individual
decision-making in family units. The studies that are critical of juries do
not compare decision-making in this context with other decisions people
routinely make.
Should health courts be adopted? This chapter has presented some
pros and cons of health courts. The most attractive feature is the poten-
tial for the court to monitor payment and care of injury victims over
time and to facilitate coordination among agencies as family and mental
 Juries and Health Courts 187

health courts do. Health courts are worth considering, but only if the
problems raised by them are viewed through a framework in which
alternative methods for addressing the same set of issues are considered.
For example, states might consider different approaches for presenting
scientific evidence to courts. Perhaps the public sector should fund
research that summarizes the empirical evidence on specific scientific
controversies that arise in medical malpractice cases. Nontechnical sum-
maries could be provided to juries.
Another possibility is using the standard of review set out in Daubert91
to disqualify incompetent witnesses whose testimony might mislead
jurors. Still another possibility is the use of expert juries. Viewed in the
context of a number of other realistic alternatives, health courts are
indeed well worth considering.
8
Patient Safety and Medical Malpractice

Patient safety has moved to the top of the health policy agenda in recent
years. Widespread interest in the topic can be traced to the publication
of an Institute of Medicine (IOM) report in 2000, titled To Err Is
Human. But no matter how competently they are done, reports tend to
gain widespread public attention only when their release occurs in an
environment receptive to the findings. Concern about patient safety
plausibly was building under the surface, and release of the IOM’s
report, with its very comprehensive review of the literature and specific
recommendations, provided an important catalyst for galvanizing the
public’s and policymakers’ attention on this important public policy
issue.
The concepts of patient safety and medical malpractice are closely
linked, at least in theory. Whether this is true in practice is quite another
matter. As discussed in chapter 3, the weight of the empirical evidence
is that the threat of being sued does not deter injuries. Yet deterrence is
at the top of nearly any list of what imposing tort liability should achieve.
But in defense of tort—admittedly a somewhat lame defense—all empiri-
cal studies have assessed deterrence under varying degrees of a threat of
tort liability. No study has compared patient safety in a world in which
the threat of tort liability was totally absent with rates of patient safety
as tort liability operates currently. Perhaps such a comparison would
have revealed that the entire removal of the threat of tort increases rates
of medical errors.
Patient safety is related to medical malpractice in that an important
purpose of imposing tort liability on health care providers is to provide
a negative incentive or a “stick” to induce provision of socially optimal
190 Chapter 8

levels of precaution. Ideally, there would also be positive incentives for

providers or “carrots” to engage in activities that result in preventing
the occurrence of diseases and adverse health outcomes.
During the first decade of the twenty-first century, increased public
interest in medical malpractice arose at the same time as increased inter-
est in patient safety—probably a coincidence. Many advocates for public
safety desired to keep the two issues separate, apparently out of a
concern that becoming muddled in medical malpractice issues would
blunt the patient safety movement’s momentum. Patient safety and
medical malpractice are in fact inextricably related, a situation that is
clearly understandable for political reasons.
Some experts and policymakers not only allege that there is no
relationship between patient safety and medical malpractice, but they
go a step further in arguing that medical malpractice is actually
counterproductive in assuring patient safety and patient welfare more
generally.

The Alleged Patient Safety-Medical Malpractice Mismatch

Arguments for a Mismatch

There are several arguments for a mismatch.
One argument is that medical malpractice is forcing health care pro-
viders out of practice, thus decreasing patient access to care. This is
negative defensive medicine.1 Putting aside whether or not the threat of
lawsuits and rising premiums have affected career choices, in a sense, if
medical malpractice is doing its job, it should be encouraging some exits
from practice.
Ideally, the tort system should identify and remove physicians who
consistently commit errors or are negligent, and upon their removal,
more qualified physicians would fill the vacancies. This applies to the
provision of services as well; when care is below standard, the office,
unit, or hospital providing it should be closed. For example, a hospital
in upstate New York voluntarily closed a cardiac surgery unit due to a
17.6 percent mortality rate, which was nearly four times the state
average.2 Before reopening the unit, the hospital took several measures
to ensure an increase in patient safety. It implemented a quality assurance
 Patient Safety and Medical Malpractice 191

program, created a full-time service chief, increased nursing staff, and

created weekly teaching conferences. In less than ten years, the hospital
had reduced the cardiac surgery mortality rate to 1.77 percent.3 This
story provides an interesting example, but perhaps not an ending that
generalizes.
A study conducted in the 1980s revealed that the distribution of
medical malpractice payments in Florida is highly skewed, with a few
physicians accounting for much of the loss, even after accounting
for factors such as specialty.4 The same study indicated that physicians
with adverse claims experience were less likely than others to subse-
quently quit practice or move to another state. And although physicians
with very adverse claims histories were more likely to have complaints
against them filed with the state’s licensing board, the rate of actual
sanctions was very low. A physician’s adverse claims history does not
necessarily imply poor quality of care but, at a minimum, it would seem
appropriate to examine his or her practices. At least in Florida, neither
the threat of tort liability nor the licensure process seems to have been
effective.
Some tort reformers are concerned that engaging in unnecessarily
heightened patient safety will put physicians at an even greater risk of a
lawsuit. There is an argument that defensive medicine is self-reinforcing;
the more physicians provide aggressive treatment for low-risk conditions
or order unnecessary tests, the higher the probability that those practices
will become the legal standard of care.
Some critics of tort liability as it applies to medical care have gone a
step further, maintaining that the threat of medical malpractice suits not
only does not improve patient safety, but it may even cause provision of
care to be less safe than it would be in its absence. The critics assert that
the threat of litigation has a chilling effect on discussions among provid-
ers about adverse outcomes. In addition, some critics argue that doctors
fear the possibility that their private discussions could be discoverable
by a plaintiff’s attorney during the course of litigation.5 It is unknown
to what extent this fear actually exists; there has been neither systematic
documentation demonstrating that liability forces physicians into silence
nor studies linking malpractice exposure with frequency of error re-
porting.6 Tort liability is relevant to the topic of patient safety; some
192 Chapter 8

would argue that the primary goal of imposing tort liability is to deter
injuries.
If “to err is human,” not wanting to admit to errors is also human.
Even if there were no threat of tort, it seems unlikely that providers
would generally be willing to make their errors known to others for
various psychological, if not professional and business, reasons.
The avoidance of error reporting in the absence of tort can be seen
with the Veterans Administration (VA) system. At VA hospitals, physi-
cians are protected from liability through the Federal Tort Claims
Act; patients are precluded from bringing suit against individual provid-
ers.7 But even without the threat of lawsuits, public regulation was
necessary to require disclosure of errors. Until a disclosure policy was
implemented in the 1990s, the VA system was “significantly under-
reporting patient safety incidents.”8 The experience of the VA indicates
that the embarrassment and shame from the publication of medical
errors is what prevents disclosure, not necessarily the fear of litigation
itself.

Markets and Market Failure

Quality differs in markets for an almost endless list of goods and services.
Some individuals are willing to pay more for a higher-quality good or
service; such higher quality is supplied if the additional number con-
sumers are willing to pay for such extra quality exceeds the marginal
cost of providing the extra quality. Higher quality may reflect a combina-
tion of better materials used to manufacture the product, more invest-
ment in product design to achieve a more attractive product, and more
resources devoted to looking for product defects during the course of
production. In addition, higher quality includes a greater redundancy of
inputs in the production process, greater product selection, more con-
sumer assistance in product selection, faster delivery, and more lenient
product exchange and return policy.
Societies have decided that for most products, quality provision and
quality assurance should not be left to markets alone, and governments
have intervened for various reasons. Governments do not, however,
regulate the quality of T-shirts or picture frames.
 Patient Safety and Medical Malpractice 193

The key rationale for government intervention in regulating quality is

asymmetric information between buyers and sellers of the product.
Sometimes, although full information is available to decision-makers in
the market, it is argued that these decision-makers form irrational deci-
sions and thus require the benefit of public regulation to constrain or
channel choices.9
Markets may fail when some parties to a transaction provide mislead-
ing or false information to the other parties or fail to mention facts that
these other parties may regard as pertinent to making a specific choice.
Consumers of medical care may not be well-positioned to evaluate char-
acteristics of alternative products, and certainly are not as well-
positioned as the seller of the product, giving rise to asymmetric
information. For such goods, an element of trust on the part of the pur-
chaser is required. Such goods are called “credence goods.”10
At least some types of medical care (e.g., emergency room visits,
complex surgery, and some kinds of hospital stays) are logically credence
goods. Since consumption of credence goods is inherently based on trust,
it seems unlikely that we would observe advertisements for B-quality
hospital care in the way that we observe advertisements for Motel 6.
Hospitals might claim that their care is of the highest quality, but they
would be reluctant to assert that their care is A-quality and their com-
petitor across the street offers B-quality care. These claims are sometimes
made in direct-to-consumer advertisements of pharmaceutical products,
but such ads are still comparatively rare (and could be the subject of a
lawsuit on the basis of false advertising, unless such claims are support-
able by strong empirical evidence).
The localized nature of hospital and physician services has limited the
amount of rating by national organizations and publications that evaluate
nationally (and internationally) marketed products, such as automobiles.
One can compare fuel consumption, speed, agility, and other characteris-
tics under controlled conditions. Frequency of repair is compared,
although different automobiles are often driven differently.11 Minivans
seem unlikely to be run in drag races as often as are some sports cars, or
if drag races are unseemly, around mountain curves, as in the television
advertisements for some cars. Yet owners of sports cars may be more
concerned about oil changes than are owners of minivans.
194 Chapter 8

Because quality comparisons are commonplace, automobile manufac-

turers need to consider marginal cost versus marginal return of increased
investments in quality, lest they market their products at uncompetitive
prices and lose market share.12 The Big Three automobile producers in
the United States have learned this lesson the hard way by losing sub-
stantial market share. This has driven them to improve the quality of
their products markedly in recent years. For automobiles, market forces
provide a strong incentive for manufacturers to make cost-quality
trade-offs.13
In the health care sector, there is nowhere near the same market for
products of varying quality as in most sectors. Quality varies, but the
relationship between product quality and price appears to be weaker
than for many goods and services. The localized nature of these markets
is only one reason for the difference. Other reasons include the cottage
industry nature of the sector,14 the complexity of the products, and the
important role of professional norms.
Frequency of repair in the context of automobiles has its counterpart
in adverse health outcomes that result from the receipt of medical care.
Yet while we are often willing to overlook differences in consumer
behavior when comparing frequency of automobile repair (e.g., fast
driving, poor maintenance), health care providers warn that comparisons
of rates of adverse outcomes may be misleading in the context of medical
care because patients differ in severity of illness in ways that are not
measured. Thus, for example, if hospitals are to be compared on the
basis of patient outcomes and some hospitals accept relatively sicker
patients for treatment, and the comparisons fail to account for these
differences, the reported differences in patient outcomes will be mislead-
ing (i.e., overstated). As a consequence, hospitals will seek patients who
will make them look good, and severely ill patients may be denied access.
Car dealers, in contrast to doctors and hospitals, would be foolish to
query potential customers about their driving records and habits.
However, this type of screening may well occur in the context of medical
care, especially when health outcomes are publicized without adjusting
the outcomes for differences in patient mix.
To the extent that patients (or even their doctors) are unable to judge
the quality of care of particular hospitals and doctors, health care provid-
 Patient Safety and Medical Malpractice 195

ers have had comparatively little to gain financially from reducing rates
of adverse outcomes. Of course, there may be nonfinancial reasons to be
concerned about quality (e.g., ethics, professional norms), but the finan-
cial incentive to improve quality seems to be less than it should be.
Health care executives often seem to view various investments in quality
as costly and unprofitable, which is consistent with the view that the
necessary financial incentives are lacking at present. Also, many hospitals
are still operated as freestanding, independent organizations, and physi-
cians’ practices are generally much smaller than hospitals. The high fixed
cost of investments in quality can be an important impediment to under-
taking such investments. Thus, absent market incentives, the role of
assuring quality is necessarily a function of public regulation and civil
litigation.

Adverse Events and Negligent Injuries

How the Estimates of Adverse Outcomes Have Been Generated

Since the mid-1970s, as noted in Chapter 1, there have been three major
studies of the epidemiology of medical injuries among patients hospital-
ized in the United States: studies conducted in California and New York,
and a follow-up to the New York study in Utah and Colorado.
In preparing its estimates of iatrogenic injury in U.S. hospitals for To
Err Is Human, the IOM used a group of experts to extrapolate the data
from these studies to the entire U.S. population. The IOM used its public
relations department to widely disseminate the report among members
of the public, bypassing the medical profession.15 Err was written to
present a message that everyone could understand (Box 8.1).
The message is indeed powerful. One can quibble with the estimates
and not agree with all of the report’s conclusions, but at the same time
not disagree with the basic message that medical errors are an important
public health concern.

The Quality of the Estimates

The estimates of deaths per annum in hospitals due to medical errors are
indeed “softer” than the IOM’s message implies. For one thing, medical
accidents, unlike motor vehicle and workplace accidents, are not discrete
196 Chapter 8

Box 8.1
Press Release for To Err Is Human

Experts estimate that as many as 98,000 people die in any year from
medical errors that occur in hospitals. That’s more than die from motor
vehicle accidents, breast cancer, or AIDS—three causes that receive far
more public attention. Indeed, more people die annually from medication
errors than from workplace injuries. Add the financial cost to the human
tragedy, and medical error easily rises to the top ranks of urgent, wide-
spread public problems.
To Err is Human breaks the silence that has surrounded medical errors
and their consequences—but not by pointing fingers at caring health care
professionals who make honest mistakes. Instead the book sets forth a
national agenda—with state and local implications—for reducing medical
errors and improving patient safety through the design of a safer health
care system. http://www.nap.edu/catalog/9728.html#description (accessed
July 3, 2006).

events. Many persons enter the hospital in a frail condition and hence
are particularly vulnerable to medical errors. But even if the error had
not occurred, many persons admitted to a hospital do not have a lengthy
life expectancy. For this reason, years of life lost would be a much more
precise characterization of the harm attributable to errors.
Negative outcomes do occur, and with an appropriate level of informed
consent, patients know or should know about these before agreeing to
undergo or to forgo a procedure. To determine whether an error has
occurred, it is necessary to parse the adverse event into a part that reflects
an error (and hence substandard care) and a part that presents an unfor-
tunate mishap that at the same time reflects an appropriate level of
care.
For each medical record, physician reviewers graded the confidence
that an adverse event occurred on a scale from 0 to 6.16 If the confidence
level exceeded 1, a judgment was made whether or not there was negli-
gence. Then, further notation was made as to the confidence level in this
judgment. There were two physician reviewers for each record, and
reviews were conducted independently. When there was disagreement
between the reviews, this was noted by a medical records analyst and
resolved through an independent review by a supervisory physician.
 Patient Safety and Medical Malpractice 197

No subspecialists in various fields served as medical record reviewers.

Thus, a general surgeon might have reviewed an adverse outcome result-
ing from the care of an ophthalmologist, for example. This was an
“implicit review,” meaning that it was up to the physician to make an
assessment of negligence without following explicit criteria. Especially
considering that these reviews were conducted at a time when quality of
care assessment was in its infancy, it would have been a practical impos-
sibility to develop explicit criteria for negligence for each type of adverse
outcome and specialty.
Courts, by contrast, have to consider only one case at a time. The
reviewers had to rely on what is written in a medical record. Courts, on
the other hand, have much more than a medical record on which to base
judgments. What is or is not recorded (or altered) in a medical record
may be a subject for consideration in a legal dispute.
Holding these medical assessments as the “gold standard” of levels of
quality of care and negligence would be to give them too much credit.
Yet what the study team accomplished was indeed very pathbreaking,
particularly at the time. Furthermore, such studies require substantial
resources and patience in order to conduct them at the standard at which
these studies were conducted.
In many cases, the underlying technology encountered in a particular
medical record review may not have been very complex, and determining
whether an adverse outcome occurred, and if that adverse event was
attributable to negligence, would not have been very difficult. The most
common adverse event in the New York study involved drug complica-
tions (19 percent), followed by wound infections (14 percent). Neither
may have required appreciable technical and specialized skills to
analyze.
For other alleged errors, expertise that is more specialized is likely
required. Errors in management were identified for 58 percent of adverse
events, with nearly half of these attributed to negligence. Failure to
diagnose was also common, as were adverse events in the emergency
room.17
A determination of whether or not there was negligence assumes that
the appropriate care standard has been defined. Sometimes there is a
consensus about an appropriate care standard, but often it is lacking, as
198 Chapter 8

attested to by the substantial geographic variation in patterns of care

that have been widely documented. Moreover, care standards often
change as technologies improve.
The reviewers could not have known about outcomes that became
evident months or even years after the hospitalization evaluated in the
study occurred. At the time of discharge from the hospital, there would
have been no adverse outcome. Many major injuries (e.g., birth injuries)
are latent for years; the fact that an injury had occurred becomes evident
only then.
Even though there are often several defendants in a tort case, liability
is generally attributed to individuals, not to organizations. But rather
than attribute most medical errors to individuals, the IOM report empha-
sizes the role of errors by medical systems. Because it focused on systemic
changes, Err states that it did not deal with tort liability or tort reform,
and deliberately excluded discussions of tort or medical malpractice. By
avoiding the legal aspect of medical error, the authors wanted to focus
the reform movement on patient safety—not on the resulting lawsuits.18
Likely an additional unstated reason for this absence is the contentious
nature of tort as a public policy issue. Focusing on system issues would
avoid some of the conflict inherent in discussions of tort.19

Err’s Recommendations and the Follow-up to Err: The Experience of

the First Five Years Post Err

The Recommendations
The report contains four basic recommendations for improving patient
safety. First, Err encourages creating a national focus on patient safety,
including establishing a Center for Patient Safety. Second, a national
mandatory reporting system would focus on serious adverse events
attributable to error. Implementation of voluntary reporting is also
encouraged for errors causing minimal or no patient harm. Information
from the reporting system would be publicly available. Such reporting
would supplement an existing system for reporting medication errors.20
Third, Err recommends raising performance standards and quality of
care expectations for health professionals, to be accomplished in part
 Patient Safety and Medical Malpractice 199

through direct public and private regulatory mechanisms, such as licens-

ing, certification, accreditation, and minimum performance levels for
providers and their employees. Err recommends that federal regulatory
agencies (e.g., the U.S. Food and Drug Administration) should increase
their performance standards.21
Fourth, Err recommends that safety measures be implemented at all
care sites, including systematic collection of data on injuries and imple-
menting systems for hospitals to track their own performance over time.
Health care providers are encouraged to implement redundancy or
backup systems, continuous training to deal with unexpected events, and
improved communication among staff. Various process measures to
offset human frailties are also encouraged, such as designing jobs for
safety (e.g., higher staffing ratios or fewer work hours), avoiding reliance
on memory and vigilance (e.g., use of checklists, protocols, automation
of procedures), employing constraints in the work process (e.g., a com-
puter in the pharmacy that requires entry of allergy medicines before the
prescription is filled), simplifying key processes (e.g., reducing the number
of handoffs or limiting choice of drugs in the pharmacy), and standard-
izing the work process (e.g., standard order forms, prescribing conven-
tions, and placement of supplies).

The First Five Years Post Err

The above recommendations all seem sensible at first glance, but
they raise the question of why it is taking health care providers so long
to implement them. Safeguards listed with the fourth recommendation
are widespread in other industries, such as in the airline and nuclear
power industries. Various forms of private and public regulation of
health care providers have been implemented, some mechanisms for
almost a century, including hospital peer review by medical staffs, certi-
fication, accreditation, and licensure. An argument advanced by propo-
nents of a pluralistic health care financing system in the United States is
that it would assure the public, or at least some of the public access to
high-quality care, assuming that this level of safety, and the additional
cost of implementing it, would be demanded and paid for by well-
informed consumers.
200 Chapter 8

Finally, though it is not discussed in Err, there is little in the recom-

mendations that would imply that the existence of tort liability prevents
medical errors. Providers oppose reporting of errors on grounds that
such reports might be used against the provider in litigation. But there
are other reasons for opposing such reporting, such as simply looking
bad to outsiders. Firms in the other industries do not admit to their errors
largely for business reasons.
Given the publicity following publication of Err, administrators in
private and public sectors, and legislators at both state and national
levels, came under substantial pressure to address and implement safety
initiatives. The pressure was amplified by the report’s ambitious objective
of a 50 percent reduction in errors within the five years following
publication.
Hospitals subsequently adopted measures to reduce and prevent
adverse events; for many this included creating a new budget dedicated
to patient safety or increasing an existing patient safety budget.22
Nevertheless, a major barrier hindering development of hospital patient
safety programs is lack of adequate funding. In large part, delays reflect
a lack of strong financial incentives for hospitals to implement patient
safety programs.
There is no way to tell whether there was a 50 percent reduction in
errors within five years of Err’s release. Administrative data collected for
other purposes are the only potential (albeit highly imperfect) data
source.23
Broad social goals are specified relatively easily. The devil lies in the
details of actual implementation and in measurement. The best one could
hope for from this recommendation is that individual health care orga-
nizations would decide that error reduction is a high priority and insti-
tute programs and measurement systems to accomplish this objective.
Despite the lack of funding, several specific patient safety measures
recommended in Err have been implemented on a limited scale, including
(1) reduced work hours for medical residents,24 (2) error reporting
systems, and (3) U.S. Food and Drug Administration (FDA)-required bar
coding.
Error reporting has received a substantial amount of attention. The
value of consistent, honest error reporting per se is not to be disputed,
 Patient Safety and Medical Malpractice 201

particularly as a first step in error reduction. And the threat of tort may
be one of the factors inhibiting honest and accurate reporting.
Error reporting is a necessary but not a sufficient condition for improv-
ing patient safety. For one thing, compiling of error reports does little
other than create an illusion of open communication and disclosure.
Error reports must be reviewed and acted upon in order to have any
value. Standing on their own, they report single adverse events, fre-
quently offering no clear remedies.
Particular types of errors are common to many institutions. In
this sense, error determination is a public good.25 It should be unneces-
sary for each and every organization to experience each particular
type of error before anything is done about it.26 If a representative
number of hospitals were to accurately report the sources of errors, this
should suffice for learning about sources of errors in hospitals. But cur-
rently, each hospital has an incentive to wait for the other hospital to
do the reporting (even if the identities of the reporting hospitals were
confidential). For this reason, there is a case for public intervention since
private markets will not supply this information in adequate amounts
unless they are paid or are required to do this by a government
authority.
State legislatures have increasingly become involved in patient safety
policy. When legislatures have been involved in requiring or encouraging
error reporting, adverse events have been defined quite narrowly, effec-
tively limiting reports to events resulting in death.
However, this captures only a fraction of adverse events. Using such
limited data, hospitals have been unable to identify patterns of error.
Also, inclusion of near misses in addition to actual errors would be
useful information to have, even if such errors were released only to
hospitals.27
A more fundamental challenge to the success of error reporting
is actually obtaining the reports. Several states had enacted statutes
creating reporting systems, but rather than expect hospitals to use
their own funds for this purpose, the decision was made by the legisla-
tures that public funding is appropriate; however, states faced funding
problems of their own during the years immediately following release of
Err.28
202 Chapter 8

Some states have had reporting systems since the early 1970s.29 But in
these states, underreporting has been a chronic problem. New York’s
reporting system, implemented in 1985, receives 20,000 reports
annually. California has had an error reporting system since 1972,
but as of the late 1990s, the state received fewer than 4,500 reports
annually, despite the fact that it has almost twice the population of
New York.30
The third important innovation since Err, with much promise for
improving patient safety, is the adoption of bar coding. Drug errors are
common, and many arise from predictable mistakes (e.g., unintelligible
written orders, pulling the wrong medication for a patient, or neglecting
to consider the patient’s history when prescribing).31
In 2004, after several proposed rules, comments, and revisions, the
U.S. Food and Drug Administration released a bar code regulation. The
regulation required that standardized bar codes be used on prescription
drugs, blood products, vaccines, and over-the-counter drugs under
medical orders in hospitals.32
However, the regulation has a major deficiency; the FDA lacks the
statutory authority to regulate hospitals. Thus, in the end, it is the hos-
pital’s decision to acquire scanning equipment. Presumably, if and when
use of bar codes becomes the community standard, hospitals that choose
not to purchase scanning equipment will be at increased risk of tort
liability.33
One study of VA hospitals reports a 24 percent decrease in drug-
related errors soon after implementation of its national bar-coding system
in 1999.34 But despite the success of the VA, diffusion of bar coding in
U.S. hospitals has been slow. By the mid-2000s, only about 5 percent of
hospitals in the United States had implemented such systems, mainly
because of their high cost.35 Also, the reduction in adverse events from
use of bar coding may be low.36
The Leapfrog Group estimated 567,000 serious medical errors
would be avoided per year with the use of computerized physician order-
entry (CPOE) systems.37 But five years post Err, only 5 to 9 percent of
U.S. hospitals had implemented CPOE.38 The slow rate of diffusion
reflects the considerable expense of acquiring equipment and software
and then making a system operational.39 Some hospitals chose to allocate
 Patient Safety and Medical Malpractice 203

their entire patient safety budget to CPOE, leaving inadequate funding

for addressing other safety efforts.40 Although implementation cost is
high, the potential for subsequent cost savings appears to be
substantial.41
Even though the purpose of CPOE is to reduce errors, it has led to
new errors being made.42 Technical aspects of the system can make use
difficult; multiple screens appear, sometimes without the patient’s name,
causing confusion as to which patient is to receive the order; and the
actual input of the order can be tricky.43
One source of error in hospital settings potentially made worse by
CPOEs is the lack of coordination of care of hospitalized patients.44
Patients inevitably are seen by multiple physicians, some of whom are
unfamiliar with the patient. With CPOE, each physician inputs instruc-
tions, which may result in contradictory treatment plans; reconciliation
by machine is difficult.

Why Is There Underreporting of Errors?

Underreporting in mandatory reporting systems has several potential

causes. However, much has been both said and written suggesting that
the major factor is the threat of tort. Legal discovery of error reports is
thought to prevent full disclosure by many providers.45 But a study of
nineteen states with mandatory reporting systems46 reveals several reasons
for underreporting other than a fear of increased liability, including lack
of effective internal systems within hospitals to identify incidents; insuf-
ficient resources to implement reporting; unclear definitions or require-
ments for what must be reported; lack of enforcement of mandatory
reporting by the state; and a perceived lack of benefit from reporting to
the facilities themselves.
The Health Care Quality Improvement Act of 1986 (HCQIA) provides
limited protection in support of disclosure and error procedures. While
it grants immunity for peer review groups, it does not extend this immu-
nity to the documents they produce.47
State laws offer slightly more defense from discovery. As of 2003,
every state except New Jersey had statutory protections for discovery of
records and deliberations of peer review committees.48 These statutes
204 Chapter 8

vary widely among states; as a result, protection from discovery may be

unreliable. Several statutes protect only the documents generated by the
peer review committee, while others protect all the information provided
to peer review committees. For example, Texas protects all materials and
information, including, but not limited to, root cause analyses, annual
hospital reports, action plans, and departmental summaries.49 Neverthe-
less, even with statutory protections, information may be discoverable
for other purposes, such as allegations of negligent supervision or
credentialing.
Kesselheim et al. (2006) assess several existing and emerging physician
clinical performance assessment (PCPA) initiatives in light of legal rules
governing the types of information that can be used as evidence in civil
litigation. Their basic conclusion is that evidence from such initiatives
cannot be introduced in court, but the conclusion is somewhat
guarded.50
For reporting to be effective, the reporting process needs to be clear
and easily done. Lack of enforcement of reporting requirements and/or
lack of effective use of the information provided reduces the incentives
for facilities to report errors. Some states offer money incentives, in the
form of grants, to entice hospitals to comply.
Even when the confidentiality of individual reports is protected, hos-
pitals often worry that release of error-reporting data will lead to loss
of reputation for hospitals with high reported error rates.51 If they are
successful in increasing their rates of error-reporting, when the state
subsequently releases a summary of error-reporting data from all the
hospitals in the state, it will appear that they are a dangerous hospital
because of the high number of error reports.
Addressing this concern, a few states have created patient safety
organizations (PSOs) to serve as a safe harbor for error reporting.52
As independent nongovernmental organizations,53 they can give
medical professionals a vehicle for reporting errors without the threat
of repercussions. Recently enacted federal legislation extends this
protection by safeguarding information that is voluntarily reported
to these organizations, thereby providing hospitals and providers
with a reliable system for confidential reporting of adverse medical
events.54
 Patient Safety and Medical Malpractice 205

The U.S. Patient Safety and Quality Improvement Act of 2005 protects
providers’ communications with a PSO, creating a $10,000 civil penalty
for each illegal disclosure and preventing accrediting bodies from taking
action against a provider’s good-faith involvement in the PSO.55 This law
prevents any information confidentially given to a PSO from being used
against medical professionals and other administrators in malpractice
suits. However, the mandatory reporting laws of the states are recog-
nized and respected by the Act, meaning providers would be required to
file using the state-mandated procedure. This significantly narrows the
broad protection offered by the Patient Safety and Quality Improvement
Act.
Regardless of whether or not there is accurate reporting, the process
of reporting to an external body is likely to be ineffective unless states
implement a system that provides timely feedback to providers and
hospitals. If a response is prompt, it conveys the message that
error reporting is valued and, more important, useful. Releasing
safety reports to the public can also create incentives for hospitals
to change their policies, or, it can create a disincentive for reporting of
errors. To prevent the latter, New York released facility-specific reports
only for facilities with very low reporting rates, in order to encourage
reporting.56

Implications
To Err Is Human has had an important effect in calling attention to
patient safety issues. The focus of these efforts has been on hospitals
because many patient safety systems involve large initial investments.
Due to the large fixed costs, mid-sized and large hospitals may be more
efficient in implementing patient safety systems than smaller hospitals
and most physicians’ practices. Thus, not surprisingly, the majority of
adopters have been hospitals.
The record post Err demonstrates practical pitfalls in implementing
systems and approaches to reduce error rates. Although policies have
emphasized public disclosure, what is really needed are incentives to
change the internal mind-set of health care organizations about
patient safety, assuming that higher levels of patient safety are indeed a
public objective worth the extra cost of achieving this goal. It seems
206 Chapter 8

unlikely that public regulation alone can achieve this objective. Medical
malpractice cannot bear the total burden of ensuring much higher levels
of patient safety.

Quality Assurance Mechanisms, Government Oversight, and Regulation

Why Other Regulatory Mechanisms May Fail

Medical malpractice is only one of several quality assurance mechanisms.
The trial bar has argued, with considerable justification, that there would
be no role for tort liability if the other quality assurance mechanisms
functioned well.57 However, absent effective private mechanisms and
government intervention, tort liability with contingent fees gives injury
victims, irrespective of their financial status, a way to address grievances
and, hopefully, prevent injuries for someone else. Based on the resolution
of the legal dispute, a warning is issued about the consequences of failure
to exercise due care. This is a contentious issue with lawyers for the
defense.
Since there has been criticism of medical malpractice, why might the
other quality assurance mechanisms fail as well? There are several
reasons.
First, public bureaucracies may be beholden to special interests or
simply unresponsive, given the internal incentives of staff or understaff-
ing. More generally, the capture theory of public regulation implies that
regulation is often used to protect existing sellers from entry rather than
to pursue the stated goals of regulation.58 Second, given the complexity
of medical care, it is difficult for public agencies or private credentialing
organizations to oversee all aspects of care that may potentially affect
quality. Third, imposing and enforcing minimum standards runs the risk
of denying care to persons in areas where care is generally inaccessible.
For example, applying minimum volume standards to hospitals in rural
areas might often result in particular types of care being geographically
inaccessible to residents of such areas. Finally, individual health care
providers lack an incentive to honestly disclose indicators of quality, in
large part because their competitors may not be as truthful and thus gain
a competitive advantage. This is an important rationale for public dis-
closure requirements.
 Patient Safety and Medical Malpractice 207

Other Regulatory Mechanisms: Licensure, Certification, Peer Review,

Credentialing, Report Cards, and Disclosure Statutes
The oldest quality safeguard is licensure by state governments. Licensing
of health professionals varies by state because the system is based on
individual licensing boards for each profession in each state. Licensure
is at best a weak safeguard of quality. Unless a physician moves to
another state, licensure is a once-and-for-all process.59 Loss of license is
rare and is in response to very major lapses and major misconduct. Sloan,
Mergenhagen, Burfield, et al. find that even physicians with very high
and persistent medical malpractice rates were investigated by the licens-
ing board only very rarely.60 The definition of what constitutes a com-
plaint and how it is handled differs according to individual licensing
board standards.
However, as Ameringer documents in his excellent historical account
of the licensure process, state licensure boards have changed, largely in
response to political pressures to serve the public interest.61 Medical
malpractice in the United States predates medical licensure and state
medical boards.
Certification, on the other hand, speaks to a physician’s educational
credentials and practical training, and imposes some requirements for
continuing medical education, but continued certification is not linked
to direct measures of physician performance. Physicians may be certified
through twenty-four different specialty boards. Not all of these boards
require recertification, and those that do, typically require it only every
seven to ten years.62
Peer review in ambulatory settings occurs informally. By contrast, hos-
pitals have a formal structure for peer review. One rationale for hospital
peer review by medical staff, albeit a comparatively modern one, is that
some internal mechanism for quality monitoring is needed if hospitals are
really to compete. However, there are risks to such peer review. In par-
ticular, medical staff peer reviewers are likely to be competitors with the
physicians they review. Banning a physician in part or in full may allow
peer reviewers to improve their own competitive position.63 Physicians
denied hospital privileges have often sought legal remedies on grounds
that the medical staff’s decision was anticompetitive. Also, to what extent
peer reviewers monitor medical errors is unknown.
208 Chapter 8

As discussed above, peer review statutes offer some protection to docu-

ments, but the protection varies by state, and information can still be
discovered in some civil actions, such as when there are allegations of
negligent supervision or credentialing. At both the state and the federal
levels there are statutory protections by way of privilege, confidentiality
provisions, and limited immunity for the participants and the work
product of the peer review committees.64 In sum, statutes provide only
limited protection.
Credentialing is another quasi-regulatory mechanism aimed at quality
assurance.65 Hospitals must be accredited by JCAHO or undergo regula-
tory review in order to participate in Medicare, and with 40 percent of
revenue from Medicare payments, hospitals have great incentive to
comply with JCAHO standards.66
Although JCAHO and other credentialing organizations, including the
U.S. Centers for Medicare and Medicaid Services, have made and con-
tinue to make important strides in quality of care assessment, impor-
tantly moving a reliance on structure and process of care indicators to
evidence-based quality of care indicators and health outcome-based indi-
cators, much work remains to be done.
Bradley et al., studying data for 2002–2003 from 962 hospitals par-
ticipating in the National Registry of Myocardial Infarction, find that
although process of care measures67 are positively correlated (if a hospital
uses one, it is more likely to use the others), all of the publicly reported
process measures for acute myocardial infarction (AMI)68 explain only
6 percent of the variation among hospitals in the thirty-day mortality
rate for AMI. Clearly, doing just a few things right is not going to make
that much difference in outcomes.
Specialty societies can play an important role in promoting patient
safety. Most prominently, anesthesiologists, faced with high mortality
rates associated with procedures they perform and high insurance pre-
miums reflecting high claims frequency, took action. They studied the
frequency of claims, identified the errors, and implemented corrective
measures.69 The American Society of Anesthesiologists created mandated
patient monitoring standards, redesigned procedures and equipment to
decrease the severity and frequency of errors, shortened residents’ hours,
set practice guidelines, standardized the operation of machines, and
 Patient Safety and Medical Malpractice 209

created safety devices for the machines These actions reduced the mortal-
ity rate from 1/10,000–20,000 to 1/200,000 in around a decade.70
Mandatory error reporting has been a popular safety initiative.71 The
reports of adverse outcomes can be used by state regulatory agencies to
identify where errors are occurring. If reporting systems provide timely
and complete information to regulators, this should, at least in principle,
improve the agencies’ role as monitors. However, reporting alone does
not alter the incentives such public agencies face. If these agencies are
not motivated to take action, reporting will not make this more likely.
The other rationale for mandatory reporting is to inform the public
about variations in quality of care so that more informed consumers can
make better health care consumption choices.72
At first glance, mandatory reporting would seem to enhance efficiency
of consumer decision-making. But there are at least two concerns. First,
health care providers worry that reporting adverse outcomes will stimu-
late medical malpractice suits. These worries appear unfounded. Marchev
reports that public officials in states with mandatory reporting did not
find a pattern of increased medical malpractice litigation following
implementation of mandatory reporting, a situation possibly attributable
to the concurrent enactment of data protection statutes.73
Another barrier to informed consumer choice is that most states that
released incident-specific information did so only upon request.74 Requests
were made after the incident occurred, implying that incident-specific
reporting cannot do much to inform consumers prospectively (i.e., in
advance of their health care decisions).
There is an additional concern relating to a consumer’s ability to make
informed decisions even if detailed facility-specific information is pro-
vided prospectively. The information revealed by observing an adverse
outcome critically depends on the risk of adverse outcome the patient
had before the procedure. There is plausibly considerable heterogeneity
in such risk. As noted at the beginning of this chapter, if adverse out-
comes are reported but there is no adequate adjustment for patient risk
ex ante, providers will have an incentive to select healthier, low-risk
patients for treatment in order to improve their report records.
Dranove et al., focusing on cardiac surgery report cards in New York
and Pennsylvania, find evidence that patients more likely to survive
210 Chapter 8

surgery tend to be selected as a consequence of public disclosure of

facility-specific mortality rates following coronary artery bypass
surgery—just as the critics of public disclosure fear.75 By contrast, Jin
and Leslie, who study the effect of introducing hygiene quality grade
cards in Los Angeles in 1998, find that grade cards caused restaurant
health inspection scores to increase, consumer demand to become sensi-
tive to changes in restaurants’ hygiene quality, and the number of food-
borne illness hospitalizations to decrease.76
A reason for the difference in findings is plausibly that restaurants have
much more difficulty gauging the risk of patrons’ getting food-borne ill-
nesses when they enter the restaurant. A cardiac surgeon is in a much
better position in advance of surgery to gauge the risk of an adverse
outcome for a patient with heart disease than is a restaurant to know,
at the time a patron is seated, whether he or she is most likely to get
sick.
As discussed earlier, a major problem with mandatory reporting
systems is underreporting. A major, often overlooked, aspect of under-
reporting is that adverse events that occur after discharge are usually not
reported. To capture adverse events that are delayed or occur after dis-
charge, the revision of the American Association for Accreditation of
Ambulatory Facilities standards requires any deaths within thirty days
after a surgical procedure at an accredited facility be reported.77 Care
provided in ambulatory settings is by law subject to regulatory scrutiny,
and states are including outpatient facilities in their mandatory reporting
laws.78 If inpatient facilities are covered, ambulatory facilities should be
covered as well.

Implications
Each quality assurance mechanism is deficient in some respect. None will
do the job of assuring patient safety by itself. What is also lacking in
health care is a market for quality. For too long, health care providers
have insisted to the public that “we are all good.” One reason that
medical malpractice is so widely despised among many health profes-
sionals is that lawsuits seem to contradict this assertion. However,
for tort law to promote patient safety, some fundamental changes are
needed.
 Patient Safety and Medical Malpractice 211

Restructuring Tort Law to Improve Patient Safety

Changes in Tort Law and Medical Malpractice Insurance Needed

Tort law does not seem to have had a major role in improving patient
safety, and some would say that tort law/medical malpractice are a con-
tradiction in terms, plausibly for several reasons.79 Two major changes
are needed to give tort law and medical malpractice insurance a positive
role in promoting patient safety. First, it seems unlikely that the scale in
many practice settings is sufficient to implement suggested patient safety
methods. Perhaps large physician groups can take advantage of scale
economies, but most physicians do not practice in large groups. Second,
there needs to be a financial penalty for making errors. Currently, the
main deterrent effect of malpractice suits for an insured provider is the
time and aggravation involved. By contrast, in other areas—motor vehicle
liability,80 workers’ compensation,81 unemployment insurance,82 and, in
a more minor area, dram shop liability83—premiums rise following a
paid claim. There is considerable empirical evidence from these other
areas that when decision-makers are exposed to higher premiums fol-
lowing paid claims, care levels increase.

Role of Enterprise Liability

Enterprise liability84 first gained prominence as a medical malpractice
reform in the 1990s. Under this approach, the focus on medical malprac-
tice litigation would be shifted from individual health care providers,
most frequently physicians, to health care organizations under whose
auspices care is delivered. The two major candidates for enterprise liabil-
ity are hospitals and health plans, the latter being the proposed enterprise
under the Clinton health plan.
Enterprise liability has several attractive features. It reduces the number
of defendants, which potentially reduces the time for dispute resolution
and litigation cost. A single defendant focuses the responsibility for
quality of care and simplifies settlement negotiations.85
Probably the most important advantage is that enterprise liability
potentially allows integration of patient safety activities with medical
malpractice insurance, implementing systems-based loss control
mechanisms and quality assurance programs not feasible for a smaller
212 Chapter 8

organization. By combining the function of preventing injuries with that

of insuring loss if and when in fact injuries do occur, the way to injury
prevention is combined with the willingness to do so (an adaptation of
an old cliché). The question of whether hospitals or health plans should
be this organization has been debated, but, given the managed care
backlash of the late 1990s, the prospects for health plan-based enterprise
liability are very slim.86
As for hospitals, their size, resources, and status as continual defen-
dants places their systems in an excellent position for liability.87 Shifting
liability to a hospital or hospital system would also create deterrence and
provide incentives to implement systemwide safety measures. Hospitals
are already structured to take precautions including monitoring of
medical staff; but hospital enterprise liability would create a greater
financial incentive for peer review at the hospital level and for overseeing
care levels of the members of its medical staff. It is likely that hospitals
would become more cautious in recruiting their medical staff members
than they are currently.
While enterprise liability provides a vehicle for creating incentives to
introduce patient safety measures from the party that is in the best posi-
tion to introduce them, there is a widespread concern among practicing
physicians about a loss of professional autonomy. While these concerns
are understandable from the perspective of individual physicians, patient
safety is incompatible with autonomy of the individual professional.
How can a hospital promote patient safety if each physician member of
the medical staff practices totally independently?

Experience Rating
Arguably the largest obstacle to producing and implementing safety
measures is the lack of financial incentives. Neither market forces nor
the threat of tort liability seems to provide sufficient incentives. An
important reason that tort liability has not been effective in promoting
patient safety is that medical malpractice insurance shields potential
defendants from the financial burdens of being sued. Such insurance
tends to be complete;88 that is, there are no deductibles or coinsurance,
and, at least for physicians, liability limits of coverage are rarely exceeded.
 Patient Safety and Medical Malpractice 213

To reduce the effect of such insurance further, experience rating of pre-

miums is rare for medical malpractice insurance.89
Thus, in general, physicians with relatively adverse medical malprac-
tice records pay the same premiums as others.90 For hospitals, however,
the situation is different. Most hospitals self-insure, at least for primary
liability coverage.91 For this reason, they bear much of the financial risk
of being sued.
There are several reasons for the lack of experience rating of
medical malpractice premiums. The most plausible one has already
been stated: that the data on losses at the level of the individual phy-
sician practice tend to be insufficiently numerous for insurers to
develop sufficiently accurate premiums. In the context of motor vehicle
liability, accidents and violations are sufficiently common to allow insur-
ers to develop fairly reliable premiums at the level of the household. By
contrast, medical malpractice claims are rare. In a given year, few physi-
cians are sued. And equally important, in contrast to motor vehicle
claims, medical malpractice claims often take many years to resolve.
Experience rating should be based on paid claims, not on claims that are
filed and subsequently dropped by claimants or lost by claimants at
verdict.
Research clearly shows that, actuarially, physicians within a specialty
should be charged appreciably different premiums. Studies have analyzed
claims frequency as a Poisson process.92 Based on the within-specialty
heterogeneity, the studies obtain very different expected claims rates for
physicians within specialties, suggesting that observed differences in
claims rates reflect more than a random process as well as systematic
physician-specific differences in the propensity to be sued.
Simulations conducted by Sloan and Hassan indicate that most hos-
pitals have a sufficient number of medical staff members for experience
rating at the hospital level to be feasible.93 Even small hospitals, given a
sufficient record of past claims (five or six years), could be an experi-
ence-rating unit. In general, experience rating is based on differences in
expected claims frequency. Sloan and Hassan find that considering dif-
ferences among physicians in claims severity as well as frequency suggests
that physicians with adverse claims frequency and severity should (on an
214 Chapter 8

actuarial basis) be paying surcharges of 400 to 500 percent. Malpractice

insurers would resist imposing such surcharges based on the survey evi-
dence presented above.
Experience rating at the level of the hospital allows hospitals to
combine financial incentives, including surcharges on medical staff, with
adverse claims experience, but given medical staff organization, it should
be possible to use other methods of deterring injuries as well.

Discussion and Conclusions

There is now a widespread consensus that patient safety in health care

is not what it should be. This conclusion is admittedly somewhat impres-
sionistic, and is not based on rigorous analysis of the benefits versus the
costs of higher levels of patient safety. Even the estimates of deaths
attributable to errors in hospitals are questionable. Whatever the optimal
level of patient safety is, imposing tort liability is at best part of the
solution.
This point is made in a very effective way by the authors of Freako-
nomics (see box 8.2). They describe the steps administrators took at
Cedars-Sinai to improve hand-washing compliance. Certainly there must
be an easier way than tort to encourage health professionals to wash
their hands. And it is a bit difficult to hold tort responsible for the current
low rate of compliance with hand hygiene regimens.
Market forces potentially play an important role, as do internal struc-
tures such as peer review by a hospital’s medical staff, private credential-
ing, and public regulation. It is inappropriate to blame tort for all of or
even most of the quality of care problems currently facing the U.S. health
care system. When tort is blamed, the effect seems to be to justify the
status quo. It seems unlikely that absent the threat of tort, physicians
would become like idealized Stakhanovite workers94 of the Soviet Union
of the 1930s—selfless and industrious individuals interested solely in
serving the public good and freely admitting errors, without being
motivated by individual financial incentives. True Stakhanovites did
not exist in practice, at least not in more than trivial numbers. Likewise,
it seems unlikely that absent the threat of tort, physicians and other
health workers would admit to their errors in large numbers. If there is
 Patient Safety and Medical Malpractice 215

Box 8.2
“Freakonomics” and Patient Safety

The notion of washing hands between patients has been around for about
a century and a half. Yet as Dubner and Levitt (2006), the authors of
“Freakonomics,” indicate, the practice has been difficult to enforce. There
was a problem with inadequate rates of hand washing at Ceders-Sinai
Hospital, a highly respected institution. The hospital decided that it needed
to devise an incentive that would increase compliance without alienating
doctors on its medical staff. At first, administrators reminded doctors
about washing, using e-mail messages, posters, and faxes. This approach
did not seem to work. Then they started a Hand Hygiene Safety Posse that
went around the hospital offering a $10 Starbucks card when they found
a physician washing up. Nurse observers reported that following imple-
mentation of this incentive, compliance with the hand-washing regimen
increased to about 80 percent from 60 percent.
The hospital’s leadership was not satisfied with this improvement since
the chief accrediting body required a compliance rate of 90 percent. Thus,
at a meeting of the Chief of Staff Advisory Committee, which consisted
of about twenty persons, mostly physicians, the hospital’s epidemiologist
cultured each of the committee members’ hands. The photographic images
were disgusting. One photograph was made into a screen saver that
appeared on every computer at the hospital. Following this, hand hygiene
compliance rose to nearly 100 percent (Dubner and Levitt 2006).

no one around to notice an error, why admit it? Individual incentives

for physicians should be aligned with the organizations with which they
work.
Increased premiums may themselves facilitate bringing physicians
under the hospital’s insurance umbrella. Mello, Kelly, Studdert, et al.
report that following sharp increases in premiums in Pennsylvania, West
Virginia, and Florida in 2002, some hospitals were directly employing
physicians in high-premium specialties.95 Other hospitals were arranging
for physicians to obtain medical malpractice insurance through the hos-
pital’s insurer and making alternative arrangements for physicians on
their medical staff to find coverage and to subsidize their premiums.
9
Medical Malpractice Insurance and
Insurance Regulation

Why Study Medical Malpractice Insurance and Insurance Regulation?

At first glance, insurance and its regulation seems to be a very unexciting

topic, one best left to the specialists. However, sudden decreases in avail-
ability and sharp increases in the price of medical malpractice insurance
are the immediate precipitators of each crisis. A frequent response to
sharp price increases is to blame the companies whose product prices
have increased. Witness, for example, accusations that multinational oil
companies are gouging consumers following oil price shocks. Respond-
ing to widespread criticism, the companies attribute the price increases
to factors beyond their control and, at the same time, offer advice to
consumers as voters and to political officials that imposing price con-
trols and/or withdrawing tax advantages will adversely affect the com-
panies’ access to capital, and hence adversely affect industry capacity in
the long run.
As with medical malpractice, there have been spells when oil products
were simply not available, especially during the 1970s. Oil supply disrup-
tions loom on the horizon, mainly due to geopolitical events. Although
they were never implemented, the federal government has had plans to
ration scarce oil products—most notably during the Carter administra-
tion and previously during World War II.
There are parallels between oil/gasoline and medical malpractice
insurance, but there are also important differences. As with oil, there
have been allegations that insurers in general, and medical malpra-
ctice insurers in particular, gouge consumers, especially following
sharp premium increases. More than oil, insurance has been subject
218 Chapter 9

to a considerable amount of regulation. Unlike oil, which in many

high-income countries, including the United States, has remained in
private hands, with for-profit being the only ownership form, public
and quasi-public provision has sometimes been substituted for private
provision of medical malpractice insurance. Even more commonly,
alternative private organizational forms have substituted for the for-
profit form.
Insurance regulation seeks to prevent insurer bankruptcies, which can
cause a disruption in the flow of funds to insured individuals and orga-
nizations that have incurred losses and to counter exploitation of market
power by insurers, which absent a countervailing force might result in
premiums set above the level needed to attract capital to insurers. To a
far greater extent than with oil, one potential source of market power
in some insurance markets is a lack of consumer information about what
is being purchased and about financial and nonfinancial characteristics
of sellers of insurance in a given market. However, lack of consumer
information is less of a policy issue for medical malpractice than for
some other lines of insurance, such as motor vehicle, homeowners, and
life insurance.
There are three important interrelated questions. First, how well do
markets for medical malpractice insurance function? Second, to the
extent that there is a case for government intervention in these markets,
how effective have such interventions been? Third, what types of policy
changes, if any, are indicated?

Important Features of Medical Malpractice Insurance and the

Regulatory Environment

Since medical malpractice insurance is almost always complete—no

cost-sharing is imposed on insured physicians when they incur medical
malpractice losses,1 and experience rating of premiums is relatively
rare2—physicians are largely protected from the financial consequences
of lawsuits. By contrast, hospitals often are self-insured for much of
medical malpractice loss, and they purchase reinsurance for protection
against very large losses; such reinsurance tends to be highly experien-
cerated. Insurers of physicians rather than physicians themselves
 Medical Malpractice Insurance and Insurance Regulation 219

purchase reinsurance.3 Thus, in contrast to physicians, hospitals do bear

the monetary costs of medical malpractice lawsuits.
Medical malpractice insurance is a form of property-casualty insur-
ance. Property insurance is first-party insurance. Casualty insurance is
third-party insurance. It covers the obligation that an insured individual,
business, or other organization incurs by negligently causing personal
injury to, and/or property loss by, another person. The most common
type of property-casualty insurance is fault-based automobile liability.
In comparison, medical malpractice insurance is a very small line of
casualty insurance.
Firms providing medical malpractice insurance may be organized in
several alternative forms. Some diversity in ownership form is seen
among property-casualty insurers more generally. As in other lines of
insurance, some firms are organized as stock companies, mutuals, recip-
rocals, and direct underwriters (e.g., Lloyd’s associations).4 Ownership
forms other than the conventional stock form are much more common
for medical malpractice insurance. Authorizing new organizational forms
of insurance was a major part of the insurance reform movement of the
1970s.
Although the provision of insurance is fundamentally a private activity
in the United States, there is also a long history of both federal and state
governments serving as insurers.5 The federal government has taken on
the largest obligations, such as covering losses from natural disasters and
insolvent banks; state governments provide workers’ compensation cov-
erage, hail insurance, unemployment insurance, and coverage plans for
uninsured motorists, to name a few. All states have insurance depart-
ments that regulate insurers within their jurisdictions and have regulated
insurance market for decades (box 9.1), though policies and practices
differ substantially among states.6
Insurance regulation is overwhelmingly the responsibility of state gov-
ernment; generally insurers must be state-licensed to conduct business.
In spite of the challenges posed by state regulation, insurers often operate
on a regional or a larger scale. By contrast, until quite recently, many
medical malpractice insurers limited their sales to the states in which they
were domiciled, especially those companies originally sponsored by state
medical societies, which until recently operated in a single state.
220 Chapter 9

Box 9.1
Insurance Market Interventions by State Governments

• Regulation of solvency, premiums, policy forms, underwriting prac-

tices
• Creation of new forms of private insurance, such as mutual and recipro-
cal companies, as an alternative to conventional stock insurers
• Authorization of pooling arrangements, such as joint underwriting asso-
ciations, to provide coverage to potential purchasers that otherwise would
have difficulty in obtaining insurance, and guaranty funds which cover
losses incurred by policyholders in the event of insurer inability to pay
(bankruptcy)
• State patient compensation programs, which provide state-issued medical
malpractice insurance above specified dollar thresholds or for persons who
have experienced particular types of medical injuries
•State-funded indemnity coverage for individuals who are in an employ-
ment relationship with the state

Two factors contributing to a state-specific orientation are (1) state

regulation and (2) legal liability is most often subject to state law inter-
preted by state courts. The same factors pertain to other lines of casualty
insurance; these other lines tend to have multistate insurers as dominant
sellers. But in these markets, industry interest groups in the states have
not generally organized their own insurance companies.
Most medical malpractice coverage is sold on an individual basis, not
to large groups or larger incorporated business entities. Relatively few
physicians are covered by the hospitals in which they practice, although
hospitals are an entity that naturally groups individual physicians
together.7

Rationale for Insurance Regulation

Regulation is an important feature of insurance markets of all types.

Four major justifications are (1) to prevent bankruptcies by maintaining
insurers’ solvency; (2) to keep premiums at nearly actuarially fair
rates by preventing insurers from exercising market power; (3) to assure
availability of coverage by implementing laws and regulations to assure
 Medical Malpractice Insurance and Insurance Regulation 221

availability; and (4) to supply information to insurance purchasers to

improve the performance of insurance markets and assure fair play.
Historically, the main roles of insurance regulation were to assure
solvency of insurers in advance of insurer bankruptcy and to mitig-
ate losses of policyholders in the event that bankruptcies occur. The pur-
pose of insurance is to mitigate a policyholder’s financial risk; thus, if
and when an insurer goes bankrupt, in the end the insured party has
not really been relieved of any risk. Additionally, individual purchasers
may have insufficient information, combined with too much complexity
for consumers of insurance to know which insurers are less likely
to become insolvent.8 Toward this end, regulators monitor whether
premiums are adequate to cover expected loss, with a reasonable
load to cover the cost of administering the plan, and whether or not
insurers are taking unreasonable financial risks with their investments.
Ex ante (before bankruptcy occurs), government oversight involves
seeing that premiums are adequate to cover anticipated losses and
expenses, and that the asset mix is not too risky. Another task of regula-
tors is to restructure insurers for which insolvency seems imminent. Ex
post mitigation of loss includes government requirements that insurers
contribute to a risk pool or guaranty fund, which compensates policy-
holders who otherwise would not receive payment for losses they incur
due to insurer bankruptcy.
Regulation of premiums is intended to prevent insurers from exercis-
ing market power in setting premiums. Governments monitor whether
premiums are too high, as well as too low, which is part of solvency
regulation, in order to assure the public access to insurance at quasi-
competitive rates. Excessive prices in the market may result from lack of
consumer knowledge and high consumer search costs, which limit com-
petition.9 In addition, competition may theoretically be diminished
because it is difficult for insurers to determine consumers’ risk level,
creating an informational advantage to existing insurers and a barrier
to entry.10
When insurance price spikes occur, governments are asked to act in
various ways to reduce prices or limit further rises in prices. On the other
hand, insurers facing competitive threats or attempting to increase their
market shares may actually set premiums too low relative to expected
222 Chapter 9

losses.11 Setting prices too low may lead to insurer bankruptcies or, more
commonly, insurers’ exit from those lines of insurance in which high
losses are being incurred. There is an analogy to the fairy tale “Gold-
ilocks and the Three Bears”: “This porridge is too hot. This porridge is
too cold. This porridge is just right.” In the long run, it seems likely that
the porridge—that is, insurance premiums—are just right. However, in
the shorter run, the porridge may either be too hot or too cold—exces-
sive prices when too hot, and inadequate prices when too cold. Both
potentially lead to disruptions in medical care markets.
But can public regulation lead to getting insurance prices right, that
is, set at levels that would prevail in competitive insurance markets? Since
governmental oversight of price is a complex process and subject to dif-
fering opinions, this fairy tale does not necessarily have a happy ending,
even though, as seen in the next section, medical malpractice insurance
premiums appear to have been adequate but not excessive during the
late 1970s and 1980s. To our knowledge, no one has accessed adequacy
of premiums more recently.
Another goal of insurance regulation is to assure that insurance is
readily available to potential purchasers. Some types of insurance are
seen as essential to socially valuable undertakings, including the delivery
of medical care. For this reason, state legislatures have sought to keep
medical malpractice insurance available by, for example, permitting sale
of insurance by companies other than those organized as for-profit com-
panies with stockholders as the owners—“stock” companies.
In a competitive insurance market, some potential policyholders may
be excluded as bad risks. It may be in the public interest that such risks
be covered. Such concerns have arisen in automobile insurance as well
as medical malpractice insurance. In the former, being able to drive may
be essential to maintaining employment, and availability of liability
insurance is directly linked to being able to drive in states with compul-
sory liability insurance laws. In the latter, a physician may be high risk,
not because he or she is a bad physician, but rather because he or she
delivers services that may result in medical malpractice suits. Joint under-
writing associations, described below, are one mechanism states have
used to assure availability of coverage.
 Medical Malpractice Insurance and Insurance Regulation 223

A final rationale for insurance regulation is information provision and

fair play. This rationale is relatively unimportant for medical malpractice
insurance. Information provision and review of policy forms are more
important when the customer is not highly educated and/or an infrequent
purchaser of coverage. The motive of fair play sometimes leads to require-
ments that insurers sell insurance to high-risk customers at premiums
below the actuarial value of the loss. This has occurred in automobile
insurance most prominently,12 which has been combined with with-
drawal restrictions forcing insurers to remain in an unprofitable market
for longer than they would voluntarily choose to do.13 Governments do
not require that medical malpractice insurers cross-subsidize high-risk
insured health care providers, but medical malpractice insurers do this
voluntarily.

Are Medical Malpractice Premiums Adequate or Excessive?

Sloan, Bovbjerg, and Githens (1991) use two alternative analytic

approaches to assess adequacy of medical malpractice premiums during
the late 1970s and 1980s, one based on a discounted cash flow frame-
work and the other based on the capital asset pricing model (CAPM). A
discussion of the technical details of their analysis would take us too far
afield. They report that during 1978–1979, medical malpractice premi-
ums yielded returns in excess of those that would prevail in a competitive
insurance market. But they do not make much of this finding, since policy
concern is raised only by persistent excess returns rather than temporary
ones. By 1982, premiums were adequately priced. The authors conclude
that during the period they study, premiums were on average neither too
high nor too low. A problem in conducting this type of analysis is that
it can be done only years after the insurance for a specific policy year
was sold. Thus, as this book is written, it is too soon to evaluate ade-
quacy of premiums after the late 1990s.
The good news that premiums are adequate does not necessarily imply
that government regulation of insurance is superfluous. What is observed
are premiums after interventions of state insurance departments, not
without it.
224 Chapter 9

Alternatives to Government Regulation of Insurers

Alternatives to direct government regulation of insurers are (1) imple-

mentation of statutory changes that permit changes in ownership form
and (2) direct public provision of insurance. If the insurer is operated
under an organizational form other than for-profit (e.g., the insurance
company is owned by policyholders), then the insurer’s incentive to
exploit policyholders by extracting maximum profit from them is plau-
sibility reduced. A stock insurer may be likely to exit a market as soon
as selling policies in that market is unprofitable. Insurers organized under
the alternative forms may be less prone to exit since the owners, the
residual claimants, pursue objectives other than obtaining maximum
profit. For example, the “bosses” of a physician-sponsored medical mal-
practice insurer may have a special interest in seeing that the insurer
remains in business in the state.
Also, some aspects of insurance may defy private contracting, justify-
ing a larger role for public insurers that do not have maximizing profit
as an organizational goal.14 For example, the quality of the legal defense
supplied to the physician policyholder by the insurer may be difficult to
specify contractually. A company owned by those insured by the company
may see that a defense is conducted more closely in accord with the
wishes of insured defendants. A for-profit insurer may calculate whether
settling the case is more profitable, but the defendant may want to con-
tinue the case as a matter of principle. Insurers organized on another
basis may not be able to continue the case without limits, but they may
be more sympathetic to views of the persons they insure.
Public insurers would not ordinarily completely avoid the principal-
agent (the insured individual being the principal and the insurer,
the agent) problems of stock companies. Direct public provision of
insurance of large claims has been motivated by reasoning that absent
government participation in the market, the volatility of both the fre-
quency and the size of these claims has increased the cost of capital to
medical malpractice insurers and has led to both premium increases and
insurer exits.
An insuring organization not subject to state regulation is the risk reten-
tion group. Groups of physicians, such as anesthesiologists, jointly pur-
 Medical Malpractice Insurance and Insurance Regulation 225

chase coverage from an existing insurer. To promote availability of

coverage, the U.S. Congress passed the Risk Retention Act in 1981
(amended in 1986), which exempts such groups from state regulation.15

Differences Between Medical Malpractice and Other Forms of

Property-Casualty Insurance

Distinctive Features of the Medical Malpractice Insurance Market

The most common loss covered by property-casualty insurance is fault-
based motor vehicle liability. Motor vehicle liability insurance pays many
claims per insured individual (claims frequency), but low amounts per
paid claim (claims severity). High claims frequency means that insurers
receive frequent signals, not only about their policyholders’ behavior but
also about how judges and juries determine liability and award damages.
This in turn allows insurers to classify insured individuals according to
their risk of incurring a loss. Medical liability insurance, by contrast, is
a low frequency/high severity line of coverage, making it more difficult
to classify individual physicians according to the probability that claims
will be filed against them in the future, one reason for the lack of experi-
ence rating of medical malpractice insurance premiums.
In motor vehicle liability, the delay from the date of an accident to the
date a claim is paid is often short. By contrast, medical liability claims
often take many years to be resolved. As a result, using data on a rela-
tively few claims, a medical malpractice insurer must estimate the likeli-
hood that a jury far in the future will hold the defendant liable, and the
amount that jury would award in damages. It is generally not possible
to compute an accurate estimate of expected future loss on an individual
physician basis.
In most states there is a compulsory minimum amount of motor
vehicle liability insurance a driver must have to make it more likely that
injurers will be able to compensate those they injure.16 Without insur-
ance, many individual drivers are “judgment proof” after being involved
in a major traffic accident (i.e., they have insufficient assets to compen-
sate the injury victim for his or her loss). Unlike motor vehicle insurance,
states, with some exceptions (e.g., Pennsylvania and Kansas), generally
do not require health care providers to purchase minimum amounts of
226 Chapter 9

medical malpractice insurance coverage. However, hospitals and physi-

cian groups may make affiliation conditional on a physician’s purchase
of adequate coverage. Because physicians tend to have substantial per-
sonal assets, few are willing to risk their career earnings on the outcome
of a single claim by “going bare” (i.e., forgoing insurance). Instead,
physicians faced with rapidly rising insurance costs may leave practice,
change their scope of practice in order to reduce premiums, or move to
another state where premiums are lower.
Although there are periodic complaints about automobile liability and
insurance, such as delays in payment, high legal fees, and the ambiguity
of fault, there is widespread acceptance of the overall litigation and
insurance system by drivers and voters. By contrast, medical malpractice
has very few defenders in the physician or health provider community.
Physicians reject, almost universally, the premise that tort liability serves
a useful role in medicine.

Types of Medical Malpractice Insurers

The market for malpractice insurance consists of three broad categories
of insurers. First are traditional “multiple-line” insurance companies,
such as the St. Paul Group of Companies, which historically was the
largest stock insurer selling medical liability coverage. It stopped writing
new policies in 2000.
A second group consists of physician-sponsored, “single-line” medical
malpractice insurance companies, many of which were chartered
during the “crisis of availability” in the 1970s. The common organiza-
tional forms are mutuals and reciprocals. The growth of these organiza-
tional forms is an institutional response to the crisis in availability
of medical malpractice insurance in the 1970s. Under the mutual form,
the roles of policyholder and owner are merged. The policyholders
supply capital to the insurance company. Policyholders (in this context,
mainly physicians) are given a say in company financial decisions
during the time their policies are in force. The mutual form aligns the
incentives of owners and policyholders since they are the same, but one
potentially important control mechanism present in the stock form
is absent: buyout of inefficient organizations. Nevertheless, lack of
buyout potential and less flexibility to leave the medical malpractice
 Medical Malpractice Insurance and Insurance Regulation 227

line of business is an advantage in terms of greater stability of insurance

supply.
Reciprocals resemble mutuals, but with two differences. First, a recip-
rocal is unincorporated and owns no capital. By contrast, a mutual is
incorporated with a stated capital and surplus. Second, the policyholders
appoint an individual or corporation as an “attorney in fact” to manage
the organization with an advisory committee to provide some oversight
of management. In a mutual, the board of directors has responsibility
for management oversight. In some reciprocals, policyholders can control
managers by exercising their option to withdraw their contribution to
surplus, or policyholders as a group can legally force dissolution of the
association. Originally chartered in one state, some physician-sponsored
insurance companies have entered markets in other states, sometimes
using a low-price strategy to gain market share, as was described in
chapter 2.
Self-insured organizations such as large hospitals assume their own
risk and set aside appropriate reserves, typically reinsuring large losses
above a high deductible. A large hospital might also use a captive insurer,
a wholly owned subsidiary that keeps formal insurance accounts but is
usually chartered abroad, and therefore exempt from domestic insurance
regulation. Risk retention groups are analogous to a buyers’ cooperative.
Limited-purpose insurers that operate under federal rather than state
law, they generally represent members of a single physician specialty or
hospitals in a particular geographic region that decide to share risk.
Private surplus line insurers offer insurance to high-risk individuals as
in other lines of insurance, most notably automobile insurance. These
companies insure only a very small number of physicians. Alongside
lower underwriting standards are higher insurance premiums.
In addition, there are alternative forms of coverage, such as joint
underwriting associations (JUAs), self-insurance vehicles, captive insur-
ance companies, and risk retention groups, in which individuals, trade
organizations, or existing insurers reduce their exposure by creating and
operating their own insurance company. In general, these alternative
insurers emphasize providing stable coverage rather than maximizing
profits; they structure their finances so that they are less likely to exit
due to adverse market conditions.
228 Chapter 9

JUAs are cooperative ventures of existing insurers that are mandated

by states to supply insurance to individuals who, in principle, are unable
to obtain insurance from other sources. JUAs are a response to a social
objective of promoting access to medical care by assuring that physicians
have access to medical malpractice insurance.17 In automobile casualty
insurance, assigned risk plans are much more common than JUAs.
Assigned risk plans provide a method for placing high risk insured
individuals with specific private insurers, often at less than an actuarially
fair rate (i.e., premiums are too low in relationship to the losses that
the insurers can expect persons assigned to them to generate).

Lack of Experience Rating

Experience rating is rare in the medical malpractice line of insurance,
not in terms of the number of insurers who engage in some form of
experience rating but rather in terms of the number of physicians
adversely affected by this practice. A national survey of insurers con-
ducted by Schwartz and Mendelson18 reports that 90.3 percent of com-
panies employed surcharges in 1985. However, only 1.6 percent of
insured physicians were subjected to surcharges and another 0.9 percent
faced other sanctions, such as restrictions on the scope of their practices
or requirements for further training and supervision. A survey of four-
teen medical malpractice insurers conducted in 1987–1988 revealed that
only one of the fourteen insurers had never implemented an experience
rating program of any type. At the time of the survey, less than 1 percent
of physician enrollees paid more than standard premiums because of
adverse claims experience.19
Due to a lack of experience rating, doctors with favorable and
unfavorable loss experience pay similar malpractice premiums. Preserv-
ing a connection between tort liability and quality of care is a major
goal of the medical malpractice system.20
The widespread view among physicians that determining liability is a
haphazard process has been a barrier to setting premiums according to
individual physicians’ claims histories.21 Low claims frequency makes it
even harder to experience-rate premiums, even though claims against a
few physicians account for a major part of overall losses to malpractice
carriers.22 This perpetuates a vicious cycle in which lack of acceptance
 Medical Malpractice Insurance and Insurance Regulation 229

of a constructive role for tort in advancing medical quality is an impedi-

ment to meaningful structural reform of the malpractice system. Govern-
ments have intervened to promote use of experience rating, but the
results have been mixed at best.23
Is the way out of this dilemma to reintroduce experience rating? Gen-
erally the unit for rating is the insured individual, but there are alterna-
tive experience rating units ranging from medical groups to a hospital
and/or its medical staff. When insured individuals are rated as part of a
larger organization, the group must allocate premiums by individual or
decide to cross-subsidize the premiums of some of its members. How the
organization allocates its premium expense among the group members
is the organization’s private concern.
Sloan and Hassan seek to answer the question of whether the hospital/
hospital’s medical staff is sufficiently large to be experience rated.24 The
alternative being considered here is whether or not a hospital’s and/or its
medical staff’s total experience contains a sufficient amount of informa-
tion to provide reliable evidence for purposes of experience rating. In their
simulations, they use data on medical staffs by hospital size from the
American Hospital Association and from Florida medical malpractice
closed claims. The closed claims were aggregated to the level of individual
physician. Then physicians were randomly selected to “staff” hospitals of
particular sizes and medical staff composition by specialty.
To gauge whether or not experience rating at the level of the hospital
would be feasible and practical, Sloan and Hassan computed the level
of initial capitalization that would be consistent with a specific bank-
ruptcy risk. More specifically, while taking the timing of premium income
versus payment for loss into account, the authors computed the probabil-
ity that losses would fall short of premium income. The problem was to
set the hospital/hospital medical staff premium in total, allowing the
organization to figure out how premiums would actually be financed.
The hospital could decide to have each physician share equally (say by
specialty) in the premium obligation, but impose other carrots/sticks to
avoid losses due to claims. That is, the task of monitoring is shifted from
the level of the insurer, where the task is not performed well, to the
hospital/hospital medical staff, where monitoring costs are likely to be
much lower. Further, it is the hospital’s legal obligation to monitor, an
230 Chapter 9

obligation typically delegated at least in part to the hospital’s medical

staff. Hospitals could improve their loss experiences by implementing
more effective peer review, improving systems which promote patient
safety, and being more selective about the physicians they retain or admit
to their medical staff.
Without discussing the details of the findings here, the message is a
positive one. The specific question the study addresses is how much the
group would need to hold in reserves to achieve a given level of insol-
vency risk. A lower insolvency risk requires higher levels of reserves.
Initially, each physician member of the hospital’s medical staff covered
under the plan would be required to make a contribution to reserves.
When the levels of required reserves is higher, the per-physician contribu-
tion would also be higher.
Setting the probability that a plan would have insufficient funds to
cover losses over a six-year period at 2.5 percent (or 0.4 to 0.5 percent
per annum), the initial contribution per physician on the staff of a hos-
pital with fewer than 100 beds would be 4.2 times the annual premium.
The ratio of initial capitalization to annual premiums was lower than
this for larger hospitals. If the bankruptcy probability percentage of 2.5
percent over six years is thought to be too high, a lower percentage could
be employed. As a consequence, the initial capitalization would be cor-
respondingly higher. A 0.4–0.5 percent per year insolvency rate does not
seem inordinately high, however, especially since contracts with indi-
vidual physicians could include a provision for ex post assessments which
would be required only very rarely.

Government-Sponsored Risk Pooling Arrangements

Government-sponsored risk-pooling arrangements address the solvency,

availability, and fair play objectives of government intervention in the
context of medical malpractice insurance.
Joint underwriting associations (JUAs) and guaranty funds are publicly
created entities that provide insurance as a last resort. In this role, state
governments facilitate transactions within the private insurance sector
that alter risk pools in order to serve the public objectives of assuring
 Medical Malpractice Insurance and Insurance Regulation 231

availability of coverage and protecting consumers against insurer insol-

vency. Without state action, it seems unlikely that insurers would coor-
dinate their activities in order to establish high-risk pools. These pools
would likely be unprofitable.

Joint Underwriting Associations

In response to the medical malpractice insurance crisis of the mid-1970s,
many states authorized the formation of JUAs. At the time, legislators
saw a need to fill the gap in coverage that arose when private insurers
withdrew from the market.25 Unlike mutual insurers with ties to medical
societies, which also were created in many states in the 1970s, JUAs
and patient compensation funds, described below and in the next
chapter, are public organizations. Under typical legislation establishing
JUAs, all companies writing liability insurance of any kind are required
to participate. JUAs are often conceived as temporary measures, but
some of those established in the 1970s remain active.26 Moreover, there
have been proposals to form JUAs for medical malpractice as recently
as 2002.27
JUAs are designed to be “insurers of last resort,” that is, to cover
providers who are unable to obtain insurance from other sources.
The Kansas JUA provides an example. As already noted, having
medical malpractice insurance coverage is mandatory in Kansas.
The Kansas statute imposing this requirement also implemented an
availability plan for backup basic coverage and an excess coverage
plan. These plans are supported by the Health Care Stabilization Fund,
which is financed by a surcharge on providers.28 The basic coverage
sets premiums in excess of those charged by private insurers and is avail-
able only to providers able to demonstrate that they cannot obtain
private coverage.
If the JUA’s premium income is insufficient to cover losses and admin-
istrative expense, each member company is assessed a pro rata share of
the shortfall. In competitive insurance markets, owners of companies
demand a reasonable rate of return on the capital they supply. The only
way to earn this return is for companies that subsidize the JUA (either
all malpractice insurers or all property-casualty insurers) to increase
232 Chapter 9

premiums to their policyholders. As a result, obtaining medical malprac-

tice insurance becomes more expensive relative to JUA coverage, leading
more physicians to substitute JUA-obtained coverage for private cover-
age. When state governments form a JUA, they facilitate transactions
among private insurers in forming risk pools. Some JUAs established
in the 1970s remain active even though they were often conceived as
temporary fixes.29
Rather than serving the intended purpose of providing medical mal-
practice insurance coverage to health care providers who cannot obtain
it from the private market, JUAs have become a main source of coverage
in a few states, such as South Carolina. Pooling arrangements therefore
may crowd out private medical malpractice insurance, as occurred in
some states with JUAs in the 1980s.30 At that time, JUAs dominated
the market in Massachusetts and Rhode Island, which like other New
England states, lacked physician-sponsored medical malpractice insur-
ers.31 Moreover, both JUAs were in substantial financial trouble,32 in
large part because the political process by which they set premiums
underestimated the actuarial value of their loss exposure.
During the crisis in Pennsylvania, which was at its worst during
2001–2004, hospitals and physicians increasingly relied on coverage
through the state’s JUA and other alternative coverage sources such as
risk retention groups. In 2002, it was estimated that Pennsylvania’s JUA
had 1,700 physician policies in force, up from 351 in 2001.33 Between
1999 and 2002, the number of health care providers in this state who
obtained coverage through the JUA increased by a factor of more than
7.34 In 2002, 12 percent of hospitals in Pennsylvania had their primary
coverage through the Pennsylvania Joint Underwriting Association.35 In
Florida, there was also a substantial increase in the number of physicians
enrolled in the state’s JUA, but the JUA in that state covered only a small
fraction of physicians in the state.36
Although JUAs can provide coverage when private insurance does not,
they also have deficiencies. Some weaknesses are common to all JUAs,
while others are unique to particular states.
Providing coverage through a JUA may conflict with deterring medical
injuries. Although some physicians who experience difficulty in obtain-
ing coverage may have clean records, particularly during a generalized
 Medical Malpractice Insurance and Insurance Regulation 233

insurance crisis, others lack access to private coverage because of a

history of repeated claims. When physicians with many past claims are
able to obtain coverage at standard rates, subsidized by physicians with
better track records, they lack an incentive to improve (or leave practice).
In addition, there is rarely any mention of loss prevention in the literature
on JUAs. The major focus is on insuring physicians, not on developing
programs to reduce the probability of claims.37
Some states have recognized that pricing should reflect underlying risk,
and offer premiums matching the insured’s expected loss. For example,
Pennsylvania’s JUA now includes a provision for premium surcharges
based on prior claims and claim expense, or based on regulatory actions
suggesting poor provider quality have been taken by licensing boards,
hospitals, Medicare or Medicaid, the federal Drug Enforcement Admin-
istration, or Pennsylvania’s controlled substance act.38 This is clearly a
step in the right direction.
However, underpricing has occurred in well-publicized cases. Not
only does pricing JUA coverage lower than that available in the private
market make it a more attractive option, theoretically, it also may
increase the prices that the private market must charge for coverage.
Each company linked to the JUA is assessed a pro rata share of the
deficit if the JUA’s administrative expenses and losses exceed its pre-
mium income. Since owners of member companies who subsidize the
JUA must provide a reasonable rate of return on their capital, pro rata
assessments force increases in premiums for policyholders of private
companies. This makes the JUA’s premiums even more appealing, result-
ing in health care providers dropping private coverage in exchange for
JUA coverage.
Overpricing is also a risk for some JUAs due to the pressure from
private insurers to assure that the JUAs do not compete with them. Like
underpricing, this is to be avoided, lest providers not buy JUA coverage
in situations in which using it would be socially optimal.

State Guaranty Funds

Another form of government-organized risk-pooling are state guaranty
funds, which, like JUAs, are institutional arrangements not unique to
medical malpractice insurance. In addition to directly promoting insurer
234 Chapter 9

solvency, insurance regulation protects policyholders’ ability to collect

on future claims by mandating guaranty funds. Between 1969 and 1981,
all states enacted laws to establish guaranty funds.39 Guaranty funds
provide a mechanism for assessing surviving insurers after the fact for
losses incurred by insolvent insurers.
Almost all states guarantee that valid property-casualty claims will be
paid, if necessary, by these state-overseen funds. When an insurer is in
receivership and management cannot meet its obligations to policyhold-
ers, the receiver can draw against the state’s guaranty fund.40 A guaranty
fund raises money to cover these payments by assessing property-
casualty insurers who do business in the state. Except for the New York
fund, they are all post-insolvency assessment funds. Members are assessed
a fixed percentage of premium volume to pay claims that exceed the assets
of the insolvent insurer.41 Thus, guaranty funds are mechanisms for
taxing all property-liability policyholders to cover losses of a particular
firm in a particular line, such as a seller of medical liability insurance.
Guaranty funds offer physicians and other health care providers who
purchase medical liability insurance the ultimate in protection from
insurer insolvency. A physician whose insurer goes bankrupt before
indemnifying an injury victim does not risk his or her personal assets.
As with JUAs, however, consumers of other types of liability insurance
may be required to subsidize the losses of physicians (of course, the
opposite may occur as well if auto liability insurers fail). These redistri-
butional effects are real, but are not transparent to consumers of health
care or to citizens as voters.
There is a frequent trade-off in insurance between risk protection and
efficiency. Guaranty funds offer protection against loss, but at a cost.
Because the state guaranty fund will honor claims against a health care
provider if the provider’s primary carrier becomes insolvent, providers
who have a choice among primary insurers may not consider financial
strength an important attribute.42 This in turn creates an incentive for
insurers to engage in risky underwriting and investment practices.43 This
effect is exacerbated by the fact that guaranty fund assessments are not
based on the riskiness of an insurer’s business strategy,44 but it seems
unlikely that this would apply to a small line such as medical liability
insurance.
 Medical Malpractice Insurance and Insurance Regulation 235

While it is the function of insurance to repay policyholders after a loss

is incurred, the government insurance pooling arrangements can shield
policyholders from avoidable as well as unavoidable losses. This occurs
when guaranty funds pay claims incurred by insurers who became insol-
vent because they assumed excessive business risk, or when JUAs insure
health care providers at less than actuarially fair premiums. Admittedly,
there may be situations in which charging actuarially fair malpractice
premiums conflicts with assuring that medical care is available in an area
underserved by physicians. But it is preferable to isolate these specific
situations and provide explicit subsidies, rather than to redistribute
resources sub rosa to the group of providers and insurers that seem to
be at highest risk of business failure.

Evaluation of Government Risk Pooling Arrangements

Aside from descriptive information about program operations, very little

is known about the performance of JUAs and guaranty funds. Lessons
have been learned and used to improve public programs over time.45 To
states considering implementing or changing their programs to tackle
medical malpractice, the experiences of other states that have programs
in place are potentially quite valuable. Nonetheless, states appear to run
their programs in isolation. As a result, so little know-how is shared that
some of the same mistakes happen again and again.
The conventional wisdom that the private sector is more efficient does
not seem to apply, as these programs have succeeded in achieving a low
administrative cost. An inherent risk of public provision is that it will
crowd out private coverage, yet with some exceptions, nothing indicates
this has occurred on a wide scale.
The lack of data collection and formal program evaluation makes it
impossible to know, for example, whether JUAs have made medical
malpractice insurance more available to health care providers or, alter-
natively, whether they have crowded out private insurance coverage.
Despite the fact that patient safety is a general concern, nothing is known
about actions JUAs have taken to improve patient safety, an especially
important consideration since these pooling arrangements often enroll
substandard risks. It is not known how frequently guaranty funds have
236 Chapter 9

been used in the medical malpractice insurance field, or what the effect
is of the existence of such funds on insurer efforts to manage their
losses.
There is no empirical evidence on the performance of state guaranty
funds. Even less is known about guaranty funds than about JUAs, either
in other lines of property-casualty insurance or in the context of medical
malpractice insurance.
One concern is that guaranty funds may not reduce insurance cycles,
but instead exacerbate them. State guaranty funds may intensify a capac-
ity shortage because the crisis phase of the insurance cycle is coupled
with large numbers of insolvencies.46 Just as insurers are leaving the
market, industry capacity is further reduced when net worth decreases
as a result of assessing funds from the remaining insurers. While some
consumers are protected from their insurer’s insolvency, guaranty funds
may cause premium levels to increase across the industry. Prefunding
the system, as in New York, could avoid the possibility of such
problems.47

Regulation of Medical Malpractice Insurance

Critique of Public Regulation

Public regulation has long been a feature of many industries. There has
been a trend toward deregulation in several sectors, but generally to a
lesser extent in insurance. The major goals of regulating medical mal-
practice insurance were described above. Although these are praisewor-
thy, they may come at a cost. Free entry and exit is the hallmark of
competitive markets.
While there is a concern about losses incurred by policyholders in the
event of insurer insolvency, solvency regulation may offer protection to
otherwise inefficient firms and confer the ability to exercise market
power on those firms that are protected. Rate regulation may reduce the
exercise of market power. Yet to the extent that rates are too low, the
supply of the product may be reduced below socially optimal levels.
Assuring availability of coverage has the benefit of potentially preserv-
ing the public’s access to medical care. But attaining this goal may con-
flict with injury deterrence and result in cross-subsidies from insured
 Medical Malpractice Insurance and Insurance Regulation 237

parties which exercise care to those which do not. The fear of losing
insurance at standard rates may be an incentive to some potential injurers
to take precautions. This incentive becomes weaker if insured parties are
guaranteed coverage, particularly at less than actuarially fair premiums,
even if they do not exercise care. Information provision is less relevant
in the medical malpractice field than for other lines of insurance. Fair
play, however, is likely to be equivalent to advocating cross-subsidization
of premiums.
Lacking specific evidence that private markets are not accomplishing
social objectives (market failure), the presumption of many economists
is that societal well-being is best served by leaving markets alone. If there
is some evidence of such market failure, then this may provide a rationale
for public intervention with one major caveat, namely, that government
intervention actually improves social welfare even in the presence of
market failure absent such intervention.
Some prominent scholars, many of whom are economists, have been
critical of the view that regulation serves the public interest.48 They argue
that there is a market for regulatory controls just as there are markets
for other goods and services. In the context of regulation, the suppliers
are government agencies and legislators. Public regulation is demanded
by special interests. Thus, for examples, farmers demand government
controls, such as limits on farm production, since such restrictions result
in higher product prices. In return, being made subject to these restric-
tions, the regulated groups make financial contributions and vote for the
suppliers.
The general electorate tends to be indifferent or even not knowledge-
able about particular regulations because it is most often not directly
affected or is affected in ways that are quite subtle. Since being actively
involved in the political process is costly, most people are not interested
in most issues most of the time. Lack of widespread public interest facili-
tates capture by the industry being regulated. Conversely, when the
public becomes more involved, capture by private interests becomes
more difficult to achieve.49
These generalizations may fit any particular industry in very general
terms, though each industry is somewhat idiosyncratic. The most in-
depth study of regulation of insurance is a book by Kenneth Meier.50
238 Chapter 9

Professor Meier emphasizes the heterogeneity of circumstances under

which regulation takes place, even within the insurance industry. A
number of lines of insurance, such as motor vehicle and homeowners
insurance, are important to a large number of persons. Other lines of
insurance, including medical malpractice insurance, are far less relevant
to the general public. Bureaucracies tend to have greater influence on
public policies when policies are complex and not particularly salient to
the public.
Salience is important because it may be difficult to arouse legislative
interest on the basis of financial contributions and personal influence
(lobbying) alone. To elicit the public’s (and hence legislators’) interest,
the issue must be made salient. For example, to motivate state legislators
to introduce tort reform legislation, the insurance industry has needed
to raise saliency enough to place the issue on the legislative agenda.
Salience has been increased by issuing news releases and other forms of
publicity warning the public and government officials about the “crisis”
and its consequences. The industry has tried to simplify the issue to
increase public participation. In dealing with the medical malpractice
crisis, the interests of insurers and health care providers are aligned. The
task is to get the public to assist in pressuring for statutory change.

Empirical Evidence on Effects of Regulation: Studies of Other Lines of

Insurance
The above general points about insurance regulation are largely theoreti-
cal or sometimes impressionistic. There is a body of empirical research
which assesses the actual impacts of regulation—showing, for example,
that even though regulation reduces premiums, it may also increase the
market shares of JUAs and other assigned-risk plans.51
The most detailed analysis of solvency regulation in the property-
casuality insurance industry is Munch and Smallwood.52 The authors find
that minimum capital requirements (maximum premium-to-surplus ratio)
reduce the number of insurer insolvencies. This was accomplished solely
by blocking entry of small, relatively risky insurers. While erecting entry
barriers has the benefit of reducing insolvencies, it also has three adverse
effects. For one, some insurance purchasers may be willing to tolerate
some additional insolvency risk in trade for a lower premium.53
 Medical Malpractice Insurance and Insurance Regulation 239

Empirical Evidence on Effects of Regulation: Studies of Medical

Malpractice Insurance
Sloan, Bovbjerg, and Githens report results of a survey they conducted of
fourteen medical malpractice insurers in 1987–1988.54 Part of the survey
dealt with regulation. Based on these findings, they concluded that for
most companies in most years, regulation has not been a major influence
on entry of insurers and, with some important exceptions, on pricing of
medical malpractice insurance. The national commercial insurers respond-
ing to the survey expressed far more concern about the regulatory climate
than did physician insurers. The one instance of an insurer leaving the
state (among the fourteen respondents) involved a national insurer.
Born, Viscusi, and Baker examine the long-run effects of major tort
reforms on insurers’ losses, using a sample of individual firms writing
medical malpractice insurance during 1984–2003.55 One of the legal
variables included in their analysis is prior approval rate regulation,
which identified states with the most stringent form of rate regulation,
but such regulation has no statistically significant effect on losses. An
earlier study by Viscusi and Born analyzed variations in the loss ratio
for medical malpractice insurers.56 The authors found premium regula-
tion had no effect on this ratio, defined as losses attributable to premiums
written in a policy year to the value of premiums earned in that year.
Viscusi and Born’s analysis of medical malpractice insurer losses, premi-
ums, and loss ratios revealed no effect of prior approval regulation on
these dependent variables.57 By contrast, they reported that certain tort
reforms did improve insurer profitability.
Between 1984 and 1991, about 1 percent of all insurance companies
in the United States failed. This represented thirty-seven firms. The insol-
vency rate had more than doubled the rate for years before this. Increas-
ing frequency of failure was only part of the problem. There was increased
severity associated with the insurer failures as well.58 Comparable statis-
tics are not available for medical malpractice insurance.

The PHICO Bankruptcy: Business Failure in Spite of Solvency

Regulation
A well-publicized failure occurred in the early 2000s in spite of
state insurance departments’ efforts to oversee financial performance
240 Chapter 9

of insurers writing coverage in their states. The failure involved the

Pennsylvania Hospital Insurance Company (PHICO).
In an attempt to provide affordable insurance, the Pennsylvania State
Hospital Association created PHICO in 1976. The company enjoyed
success, partly due to its conservative management, and paid claims up
until the time of its liquidation in 2002. During 1995, a quiescent period
for insurance, with the appointment of a new chairman to the company’s
insurance board, PHICO changed its priorities. Instead of continuing its
focus on consistently affordable insurance, the company decided to break
into the national market.
PHICO was able to obtain licenses to sell insurance in all fifty states,
the District of Columbia, Puerto Rico, and the U.S. Virgin Islands. As a
result of this expansion, premium payments from physicians rose from
$10.7 million to $130.5 million.59 This dramatic growth was not surpris-
ing; premiums were set below market value. PHICO used this as its
selling point; a 1997 advertisement boasted that premiums would be a
25 to 35 percent savings for policyholders.60 PHICO’s methods worked.
In five years the premium income rose from $94.7 million to $181.5
million.61 By 2000, PHICO had become the seventh largest medical
malpractice insurer in the country.62
The excessively low premiums set by PHICO in the mid-1990s allowed
the company’s growth to be substantial, but the growth was extremely
unstable. Insurance regulations are determined by the states, so each
state has unique requirements for operation. In order to be profitable,
companies must consider the state regulations, local markets, and
competitors before expanding.
Low premiums coupled with inadequate forethought and planning led
to financial distress. For premiums collected in 2001, PHICO paid, on
average, $1.40 for every dollar of premium it received.63
In its lawsuit against PHICO, the Pennsylvania Department of
Insurance offered a different opinion as to the company’s problems.
It alleged that mismanagement was an important factor: “The com-
pany aggressively under-priced the competition even in markets in which
the company had no previous experience. The strategy of offering low
prices in highly competitive and unfamiliar markets was fraught with
risk.”64
 Medical Malpractice Insurance and Insurance Regulation 241

Many questions remain as to how this could have happened to a stable

company with a nearly thirty-year history. It was not due to a lack of
insurance regulations; PHICO did business in all fifty states, all of which
had insurance regulations in place and requirements for licensure. Some
accuse the Pennsylvania Insurance Department of failing to adequately
monitor PHICO’s business practices and not taking steps to correct the
rapidly deteriorating situation. In response to this heavy criticism,
the Insurance Department claimed it was impossible for it to monitor
the company’s financial status because PHICO had filed misleading
reports.65 Unsurprisingly, questions have been raised as to the appropri-
ateness of existing regulations.66 There were also some political allega-
tions. State regulators who were also insurance executives and legislators
who also worked in the insurance industry drafted regulations.67
Regardless of blame, regulations may not have been the problem, as
the Pennsylvania Insurance Department argued—even though lack of
adequate oversight may have been. In 1999, PHICO’s outside auditor,
Price Waterhouse Coopers, advised the company that it was facing
massive losses and its reserves needed to be increased by $130
million.68
PHICO took no action; in 2000, accounts were renewed with no
meaningful adjustment to the underwriting terms or price.69 Instead, the
leaders of PHICO attempted to resolve the problem in a different way,
restricting the power of those in charge of the reserves. In May 2000,
the authority of the claims department was restricted, preventing it from
being able to set reserves. In addition, PHICO imposed an arbitrary
incurred loss budget, which prevented the claims department from
increasing the reserves.
Many knowledgeable observers were surprised by the extent to which
PHICO was underreserved. When state regulators placed PHICO into
rehabilitation under Pennsylvania’s insurance law, its financial statement
showed a meager $6.8 million surplus. After liquidation, it became
apparent that PHICO’s reserves had not just been dwindling, they had
been nonexistent. Regulators estimated the firm was underreserved and
insolvent by at least $250 million.
The impact of PHICO’s collapse has been far-reaching; 38,000 doctors
and nearly 900 hospitals across the country were forced to find
242 Chapter 9

alternative liability insurance, usually at much higher rates.70 Pennsylva-

nia, where PHICO was one of the major malpractice insurers, was hit
especially hard by its downfall. Other insurance companies, their share-
holders, and claimants also felt the loss; PHICO left behind 16,600
pending claims.71 Hospital systems also faced huge losses.
PHICO’s insolvency, along with several other insurance companies
leaving the market, has created debate over the underlying causes of the
insurance failures.72 Some argue a bad economy and poor business deci-
sions by company executives are to blame for the failure of so many
insurance companies during 1995–2005.
Could this have been prevented by stricter state regulations and closer
monitoring of business activities? The evidence suggests that insurance
regulation had nothing to do with either causing or preventing the com-
pany’s failure, the latter being the primary function of state insurance
regulation. Failure to prevent the company’s bankruptcy amounts to a
major criticism of the regulatory process, at least in this case. Whatever
the role of regulation, some insurance companies have fared much better,
at least to date. NORCAL Mutual Insurance Company is an example of
such companies.

NORCAL: A Physician-Sponsored Insurer with an Outcome Different

from PHICO’s
Like PHICO, NORCAL was born out of the 1970s’ medical malpractice
crisis. It, too, was started by a group of medical professionals in an
attempt to provide consistent, reasonable malpractice premiums.
NORCAL was started, owned and controlled by its member insureds.
The company initially was associated with Casualty Insurance Company
Service, Inc., but in only three years NORCAL was independent.
NORCAL, unlike PHICO, slowly expanded its business over the
course of twenty years. Between 1983 and 1990, the company focused
its efforts on doing business statewide. When the company was ready to
expand nationally, NORCAL slowly acquired other insurance compa-
nies.73 Its conservative approach was successful; it has received an A
rating from A.M. Best Co., a body which grades the performance of
insurance companies, for twenty-one years in a row. On its Web site,
NORCAL attributes its continued success to conservative business
 Medical Malpractice Insurance and Insurance Regulation 243

principles: “Our strong financial position is a result of our commitment

to financial stability and reliance on long-held principles of underwriting
to the standard of care, making conservative investments and pricing our
coverage responsibly.” In addition to sound business decisions, NORCAL,
and several tort reform advocates, maintain that the 1975 California
statute placing a $250,000 cap on attorney’s fees and pain and suffering
damages allowed the company to charge low premiums.74 NORCAL has
been able to keep the rate increase gradual and proportional, which it
attributes to decreased liabilities.
However, others point to Proposition 103, passed in 1988, which
grants the state insurance commissioner as well as outside parties the
power to challenge proposed rate hikes.75

Regulation May Adversely Affect Solvency: The Case of Savings and

Loan Banking

The Federal Deposit Insurance Corporation (FDIC) has provided insur-

ance coverage for consumer deposits in private banks. While instituted
to protect depositors against bank insolvency, a possible adverse effect
of the deposit insurance system is that it may provide a moral hazard
for excessive risk taking.76 Moral hazard arises when consumers become
much less cautious about the business practices of the banks in which
they have deposits if they have insurance in the event of bank failure.
The loss of market pressure to be prudent may correspond with less
financial scrutiny by public regulators of bank activities.
Richard Grossman compares risk-taking of insured and uninsured
savings and loan banks during the 1930s, finding an interaction between
regulation, insured savings, and risk-taking of banks.77 Those banks that
operated under relatively permissive regulatory regimes and whose
deposits were insured were more likely to undertake risky lending activi-
ties than were other banks. Rather than focus on the interaction between
regulation and deposit insurance, Michael Keeley, studying a much more
recent period, hypothesizes that moral hazard would be worse under
conditions of market competition among banks than under a regulatory
regime in which competition was restrained.78 Increased competition
may have reduced banks’ incentives to act prudently with regard to risk
244 Chapter 9

taking since the value of a bank charter fell due to comparatively free
entry of banks under deregulation which occurred in the early 1980s in
the United States. He finds that banks with more market power hold
more capital relative to assets on a market-value basis and have a lower
default risk (reflected in lower risk premiums on large, uninsured certifi-
cates of deposit). Having more capital lowers bankruptcy risk. The lower
rate on CDs demanded by investors is plausibly because investors required
a lower return since they perceived the banks’ bankruptcy risk to be
lower.
Using data from Kansas on individual banks from a period before the
FDIC was established (1910–1928), Wheelock and Wilson study deter-
minants of bank failures. Membership in the voluntary state deposit
insurance system increased the probability of bank failure.79 Richard
Cebula measures bank failure rates directly, finding that bank failure
rates rise as the fraction of deposits covered by federal deposit insurance
rises, which is direct evidence of moral hazard.80 He also reports that
bank failure rates declined with increases in the capital-to-asset ratio and
with increased competition among banks. Thus, at least under certain
circumstances, there is some reason to be concerned about the effect of
public insurance coverage on firms’ incentive to be cautious.

Conclusion

To answer the first of three questions posed in this chapter’s introduc-

tion—Is the market for medical malpractice insurance broken?—we con-
clude that the market has been fixed so often that it is difficult to
contemplate a market for medical malpractice insurance absent govern-
ment intervention. Empirical evidence from the late 1970s and the 1980s
indicates that on average, medical malpractice insurance premiums have
been adequate, not excessive; but, as acknowledged above, the correct-
ness of these premiums is measured in the presence, not the absence, of
state regulation.
The answer to the second question is more straightforward. In princi-
ple, regulation is necessary to address unique potential problems for
insurance consumers, such as access to coverage, financial insolvency
among insurers, the complexity of insurance contracts, timely payment
 Medical Malpractice Insurance and Insurance Regulation 245

of claims, and affordability of premiums, among other issues. In practice,

the record of insurance regulation is mixed. While regulation probably
has not caused more than minor harm, it is not clear that society has
been very well served by it either. Insolvencies occur in spite of regula-
tion. The PHICO bankruptcy is a case in point.
Risk-pooling programs meant to help accomplish the goals of regula-
tion often have unintended side effects. More empirical evidence is
needed on guaranty funds, yet we can already see the potential problems
with crowding out private insurance in the case of JUAs. New programs
incorporating catastrophe bonds, puts, and options are potential methods
to spread risk among investors willing to supply capital on more favor-
able terms.
Hospital self-insured medical malpractice insurance programs and
reinsurance are currently not subject to insurance regulation. To answer
the third question, in light of the evidence, we are reluctant to recom-
mend additional regulation. Publicly provided reinsurance may be worth
considering if some of the past mistakes described more fully in the next
chapter can be avoided.
10
Reinsurance

Why Reinsurance Is Important

As we indicated in chapter 2, shocks to both availability and premiums

are one precipitating cause of insurance cycles. Reinsurance is as vitally
important to primary medical malpractice insurers, especially small,
single-line insurers, as it is to self-insured hospital medical malpractice
insurance programs. Private reinsurance is not generally subject to state
or federal regulation. However, some states have implemented public
reinsurance programs, which, if important prior mistakes can be avoided,
provide a promising approach for adding stability to medical malpractice
insurance markets.

Reinsurance 101

In the medical malpractice market, physicians and hospitals are the cus-
tomers. Primary insurers sell coverage to nearly all physicians, and in
turn purchase reinsurance. Purchasing reinsurance helps a primary
insurer to reduce its exposure to losses in the aggregate above a certain
dollar threshold of loss or to very high-cost claims. Primary insurers or
self-insured entities, such as hospitals, seek to mitigate the risk from large
claims in this way because such claims can drain insurer reserves or
threaten financial solvency.1
Reinsurance is important for single-line medical malpractice insurers.
Some claims may be nearly as large as the surplus (equity) of these firms.
Without reinsurance, some smaller insurers may not be able to pay a
costly claim, or may go bankrupt.
248 Chapter 10

Rather than buying primary insurance, many hospitals, especially

larger ones, self-insure for much of their anticipated medical malpractice
loss. When hospitals self-insure, they frequently purchase excess insur-
ance, which is essentially reinsurance for hospitals. Excess insurance is
typically obtained from the same reinsurance companies that primary
insurers use. Several “layers” of reinsurance (or of excess insurance, in
the case of self-insuring hospitals) may often be purchased with a package
of coverage assembled by a broker. With a self-funded layer, the hospital
is still at risk for claims falling in that dollar range. Excess insurers
sometimes insist on this layer, known as a “risk corridor,” to help miti-
gate moral hazard on the part of hospitals. Larger facilities and health
systems rely on brokers to assemble a group of excess insurers for a
desired range involving several layers of excess coverage. In these com-
plicated arrangements, two or more carriers are often jointly responsible
for any given layer and coverage is made more affordable, if necessary,
by raising the attachment points or reducing the upper limits of coverage.
An attachment point reflects a corridor between primary coverage and
the dollar amount at which reinsurance begins to cover the loss.
Reinsurance companies face bankruptcy risks of their own. Reinsurers
located in the United States are subject to the solvency regulation in their
domiciliary state.2 However, many if not most private companies reinsur-
ing medical liability are located outside of the United States, and thus
are not subject to regulation by U.S. governmental entities. In property-
casualty insurance generally, foreign companies issue two-thirds of the
reinsurance obtained by parties in the United States.3
The most troubling difficulty for reinsurers/excess insurers with losses
at the “long tail” or high end of the claims distribution is not financial
insolvency, but the fact that large claims are rare. Large claims are vola-
tile in both their amounts and their frequency, making the frequency of
the very few large claims, especially large paid claims, difficult to predict.
Since large claims are so infrequent, private reinsurers may have difficulty
ascertaining when a true shift in the rate of these claims or in their size
has occurred, as opposed to variation due to random noise. Additionally,
from the time an injury occurs to the time a claim is filed is often years,
and the time span from filing to the date when the claim is closed can
be several more years.4 This slow resolution is in sharp contrast to insur-
 Reinsurance 249

ance lines such as automobile liability and health insurance. Since events
such as judicial decisions, high jury verdicts, or new laws can intervene
during this long resolution process, the high uncertainty associated with
premium setting is multiplied.
As a result, after a few large payouts, private-sector reinsurers may be
quick either to refuse to underwrite coverage or to substantially raise
premiums to reflect a (correctly or incorrectly) perceived increase in risk.
Volatility in the high-loss end of the claims distribution leads to volatility
in the private supply of reinsurance/excess insurance. This volatility may
then exacerbate the insurance cycle and cascade into volatility in the
availability and price of primary medical malpractice insurance as self-
insured entities and primary insurers pass on premium increases to their
customers.5 High premiums or unavailability of coverage may also occur
when insurers have difficulty estimating the ex ante probability of a loss
occurring.6
Changes in premiums are justified when the underlying distribution of
losses changes, but not justified in response to a one-time occurrence,
unless the latter signals a longer-term change in the distribution of losses.
Distinguishing between one-time shocks and more permanent changes
can admittedly be a very difficult task.
Three important aspects of reinsurance/excess coverage that reinsurers
make decisions about are (1) whether or not to cover; (2) what price to
charge for coverage; and (3) which attachment point to set for
coverage.
Attachment points have two purposes. First, if there is no corridor
between primary coverage and reinsurance, the incentive the primary
insurer has for loss mitigation may be reduced. Claims near the point at
which reinsurance applies may not be fought aggressively unless there is
a corridor at which the insured entity is at risk. In particular, primary
insurers may have less reason to exercise caution when the upside risk
is limited by the presence of reinsurance.7 Second, having a corridor
reduces the number of reinsurance claims, thereby reducing a reinsurer’s
administrative expense.8
In late 2003 and early 2004, the first author and colleagues at Duke
University conducted a survey of reinsurers. This was at a time when the
most recent insurance crisis was at its height. Even during this crisis (but
250 Chapter 10

not at the onset of the crisis, when availability problems may be at their
worst), the reinsurers surveyed either wrote or would have considered
writing hospital excess coverage in all states, but sometimes only at high
attachment points. Rather than refuse to write coverage, they quoted
premiums at a very high level in locations when insurers did not want
the business.

Research Findings on Private Reinsurance

Most of what is known about reinsurance comes from industry sources—

just speaking with people in the industry. Very little scholarly research
has been conducted on the topic, perhaps because it is such a specialized
subject, and data needed for research are largely in private hands, not
available to researchers.

Primary Insurer Decisions to Reinsure: Theory and Empirical Evidence

Hoerger, Sloan, and Hassan studied determinants of the decision to
reinsure property-casuality insurance in general and medical malpractice
insurance in particular.9 Their study considers primary medical malprac-
tice demand for reinsurance, not hospital demand for excess insurance.10
In their theoretical framework, the motive for reinsuring is to avoid
bankruptcy. Bankruptcy is costly to insurers, although such costs may
be reduced by the existence of guaranty funds, which were described in
chapter 9. When a bankruptcy occurs, policyholders receive the remain-
ing assets less the administrative cost of distributing the assets.11 The
authors use their model to assess how the insurer’s surplus, size, and
volatility of losses affect the amount of reinsurance the primary insurer
purchases.
Their empirical analysis reveals that insurers facing higher loss volatil-
ity, lower surplus-to-premium ratios (a measure of the amount of
equity the firm has relative to the premiums written in a year, with
higher ratios indicating more funds available for distribution in the
event of an unexpectedly high level of payouts), and smaller primary
insurers demand more reinsurance. These results imply that that if
primary insurers observe that volatility of their losses has increased and
experience a depletion of their net assets, as would occur if they had to
 Reinsurance 251

pay claims that were much larger than anticipated, they will demand
more reinsurance in subsequent years. In this sense, adverse experience
in a year or two will generate higher demand for reinsurance, which
causes reinsurers to bump up against their capacity constraints.12 For this
reason, crises may be exacerbated and extended. Further, it is expected
that smaller single-line medical malpractice insurers will demand more
reinsurance.
At the level of the market (as opposed to the individual firm in a
market), increased volatility and some large losses occurring to more
than one company are likely to increase the price of reinsurance. An
increased price should in turn decrease the demand for reinsurance and/
or could cause some insurers to exit the line of insurance which experi-
enced the increase in volatility. If a series of large paid claims leads
reinsurers to believe that further large payouts have become more likely,
this will increase the price of reinsurance as well. Thus, while a change
in payments on large claims will shift the demand curve for reinsurance
to the right, the resulting price increase (i.e., movement along the demand
curve) will tend to have the opposite effect.
Reinsurance premiums have become very high in recent years.
Reinsurance premium information is highly confidential; in private con-
versations with purchasers of reinsurance, the first author has been
impressed with how high excess insurance premiums charged to hospitals
have become. Not only are premiums high relative to anticipated losses,
but they seem high relative to the maximum loss that the insured entity
could incur.
This raises a question as to why any organization, such as a hospital,
would demand such insurance. Our speculation is that payment of
one or two large claims could adversely affect the hospital’s financial
performance in a year. Hospital administrators are loathe to report
adverse financial results to their boards. Thus, they would rather
essentially prepay their losses in return for a steady cash flow than
fully self-insure the risk of incurring a catastrophic expense and experi-
ence added volatility in cash flows reflecting volatility in payments
for medical malpractice claims. Primary medical malpractice insurers
may be motivated to purchase expensive reinsurance for much the same
reason.
252 Chapter 10

Why Is the Price of Reinsurance So High?

Hoerger et al.’s analysis does not include a measure of the price of
reinsurance. This type of information is very closely held by both
purchasers and suppliers of reinsurance for competitive reasons.
More recently, Froot reports data on reinsurance quantities and
prices for a broad group of catastrophe reinsurance purchases across
the insurance industry.13 He finds that between 1989 and 1998, the
observational period of his study, the ratio of reinsurance premiums to
expected loss ranged from a maximum of over 7 (in 1994) to a minimum
of slightly under 2 (in 1989). Thus, reinsurance premiums were far
greater than losses that might be anticipated, the latter being a rough
benchmark for fair value. He concludes that “Fair pricing does not
prevail in the markets for reinsurance claims, and . . . premiums are a
multiple of expected losses.”14 He concedes that his measure of expected
loss may be downward biased. But such a bias would not explain the
substantial variation in the premium/expected loss ratio between 1989
and 1998.
Does Froot’s result imply that reinsurers are charging excessively
high premiums? Not necessarily. Some markup over anticipated
loss would occur even in a competitive reinsurance market, since
there is plausibly a loading factor attributable to marketing, claims pro-
cessing, and legal costs which such companies incur. More fundamen-
tally, in view of the high volatility of large claims and positive correlations
among large claims (e.g., among jury verdicts),15 some risk premium
(premium set in excess of anticipated loss plus the loading) is certainly
justified. One reason there is no easy answer to this question is lack of
data. There are no publicly available data on premiums, losses of private
reinsurers (in contrast to primary insurers),16 or costs included in the
loading; the data available to Froot are an exception, and are for only
one company.
Froot considers possible explanations for the high price of reinsurance
as measured by the ratio of premiums to losses, and the reasons for cycles
in the ratio of premiums to losses and in the quantity of reinsurance
supplied that have been observed. The first explanation relates to the
high cost of equity capital, particularly for reinsuring the most highly
catastrophic risks. Depletion of equity capital by a large shock (e.g.,
 Reinsurance 253

a series of hurricanes or large verdicts adverse to insurers) is likely to

result in a major capital shortage, with the result that premiums for
reinsurance are raised appreciably above the expected loss of such cover-
age. Froot presents some empirical evidence on reinsurance during the
year following Hurricane Andrew which is consistent with this view.
Such shocks to capital are more likely to happen when the risks are from
earthquakes or hurricanes than from verdicts, but shocks to capital from
natural disasters, by affecting reinsurer equity, could have an impact on
reinsurance premiums in the medical malpractice insurance markets.
Individual verdicts may be correlated, but not to the extent that losses
of individual insureds from earthquakes or hurricanes in the same geo-
graphic area are correlated.17
The second explanation relates to the exercise of market power by
reinsurers. This would lead to increases in price accompanied by decreases
in the quantity of reinsurance. Although there has been some consolida-
tion in the industry, suggesting that each of the remaining participants
in the market may consequently possess more market power, Froot notes
that there was considerable entry into traditional reinsurance in the
1990s. The 1990s were not crisis years, but sellers could have been
poised for entry when and if prices of reinsurance rose. Contestability
(i.e., potential entry by new firms) in the face of increased prices would
limit exercise of market power by existing reinsurers. Moreover, there
appear to be no important barriers to entry in the reinsurance business.
In fact, some companies operate in highly unregulated environments,
such as the Caribbean.
Froot’s third explanation is that the corporate form of reinsurance
may be inefficient. In particular, it may be costly to give discretion to
managers, who may pursue objectives other than firm value maximiza-
tion. Shareholders seem to demand high returns from their investments
in companies that supply reinsurance, perhaps reflecting such “agency”
costs as managers not pursuing the interests of the principals) (the stock-
holders) but pursuing their personal financial interests. A high cost of
equity capital to reinsurers would plausibly be reflected in a higher price
of their product.
Kunreuther, Pauly, and Russell also suggest that suppliers of equity
capital may believe that the high losses they experienced are not random
254 Chapter 10

and may reflect reinsurer mismanagement.18 Although this is a possibil-

ity, neither Froot nor anyone else has provided empirical evidence to
support this agency cost hypothesis. Thus, it should be viewed as
speculative.
The fourth explanation is that reinsurance contracts are illiquid finan-
cial instruments, and investors demand high returns for this reason. In
an alternative, brokerage costs and servicing expenses are high. Froot
rejects such factors (“frictional costs of reinsurance”) as determinants of
observed price and output patterns.
Fifth, high prices of reinsurance may reflect the presence of moral
hazard and adverse selection. Moral hazard arises in this context
since it is costly for reinsurers to monitor underwriting and loss pre-
vention activities of primary insurers.19 Reinsurance agreements often
contain a provision that reinsurers will charge more following a claim,
coupled with a requirement that the reinsurance purchaser20 will con-
tinue to buy reinsurance following the claim. This type of clause might
be used to limit both moral hazard and adverse selection. To the extent
that there is repayment in the form of higher reinsurance premiums fol-
lowing payment of claims, this might explain cycles in the ratio of pre-
miums to expected loss for particular years. Even if the reinsurance
purchaser attempts to switch reinsurers, questions about past claims
experience are likely to be asked. Also, high deductibles, or equivalently
the corridor discussed above, and coinsurance in virtually all of these
contracts limit both adverse selection and moral hazard, the latter being
the incentive for insured parties to use the covered benefit because it is
covered.
Since experience rating is not common in medical malpractice insur-
ance, adverse selection can be a problem for voluntary insurance markets
if premiums do not precisely match risk. That is, those with relatively
high risk will view coverage as particularly attractive since they can
purchase insurance for the same premium as those at lower risk of loss.
Adverse selection is likely when experience rating is not used and high-
loss physicians or hospitals are able to purchase coverage at average
rates. However, premiums of private reinsurers/excess insurers are highly
experience rated and therefore avoid this problem.
 Reinsurance 255

Sixth, ex-post financing of catastrophes, such as by the U.S. govern-

ment in the event of major hurricanes, would reduce demand for private
reinsurance. However, decreased demand for private reinsurance as a
consequence of “crowding out” by public coverage would explain low
quantities of private reinsurance but not high prices. Prices would be
decreased by public provision.21 There was no trend in the percentage of
medical malpractice premiums ceded [premiums collected by primary
medical malpractice insurers, but transfused to a reinsurer, which in turn
bears the risk of poss] during 1991–2002. It remained about 15–16
percent.22
The above explanations may account for high prices of reinsurance.
However, they do much less well in explaining cyclical variation in price
and availability. Among the latter explanations, the strongest and most
compelling one for cycles may be shocks to equity resulting from natural
disasters which could adversely affect the cost of capital to these firms
and their capacity to sell reinsurance. Such explanations as exercise of
market power would seem to operate equally at all phases of the cycle
if this is an important factor at all.

Is There a Strong Case for Government Intervention in This Market?

Reinsurance and Insurance Cycles

Strains on the private reinsurance market appear to be partially respon-
sible for the recent malpractice insurance crisis.23 Medical malpractice is,
however, only one line in the global market for reinsurance and excess
insurance. The early 2000s’ “hardening” (decreased supply, increased
premiums) of the markets for reinsurance globally was experienced by
the market more generally and was not specific to medical malpractice.
Since data from the years following 2000 show increases in premiums
globally, across lines of insurance, and reductions in reinsurer capacity,
a strong argument exists that the most recent crisis was the result of an
external shock to reinsurance reserves from nonmedical catastrophes,
such as the events of September 11, 2001.24
As was discussed in Chapter 2, the phenomenon of the insurance or
underwriting cycle is an important characteristic of insurance markets.
256 Chapter 10

The return to the crisis phase of the cycle can be explained not only by
inadequate premiums, but also by external shocks to capacity resulting
in forecast errors,25 such as natural disasters or terrorism, naïve loss
forecasting by insurers, demand shocks from changes in actuarial calcu-
lations,26 regulatory lags, or public policies. While trends in judicial
decisions or in the propensity toward litigation may cause shocks that
are correlated within a country, correlations across countries for these
reasons seem unlikely.

Market Failure in the Reinsurance Market?

The question of whether the market for reinsurance and excess insurance
is “failing” or is “broken” in some fundamental way is a complicated
one to answer definitively. If a market failure could be conclusively
demonstrated (that is, circumstances in which markets do not produce
economically efficient outcomes), then there would be a rationale for
government intervention in the market.
Market failure could also reflect a deviation from the competitive
norm. Exercise of market power would lead to a higher price and lower
quantity of reinsurance than would prevail under conditions of perfect
competition. As noted in chapter 9, when purchasers lack information,
this might be a source of sellers’ market power. In the market for rein-
surance, however, purchasers (e.g., primary insurers), are sophisticated
more often than not, and, as noted above, there appear to be no impor-
tant barriers to entry. However, sound empirical evidence is lacking.27
In a market economy, absent a specific demonstration that a private
market is unable to achieve social objectives, the tendency is to leave
activities to the workings of private markets.
While at least at first glance the case for government provision of
reinsurance in medical malpractice insurance markets appears to be quite
weak, there may be a justification after all. Withdrawal of care can be
an important public concern. Having continuous medical malpractice
coverage at “affordable rates” may be a prerequisite, at least in some
cases, for maintaining patient access to affordable care. Increases in the
risk of large claims lead to higher premiums for reinsurance for primary
insurers and self-insured hospitals, causing them to raise premiums for
medical malpractice insurance for physicians.28 If physicians do not have
 Reinsurance 257

the ability to raise their fees, they may stop practicing altogether or cease
providing services perceived as high risk. If this is indeed happening, it
is an undesirable outcome and a failure of the market to provide for the
social objective.

Public Intervention in Reinsurance Markets

Public intervention offers some other potential benefits. In particular,
governments may achieve a redistribution which will not occur with
complete reliance on the private sector. For example, government
provision of reinsurance gives higher-risk individuals greater access
to coverage at “affordable” rates than would occur without govern-
ment involvement. Cross-subsidization may be seen by some as unfair,
but in some cases it may achieve an important public purpose, such as
keeping an emergency room open in an area where low-income families
live.29
Moreover, governments can diversify risks over the entire population
and spread past losses to future generations of taxpayers. This is a form
of cross-time diversification that the private market cannot achieve. Also,
the government can constrain adverse selection by enforcement of insur-
ance purchase (mandatory coverage).30
While there are benefits, there are also some arguments against govern-
ment intervention. First, a link between rising premiums and withdrawal
of care has not been demonstrated empirically. This is not to say that
this never occurs, but rather that it is not a sufficiently frequent occur-
rence to explain why people do not get care. Second, there is concern
that government provision will supplant (crowd out) private provision.
Third, as explained below, government intervention in reinsurance
markets has a mixed record—sometimes good and other times not so
good.
Fourth, publicly supplied insurance does not change the underlying
long-run risk; it merely changes the identity of the risk bearers.31
Also, as in other markets, the risk of insolvency should be reflected
in the price of premiums. Policyholders desiring inexpensive insurance
can choose to bear the associated risk of insolvency. But, in reality,
consumers are not in a position to judge an insurer’s underlying insol-
vency risk ex ante.
258 Chapter 10

The Experience of Patient Compensation Funds: Public Provision of

Reinsurance for Medical Malpractice Losses

Background
Patient compensation funds (PCFs) were initially created during the crisis
of 1975–1976 as components of comprehensive malpractice reform leg-
islation. Their goal was to assure availability of medical malpractice
insurance by paying for large losses incurred in a few cases. As discussed
above, it is difficult for medical liability insurers to achieve adequate
diversification against the adverse financial consequences of the most
severe cases. In fact, for some single-line carriers,32 a single very large
claim could result in insolvency. The rich experience of the states that
have implemented PCFs can provide guidance on (1) the utility of the
concept in general and (2) mistakes to avoid if and when a PCF is
implemented.
PCFs are often packaged with tort reforms, including limits on payment
for nonmonetary or total loss, limits on attorney contingent fees, modi-
fication of the collateral source rule, and other statutory changes that
fall under the general rubric of “tort reform.” Not surprisingly, the risk
assumed by PCFs is sensitive to the effects of these provisions, especially
damage caps that place a ceiling on an excess insurer’s dollar exposure
per claim.
As of early 2003, eleven states had established PCFs: Florida (1975),
Indiana (1975), Kansas (1976), Louisiana (1975), Nebraska (1976),
New Mexico (1978), New York (1986), Pennsylvania (1975, 2002),
South Carolina (1976), Wisconsin (1975), and Wyoming (1977).
Florida’s program closed in 1983, having underpriced coverage,33 but
was still paying claims as of April 2003. Pennsylvania passed legislation
in 2002 that schedules a phaseout of its program by 2009. Ohio consid-
ered implementing a PCF, but the Ohio Medical Malpractice Commis-
sion ultimately recommended that no further action be taken. Wyoming’s
program was never implemented.
PCFs are created by state law and organized as either a state agency
or a trust fund (table 10.1). PCF operations are monitored by the state’s
Department of Insurance or by a special Board of Governors. Adminis-
tration—actuarial reviews, claims processing, defense of claims, asset
 Reinsurance 259

management, and so on—is performed by a dedicated staff and/or by

outside organizations retained by the PCF.
Participation may be mandatory—all health care providers fitting the
statutory definition must obtain excess coverage through the PCF—or
voluntary, with providers enrolling at their option. Eligibility for PCF
coverage is triggered at levels that range from $100,000 to $1,000,000
per occurrence. Some PCFs offer unlimited excess coverage, but
most cover only an incremental layer of $500,000–$1,000,000
per occurrence.
The programs typically are funded from premium income and invest-
ment returns, not from state subsidies, a notable exception being New
York State. Providers pay premiums to the state PCF as well as to private
primary insurers. Accordingly, PCFs may improve insurance availability
but nevertheless be expensive for policyholders. Assessments are gener-
ally structured as a fraction of the premium paid for primary coverage,
and may be paid separately to the PCF or collected and passed along by
the primary insurer. PCF assessments are not experience rated except to
the extent that prior experience is reflected in the insured’s primary insur-
ance premiums. PCFs do vary premium contributions by specialty, either
mirroring physicians’ primary insurance classification (as in Pennsylva-
nia) or establishing a few specialty-based risk classes (e.g., four in
Wisconsin). Some PCFs act like insurers and maintain reserves on unpaid
claims. Others are financed on a pay-as-you-go basis, assessing premiums
as funds are expended. Like JUAs, moreover, PCFs often have the author-
ity to assess insured physicians retroactively to cover unanticipated
losses.34
The first author and Duke University colleagues surveyed nine PCFs
in late 2003 and early 2004. The survey revealed that the two major
motivations for forming PCFs were to provide (1) physicians and hospi-
tals with affordable and reliable coverage by reducing volatility in losses
from large claims, and (2) adequate compensation for injured patients.
Medical organizations were interested in the first objective, and the trial
bar was interested in the second. In addition to the survey, the authors
obtained other information available on these PCFs, including financial
data. No organizational representative surveyed could give direct evi-
dence of improvements in availability or affordability due to the state’s
260 Chapter 10

Table 10.1
Summary of Major PCF Provisions

Enabling Financial
State Legislation Structure Participation Eligibility

FL Fla. Stat. nongovernmental physicians phys./hosp.

§766.105 trust fund voluntary,
hospitals
mandatory

IN Burns Ind. separate trust voluntary phys./hosp.

Code Ann. account
§34-18-6-
and 34-18-7

KS K.S.A. state mandatory phys./hosp.

§40-3401–3419 treasury
trust

LA La. R.S. separate escrow voluntary phys./hosp.

§40:1299.41–49 fund

NE R.R.S. Neb. separate trust voluntary physicians

§44-2801–2855 account

NM N.M. Stat. Ann. state voluntary phys./hosp.

§41-5-1–29 treasury
trust

NY N.Y. C.L.S. Ins. excess voluntary physicians,

§5502 liability dentists with
pool hospital
privileges
 Reinsurance 261

Required Primary Coverage Funding Reserves

Coverage Limits Approach Authorized?

$250K/claim, Phys: annual, yes, but maximum

$500K/ $1M/$3M or semiannual, collection of
occurrence $2M/4M or quarterly $17.5M per year
(1990 levels Hosp: assessments
adjusted for $2.5M/claim paid to fund
CPI)
Phys: $250K/ $1.0M per Decided using yes
$750K occurrence actuarial
Hosp: $250K/ principles;
$5M or $7.5M collected by
primary insurer
$200K/$600K purchaser’s decided using yes
options: of actuarial
$100K/$300K, principles;
$300K/$900K, collected by
or $800K/$2.4M primary insurer
$100K/ $500K + future decided using Yes, but surcharge
$300K medical expenses actuarial must be reduced if
primary coverage principles; fund exceeds $15
collected by million
primary insurer
Phys: $1.05M per Assessments $4.5M maximum
200K/ occurrence as percentage
$600K of underlying
Hosp: premiums
$200K/$M through insurer
$200K/ unlimited decided using yes
$600K medical + actuarial
$600K principles,
noneconomic collected by
primary insurer
$1M/$3M $1M per state general not applicable
occurrence fund
262 Chapter 10

Table 10.1
(continued)

Enabling Financial
State Legislation Structure Participation Eligibility

NC* N.C. Gen. Stat. trust fund voluntary physicians

§58-47-1–50

OR** ORS trust fund voluntary physicians

§752.005–055

PA 40 P.S. state mandatory phys., hosp.,

§1303.101–910 treasury health care
P.L 154, no. 13 fund practitioners

SC S.C. Code Ann. state voluntary phys./hosp.

§38-79 treasury
fund

WI Wis. Stat. not mandatory phys./hosp.

§655.27 specified with
exceptions

WY** Wyo. Stat. trust fund voluntary physicians

§26-33-101, 105

* Repealed.
** Fund never enacted.
Notes: In 2004, Alabama was considering creating a patient compensation fund
for nursing homes as part of a larger bill that would cap damages at
$250,000.
Nevada Governor Kenny Guinn stated his support for a patient compensation
fund as part of a larger medical malpractice overhaul in a policy briefing in Roll
Call “The Health of Our Nation Policy,” Roll Call (July 2003).
 Reinsurance 263

Required Primary Coverage Funding Reserves

Coverage Limits Approach Authorized?

$100,000/yr. $2M/yr assessments yes

determined
by board,
paid to fund
$200,000/yr. unstated assessments undetermined:
determined fund must be set
by board, up by executive
paid to fund branch
phys.: $500K/ $500K/ based on no
$1.5M $1.5M primary
Hosp.: $500K/ premium
$2.5M collected by
insurer
$200K/$600K unlimited Annual yes
medical + member fees
$600K and default
noneconomic assessments
to fund
$1M/$3M unlimited assessments no
determined
by board
billed to
health care
providers
$50K/ $1M in excess decided using $4M maximum
occurrence coverage actuarial
principles,
collected by
primary insurer
West Virginia’s House of Representatives has passed H.B. 2122, a plan which,
among other things, would create a patient injury compensation study board to
design and implement a patient compensation fund.
Source: Based on “Preliminary Report on the Feasibility of an Ohio Patients
Compensation Fund,” produced for the Ohio Department of Insurance by Robert
J. Walling, of Pinnacle Actuarial Resources. Thanks to Justin Sadowsky of
Columbia Law School for updating this table.
264 Chapter 10

PCF. However, respondents did indicate that PCFs reduced the volatility
of losses experienced by insurers. Since all the PCFs, with the exception
of Pennsylvania’s, are handling the largest losses, this claim is almost
certainly valid.
Respondents to this survey also stated that PCFs have increased the
attractiveness of their states to primary insurers by limiting their expo-
sure to high losses, even in states without caps on nonmonetary or total
loss. However, no direct empirical evidence exists to verify that this has
been a factor in actually attracting primary insurers. Brokers and private
reinsurers/excess insurers, as competitors of public insurers, were not
nearly as enthusiastic about PCFs as were respondents from the PCFs.

Performance of PCFs—Do They Reduce Total Payments for Medical

Malpractice Claims?

Trends in claims frequency, from data filed with the National Practitio-
ner Data Bank (NPDB), are strikingly similar for states with PCFs and
those without such programs. During 1995–2002, PCF states had sys-
tematically higher claims frequency. There were sizable differences in
recent trends in losses paid by PCFs among the five PCF states for which
trends could be measured between 1998 and 2002.35 Unfortunately, it is
not possible to calculate total losses incurred or mean total severity of
loss by private and public insurers in PCF states because PCF payments
are not included in NPDB data. Furthermore, it is difficult to attribute
the trends solely to the presence of a PCF, as other factors may have
been responsible.
Most states with PCFs do not have damage caps, but do have an upper
limit on PCF coverage. Thus, even with PCF coverage, health care pro-
viders remain at risk for very high claims if they do not or cannot pur-
chase insurance from a private reinsurance/excess insurer. With some
notable exceptions (e.g., Pennsylvania, which covers a middle layer of
losses), it has been uncommon for juries to exceed the PCF layer of cov-
erage in their awards.36
The frequency with which private reinsurance is purchased in PCF
states is not known. PCFs understandably limit demand for private insur-
ance coverage for very large losses. According to a survey of reinsurers
 Reinsurance 265

and brokers conducted by the first author and colleagues at Duke Uni-
versity in 2004, reinsurers base the availability and price of excess cover-
age across locations on several factors: “good PCFs”; cultural factors
(for example, less sophisticated law firms, more educated juries, less liti-
gious environments); and whether or not hospitals are seen as assets to
the community or as adversaries.
But the most important determinants of premiums are exposure in
terms of size and scope of services provided, and claims experience over
a long time period (e.g., a decade). The importance of high awards in
driving premiums for excess coverage cannot be overstated, and likely is
a main factor in insurers’ decisions to differentiate based on geography.

Structural Features of PCFs

There are four main differences between private liability insurers and
state PCFs: (1) public sponsorship, which ensures availability of coverage
when commercial insurers find other lines of insurance or locations more
attractive; (2) mandatory participation in some states, though generally
with a choice of private insurer; (3) lack of regulatory oversight in some
states to assure adequacy of rates; and (4) pay-as-you-go financing in
some states. Depending on how the program is structured, PCFs either
promote or discourage entry of private insurers.

Public Sponsorship
State governments operate PCFs. Public sponsorship has an important
advantage: assuring availability of coverage. Like joint underwriting
associations (JUAs) and unlike private insurers, PCFs do not withdraw
from the market during crisis periods. Demand for private reinsurance
by primary medical malpractice insurers is directly related to the volatil-
ity of loss.37
However, financing generally comes entirely from premiums paid by
physicians and hospitals as well as investment income. Neither state
general funds nor revenue from a dedicated tax source are typically used
to support PCFs. In addition, PCFs do not reduce medical liability expo-
sure, unless they undertake specific loss prevention actions. Rather, they
transfer costs to a different funding mechanism.38
266 Chapter 10

Primary malpractice insurers that purchase private reinsurance have

an incentive to defend claims—even those that clearly exceed the primary
policy limits—because current losses are likely to be reflected in future
reinsurance premiums. This incentive is attenuated when a PCF is
involved, particularly if the PCF has the power to assess insured provid-
ers directly for losses in excess of those initially projected.

Mandatory Participation
Voluntary insurance markets are vulnerable to adverse selection if
premiums do not precisely match risk. As previously noted, medical
malpractice insurance is not usually experience rated. Compulsory par-
ticipation in a PCF can avoid adverse selection, which is otherwise likely
to occur if high-loss physicians or hospitals are able to obtain coverage
at average rates. However, requiring low-risk participants to subsidize
high-risk participants may be viewed as unfair.

Lack of Regulatory Oversight

Many arguments for regulatory oversight of insurers are thought not to
apply to PCFs. Because PCFs are public organizations, they plausibly
lack an incentive to exploit their dominant market position by charging
monopoly-level premiums. Nor are they driven by the profit motive to
engage in risky financial decisions that may lead to insolvency. PCF assets
may also be managed by a well-staffed, politically accountable unit of
state government that has responsibility for several state agencies. On
the other hand, problems do arise in practice. Because of budgetary
constraints, for example, the PCF may not be well staffed, and civil
service rules may limit its ability to compete with the private sector for
personnel.

Financing in Advance or Pay-as-You-Go Financing

PCFs may collect funds adequate to pay all losses and associated expenses
from claims occurring during the current policy year whenever those
amounts are actually spent, or they may be funded on a pay-as-you-go
basis. Under the latter approach, the PCF limits its assessments to
amounts anticipated to be spent on claims and expenses in the following
year. This approach has practical appeal, particularly in an unpredict-
 Reinsurance 267

able, “long-tail” line such as medical liability insurance, and helps solve
short-term crises in availability of excess coverage without imposing the
immediate pain of high premium assessments. In the first few years of a
PCF’s life span, losses tend to be low because most claims have not yet
been resolved, allowing assessments to be low as well. Later, however,
losses mount, and PCFs often must raise premiums sharply, incurring
the wrath of premium payers and precipitating political pressures for
reform.

Structural Weaknesses of PCFs

PCFs also have several structural weaknesses, although there is substan-

tial heterogeneity among plans.

Continued Provider Vulnerability

Because most statutes have not established upper limits on liability,
health care providers remain vulnerable to very high dollar claims.39

Reduced Incentives for Loss Prevention

The existence of a PCF may reduce incentives for loss prevention: improv-
ing patient safety, reducing the probability that a claim is filed, and
managing claims to reduce the amount of indemnity, legal fees, and other
expenses incurred by defendants.
PCFs do not surcharge primary insurers based on their loss experience.
By contrast, private reinsurers set premiums on an experience-rated
basis. A primary insurer with a poor loss history is likely to be at a dis-
advantage in the market for reinsurance, and this will likely translate
into higher premiums for its customers. A parallel argument applies to
hospitals. If hospitals can obtain excess insurance that is not experience
rated, they have a reduced incentive to be conscious about patient safety
and avoid large claims.

Voluntary PCFs and Adverse Selection

In some states, participation in the PCF is voluntary. A provider therefore
can avoid a full assessment by not renewing after the PCF becomes
expensive. Providers who are at low risk for future claims will drop out
268 Chapter 10

of the PCF at that point, leaving only high-risk providers enrolled. This
is a classic adverse selection problem.40 Unless the PCF is subsidized from
another source, it will eventually face chronic deficits or charge such high
assessments to its few remaining customers that it collapses.

Loss-Reserving Practices
PCFs differ according to whether they reserve for anticipated losses or
operate on a pay-as-you-go basis. Essential to a PCF’s long-term viability
is the method of financing the program. Several funding mechanisms
exist, including reserving for anticipated losses by making actuarial pro-
jections, and pay-as-you-go financing.41
Setting premiums based on anticipated future payments from claims
filed in a particular policy year is the standard financing approach in the
private insurance sector.42 Such a system is known as loss-reserving and
involves educated guesses of losses, with premiums being set on the basis
of predictions of an actuary. Lack of enthusiasm on the part of health
care providers for increased premiums based on less than certain projec-
tions is reasonable. Nonetheless, loss-reserving has become the time-
tested, traditional approach for private insurers.43
Loss-reserving also eliminates the main problems with pay-as-you-go
financing, which are intergenerational transfers and inevitably increasing
funding requirements. Initially several state PCFs employed pay-as-you-
go financing, which assesses providers based on the currently incurred
PCF losses. Intergenerational transfers are created under this system
because current providers can be charged premiums based on claims that
might have originated many years in the past. Health care providers who
leave or retire enjoy protection from the PCF despite not being fully
assessed for their coverage. Since trends in the cost of living are dwarfed
by increases in malpractice losses, under pay-as-you-go financing, pre-
miums unavoidably increase and providers entering practice are among
the financial losers. These speculations have not been proven, and empiri-
cal evidence is still needed to demonstrate that these intergenerational
transfers are truly causing older providers to retire early or to move, and
are deterring younger providers from entering the state.44
Pay-as-you-go financing is a common practice among social insurance
programs, such as Social Security and Medicare. In the Social Security
 Reinsurance 269

program, for example, premium taxes from currently employed persons

and their employers pay for benefits of retired workers. The implicit
contract is that although younger persons subsidize the benefits of older
persons, younger persons’ benefits will in turn be subsidized by others
when they become age-eligible.
Even though most states with PCFs do establish loss reserves, when
actuarial evaluations are performed, the recommendations are not always
followed.45 This problem is not limited to the public sector. A survey of
private primary malpractice insurers revealed much the same picture;
insurers said that they had overridden their actuaries’ recommendation
at least once in the previous five years.46 Reasons may differ between the
two groups, with private insurers more worried about losing market
share and PCFs responding to political rather than competitive
pressures.
Since it seems prudent for insurers to loss-reserve, why would legisla-
tures in some states have eschewed the practice? One reason, suggested
by Hofflander and Nye (1985), is that it simplifies administration. Instead
of having to compute reserves and invest those funds prudently, the PCF
assumes that providers are aware of the liability that is accruing and
hold reserves of their own. Reserves held by public agencies are also
vulnerable to exploitation for unrelated purposes.47
The most important explanation, however, is politics. Failure to reserve
attracts political support for a PCF because excess coverage seems inex-
pensive in the fund’s initial years. Payment obligations are delayed so
that many of the elected officials in office during the beginning years are
likely to have moved on by the time liabilities must be financed. At first,
losses are low because resolution is slow for losses that have already
occurred and claims are infrequent. Without having to pay the full bill,
pay-as-you-go financing gives the appearance of fixing a crisis. However,
over time, losses mount, another round of medical malpractice reform
is called for, and PCFs often have to raise premiums.
Proponents of pay-as-you-go financing argue that such a system may
help insulate providers from nonmeritorious claims since it does not base
assessments on a provider’s claims history. However, this holds true only
if juries often make mistakes in their determinations of damages and in
their findings of liability.48
270 Chapter 10

Since health care providers would remain at risk if the state PCF is
vulnerable to an eventual insolvency, the arguments for loss-reserving
over pay-as-you-go financing appear much more convincing. When the
unfunded liability from past policy years eventually becomes due and
payable, it is easy to label the malpractice system “out of control” instead
of confronting the design flaws in the PCF.49
Pay-as-you-go financing for medical malpractice coverage has a certain
rough justice. If juries frequently make errors in their findings of liability
and determinations of damages, a pay-as-you go system that guarantees
insurability and does not base assessments on a provider’s claims history
helps insulate providers from nonmeritorious claims.
However, the countervailing arguments are stronger, and apply equally
to public agencies and private insurers. First, the objective of insurance
is to protect policyholders against loss. If there is substantial insolvency
risk, health care providers remain vulnerable.50 Second, insurers have a
comparative advantage in loss-reserving. Unlike actuaries, health care
providers do not possess the requisite data or expertise to make such
projections.
Third, pay-as-you-go financing inevitably leads to intergenerational
transfers. The pool of health care providers in practice at a particular
time may differ substantially from the pool that existed at the time the
losses were incurred. If current premiums rise to cover past losses, this
could discourage entry by new providers and encourage exit of existing
ones. In a pay-as-you-go system, providers who retire or leave the state
are subsidized by those who remain or enter. The former realized the
benefits of excess coverage through the PCF while they were in practice
in the state, but were not assessed a full premium for such coverage.
Conversely, providers who enter practice in a PCF’s later years or are
still many years from retirement are the financial losers. As with Social
Security, one could justify this cross-subsidy in terms of a social contract
between “young” and “old” providers. Because the secular trend in
malpractice losses far exceeds changes in the cost of living, however, the
old receive a substantial net subsidy. The desire to avoid these obligations
arguably deters younger providers from entering practice in the state,
and induces older providers to retire early or move to states without PCF
assessments.
 Reinsurance 271

To Subsidize or Not to Subsidize from Sources Other than Provider

Premiums
Rather than employing a PCF to provide reinsurance directly, New York
subsidizes such coverage, and is the only state to do this. The assumption
and justification are that medical care is a merit want. A merit want is
a good that society deems should be available to all, and its allocation
should not be subject to normal market forces.51
Reinsurance excess coverage is a vital input in the production of
affordable medical care, and so it is considered necessary to subsidize it
in order to assure accessibility. In addition, the intergenerational inequi-
ties of the pay-as-you-go approach are avoided. Such subsidies might
increase the demand for coverage and result in an increase in total pre-
miums (prior to the subsidy) if implemented in a less regulated environ-
ment than New York.

Lessons Learned from the PCF Experiences

Funding Is Often Through Implicit Taxation

If a goal of government interventions is to subsidize high-cost insurance,
and if injury victims are to be protected from loss, then revenue for the
subsidy must come from somewhere else.
Unfortunately, a recurrent theme is that the taxes are implicit. Rather
than raise revenue explicitly, it is often easier to shift the burden to
unsuspecting groups, such as random classes of insurance policyholders.
Covering the shortfalls in medical liability insurance with automobile
liability insurance revenues, as in Pennsylvania, seems inequitable. If
assuring the availability of medical care is a societal priority, the tax base
should be as broad as possible, and taxpayers should understand the tax
rates they pay.
Another implicit transfer occurs under the PCFs set up on a pay-as-
you-go rather than a fully reserved basis. Here, health care providers do
not pay the full freight of their losses at the time they are incurred. The
bills come due in the end, which often coincides with the time the country
reawakens to a new medical malpractice crisis.
272 Chapter 10

Confusion about the Role of Public Insurers

In reading unpublished correspondence from public insurers, we are
struck by the lack of consensus about whether or not they truly are
insurers. Often when assessments of programs are conducted, the review-
ers remark that the agency lacks sufficient actuarial capacity. This theme
is closely related to others stated previously: enactment of programs
during crises when actuarial capacity seems a minor detail, and the avail-
ability of implicit taxes which seem to make standard loss-reserving
practices unnecessary.
Like private insurers, public insurers collect premiums and bear risk.
In a long-tail line, such as medical liability, public insurers, like their
private counterparts, face actuarial uncertainty. That the insurer is public
does not change the fact that expected loss differs among policyholders.
The rationale for experience rating and risk classification, loss-reserving,
and prudence in investing reserves therefore applies equally to public and
private insurers. The main difference is one of mission. A public agency’s
decision to supply insurance should be less responsive to immediate rates
of return, although even public insurers’ deficits eventually require an
offsetting revenue source.

Programs’ Lack of External Oversight

These programs frequently operate for years without public oversight.
Often there is no written evidence that an organization has ever been
formally evaluated.52 Lack of oversight should be the exception rather
than (almost) the rule. Sound public policy requires that program evalu-
ations be scheduled when public programs are implemented. Prompt
evaluation may identify problems in their early stages and facilitate
midcourse corrections. Evaluation also provides an opportunity to discuss
program goals, assess whether goals are being met, and determine
whether there still is a public need for the program. Program oversight
should be independent but geared to the mission of the agency.

Relevant Experiences from Other Contexts

There are parallels between losses from medical liability and losses from
catastrophes such as terrorism, earthquakes, hurricanes, and floods. The
 Reinsurance 273

record of government intervention in providing catastrophic coverage,

like that of PCFs, is mixed. Again there are important implementation
issues.

Federal Emergency Management Agency (FEMA)

In 1968 the Housing and Urban Development Act created the National
Flood Insurance Program (NFIP), to be controlled by the newly formed
Federal Insurance and Mitigation Administration (FIMA), a subagency
of FEMA.53 The NFIP was designed to provide subsidized flood insur-
ance to property owners unable to find affordable private insurance.54
Unlike almost all PCFs, except New York, a broad general tax revenue
base supports FEMA, not payments from parties at risk of incurring
losses from natural disasters or assessments from insurance companies.
Shortly after the program began, subsidized rates of insurance were
not creating incentives for communities to join the NFIP or to individu-
ally purchase flood insurance. In response, Congress created the Flood
Disaster Protection Act of 1973, which required federally regulated
lenders to make flood insurance a condition of granting or continuing a
loan.55 Over the course of years and repeated devastation from natural
disasters, the NFIP grew; as of 2003, there were 4.5 million flood policies
in place, a huge increase from the 95,000 such policies in the early
1970s.56
FEMA did not have the ability to price this insurance and reinsurance
more accurately than the market—in fact, quite the opposite; the same
high-risk properties have been covered repeatedly at premiums below
expected loss.57 By late 2005, FEMA had depleted its reserves and was
unable to continue reinsuring. Shortly thereafter, FEMA ordered private
insurers to stop paying on private claims.58 This left many policyholders
without payment or recourse. As of 2005, policyholders had filed close
to 225,000 flood claims from hurricanes Katrina, Rita, and Wilma; of
these claims, FEMA has paid only 15 percent of the $23 billion in losses
resulting from these three hurricanes.59

State-Sponsored Programs
States have intervened to provide coverage to otherwise uninsurable
owners. Florida created the Florida Windstorm Underwriting Associa-
274 Chapter 10

tion (FWUA) in 1970 to provide hurricane and windstorm coverage to

property owners who could not obtain it from private insurance com-
panies.60 In 2002, the legislature combined the residential property and
casualty underwriting association with the FWUA, forming Citizens
Property Insurance Corporation (Citizens).61 As with NFIP, Florida has
inadequate reserves from charging rates that are too low in comparison
to the risk. As a result of these unusually low rates, Citizens is undercut-
ting the private market.62

Implications for Public Intervention in Markets for Medical

Malpractice Reinsurance
The record of FEMA presented above predates Hurricane Katrina and
the problems for FEMA that followed, only some of which related to
insurance of catastrophes. However, the earlier failure to assess and
contain program costs, although partly attributable to FEMA, may also
speak to the desirability of expanding the public sector’s role as a pro-
vider of reinsurance. FWUA’s underreserving and crowding out of private
coverage should likewise be a warning that some oversight of operations
is essential.

Conclusions and Recommendations

Is the Private Market for Reinsurance and Excess Insurance Broken

and in Need of Repair?
Private reinsurance/excess insurance plays a vital role in provision of
medical malpractice insurance, yet it is often expensive, and the markets
are volatile. Although some volatility is inevitable, for personal investors
the advice is to take actions that reduce volatility. Insurance cycles exact
some cost, and volatility in reinsurance and excess insurance markets
contributes to cycles in medical malpractice insurance premiums and
availability. The fact that much of the reinsurance is not subject to
federal or state regulation is at most a secondary concern.

Are PCFs Worth Having?

A case can be made for government involvement in providing insurance
for high-cost claims. First, a public agency can provide reinsurance at a
 Reinsurance 275

lower price and, just as important, the volatility of pricing is reduced.

Thus, having a PCF may help avert at least some crises. PCFs are appeal-
ing because their decisions to supply coverage are not guided by prospec-
tive rates of return, so they can keep excess coverage available. Public
funding does not eliminate the risk but shifts the burden of risk bearing;
however, such shifts may be in accord with social preferences. The public
status of PCFs implies that they are willing to supply coverage under
circumstances that are unattractive to private, profit-seeking insurers, but
does not alter their function as risk bearers.
But most important, the shortcomings of existing PCFs, such as
Pennsylvania’s, derive not from the concept but from how it was
implemented. If a PCF funds liabilities according to actuarial principles,
implements some form of experience rating, provides corridors to miti-
gate moral hazard, and/or requires participation (is mandatory), PCFs
can play a constructive role. Although this discussion has focused on
deficiencies, there are examples of well-run public insurers. And there is
evidence that some of the errors of the past are being corrected.
In particular, pay-as-you-go financing offers short-term political
advantages, but exacts a much higher price in the long run. The fact that
prudent loss-reserving and accurate premium assessments may create a
larger fund which subsequently becomes a political target for spending
unrelated to medical liability seems a small price to pay for stability and
solvency.
Finally, other instruments for dealing with high-end losses are being
implemented. Hedge instruments are a financial mechanism to transfer
tail risk to investors. They include catastrophe bonds, catastrophe
options, and catastrophe puts.63 Insurers have recently begun to use these
instruments to protect themselves against large losses. Such losses should
be more easily absorbed in a multitrillion-dollar global capital market
as opposed to limiting them to a single property-casualty insurance
industry, as in the case of reinsurance.

Designing and Implementing an Effective Patient Compensation Fund

If implemented, a PCF should have the following characteristics. There
should be a high dollar threshold for coverage and, unlike current prac-
tice, the dollar threshold should be indexed to reflect changes in price
276 Chapter 10

levels. Since medical losses reflect the cost of medical care in part, the
price index should be a blend of the medical and the all-items Consumer
Price Index. The threshold would initially be set not to exceed the top
1 percent of the claims distribution.64 Except when dictated by state
statute in the form of a dollar limit on payment for nonmonetary or total
loss, there should be no upper limit on coverage. This would eliminate
the motive of insurers or self-insured organizations to obtain higher
layers of coverage from private sources.65
The program would use standard loss-reserving principles, not pay-as-
you-ago financing. Premiums would be paid by all health care providers
who purchase medical malpractice insurance or self-insure for such loss.
An independent committee of actuaries would oversee the process of
calculating actuarial value of the losses and setting premiums. The pre-
miums at baseline would reflect factors that are predictive of high loss
claims, including specialty and types of procedures performed to the
extent that the procedure mix affects expected loss. As done by private
reinsurers, subsequent premiums would reflect the policyholder’s loss
experience. This would be more effectively done if the experience rating
unit is a large physician group or a hospital rather than an individual
physician.
The states serve as independent laboratories for specific innovations.
It is therefore particularly important that states considering public insur-
ance programs learn from the experiences of others. Unfortunately, while
the diversity of the U.S. federal system is a strength, states seldom avail
themselves of knowledge gained in other states, and therefore tend to
reinvent the wheel.
Some public insurers also could benefit from structured technical assis-
tance, which could be financed in part by the agencies receiving it.
Research would be facilitated by data clearinghouses, which exist in
other fields but not for medical liability or liability insurance more gener-
ally. It is important to develop criteria for gauging program success
which are acceptable to the various stakeholders. Such criteria should
reflect goals in addition to cost containment.
11
No-Fault for Medical Injuries

Rationale for No-Fault

Under a tort system, compensation is based on a case-by-case determina-

tion of fault. No-fault programs for iatrogenic injuries would provide
compensation for injuries caused by medical care without regard to
fault of the medical provider. Such programs involve more than merely
discarding a critical element of tort—the negligence rule. Claims
are adjudicated by a special administrative agency rather than by courts,
and the benefits may be scheduled or computed on an individual basis,
but these is little or no payment for pain and suffering. No-fault seeks
to address deficiencies in tort, as the critique from Australia illustrates
(box 11.1)
Proponents of no-fault cite three sources of savings.1 First, benefits are
generally only for monetary loss with collateral source offsets.2 Second,
eliminating fault should reduce litigation cost. Third, when experience
rating is combined with no-fault, there should be added deterrence of
injuries. The first advantage can be obtained under tort. Although there
is a strong conceptual argument for experience rating, not all no-fault
plans are experience rated. Furthermore, a no-fault program can be
financed by a broad-based tax on medical providers, insurance compa-
nies, or general revenues.3 Financing by a broad tax would largely elimi-
nate deterrence, since losses attributed to physicians and hospitals would
be picked up by taxpayers without defendants incurring a financial
penalty. However, this option may be viewed as more socially equitable,
and having a broad tax base may ultimately be the only way to have a
278 Chapter 11

Box 11.1
Arguments Against Common-Law Negligence Action as a Basis for Compen-
sating Transport Accident Victims

• the failure of the common-law negligence action to provide compensa-

tion for a substantial proportion of transport accident victims;
• the failure of the fault principle to fulfil [sic] its stated aims and the

practical difficulties in its application;

• the deficiencies of assessing damages on a once-and-for-all basis;
•the difficulties and inconsistencies which arise in assessing damages for
noneconomic loss;
• the adverse effects of the common-law negligence action on the rehabili-
tation of many transport accident victims;
• the delays and consequent hardship experienced by many transport
accident victims in obtaining common-law damages;
• the burden on the court system, and the drain on judicial resources,
caused by deciding claims arising out of transport accidents;
• the substantial legal and administrative costs associated with common-
law negligence actions; and
• the increasing cost to the community of a compensation system relying
heavily on the common-law negligence action.

Source: New South Wales Law Reform Commission (1984), pp. 46–47.

broad no-fault program for iatrogenic injuries. As for the second reason,
eliminating negligence does not eliminate the possibility of litigating over
injury cause under no-fault.
Although there may be a saving in administrative cost and from
not paying for nonmonetary loss, one would indeed expect that compen-
sating a much larger number of injuries would increase the cost relative
to tort. Yet, advocates for no-fault alternatives contend that a major
positive attribute of no-fault is the reduction in expense involved in pay-
ments to lawyers, dispute resolution, and reduced time in resolving
claims. Rather than lawyers from both sides spending years litigating a
dispute and incurring great expense on vigorously determining negli-
gence, a no-fault system saves the legal expense and sets up predeter-
 No-Fault for Medical Injuries 279

mined awards for certain injuries, resulting in injury victims being

compensated much quicker. Other possible advantages cited include
more expert claims resolution, more effective deterrence through more
systematic case identification, and greater fairness and predictability than
under tort.4

No-Fault Programs for Iatrogenic Injuries

One set of no-fault proposals would pay for all iatrogenic injuries from
a general revenue source rather than premiums paid by health profes-
sionals. Claims adjudication would be done by an administrative agency
rather than by courts. Payment for nonmonetary loss would be elimi-
nated or capped at a low level. There would be collateral source offset
for monetary loss. Under a proposal advanced by Weiler, Hiatt, New-
house, and colleagues,5 medical malpractice claims in the United States
would be resolved under a no-fault system. A surcharge assessed to
patients for each day spent in the hospital would provide revenue for the
no-fault fund. A professional panel that would also determine monetary
losses would review the appropriate threshold for medical injury and
eligibility.
Weiler et al. argue that their no-fault plan would be no more costly,
or could be made no more costly, than medical malpractice is currently,
even though more injury victims would be compensated, and compen-
sated much more speedily.6 The lower overhead expense and speedier
compensation records of no-fault systems in Sweden and New Zealand,
and of workers’ compensation in the United States, are taken as evidence
that such a program could be financed at little or no additional cost over
what is now spent on tort. Some recent published empirical evidence
buttresses the argument that no-fault would be no more expensive than
tort.7 While these programs have indeed achieved some successes, we
question whether or not their experiences are relevant to the United
States context. Further, the experiences with the no-fault programs in
Florida and Virginia raise issues of their own. A more modest proposal
would shift the locus of liability from the individual physician to an
enterprise such as a hospital or health plan.
280 Chapter 11

U.S. Experience with Medical No-Fault

Even though policy discussions of no-fault in the United States often refer
to the experience in foreign countries, there is rich experience with no-
fault in the United States as well. No-fault has been implemented for
birth-related injuries in two states. A national no-fault program for
adverse outcomes attributable to vaccines has been in effect for two
decades. A number of states have no-fault programs for motor vehicle
accidents. And for nearly a century, states have had workers’ compensa-
tion programs for injuries occurring in the workplace.

State Programs for Neurologically Impaired Infants

In response to the medical malpractice crisis of the mid-1980s, Florida
and Virginia implemented no-fault programs for severely neurologically
impaired infants. The first no-fault program in the United States was
Virginia’s Birth-Related Injury Fund (BIF) in 1988. Shortly thereafter,
Florida enacted legislation establishing the Neurological Injury Compen-
sation Association (NICA) in 1989. The primary initial goal of both
programs was achieving savings in medical malpractice insurance pay-
ments for birth-related injuries and improved patient access to obstetrical
care.
Obstetricians faced higher and more rapidly increasing insurance pre-
miums through the 1980s. Because of the high premiums, legislators
were told that access to care was threatened, either from obstetrician-
gynecologists leaving the state for states in which premiums were lower
or from such physicians dropping the practice of obstetrics, focusing
instead on gynecology, for which medical malpractice insurance premi-
ums were considerably lower. Improving compensation for injury victims
and speedier claims resolution were secondary objectives.
Both programs were established as, and continue to be, true no-fault
programs. Fault is not a criterion for payment. To be eligible for payment,
the infant or child must satisfy quite narrow eligibility criteria and the
provider must have been a participant in the no-fault program at the
time of the delivery. The clinical criteria for eligibility for coverage are
so narrow that most birth-injured children are ineligible. Of the two
programs, BIF has the stricter eligibility criteria. Benefits are not sched-
 No-Fault for Medical Injuries 281

uled, but are set by the agencies operating the no-fault programs on a
case-specific basis. Payments are paid at the time the expense is incurred.
Eligible expenses to be covered are medical, custodial, rehabilitative, and
educational; there are also payments for special vehicles and modifica-
tions to homes necessary for care of the child. Physician participation in
these programs is voluntary in both states, and Florida makes hospital
participation mandatory.
Other similarities between BIF and NICA include operation by inde-
pendent public agencies, lower assessments by the programs for nonpar-
ticipating physicians, and the exclusion of tort for injuries covered under
the programs. Clinical criteria for injury-victim eligibility are designed
to match those cases that ordinarily result in high payments in tort. Both
programs substantially eliminate payment for nonmonetary loss, and
lawyers’ fees are much lower under no-fault than under tort.
These two no-fault programs are government-run. This is in contrast
to motor vehicle no-fault and workers’ compensation, for which the
states set the framework, but private organizations operate the programs.
Since NICA and BIF enroll so few injury victims, it seems unlikely that
private for-profit firms would have been interested in operating programs
at such a small scale. Realistically, there could have only been one private
insurer per state.
Designers of both BIF and NICA intended that no-fault would totally
replace tort for eligible cases. Applying to these programs for compensa-
tion is voluntary. Neither program has actively sought out applicants in
case finding. Case finding is the act of finding individuals for medical
treatment. Since the programs rely on a narrow premium base, case
finding would be disastrous to their finances.
No-fault programs should be evaluated by examining performance in
assuring availability of reasonably priced liability coverage, improved
injury deterrence, efficiency in administration and management of loss,
and in correctly identifying and compensating injury victims and respond-
ing to their needs.

Compensating Injuries
The primary goal of the no-fault concept is compensation of a broad
range of injuries.8 In evaluating this facet of the Florida and Virginia
282 Chapter 11

no-fault programs, we need to review the context in which the programs

were enacted. In Virginia, advocates for BIF at the time the program was
enacted were the Medical Society of Virginia, the Virginia Hospital
Association, the Virginia Hospital and Healthcare Association, the Vir-
ginia Society of Obstetricians and Gynecologists, and the Virginia Insur-
ance Reciprocal.9 These stakeholders would not be likely to have had
patient compensation as their primary goal. Understandably, they prin-
cipally represent the private interests of their memberships.
The narrow criteria for eligibility meant that the no-fault assessments
on physicians and hospitals were low. In Virginia, the first payment to
a claimant did not occur until five years after the establishment of BIF.
In BIF’s first fifteen years (1987–2002), only seventy-two claimants
received payment.10 This may seem like a windfall to providers at first
glance, but narrow eligibility criteria were likely to mean that there
would not have been much of a reduction in medical malpractice insur-
ance losses and premiums.
Only “permanently and substantially mentally and physically”
impaired infants weighting at least 2,500 grams at birth are eligible for
compensation under NICA. Hence, the experience in Florida has been
similar to Virginia’s. Until 2003, 161 claimants had been awarded com-
pensation, fewer than twelve paid claims per year.11
The number of paid claims was low relative to the pool of injuries that
might have been compensated.12 The vast majority of no-fault claims
paid in Florida were to families in which the child had been diagnosed
with cerebral palsy, although the statute does not restrict eligibility to
those infants.13 In Florida, about 2 percent of children with cerebral palsy
has been compensated by no-fault; about the same percentage has been
compensated in Virginia. In both states, injuries attributable to “genetic”
or “congenital” abnormalities are excluded. Injuries caused by “maternal
substance abuse” are excluded in Virginia. Among those families sur-
veyed by Sloan et al. who received compensation from NICA, the vast
majority indicated that they were satisfied with most aspects of the
medical no-fault program.14
Three possible reasons have been given for the paucity of claims paid
by NICA.15 First, the incidence of birth-related injuries may have
decreased sharply during the 1990s, which the authors consider unlikely.
 No-Fault for Medical Injuries 283

Second, there may have been a change in claiming behavior after NICA
was introduced. This may have been due to publication of research sug-
gesting that the link between obstetrical mismanagement and cerebral
palsy was more tenuous than previously thought.16 This the authors do
not dismiss as a possibility. The findings could have chilled interest
among both plaintiffs’ attorneys and otherwise eligible claimants.
However, if the research findings had such an influence on claiming, it
would have affected tort much more directly than no-fault claims. If
anything, this factor would make no-fault claims more, not less, numer-
ous. Third, claims for a wide range of birth-related injuries, including
some apparently compensable under NICA, may simply persist in the
tort system. This seems like the most likely explanation for the paucity
of claims paid by NICA. Sloan, Whetten-Goldstein, Entman, et al. report
that one third of all families with severely birth-injured children who
responded to their survey in Florida and filed tort claims had never
applied to NICA.17

Failure to Give Notice and the Persistence of Tort

NICA and BIF administrators had strong incentives to limit caseloads in
order to avoid raising premiums to physicians and hospitals. Neither
physicians nor hospitals had any reason to publicize the programs to
their patients before labor-delivery, since they would not have wanted
to place undue stress on the possibility of medical injury during a pre-
natal visit or to encourage tort claims from those deemed ineligible for
no-fault after labor-delivery. Many patients learned the details of the
program and their physician’s participation only after the injury
occurred.18
Trial lawyers in Florida, facing much lower compensation under NICA
than under tort, seized the opportunity to use this “failure to give notice”
as a way to steer their obstetrical cases back into the tort system. Any
bar to suing that no-fault may have erected was not airtight because of
judicial decisions favorable to the trial bar. Plaintiffs’ attorneys drove a
wedge through the cracks to avoid this limitation on suing. As Studdert,
Fritz, and Brennan conclude, “The restrictiveness of NICA compensation
criteria partly explains the persistence of tort. More precisely, the mutu-
ally reinforcing effect of a contingent approach to jurisdiction and narrow
284 Chapter 11

compensation criteria appears to have provided a ready foothold for the

tort system to assume an ongoing role in the compensation of birth-
related injuries.”19 Studdert et al. suggest that
To maintain exclusivity in the absence of a comprehensive shift to no-fault, it
may be necessary to define a scheme’s boundaries according to compensation
criteria that wholly encompass subgroups of medical injury. For example, the
scheme might be designed to cover a complete class of injury (e.g., obstetrics or
surgical), whether or not every claim for injuries in that class satisfies the com-
pensation criteria. Alternatively, jurisdiction might simply be defined according
to whether the injury in question was suffered at the hands of a participating
provider or within the walls of a participating institution.20

The authors concede, however, that public support for no-fault may
dissolve if the design leaves a large number of injured patients ineligible
for compensation in any forum.
But using broad eligibility criteria to avoid these boundary disputes
can result in a very expensive public program that is not politically
viable. However, narrow criteria for eligibility inevitably lead to disputes
about claimant’s entitlement to benefits when their claims fall at the
boundaries of the definition of eligibility. There are many approaches
available to motivated lawyers for steering claims to tort. Thus, with
narrow eligibility criteria, it is very unlikely that no-fault compensation
can be a practical substitute for tort.
One advantage of BIF’s and NICA’s small size has been the ability to
individualize the management of benefits more closely. The programs
have been careful in managing expenditures, securing favorable prices
from vendors, and questioning the benefits of unconventional therapies.21
A larger, national no-fault program would have to implement more
formal rules and procedures to reach a similar level of performance in
this regard.

Administrative Efficiency
The legislative committee in charge of assessing BIF determined that a
lack of actuarial data projecting lifetime expenses for those covered,
resulted in underestimates of the cost of care.22
Since there are dollar limits on medical malpractice awards in Virginia,
total compensation per infant/child covered under the no-fault pro-
gram may be higher than under tort.23 Under BIF, injury victims receive
 No-Fault for Medical Injuries 285

no payment for lost wages or nonmonetary loss, and payment is re-

duced by an amount equal to any collateral sources, such as health
insurance.
Compared to the tort system, overhead has been quite low in both
states’ no-fault programs.24 Of total BIF disbursements in 2001, admin-
istrative, financial service, and legal costs totaled 9 percent.25 Interview
and program data indicate very conservative fiscal administration of both
the Florida and Virginia programs.26
This high level of administrative efficiency may have come at a certain
cost to injury victims seeking legal assistance with their no-fault program
claims. If attorney compensation is too low, it becomes difficult for
worthy victims to find representation. The Florida and Virginia programs
pay a “reasonable” hourly rate rather than contingency fees. This hourly
rate may not have been sufficient to attract enough lawyers to obtain
no-fault compensation for all potentially eligible victims. There are cir-
cumstances under which a contingency fee system may be economically
efficient.27
During BIF’s initial years, assessments of physician participants were
reduced because the program appeared to be overfunded.28 Virginia’s
State Corporation Commission later found that BIF had an unfunded
liability of $88 million. The fund manager and the program staff had
failed to share financial information which would have allowed for
proper financial oversight.29 The commission decided to conduct future
reviews on an annual basis.30

Assuring Coverage and Reasonable Rates

Initially after implementation of no-fault in both states, mean medical
malpractice insurance premiums paid by obstetrician-gynecologists
decreased; this decrease exceeded the amounts individual physicians paid
in no-fault program assessments.31 The cause of the lower rates, accord-
ing to a legislative committee reviewing BIF, was fewer malpractice
claims against obstetricians and hospitals.32
In 2003, obstetricians still faced very high premiums years after the
implementation of NICA. The Florida Governor’s Select Task Force on
Healthcare Professional Liability Insurance indicated that obstetrician-
gynecologists in areas such as Jacksonville would be operating without
286 Chapter 11

insurance, retiring early, or becoming college faculty members in order

to obtain sovereign immunity.33
In Virginia, the Joint Legislative Audit and Review Commission found
reason for concern in the lower percentage of participation in BIF among
rural obstetricians.34 Barriers to access to care tend to be worse in rural
areas because most physicians locate in urban areas. Since assessments
are not based on the number of deliveries the physician performs, rural
practices may have been paying much more in no-fault assessments per
delivery. Additionally, costs of tort and levels of premiums are not as
much of a problem in rural areas since juries in such areas tend to be
less generous to plaintiffs. A narrowing of the gap between the costs of
tort and no-fault in rural areas creates a strong reason for assessments
to be based on the number of deliveries performed. While BIF does
seem to have had a favorable impact on premiums paid by physicians,
it seems to have had little effect on the availability of obstetricians.35
The experience with no-fault programs in Virginia and Florida serves
as both fascinating and important testing grounds for medical no-fault
in the United States, not because of their conceptual innovations but
rather for what they say about practical implementation of no-fault,
particularly in the U.S. context. The experience is not devastating to the
no-fault concept, but it is rather sobering.
The programs have been successful in providing individualized com-
pensation to families quicker and with less administrative cost than the
tort system. This is no small feat. But NICA did not succeed in averting
a new malpractice crisis in Florida for obstetricians after 2000. Some of
the programs’ weaknesses may be attributed to their hasty formulation
in response to the malpractice crisis of the mid-1980s, but haste is some-
times needed to achieve program implementation.
The experiences of these programs do not support the notion that a
more general medical no-fault program would be less expensive than
tort. Operating on such small scales,36 use of informal procedures, and
a small staff are possible. However, if the programs were expanded to
cover a less narrow set of injuries, administrative cost per accepted case
would surely increase because more formal administrative procedures
would be needed. If they were expanded to cover less seriously injured
children, no-fault program savings from not paying nonmonetary loss
 No-Fault for Medical Injuries 287

would be far lower than for the more serious injuries that these programs
currently cover.
The assertion that no-fault would reduce tort claims frequency for
those injuries covered is not supported by the Florida experience, although
there is some support for this in Virginia.37 One could interpret this as
a positive outcome to the extent that some injuries were compensated
that otherwise would not have been. Still, the result is inconsistent with
an important advantage asserted by proponents of no-fault. The pro-
grams did not become substitutes for tort, in part because the definitions
of covered injuries were so narrow and lawyers had strong financial
incentives to find avenues to steer no-fault claims back to tort. A desire
for retribution may have also kept some cases in the tort system. In
Florida, no-fault claimants were much less likely than tort claimants to
be motivated by a desire for retribution.38
The political context in which the no-fault programs were adopted
was not conducive to adoption of a broad-based program advocated by
leading academic supporters. The focus was overwhelmingly on medical
malpractice cost containment, not on the unmet needs of injury victims.
These programs consequently were developed to rely on a narrow funding
source.39 For an expanded no-fault program to support many more
claims, a much broader funding base would be required, including use
of general tax revenues rather than an exclusive reliance on funds from
physicians and hospitals.
Even if administrative savings from tort were large enough to fund
a broad no-fault program, there is an inherent conflict between
physicians’ and hospitals’ understandable goals of saving money on
premiums and the financial needs of a no-fault program with broad
eligibility standards. Would physicians and hospitals really be willing
to contribute their savings to a program that serves broad societal
purposes? A strong argument can be advanced that it is not their
sole responsibility to do this.

The National Vaccine Injury Compensation Program

The National Childhood Vaccine Injury Act of 1986 became effective in

October 1988. It established the National Vaccine Injury Compensation
288 Chapter 11

Program (VICP) as a no-fault alternative for resolving vaccine injury

claims. VICP was enacted when the nations’s vaccine supply was at risk;
existing manufacturers were threatening to stop selling childhood vac-
cines unless the rapidly increasing liability threat was brought under
control. The VICP is the largest medical no-fault program, and the only
such national program in the United States
VICP’s primary initial goals were to alleviate the vaccine supply prob-
lems resulting from tort claims and to compensate injuries associated
with routinely administered childhood vaccines.40 By most accounts,
VICP has succeeded, but the thimerosal issue looms on the horizon.
The U.S. Department of Health and Human Services (DHHS), the U.S.
Court of Federal Claims, and the U.S. Department of Justice (DOJ)
jointly administer the VICP. All vaccines recommended by the Centers
for Disease Control and Prevention for routine administration to chil-
dren are covered under the program.41 Since 2004, some vaccines admin-
istered to adults have also been covered.
The process begins with an individual claiming injury or death from
a vaccine filing a petition with the U.S. Court of Federal Claims. At that
point a physician at the Division of Vaccine Injury Compensation of the
DHHS reviews the petition and makes a recommendation to the DOJ.
An attorney represents the DHHS at the DOJ hearing before a “special
master,” who makes a decision on whether to compensate and, in a
separate hearing, on the amount of the award. This decision can be
appealed to the Court of Federal Claims and then to the Federal Circuit
Court of Appeals.42
A claimant qualifies for compensation by (1) demonstrating that the
injured person received a vaccine listed on the Vaccine Injury Table and
that the first symptom of the injury/condition on the table occurred
within the period listed in the table; or (2) proving that the vaccine
caused the condition; or (3) by proving the vaccine aggravated a condi-
tion existing before the vaccine was administered.43 The table lists spe-
cific injuries and the time frames in which they must occur for a claimant
to qualify for payment. While the Vaccine Injury Table allows a statutory
“presumption of causation,” and even though the injury is listed in the
table, if the court finds that the injury was unrelated to the vaccine,
compensation is not paid. Claims for table injuries are more likely to be
 No-Fault for Medical Injuries 289

compensated; yet most claims filed are for off-table injuries.44 If a claim
is found to be ineligible for compensation, the claimant may take the
case to tort. However, if the claimant accepts payment from VICP, he
or she is barred from the tort system. As of June 2006, 11,830 claims
had been filed with VICP. Of these, 1,985 were compensated, 4,340 were
dismissed, and the rest were pending.45
The number of lawsuits for just three vaccines (diphtheria, tetanus
toxoids, and pertussis) reached a peak of about 250 in 1986, then fell
dramatically after VICP was implemented.46 After about 2000, however,
concerns about liability as a barrier to vaccine research and development
(R&D) and to entry into the United States market began to reappear, as
did concerns over the thimerasol issue.47
VICP, coupled with other policies, may have increased incentives for
vaccine innovation.48 However, tort liability may still inhibit vaccine
R&D if vaccine manufacturers do not think they would gain coverage
under VICP.49
The issue of attorney fees has been a controversial one for the VICP.50
Attorneys with successful claims earn slightly more, but attorneys’ fees
and expenses are paid regardless of the outcome. An absolute ceiling of
$30,000 is set for legal expenses; the “special master” has the authority
to adjust the lawyer’s billing if it is deemed unreasonable. Contingent
fees are not permitted.
A large number of injury victims have been compensated, although
many have also been rejected, and therefore have been free to obtain
compensation from tort. Unfortunately, there are no data to indicate
how often rejection of a claim by VICP was followed by a lawsuit.
The effect of VICP on injury deterrence and patient safety is difficult
to gauge. Despite the vaccines’ being administered to such large percent-
ages of the population, vaccine-related injuries and deaths are extremely
rare.51 It is difficult to know what safety precautions might exist in the
absence of the no-fault program.
Since attorney compensation and client awards are potentially large in
the tort system and almost always limited in a no-fault program, there
are incentives to find loopholes to steer cases from no-fault to tort.
Ambiguity in the VICP regarding whether or not vaccine preservatives
fall under the VICP has been used by lawyers as a vehicle for bringing
290 Chapter 11

cases involving vaccines containing thimerosal to tort. Thimerosal has

not been explicitly covered under VICP.
Thimerosal is a mercury-containing preservative once widely used in
vaccines, and still used in some vaccines and other pharmaceuticals. Of
the 10,886 cases filed with VICP between 1988 and 2004, 4,335 were
thimerosal/autism-related.52
Thimerosal/autism claims have continued even though the Institute of
Medicine concluded there is no connection between thimerosal and
autism.53 The results of the IOM report were hotly contested by propo-
nents of the thimerosalautism link.
A major victory for plaintiffs came in March 2006 when a federal
appeals court allowed a thimerosal suit against three manufactures to
proceed in federal court (W. Davis 2006).54 The thimerosal issue serves
to highlight one of several border issues that complicate the administra-
tion of no-fault programs. A trade-off exists between ensuring compensa-
tion of all appropriate injuries and reducing the risk of compensating
injuries that are not causally related to the vaccines, and as a result
bankrupting the fund or causing it to be no less expensive than the tort
system.
The VICP created the Table of Injuries to streamline the compensation
process so that certain known side effects and injuries could serve as a
“presumption of causation” rather than needing to be independently
evaluated as to whether each claim of a similar type was caused by the
vaccine. When VICP was established, the U.S. Congress acknowledged
that the Table of Injuries would result in a presumption of causation
favoring the families, and that this would likely result in some children
being compensated who were not actually injured by a vaccine.55
The table represents not only a “line in the sand” regarding what is
to be compensated and what is not, but also the edge of what is known
in the scientific literature. Thus, claims that are off-table represent claims
either entirely illegitimate (not causally related to the vaccines) or outside
the bounds of what the scientific literature can speak to in terms of
causation.
The table was changed four times between 1998 and 2004.56 One
concern is the speed with which legitimate adverse events are recognized
and added to the table. Another issue, which seems likely to remain no
 No-Fault for Medical Injuries 291

matter how often the table is updated, is the difficulty in proving whether
or not the vaccine caused a condition not in the table. Still another is
the addition of adverse events to the list which are later found to have
no connection to vaccines. By 2004, this had occurred for four conditions
once listed on the table.
There are issues of causation in general and in specific individual cases.
Much of the debate on thimerosal has centered on such issues. Often
medical and scientific experts have been asked to testify, yet much of
their testimony has been rejected or given little consideration by the
special masters who decide these cases.57 Weight has been given to Insti-
tute of Medicine reports, broader legal notions of causation, and a
“preponderance of evidence” standard rather than more limiting legal
criteria.58
Similarly, in the context of two birth-injury no-fault programs in
Florida and Virginia, the connection between mode of obstetrical deliv-
ery and the probability of subsequently developing cerebral palsy has
been debated.59 The crux of the issue relating to the feasibility of medical
no-fault is that vaccine no-fault has worked because of lenient rules for
establishing causation. To the extent that these rules become less lenient,
it will become easier to circumvent no-fault and steer cases back to
tort.60
It is certainly daunting to imagine an expanded system implemented
in conjunction with a broader medical no-fault program. Institute of
Medicine reviews are required every two years to evaluate whether
changes to the Vaccine Injury Table are necessary in light of recent sci-
entific literature and adverse events.61 Attempting similar monitoring,
reviews, and updates across the entire spectrum of medical injuries would
represent a formidable challenge. The challenge for proponents of a
broad medical no-fault program is in the details of designing better ways
to collect, verify, and use safety and injury information.

Conclusion
All three no-fault programs demonstrate difficulties in the design deci-
sions and implementation intricacies of how broad or how narrow to
make the eligibility and causation requirements. Issues of which injuries
to include and compensate, and whether and when to compensate
292 Chapter 11

off-table injuries, are difficult and are bound to be a source of conflict.

If the no-fault program is structured to include many injuries, policymak-
ers risk bankrupting the program, especially if issues akin to the thimero-
sal issue arise. If the design is narrow, as in Florida and Virginia, then
it is easy for lawyers to chip away at the program and bring cases back
to tort. How to determine causation and the issue of how to go about
updating the list of injuries to be compensated have been especially dif-
ficult for all three medical no-fault programs. For the obstetrical pro-
grams, the link between oxygen deprivation and cerebral palsy is the
pertinent example.62 In VICP, even though no thimerosal/autism cases
had been compensated as of early 2006, the number of thimerosal claims
filed and discovery documents to be processed was financially
burdensome.

U.S. Experience with Other Types of No-Fault: Workers’

Compensation

Policy discussions of medical no-fault frequently cite workers’ compensa-

tion as a success story that demonstrates the feasibility of a medical no-
fault program in very practical terms. However, beyond broad statements
praising workers’ compensation, these discussions lack in-depth analysis
of how workers’ compensation programs in the United States actually
operate and specific lessons to be learned from these experiences.

Adoption of Workers’ Compensation by the States

Workers’ compensation represents one of the major tort reforms of the
twentieth century in the United States It shifted liability for workplace
accidents from a negligence standard to a form of shared strict liability.63
Under strict liability, payments are made to injury victims without dem-
onstration of negligence, but with evidence that a particular action (use
of a particular product) or inaction (failure to provide an adequate
warning) caused the injury in question.
Adoption of workers’ compensation served the self-interests of various
politically influential stakeholders—employees, to the extent that
expected postaccident benefits rose substantially following the switch
from tort to workers’ compensation;64 unions—the greater presence of
 No-Fault for Medical Injuries 293

unionized manufacturing industries increased speed of adoption;65 private

insurers—when states kept a role for private insurers in workers’ com-
pensation;66 manufacturers, especially large ones, since they had borne
the brunt of tort liability, which was increasing prior to adoption
of workers’ compensation;67 and state bureaucracies—because the size
of bureaucracies increased in response to a need for administrative
oversight of the programs.68

The Scope of Workers’ Compensation Benefits

Workers’ compensation covers both injuries resulting from an accident
occurring in the workplace and chronic diseases thought to have been
caused by having been in the work setting. For many chronic diseases,
there is the task of deciding whether a claim is in or out of the system.
Smith argues that workers’ compensation creates incentives for workers
to report hard-to-diagnose off-the-job injuries as having occurred on the
job.69 Ruser reports that policies enhancing the attractiveness of receiving
workers’ compensation benefits relative to work increased the ratio of
carpel tunnel syndrome cases to cuts and fractures, the latter being com-
paratively easy to diagnose and attribute to specific events.70 A decrease
in the waiting period for benefits increased the number of back sprain
cases relative to those involving fractures.

Probability of Applying for Worker’s Compensation Benefits

A criticism of the no-fault programs for neurologically impaired infants
is that most families who would be eligible for benefits do not apply.
Parallel to this, Biddle et al. estimate that most workers with work-
related illnesses do not file for benefits even though workers’ compensa-
tion has broad eligibility standards for compensation.71 Leigh and
Robbins estimate that in 1999, most of the costs of occupational disease
were not covered by workers’ compensation.72

Experience Rating
Premium setting in workers’ compensation involves three factors. First,
manual rates are established by formulas set by a rating bureau based
on the company’s industry and its occupational mix. Second, for com-
panies of sufficient size, premiums are experience rated. Experience
294 Chapter 11

rating is designed to measure whether or not an employer is better or

worse than the average risk as reflected in the manual rates. Experience
rating is typically based on actual loss patterns of the employer over the
past three years. Those with better-than-average loss experience are
awarded a credit; those with poorer experience than average receive a
debit rating.73 Third, larger companies may be permitted to self-insure
at least a part of loss. This automatically exposes these enterprises to
experience rating.
One advantage of being self-insured or insured with experience rating
relates to the incentive to take precautions. When workers are covered
by workers’ compensation, they tend to be less careful.74 But offsetting
the disincentive that workers’ compensation provides for workers to be
careful, employers may have a greater incentive to promote workplace
safety when covered by workers’ compensation, especially in states with
high compensation benefit levels. Moore and Viscusi conclude that
employers’ incentives for safety monitoring under workers’ compensa-
tion dominate any disincentive workers’ compensation may provide for
employees to be less careful in their workplaces.75

The Record of Rising Costs and Reform of Workers’ Compensation

As problems with the workers’ compensation program have emerged,
states have implemented various fixes. For example, in California, the
workers’ compensation system has been revised and reformed repeatedly
since the passage of the Boynton Act in 1913. Legislators have grappled
with issues of inflation, benefit increases, fraudulent injury claims, and
fluctuations in the economy.
In April 2004 Governor Arnold Schwarzenegger of California signed
the latest reform attempt, SB 899, into law. It remains to be seen how
this latest attempt will fare in a system that has grown increasingly
complex and expensive over the years.76 Schwarzenegger made reform
of workers’ compensation a major issue during his campaign and during
the initial part of his administration, as box 11.2 illustrates.

Implications for Medical No-Fault

Prior to its enactment in the states, various stakeholders supported
implementing workers’ compensation. In contrast, supportive constitu-
 No-Fault for Medical Injuries 295

Box 11.2
Statement by Governor Schwarzenegger Following Enactment of Workers’ Com-
pensation Reform in California, 2004

Governor Arnold Schwarzenegger made the following statement upon

signing workers’ compensation reform legislation (SB 899, Poochigian):
“Today I delivered on my promise to create real workers’ compensation
reform. This bill completes a process that brought together Republicans
and Democrats, business and labor, and all the affected parties to produce
billions of dollars in savings, protect workers, and root out fraud and
waste in the system. No longer will workers’ compensation be the poison
of our economy. Our message to the rest of the country and the world is
that California is open for business. We are making our state once again
a powerful, job-creating machine.” (Office of the Governor for the State
of California 2004)

encies for medical no-fault are lacking. Scholars support medical no-
fault, though they hardly constitute a major voting bloc. Other potential
supporters are health care providers, but their main goal is curbing the
cost of medical malpractice insurance. Patients as a group are not well
organized, as was true of unions in the case of worker’s compensation.
Boundary disputes as to what to cover have been important in workers’
compensation. There is evidence that experience rating is important for
worker safety, although not all evidence points in this direction. Workers’
compensation costs have risen, California being a case in point. The
workers’ compensation experience should serve as a warning that cost
containment may be difficult to achieve under medical no-fault.

U.S. Experience with Other Types of No-Fault: Motor Vehicle

No-Fault

Claims involving motor vehicle accidents are among the most common
tort claims.77 In Australia; Quebec, Saskatchewan, and Manitoba in
Canada; Israel; New Zealand; Sweden; and many states in the United
States, first-party liability has replaced third-party liability insurance for
motor vehicle accidents.78 That is, rather than hold the injurer liable;
each party purchases his or her own insurance, much like standard motor
vehicle collision insurance, which is first-party insurance. The rationale
296 Chapter 11

for motor vehicle no-fault is that determination of negligence is a reason

for the high cost of dispute resolution under tort and delays in compen-
sating injury victims. But eliminating the negligence standard may reduce
the incentive drivers have to be cautious, unless no-fault premiums are
experience rated (as are third-party motor vehicle insurance premiums),
which is a general concern about no-fault insurance.
The effect on deterrence of the change from third-party to first-party
liability depends on (1) the fraction of claims barred from tort and (2)
the sensitivity of first-party insurance premiums to the driver’s precaution
level. Since a primary purpose of no-fault is to eliminate costly inquiries
into fault, it seems unlikely that no-fault premiums would be highly
experience rated. In the United States, some states with no-fault for
motor vehicle accidents do permit premium surcharges based on insured
individuals’ driving records.79
Empirical research on the effects on deterrence of switching from
third-party to first-party coverage has yielded mixed results. In the earli-
est and probably most widely cited study, Landes found that implemen-
tation of no-fault in U.S. states increased rates of motor vehicle fatalities,
but Kochanowski and Young and Lund and Zador reported no effect
on fatalities.80 Sloan, Reilly, and Schenzler’s analyses obtained mixed
results on fatalities and binge drinking.81 Using data from New Zealand,
Brown did not find an increase in the amount of driving (which would
occur because the financial consequences of having an accident were
somewhat reduced by no-fault coverage) or in the accident rate following
implementation of no-fault in 1974.82
More recent research using U.S. data concludes that switching to no-
fault increases motor vehicle fatalities.83 But this is contradicted by
Loughran, who reports no statistically significant relationship between
no-fault and fatal accidents and other measures of driver care.84 Among
research studies conducted in other countries, Gaudry, using data from
Quebec, finds an increase in the number of accidents and accident victims
after implementation of no-fault, a result confirmed by Devlin.85 More
recently, Lemstra and Olszynski, using a case-control methodology, find
that conversion from tort to no-fault coverage resulted in a five-year
reduction in total injury claims per 100,000 residents in Saskatchewan
and a five-year reduction in Manitoba. This was as compared to British
 No-Fault for Medical Injuries 297

Columbia, which retained tort, and Quebec, which retained no-fault,

during the observational period.86
In sum, even after two decades of empirical research on the question
of whether or not converting from a tort to a no-fault system leads to
higher or lower injury rates, this issue is not settled. The implication for
medical no-fault is that a switch to a nonexperience-rated no-fault plan
may well decrease patient safety. While the lesson that premiums should
be experience rated seems straightforward, this is not an easy matter
under a no-fault system in which premiums come from individuals.

Experiences with Medical No-Fault in Other Countries

Overview
No-fault programs in Australia, New Zealand, Canada, and Sweden
differ appreciably from the two or three medical no-fault programs
adopted to date in the United States, especially in their relative breadth
of coverage and distribution of compensation.87 The plans in these coun-
tries do restrict eligibility for coverage to iatrogenic injuries rather than
to acute and chronic medical conditions. However, these no-fault pro-
grams include a wider range of benefits (e.g., such as special education
services) than first-party health and disability insurances typically do.

Sweden
The No Fault Patient Insurance Scheme (NFPI) was implemented in
1975.88 The program provides three separate remedies for injury victims,
each functioning independently. First, individuals may file a complaint
with the Independent Patients’ Advisory Committee; which receives
around 25,000 complaints annually.89 The committee does not have any
binding power, but it facilitates interactions between patients and physi-
cians or nurses. Second, individuals may file a grievance with the Health
and Medical Care Liability Board, asking that medical staff receive a
reprimand. The board receives around 3,000 letters per year. The last
option is financial compensation. The no-fault program, known as
Patient Compensation Insurance (PCI), receives approximately 9,000
claims per year, accounting for just 0.16 percent of health care expendi-
tures. This compares favorably with U.S. medical malpractice insurance
298 Chapter 11

premiums, which amount from 1 to 2 percent of total health care

expenditures.90
The low cost is partly attributable to its limited coverage. Only a small
subset of injuries is eligible for compensation; during its first two decades
of existence, 40 percent of around 100,000 complaints resulted in some
compensation.91 The system is funded by levies on Swedish county coun-
cils that provide medical care.92
To avoid compensating every iatrogenic injury, Sweden has imple-
mented program requirements to narrow the pool of eligibles. The fun-
damental requirement is that an injury must have been avoidable,
although the patient need not identify individuals who failed in their
duty.93 In addition, the injury must be accompanied by at least a ten-day
hospital stay, the use of thirty days of sick time, or death.94 Deductibles
are applied to the recovery amount; also, the claimant must establish a
causal relationship between the injury and the health care services
received, based on a preponderance of probability.95
Subsequent reforms have created additional exclusions to compensa-
bility. For instance, individuals may not recover on grounds of lack of
informed consent or for treatment administered during emergency situ-
ations.96 Wound infections resulting from the patient’s own bacteria are
excluded, substantially reducing the claimant pool.97 During the first
three years of the program, 22 percent of compensated injuries were for
infections that developed during the first three years of the program.98
During the program’s infancy in the mid-1970s, approximately 75
percent of claims received payment; this number fell to 18 percent during
1986–1991.99 More recent estimates place compensation rates at close
to 50 percent of all claims, suggesting a reduction in barriers to
payment.100 But 42 percent are rejected outright.101
If a claimant desires to contest rejection of his or her claim or the
amount of compensation, he or she retains a limited right to appeal.
Appeals are considered by the Patient Injury Board,102 which reviews
the claim’s compensability and also the amount of compensation. The
board’s decision is advisory, but the final outcome coincides with the
panel’s decision in about 90 percent of cases.103 If a claimant is dissatis-
fied with the board’s decision, he or she may continue the appeal process
and take the case to a binding court of arbitration, but on matters of
 No-Fault for Medical Injuries 299

process only; substantive matters may not be reviewed.104 Costs from

appeals to the court of arbitration must be borne by claimants.105 Con-
tingent fees are illegal; attorneys are paid on an hourly basis.106 With the
cost imposed on claimants and a success rate of about 20 percent,
on average the arbitration board reviews only thirty-three cases
annually.107
A major difference between the experience with medical no-fault in
the United States and in Sweden is the amount of physician involvement.
In 1997, Sweden passed a reform requiring every health care provider
to purchase no-fault insurance.108 If a provider has not purchased such
insurance, the patient receives payment nonetheless, as all insurers are
jointly and severally liable for compensating injuries attributable to
uninsured providers.109 Additionally, the physician responsible for treat-
ment is “obliged to inform the patient” if damage has occurred, and to
“assist the patient in applying for compensation.”110 This ethically
imposed duty explains why compensation forms are readily available
in all clinics and hospitals and why physicians are actively involved in
60–80 percent of claims.111 By contrast, in Florida’s NICA program, for
example, there is generally no provider involvement with the claim, and
the program has not been aggressively advertised in order to avoid
NICA’s being deluged with claims.
The administrative overhead of the PCI program is substantially lower
(18 percent) compared to the administrative costs of the U.S. tort system
(50 percent or higher), but comparing these figures directly is misleading.
The PCI does not have the tort costs for investigation, feedback, and
management, and Sweden also has social insurance programs.112
PCI’s low overhead reflects several factors. First, the speed of claims
resolution lowers overhead expenses; from the time of filing a claim to
its final determination is about six months.113 The claims process is done
with little attorney involvement, saving a significant amount in legal
fees.114 In addition, successful claims are paid using a fixed benefits
schedule, eliminating the cost of setting damages on an individualized
basis.
To compare the expense of no-fault against tort in the United States,
Studdert and Brennan use data from Utah and Colorado, assuming
(counterfactually) that the two states had no-fault programs based on
300 Chapter 11

the Swedish model.115 They estimate that Utah would have a $10–15
million increase, and Colorado a $25–35 million increase, in expense.
Although no-fault would be more expensive, far more claimants would
be compensated than under tort. Studdert and Brennan estimate that
administrative cost under a no-fault plan would be appreciably lower
than under tort.
Their estimates of spending under a Swedish-style no-fault plan may
be overly optimistic. For one thing, depending on how the program is
structured, rates of claiming could be far greater than Studdert and
Brennan predict. Further, Danzon has raised serious questions as to
whether the United States, with its very different tort and health care
institutions, could ever implement a program similar to the PCI and
obtain similar results.116 Even though individuals in Sweden have access
to both tort and no-fault, the Swedish tort system is much less favorable
to plaintiffs than in the case in the United States. The burden of proof
is roughly a 75–85 percent threshold probability of negligence, and
expert testimony is difficult to secure. Also, contingent fees are illegal,
and payments under tort in Sweden are subject to full collateral source
offset.117
Given the higher returns to pursuing a tort claim in the United States.
than in Sweden, remaining in tort would often be more attractive to
claimants and their attorneys in the United States. Substituting no-fault
for tort would be a much greater challenge in the United States than in
Sweden.
Another difference relates to level of physician involvement in Sweden
versus that in the United States. At least until it was clear that medical
no-fault would withstand constitutional challenges, it seems unlikely that
physicians in the United States. would readily inform their patients when
an error occurred.

New Zealand
Established in 1972, New Zealand’s original no-fault program was broad
and inclusive. Justice Owen Woodhouse, who spearheaded the develop-
ment of the system, held that the basic principle of no-fault should be
collective responsibility, acting as a form of social insurance.118 Since its
inception in 1972, the Accident Compensation Corporation (ACC) has
 No-Fault for Medical Injuries 301

administered New Zealand’s no-fault program.119 Initially, all accidental

personal injuries were covered, including medical malpractice, which was
labeled “medical misadventure.” However, total expenditures from the
program rose rapidly.
Originally, funding came from levies on employers, motor vehicle
owners, and subsidies from the government, with funds from each source
placed in a separate fund. The funding changed dramatically over the
course of three decades and several reforms; a pay-as-you-go financing
structure was established, a new levy for registered health professionals
was created, and seven separate funds administered by the ACC were
established. Also, the government retained the power to require risk-
rated premiums for health professionals.120 Administrative cost, however,
was only 10 percent of total expense.121 Even so, the overall costs of
New Zealand’s no-fault program have proven to be burdensome. Cost
per claim has risen considerably.122
In response to rising costs, there was a major reform in 1992 which
substantially restricted the scope of covered injuries, shortened the time
within which claims could be brought, and eliminated lump sum pay-
ments for pain and suffering.123 As in Sweden, the New Zealand program
now required a fourteen-day hospital stay or twenty-eight sick days as
a requirement for eligibility.124 The newly restricted definition of medical
misadventure introduced an element of fault, limiting claims to injuries
resulting from medical error or mishap, and thus removing the problem
of having to distinguish between injuries resulting from medical care and
unavoidable or inevitable injuries.125 The introduction of fault in the
1990s was not a surprise; courts had used fault in their analyses of
medical misadventure throughout the 1980s, and a substantial body of
case law had developed.126
In 2002, the no-fault system was reformed again with the Injury, Pre-
vention, Rehabilitation, and Compensation Act of 2001. With this act,
the ACC was restored as the sole administrator of accident compensation
schemes, and efforts were refocused to emphasize restoration.127 The
most recent reform occurred in 2005, when New Zealand modified its
no-fault program in order to establish firm boundaries and definitions
for eligibility. The new system eliminated the element of fault from the
eligibility criteria, returning New Zealand to a true no-fault system.128
302 Chapter 11

In addition to monetary compensation, claimants in New Zealand

have the option of pursuing nonmonetary remedies. In 1994, the New
Zealand Parliament created legislation establishing the Health and Dis-
ability Commission (HDC).129 The HDC is also responsible for handling
disciplinary complaints.130 However, the most common form of nonmon-
etary relief sought from the HDC is not disciplinary action; rather, it is
a request for corrective measures to address the cause of harm.131
Error reporting in New Zealand is a regular part of the medical
culture; acknowledgment of injuries in patient records is extremely
high.132 Truthful, consistent error reporting provides opportunities to
evaluate quality of care problems, a sharp contrast with the United
States, where such reporting is rare. Also, in contrast to tort in the United
States and like Sweden, compensation is delivered quickly; the ACC is
required to provide notice of its decision within nine months.133 Patients
are offered different forms of accountability—much like Sweden’s system.
Since many claimants desire only an explanation, accountability, apology,
assurance of a system change, or intervention, the HDC allows patients
to initiate patient safety measures through their complaints.
Despite these positive results, a recent survey of physicians in New
Zealand demonstrated that many physicians do not believe the ACC and
the HDC are achieving their stated goals.134 Nearly 40 percent did not
think most complaints were warranted, and another 33 percent did not
believe complainants were normal people.135 In addition, the complaint
and disciplinary system is nicknamed “death by 1000 arrows” by some
(Figure 11.1) due to its complexity.136 Some view this complexity as
standing in the way of New Zealand’s no-fault programs’ accessibility,
efficiency, and effectiveness. Also, New Zealand’s no-fault system may
lack incentives to improve patient safety, given its broad funding base
and lack of experience rating.

Implications for Medical No-Fault in the United States

The United States is far more litigious than the other countries that have
implemented no-fault programs.137 It seems likely that a substantial
amount of litigation would develop in the United States at the boundaries
of no-fault coverage, with lawyers for claimants arguing that injuries
were not covered by the no-fault program, thus making the claimants
 No-Fault for Medical Injuries 303

Inquest Complaints Assessment Committee

Informal
complaint Privacy
Patient
Commissioner

Private Ability
Medical council
proceeding Up to $200 000
Accident
Damages
Compensation Conciliation
Committee Competency
Exemplary Publicity Investigation review
damages
Health and Medical Practitioners
Disability No action
Medical misadventure Disciplinary Tribunal
Commissioner No action
Costs
No Yes
Investigation Erasure
Advocate

No further action Fine

Vindication Apology Suspension up to
Medical mishap Medical error $20 000

Conditions
Director
No action
Compensation of
Compensation Proceedings Human
and Rights Review
refer to H&DC Tribunal
Declaration of breach

Complaints and disciplinary Up to $200 000

procedures in New Zealand Restraining order to remedy breach damages

Figure 11.1
Death by 1000 Arrows: The Multiple Pathways of the Current Complaint System
in New Zealand Source: Cunningham (2004).

eligible for payment under tort. Also, Sweden and New Zealand have
much different health systems from that in the United States. Health care
services are publicly financed and lower cost in the other countries than
in the United States, in which half of such services are privately financed.
These factors tend to make no-fault in the other countries more
affordable.
Consistent error reporting and physician involvement are essential to
the success of a no-fault program. Even though some advocates for a
medical no-fault program in the United States assert that under no-fault,
health care providers would willingly admit their errors, this seems
highly questionable in the U.S. context. In New Zealand, by contrast,
the culture seems more conducive to error reporting. Finally, while the
United States has had a tumultuous relationship with its tort system, the
concept of civil action is deeply ingrained in U.S. history and culture.
304 Chapter 11

Many citizens believe that civil actions for negligence are a fundamental
right that should not be withdrawn.138 While supporters of the medical
no-fault concept in the United States often cite positive experiences of
no-fault in other countries as evidence that such programs work, others
are much more doubtful. Danzon argues that rather than providing
a prototype for other countries to adopt, the original New Zealand
no-fault program demonstrates pitfalls to avoid.139 We agree with her
assessment.
Finally, a continuing theme is the lower administrative cost and speed-
ier payment under no-fault. This positive attribute is sufficiently univer-
sal to be accepted as an important benefit of no-fault compensation
programs.

Is There a Role for No-Fault in the U.S. Context?

This chapter has described the pros and cons of no-fault programs. No-
fault programs have some important advantages. Injury victims are com-
pensated who would otherwise not be compensated, and compensation
is made at a much lower administrative cost and much more quickly, on
average, than by tort.
In the United States, we have never had a real national discussion
about the merits of providing coverage for persons who are injured
during the course of receiving or, in some cases, not receiving medical
care as opposed to other groups, such as those who do not have any
health insurance coverage. If a serious public discussion were to occur,
it seems unlikely that covering those persons with iatrogenic injuries
would receive top priority.
Instead of a broad medical no-fault program, advocates of medical
no-fault in the public arena (as opposed to discussion among academic
experts) have had another focus: to reduce payments to medical mal-
practice claimants and associated legal expenses. This priority has led to
adoption of a very few programs with narrowly defined eligibility crite-
ria, which although generally efficient in terms of administrative cost and
speed of payment, serve the compensation objective in only a limited
way. Further, although there are savings in time and expense of litiga-
tion, other problems have emerged. For one, although negligence is not
 No-Fault for Medical Injuries 305

an issue with no-fault, causation issues are. The most striking example
is over causation in the thimerosal/autism debate. Having a no-fault
program for vaccines has not stopped vaccine manufacturers from assert-
ing that they do not invest in vaccine research and development to the
extent that they otherwise would because of the looming threat of tort.
The birth-injury no-fault programs in Florida and Virginia have relied
on premium income from hospitals and physicians. This narrow funding
base, obtained from virtually the only strong advocates of medical no-
fault in these states, has inevitably led to pressure on program adminis-
trators to keep the programs small. If redistribution of income from
“someone” to families with birth-injured children is indeed an important
social objective, it seems odd that the “someone” should be hospitals
and physicians. General tax revenue would seem to be a more logical
and just alternative.
The practical reason that hospitals and physicians are taxed is
simply that these no-fault programs were adopted to relieve hospitals
and physicians from lawsuits and high premiums. Thus, it was felt that
these groups should be willing to be taxed. However, these no-fault
programs have not consistently reduced the cost of medical malpractice
to providers and manufacturers. Certainly, if the goal is lawsuit and
premium relief, there are more effective approaches, including adoption
of flat caps on nonmonetary or total loss, a solution we argue is
inequitable.
The notion that a broad no-fault program financed from general
tax revenue, such as by the personal income tax, would be adopted in
the United States seems to us to be pure fantasy. The debate over uni-
versal health insurance has occurred over many decades without progress
being made. More limited public programs that cover certain demo-
graphic and income groups appear to have higher priority; Medicaid and
SCHIP programs have been adopted. The political context of the other
countries with medical no-fault programs is different from that of the
United States; one indicator is that they have universal health
insurance.
If this type of no-fault program is unrealistic, are there alternatives?
One possibility is a mandatory no-fault program financed by an excise
tax on payments to hospitals and physicians for the services they provide.
306 Chapter 11

Presumably at least some portion of this tax would be shifted forward

to health insurance premium payers. Essentially this would be a tied sale
with hospital and physicians’ services bundled together. This seems close
to an unfunded mandate which would require health insurers to provide
no-fault insurance with health insurance coverage. If history is any guide,
health insurers would almost certainly oppose this unfunded mandate.
However, a program financed by an excise tax/unfunded mandate has a
much higher probability of adoption than does one funded from general
tax revenue.
Still another possibility is a voluntary no-fault program based at
hospitals. Both hospitals and their medical staffs would decide whether
to offer such coverage, and patients would decide whether or not to
accept the offer of no-fault coverage at the hospital or remain in tort. Such
a program would include specific injuries resulting from care within the
walls of the hospital, whether inpatient or ambulatory care. The hospital
would be the experience rating unit. Thus, hospitals with effective patient
safety programs would be able to offer such coverage at a lower cost.
Having a no-fault program could be a feature that would make the hos-
pital attractive to consumers. Details of how the program is financed
would be left to the hospital. The state Department of Insurance would
oversee the program as it currently does for other forms of insurance.
Disputes, including those about injury causation, would be resolved by a
mechanism specified in an agreement signed by the patient.
There are admittedly complex issues to be resolved even for such a
voluntary program. One of the major issues is the method of enrolling
patients into the program. Signing a document giving up tort rights in
return for no-fault coverage at the time care is received is likely not to
be viewed favorably by courts. But it should be possible to enroll patients
in advance. One occasion would be at the time employees enroll in a
health insurance plan with their employers. Hospitals in the area could
present documents describing their no-fault insurance programs, includ-
ing any payments required of enrollees and methods of resolving dis-
putes. Also, when hospitalized, patients could sign up for the program
for future stays.
Such a program remains voluntary for both providers and patients.
Trying to develop universally applicable rules as to which injuries are
 No-Fault for Medical Injuries 307

avoidable and which are not, updating these rules with changes in tech-
nology and in knowledge as to whether the injuries are avoidable or not,
as well as resolving thorny issues of causation, is best left on a decentral-
ized basis. Government should facilitate production and dissemination
of relevant information and provide general program oversight, but not
get involved in detailed rulemaking, as is inevitable under a mandatory
program.
12
Reforms: What Can Be Done

The empirical evidence and discussion of relevant health care institutions

in this book thus far has provided a foundation for proposals for achiev-
able reforms to be discussed in this chapter. Proposals for reform start
with a premise that something is broken and requires fixing. Medical
malpractice in the United States is broken, but generally not for the
reasons stated by advocates of tort reform.
One constant has been the rhetoric of tort reform. The preface to a
January 1975 symposium issue on medical malpractice began with this
statement:
The term “medical malpractice” has been an increasingly frightening one to
patients, doctors, and insurers as well. In recent months, the spectre of physician
strikes, astronomical damage awards, soaring liability insurance premiums,
and allegations of poor-quality medical care have stirred debate in state legisla-
tures, in the Congress, in the press, and in scholarly journals. The medical
malpractice crisis is real, and the problems which created the crisis remain
with us.”1

Except perhaps for the “spectre of physician strikes,” using some

white-out for the date, this statement could have been published in
January 2005. The many tort reforms since the 1970s have not altered
the tenor of the public policy discussion. Caps on damages have “worked”
to reduce payments by medical malpractice insurers and create premiums
below what they otherwise would have been (Chapters 4 and 5),2 but
caps have not altered the incentives of the participants in this market
(Chapter 1). Furthermore, if there is a benefit to caps, it is mainly in
redistributing income from injury victims and their attorneys to health
care providers rather than in improving efficiency in allocation of
resources. If there were overclaiming, one could say that by reducing
310 Chapter 12

access to legal representation, caps are efficiency enhancing. But, if any-

thing, the empirical evidence supports underclaiming and undercompen-
sation on average (chapters 1 and 5).
It seems unlikely that any savings in medical malpractice insurance
premiums would accrue to patients as taxpayers and health insurance
premium payers (chapter 3). Organized medicine plausibly supports caps
first and foremost primarily in response to pressures from its
constituency for financial relief.
The title of this chapter is “achievable reforms.” Starting with the crisis
of the mid-1970s, as the above quotation indicates, scholars, other
experts, and some policy analysts have proposed broad reforms of
the current system. Broad examples include proposals to substitute
no-fault for medical injuries for tort (chapter 11), abolition of the jury
for medical malpractice disputes as part of a plan for special health
courts (chapter 7), substitution of alternative dispute resolution (ADR)
mechanisms for traditional methods of dispute resolution used in tort,
enterprise liability, using insurers or hospitals as the enterprise, substitu-
tion of private contracts for tort,3 limits on plaintiff attorneys’ fees
(chapter 6), and scheduled damages, especially for nonmonetary loss
(chapter 5).
Many, and probably most, of these reforms are not achievable. As
Baker contends, “No-fault compensation for medical injuries turns out
to be an idea that lots of people like in theory, but almost no one likes
in practice.”4 Abolition of the jury for medical malpractice seems war-
ranted on grounds that there are cognitive limitations of individuals in
general and lack of experience in their role as jurors in particular. But
juries provide a potential check on abuse of power by medical providers,
and hence eliminating juries in medical malpractice cases will under-
standably encounter some opposition. By lowering litigation cost to
plaintiffs, statutes enacted to encourage alternative dispute resolution
(ADR), which involves mediation and/or arbitration, appears to have
increased medical malpractice payments.5
Private contracts seem attractive in principle, but the concern is that
patients are not well-positioned to be effective negotiating partners in
the contracting process. Although, as discussed below, enterprise liability
has attractive features, it has been disregarded, and even opposed, by
 Reforms: What Can Be Done 311

organized medicine.6 Similarly, while scheduled damages have attractive

features and support from the academic community,7 they are vehe-
mently opposed by the trial bar. Limits on plaintiff attorneys’ fees in
medical malpractice cases enjoy some political support from organized
medicine. These proposals focus on high incomes of plaintiff attorneys
and not on attorneys for the defense, physicians with high incomes, and
CEOs.
Ironically, but understandably from the vantage point of promoting
one’s self-interest, the same medical organizations oppose limits on
physicians’ fees. And while medical organizations are major opponents
of the status quo in medical malpractice, several members of this
community have supported imposing tort liability on managed care
organizations.
Indeed, a problem common to many of these mainly “second-
generation” reform proposals (chapter 4) is that they lack a strong politi-
cal constituency. At the state level, there are strong political constituencies
for damage caps. The U.S. Congress has not even enacted traditional
“first-generation” (not to speak of second-generation) reforms.8 Trial
lawyers and consumer advocate groups have more political clout in the
U.S. Congress than in many state legislatures. Whatever deficiencies the
second-generation proposals may have, the real reason for no action on
fundamental reform is that it lacks a sufficiently influential political
constituency.
The prudent course involves identification of the major deficiencies of
the existing system and incremental reform. Admittedly, incremental
reform is far from anyone’s conception of the ideal, but the real choice
is between no change and incremental change.

What’s Good and What’s Bad with the Existing System?

Positive Aspects
The current system has a few positive attributes. First, being able to sue
in combination with the contingent fee method for compensating plain-
tiffs’ attorneys gives patients who are unsatisfied with outcomes a mecha-
nism for addressing their grievances that may not be possible through
other channels. The regulatory apparatus is sometimes controlled or
312 Chapter 12

substantially influenced by health care providers or for bureaucratic

reasons may be unresponsive to patients’ complaints. As Mechanic
argued over three decades ago, “One of the remaining sources of
power for the patient is his ability to threaten or initiate malpractice
litigation when he feels that his interests have been abused. Although
patients rarely do so, the remote threat may to some extent control the
bounds of physician behavior.”9 Second, the American jury, in spite of
possible limitations, gives ordinary citizens a role in this dispute resolu-
tion system. Regulatory agencies may not be equally sensitive to con-
sumer interests. Third, if one views secular trends in medical malpractice
payments and premiums, as opposed to very large increases in premiums
at the onset of “hard markets,” the increases in payments and premiums
are rather modest, only slightly higher than the changes in prices in
general.

Negative Aspects
The current system has serious deficiencies, although they are not the
same as those typically depicted in the popular, trade, and medical pro-
fessional press and in testimony before state legislatures and the U.S.
Congress by various stakeholders. First, unlike other fields of personal
injury tort, there is no empirical evidence that the threat of medical
malpractice lawsuits deters injuries (chapter 3). Particularly since injury
deterrence is typically listed as the first goal of tort liability, this is a very
serious deficiency. Nearly complete insurance coverage for medical mal-
practice liability, the lack of experience rating, and the low enforcement
rate (low claims frequency relative to negligent adverse events) account
for part of the failure of tort to deter. Tort liability focuses on mistakes
of individual providers, but errors frequently reflect simultaneous omis-
sions or misjudgments on the part of several individuals. Backup systems
for correcting errors when they occur are often lacking.
When asked, almost all physicians in the United States maintain that
they practice defensively on account of the threat of being sued. To the
extent that this is so—and there is limited empirical evidence to support
the view that there is some defensive medicine—it would seem not that
physician decisions are affected by tort, but that the signal from tort is
insufficiently precise or even wrong.10
 Reforms: What Can Be Done 313

Several studies have attempted to measure the cost of defensive medi-

cine (chapter 3). To the extent that it can be accurately measured, the
effect of the threat of lawsuits on cost appears to be modest, especially
when placed in context of the appreciable increase in real expenditures
on personal health care that has occurred in the United States since the
mid-1960s, when Medicare and Medicaid programs were first imple-
mented. Defensive medicine is at best a second-tier driver of increased
real expenditures.
Perhaps just as serious, or even more serious, is the health care pro-
vider community in the United States’ almost total rejection of the view
that medical malpractice has a constructive role to play in health care
delivery. Providers generally see no link between medical malpractice
litigation and provision of high-quality care. Some public policy discus-
sions of iatrogenic errors, such as those by the Institute of Medicine
(chapter 8), have explicitly excluded considerations of medical malprac-
tice. Including medical malpractice in public discussions of patient safety
has been seen as muddling the analysis and, more important, risks pro-
voking opposition from the provider community, thus impeding intro-
duction of both high-tech and low-tech approaches to improving patient
safety.
Much commentary in both assessments of medical malpractice and
patient safety see medical malpractice as part of the problem rather than
part of the solution. The problem is said to be that the threat of medical
malpractice litigation leads to excessive secrecy about specific medical
errors, both out of fear that discussion of medical errors will lead to
more lawsuits and that the discussion could be introduced by plaintiffs
at trial as evidence of defendant liability. These concerns probably have
some validity.
Yet it seems unlikely that insulating providers from threats of liability
will lead to widespread disclosure of errors. Nor can the threat of law-
suits be the only reason that health care organizations do not adopt
patient safety measures, such as systems to prevent errors in drug delivery
in hospitals. Even under no-fault, disclosing errors would lead to an
increased number of compensated cases. As long as health care providers
are asked to fund medical no-fault insurance programs, one cannot
expect providers to be eager to disclose their mistakes. Even absent a
314 Chapter 12

financial disincentive to disclose, it is not human nature to want to be

known for one’s “bloopers,” even if the bloopers are disclosed only to
colleagues and coworkers, and not to the public. Public disclosure only
increases embarrassment.
At present, the business case for investments in patient safety leaves
much to be desired. Health care product markets do not demand high
levels of patient safety, or at most are only beginning to do so. And if
providers were to incur the additional cost of error-reducing innovations,
they would receive little financial incentive in terms of reduced medical
malpractice insurance premiums. In sum, if quality of care falls short of
socially optimal levels, there is little financial incentive to raise it to
optimal levels.
Litigation is an extremely inefficient system for compensating injury
victims. Various forms of first-party insurance are much more efficient
in distributing compensation to persons who have incurred a loss due to
personal illness or injury.
Thus, in sum, on most lists of the goals of tort, medical malpractice
does badly on the first two objectives—injury deterrence or improved
patient safety, and compensation of persons with iatrogenic injuries. Its
strongest, and most important, features lie in the category of giving
injury victims a day in court and making injurers accountable to ordinary
citizens who serve as jurors.

Creating Financial Incentives for Preventing Injuries

The highest priority is to focus on providing financial incentives for

preventing injuries, not on preventing all injuries but rather those injuries
for which the cost of prevention does not exceed cost. The literature on
prevention of medical injuries, admittedly not as rigorous as it might be,
suggests that many opportunities for injury prevention exist.11 Further-
more, the evidence is for average benefits and costs, and there is likely
to be substantial variation about these averages. Even if private incen-
tives were aligned to encourage adoption of injury-prevention innova-
tions that are often socially desirable, there will be situations in which
adoption is not appropriate. For example, it may not be appropriate for
all hospitals (e.g., for small hospitals) to adopt computerized physician
 Reforms: What Can Be Done 315

order entry systems. The case for adoption should be made on a project-
specific basis.

Enterprise Insurance for Hospitals and Their Medical Staffs: A

Promising Achievable Reform

Historical Context
Before the 1970s, medical malpractice insurance was provided by com-
mercial property-casualty insurance companies. In the mid-1970s, in
response to a lack of availability of medical malpractice insurers, states
enacted statutes permitting physician-sponsored, single-line medical
malpractice insurers.12 This solved a short-run problem of availability of
coverage, at least until the third crisis.
Both the commercial and the physician-sponsored companies have not
been as active as they should have been in loss prevention and certainly
in injury prevention. Also, insurers may be inherently inefficient in injury
and loss prevention; they are not located where services are being deliv-
ered and are not part of the delivery process.13
Also, given the lack of experienced-rated premiums, physicians also
have not had a meaningful financial incentive to allocate resources to
injury prevention. The lack of experience rating undermines tort’s objec-
tive of corrective justice since individual defendants do not suffer the
financial consequences of any injustice they may have committed.14

Definition and Rationale

Baker coined the term “enterprise insurance.”15 Physicians who work or
render services to patients in hospitals or other health care organizations
would obtain their medical malpractice insurance through such an orga-
nization. He argues that enterprise insurance avoids some of the short-
comings of enterprise liability from which the term “enterprise insurance”
was derived.
Hospitals and health plans have not welcomed the concept of enter-
prise liability; to these organizations, assuming greater liability seems to
be an added burden. Nor has organized medicine been enthusiastic about
ceding such liability to these health care organizations, plausibly because
organized medicine and its membership worry about loss of professional
316 Chapter 12

control over medical decisions. An additional unstated reason is a concern

about the loss of financial independence.
In the variant of enterprise liability that seems most sensible, medical
staff would be insured for medical malpractice by the hospital for all
care delivered within the walls of the hospital. Physicians would gener-
ally continue to purchase medical malpractice insurance to cover claims
arising from care delivered in their offices.16
There is a strong case for the enterprise to be the hospital. Most of
the medical malpractice losses arise from care delivered in hospitals.17
Large hospitals, including those affiliated with academic health centers,
provide medical malpractice insurance now.18 An example is the Duke
University Health System, which provides such insurance to hospitals in
its system and to physician faculty members (box 12.1). If Duke were to
extend coverage to other physicians who practice at its community hos-
pitals, it would be necessary to change its insurance company’s charter.
Hospitals could form their own insurance companies for the purpose of
providing liability coverage for services provided within their walls, or
they could partner with an established insurance company as a joint
venture.
Being part of an insuring group has its advantages and disadvantages.
The efficiencies inherent in combining patient safety measures and insur-
ance, including premium-setting, are definite pluses. There is a potential
to save on premium dollars. Further, since the insurer is much better able
to “poke inside” the clinical organization, it may be less likely to raise
premiums dramatically, given a change in loss experience, since it has a
better sense of what is going on. The minus side, however, is that physi-
cians often prefer to maintain their independence. Even though partici-
pating in an enterprise insurance plan need not result in loss of
independence, some physicians may regard this as a slippery slope.
A survey of hospitals conducted by the Hospital Association of Penn-
sylvania found that as of 2001–2002, 41 percent of responding hospitals
had noncommercial primary layer insurance, presumably including self-
insurance, insurance through a captive or risk retention group, leaving
only 38 percent commercially insured. Among teaching hospitals and
relatively large hospitals in the state, the leading form of self-insurance
was the single-parent captive insurer.
 Reforms: What Can Be Done 317

Box 12.1
Enterprise Insurance at an Academic Health Center: Duke University Health
System

The Duke University Health System (DUHS), headquartered in Durham,

North Carolina, consists of three hospitals, the Duke University Hospital,
and two community hospitals, as well as clinics throughout central North
Carolina and southern Virginia, ambulatory care centers, wellness centers,
home care specialists, and a hospice. As of 2006, DUHS had nearly 17,000
full-time employees. Its professional liability program is administered by
a captive insurance company, Durham Casualty Company, Ltd., domiciled
in Bermuda. The program covers DUHS hospitals and faculty physicians,
house staff, and other physicians who are members of the Private Diag-
nostic Clinic (PDC). The PDC is not part of the DUHS, but for the purpose
of provision of medical malpractice insurance, all PDC-affiliated physi-
cians are covered. Duke University Hospital has a closed medical staff,
with staff membership limited to Duke faculty. Almost 1,000 physicians
who are not faculty members have admitting privileges at one of the DUHS
community hospitals. These physicians are not covered by the plan. Duke
self-insures but purchases reinsurance coverage for its larger losses.
The plan offers several advantages. First, even though there is no experi-
ence rating of individual physicians, the cost of medical malpractice insur-
ance, as determined on an actuarial basis, is deducted from physician
billings. During periods of relatively adverse loss experience, such deduc-
tions are great. Physicians experience variations in deductions. Variations
in loss attributable to physicians are not totally borne by hospitals or
DUHS more generally. Second, DUHS has an active patient safety program.
There are efficiencies in combining incident identification and prevention,
claims management and settlement authority, physician credentialing with
indemnification of liability and associated financial incentives which can-
not be realized when liability insurance is purchased from a nonaffiliated
organization. Third, potentially lower premiums and lower volatility
should be an advantage. Fourth, while many physicians desire to own their
own facilities, enterprise insurance could provide a mechanism for attract-
ing physicians to hospitals.
318 Chapter 12

Also, hospitals and other large health organizations, such as large

medical group practices, are in a better position to prevent and manage
their losses than are individual physicians and small medical groups.
Simulations performed by Sloan and Hassan (1990) provide empirical
support for the notion that large and even medium-sized hospitals have
medical staffs of sufficient size to support this risk-bearing role.19 That
study assumed no injury and loss prevention program would be imple-
mented. With such programs, the case for enterprise insurance is much
stronger.
Two other factors also support enterprise insurance and having hos-
pitals as the enterprise. With the hospital as the insuring enterprise,
hospital medical staff would have added incentives to be selective about
the quality of physicians they admit to and retain on their medical staffs.
And medical staffs have a much more direct incentive to support adop-
tion of patient safety measures in order to reduce medical malpractice
losses at the hospital, especially if medical staff are placed at some risk
for losses above a threshold value. It would be up to individual hospitals
to negotiate cost-sharing incentives with their medical staffs.
An issue is whether or not regulatory oversight would be needed, and
if so, the form it should take. Insurance regulation is almost entirely a
state responsibility. U.S. law delegates the regulation of the business of
insurance, and every person engaged therein, to the laws of the several
states. Congress is forbidden from enacting legislation that will invali-
date, impair, or supersede any law of a state relating to insurance.20 This
has resulted in heterogeneity among states in insurance law and regula-
tion. However, there is federal law partially superseding this for certain
forms of insurance, preventing state legislatures from regulating risk
retention groups unless the group has been chartered in that state.21

Possible Objections to Enterprise Insurance

Baker lists several possible objections to enterprise insurance.22 Such
insurance may
interfere with the health care market, create a windfall for physicians with hos-
pital practices, force hospitals to bear too much of the cost of medical liability,
reduce the autonomy of physicians and other practitioners, lead to an increase
in medical malpractice claiming, and present insurmountable administrative
complications.
 Reforms: What Can Be Done 319

That such insurance would interfere with the health care market and
present insurmountable administrative complications could be said for
virtually any change. Existing insurers would not welcome the change
because enterprise insurance will probably mean that these insurers
will lose business. A rebuttal to the “insurmountable administrative
complications” is that such insurance has already existed for years, albeit
on a limited basis. Shares of premium expense borne by the hospitals
and individual physician members would depend on market factors and
would vary among hospitals with enterprise insurance.
Autonomy of physicians and other practitioners often conflicts with
the goal of improving patient safety, which requires a team approach.
Physicians must be actively involved in implementing measures to reduce
medical errors, particularly since they play a lead role in the provision
of care in hospitals. But some loss of individual physician autonomy is
inevitable if meaningful changes in patient safety are to be realized.
Major improvements require that physicians and other hospital per-
sonnel function as a team.
If enterprise insurance substantially improves loss and injury preven-
tion, it is not inevitable that there would be an increase in medical mal-
practice claims. There is substantial underclaiming now. In contrast to
enterprise liability, individual physicians would be named in medical
malpractice suits as they are now. It is not obvious that a hospital would
be viewed by jurors as having deep pockets any more than a large insurer
would.
The most serious objection is a more general one and not in Baker’s
list. If the concept of enterprise insurance is so attractive, why is it not
more common now?
There are several answers. First, in the United States, in contrast to
other high-income countries, hospital medical staffs have been largely
independent of hospitals.23 Although some physicians are employed by
hospitals, the vast majority practice in hospitals with the only formal
relationships being membership on the hospital’s medical staff and on
some committees. Physicians have resisted being under the control of
hospitals, both for financial reasons and out of concern for possible loss
of professional autonomy. As technology has become more sophisti-
cated, this independence is becoming more and more realistic.24 Since
320 Chapter 12

about 1995, physicians have increasingly joined physician-hospital orga-

nizations for purposes of negotiating contracts with health insurers,
ostensibly because such joint negotiations raise physicians’ market power.
Yet any proposal from the outside that would cede control of medical
decision-making to hospitals is likely to be resisted by many physicians.
The key will be to have active physician involvement in hospital-based
enterprise insurance with a meaningful financial incentive for individual
physicians to prevent lawsuits.
Second, there are substantial differences among hospitals in their
sophistication about risk management and medical malpractice. In part,
this is a function of hospital and medical staff size. Indeed, smaller hos-
pitals may be of an insufficient size to operate a medical malpractice
insurance plan on their own.25 Such hospitals might join regional com-
pacts, which in turn form a risk retention group.
Third, in many years, premiums charged by medical malpractice insur-
ers are below fair value. In such cases, hospitals and others have a finan-
cial incentive to purchase insurance coverage rather than to self-insure.
Finally, accountability incentives are not likely to provide sufficient
motivation for hospitals to create systems management of medical inju-
ries.26 Hospitals and physicians have many nonliability objectives and
concerns. Implementation of enterprise insurance alone may not lead to
optimal levels of patient safety in hospitals. Yet the converse does not
follow. Just because enterprise insurance is not likely to be the silver
bullet does not imply that it is a BB gun (i.e., it cannot have an important
role in promoting patient safety).

Captive Insurers and Risk Retention Groups

There is a major trend in the United States to use of captives, such as
the Durham Casualty Company (see box 12.1).27 Captives provide an
organization for performing traditional functions of insurers, such as
computing loss reserves and investing premium income from the time it
is received until loss payments are made. There are also tax advantages
to captives, as well as regulatory advantages if the captives are domiciled
offshore.28
A captive insurance company is an insurance company which is owned
by the entity it insures.29 Captive insurance involves the insured creating
 Reforms: What Can Be Done 321

a subsidiary or sister corporation (the captive), which acts as a funding

vehicle for the insured, assuming some or all of the owner’s financial
liability for medical malpractice.30 Typically captives are incorporated
offshore, and as a result they face fewer regulations, such as lower
capitalization requirements and, as mentioned before, tax advantages.31
As of 1995, 70 percent of captives worldwide were formed offshore.
There has been a recent trend to decrease regulations for formation of
onshore captives, but state laws and regulations remain demanding,
much more so than the laws of offshore domiciles. Regardless of where
they were formed, most captives are subject to state insurance regulation,
but for some tax-exempt organizations, insurance code regulatory viola-
tions may be avoided.32
Risk retention groups are an option for small hospitals, physicians not
affiliated with large medical groups, hospitals offering enterprise insur-
ance, or hospitals that decide that forming their own insurance program
is inadvisable. The Risk Retention Act was initially enacted as the Product
Liability Risk Retention Act of 1981.33 In 1986, during the liability crisis
of the mid-1980s, the law was amended to apply to all other liability
coverage, including medical malpractice.34 The act, as amended, autho-
rizes “risk retention groups” to pool similar risks for self-insurance and
risk “purchasing groups” to be formed to purchase coverage on a group
basis. These organizations must be chartered in only one state. In other
states, such entities are freed of most insurance regulation and do not
receive the protection of state guaranty funds.35
There is a danger, however, that large risk retention groups would
vitiate the important advantage of enterprise liability: the integration of
delivery, injury prevention, and insurance functions. Thus, when the
hospital is of sufficient size to support an enterprise insurance plan, this
is preferable to a risk retention group.

Established Forms of Liability Applicable to Hospitals: Stepping-stones

to Enterprise Liability?

Until the most recent medical malpractice crisis, medical malpractice was
largely viewed as a problem for physicians. Hospitals and nursing homes
were rarely mentioned as victims of tort. However, a large number of
322 Chapter 12

medical malpractice claims originate from care provided in hospital set-

tings, and the physician specialties paying the highest medical malprac-
tice premiums all deliver a substantial amount of their care in hospitals.36
Moreover, the vast majority of claims involving alleged wrongdoing on
the part of physicians for care provided in hospitals name members of
medical staffs of hospitals as defendants, not physicians who are employ-
ees of the hospital.
The doctrine of respondeat superior has been applied to staff model
health maintenance organizations, forcing them to bear legal responsibil-
ity for their employed physicians.37 Respondeat superior is a legal doc-
trine which makes employers liable for the actions of their employees
during the course of their employment, and allows plaintiffs to hold
otherwise unrelated parties vicariously liable. However, in most U.S.
hospitals, in contrast to those in the vast majority of other countries,
physicians typically are not employed by the hospital, but serve on an
independent contractor basis. This has made it difficult to use the
doctrine of respondeat superior, as it requires an employer/employee
relationship; direct employment by the hospital is needed.
In the absence of an employee/employer relationship, an alternative
theory of liability is ostensible, or apparent, agency.38 Ostensible agency
is based on the principle that an agent (the physician) acting for the
principal (the hospital) may cause a third party (the patient) to reason-
ably believe the principal has employed the agent. Ostensible agency is
a cause of action that courts have recognized in place of respondeat
superior, but in the process they have also raised the standard of proof
for plaintiffs, requiring them to prove both an apparent agency relation-
ship, and a reliance on that relationship.39 Factors that courts consider
in determining reliance on an agency relationship include, but are not
limited to, the fact the hospital supplied the facilities, equipment, and
medical personnel, and also set policies and procedures for the hospital.
Perhaps more important, the court considers whether or not the patient
looked to the institution, not the individual physician, for care and had
a practical opportunity to choose other physicians.40
Corporate liability is another cause of action that courts have used to
hold hospitals liable for the actions of their providers in the absence of
an agency relationship. Corporate liability extends the scope of a hospi-
 Reforms: What Can Be Done 323

tal’s potential liability. Under this doctrine, a hospital may be directly

liable for medical errors on the part of physicians selected by and under
contract with the hospital, a common practice for physicians in a few
specialties (e.g., emergency room medicine, anesthesiology, pathology,
and radiology). They can also be held directly liable for medical errors
of physicians who have been reviewed by the medical staff for privileges
and selected by the patient for treatment.41
The transition from the days of the charitable immunity doctrine,
which protected most hospitals from liability for actions undertaken by
medical staff, to agency and corporate liability has been gradual, with
the effect that only since about 2000 has hospital medical malpractice
been widely considered to be a major issue.42
Under corporate liability the hospital is directly liable for the negligent
acts of its medical personnel, even those working on a contract basis.
Thus, a plaintiff can sue the hospital directly, omitting the need to prove
the existence of an agency relationship.43 Unlike enterprise liability, to
be discussed next, corporate liability does not consolidate the causes of
action; a plaintiff may sue the hospital in addition to individual health
care providers. It is also necessary for the plaintiff to prove the physi-
cian’s malpractice as a prerequisite to a valid claim against the
hospital.
One of the most important elements the plaintiff needs to prove in
order for a claim to be successful under a corporate liability theory is
that the hospital breached one of its duties.44 The court in Thompson v.
Nason45 explains the hospital’s duties in its opinion. Labeled the “Thomp-
son duties,” they are “to use reasonable care in the maintenance of safe
and adequate facilities and equipment; to select and retain only compe-
tent physicians;46 to oversee all persons who practice medicine within its
walls as to patient care; and to formulate, adopt, and enforce adequate
rules and policies to ensure quality care for the patients.”47
Although the breach of one of these four duties is essential to the case,
absent obvious negligence, the prima facie case against the hospital is
made up of three parts. First, there must be a breach of duty; second,
there must be actual or constructive notice of the defects or procedures
that created the harm; and the final element is proving that the hospital’s
conduct was a substantial factor in causing the harm.48 Many states have
324 Chapter 12

adopted corporate liability through case law, and these duties have also
been codified by the Joint Commission on the Accreditation of Health-
care Organizations (JCAHO), as well as state legislators and hospitals.49
However, without specific legislation, it is unlikely that liability will
progress past corporate liability, forgoing the benefits of consolidated
claims.

Enterprise Liability for Hospitals: An Attractive Reform That Faces

Political Obstacles

Definition and Rationale

Under enterprise liability, the hospital would be the named defendant in
lawsuits brought by plaintiffs who allege they were injured as a conse-
quence of care received while they were being treated at the hospital.
Separate suits against individual physicians would not be filed. This is
in contrast to enterprise insurance, where the enterprise would be the
insurer of physician defendants, but would be named in lawsuits only
under the doctrine of vicarious liability just described. In this important
sense, enterprise liability is a larger change from the status quo than is
enterprise insurance.
The concept of enterprise liability for hospitals is at least three decades
old. In 1975, Myron Steves, then a partner in an excess and surplus lines
insurance agency in Houston, Texas, proposed “shifting liability expo-
sure and the cost of insuring it from individual practitioners to institu-
tional providers for incidents occurring within institutional settings.”50
He further proposed that exposure to nonhospital risk be insured sepa-
rately on an individual basis. He advocated that hospitals self-insure for
medical malpractice expense coupled with aggregate excess insurance
(reinsurance) for high losses.
In addition to the advantages of enterprise insurance discussed above,
Steves argued that once it is exposed to medical malpractice lawsuits for
care delivered within its walls, the hospital would have an incentive to
develop databases for medical injuries as well as claims. Hospitals are
well positioned to assemble such data, certainly compared to physician
practices (except for large medical groups). Steves also proponed that if
a case goes to trial, there is an advantage to defendants in presenting a
 Reforms: What Can Be Done 325

common defense. With 25 percent of malpractice suits involving more

than one defendant, a common defense would consolidate many of the
costs associated with filing, litigating, and defending a suit.51 There
would be fewer attorneys involved, lower court costs, and a lesser amount
of related legal expenses. In addition, with only one defendant, the need
for adversarial litigation tactics would be decreased because the defen-
dants do not have competing interests; thus negotiation and settlement
time are reduced.
Left unsaid in general discussions of enterprise liability is specifically
how the burden of hospital premiums would be shared. It would be
advisable that physicians bear some part of the premium burden to
provide some incentive to avoid claims. Hospitals could implement their
own systems of surcharging physicians with many medical malpractice
claims. Of course, hospitals, or medical staffs operating on their behalf,
would retain the option of removing physicians with adverse claims
experience or those who do not comply with hospital patient safety regi-
mens from their medical staffs. In fact, they would have a greater incen-
tive than they do currently to monitor physician performance and to
remove physicians with adverse claims experience.

Comparison with Enterprise Insurance

Enterprise insurance does not change the cause of action against physi-
cians and hospitals, nor does it change the named defendants. Enterprise
insurance merely shifts the identity of the insurer from commercial insur-
ance to an insurer linked to the enterprise. Because enterprise insurance
does not alter the nature or function of a malpractice case, it remains in
the best interest of each defendant to assert a defense that would exoner-
ate him- or herself, regardless of fellow defendants. Theories of defense
from the various defendants can be contradictory, and proceedings
between the parties, such as settlement negotiations, are adversarial. In
some cases, individual defendants settle, leaving the hospital to litigate
the claim.52
With enterprise liability, claims against various physicians and other
providers are consolidated into one cause of action against the hospital.
This shifts liability from physicians to the hospital, making the hospital
responsible for defending its choices and also the quality of its providers,
326 Chapter 12

among other things. The hospital also would be responsible for the neg-
ligence and errors of its physicians. Actions can be brought for negligent
hiring (not properly reviewing providers’ credentials), breach of duty to
provide adequate health care,53 and failure to maintain facilities. This is
by no means an exhaustive list. Regardless of what theory the case is
brought under, the defense will be common to both hospital and pro-
vider, as the hospital has a vested financial interest in vindicating itself
and its providers.

The Clinton Administration’s Attempt to Incorporate Enterprise

Liability in its Health Reform Program
To date, the only federal legislation regarding enterprise liability was
contained in the Health Security Act of 1993, a major proposal of the
Clinton administration’s first term. The Clinton plan proposed imposing
enterprise liability on health plans, thereby transferring liability from
physicians to the health plan.
Having health plans assume physician liability is deficient for several
reasons. First and foremost, the typical physician contracts with a number
of health plans. Multiplicity of plans has deterred widespread use of
plan-specific practice guidelines. Further, there has been substantial
entry, exit, and mergers of health plans. This provides an unstable anchor
for a long-tail line such as medical malpractice insurance.
Particularly following the revolt against managed care in the 1990s,
it seems highly unlikely that anyone would advocate that liability be
transferred to such plans. Physicians were never enthusiastic about this
aspect of the Clinton proposal, which promoted managed care more
generally. Both patients and providers were uncomfortable with the
notion of “corporate healing.”54

Objections to Enterprise Liability

Even with the hospital designated as the enterprise, various groups have
expressed concerns about enterprise liability. A major concern is that by
removing an individual physician’s liability, the incentive for deterrence
would be lost.55 But if this objection is at all sincere, there is little empiri-
cal evidence that the threat of tort liability—as it exists today—deters
injuries.
 Reforms: What Can Be Done 327

With hospital enterprise liability, the deterrent would be internalized

to the hospital. These organizations could impose a combination of
financial and nonfinancial incentives for individual physicians to prevent
injuries, coupled with increased surveillance measures. Also, the hospital
and physicians at the hospital collectively would have an incentive to
promote patient safety since the enterprise’s premiums would depend on
future anticipated losses from medical malpractice claims.
A second objection is that enterprise liability would create a deep
pocket which would increase claims frequency, primarily because tort
claimants would be more successful in obtaining higher compensation
from supposedly rich and faceless institutions, many of which have
headquarters in a distant city, than from individual physicians practicing
in their own communities.56 But some insurers presumably have deep
pockets as well.
Third, enterprise liability may restrict patient choice of provider.
Physicians may have to limit their admissions to the one hospital at
which they receive medical malpractice health insurance coverage. Physi-
cians frequently have privileges at more than one hospital. This potential
concern can be largely remedied by limiting physician coverage to care
delivered within the walls of the facility under the hospital’s policy. Thus,
if a physician practiced at three hospitals, he or she would be covered
under three hospital policies. In addition, the physician would need to
obtain medical malpractice insurance for care delivered in the office, but
such coverage would be at a greatly reduced premium.
Fourth, physicians complain about the trend in loss of autonomy.
Enterprise liability would probably exacerbate this trend. An argument
for autonomy is that it allows providers to use their professional skill
and judgment in particularized situations. Outsiders, such as hospitals
(or health plan review boards) may not be well positioned to know all
the details and considerations evident in a physician-patient interaction.
Furthermore, enterprise liability may increase the bargaining power of
hospitals and health plans vis-á-vis individual physicians. This may be
so, but there is no reason to believe that the current power relationship
between hospitals and individual medical staff members is ideal.
Physicians have much more autonomy in U.S. hospitals than do their
counterparts in most other countries.
328 Chapter 12

Fifth, inpatient care is shrinking as a share of the total personal health

care dollar. More care is being delivered on an ambulatory basis. So, as
Danzon writes, “It is increasingly anachronistic to view the hospital as
the focus of care and hence as the best locus of liability.”57 But this
concern disregards another trend. Hospital-provided ambulatory care is
growing, and hospital enterprise liability would encompass care at all
sites at which the hospital organization or system provides care.

Bottom Line on Advantages and Disadvantages of Enterprise Liability

Enterprise liability, despite its drawbacks, addresses many aspects of the
malpractice crisis, including patient safety. A major barrier is that a
political constituency for enterprise liability is lacking at both federal
and state levels. Health care consumers are not well organized. Providers
appear to be concerned about the “deep pockets” argument as well as
potential loss of autonomy, and at the same time hospitals and health
plans apparently see no evidence that assuming liability for physicians
and other providers would make them more attractive in the
marketplace.58

Incentives to Encourage Adoption of Enterprise Insurance or

Enterprise Liability
Given that existing market forces have not led to adoption of either
enterprise insurance or enterprise liability, what types of specific policy
interventions may be implemented to encourage its adoption by hospitals
and their medical staffs?
As already noted, legislation is needed for enterprise liability to exist.
Presumably legislation could also contain a provision forcing physicians
to participate in at least one hospital’s program. Another compulsory
approach would be for Medicare to require participating hospitals to
have patient safety programs which would include enterprise insurance
as a component.
Rather than use the stick to promote enterprise liability, there are also
carrots. One approach would be to offer public reinsurance to hospitals
with insurance plans that satisfy certain criteria. Hospitals with qualified
plans could be granted the option of offering no-fault plans as an alter-
native to tort. Abraham and Weiler contend that implementation of
 Reforms: What Can Be Done 329

enterprise liability may be a stepping-stone on the road to no-fault.59

What they seem to have in mind as an ultimate goal is a universal no-
fault of the type that has been implemented in a few countries. However,
recognizing the difficulty in moving to a mandatory strict medical
liability system, they propose an elective version of no-fault.60 What is
proposed here follows the same premise—a voluntary plan offered as an
alternative to tort by individual hospitals.

Reforms Possibly Linked to Provision of Enterprise Insurance and/or

Enterpriase Liability

Public Reinsurance/Excess Insurance

Public reinsurance or excess insurance (when applied to hospitals) plans
have worked well when premiums have been set on an actuarial basis
with loss reserving by policy year rather than on a pay-as-you-go basis,
thresholds for coverage are indexed for inflation, and there is experience
rating of reinsurance premiums (chapter 10). A rationale for public pro-
vision is that (1) high premiums on private reinsurance contracts reflect
a high cost of risk bearing and (2) there are cycles in availability and
premiums that can be very disruptive to primary insurers.
Although provision of publicly subsidized excess insurance would be
attractive to hospitals, it may be difficult to offer coverage at actuarially
fair premiums that are appreciably below premiums prevailing in the
private reinsurance market, given that a public program would face risk
of catastrophic loss similar to that of its private counterparts. However,
public monies could be used to fund a pool to cover possible shortfalls.
This public expenditure could be limited by restricting availability of
public coverage to those hospitals with qualified enterprise insurance
plans. To the extent that PCFs reduce volatility of availability and pre-
miums in the reinsurance/excess insurance market, volatility in the overall
market for medical malpractice insurance would be reduced.61

Voluntary No-Fault
No-fault insurance has several attractive features relative to tort: in
particular, low administrative expense and speedier payment. Coverage
of monetary losses may be broader, while payment for nonmonetary
330 Chapter 12

loss is likely to be more restricted. The hospital could be allowed

to charge patients a premium if they opt for the no-fault plan. This
premium would be a net savings to the institution from eliminating tort
claims.
A hospital with a good patient safety record may find it advantageous
to offer no-fault insurance. The very fact that such insurance is being
offered may be taken as a signal that the hospital/medical staff have
confidence in the safety record of the hospital. It would be expected that
hospitals would be open about events that led to injury. However, it
would be counterproductive to impose an obligation of disclosure on
such plans absent a claim, unless it was in the business interest of the
hospital to do so.
It is important, however, that coverage extend to a large number of
conditions. Exclusions from coverage would need to be broad and easily
understood by patients. Very narrow thresholds, such as those in exis-
tence in Florida and Virginia, are difficult for patients to assess in
advance of injuries. A few very costly procedures may be excluded from
coverage, but it would be important that these be listed and described
in understandable terms in advance.
Patients would contract for no-fault coverage well in advance of receiv-
ing care at the hospital, thus avoiding situations in which they have an
active health condition and face immediately at point of service, which
could be interpreted in legal terminology as an adhesion contract.62 For
employer-sponsored plans, it would be a simple matter for an employee
to designate whether or not he or she wishes to substitute no-fault for
tort. The employee would select a plan for services and receive services
from the specific hospital with which that plan contracts. Surcharges for
no-fault (if surcharges are imposed on patients) would then be built into
the premium charged. Or, alternatively, payment (presumably a higher
one) could be made at the time the employees/families present at the
hospital for care.
In the case of voluntary no-fault, since insured patients would agree
not to sue under tort, the savings in tort payments would offset at least
part of the cost of the no-fault plan. Other options, such as for the use
of alternative dispute resolution mechanisms, could also be made on this
basis.
 Reforms: What Can Be Done 331

No-fault plans would require prior regulatory approval, depending on

the applicable regulatory authority. Regulators would pay attention to
the method of enrolling persons into the plan, pricing, and issues bearing
on plan solvency.
This type of voluntary no-fault program would offer several important
advantages to hospitals and their medical staffs. First, it would relieve
providers of the threat of tort. Second, offering no-fault benefits would
be a signal that the hospital has an effective patient safety program and
low rates of medical errors and injuries. Third, to the extent that injury
victims value quick payments with little involvement of attorneys, this
should increase demand for the hospital’s services as well.
A universal compulsory no-fault program is infeasible in the United
States. Unlike the countries that have adopted no-fault for medical inju-
ries, the United States lacks universal health insurance coverage. It seems
improbable that the United States would give such priority to persons
with ioatrogenic injuries over all other possible reasons for obtaining
medical care.

Reforming Determination of Liability and Damages

Health Courts
The rationale for health courts—that lay juries and even judges lack the
extensive technical background needed to understand complex evi-
dence—is part of a much larger issue about the use of scientific evidence
in the courts more generally. Health courts represent only one of several
nonmutually exclusive alternatives for addressing this issue. These alter-
natives include (1) use of court-appointed experts, (2) bifurcated trials,
(3) use of special masters, (4) specially convened expert panels, (5) blue
ribbon juries, and (6) alternative dispute resolution.63
Under trial bifurcation, the court tries liability and damages at separate
stages of litigation. Bifurcation helps to limit the number of issues that
a judge or jury must consider in a single trial. Special masters may be
appointed at any stage in the trial process. They possess special expertise
in the issues at hand, and can aid the court in assessing important techni-
cal issues. Court-appointed experts or expert panels can be used to
present evidence in a more neutral fashion than experts appointed by the
332 Chapter 12

litigants who have an incentive to present one side of the evidence.

Unfortunately, with this method, the other side of the evidence may not
be revealed during cross examination either because the cross examining
attorney lacks the expertise to effectively probe into narrow technical
issues or because the expert simply stonewalls. Blue-ribbon juries might
be drawn from a pool of experts on a given topic.
For example, in assessing liability and damages in a case involving a
brain-damaged infant, the jury might be composed of obstetricians,
pediatricians, special educators, psychologists, social workers,
clergy, and others who have professional and practical experience
in dealing with individuals (and families) with such injuries. By contrast,
a blue-ribbon jury for a failure to diagnose cancer case might include a
radiologist, a pathologist, a sonographer, and a risk communication
expert.
States might consider health courts as an option. Nevertheless, we
are reluctant to give this option our enthusiastic endorsement. Preserv-
ing juries in some form, even if they are blue-ribbon juries, would
provide broader representation of perspectives and values than would
sole reliance on a narrow group of professionals to make judgments
on specific cases. Even a judge with health expertise will not be able
to be expert on the full range of issues health courts are likely to
confront. In the end, as Mehlman acknowledges, it is important that
any health court be viewed as legitimate by plaintiffs as well as defen-
dants.64 If the court consists entirely of or is dominated by physicians
and other health professionals, buy-in by plaintiffs seems highly
improbable.

Scheduled Damages
There is widespread concern about the variability and unpredictability
of damages, especially for nonmonetary loss. Although only a very small
fraction of legal disputes ever reach verdict, results at verdict presumably
guide settlement negotiations. The major public policy response of the
states has been to impose a cap on nonmonetary damages or on total
damages.
Such caps have two virtues. First, they are simple to administer.
Second, there is empirical evidence that imposing caps reduces payments
 Reforms: What Can Be Done 333

for damages and for administrative expense (chapter 5), which is plau-
sibly the reason that they have received so much political support.
Caps are flawed because they assume that unjustifiable variability and
unpredictability are much more common for severe injuries, a premise
for which empirical support is lacking. Also, the vast majority of
caps are not indexed for inflation. Thus, unless states explicitly
change the statutes, over time the real value of the cap will approach
zero. This approach amounts to taking away individuals’ access to recov-
ery through tort via the back door. A more direct (and honest) approach
would be to legislate an end to recovery on grounds of medical
malpractice.
As explained in chapter 5, there is some academic support for the view
that there should be no payment for nonmonetary loss, but there is also
substantial academic support for the position that such payments should
be made. Whatever the intellectual arguments are, explicitly eliminating
payment for nonmonetary loss is not an achievable reform. And there is
a long history of paying for nonmonetary loss in the United Stated, at
least back to the late nineteenth century, when Civil War veterans were
paid pensions based in part on such loss. Achieving this goal by the back
door—that is, specifying caps in nominal terms and not updating them—
seems to be feasible. A cap fixed in nominal dollars is not a policy that
the advocates of such damage caps would like to be applied to
themselves.
The U.S. Congress could hypothetically decide to freeze physicians’
fees paid by Medicare for reasons of simplicity and cost containment.
Such a policy would be vehemently opposed by the advocates of nominal
caps on damages, such as California’s, which have not been increased
since 1975. All told, if variability, predictability, and accuracy of loss
determination are to be improved, this goal should be accomplished in
a way other than flat caps fixed in nominal dollars.
Several approaches for achieving this goal, in particular with regard
to nonmonetary damages, have been proposed.65 None of them have
been implemented to date.
At one extreme, all proposals for scheduling damages will be
opposed by the organized trial bar on grounds that they limit jury
discretion in awarding damages. There is a trade-off between complete
334 Chapter 12

individualization of awards and reducing volatility and increasing pre-

dictability of awards. At the other extreme, groups representing defen-
dants will object to proposed alternatives to the present system, including
flat caps, on grounds that they are needlessly complex. These same
groups, however, tend to have no problem with complexity when their
reimbursement is involved.
Rather than select among the various proposals that have been pub-
lished or among many others that might be developed, it is the notion
of scheduling and its feasibility that is endorsed here. States should care-
fully examine the options, listening to experts on all sides. No system
for scheduling nonmonetary loss can be perfect, and states are likely to
differ in their assessments of the importance of various advantages and
deficiencies.
In any event, total loss should not be scheduled or capped. However,
it would be appropriate for states to review the instructions that are
provided to juries in order to assess whether guidelines for determination
of monetary loss should be developed. Cost-benefit and cost-effectiveness
analysis are becoming more widely used, and some developments in these
fields may be usefully applied to damage assessment in medical
malpractice.

Contracts for Services

Service contracts, described in chapter 5, represent a missed opportunity
for organizations that are sufficiently vertically integrated, and can add
some additional integration, to offer care on a comprehensive basis to
persons with severe medical injuries. Large academic health centers
(AHCs) are such organizations. The contracts combine service provision
and insurance functions. An AHC would at most very rarely have an
affiliation with an insurer. AHCs could find insurer partners and market
such contracts as a joint venture. Being involved in such contracting
would provide potential teaching and research opportunities as well.
The two small no-fault programs in Florida and Virginia have gained
experience with managing the care of severely impaired children over the
life course (to date only the first part of the life course). Admittedly, such
contracting is a great idea in principle, but in practice, AHCs may not
be ready for this. However, as cost containment pressures from payers,
 Reforms: What Can Be Done 335

including Medicare, continue to build, this type of program may look

increasingly attractive.

Final Word

In a very readable and useful book, Tom Baker exposes the myths of
medical malpractice.66 By contrast, the present book has described only
a few of these myths. Perhaps the myths have had the positive effect of
alerting the public and their elected representatives to the medical mal-
practice issue. The myths and the name calling have had the unfortunate
consequence of leading to adoption of both ineffective and often mis-
guided public policies. In particular, there is a missed opportunity in not
aligning the financial incentives for patient safety with the rhetoric that
medical care is unsafe and considering medical malpractice as something
to be limited rather than reformed in a way that it could be more effec-
tive in achieving its deterrence function. In sum, we should and can do
better.
Notes

Chapter 1

1. This is discussed in chapter 4.

2. Chapter 12 addresses the prospects for medical malpractice to have a produc-
tive role in injury prevention.
3. Kersh (2006).
4. It is helpful to outline the functions and goals of the tort system both gener-
ally and in relation to medical malpractice. These points are explained much
more fully in textbooks on tort law and on law and economics. See, e.g., Shapo
(2003) and Micelli (2004). To begin with, medical malpractice is a cause of
action within the law of torts. Torts are neither criminal nor contractual in
nature, although some torts are also crimes punishable by imprisonment. In a
civil action for tort, the plaintiff attempts to hold the defendant liable for causing
harm to his or her person or property, whether or not the defendant’s actions
were deliberate or simply careless. If the plaintiff is successful, the court shifts
the plaintiff’s loss to the defendant by granting either monetary compensation
or an equitable remedy. Tort doctrines are largely in the state domain and have
evolved over centuries from both common and statutory law.
In tort, an injury is not enough to obtain damages. The plaintiff must prove
four elements for a successful claim in medical malpractice tort cases. These ele-
ments do not apply to all tort actions. There are three different categories of
tort—intentional, negligent, and strict liability—each with different requirements
for proof of liability. The vast majority of medical malpractice cases are negligent
torts. However, battery is a type of intentional tort that has been used in medical
malpractice cases. Usually a battery case involves lack of informed consent or
failure to respect an advance directive, and requires a showing of voluntary
intentional action, not negligence.
Those who cannot demonstrate each element receive no compensation. These
elements are (1) the defendant owed a duty of care (duty), (2) he or she failed
to conform to the required standard of care (breach of duty), (3) this failure was
the proximate cause of the plaintiff’s injury (causation), and (4) the plaintiff
338 Notes to pages 3–13

suffered either pecuniary or emotional damages (damages). Most medical mal-

practice lawsuits are filed as negligent tort actions, as opposed to strict liability,
or as an intentional tort. To prove negligence, plaintiffs must present evidence
that is convincing to a jury on three criteria: an injury occurred; an action or
inaction of the defendant caused the injury; and the action or inaction represents
a failure to exercise due care. Nevertheless, even when a plaintiff is successful in
proving the negligence of the defendant, if the plaintiff suffered only a minor
injury, he or she should not expect much compensation.
5. Dram shop statutes make alcohol servers liable for subsequent accidents
caused by persons who drank in their establishment or were under the legal
drinking age (Sloan, Stout, Whetten-Goldstein, et al. 2000).
6. White (2004).
7. Sturgis (1995, p. 8). Whether compensation can be provided at a lower cost
and more equitably through public and private insurance systems is a point of
contention, and will be discussed at length later in this book.
8. This is discussed further in chapter 3.
9. This is discussed further in chapter 6.
10. State of Maryland (2004, p. 24).
11. Vidmar (1998, 2004).
12. Sloan, Bovbjerg, and Githens (1991).
13. Abraham (2001).
14. See, e.g., Mello and Studdert (2006); Sloan (1990).
15. Experience rating is discussed further in chapter 3.
16. Schwartz and Mendelson (1989).
17. See, e.g., Shavell (1980).
18. See, e.g., Perssons and Tabillini (2002); Sloan and Hsieh (1990).
19. E.g., Becker (1983); Posner (1976); Stigler (1971).
20. This is discussed in more detail in chapter 10.
21. Hickson et al. (1992, 1994, 2002); Sloan and Hsieh (1995); Farber and
White (1991); Nalebuff (1987); Nalebuff and Scharfstein (1987); May and
Stengel (1990); Sieg (2000).
22. Weiler et al. (1993); Mills, Boyden, and Rubsamen (1977).
23. E.g., Danzon and Lillard (1983); Sloan and Hsieh (1990).
24. Sloan et al. (1993).
25. Ibid.
26. Sloan, Whetten-Goldstein, et al. (1998).
27. Bovbjerg and Sloan (1998).
28. Vidmar (1995, 2003).
 Notes to pages 13–19 339

29. Helland and Taborrok (2003).

30. Bovbjerg (1991).
31. See, e.g., Doherty, Lamm-Tennant, and Starks (2003); Doherty and
Posey (1997); Nye and Hofflander (1987); Lai et al. (2000); Winter (1988,
1994).
32. Shavell (1987).
33. E.g., Fournier and McInnes (2001); Ellis, Gallup, and McGuire (1990);
Sloan and Hassan (1990).
34. Sloan (1990).
35. Schwartz and Mendelson (1989).
36. Sloan, Bovbjerg, and Githens (1991).
37. Schlesinger and Venezian (1986).
38. Hoerger, Sloan, and Hassan (1990).
39. Danzon (1985, 1994).
40. Danzon, Pauly, and Kingston (1990).
41. Kessler and McClellan (1996); Localio et al. (1993); Sloan, Entman, et al.
(1997).
42. Institute of Medicine (2000).
43. Entman, Glass, Hickson, et al. (1994).
44. Sloan, Mergenhagen, Burfield, et al. (1989).
45. Danzon (1994).
46. Sloan, Mergenhagen, and Bovbjerg (1989); Danzon (1986); Zuckerman,
Bovbjerg, and Sloan (1990); Born and Viscusi (1994); Born, Viscusin and
Carelton (1998); Sloan (1995).
47. Harrington (1994).
48. Sloan, Bovbjerg, and Githens (1991).
49. E.g., Shavell (1987).
50. See, e.g., Tinetti, Bogardus, and Agostini (2004).
51. Discussed in chapter 3.
52. E.g., Abraham (1986).
53. B. L. Smith (1992); Munch and Smallwood (1980); Hay (1996).
54. Both sides must answer questions during the discovery process. Experts on
both sides are deposed. During a trial, each witness is examined directly and is
subject to cross-examination.
55. Mills, Boyden, and Rubsamen (1977).
56. Weiler et al. (1993).
57. This is discussed further in chapter 8.
58. Studdert, Mello, and Brennan (2004).
340 Notes to pages 19–33

59. See chapter 8 for further discussion.

60. Sieg (2000).
61. Hoerger, Sloan, and Hassan (1990).

Chapter 2

1. Underwriting standards are the criteria insurers use for deciding whether or
not to insure an individual or an organization.
2. There are some differences in dating of the cycles. For example, some would
start the cycle of the mid-1970s in 1974. See graphs at http://www.iii.org/media/
facts/statsbyissue/pcinscycle/ (accessed July 21, 2006).
3. See, e.g., T. Baker (2005b).
4. Discussed in chapter 7.
5. See chapter 6.
6. Keynesian economics is based on the ideas of the twentieth-century British
economist John Maynard Keynes. Keynesian economics promotes a mixed
economy in which both the state and the private sector play an important role;
more specifically, it advocates monetary and fiscal programs run by the govern-
ment to increase employment and spending. Keynesian economics is in contrast
with laissez-faire economics, an economic theory based on the belief that markets
and the private sector can operate well on their own, without state intervention
(Doherty and Posey 1987).
7. Sloan, Bovbjerg, and Githens (1991).
8. See chapter 10 for a discussion on reinsurance.
9. Health insurance is a type of first-party insurance. It provides coverage for
losses the individual himself or herself incurs. For instance, if a car hits a fence
while backing out of the garage, automobile collision insurance covers the loss.
The threshold of evidence required for payment tends to be much lower for first-
party than for third-party insurance. For the former, it is only necessary to show
that an accident and loss co-occurred. Complex causation and negligence issues
are avoided. For this reason, the overhead of such insurance tends to be far lower
than for third-party insurance.
10. Rosenblatt (2004).
11. Sloan, Bovbjerg, and Githens (1991).
12. This may be somewhat of an overestimate, since these estimates seem to
start with the date the injury occurred, and the vast majority of policies sold
since the 1970s are sold on a claims-made basis.
13. Some medical malpractice insurers invest heavily in stocks (see Percy 2004),
but these are in the minority.
14. Kaufman and Ryan (2000, p. 3); Best (1998).
 Notes to pages 34–41 341

15. See chapter 10 for a detailed discussion of reinsurance.

16. Market risk is risk that the value of the security will fall; credit risk is the
risk that the company will be unable to pay its fixed obligations to investors.
17. Danzon (1985) suggests that the increase in medical malpractice premiums
during the 1970s may have reflected an increase in the correlation between
investment and underwriting returns, the latter reflecting the difference between
premium income and expenses attributable to payments to claimants, legal
expense, and other administrative expense associated with issuing insurance
policies. Insurers not only derive cash flow from investments but, to the extent
that returns on investments have no relationship to or are negatively related to
underwriting returns, investing provides a method of diversifying away risk. If
returns become more positively correlated, the diversification advantage of
investments diminishes. As a consequence, one would expect premiums to rise.
There is no good evidence that the correlation has been changing or, if it has, in
which direction the change has occurred.
18. This time interval is far too short for insurers to have learned anything
material about loss development during that policy year.
19. According to the U.S. General Accounting Office (June 2003, p. 15), after
adjusting for inflation, the mean annual increase in paid losses was about 3.0
percent between 1988 and 1997, but for 1998–2001, the mean annual rate of
increase in paid losses was 8.2 percent. This rate of increase was plausibly
reflected in higher initial incurred losses after 2000.
20. In much of the latter part of the twentieth century, and to a lesser extent at
the beginning of the twenty-first, economists studying both firm and consumer
behavior have typically assumed that firms are rational. Some financial models
applied to insurance pricing build on the rational expectations hypothesis. That
is, insurers combine all available information on future expected losses and
administrative expense to form informationally efficient predictors of future
losses (Fairley 1979; Moridaira, Urrutia, and Witt 1992; Kraus and Ross 1982;
Myers and Cohn 1987; Hill 1979). Further, the standard assumption is that there
are no “agency” problems. That is, insiders in the firm act to maximize the value
of their employing organizations.
However, both earlier and more recent studies relax these assumptions, allow-
ing for nonrational behavior and agency problems. Assumptions of rationality
and lack of agency problems are sometimes made for convenience.
21. Nye and Hofflander (1987, p. 502).
22. The word is derived from the Greek for “few sellers.” Because the decisions
of one firm influence and are influenced by the decisions of other firms, oligopo-
listic markets and industries are at the highest risk for collusion.
23. For simplicity, firm objectives are often assumed to be unitary in economic
analysis. That is, the firm has a clear objective, such as profit maximization,
which it pursues continually.
24. This type of argument is also made by Alice Rosenblatt (2004), an actuary,
in explaining cycles in the context of health insurance.
342 Notes to pages 42–51

25. Nye and Hofflander (1987, p. 502) provide empirical support for the loss
extrapolation hypothesis for some, but not all, lines of property-liability insur-
ance they investigated. They did not include medical malpractice insurance in
their study.
26. See Sloan, Bovbjerg, and Githens (1991) for a discussion of such behavior.
27. United States General Accounting Office (June 2003).
28. Nye et al. (1988); United States General Accounting Office (June 2003).
29. Cummins and Outreville (1987).
30. However, the presence of these internal conflicts substantially complicates
formal modeling of insurer behavior.
31. An example of behavior counter to the rational expectations hypothesis is
setting premiums based on past losses (Venezian 1985).
32. Surowiecki (2004).
33. Interest rates not adjusted for inflation rates.
34. Mankiw (2007, p. 542).
35. United States Department of Commerce (2006, table 706).
36. Such as the discounted cash flow and the capital asset pricing models.
37. See, e.g., Myers and Cohn (1987).
38. E.g., Doherty and Garven (1995); Gron (1994b); Gron and Lucas (1998);
Winter (1988, 1991, 1994).
39. Priest (1987).
40. Or, in another property-casualty insurance context, to a series of
hurricanes.
41. Fung, Lai, Peterson, et al. (1998) report that surplus changes relate to pre-
miums, but with a lag of from three to five years after the shock to surplus
occurred. These authors also find interest rate effects on premiums, but again
with a lag. An earlier study based on aggregate time series data from the prop-
erty-casualty insurance industry for 1949–1990 by Gron (1994b) finds that the
difference between the price of insurance and noncapital costs, referred to as the
price-payment margin (PPM), varies with surplus (net assets), but with a lag, and
increases and decreases in surplus have different effects on PPM.
42. Sloan, Bovbjerg, and Githens (1991).
43. Gron (1994a).

Chapter 3

1. State of Florida (2003, p. iii).

2. Ibid. (p. vi).
3. Ibid.
 Notes to pages 51–67 343

4. http://www/usatoday.com/news/washington/2005-01-05-bush-tort_x.htm,
(accessed July 7, 2006).
5. Danzon, Pauly, and Kingston (1990, p. 122).
6. See, e.g., United States General Accounting Office (1995); United States
Congress, Office of Technology Assessment (1994).
7. Microeconomics studies how individuals, households, and firms make deci-
sions to allocate limited resources, typically in markets where goods or services
are being bought and sold.
8. All economic models are abstractions. Economists evaluate the performance
of a model not on the assumptions, which are inherently unrealistic, but on the
accuracy of the predictions. The assumption of profit maximization has worked
well in predicting responses to changes in factors exogenous to individual physi-
cians, such as insurers’ change in payment rates on physicians’ fees and physi-
cians’ decisions to accept the insurers’ payments as payment in full (see, e.g.,
Sloan, Cromwell, and Mitchell 1978). More complex models can incorporate
more realism, but greater generality usually comes at a loss of the model’s predic-
tive ability.
9. United States Government Accountability Office (2005).
10. See, in particular, McGuire (2000); McGuire and Pauly (1991).
11. Bovbjerg and Berenson (2006, p. 223).
12. Sloan and Bovbjerg (1989); Sloan, Bovbjerg, and Githens (1991); Thorpe
(2004).
13. See, for example, commentary by Thomas Young, MD, in his presidential
address to the South Atlantic Association of Obstetrics and Gynecology in
January 2005. He complains that reimbursement has seriously lagged increases
in practice cost in general and the increases in medical malpractice premiums in
particular. Young (2005).
14. Rodwin, Chang, and Clausen (2006).
15. Ibid.
16. United States Department of Commerce (2006).
17. See, e.g., Blumenthal (2004).
18. See, e.g., Fuchs (1978); E. F. Hughes et al. (1972).
19. See, e.g., Newhouse et al. (1982a, 1982b); Schwartz et al. (1980) and articles
referenced therein.
20. Mello et al. (2005).
21. Brooks et al. (2005).
22. Sloan, Mergenhagen, et al. (1989).
23. Mello, Kelly, Studdert, et al. (2003, p. 227).
24. New Jersey Hospital Association (2002).
25. Luft (1980); Luft, Bunker, and Enthoven (1979).
344 Notes to pages 68–75

26. A loading factor represents a dollar amount added to the expected loss in
arriving at a premium.
27. Chapters 4 and 5.
28. United States General Accounting Office (June 2003, August 2003).
29. Mello, Studdert, and Brennan (2003).
30. United States General Accounting Office (June 2003, p. 5).
31. Sloan (1990).
32. United States Congress, Office of Technology Assessment (1994, p. 13).
33. Ibid. (p. 42).
34. Web of Science, http://scientific.thomson.com/products/wos (accessed
September 15, 2005).
35. Kessler and McClellan (1996).
36. Caps on damages are statutorily set limits on the amount of compensation
that can be awarded to the plaintiff for monetary and nonmonetary damages.
37. Punitive damages may be awarded when the defendant’s action is found to
have deliberately harmed the patient or to have represented conduct not befitting
a professional. A few states have abolished punitive damages in medical mal-
practice cases (punitive damages are rarely paid in the context of medical
malpractice).
38. Prejudgment interest refers to interest payments on the loss between the date
of injury or date a lawsuit is filed, and the date the verdict is reached. Limits on
prejudgment interest reduce such interest payments.
39. The common law allows the plaintiff to collect damages on elements of loss
for which he or she has received payment from other sources. Statutory changes
enacted by some states either have required that juries be informed of payments
from collateral sources before they assess compensation for monetary loss, or
that payments from collateral sources be deducted from compensation.
40. Caps on contingency fees limit a plaintiff’s attorney’s fees to a fixed percent-
age of the award or allow for judicial review of the proposed fee and approval
of the fee depending on the outcome of the review.
41. Rather than pay damages as a lump sum, under periodic payments for future
damages, payment is made only as long as the defendant lives or suffers from
the injury.
42. Under the common-law joint and several liability rule, individual defendants
in a tort suit can be held liable for the total amount of damages, irrespective of
their actual share of liability. Under joint and several liability reform, states have
limited an individual defendant’s liability to his or her share of total liability.
43. A patient compensation fund (PCF) is a government-operated mechanism
that pays the loss in excess of a statutorily determined amount. In one state,
Pennsylvania, the dollar ceiling on payments by the PCF is so low that the PCF
 Notes to pages 75–80 345

has in effect provided a middle layer of coverage in recent years (see Sloan,
Eesley, Conover, et al. 2005).
44. United States General Accounting Office (1993).
45. The lag may be relevant for claims frequency and premiums since there is a
delay from the date of injury to the date the claim is filed, and the constitutional-
ity of some reforms may be subsequently challenged. However, practice patterns
are likely to respond much more quickly to such statutory changes.
46. The rationale for a three-year lag is less clear than for claims frequency and
for premiums. When separate variables were defined for direct and indirect
reforms, neither had statistically significant effects on recordkeeping, diagnostic
tests, referrals for consultation, or time spent with patients. It is not clear that
less recordkeeping and time spent with patients would be desired by most well-
informed patients.
47. Sloan, Githen, Clayton, et al. (1993).
48. Localio et al. (1993).
49. Danzon (1994). Our review of the literature revealed no more recent
studies.
50. Studdert, Mello, Sage, et al. (2005) report such statements. This type of
study is subject to the same deficiencies mentioned during the discussion of physi-
cian supply.
51. In an otherwise excellent review, Danzon (2000, pp. 1370–1371) is not
cautious about extrapolating these insignificant findings. She writes: “The
point estimate is negative but not statistically significant. Taking this point
estimate at face value and extrapolating would imply that tort liability reduced
the rate of negligent injuries per admission by 29 percent (from 1.25 with no
liability to 0.89 with the current system) and reduced the overall rate of medical
injuries per admission by 11 percent (from 3.7 to 3.3). The failure to find signifi-
cant effects may be influenced by the small sample (forty-nine hospitals) and
imperfect instruments available. Moreover, at best these data would estimate the
marginal effect of changes in liability over the limited range of variation in the
New York sample. Considering these intrinsic limitations that bias against
finding significant effects, together with other evidence that physicians do per-
ceive a significant risk of suit and change their behavior in response to liability,
Weiler et al. conclude that liability plausibly does have a significant deterrent
effect.
This empirical evidence on deterrence benefits is consistent with the rough
calculations by Danzon (1985): that under quite generous assumptions about the
costs of defensive medicine, the malpractice system would pay for itself (yield
positive net benefits) if it reduced negligent injury rates by at least 20 percent,
ignoring such intangible benefits.”
These conclusions push these statistically insignificant findings too far.
52. According to Mello and Brennan (2002): “Recognizing the limitations of
the initial HMPS (Harvard Medical Practice Study), a different subgroup of
346 Notes to pages 80–87

investigators later took a second stab at modeling deterrence. One of the inves-
tigators had done some additional work on constructing measures of deterrence,
and we hoped that a more sophisticated model might yield more conclusive find-
ings than the initial models had. Over a three-year period in the mid-1990s,
several researchers in a variety of disciplines debated the proper specification of
these models, often clashing over suggested approaches. The number of different
models proliferated; by the conclusion of the project, the team had run models
with four different measures of malpractice risk, two different outcome measures,
and two different estimation strategies. In the end, we were unable to agree that
any one model was correctly specified, and also could not agree on how to
interpret the group of findings as a whole. As a result, we did not submit our
findings for publication” (p. 1611). The outcome measures were the theoretically
appropriate variables described in the text.
53. The survey data were also used in Hickson, Clayton, Entman, et al. (1994),
described earlier.
54. There were several underlying actions/inactions used: for instance, substan-
dard documentation and failure to counsel the patient about events or future
risk appropriately. In addition, errors could be coincidental and unassociated
with the outcome; associated with the adverse outcome but not related to either
cause or preventability of the outcome; or to the outcome in such a way that the
error may have caused or contributed to the adverse outcome; or missed the
opportunity to prevent it.
55. Danzon (1994, 1985); Shavell (1980).

Chapter 4

1. See, e.g., Dao (2005) for states’ responses to the most recent crisis: “The
impending battles over malpractice costs have in some states been wrapped in
the broader cloak of ‘tort reform,’ intended to restrict the civil liability of many
types of businesses. They also come at a time when President Bush has pledged
to push for federal restrictions on medical malpractice lawsuits. But in most of
the states, soaring malpractice premiums have been the driving force for the
campaigns—in part because compelling stories about doctors and their patients
have put human faces on the larger issue.”
2. See, e.g., Majoribanks et al. (1996).
3. Using data from interviews of physicians and attorneys, Peeples, Harris, and
Metzloff (2000) report that most physicians who were sued did not believe that
they were liable, and expressed a desire for vindication. The authors explain that
there is a distinct cultural difference between law and medicine. At the risk of
overgeneralizing, for lawyers, settlement is neutral and does not have a negative
connotation. For physicians, settlement implies fault.
4. Danzon (2000).
 Notes to pages 89–92 347

5. Ad damnum clauses are used in a complaint to set a maximum amount

of money that a plaintiff can recover under a default judgment if the defen-
dant fails to appear in court. An ad damnum clause can also set an absolute
limit on the amount of damages recoverable in the case, regardless of how
much loss the plaintiff is able to prove at trial. The reasoning is that a defen-
dant should not be exposed to liability greater than the ad damnum
solely because he or she does not come to court and defend himself or herself.
In most states and in the federal courts, a plaintiff can collect money damages
in excess of the ad damnum if proof can be presented at trial to support
the higher amount. http://www.answers.com/topic/ad-damnum (accessed October
2, 2006).
6. Barker (1992).
7. Both of these statutory changes are discussed in detail in chapter 6.
8. Bovbjerg (1995).
9. Both PCFs and JUAs are described in considerable detail in chapters 8
and 10.
10. In the late 1980s, Robert Rabin, a highly respected legal scholar of tort,
concluded that the “time for a substantive reform of tort doctrine has largely
passed. There is no turning back to the harsh world of wholesale exemptions
from fault responsibility. Correlatively, there is no persuasive indication of a
substantial move ‘beyond’ fault. . . . If modest improvements in the system are
to be effected, the would-be reformer must look elsewhere than substantive
doctrine” (Rabin 1988, p. 39).
11. Kinney (1995).
12. Scheduling damages is discussed in detail in chapter 5.
13. Generally, ADR is made up of any means of settling disputes outside
of the courtroom. The two most frequently used forms are mediation and
arbitration. Arbitration is basically a simplified version of litigation—without
discovery and with simpler rules of evidence. In mediation an impartial third
party facilitates an agreement in the common interest of all the parties
involved.
14. There are many advantages, as well as several disadvantages, to arbitration.
Some advantages are that its cost can be cheaper, the proceedings are generally
private, there is more flexibility than in a court, and that when the subject of
dispute is highly technical, arbitrators with the appropriate degree of expertise
can be appointed. Two disadvantages are that the parties have to pay for the
arbitrators, which adds an extra layer of legal expense, and that the rule of
applicable law is not binding.
15. Mediation is different from arbitration in that mediation sessions are not
decided in favor of one party or another. The mediator facilitates the negotiation
process between the parties. In addition, the parties are not bound to resolve
their dispute, and may pursue litigation if they are dissatisfied with the results
of mediation.
348 Notes to pages 92–96

16. See chapter 11 for a discussion of no-fault.

17. See chapter 6 for further discussion of this and related proposals.
18. Enterprise liability is discussed in greater detail in chapter 12.
19. Weiler (1991); Abraham and Weiler (1994).
20. Epstein (1978); Havighurst (1995).
21. If contracts are set at the group level, they may reflect preferences of the
median group member, but there is likely to be substantial heterogeneity within
the group, which would be lost in establishing a single standard for the group
as a whole. Furthermore, as Danzon (2000, p. 1382) points out, if a provider
tried to restrict care at the level of the individual patient in return for a lower
fee, courts may view this as a contract of adhesion because it is patient-specific;
the reasoning would be that an asymmetric power relationship exists between a
physician and an individual patient.
22. Kinney (1995); United States General Accounting Office (1993).
23. Kinney cites a president of the American Medical Association who stated
that no-fault approaches are “anti-professional” because they are “freeing physi-
cians from the consequences of their actions” (Kinney 1995, p. 123).
24. Abraham and Weiler (1994).
25. The U.S. Congress has become involved in tort reform from time to time,
but to date no important legislation has been enacted.
26. Whiteman (1985); Ray (1982); Kingdon (1981); Songer et al. (1985); Songer
(1988).
27. Songer (1988).
28. Spiller and Vanden Bergh (2003). Not only are they common, but for some
states, amendments are frequent. Louisiana’s tenth constitution, adopted in
1921, was amended over 530 times before being replaced by a new constitution
in 1972, which has subsequently been amended over fifty times (Tarr 1998). In
1998, eighteen states required only a simple majority to ratify an amendment
(Tarr 1998).
29. There have been only twenty-seven successful amendments to the U.S. Con-
stitution in 220 years.
30. Witt (2005).
31. The United States uses a common-law system as opposed to a civil-law
system (used in Europe, Latin America, Quebec, and the state of Louisiana). In
a civil-law system, judges look solely to statutes and written legal codes as the
basis for their decisions. This is in contrast with a common-law system, where
the judiciary does not rest its authority upon any express or positive statute or
other written declaration, but upon statements of principles found in the deci-
sions of courts. Some courts have defined common law as a law of necessity,
to be applied only in the absence of explicit, controlling statutory law. See
Metropolitan Life Ins. Co. v. Strnad, 255 Kan. 657 (1994); State v. Lawrence,
 Notes to page 96 349

98 Idaho 399 (1977). In practice, most courts make rulings using legisla-
tion, written legal codes, and constitutions. However, these rulings are heavily
informed by prior judicial decisions. For a detailed analysis of the difference
between the two legal systems, and their developments, see Zweigert and Kötz
(1987).
32. The benefit of a common-law system is that the common law is not static;
it is dynamic and has an inherent capacity for growth and change. As described
in American Jurisprudence: “It is to be derived from the interstices of prior
opinions and a well-considered judgment of what is best for the community. Its
development is informed by the application of reason and common sense to the
changing conditions of society, or to the social needs of the community which it
serves. It is constantly expanding and developing in keeping with advancing civi-
lization and the new conditions and progress of society, and adapting itself to
the gradual change of trade, commerce, arts, inventions, and the needs of the
country. It is said that public policy is the dominant factor in the molding and
remolding of common-law principles to the end that they may soundly serve the
public welfare and the true interests of justice. The fact that no case remotely
resembling the one at issue is uncovered does not paralyze the common-law
system, which is endowed with judicial inventiveness to meet new situations.”
(Thomas 2006)
33. V. E. Schwartz, Behrens, and Lorber (2000). For an in-depth discussion of
the politics of electoral accountability, see Tarr (2005).
34. Coolidge (1905, p. 213).
35. Statutes of repose, like statutes of limitation, place limits on how long after
an injury a suit may be brought. However, repose statutes begin at the time of
the negligent injury, not at the tie of discovery. Statutes of repose establish an
outer limit for when a suit may be brought. This prevents doctors from being
sued decades after a procedure is performed.
36. Krumlauf (2005). There were other changes in the law as well, which are
not included in the list in the text. See Am. Sub. H.B. 350, 21st Leg., Reg. Sess.
(Ohio 1996).
37. State ex rel. Ohio Academy of Trial Lawyers v. Sheward, 86 Ohio St. 3d
451; 1999 Ohio 123; 715 N.E.2d 1062.
38. The constitutionality of damage caps is explored separately in chapter 5.
39. The court cited the one-subject rule as the basis for declaring the statute
unconstitutional. Provisions in legislation must unite to form a single subject for
purposes of Section 15(D), Article II, of the Ohio Constitution. However, in a
1991 case the court held that a bill may contain more than one topic, so long
as there is a common purpose or relationship between the topics. (See Hoover,
19 Ohio St. 3d at 6; 19 Ohio B. Rep. at 5; 482 N.E.2d at 580.) This bill affected
eighteen different titles, thirty-eight different chapters, and over one hundred
different sections of the Revised Code, as well as procedural and evidentiary rules
and hitherto uncodified common law.
350 Notes to page 97

40. The constitution vests the legislative power of the state in the General
Assembly (Section 1, Article II, Ohio Constitution), the executive power in the
governor (Section 5, Article III, Ohio Constitution), and the judicial power in
the courts (Section 1, Article IV, Ohio Constitution). The constitution also speci-
fies that “the general assembly shall [not] . . . exercise any judicial power, not
herein expressly conferred” (Section 32, Article II, Ohio Constitution; State ex
rel. Ohio Academy of Trial Lawyers v. Sheward, 86 Ohio St. 3d 451; 1999 Ohio
123; 715 N.E.2d 1062). In 2005, the Arkansas Supreme Court challenged the
constitutionality of its tort reform altering the statute of limitations (Davis v.
Parham, 2005 Ark. LEXIS 300). But using the rational basis test, the court ruled
the statute had a rational relationship between the burden of proof required and
a legitimate government objective. Similarly, a 1981 case in the Alabama Supreme
court challenged the statute of limitations in the tort reform act (Reese v. Rankin
Fite Memorial Hospital, 403 So. 2d 158; 1981 Ala. LEXIS 3630). The court
upheld the statute, saying, “. . . we cannot substitute our judgment for that of
the legislative branch, whose enactments come to us clothed with a presumption
of validity.”
41. Krumlauf (2005).
42. Ohio courts have consistently held that challenges to legislation should be
brought in an action in the Court of Common Pleas rather than an extraordinary
writ to the Supreme Court. See State ex rel. Gaydosh v. Twinsburg (2001), 93
Ohio St.3d 576, 579; 2001 Ohio 1613; 757 N.E.2d 357; and Rammage v. Saros,
97 Ohio St.3d 430, 431; 2002 Ohio 6669, at P11; 780 N.E.2d 278.
43. Strahler v. St. Luke’s Hosp., 706 S.W.2d 7; 1986 Mo. LEXIS 248; 62
A.L.R.4th 735.
44. § 516.105 R.S.Mo. (2006).
45. Missouri Constitution, Art. I, §14.
46. See Reese v. Rankin Fite Memorial Hospital, 403 So.2d 158 (Ala. 1981);
Eastin v. Broomfield, 116 Ariz. 576 (1977); Gay v. Rabon, 652 S.W.2d 836
(1983); Lacy v. Green, Del. Super., 428 A.2d 1171 (1981); Pinillos v. Cedars of
Lebanon Hospital Corporation, 403 So.2d 365 (Fla. 1981); LePelley v. Grefen-
son, 614 P.2d 962 (1980); Anderson v. Wagner, 402 N.E.2d 560, appeal dis-
missed 449 U.S. 807, (1979); Johnson v. St. Vincent Hospital, Inc., 404 N.E.2d
585 (1980); Rudolph v. Iowa Methodist Medical Center, 293 N.W.2d 550 (Iowa
1980); Stephens v. Snyder Clinic Association, 631 P.2d 222 (1981); Everett v.
Goldman, 359 So.2d 1256 (La. 1978); Attorney General v. Johnson, 385 A.2d
57, appeal dismissed 439 U.S. 805 (1978); Paro v. Longwood Hospital, 369
N.E.2d 985 (1977); Linder v. Smith, 629 P.2d 1187 (Mont. 1981); Prendergast
v. Nelson, 256 N.W.2d 657 (1977); Perna v. Pirozzi, 457 A.2d (1983); Armijo
v. Tandysh, 646 P.2d 1245 (1981), overruled by Roberts v. Southwest Commu-
nity Health Servs., 837 P.2d 442 (1992); Comiskey v. Arlen, 390 N.Y.S.2d 122
(1976); Roberts v. Durham County Hospital Corporation, 289 S.E.2d 875
(1982), aff’d, 298 S.E.2d 384 (1983); Beatty v. Akron City Hospital, 424 N.
 Notes to pages 97–102 351

E.2d 586 (1981); Allen v. Intermountain Health Care, Inc., 635 P.2d 30 (1981);
Duffy v. King Chiropractic Clinic, 565 P.2d 435 (1977); State ex rel. Strykowski
v. Wilkie, 261 N.W.2d 434 (1978); Woods v. Holy Cross Hospital, 591 F.2d
1164 (5th Cir. 1979); DiAntonio v. Northampton-Accomack Memorial Hospi-
tal, 628 F.2d 287, 291–292 (4th Cir. 1980); Fitz v. Dolyak, 712 F.2d 330 (8th
Cir. 1983). The California Supreme Court held that its statutory provision
regarding periodic payment of future damages was constitutional because it did
not deny due process, equal protection, or trial by jury. American Bank and
Trust Company v. Community Hospital of Los Gatos-Saratoga, Inc., 683 P.2d
670 (1984).
47. Strahler v. St. Luke’s Hosp., 706 S.W.2d 7 (Mo. Ganc 1986). The Lochner
era is a period in American legal history (1890–1937) during which the U.S.
Supreme Court struck down many economic regulations. Many criticize the
Lochner era as a regrettable period in U.S. jurisprudence; this view is embodied
by the term “to Lochnerize,” which connotes fundamental judicial error (Cloud
1996). Lochner v. New York, 198 U.S. 45 (1905).
48. U.S. Congressional Budget Office (2004, pp. 8–9).
49. These summaries have been prepared by the Council of State Governments,
the National Governors’ Association, the National Conference of State Legisla-
tures, the U.S. General Accounting Office, and the Office of Technology Assess-
ment of the U.S. Congress (now defunct). At one time, the Robert Wood Johnson
Foundation had a program dedicated to the study of medical malpractice and
its reform. More recently, the Pew Charitable Trusts invested in such research
and evaluation. With very few exceptions, such as New York State’s involvement
in the Harvard Medical Practice Study (Weiler, Hiatt, Newhouse, et al. 1993),
states have not been involved in conducting such research or in sponsoring the
work of others.
50. See chapter 1 for a discussion of the participants in this market.
51. Nathanson (2004).
52. Shmanske and Stevens (1986).
53. U.S. Congressional Budget Office (2004); U.S. Congress, Office of Technol-
ogy Assessment (1994). See e.g., Bovbjerg (1995); Danzon (2000); Saxton (2003);
Studdert, Yang, and Mello (2004); U.S. Congressional Budget Office (2003).
54. But not earlier evidence. See Sloan (1985).
55. Sloan, Bovbjerg, and Githens (1991).
56. It is likely that this reform was evaluated in some studies in preliminary
analysis, but the reform was dropped from the results that were reported. One
cannot know this for sure, although it seems unlikely that in terms of savings,
periodic payments have anywhere near the cost-saving potential of caps.
57. They are discussed in greater detail in chapter 8.
58. Also see Gunnar (2004).
59. Sloan, Bovbjerg, and Githens (1991).
352 Notes to pages 103–110

60. Encinosa and Hellinger use the county as the observational unit.
61. Sloan, Bovbjerg, and Githens (1991).
62. The Task Force devoted three pages to an innovative medical no-fault
program it had implemented over a decade before the report was written.
63. Thorpe (2004, p. W4-27).
64. See Danzon (2000, p. 1396) for further discussion of this point.

Chapter 5

1. Glassman (2004); Kelly and Mello (2005).

2. Danzon (1984); Zuckerman, Bovbjerg, and Sloan (1990).
3. Born and Viscusi (1994); Born, Viscusi, and Carleton (1998); Sloan, Mergen-
hagen, and Bovbjerg (1989); Thorpe (2004); Zuckerman, Bovbjerg, and Sloan
(1990).
4. Sloan (1985).
5. Glassman (2004).
6. See, e.g., Browne and Wells (1999).
7. Kelly and Mello (2005).
8. Ibid.
9. The research cited in this chapter’s first paragraph evaluated the effects
of limits on damages in medical malpractice cases in general and did not
focus on California. Nor did this research—or, to our knowledge, most research
by others (exceptions being Pace, Golinelli, and Zakaras 2004; and Studdert,
Yang, and Mello 2004)—assess the other policy changes that California has
implemented that may account for the relative stability in premiums in that
state.
10. This topic was discussed more extensively in chapter 4.
11. E.g., Croley and Hanson ( 1995, p. 1803).
12. The function of insurance is to transfer money from the healthy to the sick,
but money is not a perfect substitute for irreplaceable loss (Cook and Graham
1977; Danzon 2000, p. 1372).
13. For the United Kingdom, see the Ogden Tables, 5th ed. (London:
Government Actuary’s Department 2004). http://gad.gov.uk/publications/docs/
ogdentables5thed.pdf (accessed June 20. 2007).
14. See, e.g., Rustad (1992).
15. Kelly and Mello (2005).
16. Bovbjerg (1991, p. 467).
17. See chapter 11.
18. E.g., Gan et al. ( 2001).
 Notes to pages 110–115 353

19. As explained below, economic theory offers a prediction of when an indi-

vidual would and would not want to purchase insurance in excess of anticipated
monetary loss.
20. See, e.g., Croley and Hanson (1995, pp. 1837–1841).
21. Croley and Hanson (1995). Theoretically, the amount people are willing to
pay for insurance depends on the value of an extra dollar in the healthy versus
the sick or injured state. Conclusive evidence on the marginal utility in healthy
versus sick states is lacking (see e.g., Viscusi and Evans 1990).
22. Viscusi (1979, 1992).
23. See chapter 6 for discussion of lawyer compensation issues.
24. Calfee and Winston (1993).
25. Of course, strong advocates for tort reform could argue the opposite. If we
had binding limits on lawyers’ contingent fees, then pressures to compensate
injury victims for nonmonetary loss would diminish.
26. Treaster (2003).
27. Pace, Golinelli, and Zakaras (2004).
28. Campion (2003).
29. The discussion of constitutional challenges draws extensively on Kelly and
Mello’s (2005) comprehensive analysis of this issue.
30. California’s law was challenged on this basis, but the court concluded that
there was no inherent right to collect an unlimited amount of tort damages.
31. Hallinan (2004); Niemeyer (2004).
32. Peeples and Harris (2005).
33. Vidmar, Gross, and Rose (1998).
34. See, e.g., Sieg (2000) for data on the percent of trials won by medical mal-
practice plaintiffs in Florida; and Vidmar, Robinson, and MacKillop (2006).
35. McCaffrey, Kahneman, and Spitzer (1995).
36. Helland and Tabarrok (2003).
37. Bovbjerg, Sloan, Dor, et al. (1991).
38. On undercompensation of medical injuries by tort, see the discussion later
in this chapter.
39. There is no sound empirical evidence linking an outlier award on the high
side to subsequent payments, either settlements or awards at verdict. However,
the conventional wisdom is that such a link exists and is important. For example,
in a brief critique of Hyman and Silver (2006), Ted Frank discusses a “problem
of outlier cases, with disproportionate damages that have a disproportionate
effect on the system” (2006).
40. Precedent-setting applies much less to jury trials since common practice is
not to inform jurors about previous awards. Nor is it likely that even the judge
has this type of information.
354 Notes to pages 115–118

41. Abraham (2001).

42. See chapter 2 for discussion of this point.
43. An example of a jury instruction on damages: “The purpose of the law of
damages is to award, as far as possible, just and fair compensation for the loss,
if any, which resulted from the defendant’s violation of the plaintiff’s rights. If
you find that the defendant is liable on the claims, as I have explained them,
then you must award the plaintiff sufficient damages to compensate him or her
for any injury proximately caused by the defendant’s conduct. . . . A prevailing
plaintiff is entitled to compensatory damages for the physical injury, pain and
suffering, mental anguish, shock and discomfort that he or she has suffered
because of a defendant’s conduct. . . . The damages that you award must be fair
and reasonable, neither inadequate nor excessive.” Section 77-3, Sand, Siffert,
Loughlin, et al. (2005).
44. Bovbjerg, Sloan, and Blumstein (1989).
45. Sloan, Githens, Clayton, et al. (1993).
46. See chapter 4; and Studdert, Yang, and Mello (2004).
47. Pace, Golinelli, and Zakaras (2004) quantify the effect of not allowing the
California cap to rise with increases in overall prices. Some other states that
implemented caps did not duplicate the effects in California (Mogel 2003). In
Indiana, which combines a limit on total damages with a patient compensation
fund (PCF), the mean payment for large medical malpractice claims was substan-
tially higher than the mean payment for similar claims in Michigan and Ohio,
states without PCFs and caps (Bovbjerg 1991; Gronfein and Kinney 1991).
Although this result most likely reflects idiosyncrasies of the PCF-total caps
combination in Indiana and mainly policies of the PCF, it serves as a reminder
that context and interactions between caps and other policy changes are critically
important to consider. Even though caps reduce payments on average, for cases
that would otherwise settle for much less than the cap, plaintiffs may base their
expected values of payment on the amount of the cap, thus forming unrealistic
expectations and reducing the probability of settlement (Pogarsky and Babcock
2001).
48. Hertzka (2003).
49. Pogarsky and Babcock (2001).
50. Chupkovich (1993).
51. For example, a North Dakota court found the state’s damage cap ceiling in
violation of the state equal protection guarantee. The statute benefited physicians
but denied adequate compensation to plaintiffs with proven meritorious claims,
and did nothing toward the elimination of nonmeritorious claims (de Sa e Silva
1988).
52. See, e.g., Bovbjerg (1991).
53. This lack of information applies to all tort cases, not just medical malprac-
tice. Gash (2005) proposed a national punitive damages registry to address a
 Notes to pages 118–123 355

problem particularly applicable to product liability, namely, that a defendant

who has allegedly injured multiple plaintiffs by a single action or a course of
conduct, faces multiple punitive damages for that conduct.
54. Having state registries based on jury verdicts (admittedly the minority of
claims) linked to a national registry would provide valuable data not only on
findings relative to liability but, most important, on the rationale for awards at
verdict. The following items might be recorded by the registry: (1) nature and
extent of injuries; (2) findings on each element of monetary loss, including losses
in wages; medical, custodial, and special education costs, both past and pro-
jected, with the discount rate used and an explicit account of future inflation
underlying the calculation of future loss; (3) identification of the types of non-
monetary losses and compensation levels for each; and (4) adjustments made for
comparative negligence, prior settlements by other defendants, collateral sources,
joint and several liability, and/or other factors.
55. This argument have been made by others as well. See, e.g., Studdert, Yang,
and Mello (2004).
56. Scheduled damages have been advocated by the following scholars, among
others: Bovbjerg, Sloan, and Blumstein (1989); Corrigan, Greiner, and Erickson
(2002); Mello and Brennan (2002).
57. See, e.g., Viscusi (1993).
58. European tables may provide some guidance, but they should not be adopted
uncritically.
59. Gold et al. (1996).
60. Bovbjerg, Sloan, and Blumstein (1989); Sloan and Hsieh (1990).
61. These values would be indexed by the Consumer Price Index.
62. Juries might indicate that the case involved less pain and suffering than one
scenario but more than another. Then the pain and suffering might be reduced
or increased to the midpoint of the dollar values associated with each of the
scenarios in the notebook.
63. See, e.g., Schuck (1991). The premise that jury awards for nonmonetary
loss are seriously flawed is not a generally accepted proposition. For example,
Cohen and Miller (2003) examine juries’ “willingness to award” nonmonetary
damages in over 1,200 cases of consumer product-related injuries and inten-
tional assaults by a third party. Based on data on awards for injuries with
out-of-pocket losses sustained by the plaintiff, the authors computed implicit
values of a statistical life from jury awards for nonfatal injuries. They found
that jury awards are predictable (using regression analysis), but there is a
high degree of variability. There is also substantial variability in the circum-
stances leading to the awards. The implied values of a statistical life from
the jury award data range from $1.4 to $3.8 million, well within the range
independently derived from wage-risk studies (see e.g., Viscusi 1993). In spite of
their results, the authors state that they favor scheduling awards for nonmone-
tary loss.
356 Notes to pages 123–138

64. He suggests that juries assess damages from an ex ante perspective that asks
how much a reasonable person would have paid to eliminate the risk that caused
the nonmonetary loss. Such willingness to pay would reflect both the probability
of the loss and the value of the loss, given that a loss occurred. He also proposes
that juries be asked to assume a specific probability that the injury would occur.
Then, based on their responses, one could assign values to the loss, given that it
occurred.
65. Alberini, Hunt, and Markandya (2006); Perreira and Sloan (2002); Sloan,
Viscusi, Chesson, et al. (1998); Viscusi and Evans (1990).
66. There are several possible ways to deal with this problem. One is to use a
Delphi technique. Jurors would get feedback on the mean responses in the first
round as well as the high and low values. They then would be asked to repeat
the valuation exercise. Perhaps by the second or third round, there would be
some convergence in responses. Alternatively, the court could employ a panel of
citizens who would be asked to perform the valuation task. Since the valuations
would have to be done in person, this could be an expensive exercise. On the
other hand, the cost would be low relative to the awards being contemplated in
some cases.
67. Avraham (2006, p. 103).
68. Blumstein, Bovbjerg, and Sloan (1991).
69. Farber and Bazerman (1986); Stern, Rehmus, Lowenberg (1975).
70. Würdemann Vanderbilt (1912, p. 884).

Chapter 6

1. See, e.g. Kritzer (1997).

2. Kritzer (2002) presents data on litigation rates by country which place the
United States fifth after Germany, Sweden, Israel, and Austria. The issues for
which the lawsuits are brought vary among countries.
3. Brickman (2003, p. 658).
4. Lesser et al. (2004). Lester Brickman (2003, p. 657) argues that by “pursuing
anticompetitive strategies including erecting barriers to competition from outside
the profession and promulgating ethical rules restricting price competition within
the profession, contingent-fee lawyers have not only flouted ethical rules and
fiducial protections but have also imposed substantial rents on tort claimants as
a price for tort claiming.” A deficiency of this criticism is that to the extent it
applies to contingent-fee lawyers, it applies to lawyers in general.
5. See, e.g., Olson (1991, p. 45).
6. See Tabarrok and Helland (2005, p. 11).
7. Daniels and Martin (2006).
8. Formal Opinion 94–389, Contingent Fees, December 5, 1994.
 Notes to pages 141–144 357

9. W. K. Davis (1999).
10. Taylor v. Bemiss, 110 U.S. 42 (1884).
11. W. K. Davis (1999).
12. Plaintiffs are risk-averse and attorneys are risk-neutral. (Danzon 1983).
Whatever law firms’ risk preferences, members of a firm can diversify away some
risk by taking on many cases (Garoupa and Gomez-Pomar 2002).
13. Inselbuch (2001).
14. Dana and Spier (1993); Danzon (1983); Hay (1996).
15. Hay (1997).
16. Some have argued that contingent fee arrangements are more appropriate
when there is asymmetric information between attorneys and their clients, the
former knowing more about their own ability, and clients, at least initially,
knowing more about the merits of the case (Dana and Spier 1993; Rubinfeld
and Scotchmer 1993). When there is asymmetric information, contingent fees
can allow clients to signal the quality of their cases and attorneys to signal the
quality of their advice. A client who thinks he or she has a high-quality case
should be willing to pay a high hourly fee and a low contingent fee percentage,
while an attorney would signal high quality by offering services at a relatively
high contingent fee percentage. A problem with the use of contingent fees as a
signaling device is that variation in such fee percentages is quite small—mostly
in the range of 33 to 40 percent (Sloan, Githens, Clayton, et al. 1993).
17. These are called “agency problems” in economics.
18. A theoretical paper by Choi (2003) argues that leaving the lawyer in charge
of the litigation and guaranteeing him or her a large amount of the rent can
increase the plaintiff’s return from a settlement.
19. In a theoretical study, Emons (2000) shows that fixed contingent fees may
lead to insufficient attorney effort, and paying the attorney by the task performed
always implements efficient behavior.
20. L. A. Baker (2001–2002).
21. Sieg (2000).
22. Miceli (1994).
23. L. A. Baker (2001–2002).
24. Dana and Spier (1993).
25. Brickman (2003, p. 655).
26. We have updated his estimates to 2004 dollars.
27. Kritzer (2004, p. 188).
28. Responses to the survey also indicated that tort reform public relations
campaigns and decisions of the Texas Supreme Court, rather than the statutory
changes themselves, were perceived as having the greatest negative impact on the
lawyers’ practices. Public relations campaigns were seen as influencing juror
358 Notes to pages 144–146

attitudes; following campaigns, there was a widespread perception among jurors

that the system was out of control. Not only did public relations campaigns,
judicial decisions, and legislative changes affect attitudes, but there was also a
decline in automobile liability caseload.
29. Kakalik, Hensler, McCaffrey, et al. (1998).
30. The same Wisconsin survey revealed that most lawyers in the study, 60
percent, charged a fixed 33 percent contingent fee; 31 percent employed a vari-
able fee. For variable fees, the most common pattern was a fee of 25 percent if
there was no substantial trial preparation, rising to 40 percent if the case resulted
in an appeal. There were sometimes reductions in fees below the rate specified
in the retention agreement. Kritzer concludes that “At least in Wisconsin, the
assertion by contingency fee critics that there is a uniform contingency fee is
clearly false” (2004, p. 39). But, as Brickman notes, relative to many, if not most,
states, Wisconsin is not a litigious state; as a result, data from this state may
be unrepresentative of contingent fee practices in more litigious states (2003,
p. 685).
31. Kritzer (1997).
32. Kritzer (2004, pp. 68, 71).
33. Brickman (2003, p. 697).
34. Ibid. (pp. 698, 734). Brickman cites Burnett v. Commissioner as evidence
of this. In Burnett, the plaintiff taxpayer, an attorney, attempted to deduct
advances made to his clients which were only to be repaid only if there was
recovery for the client. In this case, the court stated: “Here there was a close
correlation between the conditions of reimbursement and the primary criterion
employed by petitioner to select clients to receive financial assistance, e.g.,
although reimbursement was tied to the recovery of a client’s claim, assistance
was granted only to those whose claims would in all probability be successfully
concluded. Moreover, petitioner experienced a high rate of return on the
advances, evidenced by the fact that during the five-year period ending in 1961
in which petitioner’s conditional advances were made, only $4,417 out of
approximately $290,000 of them became worthless. These factors clearly indi-
cate that the advances were intended to, and did, operate as loans, i.e., advances
virtually certain of repayment, to petitioner’s clients” 356 F.2d 755, 760 (5th
Cir. 1966).
35. Kritzer (2004, p. 39).
36. Brickman (2003, p. 685).
37. Hawaii Rev. Stat. §607-15.5 (1986).
38. Pa. Stat. Ann. Tit. 40 §1301.604
39. Heller v. Frankston, 475 A.2d 1291, 1296 (Pa. 1984).
40. Or. Rev. Stat. §752.150 (1975).
41. Or. Rev. Stat. §18.540 (1987).
 Notes to pages 146–149 359

42. Idaho Code §39-4213

43. N.H. Rev. Stat. Ann. §507–C:8
44. Carson v. Maurer, 424 A.2d 825, 839 (N.H. 1980).
45. Ibid.
46. Fla. Stat. §768.56 (1980).
47. Florida Constitution, Art. I, §26 (2005).
48. Ibid. This amount is exclusive of reasonable and customary costs and does
not take into account the number of defendants.
49. In re Amendments to the Rules Regulating the Fla. Bar, 933 So.2d 417
(Fla. 2006).
50. T. Baker (2005b, p. 185).
51. Charles Silver states the perceived problem: “Unsophisticated laypersons
cannot shop for legal services intelligently . . .” (2002, p. 2088).
52. California (Cal. Bus. & Prof. Code §6146); Connecticut (Conn. Gen. Stat.
§52–251c); Delaware (Del. Code Ann. Tit. 18 §6865); Florida (Florida Code of
Professional Responsibility, Disciplinary Rule 2-106 and Fla. Stat. §768.595);
Illinois (Ill. Rev. Stat. ch. 110, §2–1114); Maine (Me. Rev. Stat. Ann. Tit. 24,
§2961); Massachusetts (Mass. Gen. Laws Ann. ch. 231, §60I); New Jersey
(New Jersey Court Rules §1:21-7); New York (N.Y. Jud. Law §474a);
Wisconsin (Wis. Stat. §655.013); Wyoming (Ct. Rules Governing Contingent
Fees R.5).
53. Arizona (Ariz. Rev. Stat. Ann. §12-568); Hawaii (Hawaii Rev. Stat. §607-
15.5); Iowa (Iowa Code §147.138; Kansas (Kan. Stat. Ann. §7-121b); Maryland
(Md. Cts. & Jud. Proc. §3-2A-07); Nebraska (Neb. Rev. Stat. §44-2834); New
Hampshire (N.H. Rev. Stat. Ann. §508:4-e); Tennessee (Tenn. Code Ann. §29-
26-120); Washington (Wash. Rev. Code §7.70.070); Wyoming (Ct. Rules Gov-
erning Contingent Fees, R.6).
54. Florida (Fla. Stat. §§766.207, 766.209); Indiana (Ind. Code §16-9.5-5-1);
Massachusetts (Mass. Gen. Laws Ann. ch. 231, §60I); Michigan (Michigan
Court Rules, Rule 8.121(B)); Oklahoma (Okla. Stat. Tit. 5, §7); Tennessee
(Tenn. Code Ann. §29-26-120); Utah (Utah Code Ann. §78-14-7.5).
55. While both court review and sliding-scale statutes have been enacted in ten
states, some of the statutes under the court review category are not mandatory;
rather, it is at the discretion of the court or upon the request of the parties to
review attorney fees.
56. For particulars, see chapter 11.
57. These data are from the 1992 Civil Justice Survey of State Courts.
58. These data were obtained from closed medical malpractice insurance claims
from the Department of Insurance for the State of Florida.
59. Helland and Tabarrok (2003, p. 54).
60. Sloan and Hoerger (1991).
360 Notes to pages 149–156

61. Di Pietro and Carns (1996); Stanley (2003); Vargo (1993). The description
here does not do full justice to the historical facts surrounding this issue. There
were many other factors in play, such as animosity toward lawyers, heavy restric-
tions on attorneys’ fees if they could collect them, and the use of attorneys during
this time period.
62. Vargo (1993).
63. Canter v. American Ins. Co., 28 U.S. 307 (1830).
64. Vargo (1993).
65. Di Pietro and Carns (1996).
66. Alaska R. Civ. Proc. 82.
67. Ibid.
68. Di Pietro and Carns (1996).
69. Snyder and Hughes (1990); J. W. Hughes and Snyder (1995).
70. J. W. Hughes and Snyder (1995).
71. Ibid.
72. Di Pietro and Carns (1996).
73. Ibid.; Stanley (2003).
74. O’Connell and Bryan (2000–2001).
75. Ibid.
76. Copland (2004).
77. Sloan, Githens, Clayton, et al. (1993).
78. Ibid.
79. In re: Petition for Rulemaking to Revise the Ethical Standards Relating to
Contingency Fees, Utah Supreme Court Advisory Committee on Rules of Profes-
sional Conduct (2004, p. 76).
80. The origins of legal insurance have been traced back to ancient Rome.
However, this lengthy history and development are extraneous to this discussion.
Instead, we list the years in which specific legal insurance policies were created
or offered. The following countries have had LEI since the year listed: Austria,
1955; Greece, 1971; Great Britain, 1975; Luxembourg, 1969; Sweden, Norway,
and Denmark, 1960s (Pfennigstorf 1986).
81. This capture of the market is due to uninsured loss recovery, which forms
the most significant part of the market. Uninsured loss recovery is the pursuit,
after a car accident, of the uninsured losses for which the insured is involved but
not at fault. (Kilian 2003).
82. Ibid.
83. Ibid.
84. Ibid.
85. Ibid.
 Notes to pages 156–164 361

86. Ibid.
87. Ibid.
88. Ibid.
89. Maute (2001).
90. Costich (1994).
91. 377 U.S. 1 (1964). The brotherhood’s activities also met resistance in Illinois
(1932), Ohio (1933), California (1941), New York (1933), Tennessee (1952),
and Missouri (1960). See also Riedmueller (1973–1974).
92. Maute (2001).
93. This is not to suggest that there were not many other criticisms and points
of opposition; rather, this was one area where a lot of debate and reform
centered.
94. Heid and Misulovin (2000).
95. Pfenningstorf (1975).
96. Ibid.
97. Kilian (2003).
98. Kritzer (2004, p. 254).
99. See Lerman (1999).
100. Haltom and McCann (2004, p. 136).

Chapter 7

1. See, e.g., Developments in the Law: Confronting the New Challenges of

Scientific Evidence (1995); and Helland and Tabarrok (2000).
2. Dann (2002).
3. These criticisms of juries are found in many articles. See, for example, Vidmar
(1994a).
4. But not Vidmar (1994b), whose work supports juries. He mentions that
others had alleged race and gender bias.
5. Procedures differ across states in terms of the presentation of arguments
involving punitive damages. In some states, there are separate trials for hearing
evidence and deciding on punitive damages. In others, this is done at the same
trial at which liability and compensatory damages are determined (Priest 2002,
p. 10).
6. Ellsworth and Reifman (2000).
7. Vidmar and Rose (2001); Viscusi and Born (2005).
8. Zimmerman and Oster (2002, p. A1).
9. MacCoun (1993, p. 138).
362 Notes to pages 164–171

10. Vidmar, Gross, and Rose (1998).

11. Moller, Pace, and Carroll (1999).
12. Grisham (1996).
13. Vidmar (1994a).
14. Cohen and Smith (2004).
15. Vidmar (1994a).
16. 174 Cal. App. 3d 831 (1985).
17. Sugarman (1990).
18. Taragin, Willett, Wilczek, et al. (1992).
19. Ibid. (p. 780).
20. Farber and White (1991).
21. Sloan, Githens, Clayton, et al. (1993).
22. Studdert, Mello, Gawande, et al. (2006).
23. See chapter 5 for a description.
24. The authors acknowledge in their discussion of limitations that not blinding
the physician reviewers as to the case outcomes may have caused hindsight bias.
They argue that if hindsight bias existed, it would have led the reviewers to
overestimate the prevalence and costs of claims not associated with medical error.
But more important, hindsight bias could lead one to reject the myth that claims
outcomes are random. However, reviewers were blinded as to outcome in the
third study discussed in the text. Although blinding would have been preferable,
it seems doubtful that the results would be overturned if reviewers had not
known whether or not payment was in fact made.
25. Studdert, Mello, and Brennan (2004)l; Weiler (1993); Weiler, Hiatt,
Newhouse, et al. (1993).
26. Farber and White (1991); Hoerger, Sloan, and Hassan (1990).
27. Lichtenstein, Slovic, Fischhoff, et al. (1978); Morgan, Granger, Brodsky,
et al. (1983); Viscusi (2002b, pp. 182–183).
28. Tversky and Kahneman (1973). One of the issues in mock jury research is
that individuals seem to be sensitive to the magnitude of the plaintiff’s initial
demand for compensation (sometimes called the “ask”) or for demands by the
plantiff’s attorney in closing argument at trial. However, as Vidmar (2004a)
noted, such experiments are unrealistic in that jurors would be exposed to coun-
terarguments in the defendant’s attorney’s closing statement. Presumably the
argument would be that compensation should be much lower, even if the jury
found liability.
29. Sunstein (2003); Galanter (1993, p. 84).
30. MacCoun (1993, p. 162).
31. Lempert (1993, p. 192).
 Notes to pages 171–175 363

32. A common technique in empirical research on juries is to obtain responses

to hypothetical questions from persons who are asked to assume that they are
jurors (“mock jurors”).
33. The framing of questions is the subject of a literature in its own right, as is
whether or not the questions asked of mock jurors are those that jurors would
be asked to answer in the real world. For example, Vidmar (1999) criticizes prior
research comparing judge and jury decision-making on grounds that the experi-
ments ask jurors to make decisions about law—decisions that are exclusively the
responsibility of the judge.
34. Lempert (1999).
35. Viscusi (2002b).
36. Vidmar and Rice (1993).
37. Vidmar (1993).
38. Sunstein, Schkade, and Kahneman (2000).
39. Sunstein, Schkade, and Kahneman (2002, pp. 132, 141).
40. Viscusi (2002a, p. 207; 2001).
41. Sunstein (2002a, p. 257).
42. Vidmar (2004).
43. Sloan, Githens, Clayton, et al. (1993).
44. Polinsky and Shavell (1998, p. 870).
45. Rustad (1992).
46. Moller, Pace, and Carroll (1999) analyze jury verdicts reached in 1985
through 1994 from California; New York; Cook County, Illinois; Harris County,
Texas; and the St. Louis metropolitan area. Out of 5,238 verdicts, punitive
damages were awarded in 28 of these cases (0.5 percent of total verdicts or 1.7
percent of plaintiff verdicts). In a sample of 1,156 medical malpractice verdicts
in the seventy-five largest counties, punitive damages were awarded in only seven
cases (0.6 percent of total medical malpractice verdicts). Medical malpractice
cases were 9.7 percent of all tort, contract, and real property cases in the sample
(Cohen and Smith 2004). Awards for punitive damages were 9.5 percent of total
awards for punitive damages in all of the cases. In Vidmar (2004, p. 1367), 12
out of 270 jury verdicts that were examined were medical malpractice cases. The
median punitive damage award was $1 million in 1999 dollars; the median
compensatory award was 2.5 times higher. Moreover, relative to the other case
types, compensatory awards were substantially higher relative to punitive damage
awards in medical malpractice.
47. Sunstein (2002a, p. 75).
48. Priest (2002, pp. 2–4).
49. Tabarrok and Helland (1999).
50. Tabarrok and Helland (1999). Eisenberg, LaFountain, Ostrom et al. (2002)
report that controlling for other factors, there were no differences between judges
364 Notes to pages 175–177

and juries in the rate at which punitive damages were awarded and in the
amounts, conditional on a punitive damage award being awarded, but there was
greater variability in the magnitude of punitive awards in jury trials than in judge
trials. However, the greater range of punitive damage awards in jury trials
yielded very few awards that were higher than what judges might have awarded
in similar cases. Without controlling for other factors, punitive damage awards
were far higher when awarded by juries than by judges, which is consistent with
the conventional wisdom and stories in the media that do not account for other
factors.
51. Proponents suggest that the establishment of uniform and consistent stan-
dards, as well as improved communication, will facilitate patient safety, although
it is unclear how effective this would be in practice.
52. At the time this book was written in 2006, despite concerns, it was the
reform of choice for many legislatures, including Congress. Proposals in the
U.S. House and Senate authorize the Secretary of Health and Human Services
(HHS) to award demonstration grants to the states for the development,
implementation, and evaluation of health courts. In addition to existing
proposals, Senator John Cornyn, a Republican from Texas, advocated
health court pilot projects which would shift resolution of claims from state
courts to administrative courts maintained by the HHS (Kyl 2006). Common
Good was working in partnership with the Harvard School of Public Health on
a model “compatible with the U.S. health care and legal environment” (Saltus
2005).
53. In the AMA proposal, a patient has to take his or her claim before
a reviewer, two sets of physician peer reviewers, and finally an administra-
tive law judge called a hearing examiner. After all of this, the parties may still
appeal to the highest court in the state. See (American Medical Association
1988).
54. The AMA proposal, introduced in 1988, contains several features absent
from the more recent Common Good proposal. The AMA’s model is multitiered,
using two levels of independent review before reaching judicial review. First, a
claim is assessed in a prehearing stage. During this stage, reviewers evaluate the
merit of the claim, using testimony and medical records. In the absence of an
early settlement offer, the reviewer must determine if the claim is nonmeritorious
and, if so, recommend dismissal. If the claim is determined to be meritorious, it
is further evaluated at the next level by an expert in the same field as the health
care provider. If the expert agrees that the claim has merit, the case is sent to a
hearings examiner. At this stage, should the claim be determined to have no
merit, it is given to a second expert for review.
For cases passing the prehearing and expert reviews, the claimant appears
before the hearings examiner in a triallike hearing, with both sides represented
by counsel. The examiner serves a judicial function akin to an administrative
law judge. The AMA proposal gives the claimant an appointed staff attorney,
but allows for private legal representation at the claimant’s option. When the
 Notes to page 177 365

hearing process begins, blind settlement offers are required from both parties.
The examiner may call witnesses of his or her choosing and must issue a written
opinion within ninety days after conclusion of the hearing. In this written
opinion, the hearing examiner is not only responsible for determining the defen-
dant’s liability, but also must assess damages.
Decisions from the hearing examiner can be appealed to the Medical Board,
which is responsible for making a full, independent determination regarding the
provider’s liability. The Medical Board will be made up of three members, all of
whom are selected by the governor upon advice of a nominating committee. The
state’s intermediate appellate court hears appeals from the Medical Board’s deci-
sions. The board would oversee the administrative process, review appeals, and
oversee continuing medical education. The AMA hoped that the board would
eventually restore consistency in damage awards, resolve claims more quickly,
and develop rules and substantive guidelines to complement existing statutory
standards.
In spite of the many hoops claimants must jump through before their claim
reaches judgment—which even many neutral observers would regard as a nega-
tive feature—the AMA proposal contains a few novel concepts. First, the plan
requires blind offers during the prehearing stage. If the submitted offers are
incompatible (e.g., the plaintiff’s offer is higher than the defendant’s), the offers
are rejected, and the claim continues with the administrative judge. If the defen-
dant’s offer is greater than or equal to the claimant’s demand, the case is settled
at the claimant’s amount, with payment required within thirty days. To discour-
age bad-faith offers, a party may face sanctions should the outcome of the
hearing not be an improvement over the blind offer.
A very intriguing aspect of the proposal is the provision of free legal repre-
sentation to claimants. Current proposals assume health courts will be navigable
for the injured party without an attorney. Or, alternatively, injured patients can
readily obtain counsel should they need it. However, it is questionable whether
attorneys would be willing to accept cases when the damages are scheduled or
when there is a 20 percent cap on contingency fees. Attorneys who regularly
charge contingency fees of 33–40 percent will not take cases that have a probabil-
ity of a low damage award, even if some recovery is likely. Appointing a public
defender-type lawyer for injured patients solves this problem, but it also shifts
the costs from the litigants to taxpayers. However, as discussed below, both tax
courts and family courts are accessible to litigants who represent themselves
(“pro se” litigants), and they make up a large number of the plaintiffs appearing
before these courts.
Investigation of medical records, interviews with providers, and consultations
with and the hiring of experts are all costs currently borne by the plaintiff and
defendant, but would be shifted to the state. To the extent that the new system
increased claims frequency, this would add to cost as well. On the other hand,
the dispute resolution process could be much shorter than the current tort system,
with many medical malpractice claims being resolved in months rather than in
several years, which could reduce cost appreciably.
366 Notes to pages 178–181

55. Medical no-fault is discussed more in chapter 11. Very few no-fault systems
operating in other countries are truly “no-fault”; rather, they use modified ver-
sions of liability—such as avoidable events.
56. Udell and Kendall (2005).
57. United States Constitution, Art. I, Sec. 8, Clause 3. Under principles of state
sovereignty and federalism, the constitutionality of federal regulation is question-
able. The constitution grants the federal government certain enumerated powers,
such as the right to regulate interstate commerce and the power to pass any law
that is necessary and proper for the execution of its enumerated powers. Powers
that the constitution does not grant to the federal government or forbid to the
states are reserved exclusively to the states. Should the U.S. Congress enact a law
creating and regulating health courts, it is uncertain whether it would be upheld;
the Supreme Court did not invalidate any federal statute as exceeding the states’
enumerated powers from 1938 until 1995. These powers historically have been
interpreted liberally but, in recent years, reviewed with a more critical eye.
58. Kyl (2006). While the Supreme Court has incorporated most of the Bill of
Rights to the states, they have not incorporated the Seventh Amendment. This
means that the constitutional protection of jury trials in civil cases applies only
to federal courts.
59. Chapter 1’s discussion of myth 1.
60. See Danzon (1986).
61. McCoid (1991).
62. P. Davis, Lay-Yee, Scott, et al. (2003).
63. In 2000 President Clinton signed a bill authorizing $10 million a year for a
period of four years to support the creation of up to 100 mental health courts.
(H. J. Stedman, Davidson, and Brown 2001).
64. The juvenile court system in the U.S. began in 1899 in Chicago, and family
courts were first developed in 1914, in response to a perceived societal need for
improved court performance in family matters (Babb 1998).
65. For example, a recent study showed that 40 percent of families that appear
in court for child abuse, neglect, delinquency, or divorce have been to court
previously for another family-related matter (Developments in the Law 2003).
Keeping family law issues in the general court system was seen as wasteful
because of duplicative judicial attention to the same family in different courts by
different judges.
66. The act was sponsored by the U.S. Children’s Bureau in collaboration with
the National Probation and Parole Association and the National Council of
Juvenile Court Judges.
67. Babb (1998, p. 480).
68. Permissive family court systems give litigants the choice of forum. They may
file in the family court or in the general court, giving at least two sets of courts
jurisdiction over family law claims. Mandatory systems shift all family law cases
 Notes to pages 181–182 367

from the general court to the specialized family court. Only the family court has
jurisdiction, and no other courts are allowed to hear cases dealing with domestic
relations. Currently, only a minority of states continue to process family law
cases as part of the general civil docket (Babb 1998). And none of the states that
have implemented family courts have returned to adjudicating family law cases
within the general docket (Kosanovich 2006).
69. Fields (2006).
70. Rottman (2000). For example, a judge may determine in an abuse case that
the father had sexually abused his daughter, while a second judge during divorce
proceedings excludes evidence of the sexual abuse and grants the father visitation
rights (Williams 1995). In contrast, family courts aim to have centralized case
files and rulings, giving judges and social services access to all relevant informa-
tion about the parties before them. In the context of medical malpractice, some
physicians are sued repeatedly. Consolidating these cases and coordination with
other quality-assurance systems, such as state licensure boards and health insur-
ers, may be more efficient.
71. Forum shopping occurs when the plaintiff searches among jurisdictions or
courts to identify the most favorable outcome for plaintiffs due to differences in
judges, laws, rules of evidence, and rules of the court, among other things.
72. A U.S. Department of Justice study reports that as of 1995, domestic issues
accounted for 40 percent of all civil filings, while tort cases accounted for only
10 percent (S.K. Smith, DeFrances, Langan, et al. 1995). Of the tort cases,
medical malpractice cases made up 5 percent (0.5 percent of total filings). In
addition to making up such a large part of the docket, domestic filings were not
stagnant; between 1984 and 2000, domestic relations filings increased 79 percent
(Developments in the Law 2003). On top of the dramatic increase in family law
cases, criminal dockets were getting bigger, discovery in civil cases was mounting,
and commercial cases were becoming more complex. These factors created an
atmosphere where judges were increasingly unable to devote sufficient time to
the nuances of family law (Folberg 1999). Attorneys and some pro-se litigants
took advantage of the court’s disorganization by judge shopping and repetition
of judicial efforts (Folberg 1999).
73. For example, in 2002 the District of Columbia created a unified family court
(Geraghty and Myniec 2002). Prior to this, there was a family division in place
within the court of general jurisdiction. The legislation created a one judge and
one family system separate from the general court, changed the judicial rotation
from one year to three, and added new judges and judge-magistrates (Geraghty
and Myniec 2002). The legislature appropriated $18 million for the transition,
though court officials estimated it would take $46 million to hire judges and
staff and to build courtrooms (Leonning 2001).
74. Kondo (2001).
75. Domitrovich (1998).
76. Ibid.
368 Notes to pages 182–187

77. Judges may burn out quickly through having to deal with the same difficult,
emotionally charged family cases every day. This can be detrimental not only for
a judge, but also for the parties before him. See Geraghty and Myniec (2002)
for a brief discussion of this issue.
78. Babb (1998); Geraghty and Myniec (2002).
79. Babb (1998).
80. Schepard (2002).
81. Judges are responsible for adjudicating custody disputes, child support,
divorces, domestic violence charges, juvenile delinquency, all phases of abuse,
neglect, and dependency cases, adoption, and guardianship (Kosanovich 2006).
82. See U.S. Constitution, Art. I, Sec. 8, Clause 4.
83. Laro (1995). In addition, over 95 percent of tax cases are brought before
the tax court.
84. See Laro (1995) for a discussion of the arguments for and against govern-
ment bias in tax courts. Proponents suggest there are multiple factors that
account for this large percentage, none of which include government bias.
85. The trend toward specialization is not limited to courts. For instance, there
are few “general practice” physicians or attorneys left; rapid and complex devel-
opments in both medicine and law make it difficult for professionals to remain
versed and competent in many areas of practice. Specialty courts are a reflection
of this trend.
86. See chapter 11.
87. Sage (1997).
88. The average caseload for a New York Family Court judge is 2,500 cases
(Schepard 2002).
89. “In Suits at common law, where the value in controversy shall exceed twenty
dollars, the right of trial by jury shall be preserved, and no fact tried by a jury
shall be otherwise re-examined in any Court of the United States, than according
to the rules of the common law” (Constitution of the United States, Amendment
7).
90. Leonning (2001).
91. Federal Rule of Evidence 702 codified and superseded Daubert v. Merrell
Dow Pharms., Inc., 113 S. Ct. 2786 (1993). However, Daubert’s standard of
review is the basis for determining the admissibility of an expert witness’s testi-
mony and continues to be used at both the federal and state levels (See, e.g.,
United States v. Parra, 402 F.3d 752, 758 (7th Cir. 2005); United States v.
Mahone, 453 F.3d 68 (1st Cir. 2006); United States v. Frabizio, 2006 U.S. Dist.
LEXIS 56327 (D. Mass. 2006); Alves v. Mazda Motor of Am., Inc., 2006 U.S.
Dist. LEXIS 65465 (D. Mass. 2006); Palandjian v. Foster, 446 Mass. 100 (Mass.
2006); State v. Abner, 2006 Ohio 4510 (Ohio Ct. App. 2006). Federal Rule of
Evidence 702 states, “If scientific, technical, or other specialized knowledge will
 Notes to pages 187–194 369

assist the trier of fact to understand the evidence or to determine a fact in issue,
a witness qualified as an expert by knowledge, skill, experience, training, or
education, may testify thereto in the form of an opinion or otherwise, if (1) the
testimony is based upon sufficient facts or data, (2) the testimony is the product
of reliable principles and methods, and (3) the witness has applied the principles
and methods reliably to the facts of the case.”

Chapter 8

1. See chapter 3 in particular for a discussion of defensive medicine.

2. Hyman and Silver (2005).
3. Ibid.
4. Sloan Mergenhagen, Burfield, et al. (1989).
5. Brennan and Mello (2003).
6. Hyman and Silver (2005).
7. Ibid.
8. United States General Accounting Office (1987).
9. See, e.g., Breyer (1982) for a discussion of this rationale for health and safety
regulation.
10. Those goods that can be evaluated in terms of their prices and characteristics
in advance of the purchase are called “search goods.” “Experience goods” are
those for which the characteristics can be adequately evaluated only some period
after consumption is commenced. Physician visits for routine care are plausibly
experience goods. Visits to the emergency room are more likely to be credence
goods.
11. In general, such differences in the risk of adverse outcomes (e.g., mechanical
failures attributable to how vehicles are used) are overlooked in making com-
parisons of repair frequency; however, insurers do charge different premiums for
automobile insurance based on a particular vehicle’s history of frequency of
accidents. No insurer, to our knowledge, bases health insurance premiums on
the frequency of medical errors at particular facilities or of particular
physicians.
12. Making a more reliable product may require more redundancy at various
stages of the production process or more durable materials. These additional
costs will be incurred by a profit-seeking enterprise if they yield additional
revenue at least equal to such extra cost.
13. Manufacturers are also subject to product regulation, which specifies
minimum levels of safety equipment and requires that some characteristics, such
as fuel consumption, be prominently displayed on the automobile at the time of
initial sale. There is also a role for tort liability, as seen recently in the case of
the Ford Explorers, which tended to flip over at a seemingly unacceptable rate,
370 Notes to pages 194–200

a problem that has been remedied in later models, presumably in part to avoid
being sued. Even so, no one assigns the principal role of automobile quality
assurance to tort.
14. A cottage industry consists of many small firms.
15. Mello and Brennan (2002).
16. The methodology is described in Brennan, Leape, Laird, et al. (1991).
17. Leape, Brennan, Laird, et al. (1991).
18. Mello and Brennan (2002).
19. A case in point is the IOM report that followed To Err Is Human, titled
The Quality Chasm. At the 2001 press conference at which Quality Chasm was
released, IOM committee members emphasized that their inquiry into patient
safety had sidestepped medical malpractice issues (http://www.iom.edu/
CMS/8089/5432.aspx, accessed July 3, 2006).
20. A national system for reporting medication errors or interactions, the Medi-
cation Errors Reporting Program (MER), created by the Institute for Safe Medi-
cation Practice, had been in place since 1991. Though it enjoyed some success,
the drug error information submitted to the MER program is not fully protected;
information may be shared with the FDA and the pharmaceutical companies
mentioned in the reports. This lack of confidentiality has significantly inhibited
the use of MER. As a result, the system was largely replaced in 1998 by a new
anonymous Internet reporting system, MedMARx. Hospitals can subscribe to
the program, and providers can report errors through the MedMARx Web site
with the assurance that the information will remain confidential. Information
from the MedMARx program is not reported to the FDA, and hospitals may
obtain data for their facilities and comparative information from other facilities,
but the identities are not revealed. Anonymity has proven efficacious; according
to the U.S. Pharmacopeia Web site, between 1999 and 2003 nearly 600,000 drug
errors were reported to MedMARx (http://www.usp.org/products/medMarx,
accessed June 15, 2006).
21. Err encourages the U.S. Food and Drug Administration to increase its
efforts in monitoring safe use of drugs, both pre and post marketing, by imple-
menting higher standards in drug packaging and labeling, requiring pharmaceuti-
cal companies to test potential drug names for confusion with similarly named
drugs, and to work directly with physicians, pharmacists, and consumers to
establish responses to identified problems. The report notes that during 1997
and 1998, five drugs were removed from the market—but not before almost 20
million people had been exposed to their risks. Some of these drugs were removed
by the pharmaceutical companies themselves following reports of adverse out-
comes. Err hopes to prevent this from reoccurring with more vigorous
monitoring.
22. A survey conducted after release of Err finds that the mean hospital’s annual
patient safety budget is $1.9 million, with a range from $50,000 to $15 million
(Devers, Pham, and Liu 2004).
 Notes to pages 200–201 371

23. In the United States, the only national administrative data are Medicare
claims data, which predominantly reflect care delivered to persons over the age
of sixty-five. In a country with universal health insurance, entire populations are
covered. For example, in the Canadian province of Manitoba, claims data are
being used for monitoring negligent adverse events (Bruce, Prior, Katz, et al.
2006).
24. In July 2003, the Accreditation Committee on Graduate Medical Education
(ACGME) fully implemented modest regulations limiting resident work hours to
a maximum of eighty per week or thirty hours per shift, in part a response to
studies showing a link between resident physician fatigue and poor performance
(D. Weinstein 2002). Despite the ACGME’s good intentions, implementation
and regulation of this policy have been difficult for many hospitals, and the
ACGME’s resources for investigating violations are limited. The fundamental
problem with regulation of hours is that it is not easy for residents to report
violations. Many fear the repercussions of being the whistle-blower or, even
worse, causing their residency program to lose its accreditation. As a result, hour
violations are difficult for program directors to detect. However, resident hours
and fatigue are not the only factors contributing to resident errors. Surveys of
residents have revealed other factors that may adversely affect patient safety,
including inadequate supervision, problems with handoffs, lack of knowledge,
too many other tasks, and a large patient load (Jagsi, Kitch, Weinstein, et al.
2005; Wu, Folkman, McPhee, et al. 2003).
25. Public goods have the property that individual A’s consumption does not
reduce individual B’s consumption of the good. In this context, hospital A’s
use of a report on errors does not diminish hospital B’s ability to consume
this information. By contrast, for a private good such as a peanut butter
sandwich, if individual A eats it, the same sandwich cannot be eaten by individual
B.
26. The Joint Commission on Accreditation of Healthcare Organizations
(JCAHO) and the National Quality Forum (NQF) have developed a system to
collect and classify data from all existing systems (Wood and Nash 2005). With
the intent of making national comparisons of adverse event reports, the JCAHO
and NQF plan to standardize data collected from state reporting systems using
a “Patient Safety Event Taxonomy” system. The hope is that by providing an
amalgamation of information, hospitals can learn from the mistakes of their
peers nationally, instead of relying on data restricted to their state or individual
hospital (Wood and Nash 2005).
27. For example, a mixed-up X-ray may have been discovered before an adverse
outcome actually occurred, but this should be seen as a fortuitous circumstance.
Pennsylvania, followed by Florida and Maryland, corrected this problem in 2002
by adding near-miss events to the definition of adverse event (Wood and Nash
2005). As of 2004, only twenty-four states had implemented mandatory report-
ing systems (Weinberg, Hilborne, and Ngyuen, 2005).
28. Marchev, Rosenthal, and Booth (2003).
372 Notes to pages 202–203

29. Flowers and Riley (2001).

30. Institute of Medicine (2000); United States Census Bureau (2000).
31. Wright and Katz (2005).
32. The cost of compliance for the U.S. Food and Drug Administration, drug
wholesalers, and manufacturers was estimated to be $5.1 million, and $680
million for hospitals (Warburton 2005). Advocates of the system estimate savings
of $3.9 billion annually from reductions in hospitalization days attributable to
medication errors (Warburton 2005), but these savings do not accrue to hospitals
which receive more revenue when there are more hospital days.
33. In its favor, the use of bar codes is a simple process; a physician’s electronic
order is sent directly to the pharmacy, where a bar code specific to the patient
is generated. The pharmacist proceeds to verify the order, apply the bar code to
the medicine, and send it to the appropriate floor, where nurses can compare the
bar codes of the patient and the medicine, using a scanner (Wright and Katz
2005).
34. Wright and Katz (2005).
35. Partners Healthcare deemed this to be a worthwhile expenditure. Partners
includes community hospitals in Boston, two academic medical centers, specialty
facilities, and community health centers. The cost for Partners to implement a
bar-coding system was $10 million, with an additional $1 million in annual
operating cost (Wright and Katz 2005). However, Brigham and Women’s Hos-
pital, one of Partners’ two founding academic medical centers, reported a 50
percent decrease in drug errors, which it estimated as the prevention of twenty
dispensing errors per day at $4,700 per event (Wright and Katz 2005).
36. Critics point to the human error inevitably involved in the system. For
instance, a patient without diabetes almost received a large dose of insulin after
his bar-coded wristband had been inadvertently switched with that of another
patient who had diabetes (Koppel, Metlay, Cohen, et al. 2005; McDonald 2006;
Wachter 2006).
37. Birkmeyer (2003).
38. Koppel, Metlay, Cohen, et al. (2005).
39. For cost estimates, see Bates, Leape, Cullen, et al. (1998).
40. The annual operating cost is high because the system must be continuously
updated to achieve its potential. A CPOE system, however, need not incorporate
all the bells and whistles. Even a modest CPOE system can provide great error
reduction so long as it provides for dose selection, a simple process for checking
orders against allergies and drug interactions, and a method for physicians to
clearly mark the frequency of the drug dosages (Bates, Leape, Cullen, et al. 1998).
In addition to error reduction, CPOE results in improved documentation, since
the only way the physician can place an order is through the computer system.
Also, the system provides a detailed log of all patient orders, which may be useful
in tort litigation defense.
 Notes to page 203 373

41. A randomized, controlled clinical trial showed a connection between a com-

prehensive electric medical record system and a 12.7 percent reduction in charges
per admission (Kaushal, Shojania, and Bates 2003). Cutler, Feldman, and
Horwitz (2005) report that for-profit hospitals were the least likely to invest in
CPOE systems and government hospitals (excluding federally owned hospitals)
were the most likely to do so. Apparently, the view that CPOE reduces cost is
not shared by for-profit hospitals.
42. One study found a threefold increase in mortality of critically ill children
after the installation of commercially sold CPOEs (Han, Carcillo, Venkataraman,
et al. 2005; Wachter 2006). User error accounts for much of this; in a survey 72
percent of house staff reported “difficulty in viewing all the medications on one
screen,” causing uncertainty about medications and dosages (Koppel, Metlay,
Cohen, et al. 2005). Other serious problems found with use of CPOE include
failure to provide for suspension of medication post surgery, sending medication
to the wrong room, and inflexible ordering screens that result in prescription of
incorrect medications.
43. Koppel, Metlay, Cohen, et al. (2005). One hospital reported a 20 to 40
percent loss of efficiency with the use of CPOEs (Poon, Blumenthal, Jaggi, et al.
2004).
44. One study found that more than 80 percent of malpractice suits are due
to communication problems between treating physicians (Levinson 1994).
Another major source of discontinuity exists with the use of a call system, which
requires residents to cover other teams’ patients during a twenty-four-hour
period. The handoff between the on-call resident and the patient’s regular resi-
dent is where communications may fail. X-rays and test results may be lost or
misdated, thus preventing coordination between the physicians of the patient’s
diagnosis and care (Ghandi 2005). Unsurprisingly, one study found an associa-
tion between potentially preventable adverse events and a cross-coverage system
where every four days patients are seen by the on-call resident instead of their
usual resident (Petersen, Brennan, O’Neil, et al. 1994). In addition, a typical
primary care physician may have to review up to 800 data elements from chem-
istry and hematology reports per week, spending more than seventy minutes per
day on test result management (Poon, Blumenthal, Jaggi, et al. 2004). Con-
fronted with this information overload, it is difficult for a primary care physician
to adequately communicate and follow up with other physicians caring for the
patient in the inpatient setting.
In the United States, patients’ physicians typically follow them into the hos-
pital. This is unlike most countries, where hospital-based physicians work inde-
pendently of community physicians (Ajdari and Fein 1998; Wachter and Goldman
1996). Quite independently of Err and in spite of custom, a new specialty has
emerged which alters inpatient care. Hospitalists, who care for and coordinate
the care of hospitalized patients, have grown in number since their emergence in
the early 1990s; currently there are 10,000 hospitalist physicians with a profes-
sional society of 5,000 members (Freese 1999; Wachter 2004). There have been
374 Notes to page 203

numerous studies regarding the effect of hospitalists on patient safety, and they
have several findings in common.
Most of the studies found that hospitalists reduced the average length of
hospital stay and lowered costs (Wachter and Goldman 2002). Diamond, Gold-
berg, and Janosky (1998) find that the readmission rate at a community teaching
hospital was reduced by almost 50 percent after the introduction of hospitalists.
It appears from the data that improvements in cost and outcomes increase with
time. A study done by Meltzer, Manning, Morrison, et al. (2002) demonstrated
an 8 percent reduction in length of hospital stay, 4.6 percent lower costs, an 18
percent reduction in risk for thirty-day mortality, and a 14 percent reduction in
risk for six-day mortality with the care of hospitalists, but only after their second
year of service. This same study also showed a $780 savings in adjusted costs
for the hospitalists’ second year of service. A more recent study (Kaboli, Barnett,
and Rosenthal 2004) echoed these findings. Another profession consists of inten-
sivists who focus their work in the ICU (Wachter and Goldman 1996). Twenty-
two percent of hospitals have fully implemented the use of intensivists (Devers,
Pham, and Liu 2004).
Even though hospitalists may improve coordination of care for hospitalized
patients, there has been some concern in the medical community regarding the
negative impact of a hospitalist system. A central and reoccurring concern is not
only the discontinuity in care during hospital admission and discharge, but also
increased costs for the hospital, and the effect on the relationship between
patients and their primary care physician (Showstack, Katz, and Weber 1999;
Wachter 2004). The hospitalist program at Kaiser Permanente Medical Center
in Santa Clara, California, faced the most resistance from internists reluctant to
hand over their inpatient responsibilities (Craig, Hartka, Likosky, et al. 1999).
To allay these fears, successful hospitalist systems have implemented procedures
to continue communication between the hospitalists and the primary care physi-
cians. A few of these measures are calling the primary care physician at both
admission and discharge, faxing daily progress notes, and encouraging the
involvement of the primary care physician through hospital visits or phone calls
to the patient (Wachter and Goldman 2002). After introduction of hospitalists
in 1994, Park Nicollet Hospital in Minnesota found that 89 percent of their
internists and family practitioners felt the hospitalist system was better or much
better than before; it had improved the care of patients, their call schedules, and
communication with colleagues (Freese 1999). However, some evidence shows
that when patients are seen by hospitalists during the week and see traditional
general medicine attendings over the weekend, the discontinuity of care elimi-
nates any savings a hospitalist may have accumulated during the week (Meltzer
2001).
45. Institute of Medicine (2000).
46. Marchev, Rosenthal, and Booth (2003).
47. The Act also requires that hospital peer review groups report any disciplin-
ary actions against medical staff to the National Practitioner Data Bank (42
 Notes to pages 203–207 375

USCS §11132). Such reports may be accessed by other hospitals considering a

physician for membership on their medical staff. The HCQIA was designed to
allow the removal of incompetent doctors by peer review committees without
fear of lawsuit from the physician in question; it was not designed to protect
information from discovery by malpractice plaintiffs or the public more
generally.
The court in Johnson v. Nyack stated, “The HCQIA gave qualified immun-
ity from suit to officials who conduct peer reviews that meet the standards
outlined in the statute. Yet Congress, in providing protection for those involved
in peer review, did not establish a privilege for most documents created in
that process” 169 F.R.D. 550, 560 (1996 SD NY). Another court went so
far as to say, “The Health Care Quality Improvement Act of 1986 is a clear
congressional statement that no general medical peer review privilege exists in
federal law” (Mattice v Mem’l Hosp. of S. Bend 203 FRD 381 (N.D. Ind.
2001).
48. Marchev (2003).
49. Weinberg, Hilborne, and Ngyuen (2005).
50. They state: “Reluctance to embrace [physician clinical performance assess-
ment] PCPA initiatives on grounds that they will be used as evidence against
physicians in malpractice litigation reflects perceptions of the law rather than
realities. The bar for admission of such evidence in malpractice litigation is high
and the possibility that PCPA data will reach this bar seems remote, at least for
the vast majority of injury types that prompt litigation. However, PCPA mea-
sures may be used against physicians in other medicolegal circumstances. Their
exclusion in litigation does not necessarily extend, for example, to proceedings
by state licensure boards, hospital review committees, and other adjudicatory
bodies with more relaxed rules of evidence” (Kesselheim, Ferris, and Studdert
2006 pp. 1833–1834).
51. Longo, Hewett, Ge, et al. (2005).
52. Wood and Nash (2005).
53. These are similar to the MedMARx system. See above.
54. Jacobson and Bloche (2005); Wood and Nash (2005).
55. Daly (2005).
56. State governments, irrespective of hospital responses in their states, support
public release of this information; a majority of the states with reporting systems,
fourteen of twenty-one, felt it would be useful to release it to the public. In 2003,
these states issued periodic reports or planned to (Marchev, Rosenthal, and
Booth 2003).
57. Sage (2003); White (1994).
58. These theories are discussed in greater detail in chapter 9.
59. Continuing medical education requirements may be imposed, but basically
physicians, unlike drivers, whose licenses require periodic reexaminations, are
376 Notes to pages 207–208

licensed for life (with the exception of delicensure from disciplinary

proceedings).
60. Sloan, Mergenhagen, Burfield, et al. (1989).
61. Ameringer (1999). Bovbjerg, Aliaga, and Gittler (2006) document that
medical boards are experimenting with some promising alternatives to conven-
tional processes.
62. The American Medical Association is attempting to remedy these inconsis-
tencies through a voluntary accreditation program which measures individual
physicians against national standards and peer performance (Viswanathan and
Salmon 2000).
63. Blumstein and Sloan (1988).
64. Newton (2001). A sampling of the variability of these statutes: in 1999,
forty-seven states had immunity statutes, forty-eight states had peer review stat-
utes, and thirty-one states required confidentiality and gave privilege to informa-
tion obtained during peer review (Scheutzow 1999). Of these thirty-one states,
only eight imposed civil or criminal penalties for breach of confidentiality; of the
forty-eight states granting privilege, only ten granted full protection to any
parties in a civil action involving the provider under review without requiring
release of that information to the state licensing board (Scheutzow 1999).
65. As of the late 1990s, forty-four states used evaluations from the Joint Com-
mission on Accreditation of Healthcare Organizations (JCAHO) as a require-
ment for licensure, with 85 percent of hospitals accredited. Health plans are
evaluated in a similar fashion, with accreditation a component in obtaining
licensure (Institute of Medicine 2000).
66. Devers, Pham, and Liu (2004). To address patient safety issues in
ambulatory care, Schaffer, Feldman, Fleischer, et al. (2005) suggest that all physi-
cians be accredited by a hospital, including performing a minimum number of
procedures per year. However, there is some evidence suggesting that stricter
licensing and credentialing laws may not improve outpatient safety. Based on a
study of reported adverse events in office settings in Florida during 2000–2004,
Coldiron, Fisher, Adelman, et al. (2005) report that 94 percent of the physicians
reporting adverse events were board certified, and 97 percent had hospital
privileges.
Some states and professional organizations, including Pennsylvania, Rhode
Island, Florida, California, Georgia, Texas, the American Society of Plastic Sur-
geons, and the American Society of Aesthetic Plastic Surgeons, require accredita-
tion from an outside agency (Coldiron, Fisher, Adelman, et al. 2005; Franko
2001). In addition, the JCAHO has recently taken the initiative to reduce errors
from patient handoffs. For the first time, they are requiring hospitals to establish
standards for communications during handoffs (Landro 2006).
67. Bradley, Herrin, Elbel, et al. (2006). E.g., beta-blocker use at admission and
discharge, aspirin use at admission and discharge, and angiotensin-converting
enzyme (ACE) indicator use.
 Notes to pages 208–211 377

68. Heart attack.

69. Hyman and Silver (2005).
70. Ibid. Physicians in other specialties have also recognized that they have a
positive role to play in improving patient safety in their practices. For example,
Attarian and Vail (2005) suggest to their colleagues that every surgeon document
and demonstrate four elements in the care of each patient: (1) technical knowl-
edge of the procedure; (2) competence in performance; (3) carefulness in evalu-
ation and diagnosis; and (4) care in the treatment of patients using informed
consent, professional skill and attention to the patient’s surgery, postoperative
care, and complications. The authors conclude that following this standard of
care will make proving liability nearly impossible, while improving patient
safety.
71. Thirty-nine states have introduced legislation, and six have passed laws
(R. Weinstein, Siegel, and Brennan 2005).
72. Marchev, Rosenthal, and Booth (2003).
73. Marchev (2003). Nineteen of the twenty-one states enacted legislation
designed to protect the reported data. Methods included de-identification of data,
to protect facility confidentiality, and anonymous reporting, to reduce the risk
of increased litigation. Though helpful in encouraging providers’ participation,
both of these methods limit the value of the data for individual consumer
choice.
74. Marchev, Rosenthal, and Booth (2003).
75. Dranove, Kessler, McClellan, et al. (2003).
76. Jin and Leslie (2003).
77. Coldiron, Fisher, Adelman, et al. (2005).
78. Texas includes all outpatient facilities owned or run by a hospital in its
mandatory reporting requirements (Texas Health and Safety Code Sec.241.202
(2003). However, many states extend the duty only to outpatient surgical facili-
ties (National Academy for State Health Policy 2006). See Pennsylvania (Pa. Stat.
Ann. tit. 40, §1303.301–1303.315 (2003)); Nevada (Nev. Rev. Stat. Ann.
§439.800–439.890 (2006)); and Connecticut (Conn. Gen. Stat. §19a-127n
(2006).
79. See chapter 3 for a more detailed discussion of this issue.
80. Lemaire (1985).
81. Moore and Viscusi (1989); also see chapter 10.
82. Topel (1984).
83. Sloan, Stout, Whetten-Goldstein, et al. (2000).
84. Enterprise liability is discussed in depth in chapter 12.
85. Havighurst (1997). Danzon (2000), however, cautions that the saving
in litigation cost would be minor. Individual physicians would still be called
to testify to describe the course of events leading to the injury. In addition,
378 Notes to pages 211–213

she notes that as long as the liability rule includes negligence, individuals
may be asked to describe the facts leading to a conclusion of negligent or non-
negligent acts.
86. The case for health plans as the enterprise is as follows. They are well-capi-
talized; the managed health plans may have the capacity to balance quality and
cost in their product offerings, and can incorporate alternative dispute resolution
into contractual provisions (Sage and Jorling 1994). Managed health plans
presumably are actively involved in managing care and eliminating care that is
not cost-effective. Thus, shifting liability to them seems to be a practical
alternative.
Several health plans have willingly embraced enterprise liability without any
compulsion from statutes or regulation. Kaiser-Permanente, Sharp Health Care,
the Public Health Service, the Department of Defense, the Veterans Administra-
tion, the Indian Health Service, and the Bureau of Prisons all expressly assume
liability for the medical errors of their staff physicians (Sage and Jorling 1994).
Having responsibility for the actions of physicians with which health plans have
contractual arrangements creates incentives for health plans to be more cautious
in their screening of providers (Jacobi and Huberfeld 2001).
But having health plans be the enterprises also has several important disad-
vantages. And the disadvantages are substantial enough to largely eliminate
health plan enterprise liability as a major alternative to the present system.
Unlike hospitals, health plans are typically not located where the medical care
is delivered. Except in closed staff managed care organizations, such as Kaiser,
most physicians have contractual arrangements with several health plans. It
becomes difficult to abide by various protocols of different health plans and,
often, even to know what the protocols are. There is considerable turnover in
the ownership of health plans. During the course of a single tort case, health
plan ownership may change several times. Much of employer-provided health
insurance in the United States is self-insured. It seems doubtful that many
employers would desire to take on being potentially liable for medical malprac-
tice. Nor would they likely be efficient in performing this role. Given the managed
care backlash and physician opposition to many health plans, this would amount
to moving physicians from the frying pan, the existing situation, into the fire,
enterprise liability with health plans the enterprise of choice. Finally, some per-
sonal health care services may not be covered by enterprise liability. Services such
as lasik surgery or home health or nursing home care, may not be covered by
the health plan. In principle, such services could be excluded from coverage under
the enterprise liability program, and force providers to continue to obtain medical
malpractice insurance as they do currently.
87. Mello and Brennan (2002).
88. Danzon (1985).
89. Sloan (1990). Of fourteen medical malpractice insurers, three multiple-line,
commercial stock and eleven single-line physician-sponsored mutuals or recipro-
cals specializing in medical malpractice insurance surveyed in 1987–1988 by
 Notes to pages 213–219 379

colleagues and the first author, often fewer than 1 percent of physician insureds
paid more than standard rates because of adverse prior claims experience. Among
these, only recent claims were counted against the physician since insurers
believed that old claims have little predictive value. None of the surcharges
exceeded 200 percent of standard rates.
The survey revealed several reasons that medical malpractice insurers were
reluctant to engage in experience rating. Some national medical organizations
and their physician constituents opposed it. Surcharged physicians often left the
company because they were able to secure lower premiums from competing
insurers. This occurred because each insurer used only its own experience in
setting premiums. Experience rating was seen as infeasible in medical malpractice
insurance markets in which competition prevailed, since physicians could always
find a lower-priced alternative. Some insurers indicated that experience rating is
inappropriate for a line with low claims frequency, such as medical malpractice
insurance.
90. Sloan, Bovbjerg, and Githens (1991). In some cases, physicians with adverse
records find that their coverage is not renewed and they must seek coverage from
surplus line insurers (W. B. Schwartz and Mendelson 1989) or joint underwriting
associations (see chapter 9).
91. Both insurers of physicians and self-insuring hospitals frequently purchase
reinsurance coverage (see chapter 9).
92. The Poisson process, named after the French mathematician Siméon-Denis
Poisson (1781–1840), is a stochastic process which is defined in terms of the
occurrences of events (Ellis, Gallup, and McGuire 1990; Rolph 1981; Nye and
Hofflander 1988). (http://en.wikipedia.org/wiki/Poisson_process, accessed June
20, 2007).
93. Sloan and Hassan (1990).
94. In Soviet history and iconography, a Stakhanovite follows the example of
Aleksei Grigorievich Stakhanov, employing hard work to overachieve on the job
Stakhanovite workers were honored and rewarded for exceptional diligence in
increasing production (Wren and Bedeian 2004). http://en.wikipedia.org/wiki/
Stakhanovite (accessed June 20, 2007).
95. Mello, Kelly, Studdert, et al. (2003).

Chapter 9

1. Danzon (1985).
2. Sloan (1990).
3. Sloan, Bovbjerg, and Githens (1991).
4. A Lloyd’s association is like a reciprocal (see below) in that the organization
issues no policies. Rather, it serves as a mechanism whereby members insure
themselves and others. A Lloyd’s association insures outsiders, whereas in a
380 Notes to pages 219–229

reciprocal, members insure each other. In the malpractice insurance field, the
Lloyd’s form is used only for reinsurance.
5. Greene (1976).
6. Klein (1995); Sloan, Bovbjerg, and Githens (1991).
7. Sloan and Hsieh (1990).
8. Klein (1995).
9. Harrington (2002).
10. Cummins and Danzon (1991).
11. See chapter 2 for further discussion of this point.
12. Blackmon and Zeckhauser (1991); Grabowski, Viscusi, and Evans (1989).
13. Yelen (1993).
14. Hart, Shleifer, and Vishny (1997).
15. The Risk Retention Act was first passed by the U.S. Congress in 1981 to
assist individuals or organizations seeking product liability coverage (P.L. 97–
45). It responded to specific concerns about manufacturers’ ability to purchase
coverage even prior to the crisis of the mid-1980s. In 1986, the coverage was
expanded to include liability coverage in general, including medical malpractice
insurance (P.L. 99–563).
16. States without mandatory insurance use other mechanisms to accomplish
the same objective.
17. According to Schwartz and Mendelson (1989a), in the 1980s, about 900
physicians in the United States lost coverage from a standard insurer and were
able to gain coverage through a JUA.
18. Schwartz and Mendelson (1989b).
19. Sloan, Bovbjerg, and Githens (1991, p. 177).
20. In a theoretical study, Shavell (1987) shows that insurance does not neces-
sarily interfere with the deterrence function of malpractice liability if premiums
are perfectly experience rated. In contrast to medical malpractice, experience
rating is common in other lines, e.g., automobile insurance or workers’
compensation.
21. Sloan (1990).
22. Hickson, Clayton, Githens, et al. (2002); Sloan, Mergenhagen, Burfield,
et al. (1989).
23. For example, New York State required that its Department of Insurance
institute merit rating in the mid-1980s. Merit rating is a system of surcharges
and credits based on an individual insured’s history of liability claims relative to
the average insured in his or her specialty and geographic area, and on disciplin-
ary actions by hospitals or licensing boards against the insured. Based on this
experience, James P. Corcoran, Superintendent of Insurance of New York State
concludes: “Physicians are unalterably opposed to merit rating. It is unrealistic
 Notes to pages 229–234 381

to apply a merit rating plan, or any individual risk rating plan, to a low-
frequency, high-severity coverage. Due to the length of time claims are open, it
is difficult to have enough meaningful data for merit rating of medical malprac-
tice. A merit rating plan is not intended to be used to remove poor doctors by
pricing them out of business” (1997, p. 2). An experience review plan was pro-
posed in Massachusetts, but in the face of political opposition from physicians,
the plan was never adopted (Sloan, Mergenhagen, Burfield, et al. 1989). Other
programs implemented by individual insurers have not been evaluated (Sloan,
Bovbjerg, and Githens 1991, pp. 173–176). Of the thirteen out of the fourteen
insurers that had previously implemented some form of experience rating among
the insurers surveyed in 1987–1988, most had completely abandoned the program
as of the survey date or continued a program in a very limited form (Sloan,
Bovbjerg, and Githens 1991, p. 177).
24. Sloan and Hassan (1990).
25. Nutter (1985).
26. Council of State Governments (2003); New York Department of Insurance
(1997); Robinson (1986).
27. Morton (2003).
28. Sutter (2002).
29. Council of State Governments (2003); New York Department of Insurance
(1997); Robinson (1986).
30. Kenney (1988).
31. Sloan, Bovbjerg, and Githens (1991).
32. Nutter (1985).
33. Hospital and Healthsystem Association of Pennsylvania (2002b).
34. Bovbjerg and Bartow (2003).
35. Hospital and Healthsystem Association of Pennsylvania (2002a). These
increases occurred even though coverage from the JUA tends to be relatively
expensive (Eskin 2003; Mello, Kelly, Studdert, et al. 2003).
36. State of Florida (2003).
37. One exception is legislation introduced in Missouri requiring the JUA
administrator to formulate, implement, and monitor a risk management program
for all policyholders (State of Missouri 2002, 2003).
38. Pennsylvania Professional Liability Joint Underwriting Association Manual.
http://www.pajua.com/images/200309%20U-W%20Manual.pdf.
39. Downs and Sommer (1999).
40. Sloan, Bovbjerg, and Githens (1991, pp. 55–56).
41. Lee, Mayers, and Smith (1997).
42. Hofflander and Nettesham (2001).
43. Cummins (1988); Lee, Mayers, and Smith (1997).
382 Notes to pages 234–240

44. Downs and Sommer (1999). On the other hand, a guaranty fund may induce
insurers to monitor each other and alert regulators when a competitor takes on
too much risk (Hall, Cummins, Laderman, et al. 1988).
45. See, e.g., Sloan (2004).
46. Gron (1990, 1994a).
47. Gron (1994b).
48. E.g., Breyer (1982); Peltzman (1976); Posner (1974); Stigler (1971).
49. Schneiberg and Bartley (2001).
50. Meier (1988).
51. Grabowski, Viscusi, and Evans (1989) investigated the effects of state auto-
mobile insurance regulation on price and availability of insurance. Earlier studies
on these outcomes had produced mixed results, in part because there was little
variation in state regulatory practices in the sample. Grabowski and coauthors
used a sample of thirty states for the period 1974 to 1981 and a sample of eleven
states that had deregulated such insurance. They report that regulation reduced
the auto insurance premiums relative to losses, but also increased the size of the
involuntary market in these states by 17 percent. The involuntary market con-
sisted of JUA and assigned-risk plans. The disadvantage of the involuntary
market is the extensive subsidy of unsafe drivers, which not only is inequitable
but also increase the risk of driving to the public at large.
52. Munch and Smallwood (1980).
53. Above, we used the example of a 0.4–0.5 percent insolvency risk. A lower
insolvency risk would require higher initial capitalization. Perhaps some consum-
ers are willing to tolerate a 0.4–0.5 percent insolvency risk per year. By setting
the initial capitalization standard too high, these persons are made worse off.
Also, some firms may be undesirable in terms of their insolvency risk, but may
offer other products with attributes that some consumers desire. Also, elimina-
tion of competition from small firms may result in higher premiums being
charged by those firms that are able to gain entry into the market. Winter (1988,
1991) warns that insurance regulators who use the premium-to-surplus ratio to
gauge insolvency risk may view increases in ratio as an unfavorable development.
If regulators act on this evidence to restrict the supply of insurance during down-
turns in the insurance cycle, they may exacerbate the cycle by reducing avail-
ability of coverage. Winter does not present any direct empirical evidence that
regulation has actually had this effect.
54. Sloan, Bovbjerg, and Githens (1991).
55. Born, Viscusi, and Baker (2006).
56. Viscusi and Born (1995).
57. Viscusi and Born (2005).
58. Born (2001, p. 212).
59. Goldstein (2003).
 Notes to pages 240–248 383

60. Darragh (2002b).

61. Goldstein (2003).
62. Ibid.
63. Darragh (2002a).
64. Goldstein (2003).
65. McLeod (2002).
66. Darragh (2002b).
67. Ibid.
68. McLeod (2002).
69. Ibid.
70. Goldstein (2003).
71. Ibid.
72. St. Paul stopped accepting new clients and refused to renew the policies of
its current policyholders in 2000.
73. In 1991 NORCAL acquired the assets and liabilities of the Medical Indem-
nity Corporation of Alaska (MICA), and in 1994 it acquired a portion of Premier
Alliance Insurance Company’s business located in Rhode Island. NORCAL then
opened offices in Anchorage and Providence. A few years later, NORCAL
entered into a jointly owned venture with the Medical Mutual Liability Insurance
Society of Maryland (Med Mutual). Its portion of this venture began with the
start of an insurance holding company, Medical Group Holdings, Inc. (MGH).
In 2000, NORCAL acquired 100 percent ownership of MGH from Med Mutual.
Also during this time period, NORCAL acquired the majority interest in Penn-
sylvania Medical Society Liability Insurance Company (PMSLIC). Four years
later, MGH acquired the ownership of PMSLIC, giving NORCAL 100 percent
ownership and making PMSLIC a subsidiary of NORCAL.
74. Darmiento (2005).
75. Ibid.
76. Cebula (1995); Grossman (1992); Keeley (1990); Wheelock and Wilson
(1995).
77. Grossman (1992).
78. Keeley (1990).
79. Wheelock and Wilson (1995).
80. Cebula (1995).

Chapter 10

1. Sloan, Bovbjerg, and Githens (1991).

2. Klein (1998, p. 176).
384 Notes to pages 248–251

3. Kunreuther (1998, p. 47).

4. Medical malpractice insurers have solved the problem of the lag between the
date of injury and the date of claim filing by switching to claims-made coverage
from occurrence coverage. Under claims made, the insurer covers losses from all
claims filed during the policy year. Under an occurrence policy, the insurer covers
losses from injuries occurring during the policy year.
5. Sloan, Eesley, Conover, et al. (2004).
6. Kunreuther, Pauly, and Russell (2004); Michel-Kerjan and Marcellis-Warin
(2006).
7. According to Doherty and Smetters (2005), there is more monitoring when
the primary insurer and reinsurer are affiliates than when they are unaffiliated,
presumably because affiliations reduce monitoring cost.
8. In Pennsylvania, a PCF provides what is essentially a middle layer of medical
malpractice insurance coverage for hospitals. Thus, since there is no corridor
between the maximum loss covered by the primary insurer and the point at which
the middle layer attaches to the PCF plan, the incentive that the primary insurer
has to defend claims is reduced. It is often easier for a primary insurer to close
a case when the settlement offer rises to an amount covered by the PCF than to
fight for a lower settlement at a level below which the PCF is obligated to pay
part of the claim.
9. Hoerger, Sloan, and Hassan (1990).
10. A survey of hospital administrators conducted in 2004 by the first
author and Duke University colleagues revealed that hospitals choose to purchase
excess coverage to eliminate/reduce total excess loss, to spread the risk of
catastrophic loss over time, and to address trustee concerns (Sloan, Eesley,
Conover, et al. 2004). Some might argue that medical malpractice expense is a
small percentage of total hospital revenue—3.9 percent for physicians, 1 percent
for hospitals (United States General Accounting Office 1995; United States Gov-
ernment Accountability Office 2005, p. 5). However, to the extent that raising
hospital prices is precluded, increases in premiums directly reduce profits so that
premium increases as a share of profits are likely to be substantially higher.
Factors such as market competitiveness and hospital ownership influence the
ability of hospitals to shift costs by increasing prices. The extent of this cost-
shifting to consumers appears to be stable over time (Zwanziger, Melnick, and
Bamezai 2000).
11. Given that bankruptcy is not costless, even an insurer not averse to risk
(risk-neutral insurer) would demand reinsurance. In analyzing decisions of firms,
economists typically assume that the firm is risk neutral. An alternative is to
assume that the firm is risk averse. While this complicates the analyses, in general,
the more complex assumption of risk aversion does not yield important addi-
tional insights.
12. See chapter 2 for further discussion of insurer constraints.
 Notes to pages 252–256 385

13. Froot (2001) This information was obtained from Guy Carpenter &
Company, a reinsurance subsidiary of Marsh McLennan and by far the largest
at-risk intermediary in the United States.
14. Ibid. (p. 541).
15. Weiss and Chung (2004).
16. See chapter 9 for a discussion of adequacy of primary medical malpractice
premiums for which the requisite data are available.
17. A common misconception is that insurers can just increase premiums in
response to previous losses. Such patterns are observed, but for a different
reason. Premium increases following the occurrence of a catastrophic event
may be attributable to probability updating by insurers (future events seem
more likely) and not to repayment of losses from prior claims (Froot and
O’Connell 1997; Michel-Kerjan and Marcellis-Warin 2006; Weiss and Chung
2004).
There may also be updating by potential insurance purchasers. There is
empirical evidence in other contexts (e.g., flood insurance and terrorism insur-
ance) that consumers tend to cancel their policies after a period during which an
event did not occur. After an event occurred, their beliefs and demand for insur-
ance change dramatically. See a summary of evidence in Kunreuther, Pauly, and
Russell (2004) and Kunreuther (2006).
18. Kunreuther, Pauly, and Russell (2004).
19. Moral hazard is an issue for insurance more generally. According to Kun-
reuther (2006), “If the insurer reduced the Lowes’ [a hypothetical family] hom-
eowners’ premium by $275, would the family invest in the mitigation measure?
Empirical evidence on individuals’ decision processes with respect to adoption
of protective measures suggests that they would not.”
20. The cedent.
21. Froot lists two other possible explanations that are even more speculative
than the others, and therefore are not discussed here. In the end, he calls for
more empirical evidence and does not reach definitive conclusions among the
various explanations he proposes. Overall, the high cost of externally supplied
equity capital seems to be the most compelling of the above explanations.
Subsequent research by Weiss and Chung (2004) finds empirical support for this
explanation.
22. Nordman, Cermak, and McDaniel (2004).
23. See further discussion of this issue in chapter 2.
24. Doherty, Lamm-Tennant, and Starks (2003); Bovbjerg and Bartow (2003,
p. 21).
25. Venezian (1985). These shocks can be due to factors that affect property-
casualty insurance in general (Doherty, Lamm-Tennant, and Starks 2003; Gron
1994a; Winter 1988).
26. Froot and O’Connell (1997).
386 Notes to pages 256–264

27. Sloan, Bovbjerg, and Githens (1991), using data on medical malpractice
coverage for the years 1975–1985, conclude that profitability of primary
medical malpractice insurers on a risk-adjusted basis was about at the level
one would expect in a competitive industry. This analysis is based on income
and balance sheets (convention statements), which are routinely filed with
state departments of insurance. However, given the lack of regulation of rein-
surers by these same departments, a parallel analysis cannot be conducted for
reinsurance.
28. Even self-insured hospitals pay “premiums” for excess coverage in terms of
their loss experience plus overhead costs of running a captive or risk retention
group (RRG). Regulators also require self-insured entities to hire a “fronting”
insurer licensed in the state to assure payment of claims (Bovbjerg and Bartow
2003). Despite the increase in premiums, 20 percent of hospitals in a Duke study
reported they had had difficulties collecting payments from their excess carrier
since 2000, and 33 percent reported difficulties in collecting payments from their
PCF (Sloan 2004).
Theory suggests two ways for high premiums or lack of availability of cover-
age to affect hospitals. First, hospitals unable to purchase coverage may cease
provision of services that are most likely to produce malpractice claims as a way
of reducing their exposure to risk. Second, if coverage is purchased at a high
price, hospital cash flow could be compromised. Five of the twenty-one hospitals
in the Duke University study reported closing services (obstetrics, for example)
or failing to open a service as a result of concerns about excess coverage (Sloan
and Eesley 2006). In addition, lack of excess insurance or high premiums had
affected hospital operations in the past for 40 percent of hospitals surveyed.
Many hospitals surveyed said that fewer dollars were available for expansion of
services due to high premium expenses.
29. Blumberg and Holahan (2004).
30. Katherine Swartz (2006) proposes that the federal government provide a
reinsurance program for health insurance that would take responsibility for
persons in the highest 1 percent of medical expenses but would be limited to
persons who were in the individual and small-group insurance markets. She
estimates that such a program would reduce health insurance premiums for such
coverage by 20–40 percent, with the caveat that the savings would depend on
how the reinsurance programs were structured. She argues that such a program
would have the advantage of reducing insurers’ incentives to select against high-
cost individuals which exists currently.
31. Admittedly, for equity reasons, the change in identity may be social
welfare-enhancing.
32. Insurers that sell medical liability insurance but no other type of coverage.
33. Sloan, Bovbjerg, and Githens (1991, p. 123).
34. Sloan, Bovbjerg, and Githens (1991).
35. Sloan, Eesley, and Conover (2005, fig. 2).
 Notes to pages 264–270 387

36. Bovbjerg and Bartow (2003).

37. Hoerger, Sloan, and Hassan (1990).
38. Pinnacle Actuarial Resources (2003).
39. Philadelphia juries, for example, in recent years have often awarded amounts
in medical malpractice cases in excess of $1 million (Bovbjerg and Bartow 2003).
In Pennsylvania, the presence of a PCF has not eliminated hospital demand for
private excess coverage (Hospital and Healthsystem Association of Pennsylvania
2002a).
40. See the classic article by Rothschild and Stiglitz (1976), which led to the
authors receiving a Nobel Prize in economics.
41. All PCFs except those in Pennsylvania and New York now use the reserving
for anticipated losses mechanism. Pennsylvania still uses pay-as-you-go financing
even though several state PCFs have decided to drop it (Hofflander, Nye, and
Nettesheim 2001).
42. Kansas, New Mexico, Wisconsin, and Louisiana have used loss-reserving in
funding their PCFs (Kansas Health Care Stabilization Fund 2002; Louisiana
Patient Compensation Fund, http://www.lapcf.state.la.us; New Mexico Public
Regulation Commission 2001; Wisconsin Patient Compensation Fund, http://oci
.wi.gov/pcf/htm). Still, choosing to hold reserves in and of itself does not assure
adequate financing. Though the reasons may differ, public and private insurers
which have actuarial evaluations performed do not always follow their recom-
mendations (Sloan, Bovbjerg, and Githens 1991, p. 157). Additionally, reserves
are vulnerable to utilization for unrelated purposes, as has occurred in Wisconsin
(Wisconsin Insurance Report 2001, http:/oci.wi.gov/ann_rpt/bus_2001/anrpttoc
.htm; Wisconsin Hospital Association 2003).
43. Sloan, Bovbjerg, and Githens (1991, pp. 123–144).
44. Albert (2003) tells the story of one physician’s decision to move from
Pennsylvania to Wisconsin based on the different medical malpractice environ-
ments in each state.
45. Sutter (2002). The New Mexico Department of Insurance completed an
actuarial study concluding that the state’s PCF was underfunded by $9 million,
but the state medical society recommended no increase in physician
contributions.
46. Sloan, Bovbjerg, and Githens (1991).
47. In Wisconsin, a state in which the PCF does maintain reserves at a relatively
high level (Wisconsin Insurance Report 2001), the governor proposed to tap
$200 million from the fund in order to subsidize Medicaid (Wisconsin Hospital
Association 2003).
48. For further discussion of this issue, see chapters 5 and 7.
49. Hofflander, Nye, and Nettesheim (2001).
50. In South Carolina, which historically appeared to underreserve as a
matter of policy, health care providers could be sued individually for the
388 Notes to pages 270–277

full amount claimed if the PCF’s funds were insufficient to pay its obligations.
The state’s PCF and its consulting actuary considered the PCF a risk pool
rather than an insurance company, and therefore saw no need for regulatory
oversight or standard loss-reserving practices (South Carolina Legislative Audit
2000).
51. Margolis (1982); Musgrave (1957).
52. During the course of research on this report, the first author requested
unpublished material from several state agencies. In many cases, the material was
sent. In other cases, however, data could not be obtained. In at least one case,
an employee volunteered to work overtime, for an overtime wage, to photocopy
financial documents. Apparently no records were publicly available, nor was
production of the documents seen as an appropriate function of the public
agency.
53. Anderson (1976).
54. Insurance is available to residents only after their community has completed
the requirements. The area must be declared flood prone by FEMA, and the
community must file an application with the NFIP within one year (R. Stedman
2003). After completing a thorough application process requiring approval of
flood hazard maps and the institution of new community ordinances, the com-
munity becomes part of the Emergency Program, making it eligible for limited
flood insurance (R. Stedman 2003).
55. http://www.insurancescrawl.com/archives/2005/11/stranded_withou.html.
56. http://www.fema.gov/pdf/nfip/market_pen.pdf.
57. United States Congressional Budget Office (2002).
58. http://www.insurancescrawl.com/archives/2005/11/stranded_withou.html.
59. Chu (2005).
60. United States Congressional Budget Office (2002).
61. Available at www.citizensfla.com.
62. Kunreuther (2006).
63. Doherty and Smetters (2002); United States Congressional Budget Office
(2002, Box B-1, p. 45); Varian (2001).
64. United States Congressional Budget Office (2002).
65. Varian (2001); Doherty and Smetters (2002).

Chapter 11

1. See, e.g., Danzon (1991).

2. Danzon (1994).
3. Danzon (1991).
 Notes to pages 279–285 389

4. Weiler, Hiatt, Newhouse, et al. (1993); Bovbjerg and Sloan (1998); Studdert
and Brennan (2001).
5. Weiler, Hiatt, Newhouse, et al. (1993).
6. (Weiler 1993); Weiler, Hiatt, Newhouse, et al. (1993, p. 146).
7. Studdert and Brennan (2001).
8. See, e.g., Weiler, Hiatt, Newhouse, et al. (1993).
9. Joint Legislative Audit and Review Commission of the Virginia General
Assembly (2002, p. 4).
10. Ibid. (p. 83).
11. State of Florida (2003).
12. Sloan, Whetten-Goldstein, Kulas, et al. (1999).
13. Freeman and Freeman (1989).
14. Sloan et al. (1997).
15. Studdert, Fritz, and Brennan (2000).
16. Blair and Stanley (1988); Nelson and Ellenberg (1986).
17. Sloan, Whetten-Goldstein, Entman, et al. (1997).
18. Failure to give adequate notice is understandable, in that few expectant
mothers want to contemplate life with a child with cerebral palsy before the fact,
and this outcome has a low probability of occurring.
19. Studdert, Fritz, and Brennan (2000, p. 523).
20. Ibid. (p. 524).
21. Bovbjerg and Sloan (1998, p. 112).
22. Joint Legislative Audit and Review Commission of the Virginia General
Assembly (2002, p. vi).
23. Ibid. (p. 25). Unlike Florida, where closed medical malpractice claims are
publicly available and can be compared with no-fault, there are no claims in the
public domain in Virginia. Thus there is no way to perform an independent
comparison.
24. Sloan, Whetten-Goldstein, Entman, et al. (1997).
25. Joint Legislative Audit and Review Commission of the Virginia General
Assembly (2002, p. 9) In comparison, the tort system overhead is approximately
50 percent (Kakalik and Pace 1986).
26. Bovbjerg and Sloan (1998).
27. Hay (1996); Danzon (1983); B. L. Smith (1992).
28. Joint Legislative Audit and Review Commission of the Virginia General
Assembly (2002, p. 16).
29. Ibid. (p. xv).
30. Ibid. (p. 8).
390 Notes to pages 285–290

31. Norton (1997); Bovbjerg and Sloan (1998).

32. Joint Legislative Audit and Review Commission of the Virginia General
Assembly (2002, pp. 37–38).
33. State of Florida (2003, p. 306).
34. Joint Legislative Audit and Review Commission of the Virginia General
Assembly (2002, p. 49).
35. Ibid. (p. v).
36. NICA reserved $3 million per covered child on average (State of Florida
2003).
37. Sloan, Whetten-Goldstein, Entman, et al. (1997); Sloan, Whetten-Goldstein,
Stout, et al. (1998).
38. Sloan, Whetten-Goldstein, Entman, et al. (1997).
39. Bovbjerg and Sloan (1998, p. 117).
40. Lloyd-Puryear, Ball, and Benor (1998).
41. http://www.hrsa.gov/vaccinecompensation. Influenza and hepatitis A vac-
cines have recently been added to the program.
42. http://www.hrsa.gov/vaccinecompensation/statistics_report.htm
#post_1988.
43. Ibid.
44. Advisory Commission on Childhood Vaccines (2003).
45. http://www.hrsa.gov/vaccinecompensation/statistics_report.htm
#post_1988.
46. Evans (1998).
47. Sloan, Berman, Rosenbaum, et al. (2004).
48. Finkelstein (2004).
49. Advisory Commission on Childhood Vaccines (2004).
50. Ridgway (1999).
51. Evans (1998, p. 8).
52. During the discovery phase of the proceedings, respondents have produced
92,268 documents related to thimerosal. The Department of Justice believes the
excessive number of documents is delaying the proceedings and runs counter to
the intent of the no-fault program. The U.S. Food and Drug Administration was
asked to determine which documents may be relevant. The petitions appear to
be leveling off, but the number of thimerosal claims, about ten times the number
of nonthimerosal claims between 2002 and 2004, resulted in a need for addi-
tional funding to process all of the petitions. Since 1996, the appropriations to
staff the VICP program have stayed flat. However, the president’s budget for FY
2005 sought an increase of $2,305,000 (50 percent of the past appropriations)
to handle the growth in vaccine injury claims caused by thimerosal-related claims
(Keisler 2004).
 Notes to pages 290–291 391

53. Institute of Medicine (2004).

54. Holder v. Abbott Laboratories Inc. (2006). 444 F.3d 383; see also W. Davis
(2006).
55. Finkelstein (2003, p. 5).
56. These changes include the addition of chronic arthritis from rubella vaccines,
the removal of certain disorders for the DTP vaccine, and a clarification of the
definition of encephalopathy in 1995. In 1997, modifications of the table included
the addition of brachial neuritis and removal of encephalopathy for tetanus-
containing vaccines, as well a number of other modifications. In 1998 rotavirus
vaccine was added to the table. August 2002 brought a second category of rota-
virus along with an additional injury for that vaccine, addition of pneumococcal
conjugate vaccines with no condition specified, and removal of early-onset Hib
disease and residual seizure disorder from the table (Health Resources and
Services Administration 2006).
57. Ridgway (1999, p. 69).
58. Ibid. (p. 71).
59. Ibid. One solution may be to extend the statute of limitations so that the
scientific literature has more time to catch up to the legal reality of decisions that
need to be made. However, the problem is that while the scientific literature is
coming to a consensus, there are unknown numbers of people who are incurring
and having to pay large medical bills for real injuries. The VICP addresses this
potential problem with a long retroactive window once a change is made. If an
injury or a new vaccine is added to the table, petitioners have two years from
the effective date of the table change to file a claim for a vaccine-related injury
or death that occurred up to eight years before the effective date of the table
change (Advisory Commission on Childhood Vaccines 2004).
60. An important issue not discussed in the text relates to patient safety. There
are issues regarding how safety concerns are being communicated to vaccine
manufacturers and what they are actually doing about them. Do manufacturers
have to spend too much to comply with erroneous adverse events (this may be
happening with thimerosal), resulting in a costly tax on more beneficial innova-
tions? Or are they spending too little to expeditiously correct legitimate safety
issues? If the option of tort is eliminated, where is the manufacturer’s incentive
to correct and prevent the (admittedly rare) problems caused by the vaccines?
Isn’t genetic testing or altering the vaccine possible to prevent some of these side
effects?
Part of the answer is that there are four ways vaccine safety is monitored
(Institute of Medicine 1997). One question for a broader medical no-fault
program is whether it would require implementation of a similar safety monitor-
ing system to be able to update Tables of Injuries for a wide variety of medical
procedures and pharmaceuticals. For vaccines, the four monitoring systems are
the Vaccine Adverse Event Reporting System (VAERS), the Vaccine Safety Dat-
alink (VSD), monitoring by vaccine manufacturers themselves, and monitoring
by the FDA.
392 Notes to pages 291–295

VAERS was set up in 1990 and is administered by the Centers for Disease
Control and Prevention and the FDA. Health care providers are required to
report to VAERS serious adverse events occurring within thirty days of vaccina-
tion. However, the number of doses actually given from each lot is not known,
so interpretation of VAERS data is uncertain. Five adverse events have a very
different meaning if they are the result of twenty doses given from a lot as com-
pared to two million doses having been given.
Actively capturing more complete and detailed information from Oregon,
Washington, and California, the Vaccine Safety Datalink (VSD) helps to deter-
mine whether an event is linked with a vaccine or with some other cause. The
VSD can test and evaluate signals of potential problems spotted by the VAERS
system.
Food and Drug Administration monitoring is accomplished via review of
VAERS and the manufacturer’s data. Manufacturers form the largest percentage
of contributors to VAERS, so this monitoring represents some built-in duplica-
tion. The National Childhood Vaccine Injury Act of 1986 mandates that manu-
facturers report adverse events to the Department of Health and Human
Services.
61. Advisory Commission on Childhood Vaccines (2003).
62. Freeman and Freeman (1989).
63. Fishback and Kantor (1998b, p. 305).
64. But employees in nonunionized settings paid for the added benefits in the
form of reduced wages (Arnould and Nichols 1983; Dorsey and Walzer 1983;
Fishback and Kantor 1998a, 1998b; Gruber and Krueger 1991).
65. Fishback and Kantor (1998b, p. 326).
66. Ibid. (especially p. 326).
67. Ibid. (p. 325).
68. Ibid. (p. 327).
69. R. Smith (1990).
70. Ruser (1998).
71. Biddle, Roberts, Rosenman, and Welch (1998).
72. Leigh and Robbins (2004).
73. Rhodes and Ohlsson (1997, pp. 5–6 and 5–7).
74. This has been documented by Ruser (1998) and Campolieti and Hyatt
(2006).
75. Moore and Viscusi (1989).
76. Library (2004, p. 1), see http:www.igs.berkeley.edu/library/
htWorkersCompensation.htm (accessed December 30, 2004).
77. Bovbjerg, Tancredi, and Gaylin (1991).
78. Joost (1992, chapter 7).
 Notes to pages 296–299 393

79. Sloan and Githens (1994).

80. Landes (1982); Kochanowski and Young (1985); and Lund and Zador
(1986).
81. Sloan, Reilly, and Schenzler (1994, 1995).
82. Brown (1985).
83. Cohen and Dehejia (2003); Cummins, Phillips, and Weiss (2001).
84. Loughran (2001).
85. Gaudry (1986); Devlin (1990).
86. Lemstra and Olszynski (2005).
87. Gellhorn (1988); E. Cohen and Korper (1976); Palmer (1979); Rosenthal
(1988).
88. Fallberg and Borgenhammar (1997).
89. Essinger (2005).
90. Danzon (1994).
91. Fallberg and Borgenhammar (1997).
92. Danzon (1994).
93. Ibid.
94. Ibid.; Studdert and Brennan (2001).
95. Espersson (2000).
96. Ibid.
97. Danzon (1994).
98. Compensability is available for treatment-related infections only. Examples
of uncompensable infections are postsurgical infections of the intestines, trachea,
and tissue of diminished vitality. Also, infection from prolonged catheterization
or drainage is not compensable.
99. Danzon (1994).
100. Ibid.; Fallberg and Borgenhammar (1997); Studdert and Brennan (2001);
Espersson (2005).
101. Sappideen (1993).
102. This is also referred to as the Patient Claims Panel. Claimants may skip
this step and proceed directly to a court of general jurisdiction, though it is in
their interest to have their claim evaluated by a panel of experts before proceed-
ing to a costly judicial process (Espersson 2000).
103. Danzon (1994).
104. Ibid.; Studdert and Brennan (2001).
105. Fallberg and Borgenhammar (1997).
106. Danzon (1994).
107. Ibid.; Fallberg and Borgenhammar (1997).
394 Notes to pages 299–301

108. Espersson (2000).

109. Ibid.
110. Swedish Medical Association (2003).
111. Physician involvement includes alerting patients to possible medical inju-
ries, referring patients to social workers, and aiding patients in the filing of their
complaint (Studdert and Brennan 2001).
112. Danzon (1994).
113. Studdert and Brennan (2001).
114. Ibid.
115. Ibid.
116. Danzon (1994).
117. Ibid.
118. (Miller 1993). Justice Woodhouse was also involved in the formation of
Australia’s no-fault system. Australia’s Northern Territory has used a no-fault
system for compensating motor vehicle injuries since 1979. During the early
1970s, Justice Woodhouse performed an inquiry and wrote a comprehensive
report for Australia regarding a possible national compensation scheme
(O’Connell and Partlett 1988). Woodhouse, who helped New Zealand develop
its system, recommended a scheme for compensating victims of both accidents
and illnesses. His report also suggested that Australia abolish concurrent access
to the tort system and institute more funding and better structures for accident
prevention (O’Connell and Partlett 1988). Legislation to adopt the report was
submitted to the Parliament and quickly passed in the House of Representatives.
However, approval from the Senate was also necessary. The legislation was
submitted to the Senate and was immediately sent to committee; it was not
introduced for vote until several years later, at which point the governor pro-
rogued Parliament and the bill lapsed (O’Connell and Partlett 1988). In recent
years, Australia has been reconsidering implementing a medical malpractice no-
fault system similar to New Zealand’s.
119. New Zealand has had a no-fault system for workers’ compensation since
the 1900s. In 1928 a no-fault automobile program was added, and in 1972, the
no-fault scheme was expanded to include all personal injuries.
120. Hitzhusen (2005); Weiss (2004); Flood (2000).
121. Bismark and Paterson (2006).
122. Lowes (2003).
123. Studdert, Thomas, Zbar, et al. (1997). Only a few years after the 1992
reform, the Accident Rehabilitation and Compensation Insurance Act was
repealed and replaced with the Accident Insurance Act of 1998. The substance
of the revised no-fault program remained mostly unchanged, except for its partial
privatization (Todd 2000). Employers and self-employed persons were now
allowed to purchase insurance from either private companies or the state-owned
 Notes to pages 299–312 395

company, eliminating the monopoly of the Accident Compensation Corporation

(ACC). After the 1992 and 1998 reforms, only a small percentage of injury
victims who might have been eligible for compensation actually applied for cov-
erage (Todd 2000). However, most of the claims filed were identified as adverse
events and assessed by the ACC to be compensable (P. Davis, Lay-Yee, Fitzjohn,
et al. 2002).
124. Studdert and Brennan (2001).
125. Weiss (2004).
126. Ferguson (2003).
127. Hitzhusen (2005).
128. Bismark and Paterson (2006).
129. Paterson (2004).
130. Bismark, Dauer, Paterson, et al. (2006).
131. Ibid.
132. P. Davis, Lay-Yee, Scott, et al. (2003).
133. Hitzhusen (2005). This was not always so. Accident victims were placed
ahead of those with illness for compensation, in order to encourage a fast recov-
ery which would allow them to return to work more quickly. This lengthened
the wait for those filing claims under illness (Flood 2000).
134. Cunningham (2004b, p.5).
135. Ibid.
136. Cunningham (2004a, p. 2).
137. Rosenthal (1988).
138. Kupeli (1996).
139. Danzon (2000, p. 1394).

Chapter 12

1. Symposium on Medical Malpractice (1975, p. 1177).

2. The chapters in parentheses in this chapter refer to chapters in this book.
3. Arlen (2006); Havighurst (1995).
4. T. Baker (2005b, p. 172).
5. Danzon (1986).
6. Abraham and Weiler (1994); Bovbjerg and Berenson (2006).
7. E.g., Sharkey (2006).
8. See, e.g., Kersh (2006).
9. Mechanic (1975, p. 1195).
10. Mehlman (2006) makes this point.
396 Notes to pages 314–320

11. Several promising opportunities are discussed in chapter 8.

12. See chapter 9 for further discussion of alternative ownership forms.
13. This is also a reason that physicians often object to talking to managed care
personnel on the telephone, seeking permission to perform a test or admit a
patient to the hospital.
14. These points are made by Abraham and Weiler (1994, p. 400), among
others.
15. At least to our knowledge, he is the first to use this term (T. Baker 2006).
16. Some borderline coverage issues might arise. For example, a physician may
have misdiagnosed a condition in his or her office, which led to an error while
the same patient was hospitalized. This type of error should be assigned to the
enterprise, since the hospital has a duty to implement safeguards to identify the
possibility of such errors before they occur.
17. The United States General Accounting Office (1987) reports that about 80
percent of the claims closed in 1984 involved an injury that occurred in a
hospital.
18. For example, the Risk Management Foundation has covered Harvard
medical institutions and physicians since 1976. Medical malpractice insurance
coverage is provided by Controlled Risk Insurance Company, Ltd. (CRICO) and
Controlled Risk Insurance Company of Vermont, a risk retention group (www.
rmf.harvard.edu/company/about-us.aspx, accessed August 14, 2006). Abraham
and Weiler (1994) describe CRICO’s scrutinizing of cases that had generated
suits and payments arising from anesthesia procedures in the 1980s. Initially, the
new standards that were introduced were opposed by some physicians as “cook-
book” medicine. But Harvard hospitals’ administrators required implementation
of standards which resulted in reductions of anesthesia-related mishaps and
claims.
19. The simulations were performed by Sloan and Hassan (1990).
20. McCarran-Ferguson Act, 15 USCS §1012.
21. Liability Risk Retention Act, 15 USCS §3902. Risk retention groups are
discussed further below.
22. T. Baker (2006, p. 287).
23. Starr (1982).
24. For example, is it not strange that a person brought to an emergency room
where he or she receives many diagnostic tests would be billed separately by the
hospital and by the physicians who interpret the tests? This is a common practice
at most hospitals in the U.S. In most cases, the patient will have had no direct
contact with the test interpreters. They are “retained” only in the sense that the
patient may have agreed that the tests be performed.
25. Sloan and Hassan (1990).
26. This point is made by Bovbjerg and Berenson (2006, p. 240).
 Notes to pages 320–323 397

27. Based on several citations provided in the article, Mello, Kelly, Studdert,
et al. (2003) conclude that at the time the article was published, there were 100
hospital captives operating in the United States.
28. Christopherson (1996).
29. Koviak (2004).
30. Christopherson (1996).
31. Of the 3,400 captives created worldwide, nearly 1,850 have a U.S. sponsor
(ibid.).
32. Ibid.
33. P.L. 97–45.
34. The Risk Retention Amendments of 1986.
35. Congress provides some protection to group members by giving U.S. district
courts the authority to enjoin risk retention groups from the business of insur-
ance upon finding the group in a financially hazardous condition. 15 U.S. Code
§3906.
The rationale for relaxing regulatory requirements was that the groups do
not sell insurance to consumers, but only to their own members, and regulatory
oversight by a single state is likely to be sufficient for this reason. Members of
the risk groups must be related to each other, similar to businesses being in
common through related trade, product, services, premises, or operation. Groups
cannot exclude anyone from membership solely in order to gain a competitive
advantage (Geiger 1997; Sloan, Bovbjerg, and Githens 1991).
36. Sloan, Bovbjerg, and Githens (1991).
37. Sage, Hastings, and Berenson (1994).
38. In 1995 the Court of Appeals for the District of Arizona held that even if a
hospital cannot be liable under respondeat superior, it may still be liable under
ostensible agency (Joslin v. Yuma Regional Medical Ctr., 1995 U.S. App. LEXIS
31614).
39. For example, in Menzie v. Windham Community Memorial Hospital, the
court recognized an apparent agency cause of action, but granted summary judg-
ment for the defendant because the plaintiff could not establish his reliance on
the apparent agency relationship with the physician (774 F. Supp. 91, D. Conn.
1991).
40. See Simmons v. St. Clair Memorial Hospital, 332 Pa. Super. 444 (1984);
Sword v. NKC Hospitals, Inc., 714 N.E. 2d 142 (Ind. 1999); Leconche v.
Elligers, 1991 Conn. Super. LEXIS 1693. A recent case in Arizona has criticized
the position that merely providing care is sufficient to establish an apparent
agency relationship. The court held that some action of the principal is required;
having staff privileges alone is not enough. Henry v. Flagstaff Med. Ctr., 132
P.3d 304 (Ariz. Ct. App. 2006).
41. Abraham and Weiler (1994, pp. 390–391). In Colorado, courts decided the
contrary; physicians are the principal and the hospital is the agent, leaving the
398 Notes to pages 323–330

physician liable for the acts of nurses and staff below him or her. Krane v. St.
Anthony Hospital Systems, 738 P.2d 75 (Colo. Ct. App. 1987).
42. Tappan (2005).
43. Abraham and Weiler (1994).
44. Ibid.
45. Thompson v. Nason Hospital, 591 A.2d 703 (Pa. 1991), followed by Welsh
v. Bulger, 698 A.2d 581 (Pa. 1997). See also Johnson v. Misericordia Community
Hospital, 99 Wis. 2d 708, 735 (Wis. 1981).
46. In U.S. hospitals, before a physician can secure admitting privileges, the
medical staff of the hospital is charged with reviewing the physician’s credentials
and other qualifications.
47. Beginning with Darling v. Charleston Community Memorial Hosp., 211
N.E.2d 253 (Ill. 1965), courts have held hospitals directly liable for failing to
supervise or credential physicians. Later courts have added to the duties, includ-
ing failure to ensure the safety and availability of facilities and equipment, setting
policies that interfere with a physician’s independent medical judgment, and
failing to monitor and oversee the treatment both prescribed and administered
by its physicians. Humana Medical Corp. v. Traffanstedt, 597 So. 2d 667 (Ala.
1992); Muse v. Charter Hosp., 117 N.C. App. 468, 474 (N.C. Ct. App. 1995);
Pedroza v. Bryant, 101 Wn.2d 226 (Wash. 1984).
48. Graham v. Barolat, 2004 U.S. Dist. LEXIS 23567 (E.D. Pa. Nov. 17,
2004).
49. Abraham and Weiler (1994).
50. Steves (1976, pp. 1324–1325).
51. Sage (1997).
52. Thornton v. Ware County Hosp. Auth., 215 Ga. App. 276 (1994).
53. Schleier v. Kaiser Foundation Health Plan of the Mid-Atlantic States, Inc.,
876 F.2d 174 (1989), holding an HMO liable for the actions of one of its physi-
cians, despite the fact the physician was contracted and not employed by the
hospital.
54. Sage (1997, p. 169).
55. Tappan (2005).
56. Sage, Hastings, and Berenson (1994).
57. Danzon (2000, p. 1378).
58. See, e.g., Mello and Brennan (2002).
59. Abraham and Weiler (1994).
60. Ibid. (p. 427).
61. Sloan and Eesley (2006).
62. An adhesion contract is typically a standard form or boilerplate contract
 Notes to pages 330–335 399

entered into by parties with unequal bargaining power. For example, a consumer
renting a car must sign the contract as is; the only other option is to find another
car rental agency that offers more favorable terms. In most cases, the industry
(e.g., airlines, credit cards, banks) utilizes standard contracts, giving the con-
sumer no choice as to the terms and no option to negotiate. The consumer may
be able to go to another car rental agency, but the likelihood that the terms of
the contract would be different is small. Courts treat these contracts as they treat
any other contract, and the contract is upheld unless it is determined to be
unconscionable. This has been interpreted as the “absence of meaningful choice
on the part of one party due to one-sided contract provisions, together with terms
which are so oppressive that no reasonable person would make them and no fair
and honest person would accept them.” Fanning v. Fritz’s Pontiac-Cadillac-
Buick, 322 S.C. 399, 403 (S.C. 1996).
63. Developments in the Law (1995). Readers should consult this article for a
much more detailed discussion of these issues.
64. Mehlman (2006).
65. See chapter 5.
66. T. Baker (2005b).
References

Abraham, Kenneth S. 1986. Distributing Risk: Insurance, Legal Theory, and

Public Policy. New Haven, CT: Yale University Press.
———. 2001. The Trouble with Negligence. Vanderbilt Law Review 54 (3):
1187–1224.
Abraham, Kenneth S., and Paul C. Weiler. 1994. Enterprise Medical Liability
and the Evolution of the American Health Care System. Harvard Law Review
108: 381–436.
Advisory Commission on Childhood Vaccines. 2003. 54th Meeting of the
Advisory Commission on Childhood Vaccines (ACCV) and conference call
(Mar. 5).
———. 2004. 56th Meeting of the Advisory Commission on Childhood Vaccines
(ACCV) and conference call (Mar. 16).
Ajdari, Zohreh, and Oliver Fein. 1998. Primary Care in the United Kingdom and
the United States. Archives of Family Medicine 7: 311–314.
Alberini, A., A. Hunt, and A. Markandya. 2006. Willingness to Pay to Reduce
Mortality Risks: Evidence from a Three-Country Contingent Valuation Study.
Environmental & Resource Economics 33 (2): 251–264.
Albert, Tanya. 2003. A Tale of Two States. American Medical News 46 (18):
9–10.
American Medical Association, Specialty Society Medical Liability Project.
1988. A Proposed Alternative to the Civil Justice System for Resolving Medical
Liability Disputes: A Fault-Based Administrative System. Connecticut Medicine
52 (6): 347–350.
Ameringer, Carl F. 1999. State Medical Boards and the Politics of Public Protec-
tion. Baltimore: Johns Hopkins University Press.
Anderson, Dan R. 1976. All Risks Rating Within a Catastrophe Insurance
System. Journal of Risk and Insurance 43 (4): 629–651.
Aranson, Allison F. 1992. The United States Percentage Contingent Fee System:
Ridicule and Reform from an International Perspective. Texas International Law
Journal 27: 755–794.
402 References

Arlen, Jennifer. 2006. Private Contractual Alternatives to Malpractice Liability.

In Medical Malpractice and the U.S. Health Care System, edited by William M.
Sage and Rogan Kersh. Cambridge: Cambridge University Press.
Arnould, Richard J., and Len M. Nichols. 1983. Wage-Risk Premiums and
Workers’ Compensation: A Refinement of Estimates of Compensating Wage
Differential. Journal of Political Economy 91 (2): 332–340.
Attarian, D. E., and T. P. Vail. 2005. Medicolegal Aspects of Hip and Knee
Arthroplasty. Clinical Orthopedics and Related Research 433: 72–76.
Avraham, R. 2006. Putting a Price on Pain-and-Suffering Damages: A Critique
of the Current Approaches and a Preliminary Proposal for Change. Northwest-
ern University Law Review 100 (1): 87–119.
Babb, B. A. 1998. Fashioning an Interdisciplinary Framework for Court Reform
in Family Law: A Blueprint to Construct a Unified Family Court. Southern Cali-
fornia Law Review 71 (3): 469–546.
Baicker, Katherine, and Amitabh Chandra. 2004. The Effect of Malpractice
Liability on the Delivery of Health Care. NBER Working Paper no. 10709.
Cambridge, MA: NBER. Available at http://www.nber.org/papers/w10709.
(accessed June 22, 2007).
Baker, L. A. 2001–2002. Facts About Fees: Lessons for Legal Ethics. Texas Law
Review 80: 1985–1995.
Baker, T. 2005a. Medical Malpractice and the Insurance Underwriting Cycle.
DePaul Law Review 54: 393–438.
———. 2005b. The Medical Malpractice Myth. Chicago: University of Chicago
Press.
———. 2006. Medical Malpractice Insurance Reform: “Enterprise Insurance”
and Some Alternatives. In Medical Malpractice and the U.S. Health Care System,
edited by William M. Sage and Rogan Kersh. Cambridge: Cambridge University
Press.
Barker, Drucilla K. 1992. The Effects of Tort Reform on Medical Malpractice
Insurance Markets: An Empirical Analysis. Journal of Health, Politics, Policy
and Law 17 (1): 143–161.
Barry, John E., and Bert W. Rein. 1999. The Case for Abolishing Contingent
Fee Arrangements Working Paper. Washington, DC: Washington Legal
Foundation.
Bates, David W., Lucian L. Leape, David J. Cullen, Nan Laird, Laura A. Petersen,
Jonathan M. Teich, Elizabeth Burdick, Mairead Hickey, Sharon Kleesfield, Brian
Shea, Martha Vander Vliet, and Diane L. Sieger. 1998. Effect of Computerized
Physician Order Entry and a Team Intervention on Prevention of Serious
Medication Errors. Journal of the American Medical Association 280 (15):
1311–1316.
Becker, G. S. 1983. Theory of Competition Among Pressure Groups for Political
Influence. Quarterly Journal of Economics 98: 371–400.
 References 403

Beider, P., and S. Hagen. 2004. Limiting Tort Liability for Medical Malpractice.
Congressional Budget Office. Available at http://www.cbo.gov/showdoc.cfm?
index=4968&sequence=0 (accessed June 21, 2007).
Best, A. M. 1998. Best’s Aggregates & Averages: Property-Casualty. Oldwick,
NJ: A.M. Best.
Biddle, Jeff, Karen Roberts, Kenneth D. Rosenman, and Edward M. Welch.
1998. What Percentage of Workers with Work-Related Illnesses Receive Workers’
Compensation Benefits? Journal of Occupational and Environmental Medicine
40 (4): 325–331.
Birkmeyer, J. D. 2003. The Leapfrog Group’s Patient Safety Practices: The
Potential Benefits of Universal Adoption. Available at http://www.leapfroggroup
.org/media/file/Leapfrog-Birkmeyer.pdf (accessed June 22, 2007).
Bismark, Marie, Edward Dauer, Ron Paterson, and David Studdert. 2006.
Accountability Sought by Patients Following Adverse Events from Medical Care:
The New Zealand Experience. Canadian Medical Association Journal 175 (8):
889–894.
Bismark, Marie, and Ron Paterson. 2006. No-Fault Compensation in New
Zealand: Harmonizing Injury Compensation, Provider Accountability, and
Patient Safety. Health Affairs 25 (1): 278–283.
Blackmon, B. G., and R. Zeckhauser. 1991. Mispriced Equity—Regulated
Rates for Auto Insurance in Massachusetts. American Economic Review 81 (2):
65–69.
Blair, E., and F. J. Stanley. 1988. Cerebral Palsy in Low-Birthweight Infants.
Developmental Medicine & Child Neurology 30 (4): 550–552.
Blumberg, L. J., and J. Holahan. 2004. Government as Reinsurer: Potential
Impacts on Public and Private Spending. Inquiry 41: 130–143.
Blumenthal, D. 2004. New Steam from an Old Cauldron: The Physician-Supply
Debate. New England Journal of Medicine 350 (17): 1780–1787.
Blumstein, J., and F. A. Sloan. 1988. Antitrust and Hospital Peer Review. Law
and Contemporary Problems 51: 7–92.
Blumstein, J., F., Randall R. Bovbjerg, and F. A. Sloan. 1991. Beyond Tort
Reform: Developing Better Tools for Assessing Damages for Personal Injury.
Yale Journal of Regulation 8 (1): 171–212.
Born, P., and W. K. Viscusi. 1994. Insurance Market Responses to the 1980s
Liability Reforms: An Analysis of Firm-Level Data. Journal of Risk and Insur-
ance 61 (2): 192–218.
Born, P. H., W. Kip Viscusi, and D. W. Carleton. 1998. The Distribution of
the Insurance Market Effects on Tort Liability Reforms. Brookings Papers on
Economic Activity: 1998: 55–105.
Born, Patricia, W. Kip Viscusi, and Tom Baker. 2006. The Effects of Tort
Reform on Medical Malpractice Insurers’ Ultimate Losses. Harvard Law and
404 References

Economics Discussion Paper No. 554. Available at SSRN: http://ssrn.com/

abstract=921441 (accessed June 22, 2007).
Born, Patricia H. 2001. Insurer Profitability in Different Regulatory and Legal
Environments. Journal of Regulatory Economics 19 (3): 211–237.
Bovbjerg, R. R. 1991. Lessons for Tort Reform from Indiana. Journal of Health
Politics, Policy, and Law 16 (3): 465–482.
Bovbjerg, R. R., F. A. Sloan, and J. F. Blumstein. 1989. Valuing Life and Limb
in Tort: Scheduling Pain and Suffering. Northwestern University Law Review
83 (4): 908–976.
Bovbjerg, R. R., L. R. Tancredi, and D. S. Gaylin. 1991. Obstetrics and Mal-
practice: Evidence on the Performance of a Selective No-Fault System. Journal
of the American Medical Association 265 (21): 2836–2843.
Bovbjerg, Randall. 1989. Legislation on Medical Malpractice: Further Develop-
ments and a Preliminary Report Card. University of California-Davis Law
Review 22 (2): 499–556.
———. 1991. Lessons for Tort Reform from Indiana. Journal of Health Politics,
Policy, and Law 16 (3): 465–483.
———. 1995. Medical Malpractice: Problems & Reforms. Washington, DC:
Urban Institute.
Bovbjerg, Randall R., Pablo Aliaga, and Josephine Gittler. 2006. State Discipline
of Physicians: Assessing State Medical Boards Through Case Studies. Washing-
ton, DC: Urban Institute.
Bovbjerg, Randall R., and Anna Bartow. 2003. Understanding Pennsylvania’s
Medical Malpractice Crisis: Facts About Liability Insurance, the Legal
System, and Health Care in Pennsylvania. In The Project on Medical Liability
in Pennsylvania for the Pew Charitable Trusts. New York: Columbia
University.
Bovbjerg, Randall R., and Robert A. Berenson. 2006. Enterprise Liability in the
21st Century. In Medical Malpractice and the U.S. Health Care System, edited
by William M. Sage and Rogan Kersh. Cambridge: Cambridge University
Press.
Bovbjerg, R. R., and F. A. Sloan. 1998. No-Fault for Medical Injury. University
of Cincinnati Law Review 67: 53–123.
Bovbjerg, R. R., F. A. Sloan, A. Dor, and C. R. Hsieh. 1991. Juries and Justice:
Are Malpractice and Other Personal Injuries Created Equal? Law and Contem-
porary Problems 54: 5–42.
Bradley, E. W., J. Herrin, B. Elbel, R. J. McNamara, D. J. Magid, B. K.
Nallamoutu, Y. Wang, S. L. T. Normand, J. A. Spertus, and H. M. Krumholz.
2006. Hospital Quality for Acute Myocardial Infarction: Correlation Among
Process Measures and Relationship with Short-Term Mortality. Journal of the
American Medical Association 296 (1): 72–78.
 References 405

Brennan, T. A., Lucian L. Leape, N. M. Laird, L. Hebert, R. Localio, A. G.

Lawters, J. P. Newhouse, P. C. Weiler, and H. H. Hiatt. 1991. Incidence of
Adverse Events and Negligence in Hospitalized Patients. New England Journal
of Medicine 324 (8): 370–376.
Brennan, Troyen A., and Michelle M. Mello. 2003. Patient Safety and
Medical Malpractice: A Case Study. Annals of Internal Medicine 139 (4):
267–273.
Breyer, S. 1982. Regulation and Its Reform. Cambridge, MA: Harvard Univer-
sity Press.
Brickman, Lester. 1989. Contingent Fees Without Contingencies: Hamlet Without
the Prince of Denmark? University of California-Los Angeles Law Review 37:
29–138.
———. 1996. ABA Regulation of Contingency Fees: Money Talks, Ethics Walks.
Fordham Law Review 65: 247–336.
———. 2003. Effective Hourly Rates of Contingency-Fee Lawyers: Competing
Data and Non-Competitive Fees. Washington University Law Quarterly 81 (3):
653–669.
Brooks, R. G., N. Menachemi, A. Clawson, and L. Beitsch. 2005. Availability
of Physician Services in Florida, Revisited: The Effect of the Professional Liability
Insurance Market on Access to Health Care. Archives of Internal Medicine 165
(18): 2136–2141.
Brown, C. 1985. Deterrence in Tort and No-Fault: The New Zealand Experi-
ence. California Law Review 73: 976–1002.
Browne, Mark J., and Brenda P. Wells. 1999. Claims Adjudication in the
Personal Automobile Insurance Residual Market. Journal of Risk and Insurance
66 (2): 275–290.
Bruce, S., H. Prior, A. Katz, M. Taylor, S. Latosinsky, P. Martens, C. De Coster,
M. Brownell, R, Soodeen, and C. Steinbach. 2006. Application of Patient Safety
Indicators in Manitoba: A First Look. Manitoba Centre for Health Policy.
Calfee, J. E., C. Winston, and K. Viscusi. 1993. The Consumer Welfare Effects
of Liability for Pain and Suffering: An Exploratory Analysis. Brookings Papers
on Economic Activity, Microeconomics 1993 (1): 133–196.
Campion, E. W. 2003. Death at Duke. New England Journal of Medicine 348
(12): 1083–1084.
Campolieti, M., and D. E. Hyatt. 2006. Further Evidence on the “Monday
Effect” in Workers’ Compensation. Industrial & Labor Relations Review 59 (3):
438–450.
Cebula, Richard J. 1995. The Impact of Federal Deposit Insurance on Savings
and Loan Failures: Reply [to Ira S. Saltz]. Southern Economic Journal 62 (1):
256–259. (Saltz on p. 253.)
Choi, Albert. 2003. Allocating Settlement Authority Under a Contingent-Fee
Arrangement. Journal of Legal Studies 32: 585–610.
406 References

Choi, Seungmook, Don Hardigree, and Paul D. Thistle. 2002. The Property/
Liability Insurance Cycle: A Comparison of Alternative Models. Southern
Economic Journal 68 (3): 530–548.
Christopherson, J. A. 1996. The Captive Medical Malpractice Insurance
Company. Annals of Health Law 5: 121–143.
Chu, K. 2005. FEMA Halts Flood Insurance Payments. USA Today, Nov. 17.
Chupkovich, Patricia J. 1993. Statutory Caps: An Involuntary Contribution to
the Medical Malpractice Insurance Crisis or a Reasonable Mechanism for
Obtaining Affordable Health Care? Jounal of Contemporary Health Law Policy
9: 337–376.
Cloud, Morgan. 1996. The Fourth Amendment During the Lochner Era: Privacy,
Property, and Liberty in Constitutional Theory. Stanford Law Review 48 (3):
555–631.
Cohen, Alma, and Dehejia, Rajeev. 2003. The Effect of Automobile Insurance
and Accident Liability Laws on Traffic Fatalities.
Cohen, Eva D., and Samuel P. Korper. 1976. The Swedish No-Fault Patient
Compensation Program: Provisions and Preliminary Findings. Insurance Law
Journal 637: 70.
Cohen, M. A., and T. R. Miller. 2003. “Willingness to Award” Nonmonetary
Damages and the Implied Value of Life from Jury Awards. International Review
of Law and Economics 23: 165–181.
Cohen, Thomas H., and Steven K. Smith. 2004. Civil Trial Cases and Verdicts
in Large Counties, 2001. Washington, D.C.: A.S. Department of Justice, Office
of Justice Programs, Bureau of Justice Statistics.
Coldiron, B., A. H. Fisher, E. Adelman, C. B. Yelverton, R. Balkirshnan, M. A.
Feldman, and S. R. Feldman. 2005. Adverse Event Reporting: Lessons Learned
from Four Years of Florida Office Data. Dermatologic Surgery 31 (9):
1079–1093.
Cook, Phillip J., and Daniel A. Graham. 1977. The Demand for Insurance
and Protection: The Case of Irreplaceable Commodities. Quarterly Journal of
Economics 91 (1): 143–156.
Coolidge, Thomas Jefferson. 1905. Jefferson in His Family. Vol. 15 of The
Writings of Thomas Jefferson. Edited by A. Lipscomb and A. Bergh.
Washington, DC: Thomas Jefferson Memorial Association of the United
States.
Copland, James R. 2004. Contingency Fees: Format, Links Down. Available
at http://www.pointoflaw.com/feature/fee_ding_frenzy.php. (cited Sept. 25,
2006).
Corrigan, Janet, Ann Greiner, and Shari M. Erickson, eds. 2002. Fostering Rapid
Advances in Health Care: Learning from System Demonstrations. Washington,
D.C.: National Academics Press. Also available at http://www.nap.edu/books/
0309087074/html. (accessed June 22, 2007).
 References 407

Costich, J. F. 1994. Joint State-Federal Regulation of Lawyers: The Case

of Group Legal Services under ERISA. Kentucky Law Journal 82: 627–
659.
Council of State Governments. 2003. Medical Malpractice Crisis. Trends Alert.
Lexington: The Council.
Craig, Diane E. , Liz Hartka, William H. Likosky, William M. Caplan, Paul
Litsky, and Jannalee Smithey. 1999. Implementation of a Hospitalist System in
a Large Health Maintenance Organization: The Kaiser Permanente Experience.
Annals of Internal Medicine 130 (4): 355–359.
Croley, Steven P., and Jon D. Hanson. 1995. The Nonpecuniary Costs of
Accidents: Pain-and-Suffering Damages in Tort Law. Harvard Law Review
108 (8): 1785–1898.
Cummins, J. David. 1988. Incorporating Risk in Insurance Guaranty Fund
Premiums. In Workers’ Compensation Insurance Pricing: Current Programs and
Proposed Reforms, edited by P. S. Borba and D. Appel. Dordrecht: Kluwer
Academic.
Cummins, J. David, and Patricia M. Danzon. 1991. Price Shocks and Capital
Flows in Liability Insurance. In Cycles and Crises in Property/Casualty Insur-
ance: Causes and Implications for Public Policy, edited by J. D. Cummins, S. E.
Harrington, and R. W. Klein. Kansas City, MO: National Association of Insur-
ance Commissioners.
Cummins, J. David, and J. Francois Outreville. 1987. An International Analysis
of Underwriting Cycles. Journal of Risk and Insurance 54 (2): 246–262.
Cummins, J. David, Richard D. Phillips, and Mary A. Weiss. 2001. The Incentive
Effects of No-Fault Automobile Insurance. Journal of Law and Economics 44
(2): 427–464.
Cunningham, Wayne. 2004a. The Medical Complaints and Disciplinary Process
in New Zealand: Doctors’ Suggestions for Change. Journal of the New Zealand
Medical Association 117 (1198): 1–9.
———. 2004b. New Zealand Doctors’ Attitudes Towards the Complaints and
Disciplinary Process. Journal of the New Zealand Medical Association 117
(1198): 1–9.
Cutler, David M. , N. E. Feldman, and J. R. Horwitz. 2005. U.S. Adoption
of Computerized Physician Order Entry System. Health Affairs 24 (6):
1654–1664.
D’Arcy, Stephen P. 1986. Legislative Reform of the Medical Malpractice Tort
System in Illinois. Journal of Risk and Insurance 53 (3): 538–550.
Daly, R. 2005. Voluntary System to Collect Medical-Error Data. American
Psychiatric Association 40 (17): 11.
Dana, James D., and Kathryn E. Spier. 1993. Expertise and Contingent Fees:
The Role of Asymmetric Information in Attorney Compensation. Journal of
Law, Economics, and Organization 9 (2): 349–367.
408 References

Daniels, Stephen, and Joanne Martin. 2002. It Was the Best of Times, It Was
the Worst of Times: The Precarious Nature of Plaintiffs’ Practice in Texas. Texas
Law Review 80: 1781–1828.
———. 2006. Plaintiffs’ Lawyers, Specialization, and Medical Malpractice.
Vanderbilt Law Review 59 (4): 1051–1073.
Dann, B. M. 2002. Jurors as Beneficiaries of Proposals to Objectify Proof of
Standard of Care in Medical Malpractice Cases. Wake Forest Law Review 37:
943–952.
Danzon, P. M. 1984. Tort Reform and the Role of Government in Private Insur-
ance Markets. Journal of Legal Studies 13 (3): 517–549.
———. 1985. Liability and Liability Insurance for Medical Malpractice. Journal
of Health Economics 4: 309–331.
———. 1986. The Frequency and Severity of Medical Malpractice Claims: New
Evidence. Law and Contemporary Problems 49 (2): 57–84.
———. 2000. Liability for Medical Malpractice. In Handbook of Health
Economics, edited by A. J. Culyer and J. P. Newhouse. Amsterdam: Elsevier.
Danzon, P. M., M. V. Pauly, and R. Kingston. 1990. The Effects of Malpractice
Litigation on Physicians’ Fees and Incomes. American Economic Review 80:
122–127.
Danzon, Patricia M. 1994. Alternative Liability Regimes for Medical Injuries:
Evidence from Simulation Analysis. Journal of Risk and Insurance 61 (2):
219–244.
Danzon, Patricia, and Lee A. Lillard. 1983. Settlement out of Court: The
Disposition of Medical Malpractice Claims. Journal of Legal Studies 12:
345–377.
Danzon, Patricia M. 1991. Liability for Medical Malpractice. Journal of
Economic Perspectives 5 (3): 51–69.
———. 1994. The Swedish Patient Compensation System: Myths and Realities.
International Review of Law and Economics 14: 453–466.
Danzon, Patricia M. 1983. Contingent Fees for Personal Injury Litigation. Bell
Journal of Economics 14 (1): 213–224.
Dao, James. 2005. G. O. P. Push in States to Curb Malpractice Costs. The New
York Times, Jan. 14, p. 1.
Darmiento, L. 2005. California Keeps Malpractice Rates in Check. Los Angeles
Business Journal, Feb. 14, p. 12.
Darragh, T. 2002a. Medical Malpractice Insurers Struggle in Pennsylvania. The
Morning Call, Allentown, PA, Apr. 21.
———. 2002b. Risky Insurance Company Move Also Fuels Pa. Malpractice
Crisis. The Morning Call, Allentown, PA, Apr. 21.
Davis, Peter, Roy Lay-Yee, Julie Fitzjohn, Phil Hider, Robin Briant, and Stephan
Schug. 2002. Compensation for Medical Injury in New Zealand: Does
 References 409

“No-Fault” Increase the Level of Claims Making and Reduce Social and
Clinical Selectivity? Journal of Health Politics, Policy, and Law 27 (5):
833–854.
Davis, Peter, Roy Lay-Yee, Alastair Scott, Robin Briant, and Stephan Schug.
2003. Acknowledgement of “No Fault” Medical Injury: Review of Patients’
Hospital Records in New Zealand. British Medical Journal 326 (7380):
79–80.
Davis, W. Kent. 1999. The International View of Attorney Fees in Civil
Suits: Why Is the United States the “Odd Man Out” in How It Pays Its
Lawyers? Arizona Journal of International and Comparative Law 16:
361–436.
Davis, W. N. 2003. Special Problems for Specialty Courts: Clients Get Needed
Treatment Rather Than Jail Time, but Prosecutors and Defense Lawyers Alike
Worry About Compromising Their Roles as Advocates. American Bar Associa-
tion Journal 89: 32–37.
———. 2006. No Longer Immune?: Court Opens Door to Cases Claiming Link
Between Autism and Vaccine Preservative. American Bar Association Journal 92:
19, 43.
de Sa e Silva, Marco. 1988. Constitutional Challenges to Washington’s Limit on
Noneconomic Damages in Cases of Personal Injury and Death. Washington Law
Review 63 (3): 653–675.
Developments in the Law: Confronting the New Challenges of Scientific Evi-
dence. 1995. Harvard Law Review 108 (7): 1481–1605.
Developments in the Law: VI. Unified Family Courts and the Child Protection
Dilemma. 2003. Harvard Law Review 116 (7): 2099–2122.
Devers, Kelly J., Hoangmai H. Pham, and Gigi Liu. 2004. What Is Driving
Hospitals’ Patient-Safety Efforts? Health Affairs 23 (2): 103–114.
Devlin, R. A. 1990. Some Welfare Implications of No-Fault Automobile Insur-
ance. International Review of Law and Economics 10: 193–205.
Di Pietro, S., and T. W. Carns. 1996. Alaska’s English Rule: Attorney’s Fee
Shifting in Civil Cases. Alaska Law Review 13: 33–96.
Diamond, Herbert S., Elliot Goldberg, and Janine E. Janosky. 1998. The Effect
of Full-Time Faculty Hospitalists on the Efficiency of Care at a Community
Teaching Hospital. Annals of Internal Medicine 129 (3): 197–203.
Doherty, N. A., and J. R. Garven. 1995. Insurance Cycles: Interest-Rates and
the Capacity Constraint Model. Journal of Business 68 (3): 383–404.
Doherty, N. A., and L. Posey. 1997. Availability Crises in Insurance Markets:
Optimal Contracts with Asymmetric Information and Capacity Constraints.
Journal of Risk and Uncertainty 15 (1): 55–80.
Doherty, N. A., and K. Smetters. 2005. Moral Hazard in Reinsurance Markets.
Journal of Risk and Insurance 72 (3): 375–391.
410 References

Doherty, Neil A., Joan Lamm-Tennant, and Laura T. Starks. 2003. Insuring
September 11th: Market Recovery and Transparency. Journal of Risk and Uncer-
tainty 26 (2/3): 179–199.
Domitrovich, S. 1998. Utilizing an Effective Economic Approach to Family
Court: A Proposal for a Statutory Unified Family Court in Pennsylvania.
Duquesne University Law Review 37 (1): 1–66.
Dorsey, S., and N. Walzer. 1983. Workers’ Compensation, Job Hazards, and
Wages. Industrial and Labor Relations Review 36 (4): 642–654.
Downs, David H., and David W. Sommer. 1999. Monitoring, Ownership, and
Risk-Taking: The Impact of Guaranty Funds. The Journal of Risk and Insurance
66 (3): 477–497.
Dranove, David, and Anne Gron. 2005. Has the Malpractice Crisis in
Florida Really Affected Access to Care? working paper. Northwestern
University.
Dranove, David, Daniel Kessler, Mark McClellan, and Mark Satterthwaite.
2003. Is More Information Better? The Effects of “Report Cards” on Health
Care Providers. Journal of Political Economy 111 (3): 555–588.
Dubay, Lisa, Robert Kaestner, and Timothy Waidmann. 1999. The Impact of
Malpractice Fears on Cesarean Section Rates. Journal of Health Economics 18
(4): 491.
Dubner, Stephen, and Steven D. Levitt. 2006. Freakonomics: Selling Soap. The
New York Times, Sept. 24.
Eisenberg, Theodore, John Goerdt, Brian Ostrom, David Rottman, and Martin
T. Wells. 1997. The Predictability of Punitive Damages. Journal of Legal Studies
26: 623–661.
Eisenberg, Theodore, Neil LaFountain, Brian Ostrom, David Rottman, and
Martin T. Wells. 2002. Juries, Judges, and Punitive Damages: An Empirical
Study. Cornell Law Review 87: 743–782.
Ellis, Randall P., Cynthia L. Gallup, and Thomas G. McGuire. 1990. Should
Medical Professional Liability Insurance Be Experience Rated? Journal of Risk
and Insurance 57 (1): 66–78.
Ellsworth, Phoebe C., and Alan Reifman. 2000. Juror Comprehension and Public
Policy: Perceived Problems and Proposed Solutions. Psychology, Public Policy,
and Law 6 (3): 788–821.
Emons, Winand. 2000. Expertise, Contingent Fees, and Insufficient Attorney
Effort. International Review of Law and Economics 20: 21–33.
Encinosa, William E., and Fred J. Hellinger. 2005. Have State Caps on Malprac-
tice Awards Increased the Supply of Physicians? Health Affairs 24 (Web Exclu-
sives): 250–258. Available at: http://content.healthaffairs.org/webexclusives/index
.dtl?year=2005 (accessed June 22, 2007).
Entman, S. S., C. A. Glass, G. B. Hickson, P. B. Githens, K. Whetten-Goldstein,
and F. A. Sloan. 1994. The Relationship Between Malpractice Claims History
 References 411

and Subsequent Obstetric Care. Journal of the American Medical Association

272 (20): 1588–1591.
Epstein, R. 1978. Medical Malpractice: Its Cause and Cure. In The Economics
of Medical Malpractice, edited by S. Rottenberg. Washington, DC: American
Enterprise Institute Press.
Eskin, David J. 2003. Prepared Witness Testimony, Hearing Before the U.S.
House Committee on Energy and Commerce (Feb. 10). Abington, PA: Abington
Memorial Hospital.
Espersson, Carl. 2000. The Patient Injury Act: A Comment by Carl Espersson.
Available at http://www.patientforsakring.se/infoglueDeliverLive/digital
Assets/2129_Engelska_comments.pdf.
———. 2005. The Swedish Patient Injury Act. Transcript by Federal News
Service, Washington, DC 2005. Available at www.patientforsakring.se/infoglue
DeliverLive/digitalAssets/2129_Engelska_comments.pdf.
Essinger, K. 2005. The Regions’ Mutual Insurance Company for Patient In-
juries (LOF). at. Available http://www.patientforsakring.se/infoglueDeliverLive/
digitalAssets/2129_Engelska_comments.pdf.
Evans, G. 1998. Vaccine Liability and Safety Revisited. Archives of Pediatric and
Adolescent Medicine 152: 7–10.
Fairley, W. B. 1979. Investment Income and Profit Margins in Property-Liability
Insurance: Theory and Empirical Results. Bell Journal of Economics 10:
192–210.
Fallberg, L. H., and E. Borgenhammar. 1997. The Swedish No Fault Patient
Insurance Scheme. European Journal of Health Law 4: 279–286.
Farber, Henry S., and Max Bazerman. 1986. The General Basis of Arbitrator
Behavior: An Empirical Analysis of Conventional and Final-Offer Arbitration.
Econometrica 54 (4): 819–844.
Farber, Henry S., and Michelle J. White. 1991. Medical Malpractice: An Empiri-
cal Examination of the Litigation Process. Rand Journal of Economics 22:
199–217.
Ferguson, J. 2003. Medical Misadventure Under Accident Compensation: Diag-
nosis and Treatment of a Problem. New Zealand Law Review part IV: 485.
Fielding, Stephen L. 1990. The Social Construction of the Medical Malpractice
Crisis: A Case Study of Massachusetts Physicians. Sociological Forum 5 (2):
279–295.
Fields, G. 2006. In Brooklyn Court, a Route Out of Jail for the Mentally Ill.
Wall Street Journal, Aug. 21, pp. A1, A8.
Finkelstein, Amy. 2004. Static and Dynamic Effects of Health Policy: Evidence
from the Vaccine Industry. Quarterly Journal of Economics 119 (2): 527.
Finkelstein, Joel B. 2003. Frist Bill Further Safeguards Doctors from Vaccine
Lawsuits. American Medical News 46 (16): 5.
412 References

Fishback, P. V., and S. E. Kantor. 1998a. The Political Economy of Workers’

Compensation Benefit Levels, 1910–1930. Explorations in Economic History 35
(2): 109–139.
———. 1998b. The Adoption of Workers’ Compensation in the United States,
1900–1930. Journal of Law and Economics 41 (2): 305–341.
Fitzpatrick, Sean M. 2003–2004. Fear Is the Key: A Behavioral Guide to Under-
writing Cycles. Connecticut Insurance Law Journal 10 (2): 255.
Flood, Colleen M. 2000. New Zealand’s No-Fault Accident Compensation
Scheme: Paradise or Panacea? Health Law Review 8 (3): 1–9.
Flowers, Lynda, and Trish Riley. 2001. State-Based Mandatory Reporting of
Medical Errors: An Analysis of the Legal and Policy Issues. Portland, ME:
National Academy for State Health Policy.
Folberg, J. 1999. Family Courts: Assessing the Trade-offs. Family and Concilia-
tion Courts Review 37 (4): 448–453.
Fournier, Gary M., and Melayne Morgan McInnes. 2001. The Case for Experi-
ence Rating in Medical Malpractice Insurance: An Empirical Evaluation. The
Journal of Risk and Insurance 68 (2): 255–276.
Frank, Ted. 2006. Hyman and Silver: “Medical Malpractice Litigation and Tort
Reform: It’s the Incentives, Stupid.” Available at http://www.pointoflaw.com/
archives/003181.php (cited December 20, 2006).
Franko, Frederick P. 2001. State Laws and Regulations for Office-Based Surgery.
Association of Operating Room Nurses 73 (4): 843–846.
Freeman, Andrew D., and John M. Freeman. 1989. No-Fault Cerebral Palsy
Insurance: An Alternative to the Obstetrical Malpractice Lottery. Journal of
Health Politics, Policy, and Law 14 (4): 707–718.
Freese, Robert B. 1999. The Park Nicollet Experience in Establishing a Hospital-
ist System. Annals of Internal Medicine 130: 350–354.
Froot, Kenneth A. 2001. The Market for Catastrophe Risk: A Clinical Examina-
tion. Journal of Financial Economics 60 (2): 529–571.
Froot, Kenneth, and Paul G. J. O’Connell. 1997. The Pricing of U.S. Catastrophe
Reinsurance. In National Bureau of Economic Research Working Paper 6043.
Cambridge, MA: NBER, pp. 1–36.
Fuchs, V. R. 1978. Supply of Surgeons and Demand for Operations. Journal of
Human Resources 13: 35–56.
Fung, Hung-Gay, Gene C. Lai, Gary A. Patterson, and Robert C. Witt. 1998.
Underwriting Cycles in Property and Liability Insurance: An Empirical
Analysis of Industry and By-line Data. Journal of Risk and Insurance 65 (4):
539–561.
Galanter, Marc. 1993. The Regulatory Function of the Civil Jury. In Verdict:
Assessing the Civil Jury System, edited by R. E. Litan. Washington, DC: The
Brookings Institution.
 References 413

Gan, T. J., F. Sloan, G. D. Dear, H. E. El-Moalem, and D. A. Lubarsky. 2001.

How Much Are Patients Willing to Pay to Avoid Postoperative Nausea and
Vomiting? Anesthesia and Analgesia 92 (2): 393–400.
Garoupa, Nuno, and Fernando Gomez-Pomar. 2002. Cashing by the Hour: Why
Large Law Firms Prefer Hourly Fees over Contingent Fees. UPF Working Paper
No. 639. Available at SSRN: http://ssrn.com/abstract=394305 (accessed June 22,
2007).
Gash, J. 2005. Solving the Multiple Punishments Problem: A Call for a National
Punitive Damages Registry. Northwestern University Law Review 99 (4):
1613–1686.
Gaudry, M. 1986. Measuring the Effects of the 1978 Quebec Automobile Insur-
ance Act and the DRAG Model. Montreal: Dept. of Economics, University of
Montreal, Publication #493, Sept. Mimeo.
Geiger, R. S. Risk Retention Groups: Preemption of State Law 1997. Available
at: http://library.findlaw.com/1997/Dec/1/126440.html (accessed Aug. 16,
2006).
Geistfeld, M. 1995. Placing a Price on Pain and Suffering: A Method for Helping
Juries Determine Tort Damages for Nonmonetary Damages. California Law
Review 83: 773–852.
Gellhorn, Walter. 1988. Medical Malpractice Litigation (U.S.)—Medical Mishap
Compensation (N.Z.). Cornell Law Review 73 (2): 170–212.
Geraghty, A. H., and W. J. Myniec. 2002. Unified Family Courts: Tempering
Enthusiasm with Caution. Family Court Review 40 (4): 435–452.
Ghandi, Tejal K. 2005. Fumbled Handoffs: One Dropped Ball After Another.
Annals of Internal Medicine 142 (5): 352–358.
Glassman, Adam D. 2004. The Imposition of Federal Caps in Medical Malprac-
tice Liability Actions: Will They Cure the Current Crisis in Health Care? Akron
Law Review 37: 417.
Gold, M. R., J. E. Siegel, L. B. Russell, and Milton C. Weinstein, eds. 1996. Cost-
Effectiveness in Health and Medicine. New York: Oxford University Press.
Goldstein, J. 2003. Collapse Spreads Misery. The Philadelphia Inquirer, Mar. 2,
Section E, page 1.
Grabowski, H., W. K. Viscusi, and W. N. Evans. 1989. Price and Availability
Tradeoffs of Automobile Insurance Regulation. Journal of Risk and Insurance
56 (2): 275–299.
Greene, Mark R. 1976. The Government as an Insurer. Journal of Risk and
Uncertainty 43 (3): 393–407.
Grisham, John. 1996. The Runaway Jury. New York: Doubleday.
Gron, Anne. 1990. Property-Casualty Insurance Cycles, Capacity Constraints,
and Empirical Results Ph. D. dissertation. Department of Economics. MIT,
Cambridge, MA.
414 References

———. 1994a. Capacity Constraints and Cycles in Property-Casualty Insurance

Markets. Rand Journal of Economics 25 (1): 110–127.
———. 1994b. Evidence of Capacity Constraints in Insurance Markets. Journal
of Law & Economics 37: 349.
Gron, Anne, and Deborah Lucas. 1998. External Financing and Insurance Cycles.
In The Economics of Property-Casualty Insurance, edited by D. F. Bradford.
Chicago: University of Chicago Press.
Gronfein, William P., and Eleanor DeArman Kinney. 1991. Controlling Large
Malpractice Claims: The Unexpected Impact of Damage Caps. Journal of Health
Politics, Policy, and Law 16 (3): 465–483.
Grossman, Richard S. 1992. Deposit Insurance, Regulation, and Moral Hazard
in the Thrift Industry: Evidence from the 1930’s. American Economic Review
82 (4): 800–822.
Gruber, J., and Alan Krueger. 1991. Tax Policy and the Economy. Cambridge,
MA: MIT Press.
Gunnar, William P. 2004. Is There an Acceptable Answer to Rising Medical
Malpractice Premiums? Annals of Health Law 13: 465.
Hall, Bronwyn-H., Clint Cummins, Elizabeth S. Laderman, and Joy Mundy.
1988. The R&D Master File Documentation. NBER Technical Working Paper
0072. Cambridge, MA: NBER.
Hallinan, Joseph T. 2004. In Malpractice Trials, Juries Rarely Have the Last
Word. Wall Street Journal, Nov. 30.
Haltom, William, and Michael W. McCann. 2004. Distorting the Law: Politics,
Media, and the Litigation Crisis. Chicago: University of Chicago Press.
Han, Yong Y., Joseph A. Carcillo, Shekhar T. Venkataraman, Robert S. B. Clark,
R. Scott Watson, Trung C. Nguyen, Hulya Bayir, and Richard A. Orr. 2005.
Unexpected Increased Mortality After Implementation of a Commercially Sold
Computerized Physician Order Entry System. Pediatrics 116 (6): 1506–1512.
Harrington, Scott E. 1994. State Decisions to Limit Tort Liability: An Empirical
Analysis of No-Fault Automobile Insurance Laws. Journal of Risk and Insurance
61 (2): 276–294.
———. 2002. Repairing Insurance Markets. Regulation 25 (2): 58–63.
Harrington, Scott E., and Robert E. Litan. 1988. Causes of the Liability Insur-
ance Crisis. Science 239: 737–741.
Hart, O., A. Shleifer, and R. W. Vishny. 1997. The Proper Scope of Government:
Theory and an Application to Prisons. Quarterly Journal of Economics 112 (4):
1127–1161.
Havighurst, C. C. 1995. Health Care Choices: Private Contracts as Instruments
of Health Reform. Washington, DC: American Enterprise Institute Press.
Havighurst, Clark C. 1997. Making Health Plans Accountable for the Quality
of Care. Georgia Law Review 31: 587–649.
 References 415

Hay, Bruce L. 1996. Contingent Fees and Agency Costs. Journal of Legal Studies
25 (2): 503–533.
———. 1997. Optimal Contingent Fees in a World of Settlement. Journal of
Legal Studies 26: 259–278.
Health Resources and Services Administration. 2006. Vaccine Injury Table.
Available at http://www.hrsa.gov/vaccinecompensation/table.htm (accessed Sept.
2006).
Heid, B., and E. Misulovin. 2000. The Group Legal Plan Revolution: Bright
Horizon or Dark Future? Hofstra Labor and Employment Law Journal 18:
335–365.
Helland, Eric, and Alexander Taborrok. 2000. Runaway Judges? Selection
Effects and the Jury. Journal of Law, Economics, and Organization 16 (2):
306–333.
———. 2003. Race, Poverty, and American Tort Awards. Journal of Legal
Studies 32: 27–58.
Hellinger, Fred J., and William E. Encinosa. 2003. The Impact of State Laws
Limiting Malpractice Awards on the Geographic Distribution of Physicians.
Washington, DC: U.S. Department of Health and Human Services.
Hersch, Joni, and W. Kip Viscusi. 2004. Punitive Damages: How Judges and
Juries Perform. Journal of Legal Studies 33: 1–29.
Hertzka, Robert E. 2003. State Medical-Malpractice Law Works. The San Diego
Union-Tribune, B7.
Hickson, G. B., E. W. Clayton, S. S. Entman, C. S. Miller, P. B. Githens, K.
Whetten-Goldstein, and F. A. Sloan. 1994. Obstetricians’ Prior Malpractice
Experience and Patients’ Satisfaction with Care. Journal of the American Medical
Association 272 (20): 1583–1587.
Hickson, G. B., E. W. Clayton, P. B. Githens, and F. A. Sloan. 1992. Factors
That Prompted Families to File Medical Malpractice Claims Following
Perinatal Injuries. Journal of the American Medical Association 267 (10):
1359–1363.
Hickson, Gerald B., Charles F. Federspiel, James W. Pichert, Cynthia S. Miller,
Jean Gauld-Jaeger, and Preston Bost. 2002. Patient Complaints and Malpractice
Risk. Journal of the American Medical Association 287 (22): 2951–2957.
Hill, R. D. 1979. Profit Regulation in Property-Liability Insurance. Bell Journal
of Economics 10: 172–191.
Hitzhusen, M. 2005. Crisis and Reform: Is New Zealand’s No-Fault Compensa-
tion System a Reasonable Alternative to the Medical Malpractice Crisis in
the United States? Arizona Journal of International and Comparative Law 22:
649–689.
Hoerger, T., F. A. Sloan, and M. Hassan. 1990. Loss Volatility, Bankruptcy,
and Insurer Demand for Reinsurance. Journal of Risk and Uncertainty 3:
221–245.
416 References

Hofflander, Alfred E., and Blaine F. Nye. 1985. Medical Malpractice Insurance
in Pennsylvania. Menlo Park, CA: MAC Group.
Hofflander, Alfred E., and Jane D. Nettesham. 2001. Report on the Medical
Malpractice Insurance Delivery System in Pennsylvania. Redwood City, CA:
Standford Consulting Group.
Hospital and Healthsystem Association of Pennsylvania. 2002a. Survey of
Professional Liability Coverage: Findings of Statewide Survey. Harrisburg, PA:
Health and Hospital Association of Pennsylvania.
———. 2002b. Professional Liability Coverage in Pennsylvania: Findings of a
Statewide Survey. Harrisburge, PA: Health and Hospital Association of
Pennsylvania.
Hughes, E. F., V. R. Fuchs, J. E. Jacoby, and E. M. Lewit. 1972. Surgical Work
Loads in a Community Practice. Surgery 71 (3): 315–327.
Hughes, James W., and Edward A. Snyder. 1995. Litigation and Settlement Under
the English and American Rules: Theory and Evidence. Journal of Law and
Economics 38 (1): 225–250.
Hyman, David A., and Charles Silver. 2005. The Poor State of Health Care
Quality in the U.S.: Is Malpractice Liability Part of the Problem or Part of the
Solution? Cornell Law Review 90 (4): 893–994.
———. 2006. Medical Malpractice Litigation and Tort Reform: It’s the Incen-
tives, Stupid. Vanderbilt Law Review 59 (4): 1085–1136.
Inselbuch, Elihu. 2001. Contingent Fees and Tort Reform: A Reassessment
and Reality Check. Law and Contemporary Problems 64 (Spring/Summer):
175–196.
Institute of Medicine. 1997. Board on Health Promotion and Disease Prevention:
Vaccine Safety Forum, Summaries of Two Workshops.
———. 2000. To Err is Human: Building a Safer Health System. Washington,
DC: National Academies Press.
———. 2004. Immunization Safety Review: Vaccines and Autism. Washington,
DC: National Academy Press. Also available at http://books.nap.edu/
catalog/10997.html.
———. 2001. Crossing the Quality Chasm: A New Health System for the 21st
Century. Washington, D.C.: National Academy Press.
Institute of Medicine, Janet Corrigan, Ann Greiner, and Shari M. Erickson. 2002.
Fostering Rapid Advances in Health Care Learning from System Demonstra-
tions. Washington, DC: National Academies Press.
Jacobi, John V., and Nicole Huberfeld. 2001. Quality Control, Enterprise Liabil-
ity, and Disintermediation in Managed Care. Journal of Law, Medicine and
Ethics 29 (3/4): 305–322.
 References 417

Jacobson, Peter D., and M. Gregg Bloche. 2005. Improving Relations Between
Attorneys and Physicians. Journal of the American Medical Association 294 (16):
2083–2085.
Jagsi, Reshma, Barrett T. Kitch, Debra F. Weinstein, Eric G. Campbell, Matthew
Hutter, and Joel S. Weissman. 2005. Residents Report on Adverse Events and
Their Causes. Archives of Internal Medicine 165 (22): 2607–2613.
Jin, Ginger Zhe, and Phillip Leslie. 2003. The Effect of Information on Product
Quality: Evidence from Restaurant Hygiene Grade Cards. Quarterly Journal of
Economics 118 (2): 409–451.
Joint Legislative Audit and Review Commission of the Virginia General Assem-
bly. 2002. Review of the Virginia Birth-Related Neurological Injury Compensa-
tion Program. Richmond: Common wealth of Virginia.
Joost, Robert H. 1992. Automobile Insurance and No-Fault Law. 2nd ed. West
Publishing Co. Deerfield, IL. Clark Boardman Callaghan.
Kaboli, Peter J., Mitchell J. Barnett, and Gary E. Rosenthal. 2004. Associations
with Reduced Length of Stay and Costs on an Academic Hospitalist Service.
American Journal of Managed Care 10 (8): 561–568.
Kagan, Robert A. 2001. Adversarial Legalism: The American Way of Law.
Cambridge, MA: Harvard University Press.
Kakalik, J. S., D. R. Hensler, D. McCaffrey, M. Oshiro, N. M. Pace, and M. E.
Valana. 1998. Discovery Management: Further Analysis of the Civil Justice
Reform Act Evaluation Data. Santa Monica, CA: Rand Corporation.
Kakalik, J. S., and Nicholas M. Pace. 1986. Costs and Compensation Paid in
Tort Litigation. Santa Monica, CA: RAND Corporation.
Karpoff, Jonathan M., and John R. Lott. 1999. On the Determinants and
Importance of Punitive Damage Awards. Journal of Law and Economics 42:
527–572.
Kaufman, Allan M., and Thomas A. Ryan. 2000. Strategic Asset Allocation for
Multi-Line Insurers Using Dynamic Financial Analysis. Casualty Actuarial
Society Forum: Dynamic Financial Analysis Call Papers, 1–20 Available at http://
www.casact.org/pubs/forum/00sforum/00sf001.pdf (accessed Jan. 20, 2005).
Kaushal, R., K. G. Shojania, and D. W. Bates. 2003. Effects of Computerized
Physician Order Entry and Clinical Decision Support Systems on Medication
Safety. Archives of Internal Medicine 163: 1409–1416.
Keeley, Michael C. 1990. Deposit Insurance, Risk, and Market Power in Banking.
American Economic Review 80 (5): 1183–1200.
Keisler, Peter D. 2004. Statement Before the Subcommittee on Commercial and
Administrative Law, Committee on the Judiciary, on Budget and Resource Needs
of the Justice Department Civil Division for Fiscal Year 2005. United States
House of Representatives, Mar. 9. Available at http://www.house.gov/judiciary/
keisler030904.htm (accessed Dec. 21, 2004).
418 References

Kelly, C. N., and M. M. Mello. 2005. Are Medical Malpractice Damages Caps
Constitutional? An overview of state litigation. Journal of Law, Medicine and
Ethics 33 (3): 515–534.
Kenney, Roger K. 1988. Financial Condition of Medical Malpractice JUAs.
Schaumburg, IL: Alliance of American Insurers.
Kersh, R. 2006. Medical Malpractice and the New Politics of Health Care. In
Medical Malpractice and the U.S. Health Care System, edited by W. M. Sage
and R. Kersh. Cambridge: Cambridge University Press.
Kesselheim, A. S., T. G. Ferris, and D. M. Studdert. 2006. Will Physician-Level
Measures of Clinical Performance Be Used in Medical Malpractice Litigation?
Journal of the American Medical Association 295 (15): 1831–1834.
Kessler, D. P., W. M. Sage, and D. J. Becker. 2005. Impact of Malpractice
Reforms on the Supply of Physician Services. Journal of the American Medical
Association 293 (21): 2618–2625.
Kessler, Daniel, and Mark McClellan. 1996. Do Doctors Practice Defensive
Medicine? The Quarterly Journal of Economics 111 (2): 353–390.
Kessler, Daniel P., and Mark McClellan. 1997. The Effects of Malpractice Pres-
sure and Liability Reforms on Physicians’ Perceptions of Medical Care. Law and
Contemporary Problems 60 (1): 81–106.
Kilian, M. 2003. Alternatives to Public Provision. The Role of Legal Expenses
Insurance in Broadening Access to Justice: The German Experience. Journal of
Law and Society 30 (1): 31–48.
Kingdon, John. 1981. Congressmen’s Voting Decisions, 2nd ed. New York:
Harper & Row.
Kinney, Eleanor D. 1995. Malpractice Reform in the 1990’s: Past Disappoint-
ments, Future Success? Journal of Health Politics, Policy, and Law 20 (1):
109–135.
Klein, Robert W. 1995. Insurance Regulation in Transition. Journal of Risk and
Insurance 62 (3): 363–404.
———. 1998. Regulation and Catastrophe Insurance. In Paying the Price: The
Status and Role of Insurance Against Natural Disasters in the United States,
edited by H. Kunreuther and R. J. Roth, Sr. Washington, DC: Joseph Henry
Press.
Klick, Jonathan, and Thomas Stratmann. 2003. Does Medical Malpractice
Reform Help States Retain Physicians and Does It Matter? Working paper under
review by Journal of Legal Studies.
Kochanowski, Paul S., and Madelyn V. Young. 1985. Deterrent Aspects of
No-Fault Automobile Insurance: Some Empirical Findings. Journal of Risk and
Insurance 52 (20): 269.
Kondo, LeRoy L. 2001. Advocacy of the Establishment of Mental Health Spe-
cialty Courts in the Provision of Therapeutic Justice for Mentally Ill Offenders.
American Journal of Criminal Law 28 (3): 255–336.
 References 419

Koppel, R., J. P. Metlay, A. Cohen, B. Abaluck, A. R. Localio, S. E. Kimmel,

and B. L. Strom. 2005. Role of Computerized Physician Order Entry Systems in
Facilitating Medication Errors. Journal of the American Medical Association 293
(10): 1197–1203.
Kosanovich, A. 2006. One Family in Two Courts: Coordination for Families in
Illinois Juvenile and Domestic Relations Courts. Loyola University Chicago Law
Journal 37 (3): 571–616.
Koviak, N. W. 2004. An Insurance Perspective on the Medical Malpractice
Crisis. Annals of Health Law 13 (2): 607–616.
Kraus, Alan, and Stephen A. Ross. 1982. The Determination of Fair Profits for
the Property-Liability Insurance Firm. Journal of Finance 37 (4): 1015–1028.
Kritzer, Herbert M. 1997. Contingency Fee Lawyers as Gatekeepers in the
American Civil Justice System. Judicature 81 (1): 22–29.
———. 2002. Lawyer Fees and Lawyer Behavior in Litigation: What Does the
Empirical Literature Really Say? Texas Law Review 80: 1943–1983.
———. 2004. Risks, Reputations, and Rewards: Contingency Fee Legal Practice
in the United States. Stanford, CA: Stanford University Press (Stanford Law and
Politics).
Krumlauf, Clare Elizabeth. 2005. Ohio’s New Modified Joint and Several Liabil-
ity Laws: A Fair Compromise for Competing Parties and Public Policy Interests.
Cleveland State Law Review 53 (2): 333–358.
Kunreuther, Howard. 1998. Insurability Conditions and the Supply of Coverage.
In Paying the Price: The Status and Role of Insurance Against Natural Disasters
in the United States, edited by H. Kunreuther and R. J. Roth, Sr. Washington,
DC: Joseph Henry Press.
Kunreuther, Howard, M. Pauly, and T. Russell. 2004. Demand and Supply Side
Anomalies in Catastrophe Insurance Markets: The Role of the Public and Private
Sectors. Paper prepared for the MIT/LSE/Cornell Conference on Behavioral
Economics, London, May 2004.
Kunreuther, Howard C. 2006. Has the Time Come for Comprehensive Natural
Disaster Insurance? In On Risk and Disaster: Lessons from Hurricane Katrina,
edited by R. Daniels, D. F. Kettl, and H. C. Kunreuther. Philadelphia: University
of Pennsylvania Press.
Kupeli, M. A. 1996. Tort Law = No Fault Compensation: An Unrealistic Elixir
to the Medical Malpractice Ailment. Suffolk Transnational Law Review 19:
559–572.
Kyl, J. 2006. Meaningful Health Care Reform Begins with Health Courts.
Washington, D.C.: Republican Policy Committee.
Lai, Gene C., Robert C. Witt, Hung-Gay Fung, Richard D. MacMinn, and
Patrick L. Brockett. 2000. Great (and Not So Great) Expectations: An Endoge-
nous Economic Explication of Insurance Cycles and Liability Crises. Journal of
Risk and Insurance 67 (4): 617–652.
420 References

Landes, Elisabeth M. 1982. Insurance, Liability, and Accidents: A Theoretical

and Empirical Investigation of the Effect of No-Fault Accidents. Journal of Law
and Economics 25 (1): 49–65.
Landro, Laura. 2006. Hospitals Combat Errors at the ‘Hand-Off’. The Wall
Street Journal, June 28, pp. D1–D2.
Laro, D. 1995. The Evolution of the Tax Court as an Independent Tribunal.
University of Illinois Law Review 1995 (1): 17–29.
Leape, Lucian L., T. A. Brennan, N. Laird, A. G. Lawthers, A. R. Localio, B. A.
Barnes, L. Hebert, J. P. Newhouse, P. C. Weiler, and H. Hiatt. 1991. The
Nature of Adverse Events in Hospitalized Patients: Results of the Harvard
Medical Practice Study II. New England Journal of Medicine 324 (6):
377–384.
Lee, Soon-Jae, David Mayers, and Clifford W. Smith. 1997. Guaranty Funds and
Risk-taking Evidence from the Insurance Industry. Journal of Financial Econom-
ics 44 (1): 3–24.
Leigh, J. Paul, and John A. Robbins. 2004. Occupational Disease and Workers’
Compensation: Coverage, Costs, and Consequences. The Milbank Quarterly 82
(4): 689–721.
Lemaire, Jean. 1985. Automobile Insurance: Actuarial Models. Boston:
Kluwer-Nijhoff.
Lempert, Richard. 1993. Civil Juries and Complex Cases: Taking Stock After
Twelve Years. In Verdict: Assessing the Civil Jury System, edited by R. E. Litan.
Washington, DC: The Brookings Institution.
———. 1999. Juries, Hindsight, and Punitive Damage Awards: Failures of a
Social Science Case for Change. DePaul Law Review 48: 867.
Lemstra, M., and W. P. Olszynski. 2005. The Influence of Motor Vehicle Legisla-
tion on Injury Claim Incidence. Canadian Journal of Public Health 96 (1):
65–68.
Leonning, C. D. 2001. D. C. Family Court Funding Splits Officials: Norton,
DeLay Say Budget Request to Congress Was Inadequate, Mishandled. The
Washington Post, Sept. 22, p. B2.
Lerman, Lisa G. 1999. Blue-Chip Bilking: Regulation of Billing and Expense
Fraud by Lawyers. Georgetown Journal of Legal Ethics 12: 205–366.
Lesser, Steven B., Edward R. Blumberg, Janice P. Brown, Michael V. Ciresi,
Thomas A. Demetrio, Lewis H. Goldfarb, Michael S. Hull, Perry K. Huntington,
Robert Johnson, Daniel M. Klein, Marc S. Moller, Charles M. Silver, Patrick E.
Longan, and D. Christopher Wells. 2004. Report on Contingent Fees in Medical
Malpractice Litigation. American Bar Association.
Levinson, W., D. L. Roter, J. P. Mullooly, V. T. Dull, and R. M. Frankel. 1997.
Physician-Patient Communication: The Relationship with Malpractice Claims
Among Primary Care Physicians and Surgeons. Journal of the American Medical
Association 277 (7): 553–559.
 References 421

Levinson, Wendy. 1994. Physician-Patient Communication: A Key to Malprac-

tice Prevention. Journal of the American Medical Association 272 (20):
1619–1620.
Library, Institute of Governmental Studies, University of California. 2004.
Workers’ Compensation in California 2004. Available at http://www.
igs.berkeley.edu/library/htWorkersCompensation.htm (accessed Dec. 30,
2004).
Lichtenstein, Sarah, Paul Slovic, Baruch Fischhoff, Mark Layman, and Barbara
Combs. 1978. Judged Frequency of Lethal Events. Journal of Experimental
Psychology: Human Learning and Memory 4 (6): 551–581.
Lloyd-Puryear, Michele A., Leslie K. Ball, and David Benor. 1998. Should the
Vaccine Injury Compensation Program Be Expanded to Cover Adults? Public
Health Reports 113: 236–242.
Localio, A. R., A. G. Lawthers, J. M. Bengtson, L. E. Hebert, S. L. Weaver,
T. A. Brennan, and J. R. Landis. 1993. Relationship Between Malpractice Claims
and Cesarean Delivery. Journal of the American Medical Association 269 (3):
366–373.
Longo, Daniel R., John F. Hewett, Bin Ge, and Shari Schubert. 2005. The Long
Road to Patient Safety: A Status Report on Patient Safety Systems. Journal of
the American Medical Association 294 (22): 2858–2865.
Loughran, David S. 2001. The Effect of No-Fault Automobile Insurance on
Driver Behavior and Automobile Accidents in the United States. Santa Monica,
CA: RAND Institute for Civil Justice.
Lowes, Robert. 2003. Malpractice: Do Other Countries Hold the Key? Medical
Economics 80 (14): 58–60.
Luft, H. S. 1980. The Relation Between Surgical Volume and Mortality: An
Exploration of Causal Factors and Alternative Models. Medical Care 18 (9):
940–959.
Luft, H. S., J. P. Bunker, and A. C. Enthoven. 1979. Should Operations Be
Regionalized: Empirical Relation Between Surgical Volume and Mortality. New
England Journal of Medicine 301 (25): 1364–1369.
Lund, Adrian, and Paul Zador. 1986. Re-Analyses of the Effects of No-Fault
Auto Insurance on Fatal Crashes. Journal of Risk and Insurance 53: 226–
342.
MacCoun, Robert. 1993. Inside the Black Box: What Empirical Research Tells
Us About Decisionmaking by Civil Juries. In Verdict: Assessing the Civil Jury
System, edited by R. E. Litan. Washington, DC: The Brookings Institution.
Mankiw, N. Gregory. 2007. Principles of Economics, 4th ed. Mason, OH:
Thomson South-Western.
Marchev, Mimi. 2003. Medical Malpractice and Medical Error Disclosure:
Balancing Facts and Fears. Portland, ME: National Academy for State
Health Policy.
422 References

Marchev, Mimi, J. Rosenthal, and M. Booth. 2003. How States Report Medical
Errors to the Public: Issues and Barriers. Portland, ME: National Academy for
State Health Policy.
Margolis, Howard. 1982. A Thought Experiment on Demand-Revealing Mecha-
nisms. Public Choice 38 (1): 87–91.
Marjoribanks, Timothy, Mary-Jo Delvecchio Good, Ann G. Lawthers, and Lynn
M. Peterson. 1996. Physicians’ Discourses on Malpractice and the Meaning of
Medical Malpractice. Journal of Health and Social Behavior 37 (2): 163–178.
Matsa, D. 2007. Does Liability Keep the Doctor Away? Evidence from Tort
Reform Damage Caps. Journal of Legal Studies 36.
Maute, Judith L. 2001. Pre-Paid and Group Legal Services: Thirty Years After
the Storm. Fordham Law Review 70: 915–944.
May, M. L., and D. B. Stengel. 1990. Who Sues Their Doctors? Law and Society
Review 24: 105–120.
McCaffery, Edward J., Daniel J. Kahneman, and Matthew L. Spitzer. 1995.
Framing the Jury: Cognitive Perspectives on Pain and Suffering Awards. Virginia
Law Review 81 (5): 1341–1420.
McCoid, J. C. II. 1991. Right to Jury Trial in Bankruptcy: Granfinanciera, S. A.
v. Nordberg. American Bankruptcy Law Journal 65 (1): 15–42.
McDonald, C. J. 2006. Computerization Can Create Safety Hazards: A Bar-
Coding Near Miss. Annals of Internal Medicine 144 (7): 510–516.
McGuire, P. 2000. Futurology as Further Ideology: Reflections on Pryor’s
Millennium Survey of Economists. American Journal of Economics and
Sociology 59 (1): 35–38.
McGuire, T. G., and M. V. Pauly. 1991. Physician Response to Fee Changes
with Multiple Payers. Journal of Health Economics 10 (4): 385–410.
McLeod, D. 2002. PHICO Execs Deny Fault for Insolvency. Business Insurance
36 (1): 3–5.
Mechanic, D. 1975. Some Social Aspects of the Medical Malpractice Dilemma.
Duke Law Journal 1975 (6): 1179–1196.
Mehlman, Maxwell J. 2006. The Shame of Medical Malpractice. Journal of
Legal Medicine 27: 17–32.
Meier, Kenneth J. 1988. The Political Economy of Regulation: The Case of
Insurance. Albany: SUNY Press.
Mello, M. M., and D. M. Studdert. 2006. The Medical Malpractice System:
Structure and Performance. In Medical Malpractice and the U.S. Health Care
System, edited by W. M. Sage and R. Kersh. Cambridge: Cambridge University
Press.
Mello, M. M., D. M. Studdert, C. M. DesRoches, J. Peugh, K. Zapert, T. A.
Brennan, and W. M. Sage. 2005. Effects of a Malpractice Crisis on Specialist
Supply and Patient Access to Care. Annals of Surgery 242 (5): 621–628.
 References 423

Mello, Michelle M., and Troyen A. Brennan. 2002. Deterrence of Medical

Errors: Theory and Evidence for Malpractice Reform. Texas Law Review 80 (7):
1595–1637.
Mello, Michelle M., Carly N. Kelly, David M. Studdert, Troyen A. Brennan, and
William M. Sage. 2003. Hospitals’ Behavior in a Tort Crisis: Observations from
Pennsylvania. Health Affairs 22 (6): 225–233.
Mello, Michelle M., David M. Studdert, and Troyen A. Brennan. 2003. The New
Medical Malpractice Crisis. New England Journal of Medicine 348 (23):
2281–2284.
Meltzer, David. 2001. Hospitalists and the Doctor-Patient Relationship. Journal
of Legal Studies 30 (1): 589–606.
Meltzer, David, W. G. Manning, J. Morrison, M. N. Shah, L. Jin, T. Guth, and
W. Levinson. 2002. Effects of Physician Experience on Costs and Outcomes on
an Academic General Medicine Service: Results of a Trial of Hospitalists. Annals
of Internal Medicine 137 (11): 866–875.
Miceli, Thomas J. 2004. The Economic Approach to Law. Stanford, CA:
Stanford University Press.
———. 1994. Do Contingent Fees Promote Excessive Litigation? Legal Studies
23: 211–224.
Michel-Kerjan, Erwann, and Nathalie de Marcellis-Warin. 2006. Public-Private
Programs for Covering Extreme Events: The Impact of Information
Distribution on Risk-Sharing. Asia-Pacific Journal of Risk and Insurance 1 (2):
21–49.
Miller, Richard S. 1993. An Analysis and Critique of the 1992 Changes to
New Zealand’s Accident Compensation Scheme. Maryland Law Review 52:
1070–1092.
Mills, D. H., J. S. Boyden, and D. S. Rubsamen. 1977. California Medical Asso-
ciation Medical Insurance Feasibility Study. San Francisco: Sutter Publications.
Mogel, Gary S. 2003. Damage Caps Tied to Lower Loss Costs: Milliman.
National Underwriter Property and Casualty-Risk and Benefits Management
107 (17): 22–24.
Moller, Erik K., Nicholas M. Pace, and Stephen J. Carroll. 1999. Punitive
Damages in Financial Injury Jury Verdicts. Journal of Legal Studies 28:
283–339.
Moore, Michael J., and W. Kip Viscusi. 1989. Promoting Safety Through
Workers’ Compensation: The Efficacy and Net Wage Costs of Injury Insurance.
Rand Journal of Economics 20 (4): 499–515.
Morgan, M. Granger, S. F. Brodsky, D. A. Butler, D. W. Ditz, L. I. Ezekoye,
H. K. Florig, D. F. Geisler, D. J. Morgan, J. D. Moteff, K. A. Perusich, S. R.
Rod, M. S. Sandilya, M. B. H. Weiss, C. F. Wiecha, B. P. Wise, L. A. Wojcik,
and H. R. Zane. 1983. On Judging the Frequency of Lethal Events: A Replica-
tion. Risk Analysis 3 (1): 11–16.
424 References

Moridaira, Soichiro, Jorge L. Urrutia, and Robert C. Witt. 1992. The Equilib-
rium Insurance Price and Underwriting Return in a Capital Market Setting.
Journal of Risk and Insurance 59 (2): 291–300.
Morton, Tom. 2003. Panel Dumps State Malpractice Insurance Idea. The Casper
Star Tribune, October 29.
Munch, P., and D. Smallwood. 1980. Solvency Regulation in the Property/
Casualty Insurance Industry. Bell Journal of Economics 11: 261–279.
Musgrave, R. 1957. Review of an Expenditure Tax. American Economic Review
47: 200–205.
Myers, Stewart C., and Richard A. Cohn. 1987. A Discounted Cash Flow
Approach to Property-Liability Insurance: Theory and Empirical Results. In Fair
Rate of Return in Property-Liability Insurance, edited by David Cummins and
Scott E. Harrington. Boston: Kluwer Academic.
Nalebuff, B. 1987. Credible Pretrial Negotiation. Rand Journal of Economics
18: 198–210.
Nalebuff, B., and D. Scharfstein. 1987. Testing in Models of Asymmetric Infor-
mation. Review of Economics and Statistics 54: 265–277.
Nathanson, Mitchell J. 2004. It’s the Economy (and Combined Ratio), Stupid:
Examining the Medical Malpractice Litigation Crisis Myth and the Factors Criti-
cal to Reform. Pennsylvania State Law Review 108: 1077.
National Association of Insurance Commissioners. 1980. Malpractice Claims:
Final Compilation, edited by M. P. Sowka.
Nelson, K. B., and J. H. Ellenberg. 1986. Antecedents of Cerebral Palsy:
Multivariate Analysis of Risk. New England Journal of Medicine 315 (2):
81–86.
New Jersey Hospital Association. 2002. Hospitals Share Stories of Doctors
Leaving and Patients’ Access to Care Threatened; NJHA Calls for Relief from
Medical Malpractice Crisis. Press Release.
New South Wales Law Reform Commission. 1984. Accident Compensation
Final Report: Transport Accidents Scheme for New South Wales, report 43. New
South Wales Government Printer, Sydney.
New York Department of Insurance. 1997. The Status of the Primary and Excess
Medical Malpractice Market and the Future Need for the Medical Malpractice
Insurance Association. Albany: New York Department of Insurance.
Newhouse, J. P., A. P. Williams, B. W. Bennett, and W. B. Schwartz. 1982a.
Does the Geographical Distribution of Physicians Reflect Market Failure? Bell
Journal of Economics 13: 493–505.
———. 1982b. Where Have All the Doctors Gone? Journal of the American
Medical Association 247 (17): 2392–2396.
Newton, G. E. 2001. Maintaining the Balance: Reconciling the Social and Judi-
cial Costs of Medical Peer Review Protection. Alabama Law Review 52 (2):
723–742.
 References 425

Niemeyer, Paul V. 2004. Awards for Pain and Suffering: The Irrational Center-
piece of Our Tort System. Virginia Law Review 90: 1401–1422.
Nordman, E., D. Cermak, and K. McDaniel. 2004. Medical Malpractice
Insurance Report: A Study of Market Conditions and Potential Solutions to
the Recent Crisis. National Association of Insurance Commissioners: Kansas
City, KS.
Norton, Stephen A. 1997. The Medical Malpractice Premium Costs of Obstet-
rics. Inquiry 34 (1): 62–69.
Nutter, Franklin W. 1985. The Second Time Around. Best’s Review 86 (4):
22.
Nye, Blaine F., and Alfred E. Hofflander. 1987. Economics of Oligopoly: Medical
Malpractice Insurance as a Classic Illustration. Journal of Risk and Insurance
54 (3): 502–519.
———. 1988. Experience Rating in Medical Professional Liability Insurance.
Journal of Risk and Insurance 55 (1): 150–157.
Nye, David J., Donald G. Gifford, Bernard L. Webb, and Marvin A. Dewar.
1988. The Causes of the Medical Malpractice Crisis: An Analysis of Claims
Data and Insurance Company Finances. Georgetown Law Journal 76:
1495–1561.
O’Connell, J. 1979. The Lawsuit Lottery: Only the Lawyers Win. New York:
Free Press.
O’Connell, J., and P. B. Bryan. 2000–2001. More Hippocrates, Less Hipocracy:
“Early Offers” as a Means of Implementing the Institute of Medicine’s Recom-
mendations on Malpractice Law. Journal of Law and Health 15: 23–52.
O’Connell, Jeffrey. 1982. Offers That Can’t Be Refused: Foreclosure of Personal
Injury Claims by Defendants’ Prompt Tender of Claimants’ Net Economic Losses.
Northwestern University Law Review 77: 589–632.
O’Connell, Jeffrey, and D. F. Partlett. 1988. An America’s Cup for Tort Reform?
Australia and America Compared. University of Michigan Journal of Law and
Reform 21: 443–487.
Office of the Governor for the State of California. 2004. Governor Schwarzeneg-
ger Signs Workers’ Compensation Legislation. Available at http://gov.ca.gov/
index.php?/press-release/3109/ (April 14, 2004; accessed Sept. 26, 2006).
Olson, Walter K. 1991. The Litigation Explosion: What Happened When
America Unleashed the Lawsuit. New York: Truman Talley BooksDutton.
———. 2003. The Rule of Lawyers: How the New Litigation Elite Threatens
America’s Rule of Law. New York: St. Martin’s Press.
Pace, Nicholas M., Daniela Golinelli, and Laura Zakaras. 2004. Capping Non-
economic Awards in Medical Malpractice Trials : California Jury Verdicts Under
MICRA. Santa Monica, CA: RAND Institute for Social Justice.
Palmer, Geoffrey. 1979. Compensation for Incapacity. Wellington, NZ: Oxford
University Press.
426 References

———. 1994. New Zealand’s Accident Compensation Scheme: Twenty Years

On. University of Toronto Law Journal 44 (3): 223–273.
Paterson, Ron. 2004. Complaints and Quality: Handle with Care! Journal of the
New Zealand Medical Association 117 (1198): 1–9.
Patient Safety Toolbox for States. 2006. National Academy for State Health
Policy. Available at http://www.pstoolbox.org/ (accessed June 22, 2007).
Pauly, M. V. 2002. Is Medical Care Different? In Competition in the Health
Care Sector: Past, Present, and Future, edited by W. Greenberg. Washington,
DC: Beard Books.
———. 2006. Who Pays When Malpractice Premiums Rise? In Medical Mal-
practice and the U.S. Health Care System, edited by W. M. Sage and R. Kersh.
Cambridge: Cambridge University Press.
Peeples, R., and C. T. Harris. 2005. Learning to Crawl: The Use of Voluntary
Caps on Damages in Medical Malpractice Litigation. Catholic University Law
Review 54 (3): 703–746.
Peeples, Ralph, Catherine T. Harris, and Thomas B. Metzloff. 2000. Settlement
Has Many Faces: Physicians, Attorneys and Medical Malpractice. Journal of
Health and Social Behavior 41 (3): 333–346.
Peltzman, Sam. 1976. Toward a More General Theory of Regulation, NBER
Technical Working Paper 133. Cambridge, MA: National Bureau of Economic
Research.
Percy, E. Farish. 2004. Checking Up on the Medical Malpractice Liability Insur-
ance Crisis in Mississippi: Are Additional Tort Reforms the Cure? Mississippi
Law Journal 73: 1001.
Perreira, K. M., and F. A. Sloan. 2002. Living Healthy and Living Long: Valuing
the Nonpecuniary Loss from Disability and Death. Journal of Risk and Uncer-
tainty 24 (1): 5–29.
Persson, Torsten, and Guido Tabillini. 2002. Political Economics and Public
Finance. In Handbook of Public Economics, edited by A. J. Auerbach and M.
Feldstein. Amsterdam: Elsevier Science.
Petersen, Laura A., Troyen A. Brennan, Anne C. O’Neil, E. Francis Cook, and
Thomas H. Lee. 1994. Does Housestaff Discontinuity of Care Increase the
Risk for Preventable Adverse Events? Annals of Internal Medicine 121 (11):
866–872.
Pfenningstorf, W., and Alec M. Schwartz, eds. 1986. Legal Protection Insurance:
American and European Approaches. Chicago: American Bar Foundation,
American Prepaid Legal Services Institute.
Pfenningstorf, Werner. 1975. Legal Expense Insurance: The European Experi-
ence in Financing Legal Services. Chicago: American Bar Association.
Pinnacle Actuarial Resources. 2003. Final Report on the Feasibility of an
Ohio Patient Compensation Fund. Available at http://www.ohioinsurance.gov/
Documents/05-01-03FinalReport.pdf (accessed June 22, 2007).
 References 427

Pogarsky, G., and L. Babcock. 2001. Damage Caps, Motivated Anchoring, and
Bargaining Impasse. Journal of Legal Studies 30 (1): 143–159.
Polinsky, A. M., and S. Shavell. 1998. Punitive Damages: An Economic Analysis.
Harvard Law Review 111 (4): 869–962.
Poon, E. G., D. Blumenthal, T. Jaggi, M. M. Honour, D. W. Bates, and R.
Kaushal. 2004. Overcoming Barriers to Adopting and Implementing Computer-
ized Physician Order Entry Systems in U.S. Hospitals. Health Affairs 23 (4):
184–190.
Posner, R. A. 1974. Theories of Economic Regulation. Bell Journal of Economics
and Management Science 5 (2): 335–358.
———. 1976. Toward a More General Theory of Regulation. Journal of Law
and Economics 19: 211–240.
Priest, George L. 1987. The Current Insurance Crisis and Modern Tort Law. Yale
Law Journal 96: 1521–1590.
———. 2002. Introduction: The Problem and Efforts to Understand It. In Puni-
tive Damages: How Juries Decide, edited by C. R. Sunstein, R. Hastie, J. W.
Payne, D. A. Schkade, and W. K. Viscusi. Chicago: University of Chicago
Press.
Rabin, Robert L. 1988. Some Reflections on the Process of Tort Reform. San
Diego Law Review 25: 13–48.
Ray, David. 1982. The Sources of Voting Cues in Three State Legislatures.
Journal of Politics 44: 1074–1087.
Reynolds, R. A., J. A. Rizzo, and M. L. Gonzalez. 1987. The Cost of Medical
Professional Liability. Journal of the American Medical Association 257 (20):
2776–2781.
Rhodes, M. S., and G. L. Ohlsson. 1997. Workers’ Compensation Answer Book.
New York: Panel Publishers.
Ridgway, Derry. 1999. No-Fault Vaccine Insurance: Lessons from the National
Vaccine Injury Compensation Program. Journal of Health Politics, Policy and
Law 24 (1): 59–90.
Riedmueller, N. J. 1973–1974. Group Legal Services and the Organized Bar.
Columbia Journal of Law and Social Problems 10: 228–263.
Robinson, Glen O. 1986. The Medical Malpractice crisis of the 1970’s: A
Retrospective. Law and Contemporary Problems 49 (2): 5–35.
Rodwin, M. A., H. J. Chang, and J. Clausen. 2006. MARKETWATCH. Mal-
practice Premiums and Physicians’ Income: Perceptions of a Crisis Conflict with
Empirical Evidence. Health Affairs 25 (3): 750–758.
Rolph, J. E. 1981. Some Statistical Evidence on Merit Rating in Medical
Malpractice Insurance. Journal of Risk and Insurance 48 (2): 247–260.
Rosenblatt, A. 2004. The Underwriting Cycle: The Rule of Six. Health Affairs
23 (6): 103–106.
428 References

Rosenthal, Marilynn M. 1988. Dealing with Medical Malpractice: The British

and Swedish Experience. Durham, NC: Duke University Press.
Rothschild, Michael, and Joseph E. Stiglitz. 1976. Equilibrium in Competitive
Insurance Markets: An Essay on the Economics of Imperfect Information.
Quarterly Journal of Economics 90 (4): 630–649.
Rottman, David. 2000. Does Effective Therapeutic Jurisprudence Require
Specialized Courts (and Do Specialized Courts Imply Specialist Judges?).
Court Review 37: 22–27.
Rubinfeld, Daniel L., and Suzanne Scotchmer. 1993. Contingent Fees for Attor-
neys: An Economic Analysis. Rand Journal of Economics 24 (3): 343–356.
Ruser, John W. 1998. Does Workers’ Compensation Encourage Hard to Diag-
nose Injuries? Journal of Risk and Insurance 65 (1): 101–124.
Rustad, Michael. 1992. In Defense of Punitive Damages in Products Liability:
Testing Tort Anecdotes with Empirical Data. Iowa Law Review 78: 1–88.
Sage, W. M. 1997. Enterprise Liability and the Emerging Managed Health Care
System. Law and Contemporary Problems 60 (2): 159–210.
Sage, W. M., K. E. Hastings, and R. A. Berenson. 1994. Enterprise Liability for
Medical Malpractice and Health Care Quality Improvement. American Journal
of Law and Medicine 20 (1–2): 1–28.
Sage, William M. 2003. Medical Liability and Patient Safety. Health Affairs 22
(4): 26–36.
Sage, William M., and James M. Jorling. 1994. A World That Won’t Stand Still:
Enterprise Liability by Private Contract. DePaul Law Review 43: 1007–1043.
Saks, Michael J. 1992. Do We Really Know Anything About the Behavior of the
Tort Litigation System—and Why Not? University of Pennsylvania Law Review
140: 1147–1292.
Saltus, R. 2005. Ending Malpractice Roulette: Do Health Courts Offer a Fairer
Way to Settle Patients’ Injury Claims? Available at http://www.hsph.
harvard.edu/review/review_fall_05/rvwfall05_malpractice.html (accessed June
22, 2007).
Sand, Leonard B., John S. Siffert, Walter P. Loughlin, Steven A. Reiss, and Nancy
Batterman. 2005. General Civil Instructions. Edited by Matthew Bender and
Company. 3rd ed. Vol. 4, Modern Federal Jury Instructions: Matthew Bender.
Original edition, 1984.
Sappideen, C. 1993. No-Fault Compensation for Medical Misadventure—
Australian Expression of Interest. Journal of Contemporary Health Law and
Policy 9: 311–322.
Saxton, Jim. 2003. Liability for Medical Malpractice: Issues and Evidence.
Washington, DC: Joint Economic Committee, United States Congress.
Schaffer, C. L., S. R. Feldman, A. B. Fleischer, M. J. Huether, and G. J. Chen.
2005. The Cutaneous Surgery Experience of Multiple Specialties in the Medicare
 References 429

Population. Journal of the American Academy of Dermatology 52 (6):

1045–1048.
Schepard, A. 2002. Law Schools and Family Court Reform. Family Court Review
40 (4): 460–472.
Scheutzow, Susan O. 1999. State Medical Peer Review: High Cost but No
Benefit—Is it Time for a Change? American Journal of Law and Medicine 25
(1): 7–60.
Schlesinger, H., and E. Venezian. 1986. Insurance Markets with Loss-Prevention
Activity Profits, Market Structure, and Consumer Welfare. Rand Journal of
Economics 17 (2): 227–238.
Schneiberg, M., and T. Bartley. 2001. Regulating American Industries: Markets,
Politics, and the Institutional Determinants of Fire Insurance Regulation.
American Journal of Sociology 107 (1): 101–146.
Schuck, Peter H. 1991. Scheduled Damages and Insurance Contracts for Future
Services: A Comment on Blumstein, Bovbjerg, and Sloan. Yale Journal of Regula-
tion 8 (1): 213–221.
Schwartz, Victor E., Mark A. Behrens, and Leah Lorber. 2000. Tort Reform
Past, Present, and Future: Solving Old Problems and Dealing with “New Style”
Litigation. William Mitchell Law Review 27 (1): 237–269.
Schwartz, W. B., and D. N. Mendelson. 1989a. Physicians Who Have Lost Their
Malpractice Insurance: Their Demographic Characteristics and the Surplus-Lines
Companies That Insure Them. Journal of the American Medical Association 262
(10): 1335–1341.
Schwartz, W. B., and D. N. Mendelson. 1989b. The Role of Physician-Owned
Insurance Companies in the Detection and Deterrence of Negligence. Journal of
the American Medical Association 262 (10): 1342.
Schwartz, W. B., J. P. Newhouse, B. W. Bennett, and A. P. Williams. 1980. The
Changing Geographic Distribution of Board-Certified Physicians. New England
Journal of Medicine 303 (18): 1032–1038.
Shapo, Marshall S. 2003. Tort Law and Culture. Durham, NC: Carolina
Academic Press.
Sharkey, C. M. 2006. Caps and the Construction of Damages in Medical Mal-
practice Cases. In Medical Malpractice and the U.S. Health Care System, edited
by William M. Sage and Rogan Kersh. Cambridge: Cambridge University
Press.
Shavell, S. 1980. Strict Liability vs. Negligence. Journal of Legal Studies 9 (1):
1–25.
———. 1987. Economic Analysis of Accident Law. Cambridge, MA: Harvard
University Press.
Shmanske, Stephen, and Tina Stevens. 1986. The Performance of Medical
Malpractice Review Panels. Journal of Health Politics, Policy, and Law 11 (3):
525–535.
430 References

Showstack, Jonathan, Patricia P. Katz, and Ellen Weber. 1999. Evaluating the
Impact of Hospitalists. Annals of Internal Medicine 130 (4): 376–381.
Sieg, H. 2000. Estimating a Bargaining Model with Asymmetric Information:
Evidence from Medical Malpractice Disputes. Journal of Political Economy 108:
1006–1021.
Silver, Charles. 2002. Does Civil Justice Cost Too Much? Texas Law Review
80: 2073–2113.
Sloan, F. A., S. Berman, S. Rosenbaum, R. A. Chalk, and R. B. Giffin. 2004.
The Fragility of the US Vaccine Supply. New England Journal of Medicine 351
(23): 2443–2447.
Sloan, F. A., and Randall R. Bovbjerg. 1989. Medical Malpractice: Crises,
Response, and Effects. Washington, DC: Health Insurance Association of
America Research Bulletino.
Sloan, F. A., S. S. Entman, B. A. Reilly, C. A. Glass, G. B. Hickson, and H. H.
Zhang. 1997. Tort Liability and Obstetricians’ Care Levels. International Review
of Law and Economics 17 (2): 245–260.
Sloan, F. A., P. B. Githens, E. W. Clayton, G. B. Hickson, D. A. Gentile, and
D. F. Partlett. 1993. Suing for Medical Malpractice. Chicago: University of
Chicago Press.
Sloan, F. A., and M. Hassan. 1990. Equity and Accuracy in Medical Malpractice
Insurance Pricing. Journal of Health Economics 9: 289–319.
Sloan, F. A., and C. R. Hsieh. 1990. Variability in Medical Malpractice Payments.
Law and Society Review 24: 601–650.
———. 1995. Injury, Liability, and the Decision to File a Medical Malpractice
Claim. Law and Society Review 29 (3): 413–435.
Sloan, F. A., P. M. Mergenhagen, and R. R. Bovbjerg. 1989. Effects of Tort
Reforms on the Value of Closed Medical Malpractice Claims. Journal of Health
Politics, Policy, and Law 14: 63–89.
Sloan, F. A., P. M. Mergenhagen, W. B. Burfield, R. R. Bovbjerg, and M. Hassan.
1989. Medical Malpractice Experience of Physicians: Predictable or Haphazard?
Journal of the American Medical Association 262 (23): 3291–3297.
Sloan, F. A., K. Whetten-Goldstein, S. S. Entman, E. D. Kulas, and E. M. Stout.
1997. The Road from Medical Injury to Claims Resolution: How No-Fault and
Tort Differ. Law and Contemporary Problems 60: 35–70.
Sloan, Frank, Jerry Cromwell, and Janet Mitchell. 1978. Private Physicians and
Public Programs. Lexington, MA: Lexington Books.
Sloan, Frank A. 1985. State Responses to the Malpractice Insurance “Crisis” of
the 1970s: An Empirical Assessment. Journal of Health Politics, Policy, and Law
9 (4): 629–646.
———. 1990. Experience Rating: Does it Make Sense for Medical Malpractice
Insurance? American Economic Review 80 (2): 128–133.
 References 431

———. 2004. Public Medical Malpractice Insurance. Funded by The Project

on Medical Liability in Pennsylvania. Pew Charitable Trust. Available at
http://www.pewtrusts.org/pdf/medical_malpractice_sloan_030904.pdf (accessed
June 22, 2007).
———, ed. 1995. Valuing Health Care: Costs, Benefits, and Effectiveness of
Pharmaceuticals and Other Medical Technologies. New York: Cambridge Uni-
versity Press.
Sloan, Frank A., Randall R. Bovbjerg, and Penny B. Githens. 1991. Insuring
Medical Malpractice. New York: Oxford University Press.
Sloan, Frank A., and Charles E. Eesley. 2006. Governments as Insurers in Profes-
sional and Hospital Liability Insurance Markets. In Medical Malpractice and the
U.S. Health Care System—New Century, Different Issues, edited by W. M. Sage
and R. Kersh. Cambridge: Cambridge University Press.
Sloan, Frank A., Charles E. Eesley, and Chris J. Conover. 2006. Public Provision
of Medical Malpractice Insurance: Pennsylvania’s Experience. Working Paper,
Duke University Center for Health Policy.
Sloan, Frank A., Charles E. Eesley, Christopher J. Conover, Carrie A. Mathews,
and William M. Sage. 2005. Public Medical Malpractice Insurance: An Analysis
of State-Oriented Patient Compensation Funds. DePaul Law Review 54 (2):
247–276.
Sloan, Frank A., Stephen S. Entman, Bridget A. Reilly, Cheryl A. Glass, Gerald
B. Hickson, and Harold H. Zhang. 1997. Tort Liability and Obstetricians’ Care
Levels. International Review of Law and Economics 17 (2): 245–260.
Sloan, Frank A., and Penny B. Githens. 1994. Drinking, Driving, and the Price
of Automobile Insurance. Journal of Risk and Insurance 61 (1): 33–58.
Sloan, Frank A., and Thomas J. Hoerger. 1991. Uncertainty, Information and
Resolution of Medical Malpractice Disputes. Journal of Risk and Uncertainty 4:
403–423.
Sloan, Frank A., Bridget A. Reilly, and Christoph Schenzler. 1994. Effects of
Prices, Civil and Criminal Sanctions, and Law Enforcement on Alcohol-Related
Mortality. Journal of Studies on Alcohol 55 (4): 454–466.
———. 1995. Effects of Tort Liability and Insurance on Heavy Drinking and
Drinking and Driving. Journal of Law and Economics 38 (1): 49–77.
Sloan, Frank A., Emily M. Stout, Kathryn Whetten-Goldstein, and Lan Liang.
2000. Drinkers, Drivers, and Bartenders: Balancing Private Choices and Public
Accountability. Chicago: University of Chicago Press.
Sloan, Frank A., W. Kip Viscusi, Harrell W. Chesson, Christopher J. Conover,
and Kathryn Whetten-Goldstein. 1998. Alternative Approaches to Valuing
Intangible Losses: Evidence for Multiple Sclerosis. Journal of Health Economics
17 (4): 475–497.
Sloan, Frank A., Kathryn Whetten-Goldstein, Penny B. Githens, and Stephen S.
Entman. 1995. Effects of the Threat of Medical Malpractice Litigation and Other
Factors on Birth Outcomes. Medical Care 33 (7): 700–714.
432 References

Sloan, Frank A., K. Whetten-Goldstein, E. Kulas, G. Hickson, and S. Entman.

1999. Compensation for Birth Related Injury: No-Fault Compared to Tort
Systems. Archives of Pediatrics and Adolescent Medicine 153: 41–48.
Sloan, Frank A., Kathryn Whetten-Goldstein, Emily Stout, Stephen Entman,
and Gerald Hickson. 1998. No Fault System of Compensation for Obstetric
Injury: Winners and Losers. Obstetrics and Gynecology 91 (March): 437–
443.
Smarr, Lawrence. 2003. Medical Malpractice Insurance Myths: American College
of Surgeons.
Smith, B. L. 1992. Three Attorney Fee-Shifting Rules and Contingency Fees.
Michigan Law Review 90: 2154–2189.
Smith, Robert S. 1990. Mostly on Mondays: Is Worker’s Compensation Cover-
ing Off-the-Job Injuries? In Benefits, Costs, and Cycles in Workers’ Compensa-
tion, edited by P. S. Borba and D. Appel. Boston: Kluwer Academic.
Smith, Steven K., Carol J. DeFrances, Patrick A. Langan, Bureau of Justice Sta-
tistics Statisticians, and John Goerdt. 2006. Tort Cases in Large Counties.
Washington, DC: U.S. Department of Justice. Available at http://www.ojp.gov/
bjs/pub/ascii/tcilc.txt (accessed Sept. 25, 2006).
Snyder, Edward A., and James W. Hughes. 1990. The English Rule for Allocating
Legal Costs: Evidence Confronts Theory. Journal of Law, Economics, and Orga-
nization 6 (2): 345–380.
Songer, Donald R. 1988. The Influence of Empirical Research: Committee vs.
Floor Decision Making. Legislative Studies Quarterly 13 (3): 375–392.
Songer, Donald R., James M. Underwood, Sonja G. Dillon, Patricia E. Jameson,
and Darla W. Kite. 1985. Voting Cues in Two State Legislatures: A Further
Application of the Kingdon Model. Social Science Quarterly 66: 983–990.
South Carolina Legislative Audit Council. 2000. A Review of the Medical
Malpractice Patients’ Compensation Fund. Available at http://www.lac.sc.gov/
NR/rdonlyres/φE4FφE92-6A69-441φ-AE41-29DE784D58DE/φ/pcf.pdf (assessed
June 22, 2007).
Spence, M. 1977. Consumer Misperceptions, Product Failure and Product Liabil-
ity. Review of Economic Studies 64: 561–572.
Spiller, Pablo T., and Richard G. Vanden Bergh. 2003. Toward a Positive Theory
of State Supreme Court Decision Making. Business and Politics 5 (1): 7–43.
Stanley, R. 2003. Buckhannon Board and Care Home, Inc. v. West Virginia
Department of Health and Human Resources: To the Prevailing Party Goes the
Spoils . . . and the Attorneys’ Fees! Akron Law Review 36: 363–409.
Starr, P. 1982. The Social Transformation of American Medicine. New York:
Basic Books.
State of Florida. 2003. Report of the Governor’s Select Task Force on Healthcare
Professional Liability Insurance. Tallahassee: Florida Insurance Council.
 References 433

State of Maryland. 2004. Governor’s Task Force on Medical Malpractice and

Health Care Access, Final Report. Available at http://www.governor.maryland
.gov/pdfs/medmal_112404.pdf (accessed Jan. 4, 2005).
State of Missouri. 2002. Senate Bill no. 1204.
———. 2003. Senate Bill no. 551.
Stedman, H. J., S. Davidson, and C. Brown. 2001. Mental Health Courts:
Their Promise and Unanswered Questions. Law and Psychiatry 52 (4):
457–458.
Stedman, Richard R. II. 2003. Of Hurricanes and Airplanes: The Congressional
Knee-Jerk Reaction to September 11. Loyola Law Review 49 (4): 991–1023.
Stern, J., C. Rehmus, J. Lowenberg, H. Kasper, and B. Dennis. 1975. Final-Offer
Arbitration. Lexington, MA: Lexington Books.
Steves, Myron F., Jr. 1976. A Proposal to Improve the Cost to Benefit Relation-
ships in the Medical Professional Liability Insurance System. Duke Law Journal
1975 (6): 1305–1333.
Stigler, G. J. 1971. Theory of Economic Regulation. Bell Journal of Economics
and Management Science 2: 3–21.
Studdert, D. M., and T. A. Brennan. 2000. The Problems with Punitive Damages
in Lawsuits Against Managed-Care Organizations. New England Journal of
Medicine 342 (4): 280–284.
Studdert, D. M., M. M. Mello, and T. A. Brennan. 2004. Medical Malpractice.
New England Journal of Medicine 350: 283–292.
Studdert, D. M., M. M. Mello, A. A. Gawande, T. K. Gandhi, A. Kachalia,
C. Yoon, A. L. Puopolo, and T. A. Brennan. 2006. Claims Errors and Compen-
sation Payments in Medical Malpractice Litigation. New England Journal of
Medicine 354 (19): 2024–2033.
Studdert, D. M., M. M. Mello, W. M. Sage, C. M. DesRoches, J. Peugh,
K. Zapert, and T. A. Brennan. 2005. Defensive Medicine Among High-Risk
Specialist Physicians in a Volatile Malpractice Environment. Journal of the
American Medical Association 293 (21): 2609–2617.
Studdert, David M., and Troyen A. Brennan. 2001. No-Fault Compensation for
Medical Injuries: The Prospect for Error Prevention. Journal of the American
Medical Association 286 (2): 217–223.
Studdert, David M., Lori A. Fritz, and Troyen A. Brennan. 2000. The Jury Is
Still In: Florida’s Birth-Related Neurological Injury Compensation Plan After a
Decade. Journal of Health Politics, Policy and Law 25 (3): 499–526.
Studdert, David M., Eric J. Thomas, Brett I. W. Zbar, Joseph P. Newhouse,
et al. 1997. Can the United States Afford a “No-Fault” System of Compensation
for Medical Injury? Law and Contemporary Problems 60 (1–2): 1.
Studdert, David M., Y. T. Yang, and Michelle M. Mello. 2004. Are Damages
Caps Regressive? A Study of Malpractice Jury Verdicts in California. Health
Attairs 23 (4): 54–67.
434 References

Sturgis, Robert W. 1995. Tort Cost Trends: An International Perspective. St.

Louis, MO: TillinghastTowers Perrin.
Sugarman, Stephen D. 1990. The Need to Reform Personal Injury Law Leaving
Scientific Disputes to Scientists. Science 248 (4957): 823–828.
Sunstein, Cass R. 2002a. To Punish or Not? In Punitive Damages: How Juries
Decide, by C. R. Sunstein, R. Hastie, J. W. Payne, D. A. Schkade, and W. K.
Viscusi. Chicago: University of Chicago Press.
———. 2002b. What Should Be Done? In Punitive Damages: How Juries Decide,
by C. R. Sunstein, R. Hastie, J. W. Payne, D. A. Schkade, and W. K. Viscusi.
Chicago: University of Chicago Press.
———. 2003. Terrorism and Probability Neglect. Journal of Risk and Uncer-
tainty 26 (2/3): 121–136.
Sunstein, Cass R., David Schkade, and Daniel Kahneman. 2000. Do People Want
Optimal Deterrence? Journal of Legal Studies 29: 237–253.
———. 2002. Do People Want Optimal Deterrence? In Punitive Damages: How
Juries Decide, by C. R. Sunstein, R. Hastie, J. W. Payne, D. A. Schkade, and
W. K. Viscusi. Chicago: University of Chicago Press.
Surowiecki, J. 2004. The Wisdom of Crowds. New York: Doubleday.
Sutter, R. L. 2002. Kansas Health Care Stabilization Fund, Indicated Liabilities
at June 30, 2002: Surcharge Determination for FY2003. St. Louis, MO:
Tillinghast Towers Perrin.
Swartz, Katherine. 2006. Reinsuring Health: Why More Middle-Class People
Are Uninsured and What Government Can Do. New York: Russell Sage Founda-
tion Press.
Swedish Medical Association, National Board of Health and Welfare. 2003.
Working in Sweden: Information for Doctors from EU/EEA Countries. Available
at http://www.lg.se/upload/epi/lg/pdf/2005/working_in_sweden.pdf (accessed
June 22, 2007).
Symposium on Medical Malpractice: Preface. 1975. Duke Law Journal 1975 (6):
1177.
Tabarrok, Alexander, and Eric Helland. 1999. Court Politics: The Political
Economy of Tort Awards. Journal of Law and Economics 42: 157–188.
———. 2005. Two Cheers for Contingent Fees. Washington DC: American
Enterprise Institute Press.
Tappan, Kristie. 2005. Medical-Malpractice Reform: Is Enterprise Liability
or No-Fault a Better Reform? Boston College Law Review 46: 1095–
1130.
Taragin, Mark I., Laura R. Willett, Adam P. Wilczek, Richard Trout, and Jeffrey
L. Carson. 1992. The Influence of Standard of Care and Severity of Injury on
the Resolution of Medical Malpractice Claims. Annals of Internal Medicine 117
(9): 780–784.
 References 435

Tarr, G. Alan. 1998. Interest Groups and Judicial Federalism: Organizational

Litigation in State Judiciaries. Publius 28 (4): 167–168.
———. 2005. Judicial Process and Judicial Policymaking. 4th ed. Belmont, CA:
Wadsworth Publishing.
Thomas, Tim A. 2006. Common Law, 2d, vol. 15A of American Jurisprudence:
Thomas West Publishing.
Thorpe, Kenneth E. 2004. The Medical Malpractice “Crisis”: Recent Trends and
the Impact of State Tort Reforms. Health Affairs. Available at http://content.
healthaffairs.org/cgi/reprint/hlehaff.w4.20vl?maxtoshow=&HITS=10&hits=10
&RESULTFORMAT=&authorl=thorpe&andorexactfulltext=and&searchid=1
&FIRSTINDEX=0&resourcetype=HWCIT (accessed June 22, 2007).
Tinetti, Mary E., Sidney T. Bogardus, Jr., and Joseph V. Agostini. 2004. Potential
Pitfalls of Disease-Specific Guidelines for Patients with Multiple Conditions. New
England Journal of Medicine 351 (27): 2870–2874.
Todd, S. 2000. International Torts: A Comparative Study. Privitization of Acci-
dent Compensation: Policy and Politics in New Zealand. Washburn Law Journal
39: 404.
Topel, Robert H. 1984. Experience Rating of Unemployment Insurance and
the Incidence of Unemployment. Journal of Law and Economics 27 (1):
61–90.
Treaster, Joseph B. 2003. Malpractice Insurance: No Clear or Easy Answers.
The New York Times. New York, N.Y., March 5, C1.
Tversky, A., and D. Kahneman. 1973. Availability: A Heuristic for Judging
Frequency and Probability. Cognitive Psychology 5: 207–232.
U.S. Congress, Office of Technology Assessment. 1994. Defensive Medicine and
Medical Malpractice. OTA-H-602. Washington, DC: U.S. Congress.
U.S. Congressional Budget Office. 2002. Federal Reinsurance for Disasters.
Washington, DC: U.S. Congress.
———. 2003. The Economics of U.S. Tort Liability: A Primer. Washington, DC:
U.S. Congress.
———. 2004. The Effects of Tort Reform: Evidence from the States. Washing-
ton, DC: U.S. Congress.
Udell, Nancy, and David B. Kendall. 2005. Health Courts: Fair and Reli-
able Justice for Injured Patients. Progressive Policy Institute Policy Report
(Feb.). Available at http://www.ppionline.org/documents/healthcourts_0217.pdf
(accessed June 22, 2007).
United States Census Bureau. 2000. State and County QuickFacts. Available at
http://quickfacts.census.gov/qfd (accessed Sept. 22, 2006).
United States Department of Commerce. 2006. Statistical Abstract of the United
States: The National Data Book, 125th ed. Washinton, DC: United States Census
Bureau.
436 References

United States Government Accountability Office. 2005. Medicare Physician Fees:

Geographic Adjustment Indices Are Valid in Design, but Data and Methods
Need Refinement. Washington, DC: United States General Accountability
Office.
United States General Accounting Office. 1987. VA Health Care, VA’s Patient
Injury Control Program Not Effective. Washington, DC: U.S. Government Print-
ing Office.
———. 1993. Medical Malpractice: Medicare/Medicaid Beneficiaries Account
for a Relatively Small Percentage of Malpractice Losses. Washington, DC: U.S.
Government Printing Office.
———. 1995. Medical Liability: Impact on Hospital and Physician Costs
Extends Beyond Insurance. Washington, DC: U.S. Government Printing
Office.
———. May 1993a. Medical Malpractice: Experience with Efforts to Address
Problems. GAO/T-HRD-93–24 (May).
———. 2003b. Medical Malpractice Insurance: Multiple Factors Have Contrib-
uted to Increased Premium Rates. Washington, DC: United States General
Accounting Office.
———. 2003c. Medical Malpractice: Implications of Rising Premiums on Access
to Health Care. Washington, DC: United States General Accounting Office
(Aug.).
Vargo, John F. 1993. The American Rule on Attorney Fee Allocation: The
Injured Person’s Access to Justice. American University Law Review
42:1567–1636.
Varian, Hal R. 2001. Catastrophe Bonds Could Fill the Gaps in Reinsurance.
New York Times. New York, N.Y., October 25, C2.
Venezian, Emilio. 1985. Ratemaking Methods and Profit Cycles in Property and
Liability Insurance. Journal of Risk and Insurance 52: 477–500.
Vidmar, Neil. 1993. Empirical Evidence on the Deep Pockets Hypothesis: Jury
Awards and Suffering in Medical Malpractice Cases. Duke Law Journal 43 (2):
217–266.
———. 1994a. Are Juries Competent to Decide Liability in Tort Cases Involving
Scientific/Medical Issues? Some Data from Medical Malpractice. Emory Law
Journal 43 (3): 885–911.
———. 1994b. Pap and Circumstance: What Jury Verdict Statistics Can Tell Us
About Jury Behavior and the Tort System. Suffolk University Law Review 28
(4): 1205–1234.
———. 1995. Medical Malpractice and the American Jury. Ann Arbor: Univer-
sity of Michigan Press.
———. 1998. The Performance of the American Civil Jury: An Empirical Per-
spective. Arizona Law Review 40 (3): 849–899.
 References 437

———. 1999. Juries Don’t Make Legal Decisions! And Other Problems: A
Critique of Hastie et al. On Punitive Damages. Law and Human Behavior 23
(6): 705–714.
———. 2003. Juror Discussions During Civil Trials. Arizona Law Review 45:
1–82.
Vidmar, Neil, Felicia Gross, and Mary Rose. 1998. Jury Awards for Medical
Malpractice and Post-Verdict Adjustments of Those Awards. DePaul Law
Review 48: 265–300.
Vidmar, Neil, and Jeffrey J. Rice. 1993. Assessments of Noneconomic Damage
Awards in Medical Negligence: A Comparison of Jurors with Legal Profes-
sionals. Iowa Law Review 78: 883–912.
Vidmar, Neil, Russell M. Robinson III, and Kara MacKillop. 2006. “Judicial
Hellholes”: Medical Malpractice Claims, Verdicts and the “Doctor Exodus” in
Illinois. Vanderbilt Law Review 59 (4): 1309–1342.
Vidmar, Neil J. 2004. Experimental Simulations and Tort Reform: Avoidance,
Error and Overreaching in Sunstein et al.’s Punitive Damages. Emory Law
Journal 53 (1): 1359–1403.
Vidmar, Neil J., and M. R. Rose. 2001. Punitive Damages by Juries in Florida:
In Terrorem and in Reality. Harvard Journal on Legislation 38:487–513.
Viscusi, W. K., and W. N. Evans. 1990. Utility Functions That Depend on
Health-Status Estimates and Economic Implications. American Economic Review
80 (3): 353–374.
Viscusi, W. Kip. 1979. Employment Hazards: An Investigation of Market Per-
formance. Cambridge, MA: Harvard University Press.
———. 1992. Fatal Tradeoffs: Public and Private Responsibilities for Risk. New
York: Oxford University Press.
———. 1993. The Value of Risks to Life and Health. Journal of Economic
Literature. 31 (4): 1912–1946.
———. 2001. Jurors, Judges, and the Mistreatment of Risk by the Courts.
Journal of Legal Studies 107: 107–142.
———. 2002a. Corporate Risk Analysis: A Reckless Act? In Punitive Damages:
How Juries Decide, by C. R. Sunstein, R. Hastie, J. W. Payne, D. A. Schkade
and W. K. Viscusi. Chicago: University of Chicago Press.
———. 2002b. Judging Risk and Recklessness. In Punitive Damages: How Juries
Decide, by C. R. Sunstein, R. Hastie, J. W. Payne, D. A. Schkade, and W. K.
Viscusi. Chicago: University of Chicago Press.
Viscusi, W. Kip, Richard J. Zeckhauser, Patricia Born, and Glenn Blackmon.
1993. The Effect of 1980s Tort Reform Legislation on General Liability and
Medical Malpractice Insurance. Journal of Risk and Uncertainty 6: 165–186.
Viscusi, W. Kip, and P. H. Born. 1995. Medical Malpractice Insurance in the
Wake of Liability Reform. Journal of Legal Studies 25: 463–490.
438 References

———. 2005. Damage Caps, Insurability, and the Performance of Insurance.

Journal of Risk and Insurance 72 (1): 23–42.
Viswanathan, Hema N., and J. Warren Salmon. 2000. Accrediting Organizations
and Quality Improvement. American Journal of Managed Care 6 (10):
1117–1130.
Wachter, R. M. 2004. The End of the Beginning: Patient Safety Five Years After
“To Err is Human.” Health Affairs W4-534–W4-545.
———. 2006. Expected and Unanticipated Consequences of the Quality and
Information Technology Revolutions. Journal of the American Medical Associa-
tion 295 (23): 2780–2783.
Wachter, Robert M., and Lee Goldman. 1996. The Emerging Role of “Hospital-
ists” in the American Health Care System. New England Journal of Medicine
335: 514–517.
———. 2002. The Hospitalist Movement 5 Years Later. Journal of the American
Medical Association 287 (4): 487–494.
Warburton, Rebecca Nunn. 2005. Patient Safety—How Much Is Enough? Health
Policy 71: 223–232.
Weiler, P. C. 1991. Medical Malpractice on Trial. Cambridge, MA: Harvard
University Press.
———. 1993. The Case for No-Fault Medical Liability. Maryland Law Review
52: 908–950.
Weiler, Paul C., Howard H. Hiatt, Joseph P. Newhouse, William G. Johnson,
Troyen Brennan, and Lucian L. Leape. 1993. A Measure of Malpractice: Medical
Injury, Malpractice Litigation, and Patient Compensation. Cambridge, MA:
Harvard University Press.
Weinberg, J., L. H. Hilborne, and Q. T. Ngyuen. 2005. Regulation of
Health Policy: Patient Safety and the States. Advances in Patient Safety 1:
405–422.
Weinstein, Debra F. 2002. Duty Hours for Resident Physicians: Tough Choices
for Teaching Hospitals. New England Journal of Medicine 347: 1275–1278.
Weinstein, R. A., J. D. Siegel, and T. A. Brennan. 2005. Infection-Control Report
Cards: Securing Patient Safety. New England Journal of Medicine 353 (3):
225–227.
Weiss, Gail Garfinkel. 2004. Malpractice: Can No-Fault Work? Medical Eco-
nomics 81 (11): 66–71.
Weiss, M. A., and J. H. Chung. 2004. U.S. Reinsurance Prices, Financial
Quality, and Global Capacity. Journal of Risk and Insurance 71 (3): 437–
467.
Wheelock, David C., and Paul W. Wilson. 1995. Explaining Bank Failures:
Deposit Insurance, Regulation, and Efficiency. Review of Economics and Statis-
tics 77 (4): 689.
 References 439

White, Michelle J. 1994. The Value of Liability in Medical Malpractice. Health

Affairs 13 (4): 75–87.
———. 2004. The “Arms Race” on American Roads: The Effect of Sport Utility
Vehicles and Pickup Trucks on Traffic Safety. Journal of Law and Economics
47 (2): 333–356.
Whiteman, David. 1985. The Fate of Policy Analysis in Congressional Decision
Making: Three Types of Use in Committees. Western Political Quarterly 38 (2):
294–311.
Williams, P. A. 1995. Children and the Law: A Unified Family Court for
Missouri. University of Missouri Kansas City Law Review 63 (3): 383–428.
Winter, R. A. 1988. Liability Crisis and the Dynamics of Competitive Insurance
Markets. Yale Journal of Regulation 5: 455–500.
———. 1991. The Liability Insurance Market. Journal of Economic Perspectives
5 (3): 115–136.
———. 1994. Dynamics of Competitive Insurance Markets. Journal of Financial
Intermediation 3: 379–415.
Winter, Ralph. 1991. Solvency Regulation and the Insurance Cycle. Economic
Inquiry XXIX (3): 458–472.
Wisconsin Hospital Association. March Inc. 2003. 2003–2005 State
Budget: Impacts on Wisconsin Hospitals. Available at http://www.wha.org/
pubArchive/position_Statements/pp2003issuesummary.pdf (accessed June 22,
2007).
Wisconsin Insurance Report. 2001. Madison: State of Wisconsin, Office of the
Commissioner of Insurance.
Witt, John Fabian. 2005. The Long History of State Constitutions and American
Tort Law. Rutgers Law Review 36 (4): 1159–1199.
Wood, Kathryn E., and David B. Nash. 2005. Mandatory State-Based Error-
Reporting Systems: Current and Future Prospects. American Journal of Medical
Quality 20 (6): 297–303.
Wren, Daniel A., and Arthur G. Bedeian. 2004. The Taylorization of Lenin:
Rhetoric or Reality? The International Journal of Social Economics 31 (3):
87–299.
Wright, A. A., and I. T. Katz. 2005. Bar Coding for Patient Safety. New England
Journal of Medicine 353 (4): 329–331.
Wu, A. W., S. Folkman, S. J. McPhee, and B. Lo. 2003. Do House Officers Learn
from Their Mistakes? Journal of the American Medical Association 265:
2089–2094.
Würdemann Vanderbilt, Henry. 1912. Injuries of the Eye. Chicago: Cleveland
Press.
Yelen, S. 1993. Withdrawal Restrictions in the Automobile Insurance Market.
Yale Law Journal 102 (6): 1431–1455.
440 References

Young, T. 2005. Presidential Address. Human Error, Patient Safety and the Tort
Liability Crisis: The Perfect Storm. American Journal of Obstetrics and Gynecol-
ogy 193 (2): 506–511.
Zimmerman, Rachel, and Christopher Oster. 2002. Assigning Liability: Insurers’
Missteps Helped Provoke Malpractice “Crisis.” The Wall Street Journal, June
24, A1, 8.
Zuckerman, Stephen, Randall R. Bovbjerg, and Frank Sloan. 1990. Effects of
Tort Reforms and Other Factors on Medical Malpractice Insurance Premiums.
Inquiry 27: 167–182.
Zwanziger, Jack, Glenn A. Melnick, and Anil Bamezai. 2000. Can Cost Shifting
Continue in a Price Competitive Environment? Health Economics 9 (3):
211–226.
Zweigert, Konrad, and Heinz Kötz. 1987. An Introduction to Comparative Law,
2nd rev. ed. Oxford: Clarendon Press.
Index

Abraham, Kenneth S., 396n18 on physician supply, 69

Academic health centers (AHCs), 334 premium trends, 58
Access to care on runaway juries, 164
barriers, 53, 55 American Society of
patient travel time and, 70 Anesthesiologists, 208
physician supply and, 69–71 Ameringer, Carl F., 207
Accident Compensation Corporation Anesthesiologists, 208–9, 224
(ACC), 300–303, 394–95n123 Anticipated losses, 37, 387n41
Adams, Mark L., 13 Aranson, Allison F., 144
Adhesion contracts, 398–99n62 Arbitration, 128–29, 298–99,
Adjusted loss expense, 32, 44 347n13–14
ADR. See Alternative dispute Arizona, 63, 65
resolution (ADR) Association of British Insurers, 156
Adverse outcomes. See Errors, Attachment points, reinsurance,
medical; Injuries, medical 249–50
Alaska, 150–51, 152, 383n73 Attorneys. See Lawyers
ALE. See Adjusted loss expense Australia, 295, 394n118
Alternative dispute resolution (ADR), Autism, 290, 305
92, 310 Avraham, R., 124
A.M. Best, 47, 242 Awards. See Compensation, patient
American Association for
Accreditation of Ambulatory Baicker, Katherine, 68
Facilities, 210 Baker, Tom, 36, 86
American Bar Association, 138 on reforms, 316, 318, 319, 335
on prepaid legal services, 157, 158 on regulation, 239
American Hospital Association, 229 Balance sheets, insurer’s, 33–34
American Medical Association Banking industry, 243–44
on damage caps, 109 Bankruptcy
on effects of tort reforms, 76 courts, 179–80
on health courts, 177, hospital, 230
364–65n53–54 insurer, 129, 218, 220–21, 239–42,
on physicians’ incomes, 62 384n11
442 Index

Bankruptcy (cont.) legal challenges to, 113–14

reinsurance company, 248 no-fault systems and, 332–34
savings and loans, 243–44 Captive insurers, 320–21
Bar coding, 202 Cardiac care, 74–76, 190–91, 208,
Barker, Drucilla K., 102 209–10
Becker, D. J., 103 Care, medical
Bedside manner, 77–79 cardiac, 74–76, 190–91, 208,
Better Business Bureau, 129 209–10
Biddle, Jeff, 293 compensation in form of provision
Birth injuries, 280–87, 305 of, 125–31
Blumstein, J. F., 119, 130 costs and malpractice insurance,
Born, Patricia, 102, 239 51–52, 72–76, 125–31
Bovbjerg, Randall R., 99, 119, 130 costs of diagnostic, 53–54
on premiums, 223 defensive, 14–15, 17, 54–55
on regulation, 239 diagnostic, 53–54, 396n24
Bradley, E. W., 208 market failure and government
Brennan, Troyen A., 80, 109, intervention, 192–95
299–300 medical malpractice reducing patient
Brickman, Lester, 143 access to, 62
Brotherhood of Railroad Trainmen, for neurologically impaired infants,
157 280–87
Brotherhood of Railroad Trainmen v. nontangible benefits of, 54
Virginia ex rel. Virginia State Bar, patient access to, 53, 55, 69–71
157 quality, 194, 197–98
Bryan, Patrick, 153 socially optimum levels of, 15–16
Burfield, W. B., 207 standards of, 90, 91, 198–99
Bush, George W., 51 threat of tort liability effect on,
Bush, Jeb, 51 14
waste and, 54
California, 18–19 Carleton, D. W., 102
award limits in, 102 Carns, T. W., 152
damage caps, 117 Centers for Medicare, U. S., 208
error reporting in, 202 Certification, physician, 207, 371n24,
hygiene quality grade cards in, 210 375–76n59
medical error studies, 195 Chandra, Amitabh, 68
workers’ compensation, 294–95, Choi, Seungmook, 47
295 Civil rights suits, 157–58
Canada, 295, 296–97 Claims, insurance
Capital asset pricing model (CAPM), effects of first-generation reforms
223 on, 100–102
Caps, damage, 92, 99, 132–33, frequency and severity of, 31,
344n36 52–53, 59, 100–102, 345n45–46
benefits to nonclaimants, 118 length of time to resolve, 32
flexible ranges in lieu of, 122 monetary costs of, 173–74
indexed to rate of inflation, 117 patient compensation funds effects
laws limiting, 102–4, 105, 354n47 on, 264–65
 Index 443

payments as transfer payments, 354–55n53, 361n5, 363–64n50,

73 363n46
payouts, 61, 61 schedules, 92, 119–25, 132–34
physician assessments of, 168–70 traditional rules regarding, 87, 89
premiums paid before, 31–32 for vaccine injuries, 287–91
reinsurance and, 48, 248–49 Competition among insurers, 39–44
settled without litigation, 114 Computerized physician order-entry
volume of medical malpractice, systems, 202–3, 372–73n40–41
18–20 Consent forms, 89–90
Claims-made policies, 32 Consumer Price Index, 29, 276
Class action suits, 140 damage caps and, 117
Clayton, E. W., 138, 167, 170 Consumers Union, 111
Clinton, Bill, 326 Contingent fee system
Colorado, 195, 299–300 advantages of, 141–42, 357n16
Common Good, 177, 178 criticism of, 135–36
Common law. See English common disadvantages of, 142–43
law early offers and, 153–55
Compensation. See also Fees, lawyer; fee shifting and, 149–52
Patient compensation funds historical context, 141
(PCFs) lawyer earnings, 143–45
for birth injuries, 280–87 legal expense insurance and, 155–60
caps, 92, 99, 102–4, 105, 107–9, legislation, 146–48
113–14, 116–17, 132–33, 332– regulation, 148–49
34, 344n36, 354n47 specialization by attorneys and, 138
computing future loss for, 109 state policy and, 146–49
default judgment, 347n5 support for, 136–40
effect on physician supply, 102–4 weaknesses of, 139–40
effects of first-generation reforms Contracts, adhesion, 398–99n62
on, 100–102 Controlled Risk Insurance Company,
in form of provision of service Ltd., 396n18
benefits, 125–31 Corcoran, James P., 380–81n23
future losses, 129–30 Corporate liability, 322–23
imperfections in current system of Costs. See also Premiums
ascertaining payment for, 114–16 in clinical scenarios and, 73–74
juror errors and, 172–73 diagnostic care, 53–54, 396n24
laws, 107–9 of injuries, 173–74
legal challenges to limits on, medical malpractice, 51–52, 72–73
113–14 no-fault program, 277–78, 299–300,
myths about, 17–18 301
for nonmonetary loss, 109–13, physicians fees, 55
116–18, 119–31, 355n63, workers’ compensation, 294
356n64 Court. See also Judges; Juries
in periodic payments, 102, 126–27 bankruptcy, 179–80
prejudgment interest, 344n38 drug, 180
public policy responses to, 116–18 family, 180–83, 366–67n68,
punitive damages, 174–76, 344n37, 367n70–73
444 Index

Court. (cont.) Deterrence of medical injuries, 79–81,

health, 176–79, 183–84, 183–86, 314–15
183–87, 331–32, 364–65n53–54, Developed losses, 37–38
364n52 Devlin, R. A., 296
juvenile, 180–81 Diagnostic care costs, 53–54, 396n24
and legal service plans, 157–58 Di Pietro, S., 152
mental health, 180, 181 Disclosure statutes, 207–10
specialty, 179–84, 368n85 Discovery rules, 89, 203–4, 390n52
tax, 179–80, 368n84 District of Columbia Bar Association,
CPI. See Consumer Price Index 186
CPOE. See Computerized physician Diversification of risk, 35–36
order-entry systems Doherty, N. A., 47
Credentialing, hospital, 208 Domitrovich, S., 182
Croley, Steven P., 111 Doyle, Jim, 13
Cycles, insurance Dram shop statues, 338n5
dynamics, 36–38 Dranove, David, 70, 209
external factors and, 44–49 Drug courts, 180
and frequency and severity of Dubay, Lisa, 76
claims, 31 Dubner, Stephen, 215
internal factors and, 39–44 Duke University Health System, 316,
medical malpractice and, 29–31 317, 384n10
nature of, 28–29 Durham Casualty Company, 320
noncauses of, 49–50
reinsurance and, 255–56 Early offers, 153–55, 162
Earnings volatility, 43
Damages. See Compensation, Economics
patient assumptions, 343n8
Daniels, Stephen, 144 Keynesian, 340n6
Danzon, Patricia Munch Eisenberg, Theodore, 175
on contingent fees, 148 Employers, experience rating of,
on enterprise liability, 328 293–94
on insurance cycles, 51 Encinosa, William E., 68, 103
on nonmonetary loss, 110 English common law, 96, 141,
on physician quality, 80, 345n51 149–52, 161, 344n39,
on premium changes, 60, 61, 348–49n31–32
341n17 Enterprise insurance
on private contracting, 132 definition and rationale, 315–18
D’Arcy, Stephen P., 94 historical context, 315
Default judgments, 347n5 incentives to encourage adoption of,
Defensive medicine, 14–15, 17, 328–29
313–14, 345n51 objections to, 318–24
defining, 71–72 reforms linked to provision of,
negative, 54–55, 70–71 329–31
outcomes, 75–76 stepping stones to, 321–24
positive, 54–55, 58, 72–73, 74–76 Enterprise liability, 92–93, 211–12,
premium changes and, 57–58 378n86
 Index 445

advantages and disadvantages of, Federal Emergency Management

328 Agency (FEMA), 273, 388n54
comparison with enterprise Federal government as insurer,
insurance, 325–26 219–20, 366n57
definition and rationale, 324–25 Federal Tort Claims Act, 192
incentives to encourage adoption of, Fees
328–29 lawyer, 112, 131 (see also Lawyers)
legislation, 326 caps, 344n40
objections to, 326–28 commission-based, 140
reforms linked to provision of, contingent fee system, 135–40, 141–
329–31 49, 357n16
Entman, Stephen S., 76, 81, 283 early offers and, 153–55, 162
Errors, medical. See also Safety, fixed, 140
patient legal expense insurance and, 155–60
barriers to patient access to shifting, 149–52
information about, 209 statutory, 140
common types of, 197, 201 task-based, 140
mandatory reporting, 209 time-based, 140, 142
no-fault systems and, 303, 306 types of arrangements for, 140
physician/hospital communication vaccine injury lawsuits and, 289
problems and, 373n44–45 value-based, 140
quality assurance mechanisms for physician, 55, 333
avoiding, 206–10 Fielding, Stephen L., 85–86
reporting, 198, 200–201, 303, 306, First-generation reforms, 86–90,
370n20 98–100
state legislation regarding, 201–2 Fitzpatrick, Sean M., 40
underreporting of, 203–6 Fixed lawyer fees, 140
Errors made by juries, 170–73, 310, Flood insurance, 273
362n28 Florida
Europe, legal expense insurance in, costs of injuries in, 173–74
155–57, 159 drug courts, 180
Excess insurers. See Reinsurance experience rating in, 229
Experience rating, 212–14, 228–30, fee shifting in, 152
378–79n89 laws on contingent fees, 146–47,
workers’ compensation programs 161
and, 293–94 medical malpractice in, 51, 62, 63,
118, 167
Fager, Donald J., 42 mental health courts, 180
Failures no-fault system, 92, 279, 281–87,
insurer, 239, 382n53 330, 334
market, 192–95, 256–57, 274 obstetrics in, 76, 77–78, 81–82
Family courts, 180–83, 366–67n68, patient compensation fund, 258
367n70–73 patient travel time in, 70
Farber, Henry S., 167 physicians directly employed by
Federal Deposit Insurance hospitals in, 215
Corporation, 243–44 physician supply, 65, 67, 69, 191
446 Index

Florida (cont.) oversight of patient safety, 206–10

second-generation tort reform, 104 reinsurance and, 255–56, 274, 329,
Windstorm Underwriting 386n30
Association, 273–74 -sponsored risk pooling associations,
Food and Drug Administration, U. S., 230–36
202, 370n21 state guaranty funds, 230–31,
Freakonomics, 214, 215 233–35
Frequency and severity of claims, 31, vaccine administration guidelines,
52–53, 59, 100–102, 345n45–46 288
Froot, Kenneth A., 252–53 Grisham, John, 165
Functions and goals of the tort Gron, Anne, 70, 342n41
system, 337–38n4 Grossman, Richard, 243
Fung, Hung-Gay, 342n41 Guaranty funds, 230–31, 233–35,
236
GAO. See General Accounting Office,
U. S. (GAO) Haltom, William, 160
Gaudry, M., 296 Hand washing, 214, 215
Gawande, A. A., 168 Hanson, Jon D., 111
Geistfeld, M., 123 Hardigree, Don, 47
General Accounting Office, U. S. Harrington, Scott E., 44
(GAO), 47 Harvard Medical Practice Study, 79,
on insurer’s balance sheets, 33 170
on length of time to resolve claims, Hassan, M., 213, 229, 250, 318
32 Health Care Quality Improvement
on paid losses, 341n19 Act of 1986 (HCQIA), 203
on patient access to medical care, Health courts, 176–79, 183–87,
69–70 331–32, 364–65n53–54, 364n52
on tort reforms, 91–92 Health insurance, 340n9
Georgia, 63 Health maintenance organizations,
Germany, 156–57 322–23
Gifford, Donald G., 41 Heart disease. See Cardiac care
Githens, Penny B. Helland, Eric, 148, 175
on the contingent fee system, 138 Hellinger, Fred J., 68, 103
on jury evaluations of liability, 167, Hersch, Joni, 175
170 Hiatt, Howard H., 279
on premiums, 223 Hickson, G. B., 77, 78
on quality of physicians, 82 Hindsight bias, 362n24
on regulation, 239 HMOs, 322–23
Goerdt, John, 175 Hoerger, T., 250, 252
Golden, David, 164 Hofflander, Alfred E., 39, 269
Government Hospitals
activity market, 11–12 bankruptcy, 230
alternatives to insurer regulation by, computerized physician order-entry
224–25 systems, 202–3
interventions, 192–95, 255–57, 272, corporate liability of, 322–23
366n57, 386n30 credentialing, 208
 Index 447

enterprise insurance and, 315–24, basing compensation for

328–31 nonmonetary loss on
enterprise liability and, 92–93, compensation for, 124–25
211–12, 324–31 compensation in form of provision
error reporting by, 201–3 of service benefits for, 125–31
experience rating of, 213–14, deterrence of, 79–81, 314–15
229–30 eligibility criteria under no-fault
hand washing in, 214, 215 systems, 281–84
health maintenance organizations in hospitals, 195–98
and, 322–23 medical errors and, 195–98
jury evaluations of liability of, 167 National Vaccine Injury
liability insurance, 10 Compensation Program and, 287–
mandatory error reporting by, 209 91, 305, 391–92n59–60, 391n56
medical errors in, 195–98 neurological, 280–81, 305
no-fault systems and, 306 no-fault programs for, 279
physician communication problems permanent, 108–9, 120
with, 373n44–4 quality of physicians and, 79–82
self-insurance by, 218, 227, 248, scenarios and nonmonetary loss,
386n28 122–23
supply, 62, 65 severity scale, 120
underreporting of medical errors by, temporary, 108
203–6 Insolvency, insurer, 129, 218, 220–21,
Hughes, James W., 151 239–42, 382n53, 384n11
Institute of Medicine, 104, 189, 195,
IBNR. See Incurred but not reported 196
loss recommendations for improving
Idaho, 146 patient safety, 198–203
Illinois, 94–95 on vaccine injuries, 291
Implicit taxation and PCFs, 271 Insurance
Incomes health, 340n9
lawyer, 143–45 legal expense, 155–60
physician, 11, 52, 55, 333 medical malpractice (see Premiums;
Income statements, insurer’s, 32 Reinsurance)
Incurred but not reported loss, 33 compared to other forms of
Indiana, 258 property-casualty insurance,
Inflation 225–30
damage caps indexed to rates of, cycles, 27–31
117 external factors affecting, 44–49
interest rates and, 45 federal and state governments as
losses and, 43 providers of, 219–20
shocks, 45 hospital liability, 10
Information provision and regulation, important features of, 218–20
223 internal factors affecting, 39–44
Initial incurred losses, 37 market, 9–11, 46, 130, 252–53
Injuries, medical. See also Safety, patient compensation funds and, 13,
patient 87, 89
448 Index

Insurance (cont.) operating profits, 36–37

policies, 31–32, 223 physician-sponsored, 9, 226–27
portion of total health care price wars among, 41, 221–22
expenses, 51–52 reciprocal, 227
premiums, 5–6, 15, 29, 223 returns on interest-bearing securities,
price war of 1979–1981, 41 44–45
reforms, 90 self-insured organizations and, 218,
regulation, 43, 48, 49, 217–18, 227
236–43, 239, 382n51, 397n35 state guaranty funds, 230–31,
risk classification, 9–10 233–35, 236
property-casualty, 13, 47, 219, stock, 222, 224
342n25 surplus lines, 10, 14, 47, 227
compared to medical malpractice types of, 219, 226–28
insurance, 225–30 underwriting standards, 340n1
compared to PCFs, 272–74 Interest rates and premiums, 44–45,
motor vehicle, 225–26, 295–97 342n41
regulation, 238 IOM. See Institute of Medicine
Insurers, medical malpractice. See Israel, 295
also Losses, insurer
alternatives to government Jefferson, Thomas, 96
regulation of, 224–25 Jin, Ginger Zhe, 210
balance sheets, 33–34 Johnson & Johnson, 166
bankruptcy, 129, 218, 220–21, Joint Commission on the
239–42, 384n11 Accreditation of Healthcare
blamed for crises, 30 Organizations (JCAHO), 208,
capacity constraints, 45–47 324, 371n26
captive, 320–21 Joint underwriting associations, 90,
ceasing to sell, 42, 43 227–28, 230–33, 235, 265
coverage of nonmonetary loss, JUAs. See Joint underwriting
111–12 associations
decision to reinsure, 250–51 Judges
developed losses, 37–38 campaign contributions from
enterprise liability and, 92–93 lawyers, 175
failures, 239, 382n53 family court, 180
federal and state governments as, health court, 177–78, 331–32
219 juror errors and, 172–73
income statements, 32 punitive damages and, 175–76
incurred but not reported losses, 34 role in tort reform, 95–97
inflation shocks and, 45 shopping among litigants, 181
initial incurred losses, 37 Juries
interest rate changes and, 44–45 blue ribbon, 331–32
internal conflict within, 40–41 collective wisdom of, 186
joint underwriting associations, 90, criticisms of, 163–65
227–28, 230–33, 235 errors made by, 170–73, 310,
limited-purpose, 227 362n28
oligopolistic behaviors, 39–44 guidance for, 118–19, 354n43
 Index 449

health courts and, 177–78, 185–86, Leapfrog Group, 202

331–32 Legal expense insurance, 155–60
injury scenarios and valuing Legal market, medical malpractice
nonmonetary loss, 122–23 blamed for crises, 30
liability determinations by, 165–72 lawyers role in, 8–9, 16
nontechnical nature of cases and, Legislation
165–66 common law and, 96, 141, 149–52,
punitive damage determinations by, 161, 344n39, 348–49n31
174–76 contingent fee, 141, 146–48
reaction to testimony, 117 disclosure, 207–10
survey approach to nonmonetary dram shop, 338n5
loss, 123–24 enterprise liability, 326
verdict registries, 355n54 Federal Tort Claims Act, 192
Juvenile courts, 180–81 FEMA, 273
Health Care Quality Improvement
Kahneman, Daniel, 173 Act of 1986, 203
Kansas, 231, 244 licensing, 207–10
patient compensation fund, 258 medical malpractice, 11–12
Karpoff, Jonathan M., 175 no-fault, 280
Keeley, Michael, 243 Patient Safety and Quality
Kelly, Carly N., 215 Improvement Act of 2005, 205
Kesselheim, A. S., 204 risk retention, 321
Kessler, D. P., 68, 74, 76, 103 Standard Family Court Act,
Keynes, John Maynard, 340n6 180–81
Keynesian economics, 340n6 state level tort reform, 94–97,
Kingston, R., 51 364n52
Kinney, Eleanor D., 94 tort award limits, 107
Klick, Jonathan, 102 workers’ compensation, 294
Kochanowski, Paul S., 296 LEI. See Legal expense insurance
Kritzer, Herbert M., 140, 144, 160, Lemstra, M., 296
358n30 Leslie, Phillip, 210
Kunreuther, Howard C., 253–54 Levitt, Steven D., 215
Licensure, physician, 207–10,
La Fountain, Neil, 175 371n24, 375–76n59
Lai, Gene C., 342n41 Life
Law-as-market, 11–12 expectancy, 126
Laws. See Legislation quality of, 121
Lawyers. See also Fees, lawyer value of, 121
campaign contributions to judges, Lillard, Lee A., 148
175 Limitations, statutes of, 89
current contingent fee system, Limited-purpose insurers, 227
135–40 Litan, Robert E., 44
incomes, 143–45 Litigation, 6. See also Judges; Juries
public opinion of, 135–36 arbitration versus, 128–29, 298–99,
role in malpractice litigation, 8–9, 16 347n13–14
specialization by, 138 bedside manner and, 77–79
450 Index

Litigation (cont.) Lott, John R., 175

challenges to damage caps, 113–14 Loughran, David S., 296
civil rights, 157–58 Louisiana, 258
class action suit, 140 LSPs. See Prepaid legal service plans
defensive medicine and, 54–55, 71– Lund, Adrian, 296
72, 313–14
deterrence of medical injuries and Maine, 91, 93
threat of, 79–81, 313–14 contingent fee system in, 141
discovery rules, 89, 203–4 Malpractice. See Medical malpractice
early offers and, 153–55, 162 Managed care organizations, 56–57
health courts and, 176–79, 183–84, Mandatory error reporting, 209
183–86, 183–87, 331–32, 364– Markets
65n53–54, 364n52 failures, 192–95, 256–57, 274
imperfections in current system of government activity, 11–12, 192–95
ascertaining payment from, insurance, 9–11, 46, 130, 252–53
114–16 knowns and unknowns in, 12–16
juror behavior in determination of legal, 8–9
liability in, 165–72, 354n43 medical care, 11
mediation versus, 347n15 power, 253
product liability, 140, 166, Martin, Joanne, 144
369–70n13 Massachusetts, 232
stages, 163–64 Matsa, D., 103
statutes of limitations, 89 McCann, Michael W., 160
threat and premiums, 53 McClellan, Mark, 74, 76, 103
tobacco, 165 MCO. See Managed care
U. S. culture and, 303–4, 312–13 organizations
vaccine injury, 287–91 Mechanic, D., 312
verdicts and post verdicts, 116, Mediation, 347n15
171–72 Medical care market, 11
Losses, insurer Medical Liability Monitor, 59
adjusted loss expense, 32, 44 Medical loss. See Injuries, medical
anticipated, 37, 387n41 Medical malpractice
developed, 37–38 claims volume, 18–20
earnings volatility and, 43 crises, 1–3, 7–8, 28, 30
effect on premium setting, 34–36, cycles, 27–31
252–53 deterrent effects of, 10
forecasting, 42–43, 60 insurance market, 9–11
incurred but not reported, 33 insurance premiums, 5–6
inflation and, 43 known and unknown variables in,
initial incurred, 37 12–16
internal factors affecting, 39–44 legal market and, 8–9
loss-reserving practices and, 268–70, market for government activity,
387n41–42, 387n45, 387n47 11–12
patient compensation fund effect on, medical care market, 11
267 mismatch between patient safety
uninsured, 360n81 and, 190–95
 Index 451

myths of, 16–20, 166–70, 173–74 National Association of Independent

nontechnical nature of disputes Insurers, 164
under, 165–66 National Flood Insurance Program,
patient access to medical care and, 273
69–71 National Practitioner Data Bank, 82,
patient travel time and, 70 118, 264
physician supply link to, 68–69 National Registry of Myocardial
positive defensive medicine and the Infarction, 208
cost of, 72–73 National Vaccine Injury
quantifying the costs of, 72–76 Compensation Program, 287–91,
reasons for studying, 217–18 305, 391n56, 391n59
tort law and, 3–4 Nebraska, 258
Medicare Negative defensive medicine, 54–55,
cardiac care, 74–76 70–71
claims data, 371n23 Negligence. See Errors, medical;
credentialing, 208 Safety, patient
indexing of fees by, 117 Neurological impaired infants,
pay-as-you-go financing, 280–87
268–69 Nevada
physician fees, 57, 333 medical malpractice in, 63
physician supply and, 57 physician supply in, 69
Meier, Kenneth, 237–38 New Hampshire, 146
Mello, Michelle M., 80, 102, 105 New Jersey, 63, 67
on hospitals directly employing medical error reporting in, 203–4
physicians, 215 Medical Inter-Insurance Exchange,
on medical evaluations of liability, 166
168 specialized courts in, 182
Mendelson, D. N., 228 New Mexico, 258
Mental health courts, 180, 181 New York
Mergenhagen, P. A., 207 experience rating in, 380–381n23
Mississippi, 69 medical error studies, 195, 197
Missouri, 97 medical malpractice in, 18–19
Moral hazard, 160, 243–44, 254, obstetrics in, 76
385n19 patient compensation fund, 258,
Mortality and anesthesiologists, 259
208–9 patient safety in, 190–91
Motor vehicles state guaranty fund, 234, 236
liability insurance, 225–26 New Zealand, 279, 295, 296,
no-fault programs, 295–97 300–302
Munch, P., 238 No Fault Patient Insurance Scheme
Myths of medical malpractice, 16–20, (NFPI), 297–300
166–70, 173–74 No-fault systems, 12–13, 92
administrative efficiency of, 284–85
National Association for the advantages of, 304–7
Advancement of Colored People, assuring coverage and reasonable
157 rates, 285–87
452 Index

No-fault systems (cont.) Nontangible benefits of medical care,

for birth injuries, 280–87 54
cost savings and, 277–78 NORCAL, 242–43, 383n73
early offers and, 153 North Carolina, 165
eligibility of injuries for, 281–83, North Dakota, 354n51
298 Nye, Blaine F., 39, 269
error reporting under, 303 Nye, David J., 41
failure to give notice and persistence
of tort in, 283–84 Obstetrics, 55–56, 59, 65, 76
health courts and, 331–32 bedside manner and, 77–78
for iatrogenic injuries, 279 no-fault systems and, 280, 282–83
insurance premium rates under, patient travel time for, 70
285–87 quality of physicians and, 81–82
motor vehicle, 295–97 reduction in services, 67
New Zealand, 279, 295, 296, Occurrence policies, 32, 384n4
300–302 O’Connell, Jeffrey, 92, 153
rationale for, 277–79 Office of Technology Assessment,
regulation, 331 U. S. (OTA), 71–72
scheduled damages and, 332–34 Ohio, 96–97
state programs for neurologically patient compensation fund, 258
impaired infants, 280–81 Oligopolist behaviors, 39–44
Swedish, 279, 295, 297–300 Olszynski, W. P., 296
for vaccine injuries, 287–91 Operating profits, 36–37
voluntary, 306, 329–31 Oregon, 150
workers’ compensation, 292–95 Ostensible agency, 322
Nondiversifiable risk, 35–36 OTA. See Office of Technology
Nonmonetary loss Assessment, U. S. (OTA)
award matrix for, 119–21
calculating awards for, 109–13, Pace, Nicholas, 117
355n63, 356n64 Partners Healthcare, 372n35
caps on awards for, 116–18 Patient compensation funds (PCFs),
compensation based on 13, 90, 344–45n43, 384n8
compensation for medical loss, adverse selection and, 267–68
124–25 compared to other types of
compensation in form of provision insurance, 272–74
of service benefits for, 125–31 continued provider vulnerability
flexible award ranges for, 122 and, 267
imperfections in current system of creation of, 258–59
ascertaining payment for, 114–16 effective, 275–76
injury scenarios and, 122–23 financing in advance or pay-as-you-
jury guidance and payment for, go financing of, 266–67, 269–70
118–19 funding of, 259, 265, 271,
jury survey approach to, 123–24 387n41–427
legal challenges to caps on, 113–14 implicit taxation funding of, 271
scheduling damages and, 119–25, lack of regulatory oversight in, 266,
132–34 272
 Index 453

loss-reserving practices, 268–70 Physicians

major provisions, 260–63 anesthesiology, 208–9, 224
mandatory participation in, 259, assessments of claims, 168–70
266 bedside manner, 77–79
as part of tort reforms, 258–59 cardiac care, 74–76
performance of, 264–65, 274–76 care levels and tort liability, 14–15
public sponsorship of, 265–66, 272 communication problems with
reduced incentives for loss hospitals and other physicians,
prevention, 267 373n44–45
structural features of, 265–67 computerized physician order-entry
subsidized from sources other than systems, 202–3, 372–73n40–41
provider premiums, 271 effects of premium increases on fees
Sweden, 297–300 charged by, 55
voluntary, 267–68 enterprise insurance and, 319–20
weaknesses of, 267–71 enterprise liability and, 327
Patients. See also Safety, patient evaluations of liability by, 167–68
access to information about medical experience rating of, 212–14,
errors, 209 228–30, 378–79n89
access to medical care, 53, 55, incomes, 11, 52
69–71 licensure and certification, 207–10,
consent forms, 89–90 371n24, 375–76n59
as consumers, 11 obstetric, 55–56, 59, 65, 67,
safety organizations (PSOs), 204–5 70, 76, 77–78, 81–82, 280,
travel time, 70 282–83
Patient Safety and Quality over-deterrence behavior, 11
Improvement Act of 2005, 205 peer review, 207–9, 374–75n47
Pauly, M. V., 51, 61, 253–54 planned exits from practice, 66
PCFs. See Patient compensation funds political activism by, 67–68
(PCFs) premiums as a fixed cost to, 55–56
Peer review, 207–9, 374–75n47, premiums as variable cost to, 57
376n64 quality of, 17, 81–82, 345n51
Pennsylvania relocation by, 67
enterprise insurance in, 316 report cards, 207–10
insurer solvency in, 240–42 risk classification of, 222
joint underwriting association, 232 and second-generation reforms, 94
laws on contingent fees, 146 specialties and geographic locations,
medical malpractice in, 7, 63, 62, 377n70
383n73 -sponsored medical malpractice
patient compensation fund, 258, insurers, 9, 226–27
259, 271, 275, 384n8 supply, 56–57, 62–69, 102–4, 191
physicians directly employed by surgeon, 65, 197
hospitals in, 215 surveys, 65–66
physician supply in, 65, 67, 69 Swedish, 299
Periodic payments, 102, 126–27, Policies, insurance, 31–32, 384n4
344n41 adhesion, 398–99n62
PHICO, 239–42 capacity constraints on, 45–47
454 Index

Politics of tort reform, 93–95 trends in, 58–60

Posey, L., 47 as a variable cost to physicians
Positive defensive medicine practice, 57
in cardiac care, 74–76 Prepaid legal service plans, 157–59
cost of medical malpractice and, Price wars among insurers, 41,
72–73 221–22
effects of tort reforms on, 76–77 Private contracting, 93, 131–32, 224,
versus negative defensive medicine, 310–11
54–55, 58 Private reinsurance, 250–55
in obstetrics, 76 Product liability litigation, 140, 166,
patient outcomes, 75–76 369–70n13
Prejudgment interest, 344n38 Profits, operating, 36–37, 343n8
Premiums, 5–6, 15 Property-casualty insurance. See
assessing adequacy of medical Insurance, property-casualty
malpractice, 223 Public policy
capacity constraints and, 45–47 punitive damages and, 174–76
claim severity and, 52–53 responses to awards, 116–18
competition among oligopolists and, Public Policy Institute, 177, 179
39–44 Public regulation, critique of,
effects of first-generation reforms 236–38
on, 100–102 Punitive damages, 174–76, 344n37,
empirical evidence on effects of 354–55n53, 361n5, 363–64n50,
changes in, 60–62 363n46
as a fixed cost to physician’s
practice, 55–56 Quality
inflation shocks and, 45 assurance mechanisms, 206–10
insurance cycles and, 29, 30 of care and negligence, 197–98
insurer’s income statements and, 32, of life, 121
341n17 market forces and, 194
interest rates and, 44–45, of physicians, 17, 81–82, 345n51
342n41
under no-fault systems, 285–87 Rabin, Robert L., 95, 347n10
paid before claims paid, 31–32 Rand Corporation, 144, 164–65
patient compensation funds and, Reciprocal insurers, 227
259, 265 Reform
physician supply changes reflected insurance
in, 56–57 enterprise liability, 326
positive defensive medicine and evaluating, 98ñ100
changes in, 57–58 first-generation, 90
prepaid legal services plans, 158 second-generation, 90ñ93
regulation of, 43, 221–22 tort
reinsurance, 249, 251 achievable, 2
setting, 34–36 alternative dispute resolution and, 92
terrorism effect on, 46, 47–48, 49, centrist, 2ñ3
255 claims frequency, damages awarded,
threat of litigation and, 53 and premium effects of, 100ñ102
 Index 455

court challenges of, 95ñ97 Reilly, Bridget A., 76, 296

differences in recommendations on, 2, Reinsurance, 47–49, 219
309ñ11 attachment points, 249–50
difficulties in evaluating, 98ñ100 defined, 247
effect on physician supply, 102ñ4 government intervention in, 255–57,
enterprise liability in, 92ñ93, 211ñ12 274, 329, 386n30
first-generation, 86ñ90, 98ñ100 importance of, 247
laws, 82ñ83 insurance cycles and, 255–56
motivations for, 85ñ86 large claims and, 248–49
patient compensation funds as part of, market failures in, 256–57, 274
258ñ59 policies as illiquid financial
patient safety and, 191ñ92, 211ñ14 instruments, 254
politics of, 93ñ95 premiums, 249, 251
positive defensive medicine and, public intervention in, 257, 329
74ñ77 reasons for the high cost of, 252–55
second-generation, 90ñ93, 104 Rendell, Edward G., 7
as a state issue, 94ñ97 Report cards, physician, 207–10
unintended consequences of, 98ñ100 Reserves, loss, 268–70
Regulation Respondeat superior, 322
alternatives to government, 224–25 Restaurant hygiene grading, 210
banking, 243–44 Reynolds, R. A., 72, 73, 79
certification, 207–10 Rhode Island, 232, 383n73
contingent fee, 148–49 Risk
credentialing, 207–10 classification, 9–10, 222
critique of public, 236–38 diversification of, 35–36
disclosure statutes, 207–10 pooling, government, 230–36
effects of, 238–43 retention groups, 227, 320–21
important features of, 218–20 setting premiums and, 34–36
for information provision and fair Roane, Spencer, 96
play, 223 Rodwin, M. A., 58
insurance, 43, 217–18, 236–43, Runaway juries, 164–65, 174
382n51, 397n35 Russell, T., 253–54
insurer failures and, 239–42
licensure, 207–10 Safety, patient. See also Errors,
no-fault programs, 331 medical; Injuries, medical; Patients
patient compensation funds, 266, data tracking, 199
272 Freakonomics and, 214, 215
patient safety, 199 government oversight of, 206–10
peer review, 207–10 in hospitals, 195–98
premium, 43, 221–22 Institute of Medicine’s
quality assurance, 206–10 recommendations regarding,
rationale for, 220–23 198–203
reinsurer, 48, 49 market failure in medical care and,
report cards, 207–10 192–95
solvency, 238, 239–44 mismatch between medical
state government role in, 219–20 malpractice and, 190–95
456 Index

Safety, patient (cont.) on primary insurer decisions to

organizations (PSOs), 204–5 reinsure, 250
outpatient, 376n66 on the quality of physicians, 82
public awareness of, 189–90 on regulation, 207, 239
restructuring tort law to improve, Smallwood, D., 238
211–14 Snyder, Edward A., 151
state legislatures involvement in, Social Security, 117, 268–69
201–2 Solvency regulation, 238, 239–44
voluntary reporting on, 198 Songer, Donald R., 94
Sage, W. M., 103 South Carolina, 94
Saks, Michael, 136 joint underwriting association,
Savings and loan banking, 243–44 232
Scheduled damages, 92, 132–34 patient compensation fund, 258,
no-fault systems and, 332–34 387–88n50
nonmonetary, 119–25 Specialty courts, 179–84, 368n85
Schenzler, Christopher, 296 St. Paul Group of Companies, 42, 43,
Schkade, David, 173 226
Schuck, Peter, 131 Stages of litigation, 163–64
Schwartz, W. B., 228 Standard Family Court Act, 180–81
Schwarzenegger, Arnold, 294, 295 Standards of care, 90, 91, 198–99
Second-generation reforms, 90–93, State governments
104 contingent fees and, 146–49
Self-insured organizations, 218, 227, insurance regulation by, 219–20
248, 386n28 as insurers, 219
September 11, 2001 terrorist attacks, joint underwriting associations,
47–48, 49, 255 230–33
Severity patient safety policies, 201–2
and frequency of claims, 31, 52–53, property-casualty insurance provided
100–102 by, 273–74
injury, 120 tort reform, 94–97, 364n52
Shifting, fee, 149–52 workers’ compensation programs,
Shocks, inflation, 45 292–93
Sloan, Frank A. State guaranty funds, 230–31, 233–
on contingent fees, 148 35, 236
on the contingent fee system, 138 Statutes of limitations, 89
on damage caps, 99 Statutory lawyer fees, 140
on enterprise insurance, 318 Steves, Myron, 324
on experience rating, 213, 229 Stock insurers, 222, 224
on jury evaluations of liability, 167, Stratmann, Thomas, 102
170 Studdert, David M., 102, 105, 109
on the malpractice insurance on hospitals directly employing
market, 130 physicians, 215
on no-fault systems, 283, 296 on medical evaluations of liability,
on nonmonetary loss, 119 168
on positive defensive medicine, 76 on no-fault, 284, 299–300
on premiums, 223 Sunstein, Cass R., 173
 Index 457

Supply no-fault approaches to, 12–13, 92,

hospital, 65 277–79
physician, 56–57, 62–69, 102–4 patient safety and restructuring of,
Supreme Court, U. S. See Courts 211–14
Surgeons, 65, 197 physician care levels and, 14
Surplus lines carriers, 10, 14, 47, private contracting and, 93,
227 131–32
Sweden, 279, 295, 297–300, 301–2 standards of care under, 90, 91
Switzerland, 156 statutes of limitations, 89
tort-feasors, 83, 89
Tabarrok, Alexander, 148, 175 Travel time, patient, 70
Taragin, Mark I., 166, 167 Trial lawyers. See Lawyers
Task-based lawyer fees, 140
Taxation, implicit, 271 Underreporting of medical errors,
Tax courts, 179–80, 183, 368n84 203–6
Terrorism, 46, 47–48, 49, 255, 272 Underwriting standards, 340n1
Texas, 64, 118, 357–58n28 Uninsured loss recovery, 360n81
enterprise liability in, 324 Utah, 195, 299–300
lawyer incomes in, 144–45
medical error reporting in, 204 VA. See Veterans Administration (VA)
Thimersol, 290, 305 Vaccine Adverse Event Reporting
Thistle, Paul D., 47 System (VAERS), 391–92n60
Thompson v. Nason, 323 Vaccine injuries, 287–91, 305,
Thorpe, Kenneth E., 102, 105 391–92n59–60, 391n56
Time-based lawyer fees, 140, 142–43 Vaccine Safety Datalink (VSD),
Tobacco litigation, 165 392n60
To Err is Human (Institute of Value-based lawyer fees, 140
Medicine), 189, 195, 196, 198–99 Value of life, 121
impact of, 205–6 Veterans Administration (VA), 192,
recommendations for improving 202
patient safety, 198–203 Vidmar, Neil J., 165, 173
Tort system, the. See also Reform, Virginia, 92, 279, 280–87, 330, 334
tort Viscusi, W. Kip, 102, 171, 175, 239
advocates of, 5 Voluntary no-fault systems, 306,
compensation under, 4, 87, 89 329–31
consent forms and, 89–90 Voluntary patient compensation
current, 311–14 funds, 267–68
experience rating and, 212–14
first- and second-generation reforms Wall Street Journal, 164
of, 86–93 Washington, 64
functions and goals of, 337–38n4 Waste, care, 54
indicators of problems with, 4–6 Weaknesses of patient compensation
joint and several liability under, 89 funds, 267–71
laws, 3, 82–83, 344n42 Webb, Bernard L., 41
motivations for reforming, 85–86, Weiler, Paul C., 80, 279, 396n18
309–11 West v. Johnson & Johnson, 166
458 Index

West Virginia
medical malpractice in, 64
physicians directly employed by
hospitals in, 215
physician supply in, 67, 69
Wheelock, David C., 244
Whetten-Goldstein, K., 82, 283
White, Michelle J., 167
Wilson, Paul W., 244
Wisconsin, 13, 358n30
lawyer incomes in, 144, 145
patient compensation fund, 258, 259
Woodhouse, Owen, 300, 394n118
Workers’ compensation, 292–95
Wyoming, 258

Yang, Y. T., 102, 105

Young, T., 296

Zador, Paul, 296

Zeckhauser, Richard J., 102