Mnsure PDF

OLA
OFFICE OF THE LEGISLATIVE AUDITOR

STATE OF MINNESOTA
EVALUATION REPORT
Minnesota Health
Insurance Exchange
(MNsure)
FEBRUARY 2015
PROGRAM EVALUATION DIVISION
Centennial Building Suite 140
658 Cedar Street St. Paul, MN 55155
Telephone: 651-296-4708 Fax: 651-296-4712
E-mail: legislative.auditor@state.mn.us Website: www.auditor.leg.state.mn.us
Through Minnesota Relay: 1-800-627-3529 or 7-1-1
Program Evaluation Division
Evaluation Staff
The Program Evaluation Division was created

within the Office of the Legislative Auditor (OLA)
in 1975. The divisions mission, as set forth in law,
is to determine the degree to which state agencies
and programs are accomplishing their goals and
objectives and utilizing resources efficiently.
James Nobles, Legislative Auditor
Topics for evaluations are approved by the

Legislative Audit Commission (LAC), which has
equal representation from the House and Senate
and the two major political parties. However,
evaluations by the office are independently
researched by the Legislative Auditors professional
staff, and reports are issued without prior review by
the commission or any other legislators. Findings,
conclusions, and recommendations do not
necessarily reflect the views of the LAC or any of
its members.
OLA also has a Financial Audit Division that
annually audits the financial statements of the State
of Minnesota and, on a rotating schedule, audits
state agencies and various other entities. Financial
audits of local units of government are the
responsibility of the State Auditor, an elected office
established in the Minnesota Constitution.
OLA also conducts special reviews in response to
allegations and other concerns brought to the
attention of the Legislative Auditor. The Legislative
Auditor conducts a preliminary assessment in
response to each request for a special review and
decides what additional action will be taken by OLA.
For more information about OLA and to access its
reports, go to: www.auditor.leg.state.mn.us.
Joel Alter
Caitlin Badger
Valerie Bombach
Sarah Delacueva
Jody Hauer
David Kirchner
Laura Logsdon
Carrie Meyerhoff
Ryan Moltz
Judy Randall
Catherine Reed
Jodi Munson Rodriguez
Laura Schwartz
KJ Starr
Jo Vos
To obtain reports in electronic ASCII text, Braille,
large print, or audio, call 651-296-4708. People with
hearing or speech disabilities may call through
Minnesota Relay by dialing 7-1-1 or 1-800-627-3529.
To offer comments about our work or suggest an
audit, investigation, or evaluation, call 651-296-4708
or e-mail legislative.auditor@state.mn.us.
Printed on Recycled Paper
OL A
OFFICE OF THE LEGISLATIVE AUDITOR

STATE OF MINNESOTA James Nobles, Legislative Auditor
February 2015
Members of the Legislative Audit Commission:
The federal Affordable Care Act aimed to reduce the number of people without health insurance
and improve the ability of consumers to compare their insurance options. Rather than relying on
the federal governments online health insurance exchange for this purpose, Minnesota chose to
develop its own exchange. The Legislature created an agency (MNsure) to develop the
exchange, and MNsures web-based enrollment system opened in October 2013.
In Spring 2014, the Legislative Audit Commission directed the Office of the Legislative Auditor
to evaluate MNsure. In our view, the establishment of Minnesotas health insurance exchange
was well intended, but it was not well executed. For a variety of reasons, the online enrollment
system was seriously flawed, and MNsures customer service was inadequate. Our report
focuses mainly on MNsures first year of enrollment, and we recommend changes in the
governance of both the agency and its enrollment system.
Our evaluation was conducted by Joel Alter (project manager), Ryan Moltz, and Laura Schwartz.
We received full cooperation from MNsure, the Department of Human Services, and the Office
of MN.IT Services.
Sincerely,
James Nobles
Legislative Auditor
Room 140 Centennial Building, 658 Cedar Street, St. Paul, Minnesota 55155-1603 Phone: 651-296-4708 Fax: 651-296-4712
E-mail: legislative.auditor@state.mn.us Website: www.auditor.leg.state.mn.us Minnesota Relay: 1-800-627-3529 or 7-1-1
Table of Contents
Page
SUMMARY
ix
INTRODUCTION
1. BACKGROUND
Federal Health Care Reform
Minnesotas Health Insurance Exchange
Context for Minnesotas Exchange
5
5
8
23
2. EXCHANGE DEVELOPMENT AND IMPLEMENTATION

Timeliness Issues
Website Functionality Issues
25
26
34
3. GOVERNANCE
Board Role
Board Policies
Compliance Issues
Governance of the MNsure Enrollment System
Communications Issues
General Governance Recommendations
47
47
48
50
51
53
55
4. ENROLLMENT
Total Enrollment
Enrollee Characteristics
Enrollment Projections
MNsures Impact on the Uninsured Population
Comparison with Other States
59
59
63
67
70
74
5. USER EXPERIENCES
User Expectations
Consumer Enrollment Process
Consumer Satisfaction with Products Purchased
Impacts on Key Stakeholders
77
77
79
92
95
6. OPERATIONS
Consumer Assisters
Contact Center
Advertising
Data Security
101
101
113
125
127
LIST OF RECOMMENDATIONS
131
APPENDIX A: Glossary of Terms

APPENDIX B: Key Events in MNsures Development and
Implementation
133
139
vi
MINNESOTA HEALTH INSURANCE EXCHANGE (MNsure)
Page
APPENDIX C: Health Insurance Rating Areas in Minnesota
APPENDIX D: 2014 Federal Poverty Guidelines
APPENDIX E: Current Assessment of Gruber-Gorman
Projections
APPENDIX F: MNsure Grants to Promote Enrollment and
Consumer Outreach, October 2013 to September 2014
143
145
AGENCY RESPONSE
155
RECENT PROGRAM EVALUATIONS
169
147
151
List of Exhibits
Page
1. BACKGROUND
1.1 Types of State Health Insurance Exchanges for Individuals, 2015
1.2 Statutory Purposes of MNsure
1.3 Key Differences between MNsure and Other State Agencies
1.4 Statutory Requirements for Composition of the MNsure Board
1.5 Nonelderly Adults Income Eligibility for Minnesotas Public
Health Care Programs and Federal Subsidies
1.6 MNsure 2014 Budget and Draft Budgets for Fiscal Years 2015
to 2017
1.7 Types of MNsure-Certified Consumer Assisters and State
Compensation per Enrollee, 2014
2. EXCHANGE DEVELOPMENT AND IMPLEMENTATION
2.1 Key Actions Related to Creation of a Health Insurance Exchange
in Minnesota
2.2 Federal Grants Awarded to Minnesota to Plan and Establish a
Health Insurance Exchange
2.3 Key Steps Related to Health Insurance Exchange Contracts for
Technical Infrastructure
2.4 Subcontractors Responsible for Technical Components of
Minnesotas Exchange
2.5 Examples of Key Federal Rules Governing Health Exchanges
Issued in 2013
2.6 Independent Assessments of Overall Project Status for MNsure,
April 2013 to January 2014
2.7 Target and Actual Dates for Completion of MNsure System
Testing
2.8 Examples of MNsure Technical Problems Cited by Deloitte,
June 2014
3. GOVERNANCE
3.1 Reasons for Changing or Keeping the MNsure Boards Role
4. ENROLLMENT
4.1 Characteristics of MNsure Enrollees (October 1, 2013, to
June 30, 2014)
4.2 Enrollment of Nonelderly Adults as a Share of the Uninsured
Population by Region (October 1, 2013, to June 30, 2014)
4.3 Median Monthly Amount that MNsure Enrollees Selected for
Advanced Premium Tax Credit (October 1, 2013, to
June 30, 2014)
4.4 Projected and Actual Enrollments for the First Open Enrollment
Period (October 1, 2013, to March 31, 2014)
7
9
10
12
16
20
22
27
28
29
30
32
37
39
42
56
64
65
68
69
viii
Page
5. USER EXPERIENCES
5.1 Examples of Potentially Confusing Application Questions from
the First Open Enrollment Period
5.2 Perceptions of MNsure Enrollees Who Had Prior Experience
Buying Insurance, 2014
5.3 Satisfaction of Consumer Assisters with Tools MNsure Provided
to Consumers, 2014
5.4 User Perceptions of Information MNsure Provided on Health
Insurance Products, 2014
5.5 Consumer Assister Satisfaction with Information Provided by
MNsure on Specific Elements of Insurance Products, 2014
5.6 Enrollees Reasons for Buying Insurance through MNsure Rather
than Directly from an Insurer
5.7 Enrollee Perceptions of Positive and Negative Effects of MNsure,
2014
5.8 MNsure Enrollees Views of Certain Aspects of the Insurance
They Purchased, 2014
5.9 Enrollees Comparison of Insurance Purchased Through MNsure
with Insurance They Had Previously, 2014
6. OPERATIONS
6.1 Overview of MNsure Consumer Assisters
6.2 Average Amount of Time Consumers Spent Waiting on Hold,
October 2013 to August 2014
6.3 Percentage of Phone Calls to the MNsure Contact Center
Abandoned by Callers, October 2013 to August 2014
6.4 Projected and Actual Contact Center Staffing and Performance,
October 2013 to August 2014
6.5 Contact Center Staffing Levels, September 2013 to August 2014
6.6 Contact Center Call Volume, October 2013 to August 2014
6.7 MNsures Key Marketing and Outreach Contracts for First Open
Enrollment Period
85
86
87
88
89
90
93
94
95
102
114
115
118
119
121
126
Summary
But the initial enrollment process
was often lengthy, and technical
problems frustrated consumers,
insurers, and counties. (pp. 80, 81,
92, 95)
Key Facts
In its first year of

operations,
MNsures failures
outweighed its
achievements.
Minnesota is 1 of 13 states that are

enrolling individuals in health
insurance through state-based
exchanges, which facilitate the
comparison and purchase of health
insurance. (pp. 6-7)
Minnesotas health insurance
exchangecalled MNsureshares
some similarities with executive
branch state agencies. But the
exchange also has important
differences, such as its governance
by a board. (p. 9)
About 28 percent of MNsure

enrollees said they were uninsured
immediately before enrolling.
(p. 73)
During its first year, MNsure failed

to provide adequate customer
service through its call center.
Also, the roles played by consumer
assisters were not sufficiently clear,
so consumers were often referred
back and forth among them.
(pp. 106, 113-122)
Key Findings
MNsure implemented its enrollment

website in 2013 with serious
technical problems. It did not
adequately test the site, and it made
insufficient use of state government
technology experts. (pp. 25, 35, 38)
Federal law imposed an ambitious
timeline on states developing
exchanges. This challenge was
heightened by late federal rules,
delays in passing state legislation,
and problems with vendor selection
and performance. (pp. 26-33)
MNsure staff withheld information

from the MNsure Board and other
key officials before the enrollment
website was launched. (p. 53)
MNsure reported that it met its

overall enrollment target in its first
open enrollment period, but the
target was seriously flawed. The
target contained an error that
resulted in an unrealistically low
estimate. (p. 69)
Many people who bought insurance

through MNsure have been satisfied
with the products they purchased.
Key Recommendations:
The Legislature should amend state

law to give the governor, rather
than the MNsure Board, authority to
appoint the MNsure chief executive
officer. In addition, the Legislature
should consider whether to retain
the MNsure Board as a governing
body or to make it purely advisory.
(p. 55)
The Legislature should amend

statutes to formally create a
governance structure for MNsures
enrollment system and ensure that
MNsures future information
technology work is subject to
oversight from the Office of MN.IT
Services. (pp. 36, 52)
MNsure and DHS should ensure

that insurance brokers are fairly
compensated for enrolling
consumers through MNsure.
(p. 109)
MNsure should improve its ability

to access and analyze the applicant
and enrollee data it collects. (p. 62)
Report Summary
In 2010, President Obama signed into
law the Affordable Care Act. Among
other things, the act authorized the
establishment of health insurance
exchanges to help people compare
and purchase insurance online.
Many states rely on the federal
governments exchange for this
purpose, but Minnesota established its
own exchange, called MNsure. In 2011,
Governor Dayton directed the
Department of Commerce to develop a
state-based exchange. Minnesota has
received $189 million in federal grants
for this purpose.
The exchange began enrolling
individuals in health insurance in
October 2013. Individuals who
purchase commercial insurance through
MNsure may qualify for tax credits that
are not available to individuals who
purchase insurance outside of MNsure.
Also, unlike most other states,
Minnesota relies on its exchange to
make eligibility determinations for its
publicly funded health care programs,
mainly Medical Assistance and
MinnesotaCare.
Multiple factors complicated the
already difficult challenge of building
a health insurance exchange by
October 2013.
In mid-2011, the Department of
Commerce solicited vendors to build an
online enrollment website. The
contracting process took longer than
expected, and exchange officials grew
dissatisfied with the lead vendor just
months into the contract.
The federal government required
Minnesota to establish legal authority
for its exchange. However, the
Minnesota Legislature did not formally
create MNsure until March 2013
about six months before the exchange
began enrolling people. In addition, the
federal governments rules that

indicated how exchanges should operate
trickled out piecemeal, well into 2013.
These delays created additional time
pressures on state officials and vendors
developing the online enrollment
system.
MNsure sought limited technical
advice from state experts and did too
little testing of the system.
Experience has shown that it is difficult
to implement big information
technology projects on time, within
budget, and with all the expected
features. The development of the
MNsure enrollment system was large
and complicated, and state officials
undertook this project with limited
technical expertise.
The 2011 Legislature created the Office
of MN.IT Services to oversee all
executive branch information
technology projects. But state officials
building the health insurance exchange
initially shunned this office, and the
Legislature later exempted MNsure
from most oversight by the Office of
MN.IT Services. MNsures limited use
of this agency may have contributed to
technical problems that arose during
development of the exchange. The
Office of MN.IT Services became more
involved in MNsure activities in 2014,
but we recommend a statutory change to
ensure this offices continued role in
overseeing MNsures technology
development.
Various red flags in the weeks and
months before October 2013 suggested
that the launch of the online exchange
might not go well. For example, an
independent contractors reviews of the
exchange raised serious doubts about its
readiness. MNsure staff did not share
this information with MNsures
governing board.
MNsure failed to adequately test the
exchanges website before enrollment
SUMMARY
xi
began. The tests were limited in

number and scope. The tests showed
many problems, and there was little
time to address them.
MNsures technical problems escalated
in late 2013 and continued well into
2014. For example, applications got
stuck in the system, and MNsure could
not easily make changes to individuals
insurance coverage in response to
events such as births or income changes.
MNsure enrolled many people in its
first year, but its overall enrollment
target was flawed.
MNsure reported 371,000 health
insurance enrollments during its first
year. About 56,000 (15 percent) of
these were in commercial insurance, and
the remainder were in public health care
programs.
MNsure set enrollment targets in
October 2013. The overall target for the
first open enrollment period (through
March 2014) included an erroneous
Department of Human Services
projection. Specifically, the department
estimated that only about 12,000 people
would enroll in Medical Assistance over
a six-month period. If a more realistic
estimate of Medical Assistance
enrollment (perhaps over 100,000) had
been included in MNsures overall
target, actual enrollments for the first
open enrollment period would have
fallen far short of the overall target.
The online enrollment system was built
without authoritative documentation of
consumers enrollment choices. As a
result, it was difficult during the first
year to use MNsure records to
definitively determine who enrolled and
in which insurance products. MNsures
enrollment system lacks good reporting
capabilities, making it difficult for
MNsure staff to extract data for
management and decision-making
purposes. We recommend that MNsure
address this weakness.
A 2014 analysis by University of

Minnesota researchers indicated that the
number of uninsured Minnesotans fell
significantly after MNsure opened for
business. The impact of MNsure on this
reduction is unclear; other factors, such
as Minnesotas expansion of its Medical
Assistance program, may have played a
role in this reduction. We surveyed
individuals who enrolled in commercial
insurance through MNsure, and
28 percent said they were uninsured
immediately prior to buying insurance
through MNsure.
MNsures technical problems
frustrated consumers, although many
were satisfied with the products
available through MNsure.
When MNsure was in development,
state officials said enrollment would be
simple and user-friendly. But a
majority of people we surveyed who
bought commercial insurance through
MNsure said it took more than four
hours to do so. Most said they
experienced significant technical
problems. Insurers and counties also
had major problems using the MNsure
system to manage cases.
MNsures technical problems caused the
Department of Human Services to defer
until 2015 its plans to use MNsure to
reexamine the eligibility of most
individuals who had been enrolled
before October 2013 in the states
public health care programs. These
eligibility reviews are supposed to occur
annually, but some individuals in
Minnesotas public programs had gone
two years without them, as of late 2014.
In our survey, most people who
purchased commercial insurance
through MNsure told us they would
choose the same product again, if given
the chance. Survey respondents
reported mixed views when asked
whether the premiums and out-ofpocket costs of products purchased
through MNsure were better than
xii
insurance they had immediately prior to

buying insurance through MNsure.
The Legislature should reconsider

MNsures governance arrangement.
During MNsures first year, many

consumers were not notified about
(1) the status of their applications for
insurance or (2) their eligibility for
public programs or tax credits. Thus,
many did not know whether they had
obtained insurance through MNsure or
what they needed to do to complete
their applications.
MNsure is governed by a seven-person

boardthe Department of Human
Services Commissioner plus six
members appointed by the governor.
The board appoints MNsures chief
executive officer.
MNsure provided inadequate

consumer assistance during its first
year.
MNsures customer service center did
not answer calls within an acceptable
amount of time. MNsure understaffed
the center, did not have a plan for
handling technical questions, and
provided insufficient training to staff.
In MNsures first 11 months, about
one-third of calls were abandoned.
MNsure also arranged for in-person
enrollment assistance from networks of
MNsure-certified navigators and
insurance brokers, but there were
significant problems. A majority of
consumer assisters we surveyed said the
training they received from MNsure was
inadequate. MNsure certified a limited
number of assisters before open
enrollment started in October 2013.
MNsure provided weak oversight of
grants it gave to organizations that
helped consumers.
The roles played by various types of
consumer assisters were, at times,
confusing and inefficient. For example,
some assisters were not authorized to
offer advice on insurance productsso
consumers seeking advice had to be
referred to other assisters who could
provide this. Some insurance brokers
helped consumers enroll, only to find
they did not qualify to be compensated
for their work.
The Governor appointed MNsures

board members in late April 2013. By
law, the board could not assume its full
authority until it adopted internal
policies and bylaws, which it finished in
mid-August 2013. As a result, the
board had little influence over exchange
operations before the launch of
MNsures website. Also, MNsure staff
provided board members, the Governor,
and others with limited information in
2013 about the exchanges operational
readiness.
In our view, an agency with MNsures
impact and visibility should be directly
accountable to the governor. There is
some precedent for an agency having
both a governor-appointed administrator
and board (for example, the Minnesota
Pollution Control Agency), but the
Legislature should consider what future
role it wants the MNsure Board to play.
The Legislature should create in law a
governance structure for MNsures
enrollment system; the current multiagency structure is entirely informal.
MNsures online system is used for
enrollments in both commercial
insurance and public health care
programs. The Department of Human
Services, which administers the public
programs, wants more explicit authority
to participate in decisions about this
system.
Introduction
T
he Affordable Care Act of 2010 significantly changed the federal

governments involvement with health care insurance. First, it mandated
that most people have coverage through a public program or a private health
insurance policy. In addition, the law authorized government-created, online
health insurance exchanges (or marketplaces) to help individuals and
businesses compare and purchase health insurance. Each state had the option of
creating its own exchange, relying on the federal governments exchange, or
sharing exchange responsibilities with the federal government. Minnesota chose
to develop its own online exchange, known as MNsure. Consumers could begin
enrolling in public or private health insurance through MNsure on October 1,
2013, with coverage starting January 1, 2014, or later.
While many people have successfully enrolled in insurance through MNsure,
many others have encountered problems. In April 2014, the Legislative Audit
Commission directed the Office of the Legislative Auditor to evaluate the
development and implementation of MNsure. We asked the following questions:
To what extent has Minnesotas state-based health insurance

exchange helped consumers make informed choices?
What are the characteristics of MNsure enrollees? Has the exchange

reduced the number of Minnesotans who lack insurance?
Was the exchange adequately planned and implemented? Has

MNsure provided sufficient outreach, marketing, and applicant
assistance? What has contributed to MNsures technical and
operational problems?
Has MNsures governing board exercised sufficient oversight of the

exchanges operations?
To address these questions, we solicited input through interviews and written

information requests from MNsure staff and board members; staff from the
Office of MN.IT Services (which provides information technology services to
state agencies) and the Minnesota departments of Human Services, Health, and
Management and Budget; representatives of key MNsure stakeholders, such as
counties and health plans; and federal human services officials. We requested to
interview the person who served as director of Minnesotas exchange from 2011
until her resignation in December 2013; she did not respond directly to our
requests.1
We invited the former MNsure executive director to speak with us, and a second time we sent her
a subpoena, but she did not respond either time. Her attorney contacted us on several occasions,
but his conditions for an interview with the former director were not acceptable. Primarily, her
attorney said the former director would only cooperate if the Office of the Legislative Auditor paid
all expenses she incurred by cooperating, including her attorneys fees.
MINNESOTAS HEALTH INSURANCE EXCHANGE (MNsure)
We analyzed a variety of documents, including statutes, contracts, planning

documents, federal grant applications, marketing materials, board minutes, and
many others. We considered the results of three prior reports by our office: two
financial audits and a special review.2 We examined and critiqued the MNsure
enrollment website, but we did not contract with technical experts to assess its
functionality; the websites technical issues have received extensive scrutiny
from MNsure, the Office of MN.IT Services, and various contractors. We also
reviewed key decisions and developments leading to the enrollment systems
implementation, plus post-implementation reports that evaluated the systems
adequacy. We did not independently test the software components of MNsures
enrollment system or their security.
We administered several questionnaires. We surveyed representative random
samples of three types of MNsure-certified consumer assisters: navigators,
certified application counselors, and insurance brokers.3 We also surveyed a
representative random sample of nonelderly adult heads of household who
enrolled through MNsure in commercial insurance from October 2013 through
June 2014.4
We obtained data on all individuals who enrolled through MNsure as of June 30,
2014. Among other things, we used these data to analyze the characteristics of
enrollees, such as where they lived and the types of insurance they enrolled in
through MNsure. We also obtained data about the operation of MNsures
customer service center, as well as data on grants made by MNsure.
We reviewed the MNsure enrollment website, using a test version of the site. We
were told that this version generally reflected the content and appearance that
users experienced during the first open enrollment period. However, because
there had been some changes to improve the functionality of the website after it
opened October 1, 2013, and because the test version did not directly interact
2
Office of the Legislative Auditor, Minnesota Health Insurance Exchange: MNsure; Internal
Controls and Compliance Audit, July 2011 through December 2013 (St. Paul, October 28, 2014),
http://www.auditor.leg.state.mn.us/fad/pdf/fad1421.pdf; Department of Human Services:
Oversight of MNsure Eligibility Determinations for Public Health Care Programs, October 2013
through April 2014 (St. Paul, November 12, 2014), http://www.auditor.leg.state.mn.us/fad/pdf
/fad1422.pdf; and MNsure: An Unauthorized Disclosure of Private Data; Special Review (St. Paul,
November 7, 2013), http://www.auditor.leg.state.mn.us/fad/pdf/fad1327.pdf.
3
Not counting members of our samples whose contact information was incorrect, we sent survey
invitations by e-mail to 313 navigators, 292 certified application counselors, and 547 brokers. We
received responses from 222 navigators (71 percent), 167 certified application counselors
(57 percent), and 295 brokers (54 percent). The survey results reported in this evaluation only
include those of respondents who said they did, in fact, work with MNsure applicants during the
first open enrollment period. The samples of respondents were large enough that we can be
95 percent confident that the true percentage of the population who would have selected a
particular response to a survey question was within 5.4 to 7.5 percentage points of the survey
respondents answers, depending on the survey question and the group surveyed.
Not counting members of our sample whose contact information was incorrect, we sent survey
invitations by U.S. mail to 975 MNsure enrollees. We received responses from 281 enrollees
(29 percent). The sample of respondents was large enough that we can be 95 percent confident that
the true percentage of the population who would have selected a particular response to a survey
question was within about 6 percentage points of the survey respondents answers.
INTRODUCTION
with federal databases, our use of the test version did not fully mirror the
experiences of actual users.
In addition, we examined existing literature regarding the design and
development of health exchanges under the Affordable Care Act. For other
states that operate exchanges with governing boards, we examined provisions in
state laws regarding those boards.
Our evaluation focused largely on development and implementation of
Minnesotas health insurance exchange up to Fall 2014. This represented the
period through MNsures first year of enrollment. We did not evaluate in depth
MNsure operations during the second open enrollment period (which started
November 15, 2014), and we did not independently assess the functionality of
MNsures enrollment system during that period. However, our report offers
comments about MNsure activities and operations as of late 2014 or early 2015
in a few areas we were able to examine.
Our review occurred too soon to fully assess some aspects of Minnesotas health
insurance exchange. For example, many of MNsures initial enrollment targets
were for 2016. We examined enrollments through 2014 and available evidence
about changes in Minnesotas number of uninsured individuals, but additional
researchover longer periodswill be necessary. Likewise, the achievement of
some statutory goals for MNsure, such as improvements in health insurance
affordability, quality, and competition, are not apparent from a single years
experience.
We focused primarily on MNsures efforts to enroll individuals and families
rather than businesses. MNsure has provided opportunities for small businesses
(with 2 to 50 full-time-equivalent employees) to sign up for insurance, but
relatively few did so during MNsures first year.
Our report briefly describes the ongoing exchange-related activities of the
departments of Commerce and Health, but we did not evaluate these activities.
For example, the Department of Commerce approves rates for insurers that offer
products through MNsure.
We have included a glossary of key terms at the end of this report (Appendix A),
as well as a timeline of selected events in MNsures history (Appendix B). These
may be of particular interest to general readers.
Chapter 1: Background
A
ccording to a recent analysis, 366,000 Minnesotansor 7 percent of

Minnesotas total populationwere uninsured in 2013.1 The 2010
Affordable Care Act, among other goals, aimed to reduce the number of
uninsured people nationally. The act authorized states to establish health
insurance exchanges to enroll people in insurance and improve their ability to
compare insurance products. This chapter briefly reviews the portions of the
federal law relevant to health insurance exchanges, and it provides an overview
of Minnesotas health insurance exchange.
KEY FINDINGS IN THIS CHAPTER
Minnesotas state health insurance exchange (MNsure) has some

important differences from other executive branch state agencies.
In contrast to practices in many other states, Minnesotas exchange

provides a single website at which individuals eligibility for tax
credits and public health care programs can be determined.
FEDERAL HEALTH CARE REFORM

In March 2010, President Obama signed into law a set of federal health care
reforms often called the Affordable Care Act (ACA).2 The law was intended to
make health insurance accessible to more Americans, while decreasing health
cares cost and improving its quality.
Starting in 2014, the law requires most people to obtain health insurance or pay a
financial penalty. To comply with the requirement, individuals may obtain
insurance through an employer or a public program (such as Medicaid or
Medicare), or purchase individual or family insurance. The box on the next page
shows the types of insurance Minnesotans had as of May 2014; most had
insurance through employer-sponsored or other types of group insurance.
Federal law exempts some people from the insurance mandate, such as those
with very low incomes and those for whom the cost of insurance premiums
would exceed 8 percent of household income.
The ACA also authorized an expansion of the federal Medicaid program, which
insures low-income people at little or no expense to them. Previously, the
1
Henry J. Kaiser Family Foundation, Health Insurance Coverage of the Total Population,
http://kff.org/other/state-indicator/total-population/#, accessed November 22, 2014. Minnesotas
percentage of uninsured people was the third lowest among states.
2
Patient Protection and Affordable Care Act, Public Law 111-148, 124 Stat. 119 (March 23, 2010),
as amended by the Health Care and Education Reconciliation Act of 2010, Public Law 111-152,
124 Stat. 1029 (March 30, 2010).
program covered specific categories

of low-income people, but there were
gaps in the coverage. For example,
Medicaid covered many low-income
adults with dependent children; it
often did not cover low-income adults
without dependent children. Under
the ACA, states have the option to
expand their Medicaid programs to
cover nearly all nonelderly adults at or
below 138 percent of the federal
poverty level.3 As of the end of 2014,
27 states (including Minnesota) and
the District of Columbia had decided
to implement the expansion.4 The
federal government pays states for
100 percent of their Medicaid
expansion costs from 2014 through
2016. In 2017, the federal
contribution begins to decline
gradually, stabilizing at 90 percent of
the expansion costs by 2020.
Minnesotans Sources of Health

Insurance, May 2014
Type
Group insurance
Medicare
Other public programs
Individually purchased
(Includes 5.0% who
purchased directly from
insurers and 0.8% who
purchased through
Minnesotas health
insurance exchange)
Other
TOTAL:
No health insurance
Percentage
Covered
56%
16
17
6
<1
95%
5%
SOURCE: Julie Sonier, Elizabeth Lukanen, and

Lynn Blewett, Early Impacts of the Affordable
Care Act on Health Insurance Coverage in
Minnesota (Minneapolis: State Health Access
Data Assistance Center, University of Minnesota,
June 2014).
The U.S. Congressional Budget Office estimated that the ACA would reduce the
number of nonelderly people without health insuranceby 12 million in 2014,
19 million in 2015, and 25 million in 2016.5 Most of the increase in insurance
coverage was projected to occur as a result of enrollments in insurance exchanges
established by the ACA, discussed below.6
A health insurance exchange is a marketplace established under the Affordable Care Act to
facilitate the comparison and purchase of health insurance.
According to the federal government, exchanges (sometimes called
marketplaces) are:
As an example of 138 percent of the federal poverty level, this threshold for a family of four in
2014 was $32,913. The ACA said the federal government could withhold all of a states Medicaid
funds if the state did not expand the Medicaid program; however, the Supreme Court said this was
unconstitutionally coercive. See National Federation of Independent Business et al. vs. Sebelius,
Secretary of Health and Human Services, 567 U.S. ____ (2012), 132 S.Ct 2566.
In January 2011, Governor Dayton issued an executive order that, starting by March 1, 2011,
expanded Medicaid to adults without dependent children who had incomes up to 75 percent of
federal poverty guidelines. In 2013, the Legislature further expanded Minnesotas Medicaid
program to cover additional populations.
U.S. Congressional Budget Office, Updated Estimates of the Effects of the Insurance Coverage
Provisions of the Affordable Care Act (Washington, DC, April 2014), 3-4.
6
For 2016, the Congressional Budget Office estimated that 24 million additional people will be
covered by insurance obtained through insurance exchanges, and an additional 12 million through
Medicaid and the Childrens Health Insurance Program. These increases will be partially offset by
reductions in employment-based coverage and other coverage.
BACKGROUND
A resource where individuals, families, and small businesses

can: learn about their health coverage options; compare health
insurance plans based on costs, benefits, and other important
features; choose a plan; and enroll in coverage. The [exchange]
also provides information on programs that help people with low
to moderate income and resources pay for coverage. This
includes ways to save on the monthly premiums and out-ofpocket costs of coverage available through the [exchange], and
information about other programs, including Medicaid and the
Childrens Health Insurance Program (CHIP). The [exchange]
encourages competition among private health plans, and is
accessible through websites, call centers, and in-person
assistance.7
Thirteen states (including Minnesota) and the District of Columbia are enrolling individuals
in health insurance coverage for 2015 through their own exchanges, while most other states
are relying entirely on the federal exchange.
Each state had the option to (1) develop its own exchange, (2) enter into a statefederal partnership exchange, or (3) rely on the federal governments exchange.
Exhibit 1.1 shows the types of exchanges that are currently used by each state to
enroll individuals. Two states (Oregon and Nevada) that operated their own
Exhibit 1.1: Types of State Health Insurance

Exchanges for Individuals, 2015
Type of Exchange
States
State-Based Exchange (14)
California, Colorado, Connecticut, District of Columbia,

Hawaii, Idaho, Kentucky, Maryland, Massachusetts,
Minnesota, New York, Rhode Island, Vermont,
Washington
State-Federal Partnership (5)
Arkansas, Delaware, Illinois, Iowa, New Hampshire
Federal Exchange (32)
Alabama, Alaska, Arizona, Florida, Georgia, Indiana,

Kansas, Louisiana, Maine, Michigan, Mississippi,
Missouri, Montana, Nebraska, Nevada,a New Jersey, New
a
Mexico, North Carolina, North Dakota, Ohio, Oklahoma,
Oregon,a Pennsylvania, South Carolina, South Dakota,
Tennessee, Texas, Utah, Virginia, West Virginia,
Wisconsin, Wyoming
These states are considered to have state-based exchanges, but they are relying on the federal
governments online exchange for 2015 enrollments.
SOURCES: National Conference of State Legislatures, http://www.ncsl.org/Portals/1/Documents
/Health/Health_Insurance_Exchanges_State_Profiles.pdf, accessed January 4, 2015; Henry J. Kaiser
Family Foundation, State Health Insurance Marketplace Types, 2015, http://kff.org/health-reform
/state-indicator/state-health-insurance-marketplace-types/, accessed January 2, 2015.
Definition of Health Insurance Marketplace, https://www.healthcare.gov/glossary/health

-insurance-marketplace-glossary, accessed December 29, 2014.
exchanges in 2014 are now enrolling individuals using the federal exchange, and
one state (Idaho) that had a federal-state partnership in 2014 is using a statebased exchange in 2015.
MINNESOTAS HEALTH INSURANCE EXCHANGE

In a 2011 executive order, Governor Mark Dayton directed the Commissioner of
Commerce, in consultation with a legislatively created health care reform task
force, to [d]esign and develop a Minnesota health insurance exchange to ensure
access to affordable, high-quality health coverage that maximizes consumer
choice and minimizes adverse selection.8 Pursuant to that direction, the
executive branch obtained federal grants and did considerable planning for
building the exchange. In December 2012, when the federal government
authorized Minnesota to establish a state-based exchange, it required Minnesota
to develop sufficient legal authorityto operate the exchange.9 We discuss the
development process for Minnesotas exchange further in Chapter 2.
Statutory Authority
In 2013, the Legislature authorized in law the states health exchange, which was
named MNsure.10 Exhibit 1.2 shows the purposes of MNsure specified in state
law. For example, the law says MNsure should simplify the comparison,
choice, enrollment, and purchase of health insurance.11
MNsure is governed by a seven-member board: six members appointed by the
governor, and the Commissioner of Human Services (or the commissioners
designee). The board is authorized by statute to:
employ staff and delegate responsibilities to them;
establish the MNsure budget;
seek revenues to fund MNsure operations;
contract for the receipt and provision of goods and services;
enter into information-sharing agreements with entities while ensuring

adequate protection of the information shared; and
exercise all powers reasonably necessary to implement and administer

the requirements of the ACA and the state MNsure law.12
Governor Mark Dayton, Executive Order 11-30, Establishing a Vision for Health Care Reform in
Minnesota, October 31, 2011. The term adverse selection refers to the greater likelihood of
unhealthy individuals than healthier individuals to purchase health insurance.
Kathleen Sebelius, Secretary, U.S. Department of Health and Human Services, letter to Governor
Mark Dayton, December 20, 2012.
10
Laws of Minnesota 2013, chapter 9.
11
Minnesota Statutes 2014, 62V.03, subd. 1.
12
Minnesota Statutes 2014, 62V.05, subd. 1(b)(6).
BACKGROUND
Exhibit 1.2: Statutory Purposes of MNsure

(1) Promote informed consumer choice, innovation, competition, quality, value, market
participation, affordability, suitable and meaningful choices, health improvement, care
management, reduction of health disparities, and portability of health plans;
(2) Facilitate and simplify the comparison, choice, enrollment, and purchase of health
plans for individuals purchasing in the individual market through MNsure and for
employees and employers purchasing in the small group market through MNsure;
(3) Assist small employers with access to small business health insurance tax credits and
to assist individuals with access to public health care programs, premium assistance
tax credits and cost-sharing reductions, and certificates of exemption from individual
responsibility requirements;
(4) Facilitate the integration and transition of individuals between public health care
programs and health plans in the individual or group market and develop processes
that, to the maximum extent possible, provide for continuous coverage; and
(5) Establish and modify as necessary a name and brand for MNsure based on market
studies that show maximum effectiveness in attracting the uninsured and motivating
them to take action.
SOURCE: Minnesota Statutes 2014, 62V.03, subd. 1.
MNsure shares some similarities with executive branch state agencies, but it also has
important differences.
MNsure is subject to a number of state requirements that apply to other state
agencies. MNsure must comply with the Government Data Practices Act, and it
must employ individuals in accordance with the states personnel management
law and the Public Employee Labor Relations Act. MNsure Board members are
subject to state laws regarding conflicts of interest, and board members and
employees are subject to laws that govern receipt of gifts. Also, the MNsure
Board is subject to state laws requiring that it conduct meetings that are open to
the public, with some exceptions.13
Exhibit 1.3 shows key differences between laws that apply to MNsure and those
that apply to other state agencies. The governance of MNsure by a board is one
important difference. In contrast to MNsure, nearly all state agencies that have
far-reaching policy and administrative responsibilities are headed by a governorappointed commissioner.14 State law has established 17 departments of the
13
MNsure is subject to the state open meeting law, but the following types of MNsure meetings
may be closed: (1) meetings regarding compensation negotiations with the director or managerial
staff; (2) meetings regarding contract negotiation strategy; and (3) meetings regarding not-public
data or trade secret information.
14
The Legislature has also created agencies that have more limited functions. For example, some
state agencies license and regulate individuals in certain occupations, and these agencies are
generally governed by boards. In some instances, the Legislature has created corporations (such as
Enterprise Minnesota, Inc., and Minnesota Business Finance, Inc.) that follow laws established for
nonprofit corporations or business corporations.
10
state, each headed by a commissioner.15 Two agencies that are not

departments of the statethe Pollution Control Agency and the Housing
Finance Agencyare the only cabinet-level agencies serving statewide missions
that are governed by boards. Those agencies, unlike MNsure, also have
governor-appointed commissioners.
Exhibit 1.3: Key Differences between MNsure and

Other State Agencies
Governance: Most state departments do not have governing boards, although

many state regulatory entities (such as occupational licensing boards) have them.
Also, only one other state agency (the Board of Campaign Finance and Public
Disclosure) has its board members approved by both the Minnesota House and
Senate. In addition, until 2016, the compensation of MNsure board members is more
generous than compensation for members of most other state-created boards.
Rules: MNsure is authorized to indefinitely use an expedited process for adopting

rules; usually statutes grant this authority to agencies only in selected cases.
Exemptions from certain laws: MNsure is exempt from many laws pertaining to
state procurement, Office of MN.IT Services oversight of IT projects, and various
Department of Administration requirements.
Conflict of interest: Certain elements of MNsures conflict of interest provisions are

more restrictive than those that apply to most state boards. For example, a spouse
of a board member may not be a health carrier executive. Also, board members are
prohibited not only from being employed by a health insurer or provider but also from
representing them.
Employee classification: State law specifies that all of MNsures managerial staff
are in the unclassified service.
SOURCES: Minnesota Statutes 2014, 62V, and Minnesota House of Representatives research staff.
The Legislature exempted MNsure from some laws that apply to most other
agencies. As discussed further in Chapter 2, MNsures information technology
projects are not subject to the full range of oversight by the states information
technology agency (the Office of MN.IT Services) that applies to other agencies
projects. Also, MNsure is authorized to follow an expedited rulemaking process
and is exempt from many of the states purchasing laws.16
Governing Board
We assessed how MNsures governing board compares with exchange boards in
other states. Specifically, we reviewed statutes and documents from the nine
states that, starting in 2013, operated exchanges governed by a board.17
15
The departments of the state are established in Minnesota Statutes 2014, 15.01.
16
Agencies may use an expedited rulemaking process only when authorized to do so by state law.
This process is shorter than the regular rulemaking process. For example, the expedited process
unlike the regular processprovides no opportunity for a public hearing on the proposed rule,
unless such a hearing is specifically required by law.
17
Besides Minnesota, these states were California, Colorado, Connecticut, Hawaii, Maryland,
Massachusetts, Oregon, and Washington.
BACKGROUND
11
Minnesotas exchange boardwith seven voting memberswas one of the

smaller boards. Only Californias (with 5 voting members) was smaller;
Hawaiis board (with 15 voting members) was the largest.
All members of the MNsure Board were appointed by the Governor.
In Minnesota, the governor (subject to legislative confirmation) controls the
make-up of the exchange board. Governor Dayton made the initial six board
appointments in April 2013; in accordance with the law, the Department of Human
Services Commissioner served as the seventh member. The six appointed
members have staggered terms, which limits the number of members a governor
can appoint in the future at a given time.18 During the 2013-2014 open enrollment
period, four other states (Hawaii, Maryland, Oregon, and Washington) had
exchange boards in which the governor had authority to appoint each member
whose service on that states board was not specifically designated by statute.
In three states (California, Colorado, and Connecticut), statutes require the
appointment of a portion of the exchange boards members by legislative leaders
or legislative committees. For example, of Californias five exchange board
members, one must be appointed by the Senate Committee on Rules, and one
must be appointed by the Assembly speaker.
Three states (Colorado, Connecticut, and Washington) had statutory provisions
that authorized members of both political parties to play a part in governing
board appointments; Minnesota did not. For example, all of Washingtons nine
board members were appointed by the governor, but that states statutes required
the governor to select two members nominated by the Senate majority caucus,
two nominated by the Senate minority caucus, two nominated by the House
majority caucus, and two nominated by the House minority caucus.
Minnesota was the only state that required both the House and Senate to approve
exchange board members. This requirement for confirmation by both houses is
unusual among Minnesota agencies, too. The Board of Campaign Finance and
Public Disclosure is the only other Minnesota entity that requires separate
confirmation by both the House and Senate.19
Unlike some states, Minnesota prohibits people employed by or representing the health
insurance industry from being appointed to the exchange board.
An appointed member of the MNsure Board may notduring the board term or
within one year prior to appointmentbe an employee of a health care insurer,
18
The terms of the initial appointees to the MNsure Board differ, which allow for staggered
appointments subsequently. For future appointments, board members will serve four-year terms,
which is the most common length of terms in states with exchange boards.
19
Minnesota Statutes 2014, 10A.02, subd. 1. Members of the Board of Campaign Finance and
Public Disclosure must be confirmed by 60 percent of House and Senate members, while MNsure
Board members must be confirmed by a simple majority. For other Minnesota agencies with
officials subject to confirmation, the Senate is typically responsible for these approvals.
12
an insurance agency, or a health care provider.20 Of the nine states that had
exchange boards starting in 2013, two others (California and Connecticut)
prohibited employees or representatives of these professions from appointment to
their exchange boards. Three other states prohibited certain industry
professionals from being appointed to their exchange boards, but these states
provisions were less restrictive than Minnesotas.21
The other three states with governing boards did not ban representatives of any
professions from serving on their boards. Rather, they limited the number of
industry-affiliated people who could serve on the exchange board (Oregon and
Colorado) or simply required them to abide by conflict of interest requirements
(Hawaii).
Minnesota law requires appointed members of the MNsure Board to represent
certain groups or to have certain types of expertise, as shown in Exhibit 1.4.
Exhibit 1.4: Statutory Requirements for Composition

of the MNsure Board
One member representing the interests of individual consumers eligible for individual
market coverage;
One member representing individual consumers eligible for public health care
program coverage;
One member representing small employers;
One member with demonstrated expertise, leadership, and innovation in health
administration, health care finance, health plan purchasing, and health care delivery
systems;
One member with demonstrated expertise, leadership, and innovation in public
health, health disparities, public health care programs, and the uninsured;
One member with demonstrated expertise, leadership, and innovation in health policy
issues related to the small group and individual markets; and
The Commissioner of the Department of Human Services or a designee.
SOURCE: Minnesota Statutes 2014, 62V.04, subd. 2.
Organization
Minnesotas health insurance exchange has had three different locations within
the executive branch of state government during its short history. When planning
for the exchange started in January 2011, the exchange was part of the Minnesota
Department of Commerce. In September 2012, the Governor announced that he
20
Minnesota Statutes 2014, 62V.04, subd. 4. The law also prohibits lobbyists, navigators (defined
later in this chapter), and spouses of health insurer executives from serving on the board.
21
Maryland did not prohibit health insurance providers from being appointed, and Massachusetts
did not prohibit providers or insurance agents from being appointed. Washington prohibited
appointment of anyone whose board membership could benefit their financial interests or those of
the entities they represent.
BACKGROUND
13
would be moving the exchange to the Department of Management and Budget.

According to the Governor,
This move will address the concern that certain core functions of the
Exchangeproviding health insurance options to individuals and
small employersare potentially in conflict with the Department of
Commerces role in regulating insurance companies and the sale of
health insurance products. Additionally, [the Department of
Management and Budget] has the capacity and experience with
financial oversight and human resources required for the next phase
of designing and developing the Exchange.22
Exchange employees transferred to the Department of Management and Budget
in December 2012, following the execution of an interagency agreement.
When the Governor signed legislation in March 2013 that created MNsure,
MNsure became a stand-alone state agency. MNsures governing board did not
assume its statutory authority and responsibilities until August 2013, when it had
approved a set of policies and bylaws for the organization. Until that time, the
boards duties were performed by the Commissioner of the Department of
Management and Budget.
Since exchange planning began in January 2011, MNsure has had two
administrative directors. April Todd-Malmlov served as executive director until
her resignation in December 2013. She was replaced by Scott Leitz, whose title
was changed to chief executive officer.
Types of Coverage Offered Through MNsure

MNsures online enrollment system has three primary functions: (1) determining
consumers eligibility for public health insurance programs, such as
Medicaid; (2) serving as a marketplace that facilitates the comparison and
purchase of certain commercial health insurance products; and (3) determining
consumers eligibility for federal tax credits and cost-sharing reductions to lower
the premiums and out-of-pocket expenses associated with commercial plans.23
The MNsure enrollment system makes eligibility determinations for both commercial
insurance and public health care programs, which distinguishes it from the exchanges used
in a majority of states.
For many insurance seekers, Minnesotas exchange provides a one-stop shopping
experience. Applicants who do not qualify for a public insurance program can learn
through the exchange whether they qualify for federal tax credits when they purchase
commercial insurance. Also, if a household contains some members who qualify for
a public program and others who do not, MNsures enrollment system can assess
eligibility for public programs or public subsidies for all household members.
22
Governor Mark Dayton, letter to Senator David Senjem, Senator Thomas Bakk, Representative
Kurt Zellers, and Representative Paul Thissen, September 18, 2012.
23
MNsure also sells health insurance for employees of small businesses and stand-alone dental
insurance. These types of insurance are not discussed in this report.
14
Nationally, all of the health insurance exchanges implemented under the ACA are
expected to determine whether an individual purchasing commercial insurance is
eligible for federal tax credits and cost-sharing reductions. However, most states
have separate processesinvolving agencies outside of the exchangethat
determine whether individuals qualify for public programs such as Medicaid.
Specifically, in 27 states, the exchange makes only a preliminary assessment of
whether an individual qualifies for a public program; the individual is then
transferred to a human services agency for a final determination. In contrast,
Minnesotas enrollment system determines eligibility for public programs, in
addition to determining eligibility for federal tax credits and subsidies.
Commercial Insurance Through Qualified Health Plans

Insurers may apply to MNsure to have their commercial insurance products sold
through the state exchange. Federal law requires that products sold through
exchanges be certified as qualified health plans.24
Among the certification requirements,
qualified health plans must offer a
comprehensive set of services. These are
commonly referred to as essential health
benefits.25 Plans must offer services in at
least the ten benefit categories shown at the
right. All qualified health plans must offer
benefits similar to those offered by a
benchmark plan. In Minnesota, the small
group insurance plan (that is, an insurance
product sold to businesses with 2 to 50 fulltime employees) with the largest enrollment
statewide provides the benchmark for
evaluating the benefits in other plans.
Types of Essential Health

Benefits
Ambulatory patient services

Emergency services
Hospitalization
Maternity and newborn care
Mental health and substance
abuse services
Prescription drugs
Rehabilitative and habilitative
services and devices
Laboratory services
Preventive and wellness
services, and chronic
disease management
Pediatric services
Qualified health plans are classified by metal levels: bronze, silver, gold, or
platinum. Assigning metal levels to qualified health plans is intended to help
consumers compare plans offering similar coverage. These metal levels reflect
the percentage of total essential health benefit costs the plan is expected to cover,
based on actuarial assessments. For example, a bronze plan would cover 60
percent of the expected value of the essential health benefits; at the other
extreme, a platinum plan would cover 90 percent.26 Chapter 4 discusses the
percentage of people who bought various types of qualified health plans through
MNsure.
24
Also, with certain exceptions, health plans that are sold in the small group and individual markets
outside of MNsure must meet market rules specified in Minnesota Statutes 2014, chapter 62K.
25
Certain insurance products that already existed at the time the ACA was signed into law are not
required to cover the essential health benefits mandated by the ACA.
26
Health insurers may also sell plans that offer catastrophic coverage, which may have an
actuarial value below 60 percent. Catastrophic plans meet the same requirements as other qualified
health plans, but they do not cover any benefits other than three primary care visits per year until a
high deductible is reached. Catastrophic plans are only available to individuals who are (1) under
the age of 30 or (2) deemed by MNsure to have a hardship exemption from the ACAs individual
mandate to purchase health insurance because the available coverage is unaffordable.
BACKGROUND
15
Insurers who want to sell products through MNsure must offer qualified health
plans in all of the metal levels and insurance markets in which they offer
coverage outside of MNsure.27 Also, premiums must be the same for the same
plans sold inside and outside of MNsure.
Minnesota law required the MNsure Board to allow all insurance products
meeting federal certification requirements to be offered through MNsure in
2014.28 In subsequent years, the board may limit which products are sold
through MNsure.29 During each of MNsures first two years of operations, five
insurers offered products through the exchange to individuals.30 There was a
small increase in the total number of qualified health plans offered to individuals
and families through MNsure in the second year of open enrollment (84)
compared with the first (78).
People purchasing commercial insurance through an ACA exchange (such as MNsure) may
qualify for tax credits or other subsidies that are not available to people buying insurance
outside of the exchange.
The Affordable Care Act authorized the federal government to provide subsidies
to make commercial health insurance purchased through exchanges more
affordable. As shown in Exhibit 1.5, households with annual incomes between
201 percent and 400 percent of the federal poverty level may be eligible for an
advanced premium tax credit.31 In such cases, the U.S. Treasury Department
sends monthly payments to the individuals health plan to cover all or part of the
persons monthly insurance premium. Alternatively, an individual may receive
the tax credit when filing federal income taxes. A person who buys the same
insurance product outside of an exchange is not eligible to receive a tax credit.
People with incomes between 201 and 250 percent of the federal poverty level
may also qualify for cost-sharing reductions authorized by the Affordable Care
Act. With a cost-sharing reduction, a persons health insurance product has
27
As used here, insurance markets refers to the individual market and small group market.
The individual market is insurance not associated with a group health plan. Employers with 50 or
fewer employees may obtain employee coverage through small group market health plans.
28
Minnesota Statutes 2014, 62V.05, subd. 5(f).
29
Minnesota Statutes 2014, 62V.05, subd. 5(e). The MNsure Board decided to allow all insurance
products meeting federal certification requirements to be offered through MNsure in 2015. State
law requires the board to adopt policies and procedures that will govern its selection of products to
be offered through the exchange.
30
The most noteworthy change from the first to second year was that the insurer (PreferredOne)
that had the most enrollees in qualified health plans during the first years open enrollment chose
not to participate in MNsure during the second year. Four insurers that offered products through
MNsure in 2014 (Blue Cross and Blue Shield of Minnesota, HealthPartners, Medica, and UCare)
continued to do so in 2015, and one additional insurer (BluePlus) also did so.
31
Besides income requirements, there are other criteria to qualify for an advanced premium tax
credit through MNsure. Notably, one must not be eligible for public health care programs; must
not have access to employer-sponsored insurance that is affordable and covers at least 60 percent of
the employees medical costs within the benefit categories covered by the employer; must file
income taxes jointly, if married; must be a U.S. citizen or qualified noncitizen; and must be a
Minnesota resident.
16
Exhibit 1.5: Nonelderly Adults Income Eligibility for

Minnesotas Public Health Care Programs and
Federal Subsidies
0-138%
Medical Assistance
139200%
MinnesotaCare
Advanced Premium Tax Credit
201-400%
201250%
Cost-Sharing Reductions
50
100
150
200
250
300
350
400
Income as a Percentage of Federal Poverty Guidelines

NOTES: The income criteria are based on household modified adjusted gross income. For
coverage year 2015, 138 percent of the federal poverty level for a family of four is $32,913;
200 percent is $47,700; and 400 percent is $95,400.
SOURCE: Minnesota House Research Department, Eligibility for Subsidized Health Coverage in
Minnesota: 2013 Session Changes (St. Paul, October 2013).
lower out-of-pocket health care expensescopayments, deductibles, and

coinsurancethan it would otherwise have.32
Individuals enroll in qualified health plans during open enrollment periods.
The first open enrollment period for Minnesotas health insurance exchange ran
from October 1, 2013, through March 31, 2014. Enrollments during this period
provided coverage during calendar year 2014. For calendar year 2015, the open
enrollment period lasted from November 15, 2014, through February 15, 2015.
Outside of open enrollment periods, individuals may enroll in qualified health
plans through an exchange only if they have qualifying life events, such as
marriage, divorce, birth or adoption of a child, or a job loss, or they are members
of a federally recognized American Indian tribe.
32
Cost-sharing reductions are available to people who purchase a silver planthat is, one that
would cover 70 percent of the expected value of the essential health benefits. The subsidy
effectively increases the silver plans expected coverage of health care costs from 70 to 73 percent.
Also, people who qualify for cost-sharing reductions have lower out-of-pocket maximum costs than
other people.
BACKGROUND
17
Public Health Care Programs

The Minnesota Department of Human Services (DHS) administers three public
health care programs. Medical Assistance (MA) is Minnesotas version of the
federal Medicaid program. In any given month, MA provides health insurance to
about one million Minnesotans. The Childrens Health Insurance Program
(CHIP) is a federal program that provides coverage to uninsured children under
age two and uninsured pregnant women. It is by far the smallest of Minnesotas
public health care programs.33 The Minnesota Legislature created
MinnesotaCare in 1992 to offer low-cost coverage to people not eligible for
MA; as discussed below, it is now Minnesotas version of a Basic Health
Program, as defined by the ACA.34 As of October 2014, MinnesotaCare had
about 76,000 enrollees. These public health insurance programs are supported by
a combination of federal and state funding.
Since October 2013, most Minnesotans seeking to enroll in state public health care
programs have been required to apply through MNsure.
Federal and state legislation changed the way eligibility is calculated for
Minnesotas public health care programs, starting January 1, 2014. Specifically,
the Affordable Care Act changed the method used to determine whether certain
individuals are eligible for Medicaid and CHIP, and the Minnesota Legislature
enacted similar changes for MinnesotaCare.35 For people who qualify for these
programs based on income, the ACA and Minnesota law require the use of an
income measure (modified adjusted gross income) that is also used for federal
tax purposes.36 Earlier, Exhibit 1.5 showed the income guidelines for persons in
these programs who are required to enroll through MNsure. As of January 2014,
for individuals who are parents or adults under age 65 without Medicare, assets
are no longer considered when determining eligibility for Minnesotas public
health care programs.
The Affordable Care Act gave states the option of implementing a Basic Health
Program, starting in 2015.37 Under this option, states receive federal assistance
to provide subsidized coverage for individuals with modified adjusted gross
incomes between 139 and 200 percent of the federal poverty level.38 A Basic
33
Only 213 people enrolled in CHIP through MNsure during MNsures first seven months.
34
Laws of Minnesota 1992, chapter 549, art. 4. MinnesotaCare was originally called Health Right.
35
Minnesota Department of Human Services, Bulletin: Legislative Changes to Medical Assistance

and MinnesotaCare Effective January 1, 2014 (St. Paul, December 11, 2013).
36
Modified adjusted gross income is a households adjusted gross income (AGI) plus any
tax-exempt Social Security, interest, or foreign income. For some people covered by Medical
Assistancesuch as the elderly and people with disabilitieseligibility is not based on modified
gross income, so they are not required to enroll through MNsure.
37
The 2013 Legislature made changes in MinnesotaCare eligibility, covered services, and service
delivery that would allow it to qualify as a Basic Health Program. Many of these changes took
effect in January 2014.
38
For states implementing a Basic Health Program, the federal government pays 95 percent of the
premium tax credit and cost-sharing reductions for which the individuals would have qualified if
they had enrolled through a health insurance exchange in a qualified health plan.
18
Health Program allows individuals to keep health insurance if their incomes rise
above the maximum for Medicaid eligibility (138 percent of the federal poverty
level). In Minnesota, the MinnesotaCare program functions as the states Basic
Health Program. Most MinnesotaCare recipients pay a sliding-scale premium,
ranging from $4 to $50 monthly. As of early 2015, Minnesota was the only state
with a Basic Health Program.39
DHS is accountable for ensuring that individuals eligibility for each of the
states three public health care programs is determined correctly.40 DHS worked
closely with other state officials on the development of the MNsure information
technology system that enrolls and determines the eligibility of applicants for
public programs. DHSs goal has been to use the MNsure system to validate
applicant-provided information, including income, Social Security number,
citizenship or immigration status, and Medicare enrollment.41
MNsures online system started processing new applications for the states public
health care programs on October 1, 2013. Individuals who were already enrolled
in these programs on that date were supposed to have their eligibility reviewed
annually through MNsure, but many of these reviews were postponed (see
discussion in Chapter 2).
Roles of Other State Agencies

Besides MNsure and DHS, two state agenciesthe Minnesota departments of
Commerce and Healthplay important roles in Minnesotas health insurance
exchange.
Insurance companies that wish to sell products in Minnesota (through or outside
MNsure) must annually submit proposed premium rates to the Department of
Commerce. If the department concludes that a proposed rate is justified, it
approves the rate. In the case of proposed rates for health maintenance
organizations, the Department of Commerce makes recommendations on the
rates to the Department of Health, which has final authority to approve these
rates.42
Minnesota law also establishes market rules that apply to individual and small
group health insurance products that are offered, sold, issued, or renewed in
39
Minnesota law clarifies that a person who meets MinnesotaCare eligibility requirements is not
eligible to enroll in a qualified health plan offered through MNsure. See Minnesota Statutes 2014,
256L.04, subd. 1c.
40
DHS has not delegated responsibility for public program eligibility determinations to MNsure.
However, DHS uses a shared information technology systemthe MNsure enrollment systemto
determine eligibility for public programs. For people enrolling in commercial insurance, this
system also determines eligibility for tax credits and cost-sharing reductions.
41
For the public programs that enroll individuals through MNsure, Minnesota Statutes 2014,
62V.05, subd. 7(a)(2), requires MNsure to have an agreement with DHS that outlines services and
cost responsibilities for eligibility and enrollment activities.
42
A health maintenance organization is a type of insurer defined in Minnesota Statutes 2014,

62D.02, subd. 4. It provides or arranges for health services on the basis of a fixed, prepaid sum,
without regard to the amount of services provided to a given enrollee.
BACKGROUND
19
Minnesota through or outside of MNsure.43 The departments of Commerce and

Health share responsibility for enforcing these requirements. For example, the
Department of Commerce requires insurers to justify the metal levels (actuarial
values) of their products. It also enforces laws governing the marketing practices
of insurers. The Department of Health, among other duties, reviews the
adequacy of insurers networks of health care providers, and enforces statutory
requirements for insurers to have quality assurance and improvement programs.
Budget
Exhibit 1.6 shows MNsures adopted budget for calendar year 2014 and its draft
budgets for fiscal years 2015 through 2017. MNsure recently changed from a
calendar year budget to a state fiscal year budget (from July to June). The exhibit
reflects the budget revisions MNsure proposed after it received an additional
$34 million in federal revenues in December 2014. Below, we further discuss
MNsures revenues and expenditures.
Revenues
To plan and build health insurance exchanges, states have relied largely on
federal funds. Initially, states could apply for $1 million planning grants;
49 states (including Minnesota) and the District of Columbia received such a
grant. These grants helped states research and plan how their exchanges would
operate. Later, states could apply for larger establishment grants to continue
planning for their exchanges; 37 states (including Minnesota) and the District of
Columbia received such grants. Most states that received establishment grants
eventually decided not to operate their own exchanges. Other states advanced
further in their planning and established governance structures for state-based
exchanges; they were eligible to receive additional (Level 2) establishment
grants. Fourteen states (including Minnesota) and the District of Columbia
received Level 2 establishment grants.
Minnesota has received $189 million in federal grants to plan and establish its health
insurance exchange, but these grants will be a diminishing revenue source for MNsure.
Chapter 2 provides details on the timing and amount of individual federal grants
Minnesota has received. The total amount Minnesota received (through October
2014) was the fourth lowest among the 14 states (plus the District of Columbia)
that opened state-based exchanges in 2013.44 Also, Minnesotas total federal
grants for its exchange per capita ($28.60) was among the lowest for states that
built their own exchanges.45
43
Minnesota Statutes 2014, chapter 62K.
44
As of October 2014, total federal grants awarded for state-based exchanges ranged from
$90.8 million (Nevada) to $1.07 billion (California); the median among states that opened
exchanges in 2013 was $178.9 million.
45
As of October 2014, total federal grants per capita ranged from $26.02 (New York) to $275.51
(Vermont); the median among states that opened exchanges in 2013 was $38.16.
20
Exhibit 1.6: MNsure 2014 Budget and Draft Budgets

for Fiscal Years 2015 to 2017
Revenues (in thousands)
Calendar
Year 2014
July 2014June 2015
July 2015June 2016
July 2016June 2017
Federal Grants
Premium Withhold
Department of Human Services
Carry Forward
Total
$ 94,162
2,194
36,535
0
$132,891
$ 60,668
5,314
52,507
537
$119,026
$ 28,278
10,647
56,103
405
$95,434
$
0
16,003
28,713
400
$45,116
Expenditures (in thousands)
Calendar
Year 2014
July 2014June 2015
July 2015June 2016
July 2016June 2017
$ 10,744
$ 13,337
$ 8,437
$ 8,437
3,794
8,510
29,279
68,692
$121,019
1,805
5,714
24,608
73,157
$118,621
650
3,887
20,585
61,475
$95,034
650
3,887
16,745
14,134
$43,853
Administrative
Regulatory (Departments of
Health and Commerce)
Communications
Customer Servicea
Technology System
Total
NOTES: The MNsure Board adopted a fiscal year 2015 budget in October 2014, but MNsure
received $34 million in additional federal grant money in December 2014. In January 2015, the board
considered the revised fiscal year 2015 budget and the fiscal year 2016 and 2017 budgets shown
here. As of early February 2015, the board had not taken action on the proposed financial plan
covering these three years.
a
Customer service includes the MNsure contact center, the assister programs, eligibility and
enrollment services, the small business options program, and plan management and reporting.
SOURCE: Office of the Legislative Auditor, summary of information from MNsure, October 2014 and
January 2015.
The ACA required states running their own exchanges to be financially selfsustaining by 2015. However, the federal government has authorized some states
(including Minnesota) to continue using unspent federal grants beyond 2014. As
shown in Exhibit 1.6, MNsure estimated that revenues from federal grants would
decline from $94.1 million in 2014 to $28.3 million in fiscal year 2016, with no
federal grant revenues in fiscal year 2017.
MNsure is authorized in state statute to retain or collect a portion of the
premiums for products sold through MNsure.46 Initially, MNsure was authorized
by statute to collect up to 1.5 percent of premiums for individual and small group
market health plans and dental plans. For calendar years 2015 and beyond, the
statutes authorize MNsure to collect up to 3.5 percent of health plan premiums.
Exhibit 1.6 shows that revenues from the premiums sold through MNsure were
projected to increase from $2.2 million in 2014 to $10.6 million in fiscal year
2016 and $16.0 million in fiscal year 2017. This increase reflects that (1) a larger
portion of the premium will be withheld in the years following 2014 and
(2) MNsure projects growth in the number of enrollees whose premium revenues
will be withheld.
46
Minnesota Statutes 2014, 62V.05, subd. 2(a) through (c).
BACKGROUND
21
Starting in fiscal year 2016, the largest revenue source for MNsure is expected to
be DHS payments to MNsure. These payments help to pay for MNsures
assistance in enrolling individuals in public programs. As shown in Exhibit 1.6,
DHS payments are projected to account for over half of MNsures revenues in
fiscal years 2016 and 2017.
Spending
In the period leading up to the launch of MNsures online enrollment system,
MNsure spent most of its money on technology. For example, in fiscal year
2013, information technology contracts, software support, and software licenses
accounted for 75 percent of MNsures spending.47
MNsure anticipates spending reductions in all areas of its budget through fiscal year 2017.
As shown earlier in Exhibit 1.6, technology-related spending is expected to
remain a significant share of MNsures total spending in the future, but the dollar
amount of these expenditures is declining. MNsure budgeted $68.7 million for
its technology system in 2014; it anticipates technology spending of
$14.1 million in fiscal year 2017. Likewise, as shown in Exhibit 1.6, MNsure is
in the midst of declining spending for customer service, communications,
administrative services, and regulatory services (provided by the departments of
Health and Commerce).
Consumer Assisters
As required by federal rules, MNsure uses various consumer assisters to help
individuals enroll through MNsure. Exhibit 1.7 provides an overview of the
different types of assisters and the compensation they receive for enrolling
someone through MNsure. Chapter 6 discusses MNsures assister program in
more detail. Assisters must obtain certification from MNsure before helping
individuals enroll in health insurance through MNsure. All three types of
assisters are required to complete specialized training before they can become
certified by MNsure.
Navigators receive state payments for helping consumers enroll through MNsure,
while brokers receive compensation from insurers for enrolling consumers
through MNsure. Certified application counselors can, as a part of their jobs,
help individuals enroll in MNsure, but they do not receive compensation from the
state or insurers for doing so.
47
Office of the Legislative Auditor, Minnesota Health Insurance Exchange: MNsure, Internal
Controls and Compliance Audit, July 2011 through December 2013 (St. Paul, October 28, 2014), 6.
This analysis did not include $33 million in MNsure spending funded from DHS revenues, all of
which MNsure spent for information technology systems development.
22
Exhibit 1.7: Types of MNsure-Certified Consumer

Assisters and State Compensation per Enrollee, 2014
Type of Assister
Navigator
Certified Application
Counselor (CAC)
Insurance Brokerb
Typical
Employer
Community
organization
Hospital
Insurance
agency
Compensation per Enrollment

MNsure paid $70 per qualified health plan
enrollment. DHS paid $70 per MinnesotaCare
enrollment and $25 per Medical Assistance
enrollment.
None.
A MNsure-certified broker appointed by an insurer
to sell insurance on behalf of that insurer was paid
by the insurer for enrollments in the insurers
products sold through MNsure.c
In this table and throughout our report, we use the term navigators to refer to (1) individuals who
help enroll people in public programs and are not eligible to be paid from federal grants; and
(2) individuals (sometimes called in-person assisters) who help enroll people in qualified health
plans and are eligible to be paid from federal grants. This is consistent with how the term has been
used by MNsure.
b
In this report, we use the term broker to refer to an insurance agent or producer licensed under
Minnesota Statutes 2014, chapter 60K.
According to the Henry J. Kaiser Family Foundation, in 2010 brokers in Minnesota earned an
average of $97 per enrollment or renewal, or 3.9 percent of premiums in the individual market.
SOURCE: Office of the Legislative Auditor.
Oversight
The 2013 Legislature created the MNsure Legislative Oversight Committee to
monitor MNsure. According to statute, the committee shall review the
operations of MNsure at least annually and shall recommend necessary changes
in policy, implementation, and statutes to the board and to the legislature.48 The
committee consists of five Senators and five House members, with three
members appointed by each bodys majority leadership and two members
appointed by each bodys minority leader. Between MNsures creation in state
law in March 2013 and the end of 2014, this committee met eight times.
State law also requires the Office of the Legislative Auditor (OLA) to review
MNsure. The law says OLA shall audit the books, accounts, and affairs of
MNsure once each year or less frequently as the legislative auditors funds and
personnel permit.49 Since 2013, OLA has issued a financial audit of MNsures
expenditures of federal funds, a financial audit of MNsure eligibility
48
Minnesota Statutes 2014, 62V.11, subd. 1(b).
49
BACKGROUND
23
determinations for public health care programs, and a special review of a MNsure
data security breach, in addition to this evaluation.50
CONTEXT FOR MINNESOTAS EXCHANGE

Minnesotas health insurance exchange is a complicated information technology
project. The enrollment website must link to the Federal Data Services Hub,
which provides access to information such as citizenship status, immigration
status, and tax records. The exchange is supposed to offer multiple access
pointsor portalsfor consumers, insurers, county staff, MNsure finance
staff, and persons assisting consumers with the enrollment process, thus allowing
these groups to see various information from the online application. Because
consumers are asked to provide private information, such as income and Social
Security numbers, it is critical for the website to be secure. Furthermore, the
websites software needs to perform complex calculations to determine
individuals eligibility for tax credits and publicly funded health care programs.
To understand the development of MNsure, it is useful to consider several points
of context related to information technology projects.
Many organizations struggle to successfully complete large information technology
projects.
One international firm that researches information technologythe Standish
Grouphas examined public and private projects and concluded that large
projects are particularly risky. It found that only 10 percent of information
technology projects with more than $10 million in labor costs succeededthat is,
were delivered on time, within budget, and with the required features and
functions.51 In contrast, it reported that 76 percent of small projects (under
$1 million in labor costs) succeeded.
Similar findings were reported in a review of 5,400 information technology
projects by another firm, McKinsey & Company. This analysis showed that, on
average, large projects (defined as those with initial price tags over $15 million)
ran 45 percent over budget and 7 percent over schedule, while delivering
56 percent less value than predicted.52
50
Office of the Legislative Auditor, Minnesota Health Insurance Exchange: MNsure; Internal
http://www.auditor.leg.state.mn.us/fad/pdf/fad1421.pdf; Department of Human Services:
Oversight of MNsure Eligibility Determinations for Public Health Care Programs, October 2013
through April 2014 (St. Paul, November 12, 2014), http://www.auditor.leg.state.mn.us/fad/pdf
/fad1422.pdf; and MNsure: An Unauthorized Disclosure of Private Data; Special Review (St. Paul,
November 7, 2013), http://www.auditor.leg.state.mn.us/fad/pdf/fad1327.pdf.
51
The Standish Group, Chaos Manifesto 2013: Think Big, Act Small (Boston, 2013), 1 and 4. The
Standish Group recommended that organizations find ways to break large projects into a series of
smaller ones.
52
Michael Bloch, Sven Blumberg, and Jurgen Laartz, Delivering Large-Scale IT Projects
On Time, On Budget, and On Value, http://www.mckinsey.com/insights/business_technology
24
The Minnesota Department of Human Services (DHS) failed at a previous attempt to develop
a large automated system to determine eligibility for the states health care programs.
In 2003, DHS entered into a contract with a vendor to develop a system called
HealthMatch. This automated system was supposed to determine applicant
eligibility for nearly all of Minnesotas publicly funded health care programs.
But DHS and its vendor underestimated the projects complexity, resulting in
implementation delays.53 DHS terminated its contract with the vendor in 2008.
The system was never implemented, and DHS continued to rely on manual
enrollment processes and decades-old information technology systems. When
Minnesota later decided to build a health insurance exchange that would include
automated methods for determining eligibility for publicly funded health care,
state officials wanted to avoid the problems that HealthMatch encountered.
As Minnesota began developing its health insurance exchange, state government
consolidated its information technology operations.
The 2011 Legislature passed a law that brought all of the executive branchs
information technology resources into one department under the direction of a
chief information officer.54 Previously, information technology staff were
employed by a variety of state agencies. Under the new law, these staff were
employed by the Office of Enterprise Technology (later renamed to the Office of
MN.IT Services). The law required a transfer of authority, duties, staff, and
assets to the consolidated office by October 1, 2011. Among the responsibilities
of the office was oversight of state agency information technology projects, to
ensure their successful completion and avoid the fate of projects such as
HealthMatch.55 Despite this, the Office of MN.IT Services played a limited role
in the development of the MNsure enrollment system, as discussed in Chapter 2.
/delivering_large-scale_it_projects_on_time_on_budget_and_on_value, accessed October 1, 2014.

53
Office of the Legislative Auditor, Follow-up Review: MinnesotaCare Eligibility Determination

(St. Paul, April 2007), 12.
54
55
Laws of Minnesota 2011, First Special Session, chapter 10, article 4.
Office of Enterprise Technology, Consolidation of Information Technology Systems and

Services: Legislative Report (St. Paul, January 19, 2012), 6.
Chapter 2: Exchange Development

and Implementation
M
innesota opted to develop its own health insurance exchange under the
federal Affordable Care Act, rather than rely on the federal exchange.
This chapter examines planning and implementation of the exchanges automated
enrollment process. Later in the report, we discuss the role the MNsure Board
played in this process (Chapter 3) and the development of other MNsure
operations, such as the consumer assister programs, customer service activities,
and marketing (Chapter 6). Appendix B provides a timeline of key events in
MNsures development and implementation.
Minnesotas efforts to meet an ambitious federal deadline were

hindered by late federal rules, delays in passing state legislation,
and problems with vendor selection and performance.
Serious technical deficiencies plagued MNsures enrollment

system throughout its first year of operations.
MNsure did too little testing of the technology it developed, and

it did not make sufficient use of state governments information
technology experts.
Because of technical problems with MNsures online enrollment

system, many Medical Assistance recipients did not receive
timely reviews of their eligibility.
Minnesotas health insurance exchange has been one of the largestand most
troubledinformation technology projects in the history of Minnesota state
government. This chapter discusses many factors that contributed to the
exchanges difficulties. Lack of effort was not among them.
Minnesotas health insurance exchange encountered numerous problems despite hard work
by many staff.
Staff from MNsure, the Office of MN.IT Services, the Department of Human
Services, and vendors worked diligently to develop the exchange.1 We heard
accounts (and saw evidence) of staff working long days and nights to build and
1
Many otherssuch as insurers, counties, community organizations, and the state departments of
Health and Commercealso devoted considerable time to the exchanges planning and
implementation.
26
implement the exchange. One exchange official described working 22-hour days
after the MNsure enrollment system starteddoing her regular job for part of
that time, and addressing consumer problems during the remainder. Staff and
vendors sometimes held daily conference calls in the middle of the night. After
problems with the exchanges functionality came to light, staff and vendors
worked hard to identify possible solutions.
TIMELINESS ISSUES
When the Affordable Care Act (ACA) was signed into law in March 2010, it
required state-based health insurance exchanges to be operational by January 1,
2014.2 Subsequent federal guidance said that exchanges would need to be ready
to begin operations October 1, 2013, for the purpose of accommodating an open
enrollment period that would start on that date.3 This section examines issues
that have affected the timeliness of Minnesotas development of its exchange.
Passage of State Legislation

Exhibit 2.1 shows significant events leading to the passage of a 2013 state law
that authorized Minnesotas exchange. In 2006, Massachusetts was the first state
to pass legislation to create a health insurance exchange. Just one year later,
Minnesota Governor Tim Pawlenty advocated development of an exchange in his
State of the State speech. The Governor said an exchange would allow
uninsured individuals access to health insurance that will lower premium costs by
roughly 30 percent.4 The 2007 Legislature subsequently funded a study, issued
in 2008, that discussed how a health insurance exchange would function.5
The ACA passed in 2010, but it took three more years for Minnesota to pass legislation
creating a state exchange.
In 2010, after the ACA passed, the federal government made grants available for
states wishing to explore the creation of an exchange. However, later that year,
Governor Pawlenty issued an executive order that prohibited state agencies from
applying for discretionary grants related to the ACA, unless approved by the
governor or required by law.6 Consequently, Minnesota was one of only two
states that was not awarded a federal planning grant in September 2010.
Patient Protection and Affordable Care Act, Public Law 111-148, sec. 1321(c)(1).
76 Fed. Reg. 136 (2011), Proposed Rules, 41870.
Governor Tim Pawlenty, State of the State Address: A Better Minnesota, January 17, 2007.
Laws of Minnesota 2007, chapter 147, art. 19, sec. 3, subd. 6.
Governor Tim Pawlenty, Executive Order 10-12, Directing State Departments and Agencies
Regarding Discretionary Participation in the Federal Health Care Law, August 31, 2010.
EXCHANGE DEVELOPMENT AND IMPLEMENTATION
27
Exhibit 2.1: Key Actions Related to Creation of a

Health Insurance Exchange in Minnesota
Date
January 2007
May 2007
March 2010
May 2010
August 2010
January 2011
October 2011
March 2013
a
Action
Governor Pawlenty advocated creation of a health insurance exchange
in his State of the State speech.
The Legislature mandated a study (issued in 2008 by the Department of
Health) to examine how an exchange would function.
President Obama signed the federal Affordable Care Act (ACA) into law.
The Legislature established a health care reform task force in response
a
to the ACA.
Through executive order, Governor Pawlenty prohibited state agencies
from applying for federal planning grants related to the ACA.
Governor Dayton rescinded Governor Pawlentys August 2010 order.
Minnesota applied for (and subsequently received) a federal planning
grant to explore the possibility of establishing an exchange.
Through executive order, Governor Dayton directed the Commerce
Commissioner to develop a health insurance exchange, and he restarted
the health care reform task force that met in 2010.
The Legislature passed and Governor Dayton signed legislation that
created MNsure in state statutes.
The task force met once in 2010 and made no recommendations.
Shortly after taking office in 2011, Governor Mark Dayton rescinded Governor
Pawlentys order.7 Minnesota subsequently received a $1 million federal
planning grant in February 2011. Later in 2011, Governor Dayton issued an
executive order that directed his Commerce Commissioner to develop an
exchange.8 The Governor also directed his Commerce Commissioner to convene
a task force to provide advice on health care reforms, including the creation of a
health insurance exchange. Legislators from both parties were among those
invited to serve on the task force, but only Democrats did so. During 2011 and
2012, several bills were introduced in the Legislature that would have created a
health insurance exchange. Only one received a hearing (in 2012), and none
passed. At the state and national levels, there were strong philosophical
disagreements (often along party lines) about the merits of the ACA.
In addition, many states halted exchange planning or postponed efforts to get
legislation for exchanges while the U.S. Supreme Court reviewed the federal
laws constitutionality. In June 2012, the court upheld Congress authority to
enact most provisions of the ACA.9
Governor Mark Dayton, Executive Order 11-02, Removing Ban on Requests for Federal
Assistance; Rescinding Executive Order 10-12, January 5, 2011.
Governor Mark Dayton, Executive Order 11-30, Establishing a Vision for Health Care Reform in
Minnesota, October 31, 2011.
National Federation of Independent Business et al. vs. Sebelius, Secretary of Health and Human
Services, 567 U.S. ____ (2012), 132 S.Ct 2566.
28
Minnesota executive branch officials did a considerable amount of planning for an exchange
before MNsure was statutorily created in 2013.
After receiving the initial federal planning grant in February 2011, Minnesota
applied for and received four federal exchange establishment grantstotaling
$113 millionbefore the Legislature created MNsure in law. Exhibit 2.2 shows
the history of federal grants for Minnesotas exchange. The largest areas of
expenditure from these grants were for information technology contracts and
software support and licenses.10
Minnesota also submitted a blueprint for its exchange to the federal
government in November 2012. This blueprint was based partly on input from
the states Health Insurance Exchange Advisory Task Force and its technical
work groups, which met dozens of times. The blueprint outlined how
Minnesotas exchange would comply with federal requirements. The federal
government approved Minnesotas blueprint in December 2012, contingent on
several conditions.11
Exhibit 2.2: Federal Grants Awarded to Minnesota to

Plan and Establish a Health Insurance Exchange
Type of Exchange Grant
Planning
Establishment, Level 1, #1
Additions to prior grants
Total
Award Date
February 25, 2011
August 12, 2011
February 22, 2012
September 27, 2012
January 17, 2013
October 23, 2013
December 22, 2014
Award Amounta
$
1,000,000
4,168,071
26,148,929
42,525,892
39,326,115
41,851,458
34,343,062
$189,363,527
NOTE: States received planning grants to assess whether to establish a state-based exchange.
States received Level 1 Establishment grants for activities associated with the states exchange
model. States received Level 2 Establishment grants if they had committed to establishing a statebased exchange and had completed certain steps, such as getting legal authority for the exchange.
a
Does not include federal money provided separately to the Department of Human Services for
modernization of systems to determine eligibility in publicly funded health care programs.
SOURCE: U.S. Department of Health and Human Services, notices of awards.
10
Between July 1, 2011, and December 31, 2013, 26 percent of exchange expenditures were for
information technology contracts, and 32 percent were for software support and services. See
Office of the Legislative Auditor, Minnesota Health Insurance Exchange (MNsure), Internal
Controls and Compliance Audit (St. Paul, October 28, 2014), 6.
11
The federal government required that Minnesota (1) demonstrate the ability to perform activities
outlined in the blueprint, (2) comply with regulations and progress milestones, and (3) demonstrate
by March 31, 2013, legal authority (beyond existing authority) for the exchange to operate.
29
Contracting for the Exchanges Technology

As shown in Exhibit 2.3, the Department of Commerce issued a request for
proposals in June 2011 for development of software and hardware related to what
it called the exchanges technical infrastructure.12 Respondents were asked to
propose prototypesfor implementation of a fully functioning exchange and/or
component modules.13 Exchange staff intended to evaluate the prototypes and
pick the best vendors to separately develop each component of the exchange.
Choosing a vendor to build each component was known as a best of breed
approach.
Exhibit 2.3: Key Steps Related to Health Insurance

Exchange Contracts for Technical Infrastructure
Date
June 2011
November 2011
February 2012
May 2012
July 2012
February 2013
May 2013
Action
Department of Commerce issued request for proposals seeking
vendors to bid on development of prototypes of various pieces of the
health insurance exchange technology.
Department of Commerce selected several vendors to develop
prototypes.
Department of Commerce informed three vendors they had been
selected to develop different parts of the exchange.
State officials changed the contracting strategy. They decided to
seek a contract with one primary vendor, rather than the previous
plan of multiple vendors.
After trying to contract with Deloitte Consulting as the primary vendor,
exchange officials selected a different primary vendor (Maximus,
Incorporated).
Exchange officials began the process of making contract changes to
downgrade Maximus role and enhance the role of state staff and a
subcontractor (EngagePoint, Incorporated).
Contract amendments initiated in February 2013 took effect.
Exchange officials had problems (1) selecting vendors to develop the exchanges technical
components and (2) getting the performance they expected from the lead vendor.
After reviewing the prototypes, the Department of Commerce informed three
companies in early 2012 that they had been selected to build parts of the
exchange.14 But two months later, the department suspended the best of breed
approach and decided instead to seek a contract with a single primary vendor.
12
Minnesota Department of Commerce, Request for Proposals: Health Benefit Exchange

Technical Infrastructure Prototypes, published in State Register on June 20, 2011, p. 1998.
13
14
Ibid. The department selected certain respondents to receive stipends to develop the prototypes.
Maximus, Incorporated, was selected to develop four modules of the exchange, while IBMCuram Software was selected for two modules, and Deloitte Consulting was selected for one. The
letters, dated February 27, 2012, said the department would begin negotiations with each company
for purposes of entering into contracts.
30
The department negotiated with Deloitte Consulting to be the primary vendor but
could not reach agreement, partly because it considered Deloittes price to be too
high. The department turned its attention to a different vendor, and in July 2012
it entered into a contract with Maximus, Incorporated. Maximus subcontracted
with various vendors to help develop parts of the exchange (see Exhibit 2.4). A
Maximus official told us the selection of subcontractors was an arranged
marriage by state officials; in other words, Maximus did not have full latitude to
pick its subcontractors.
Exhibit 2.4: Subcontractors Responsible for

Technical Components of Minnesotas Exchange
Component of the Exchange
Subcontractor
Individual eligibility and exemption

Individual enrollment
Small employer eligibility and enrollment
Health plan display and navigator/broker certification
Provider display
Fund aggregation and payment
Account administration
Mobile application or accessibility
IBM-Curam
Connecture
Connecture
Connecture
Connecture
EngagePoint
EngagePoint
No contract
NOTES: These are the eight modules for which state officials sought prototypes in a July 2011
request for proposals. Ultimately, state officials chose not to pursue development of a mobile
application and did not enter into a contract for one. Responsibility for integrating the various
components was initially the responsibility of the general contractor (Maximus); following a May 2013
contract amendment, it was the responsibility of EngagePoint.
SOURCE: Office of the Legislative Auditor, based on a review of MNsure documents.
Overall, this contracting process took somewhat longer than expected. This
mostly reflected (1) the Department of Commerces decision to change from an
approach that sought multiple vendors (the best of breed) to one that relied on a
single primary vendor, and (2) the time it took to reach agreement with Maximus
following the aborted efforts to contract with Deloitte.
During the first six months of the Maximus contract, state officials working on
the exchange developed doubts about Maximus ability to deliver. For example,
state officials did not believe Maximus produced a satisfactory project plan, and
they thought that Maximus progress on project details was too slow. Some
vendor and state staff told us that Maximus focused on identifying system
requirements from scratch, despite the fact that subcontractors had been selected
largely on the basis of the off-the-shelf software they could provide to the
exchange.15 Thus, there was concern that Maximus was not spending scarce time
15
Off-the shelf products are generally presumed to be ready for use with limited tailoring. Such
products are an alternative to developing fully customized products from scratch.
31
efficiently.16 Also, exchange officials began working with subcontractors

without going through the exchanges general contractor (Maximus), and
Maximus officials believed this undermined their authority.
In May 2013, exchange staff amended the Maximus contract to revise the
vendors duties. The amendment enhanced the states role in overseeing the
project, reduced Maximus responsibilities, and revised duties of the
subcontractors.17 According to the Commissioner of the Department of Human
Services, this decision was necessary to ensure that the state could implement the
business functions that needed to be in place October 1, 2013. The
Commissioner said: Maximus was not in a position, the state thought in early
2013, to make sure we met those goals. [W]hen people learned that the prime
contractor was not in a position to execute, roles were shifted.18
About seven months after the contract amendment, MNsure staff told the
MNsure Board that the amendment was not related to dissatisfaction with
Maximus work.19 This was not true; state dissatisfaction with Maximus was
clearly a factor in this decision.
The inability of the exchanges prime contractor to execute was a setback in a
project with a tight timeline. Some state officials suggested to us that Maximus
may have lacked the expertise to oversee a project of this sort. Others suggested
that state officials failed to provide sufficient direction to Maximusnoting, for
example, that a project governance structure was not in place when Maximus
started its work. Regardless, progress on the technical components of the
exchange was slower than expected between July 2012 and Spring 2013.
Federal Guidance
The Department of Commerces Health Insurance Exchange Advisory Task
Force began meeting and making plans for the exchange in November 2011, and
it continued into early 2013. Much of its work occurred before the federal
16
Maximus acknowledged to us that it had differences of opinion with exchange staff about
development of system requirements. Maximus said exchange staff focused on the requirements
for each module, while Maximus wanted to look at processes end-to-end to make sure the various
products would fit together effectively. Maximus also told us it submitted a timely project plan but
that exchange staff provided inadequate feedback.
17
The amendment said the State will take over program management, methodology, and
responsibility for building the Solution.
18
Lucinda Jesson, Commissioner, Department of Human Services, e-mail message to James

Schowalter, Commissioner, Department of Management and Budget, Maximus Talking Points,
January 9, 2014. In January 2013, federal officials conveyed to states a list of 70 business
functions that needed to be in place.
19
MNsure board and federal relations manager, e-mail to MNsure Board members, MNsure
Maximus Background for Committee, January 3, 2014.
32
government issued any rules regarding implementation of health insurance

exchanges (in March 2012).20
The lack of timely federal regulations and guidance slowed Minnesotas pace of building its
exchange.
States needed additional federal direction to fully implement exchange-related
provisions of the federal law, but many of these rules were not issued until 2013.
Exhibit 2.5 shows a sampling of the rules. As federal requirements changed in
the weeks and months prior to the beginning of open enrollment, state officials
and software vendors scrambled to make adjustments. According to one
Minnesota official who was involved in developing Minnesotas exchange, We
were building software in quicksand.
In addition, states developing their own health insurance exchanges faced
challenges linking to the federal governments Data Services Hub. The federal
hub is a critical part of the enrollment process; it verifies applicant information
for example, regarding income and citizenship. According to the U.S.
Government Accountability Office, state officials developing eligibility and
enrollment rules for their exchanges said they did not have complete
information on the requirements of the federal data services hub.21 The U.S.
Centers for Medicare and Medicaid Services did not authorize the newly
developed federal hub to operate until early September 2013. MNsure received
federal authorization to connect to the hub on September 27, 2013, just days
before open enrollment began. The federal hub was not available for MNsure to
connect with until October 1, 2013, the first day of open enrollment.
Exhibit 2.5: Examples of Key Federal Rules

Governing Health Exchanges Issued in 2013
Essential health benefits that must be offered by products on health insurance

exchanges (February 2013)
Health insurance premium tax credits (February 2013)
Health insurer rate review (February 2013)
Health insurance exchanges for small businesses (June 2013)
Medicaid eligibility notices and appeals (July 2013)
Exchange eligibility and enrollment (July 2013)
Consumer assistance tools and standards for navigators (July 2013)
SOURCE: Health Reform GPS, http://www.healthreformgps.org/wp-content/uploads/ACA-Rule

-Appendix-7312013.pdf, accessed October 27, 2014.
20
77 Fed. Reg. 18,310, Patient Protection and Affordable Care Act; Establishment of Exchanges
and Qualified Health Plans; Exchange Standards for Employers (2012). In the meantime, state
governments and private sector officials said that uncertainty about the federal rules contributed to
lack of progress on the exchanges; see Engelberg Center for Health Care Reform at Brookings,
Aligning Public and Private Sector Timelines for Health Insurance Exchange Implementation
(Washington, DC, December 2011), 2-3.
21
Government Accountability Office, Seven States Actions to Establish Exchanges Under the
Patient Protection and Affordable Care Act (Washington, DC, April 2013), 24.
33
Deadline for Implementation

From the time the Affordable Care Act passed in 2010, the timeframe for
implementing state-based health insurance exchanges presented a significant
challenge. In addition to designing and implementing exchanges, states had to
consider upgrades to their Medicaid eligibility determination systems to comply
with the act. As one analysis of exchange implementation observed:
Of all the [Affordable Care Act] implementation threads,
[information technology (IT)] infrastructure development is
perhaps the most challenging. Development of IT systems for
public programs typically takes years. As a result of the [acts]
requirements and tight deadlines, both the federal and state
governments are developing IT systems simultaneously, in mere
months, often in advance of key policy decisions.22
Developing a complex health insurance exchange by the deadline of October 1, 2013,
presented a major challenge to Minnesota.
With a challenging timeline for implementation, Minnesotas exchange relied
considerably on commercial off-the-shelf software products. Computer
systems for the states public health care programs were old, and building new
systems from scratch would have taken a long time. Despite using off-the-shelf
products for the exchange, the software vendors did a lot of customization. This
reflected the need to adapt to federal exchange regulations and Minnesotas
complex eligibility rules for public programs.
It also took time and was a major challenge to integrate off-the-shelf software
products that had not been built to work together. A MNsure consultant observed
that the integration of several software productseach with separate and unique
user interfaces, application logic, rules engine(s) and databases resulted in a
complex system environment.23 Partly because of time spent trying to integrate
these pieces, this complex environment was not adequately tested before October
2013 (as discussed below).
State officials we interviewed generally believed that the federal governments
October 1, 2013, deadline for starting online enrollment was nonnegotiable. We
asked U.S. Department of Health and Human Services officials about this. We
were told that the department strongly wanted states that had committed to
building state-based exchanges to have those exchanges operating on October 1,
2013. However, there were two options for states unable to make this deadline.
First, a state could have sought federal approval to use the federal exchange until
22
Michael Tutty and Jay Himmelstein, Establishing the Technical Infrastructure for Health
Insurance Exchanges Under the Affordable Care Act: Initial Observations from the Early
Innovator and Advanced Implementation States (Washington, DC: National Academy of Social
Insurance, September 2012), 5.
23
Deloitte Consulting, MNsure Phase II Project: Deliverable #5, Technical Assessment, July 16,
2014.
34
its state-based exchange was completed. Federal authorization for such a plan
would have been required several months in advance of October 1, 2013, to
ensure that the state could properly connect to the federal exchange. Two states
that had committed to building state exchanges (Idaho and New Mexico) used the
federal exchange instead during the first year of open enrollment. Second, a state
could have implemented a manual system of enrollment, without a functional
online enrollment option. Oregonwhich had an inoperable enrollment website
and never enrolled anyone through itdid all of its enrollments through manual
processes. We are not aware of federal guidance that clearly communicated the
option of an entirely manual enrollment process.
Minnesota never requested an exemption from its plan to start online enrollment
through its exchange in October 2013. State officials made some decisions in
early 2013 that narrowed the exchange projects scope somewhat, and they
remained hopeful that technical problems could be addressed prior to or shortly
after October 1, 2013. In the end, however, they miscalculated their ability to
deliver a well-functioning exchange by the federal deadline.
WEBSITE FUNCTIONALITY ISSUES

MNsures online enrollment system experienced serious problems during the
year following its launch. These problems affected the experience of users who
tried to enroll, as we discuss in Chapter 5. In addition, an analysis by a Yale
University economist suggested that MNsures technical problems may have
increased the costs of insurers for their enrollees in the individual insurance
market.24 The study said that states that experienced severe glitches in their
exchange operations saw an increase in average insurance costs relative to states
that had well-functioning exchanges, after accounting for other factors.25 The
study suggested that technical problems experienced by exchange users in
Minnesota and five other states may have stopped enough healthy people from
enrolling in insurance to cause adverse selection, a problem the Affordable
Care Act was intended to mitigate.26 (Adverse selection is the tendency of
healthy people to defer buying insurance. In general, insurance is more
affordable if it covers a wide array of people, not just those who are less healthy.)
Below, we discuss key causes and implications of MNsures technical problems.
24
Amanda Kowalski, The Early Impact of the Affordable Care Act by State (Washington, DC:
Economic Studies at Brookings, September 2014), 28-29.
25
Kowalski found that insurance costs in the individual market were higher than they would have
been had the Affordable Care Act not been implemented and state-level trends in coverage,
premiums, and costs since 2008 had persisted. The study also assumed no change in the generosity
of the insurance products sold. The individual market is insurance sold directly to individuals
rather than insurance sold to a group or provided through an employers health plan.
26
The other states experiencing severe glitches were Hawaii, Maryland, Massachusetts, Nevada,
and Oregon. Due to data anomalies, Massachusetts was omitted from the statistical analysis. The
author estimated that individual market participants in the states that experienced the worst
technical problems, including Minnesota, were $750 worse off on an annualized basis compared
with individual market participants in other states with their own exchanges.
35
Involvement of State Technology Experts

Development of the exchange (and simultaneous upgrading of the states
eligibility system for public health care programs) was a large, complicated
information technology project, and exchange staff needed all the help they could
get. When the Department of Commerce issued a request for proposals to
develop the exchanges technical infrastructure in June 2011, the only
information technology specialist working for the exchange was a Department of
Health employee.
MNsures limited use of state government information technology experts may have
hindered development of the exchange.
In 2011, the Legislature created a consolidated agency (now called the Office of
MN.IT Services) to oversee information technology development throughout
state governments executive branch, but this office had a limited role in
MNsures development. The former Commissioner of the Office of MN.IT
Services told us that her initial efforts to involve her office in planning for the
exchanges software components were resisted by exchange staff. She said she
was concerned about the lack of information technology experience among the
exchange staff, so she muscled the Office of MN.IT Services into some aspects
of the software planning process. However, her offices role in overseeing
development of the MNsure enrollment system remained limited; exchange
officials played the lead role in negotiating contracts with information technology
vendors, and they managed those vendors while building the exchange.
The Office of MN.IT Services tried, at its own initiative, to implement structures
for making exchange-related information technology decisions, but exchange
officials did not sustain these over time. The states contract with Maximus
specified that the Department of Commerce should establish a project
governance structure, which is important for soliciting input from stakeholders,
setting priorities, monitoring progress, and controlling changes in project
components. The department did not, so the Office of MN.IT Services
established one in late 2012.27 Exchange officials initially used the governance
structure to make decisions. Later, when technical problems arose, exchange
officials went into what one participant described as firefighting mode and
stopped using the decision-making structures. In a July 2014 report, an external
consultant to MNsure said: The cumulative effect [of a diluted project
governance structure] has been to create confusion among most leads and
stakeholders, inconsistent adherence to processes, [and] untimely decision
making and issue resolution.28
27
The governance structure included an Executive Steering Committee, which provided overall
project direction and oversight. The structure also included a Solutions Architecture Team, a
Business Architecture Committee, a Technical Architecture Committee, and a Security
Architecture Committee.
28
Deloitte Consulting, MNsure Phase II Project: Deliverable #2, Program and Project
Management Assessment, July 11, 2014.
36
Although the Office of MN.IT Services inserted itself into many exchange
discussions and decisions, the offices role was eclipsed in 2013 by a change in
state law. Legislators passed statutory language that exempted MNsure from
much of the Office of MN.IT Services authority. For example, the Office of
MN.IT Services is required by law to evaluate whether information technology
projects proposed by state agencies will meet users needs and assess the
projects costs and benefits against other options.29 This statutory requirement
was part of the 2011 Legislatures effort to ensure proper oversight of large
information technology projects in state government. However, the Legislature
passed legislation in 2013 that said this section (and many others related to the
Office of MN.IT Services) did not apply to MNsure.30
Because the Office of MN.IT Services ultimately assumes responsibility for
supporting the software developed by agencies and vendors, it is important for
this office to be a key partner in the development process. But, during the
development of MNsure, vendor staff were generally not paired with
counterparts at the Office of MN.IT Services so the state staff could learn about
the vendor-designed systems; this was in contrast to practices on many other state
information technology projects. In addition, some Office of MN.IT Services
staff told us that the MNsure vendors did not fully communicate the nature of the
data that would reside in the systems. Thus, as of late 2014, Office of MN.IT
Services staff were still learning what they needed to know to support this system
when the vendors finish their work.
RECOMMENDATION
The Legislature should amend Minnesota Statutes 2014, chapter 62V, to ensure that MNsures
future information technology work is subject to oversight from the Office of MN.IT Services.
The 2011 Legislature integrated the states information technology staff and
authority into a single agency, partly to help ensure the success of difficult
projects. We think it was unwise for MNsure to undertake a large information
technology project in 2011 without forging a close working relationship with the
Office of MN.IT Services. It was also unwise for the 2013 Legislature to exempt
MNsure from ongoing oversight by the Office of MN.IT Services. While there
were many reasons for the technical problems MNsures enrollment system
encountered, a state law that limited the Office of MN.IT Services project
oversight and decision-making authority did not help.
During the past year, a stronger working relationship has developed between
MNsure, DHS, and the Office of MN.IT Services. These agencies have jointly
designed an information technology governance structure to facilitate orderly
decisions on the health insurance exchange and modernization of Minnesotas
enrollment systems for public health care programs.
29
Minnesota Statutes 2014, 16E.03, subds. 3 and 4.
30
Laws of Minnesota 2013, chapter 9, sec. 5.
37
MNsure is an organization that relies considerably on its online enrollment

system. This system will need to change over time. The involvement of the
states technology experts in these changes should not be dependent solely on the
informal working relationships of officials at MNsure and the Office of MN.IT
Services. We suggest that the Legislature amend MNsures governing statute to
subject MNsures information technology projects to the oversight that other
state agencies receive.
Pre-Launch Indicators of System Readiness

Because development of the health insurance exchange was a difficult project
with an ambitious timeline, we looked at documents that showed the progress of
this project as the first open enrollment period neared.
Various red flags in the weeks and months prior to October 1, 2013, suggested that the
launch of the exchange might not go well.
Some early red flags came in reports prepared by an independent contractor for
exchange officials and the federal government. In February 2013, the exchange
contracted with a vendor to provide independent verification and validation
(IV&V) that the exchanges software and systems met expectations and
requirements. The vendor produced quarterly reports that assessed the exchange
in several areas, as summarized in Exhibit 2.6. In the areas of scope, schedule,
staffing, and quality, the vendor indicated marginal or unsatisfactory
performance in reports issued prior to October 1, 2013 (and thereafter).31
Exhibit 2.6: Independent Assessments of Overall

Project Status for MNsure, April 2013 to January 2014
Project Areas
Assessed
Scope
Schedule
Cost
Staffing
Quality
4/19/2013
Marginal
Unsatisfactory
Good
Marginal
Marginal
Overall Project Status, as of:

7/31/2013
10/31/13
Marginal
Unsatisfactory
Good
Marginal
Marginal
Unsatisfactory
Unsatisfactory
Good
Marginal
Marginal
1/31/2014
Unsatisfactory
Unsatisfactory
Good
Marginal
Marginal
NOTES: The reports that assessed the overall health of the MNsure project also rated the health of
the project aimed at modernizing the Department of Human Services eligibility and enrollment
systems. The ratings shown in this table are only the ratings for the MNsure project.
SOURCES: Software Engineering Services, Minnesota Insurance Exchange and Eligibility and
Enrollment System Modernization Internal Verification and Validation (Bellevue, NE, May 28, 2013,
August 30, 2013, December 11, 2013, and March 20, 2014).
31
In addition to these ratings of overall project health, the reports presented ratings for a variety
of more specific project activities.
38
In addition, the IV&V vendor prepared an urgent notice to MNsure on

September 12, 2013, which said:
Critical dates for [an October 1] deployment of MNsure are not
being met and [we are] very concerned that the functionality of
the system being deployed on [October 1] is not well understood
and the system will be deployed without adequate testing or
contingencies in place.32
Another series of red flags were the project status reports prepared by the
Office of MN.IT Services during much of 2013. Each week, this office
developed color-coded reports that summarized the status of exchange elements
related to information technology. Items were coded green if they were
performing as expected; yellow indicated caution regarding an items
performance, and red indicated that there were critical issues. From June 2013
through September 2013, the Office of MN.IT Services consistently coded as red
the overall status of the exchange project. This reflected concerns about the
projects scope, schedule, and risk as October 1, 2013, approached.
In addition, the federal Centers for Medicare and Medicaid Services expressed
concern about some issues as the system prepared to go live. Federal officials
conducted reviews of Minnesotas progress at several points during exchange
development. In a September 2013 review, federal officials identified many
areas in which they said MNsure had demonstrated the exchanges functionality.
But the federal report said that, as of September 12, the system had 270
unresolved defects, including 97 major defects with no workarounds.33 The
report said that resource availability to address these concerns before October 1,
2013, was a big concern. Further testing on September 19 led federal officials
to state: At this time, Minnesota is unsuccessful in completing multi-tax
household eligibility determinations, cannot generate systems logs for validating
appropriate access to the [Federal Data Services] Hub, and is having significant
difficulty with systems testing and defects. Although the reports identified areas
of concern, the federal government authorized Minnesotas exchange to connect
to the Federal Data Services Hub.
MNsure failed to adequately test the exchanges website and software before enrollment
started in October 2013.
When a complex information technology system is developed, it is important to
ensure that it will work as expected. Thus, system developers plan for various
types of testing. For example, integration testing examines whether various
system components work together effectively. Load testing examines whether
a system can handle particular numbers of users or amounts of data. User
acceptance testing determines whether a system works as intended for the user.
32
33
Software Engineering Services, MNsure IV&V Urgent Time Critical Notice, September 12, 2013.
Center for Medicare and Medicaid Services, Minnesota Operation Readiness Review Summary
(September 2013), unpublished.
39
The project schedule for Minnesotas health insurance exchange called for the
state to approve a testing plan by mid-December 2012, but the testing plan that
Maximus produced was not finalized until February 26, 2013. It was then
reworked by a different vendor (EngagePoint) when exchange staff amended the
duties of the contractors. The revised testing plan was completed in late May
2013.
Actual tests of the exchanges enrollment system also occurred later than
expected. This reflected delays in development of the software and other parts of
the system, as vendors made ongoing changes to their products and the federal
government continued to issue new exchange rules well into 2013. Key tests
were not completed until immediately prior to the start of open enrollment, and
the results were not summarized until open enrollment was already underway.
Exhibit 2.7 shows the dates when tests were supposed to be completed and when
they actually were.
Exhibit 2.7: Target and Actual Dates for Completion

of MNsure System Testing
Type of Testing
Integration Testing
Load Testing
User Acceptance Testing
Target Dates for Completion of Testing

MNsure Project
Maximus
Schedule
Testing Plan EngagePoint
Testing Plan
(November
(February
(May 2013)
2012)
2013)
6/28/13
7/5/13
9/9/13
6/28/13
7/5/13
9/9/13
8/15/13
9/15/13
9/15/13
Actual
Completion
10/1/13
10/2/13
10/1/13
NOTES: Testing is completed when the test results have been compiled into a report. According to
the testing reports, load testing and user acceptance testing ended September 30, 2013, while
integration testing ended August 5, 2013.
SOURCES: MNsure, Project Schedule, http://www.mnsure.org/images/MNHIX-ProjectSchedule11
-19-12.pdf; accessed June 15, 2014; Minnesota Health Insurance Exchange (MNsure) Test Report,
Release 1 (October 1, 2013); Nick Richardson, SOASTA CloudTest Summary Report (October 2,
2013); Maximus, Inc., Minnesota Health Insurance Exchange Final Testing Plan (February 26, 2013);
and EngagePoint, Testing Plan, Version 2.04 (May 30, 2013).
The results of the tests showed significant problems. For example, load testing in
late September 2013 found that the MNsure website could not handle more than
1,000 simultaneous users. This was well short of the goal (10,000 users) set by
system developers. Tests showed that when the number of users surpassed
1,000, people trying to access the system would receive error messages or be
logged out of the system. Another test looked at the enrollment systems ability
to accurately handle a variety of scenarios. For example, a scenario might test
whether the system worked as intended for a user with specific demographic
characteristics. Of 143 scenarios that testers examined, 53 (37 percent) failed.
Based on this and other testing results, an October 1, 2013, report concluded:
[User Acceptance Testing] overall status is red. This meant that the system
defects shown by the testing were a significant problem.
People involved in the exchanges development generally agreed that an
inadequate amount of testing occurred. A top official with the Office of MN.IT
40
Services said that perhaps 40,000 scenarios should have been tested before
October 1, 2013, given the wide variety of individual and household
characteristics among potential applicants. The number of scenarios actually
tested was less than 1 percent of this amount. Consequently, as one MNsure
Board member told us, MNsure ended up testing on the citizens in
Minnesota.34
The inadequacy of testing partly reflected time constraints, but it also reflected a
general inadequacy of MNsures provisions for quality assurance. The exchange
did not have a quality assurance manager until May 2013, just five months before
open enrollment started. According to the Office of MN.IT Services, the quality
assurance manager and his staff largely performed business support tasks rather
than quality assurance work. An Office of MN.IT Services manager told us that
the quality assurance leadership was inexperienced and hindered by constant
changes in the exchange software.
Problems After the Exchange Opened

The MNsure website did not experience serious technical problems when it was
initially launched on October 1, 2013.35 This contrasted with the websites of the
federal governments exchange and some other state-based exchanges. Still,
there were some problems with the functionality of MNsure online enrollment
process in early October. MNsure customers encountered difficulties getting
their identities verified through the MNsure system; state staff told us this
reflected problems with the design of system elements at both the federal and
state levels. Also, the eligibility part of MNsures enrollment system initially
allowed people to submit multiple applications. For example, an applicant might
submit a second application to correct an error made in the first one; the MNsure
system did not have the ability to accept changes. When MNsure received
multiple applications, it was unclear which application was the correct one.
MNsures technical problems escalated through Fall 2013 and continued well into 2014.
Significant problems with the MNsure enrollment system emerged in midNovember. MNsure made system changes in an effort to fix errors that had been
programmed into the system, but these changes triggered a problem that caused
some applications to get trapped in the MNsure system. In such cases, it was
difficult for staff to view or retrieve the applicants information. Based on our
34
In addition, testing of the MNsure site often occurred in what is called the production
environment, which is a version of the system that consumers use. Testing in this environment
can disrupt users, and the cost of fixing a problem identified in this environment is typically greater
than the cost of fixing a problem identified earlier in the process.
35
MNsure delayed opening its website until the afternoon of October 1 so that it could conduct
additional testing of the system.
41
document review, the number of MNsure cases affected by this problem peaked
in December 2013 and again in March 2014.36
In mid-December 2013, Governor Dayton sent a letter to the chief executive
officer of the IBM Corporation, expressing concerns about the functionality of
the IBM product used for MNsures eligibility determinations. The Governor
asserted that IBMs product was not as ready out-of-the-box as IBM had
conveyed to exchange officials in 2011, and other state officials agreed that the
product needed many modifications. The Governor said the IBM product did
not properly perform eligibility determinations or verify individuals application
information, as required under federal law.37 Office of MN.IT Services experts
we spoke with said the software used for IBMs portion of the exchange was
immature, suggesting that it needed substantial modification for use in
MNsure. IBM said that most of the concerns raised in the Governors letter were
not attributable to IBMs product or had been resolved prior to the Governors
letter. However, IBM sent a team of staff to St. Paul in December 2013 to help
address issues with the functionality of its software.38
In January 2014, MNsure contracted with a health care consulting firm (Optum)
to give an external assessment of MNsures problems. Regarding MNsures
software, Optum said a large gap exists between required functionality and what
has been delivered.39 Optum highlighted weaknesses in the MNsure systems
technical capabilities, such as the following:
MNsure did not have a way to link all data related to a given consumer.
It is important for an enrollment system to specify an authoritative data
sourceoften called a system of recordespecially when the system
relies on data from multiple sources. This helps to ensure the integrity of
the data. A system of record was not built into the MNsure enrollment
database until late 2014.
MNsure lacked the ability to process insurance changes due to life

events. Changes in income or family size, for example, are common
and can affect someones eligibility for public health care programs and
tax credits. Thus, it is important for a health insurance exchange to
efficiently make such changes. MNsure manually processed changes due
to life events through 2014; it hopes to have a more automated process
sometime in 2015.
36
This application limbo was known as the Process Instance Error (PIE) queue. Records we
obtained from the Office of MN.IT Services were not available for all dates, but there were 1,800
cases in the PIE queue as of December 15, 2013. Records show that the number declined sharply
in January 2014, but it climbed to more than 2,500 in March 2014 before declining to near zero in
late April 2014.
37
Governor Mark Dayton, letter to Virginia Rometty, Chairman, President, and CEO, IBM
Corporation, December 13, 2013.
38
Although IBM asked the state to have 200 workstations available for its staff, and some media
reports said IBM sent dozens of staff to St. Paul, we were told the actual number sent was about
nine.
39
Optum, MNsure Assessment Summary: Contact Center; Technical Program Management;

Software and Data, January 17, 2014.
42
In April 2014, MNsure contracted with Deloitte Consulting for a more in-depth
assessment of MNsures technical functionality and its information technology
governance and management. In addition, Deloitte became the lead primary
point of contact with all vendors as it helped MNsure make necessary fixes.40
Deloitte assessed the status of 73 functions that it said are expected in a robust
Health Insurance Exchange.41 Some of these functions (such as renewals of
insurance coverage) were ones MNsure did not need to implement for its first
enrollment year, but many should have been functional in October 2013. In June
2014, Deloitte reported that only 26 of the 73 functions worked as expected.
Exhibit 2.8 shows examples of issues Deloitte identified.
Deloitte observed that:
Several aspects of the MNsure system architecture are
contemporary and consistent with industry practices. The
systems foundation, however, centers on the integration of four
unique and independent [commercial off-the-shelf] products.
The system architecture is complex, with eachproduct
presenting a separate user interface, a separate rules engine, and
a separate database architecture. The resulting architecture
requires a high level of effort to maintain the system on an
ongoing basis.42
Exhibit 2.8: Examples of MNsure Technical Problems

Cited by Deloitte, June 2014
In the enrollment website, applicants could not edit data they had entered, view their
eligibility status, or view their enrollment information.
The system used a less-than-ideal process for applicant identity matching, potentially
resulting in duplicate cases for the same person.
The system lacked functionality to process renewals of already-enrolled individuals.
Changes made by caseworkers did not necessarily result in the expected eligibility
determinations.
The system incorrectly indicated that some applicants did not qualify for a tax credit.
The system did not process effectuation and termination notices from insurers, so MNsure
enrollment information may have been out-of-date.
Outside of open enrollment, the system lacked functionality to handle new applicants or
make changes in existing cases due to life events (births, deaths, marriage, etc.).
Caseworkers processed paper applications as time permitted but did not always have the
ability to backdate eligibility to the date the application was received.
The system could not authorize navigators or brokers to complete an online application on
behalf of a client.
MinnesotaCare clients could not make online payments.
SOURCE: Deloitte Consulting, MNsure Phase II Project, Deliverable #3, Phase 1 Functional and
Technical Assessment, June 18, 2014.
40
MNsure, Professional and Technical Services Contract with Deloitte Consulting LLP, April
2014.
41
Deloitte Consulting, MNsure Phase II Project: Deliverable #3, Phase 1 Functional and
Technical Assessment, June 18, 2014, 7.
42
Deloitte Consulting, MNsure Phase II Project, Deliverable #5, Technical Assessment, July 16,
2014, 7 and 13.
43
Deloitte provided a roadmap for addressing the deficiencies it cited. MNsure

then worked with Deloitte, other vendors, DHS, and the Office of MN.IT
Services on these issues over the next several months. By most accounts,
applicants during MNsures second open enrollment period (starting in
November 2014) had a better user experience than applicants in the first open
enrollment period.
MNsures technical problems resulted in a wide variety of time-consuming workarounds.
When the online processes for enrolling individuals through MNsure did not
work as smoothly as anticipated, other agencies and individuals stepped in to
help. Often, it was necessary to process enrollments manually rather than
through an automated process. Also, many individuals completed paper
applications, often because of difficulties encountered with the online application
process.
There are no definitive estimates of the fiscal impact of the workarounds that
were developed in response to technical problems, but some examples of these
impacts are discussed below.
43
MNsure has devoted significant numbers of its customer service staff to

manually process online applications that got stuck in the MNsure
system. Even high-ranking MNsure officials spent many hours during
2013 and 2014 trying to address individual cases that encountered
technical problems.
Due to technical problems with the MNsure website, the volume of paper
applications MNsure received was larger than anticipated.43 During
MNsures first year of enrollment, it received more than 85,000 paper
applications. Until November 2014, staff from the Minnesota
Department of Human Services manually processed most of these paper
applications. The department billed MNsure for a portion of the costs.
MNsure-certified navigators provided follow-up services to an

unexpectedly large share of consumers. Follow-up often occurred when
the consumers identities could not be verified in an automated way or
their applications could not be readily retrieved from MNsure's
enrollment system.
As we discuss in Chapter 5, health insurers and counties devoted a

significant amount of time to MNsure processes that were supposed to
have been automated.
The creation of an online health insurance exchange was never intended to eliminate the option
of paper applications for people seeking to enroll in public health care programsfor example, for
people without access to a computer. However, exchange and DHS officials hoped that the ability
of individuals to access MNsures online enrollment process at counties or by working with
assisters (such as navigators) would limit the number of cases requiring paper applications.
44
Eligibility Determination Issues

The Minnesota Department of Human Services (DHS) is responsible for ensuring
that individuals eligibility to receive benefits from the states public health care
programs is determined correctly. These programs include Medical Assistance
(the states Medicaid program), MinnesotaCare, and the Childrens Health
Insurance Program.
In recent years, DHS has explored ways to modernize and integrate the eligibility
systems for these and other programs. DHS undertook these efforts, sometimes
with direction from the Legislature, separate from implementation of
Minnesotas health insurance exchange.44 In 2011, DHS rated vendor
demonstrations of products that could be used to redesign the states eligibility
determination systems; IBM Curam was ranked the highest. The following year,
the exchange selected IBM Curam to develop a portion of its enrollment system,
and the modernization of DHSs eligibility systems and the development of the
exchange became, in effect, a joint venture.45
Eligibility determinations for public health care programs are a critical function
of MNsure. As of November 2014, enrollees in these programs accounted for
85 percent of the people who enrolled through MNsure.
There have been serious problems with eligibility determinations for persons newly enrolling
in public health care programs through MNsure.
For more than ten years, the Office of the Legislative Auditor (OLA) has raised
concerns about DHSs ability to ensure that its eligibility determinations for
public programs are correct. The development of MNsure provided the agency
with an opportunity to improve in this area. However, in November 2014, OLA
issued a report on DHSs oversight of eligibility determinations that had occurred
through MNsures enrollment system, and the report still found significant
problems.46 For example:
In a sample of persons who enrolled through MNsure, 17 percent were

not eligible for the program in which they were enrolled. For cases in
which individuals had duplicate accounts in public health care programs,
the state paid too much to insurers for health care benefits.
44
Laws of Minnesota 2011, First Special Session, chapter 9, art. 9, sec. 17, directed DHS to issue a
request for information regarding an integrated service delivery system. DHS entered into a
contract with a vendor in August 2012 to conduct planning related to system modernization. In
earlier years, DHS initiated development of a failed eligibility system without legislative direction.
45
Minnesota Department of Human Services, Redesigning Service Delivery: Planning for an

Integrated Service Delivery System, A Report to the 2012 Minnesota Legislature (St. Paul, January
2012), 39.
46
Office of the Legislative Auditor, Department of Human Services, Oversight of MNsure

Eligibility Determinations for Public Health Care Programs, October 2013 through April 2014 (St.
Paul, November 12, 2014).
45
DHS did not ensure that the records of individuals who enrolled through
MNsure in public programs were accurately and completely transferred
from MNsure to DHSs medical payment system. Inadequate controls
over such data transfers increased the risk that data could have been lost
or altered during the transfer process or accessed by people without
authorization.
DHS did not charge premiums for MinnesotaCare recipients for the first
three months of 2014, and MNsures enrollment system did not properly
calculate premiums after this time.47
DHS did not have an effective process to resolve discrepancies between

Social Security, citizenship, and immigration information reported by
MNsure applicants and information maintained by the state and federal
governments.
DHS attributed many of these issues to the implementation of MNsures new

eligibility determination system. The DHS Commissioner described the OLA
findings as serious and said DHS is working to improve compliance as we
move forward with technology improvements.48 She said some of the problems
cited in the audit had been addressed already or would be soon. Nevertheless, the
audit showed that MNsures new eligibility system had significant weaknesses,
and future audits will determine the extent to which these have been addressed.
Due to problems with MNsures functionality, DHS postponed annual eligibility reviews for
people who had enrolled in public health care programs before October 2013. Some
recipients did not have their eligibility reviewed for two years.
Initially, DHS intended to use the new MNsure system to enroll or re-enroll most
of its public health care programs participants.49 Fiscal notes for the 2013
legislation that created MNsure estimated that 621,000 individuals would enroll
or re-enroll in Medical Assistance through MNsure during 2014.50 The actual
total, as of October 2014, was much lower (233,000). All of the Medical
Assistance recipients who enrolled through MNsure from October 1, 2013, to late
2014 were new enrollees; none were renewals.
DHS delayed re-enrolling existing Medical Assistance cases through MNsure
until the MNsure system worked better. DHS told counties that the extent of the
47
Ibid., 29-31. The audit said DHS had not billed MinnesotaCare recipients for the January to
March 2014 premiums as of September 2014. In its response to the audit, DHS said it intended to
identify calendar year 2014 billing errors and send statements to enrollees by June 2015.
48
Ibid., 51.
49
The Affordable Care Act and state legislation changed the basis on which income is evaluated to
determine eligibility for public health care programs. Effective January 1, 2014, modified
adjusted gross incomeas defined for federal income tax purposesis the income measure used.
Some people qualify for public health care programs using other criteria, such as disabilities; these
individuals do not enroll in these programs through MNsure.
50
See fiscal notes for H.F. 5-8E, 2013 Leg., 88th Sess. (MN); and S.F. 1-7E, 2013 Leg., 88th Sess.
(MN).
46
delay would depend on MNsures ability to process eligibility changes related to

consumers life events, such as changes in income, household size, address, or
citizenship. DHS also said it would consider the overall functionality of the
MNsure system before moving existing cases to this system.
In addition, DHS asked the federal government for permission to defer eligibility
re-evaluations for Minnesotas existing Medical Assistance recipients; these
reviews are required annually. The federal government approved this request.
This meant that individuals could continue to receive Medical Assistance without
having their eligibility redetermined. Initially, this waiver of the federal renewal
requirement was for January through March 2014, but it was extended and was
still in effect in late 2014.
So far, relatively few of the people who enrolled in public programs before
October 2013 have had their eligibility reviewed since MNsures enrollment
system opened. Through the end of 2014, DHS had renewed about 63,000
MinnesotaCare enrollees through MNsure. In November 2014, DHS postponed
the review of about 600,000 Medical Assistance recipients who enrolled before
October 1, 2013. DHS hopes to conduct these reviews between March and
December 2015.51
The postponement of eligibility redeterminations has undoubtedly had fiscal
implications, although the exact magnitude is hard to estimate. Inevitably,
redeterminations find that some people no longer meet eligibility criteria, and
they are terminated from public programs or transferred to different ones. But, as
of late 2014, some individuals in public health care programs had not had their
eligibility reviewed for two years. This means that some people who should have
been deemed ineligiblefor example, due to increases in income or reductions in
family sizeare still enrolled in public programs.
In an effort to consider the possible impacts of deferred eligibility redeterminations,
we obtained monthly trend data on what DHS calls disenrollments of Medical
Assistance recipients. For this analysis, a disenrollment was defined as an
individual who was not eligible for Medical Assistance in the current month but
was in the previous month.52 We observed that the total number of disenrollments
for January to August 2014 was less than half the number that occurred during
comparable periods in the previous two years. The deferral of eligibility
redeterminations may be a factor in Minnesotas reduced number of Medical
Assistance case closures, although there may have been other factors, too. If fewer
Medical Assistance cases were closed due to the deferral of annual case reviews
through MNsure, this may have resulted in higher program costs than would
otherwise have been the case.
51
DHS has obtained federal approval to temporarily keep cases in its old eligibility system
(MAXIS) and, starting in late 2014, conduct determinations there until cases can be transitioned to
MNsure.
52
The analysis was limited to individuals (1) whose enrollment occurred prior to October 1, 2013,
and (2) whose continuing eligibility would likely be based on modified adjusted gross income.
As noted in Chapter 1, this income measure is used to determine eligibility for most enrollees in
public health care programs, as of January 2014.
Chapter 3: Governance
M
Nsure is governed by a seven-member board. The board consists of six

members appointed by the governor, plus the Commissioner of the
Department of Human Services (or the commissioners designee). The board
appoints the agencys chief executive officer.
This chapter discusses the role the MNsure Board has played in the development
of Minnesotas exchange. It also discusses MNsures compliance with board
policies and state law, and communication between MNsures administrative
leadership and other state officials. We recommend a change in the process for
appointing the MNsure chief executive officer, and we suggest that the
Legislature consider whether to keep a governing board for MNsure.
The MNsure Board had little influence over exchange operations

prior to the launch of the MNsure enrollment website.
MNsure staff withheld key information from the board and other
state officials during 2013.
The multi-agency governing structure for MNsures online

enrollment system lacks formal authority.
MNsure leadership has not implemented some internal policies

and state requirements.
BOARD ROLE
MNsures statutory authority resides with the MNsure Board, except in cases
where this authority has been formally delegated to staff. However, state law did
not give the board significant authority until several months after the board was
created.
Due to statutory provisions, the MNsure Board had little influence over exchange operations
prior to the start of open enrollment in October 2013.
The Governor signed MNsures enabling legislation in March 2013 and
appointed board members on April 30, 2013. However, the board was required
by statute to adopt internal bylaws, policies, and procedures before it could
assume its authority.1 The board developed various policiesfor example, a
fiscal policy and a policy on public involvement in the boards decision-making
Laws of Minnesota 2013, chapter 9, sec. 14.
48
process. The board assumed its statutory authorities and responsibilities on

August 21, 2013. This was less than six weeks before the exchanges enrollment
website opened.
The months prior to when the board assumed its statutory authority were a
critical time in MNsures development. Vendors developed key software
components, and staff made plans for MNsures contact center, outreach grants,
and marketing. But the board complied with statutes and adopted a series of
internal policies, rather than giving its full attention to operational issues. During
this time, MNsure administrative staff made key system development decisions,
and the Commissioner of Management and Budget temporarily fulfilled the
boards role.
After assuming full governance responsibilities, the board played a more active role in
oversight of MNsure operations.
State law requires the MNsure Board to meet at least quarterly.2 In fact, the
board has met much more frequently than this. During the boards first four
monthsthe period when it adopted policies and bylawsthe board averaged
two meetings per month. From September 2013 through the first open
enrollment period (ending in March 2013), the board averaged about 2.6
meetings per month. From April 2014 through November 2014, the board
averaged 1.5 meetings per month.
When problems with the MNsure enrollment systems functionality occurred in
Fall 2013, the board explored ways to fix the problems. The board chair began
going to the MNsure offices daily during the period when enrollment difficulties
peaked. A consultants report said the board essentially stepped into the
[executive director] role, responsible primarily for daily operations.3 In
addition, the board initiated two external reviews of MNsure that were conducted
in 2014.4 While a part-time board cannot be expected to oversee all of the details
of MNsures day-to-day operations, board members rightly raised questions
about MNsures performance and asked appropriate questions of staff.
BOARD POLICIES
Because the boards main focus between May 2013 and September 2013 was
adopting a set of internal policies, we examined whether these policies have been
followed.
Some of the MNsure Boards initial policies have not been implemented on schedule.
Deloitte Consulting, MNsure Phase II Project: Deliverable #2, Program and Project
Management Assessment, July 11, 2014, 18.
These were the 2014 reviews by Optum and Deloitte Consulting.
GOVERNANCE
49
First, MNsure staff have not prepared monthly and quarterly financial reports for
the board, contrary to board policy.5 As of September 2014, the board had
received only one quarterly financial report (for January to March 2014), and it
had received no monthly financial reports. MNsure finance staff told us that,
initially, their time was needed to help individuals who were encountering
enrollment problems. However, MNsures financial status merits close and
ongoing attention. Legislators of both parties expressed concern to us about the
brevity of the budget MNsure presented to legislators in 2014. Also, MNsures
financial viability will be important to monitor as MNsures federal grants are
exhausted.
Second, board policy called for staff to submit a set of annual goals for MNsure
to the board, but this has not occurred.6 Measures were supposed to address at
least the following categories: (1) access to health insurance, (2) affordability of
health insurance, (3) consumer experience, (4) health plans, and (5) finance.
Staff were supposed to set start-up goals initially, with longer-term goals
related to strategic priorities in subsequent years. In October 2013, the board
heard suggestions from several external groups about possible measures. As of
early 2015, MNsure leadership has not proposed goals in each of the required
areas to the board.7
Third, board policy requires that the board receive, for approval, a three-year
financial plan based on critically evaluated assumptions that are provided to the
Board along with an operations plan.8 The three-year plan presented to the
board in December 2014 provided limited information on the plans underlying
assumptions. The plan included assumptions regarding MNsures number of
enrollments in qualified health plans (which affect the amount of premium
revenues MNsure collects), but it did not discuss other assumptions regarding
revenues or expenditures.
Fourth, the board adopted a policy in August 2013 that required the board to
initially evaluate its own performance no later than April 2014.9 The evaluation
was supposed to consider ten dimensions of performance and would involve
either a board survey or structured interviews. As of early 2015, the board has
not conducted a self-evaluation.
MNsure Board Policy #07.1 (Fiscal Policy: Financial Planning, Budgeting, Financial Reporting
and Asset Protection).
MNsure Board Policy #09 (Reporting, Measurement and Evaluation for MNsure).
MNsure staff presented a three-year financial plan to the board in December 2014, which included
forecasts of revenues and expenditures. Because the plan was based partly on assumptions about
enrollments in qualified health plans, it contained some measures related to health insurance access
and finance, as required by the boards policy.
MNsure Board Policy #07.1 (Fiscal Policy: Financial Planning, Budgeting, Financial Reporting
and Asset Protection).
MNsure Board Policy #10 (Reporting, Measurement and Evaluation for Board of Directors).
50
RECOMMENDATION
The MNsure Board should ensure implementation of its policies or, if necessary, revise policies
that are not realistic to implement.
After October 1, 2013, the board spent much of its time trying to ensure that
MNsures operational issues were addressed. This was appropriate, in our view.
Still, we are concerned that implementation of some of the boards key policies
related to financial management and organizational accountability have been
deferred. If board members believe that some of the previously adopted policies
are no longer reasonable, this should be the subject of board discussion. Where
necessary, the board should adopt new schedules for implementing its policies
and adhere to them.
COMPLIANCE ISSUES
MNsures governing statutesadopted by the Legislature in 2013establish a
framework for the exchanges operations. We examined MNsures compliance
with key requirements.
Three state requirements of the MNsure Board were not properly implemented.
First, the board has not yet established formal policies for the operation of
MNsures Navigator Program, call center, and customer service provisions
(discussed in Chapter 6), contrary to state law. State law required the
implementation of board policies and procedures for these activities by
January 1, 2015.10 The board delayed establishing navigator policies to
(1) prevent a disruption to the Navigator Program during the open enrollment
period that ended February 15, 2015, and (2) allow for potential legislative action
in 2015.11 It intends to initiate a rule-making process regarding navigator
policies later in 2015, with rules to take effect in Fall 2015. The board has not
outlined a plan for establishing call center or customer service policies.12
Second, the board has not fully complied with state rules regarding assister
compensation rates. State rules require the MNsure Board to (1) annually set
compensation rates for consumer assisters and (2) publish the initial
compensation rates and any changes in rates in the State Register.13 The MNsure
Board did not set the assister compensation rates used in the second open
10
Minnesota Statutes 2014, 62V.05, subd. 4(a).
11
MNsure Board Policy #15 (Consumer Assistance Program: Roadmap for Designing a Navigator
Program for the Future).
12
The board adopted a call center policy related to vendor contracting, but it has not established
policies related to training standards, working with other state customer service resources, or
customer service measures.
13
Minnesota Rules, 7700.0090, subp. 1, posted July 11, 2013.
GOVERNANCE
51
enrollment period.14 Also, for the first year of open enrollment, a notice that the
rates would be posted on MNsures website was published in the State Register,
but there was no subsequent publication in the State Register when MNsure staff
changed the rates prior to the first open enrollment period. MNsure belatedly
published rates in the State Register for the second open enrollment period in
early February 2015.
Third, MNsure did not establish until 2015 an interagency agreement with the
Office of MN.IT Services required by state law. MNsures enabling statute,
which was enacted in March 2013, required the MNsure Board to establish an
agreement with the chief information officer of the Office of MN.IT Services for
information technology services that ensures coordination with public health care
programs.15 An agreement between the two agencies was executed in February
2015.16
RECOMMENDATIONS
The MNsure Board should:
Adopt consumer assister compensation rates annually and ensure that changes in the rates
are published in a timely manner in the State Register; and
Adopt navigator, call center, and customer service policies in 2015.
GOVERNANCE OF THE MNSURE ENROLLMENT

SYSTEM
As our offices auditors and evaluators have worked with staff at MNsure, DHS,
and the Office of MN.IT Services on this and other recent reviews, we have been
asked to make a clear distinction between the agency called MNsure and the
MNsure enrollment system these agencies have worked to develop. In addition,
we have been asked to acknowledge that while the MNsure Board governs the
agency, the MNsure enrollment system needs a different governance structure
one that provides DHS and the Office of MN.IT Services explicit authority to
participate with MNsure in decisions about the systems maintenance and
14
For the first open enrollment period, MNsure issued the initial assister compensation rates in
April 2013, before the MNsure Board was appointed. MNsure officials are not sure whether the
Commissioner of Management and Budgetin his capacity to fulfill the duties of the MNsure
Board before that board assumed its full statutory dutiesauthorized those rates.
15
16
Minnesota Statutes 2014, 62V.05, subd. 7(a)(1).
Minnesota Statutes 2014, 62V.05, subd. 7(a)(2), separately requires the MNsure Board to
establish an agreement with DHS for cost allocation and services regarding eligibility
determinations and enrollment for public health care programs. Also, MNsure has a data-sharing
agreement with the Office of MN.IT Services. However, these agreements are different than the
one between MNsure and the Office of MN.IT Services required in statute regarding coordination
of public health care programs.
52
modification. In our view, a clearly defined governance structure is necessary to

manage changes to the enrollment system and hold state officials accountable for
that systems performance.
Recognizing this need, administrators from MNsure, DHS, and the Office of
MN.IT Services jointly established in late 2014 a governance structure for
information technology decisions related to the health insurance exchanges
enrollment system. There is now a decision-making body (the Executive
Steering Committee) that includes representatives of each of these agencies and
is chaired by a DHS deputy commissioner. There are other governance bodies in
this structure, and there is also a Project Management Office staffed by the Office
of MN.IT Services.17
We think that the creation of this structure was a worthy attempt to ensure more
coordinated decision making on exchange-related information technology issues.
But, in our view, the underpinnings of the current structure are inadequate.
The multiagency governing structure that began overseeing MNsures enrollment system in
late 2014 has no formal authority.
This information technology governance structure is not established in state law,
so there is no assurance that it will continue as agency leadership changes. The
affected agencies did not adopt bylaws, policies, or interagency agreements when
they created this structure for MNsures enrollment system. As a result, the
scope of this structures authority and the duties of its bodies are not clearly
specified.
It is important to note that DHS has expressed concerns about the adequacy of
this structure for overseeing the MNsure enrollment system. Individuals in
public programs account for most of the systems enrollees, and DHS thinks that
it should have legally defined authority and play a large role (along with
counties) in the structure. DHS relies on the MNsure enrollment system for
proper eligibility decisions for public health care programs, and DHS is held
accountable for these decisions by both the federal and state governments.
RECOMMENDATION
The Legislature should establish in state law a structure for governing MNsures online
enrollment system.
At a minimum, this structure should include representatives from MNsure, DHS,
the Office of MN.IT Services, and counties. We offer no recommendations on
the exact make-up of this structure, but the Legislature should carefully consider
its composition and leadership. Even if the Legislature adopts a structure that
closely resembles the one informally implemented by executive branch agencies
17
The other governance bodies include a Project Management Team and Change Control
Board, each comprised of MNsure, DHS, and Office of MN.IT Services staff. As of late 2014,
there were no written policies outlining the functions of the various governance components.
GOVERNANCE
53
in 2014, there will be value in formalizing such an arrangement in law. This

structure will make key decisions about information technology issues, and how
to deploy resources to address them.
COMMUNICATIONS ISSUES
The development of a state-based health insurance exchange was a large
undertaking, with many risks. For this reason, it was important for exchange
staff to effectively convey information about the projects status to state leaders
and project stakeholders.
Before the first open enrollment period, MNsure staff withheld key information from the
board and other state officials.
Key officials told us they did not know until shortly before October 1, 2013, (or
later, for some officials) about many of the system difficulties that staff were
struggling to address. For example, the MNsure Board chair and the
Commissioner of the Department of Management and Budget told us they do not
recall being shown the Independent Verification and Validation reports that
raised concerns about the exchanges readiness. (We discussed these reports in
Chapter 2. MNsure hired a contractor to prepare these reports for the federal
government on system readiness.) Some MNsure Board members said that the
information on exchange operations they received from staff in the early months
following their appointment was not very detailed; one said that staff responses
to deep questions posed by the board tended to be brief. The board was
unaware of some key staff decisionssuch as the decision not to contract with
an overflow call center before the online enrollment system started.18
In public meetings just prior to October 1, MNsures executive director did not
emphasize the systems remaining risks or incomplete testing. There was no
mention in these meetings of contingency plans, in case the system did not work
as intended. When asked by the co-chair of the MNsure Legislative Oversight
Committee if the system was ready, MNsures executive director said she was
aware of no smoking guns that would argue for deferring the start of online
enrollment within the next week.19 The next day, when asked by the chair of the
MNsure Board whether the system was ready, the executive director said: At
this point, we feel that weve mitigated all of the areas that weve been tracking,
and that we will be ready to go.20 People we talked with said the executive
director usually presented a fairly positive view of the systems readiness in
public and private discussions. Board members did not believe they received
18
In Summer 2013, MNsure sought an overflow call center to assist MNsures contact center if
there was a large volume of calls. However, MNsure staff decided not to enter into a contract. As
we discuss in Chapter 6, MNsures contact center was overwhelmed by calls in Fall 2013.
19
MNsure Executive Director April Todd-Malmlov, comments to MNsure Legislative Oversight
Committee (September 24, 2013). She also said staff would seriously consider any issues that
came to light in their final reviews of the system.
20
MNsure Executive Director April Todd-Malmlov, comments to MNsure Board (September 25,
2013).
54
enough information about system readiness, and some felt misled by MNsure
staff leadership.
Officials in the Governors Office told us they were surprised to learn in the
weeks leading up to October 1 that there were significant issues that could
threaten the new enrollment systems functionality. They said that, before the
Governor was told in September 2013 of possible technical problems with the
website, the exchanges executive director provided limited information about
the exchanges progress. An official in the Governors Office said the
exchanges statutory governance structurewhich gave the Governor less direct
authority over the exchange than he had for other state agenciesmay have
contributed to MNsures weak communication with the Governors Office.
We also heard concerns from key stakeholders about communication during the
exchanges planning stages. For example, insurers said exchange staff were not
always forthcoming about known problems and did not listen enough to the
advice insurers offered.21 Likewise, county officials said the functionality of
MNsures new enrollment system was more limited than what they were
expecting, and some said their concerns had not received sufficient attention
from MNsure and DHS.
MNsure leaders made greater efforts in 2014 to keep the MNsure Board and key
stakeholders apprised of exchange developments than MNsure leaders did in 2013.
This largely reflected the change in MNsures executive director in December
2013. Many people praised the intelligence and hard work of MNsures first
executive director but said she did not always communicate effectively with
others about the exchange. Her replacement received better marks from people
we talked with for soliciting input from various groups and trying to make
MNsures activities more transparent.
Counties expressed continued frustration to us well into 2014 that MNsure and
DHS were not giving sufficient attention to their concerns about the exchanges
functionality.22 However, MNsure, DHS, and the Office of MN.IT Services
included a county representative on the Executive Steering Committee of the
information technology project governance structure they recently established,
which was a useful step.
21
While insurers expressed concern about their communication with exchange staff at MNsure and
its predecessor agency (the Department of Management and Budget), they expressed no significant
concern about communicationsbefore or after the enrollment system openedwith the
departments of Commerce and Health, which also perform exchange-related tasks.
22
During planning for the MNsure enrollment system and after its implementation, counties had
opportunities to discuss the system at DHSs County Roles and Responsibilities Workgroup. In
early 2014, DHS hired a MNsure Implementation Director who had extensive experience working
with counties.
GOVERNANCE
55
GENERAL GOVERNANCE RECOMMENDATIONS

The Legislature established MNsure as a state agency, but it gave MNsure a
different structure than most departments of state government. First, MNsure is,
by law, governed by a part-time board rather than a commissioner. When a state
agency is governed by a multi-member board, the agencys accountability is
more diffuse than it would be if it were governed by a single administrator.
Second, MNsures chief executive officer is appointed by the MNsure Board and
not the governor. While the board has statutory authority to govern MNsure, the
staff have played a large role in MNsures development and ongoing
operationsperhaps a larger role than the board. However, the accountability of
staff to the governor is indirect, occurring only through the selection of the chief
executive officer by the governor-appointed board. In our view, the Legislature
should amend MNsures governance structure.
RECOMMENDATION
The Legislature should amend state law to give the governor, rather than the MNsure Board,
authority to appoint the MNsure chief executive officer.
In our view, an agency with the importance and visibility of MNsure should be
directly accountable to the governor. At a minimum, we recommend that the
Legislature make MNsures top administrative position one that is appointed by
the governor. MNsure is an agency with widespread public impacts, and some of
MNsures problems in its short history have had profound impacts on consumers
trying to apply for health insurance. We think the staff leader of MNsure should
be accountable to the states highest elected official and not just to an unelected
board. Establishing this direct line of accountability might also promote better
communication between the governor and MNsure.
OPTIONS FOR CONSIDERATION

The Legislature should consider whether to retain the MNsure Board as a governing body or to
make it advisory.
If the Legislature makes the chief executive officer position directly accountable
to the governor, the Legislature should also re-evaluate the status of the MNsure
Board. Exhibit 3.1 discusses two options.
One option would be for the MNsure Board to serve in a purely advisory capacity
to MNsures top administrator. This approach would maintain a mechanism for
MNsure to receive public input, while giving greater authority to a MNsure
administrator who is directly accountable to the governor. This optionvesting
authority in an administrative appointee rather than a boardwould also be
56
Exhibit 3.1: Reasons for Changing or Keeping the

MNsure Boards Role
Reasons to Change the MNsure Board
to Serve as an Advisory Role, Not a
Governing Role (Option 1)
Most state departments are governed

by individual administrators, not
governing boards. This is because
accountability is more diffuse and less
direct when governance rests with a
multi-member board rather than one
administrator.
Administrative staff have direct, daily

involvement in issues that may be less
familiar to members of a part-time
board.
For issues requiring swift action, an

agency administrator can usually act
more quickly than a part-time board.
It may be more challenging to recruit

strong board members in the future.
Under state law, MNsure Board
members now receive a salary;
starting in 2016, they will receive more
a
limited compensation.
Ibid., subd. 2.
Reasons to Maintain the MNsure Boards

Role as a Governing Body (Option 2)
MNsure policy decisions should be

made by individuals who understand
the needs of key users (consumers
and small employers) or have
specialized expertise (for example, in
health care purchasing or health care
delivery systems). State law requires
MNsure Board members to meet such
requirements;b there are no specific
statutory requirements for MNsures
top administrator.
Board members may provide a more

independent perspective regarding
policy issues and agency oversight
than an administrator who is an
agency employee.
A governing board that holds public

meetings may provide transparency
that would be difficult to achieve if
governing authority rested with
MNsure administrative staff.
consistent with the governance structure of most of Minnesotas executive branch

departments.
A second option would be to have both a governor-appointed MNsure
administrator and a governor-appointed governing board. This model is used by
the Minnesota Pollution Control Agency (MPCA). The MPCAs commissioner
directs that agencys day-to-day administrative activities and is a member of the
governors cabinet.23 The board consists of eight members appointed by the
governor with the advice and consent of the Senate, plus the MPCA
commissioner. The MPCA board sets policy and agency direction, and it makes
final decisions on important or controversial issues. Having a governing board
may help to ensure that an agency makes decisions through a public, deliberative
process.
We offer no recommendation about which option to adopt. As described earlier,
the MNsure Board did not yet have formal authority at the time when most key
decisions related to the 2013 launch of the exchange occurred. Thus, within
23
Minnesota Statutes 2014, 116.02, subd. 1.
GOVERNANCE
57
MNsure, primary responsibility for MNsures shortcomings in the initial

enrollment period rested with staff, not the board. Since the launch, the board
has exercised more active oversight of MNsure. In our view, however, it would
make sense to reconsider the boards role if the MNsure chief executive officer
becomes a governor-appointed position.
MNsures governing officialswhether the top administrator, the board, or bothwill face
important challenges in coming months and years.
First, hundreds of thousands of persons currently enrolled in public health care
programs may start using MNsure to re-enroll in 2015. This transition has been
delayed several times already, due to the problems with MNsures enrollment
system. It remains to be seen whether MNsures technology can accommodate
this influx of enrollees, and whether its enrollment system can accurately
determine eligibility.
Second, MNsures future financial viability will depend partly on the number of
people who choose to enroll in commercial products through MNsure.
Expenditures from MNsures large federal grants are expected to decline. Thus,
MNsures expenditures will increasingly be paid by a withhold of a portion of
premiums for insurance products sold through MNsure. MNsures number of
2014 enrollments in commercial products was lower than the target set by staff in
October 2013, and in late 2014 the MNsure Board revised its budget projections
based on new, less ambitious assumptions regarding the number of future
enrollees in commercial programs.
Third, the MNsure Board must decide whetherand howto restrict which
products are sold through MNsure. Through 2015, insurers participating in
MNsure have been allowed to sell any products that meet the certification
requirements of the Affordable Care Act. However, starting in 2015, state law
authorizes the MNsure Board to determine which products may be offered
through MNsure.24
Fourth, MNsure must improve the consumer enrollment experience. Although
various fixes to some of the problems with the MNsure website and customer
service center have already been implemented, MNsure officials told us the
consumer experience needs additional attention. This may include enhancements
to MNsures own technology and customer service strategies.25
Fifth, the MNsure Board is considering whether there is a need for changes in
MNsures relationship with the Department of Human Services (DHS). A board
work group recently identified two options for organizing the exchange. On the
one hand, it said, MNsure could continue to be the lead agency for enrolling
24
25
Minnesota Statutes 2014, 62V.05, subd. 5(e).
Also, MNsure could consider ways to use private health insurance exchanges or web-based
brokers, which sell insurance from multiple insurers through their websites, in the enrollment
process. Federal regulations authorize private exchanges and web-based brokers to provide
enrollment assistance, although exchanges still must perform eligibility determinations before
people can enroll in qualified health plans.
58
individuals into public health care programs and qualified health plans.
Alternatively, the work group said, DHS could become the lead agency for
enrolling individuals into public programs. The latter option would be a
fundamental governance modification and would require statutory changes. As
of late 2014, these options were still in the early stages of board consideration.
If the Legislature retains the MNsure Board in either a governing or advisory
capacity, it may wish to consider whether there is any need to change the boards
composition. We offer no recommendation on the proper composition of the
board. There was legislative debate in 2013 about whether health care industry
professionals (such as people working for insurers) should be allowed to serve on
the board. This remains a reasonable topic for discussion. Current MNsure law
prohibits certain categories of people from serving on the MNsure Board.26
Some people maintain that such a prohibition is unnecessary because state laws
regarding conflict of interest provide sufficient protection against board members
making decisions based on their economic interests. For example, the statutes
governing the Minnesota Pollution Control Agency do not prohibit
representatives of regulated businesses from serving on the MPCA Board, but
members of this board are required by state rules to recuse themselves from
certain decisions in which they have a financial or employment interest.27 In
addition, some people think there would be value in having knowledgeable
people from the health care industry serving on the MNsure Board.
26
Minnesota Statutes 2014, 62V.04, subd. 4(a), states: Within one year prior to or at any time
during their appointed term, board membersshall not be employed by, be a member of the board
of directors of, or otherwise be a representative of a health carrier, institutional health care provider
or other entity providing health care, navigator, insurance producer, or other entity in the business
of selling items or services of significant value to or through MNsure.
27
Minnesota Statutes 2014, 116.02, subd. 3. The law says that no MPCA Board member other than
the commissioner shall be an officer or employee of the state or federal government. The law
limits to two the number of members who are officials or employees of a municipality or
governmental subdivision; neither may be a member ex officio or otherwise on the management
board of a municipal sanitary sewage disposal system. Minnesota Rules, 7000.9000, subp. 1,
posted April 20, 2004, established a policy for MPCA members regarding conflict of interest.
Chapter 4: Enrollment
M
Nsures success depends on its ability to enroll people in health insurance.

MNsures enrollment system has three primary responsibilities: (1) to
determine consumers eligibility for the public health insurance programs
administered by the Department of Human Serviceschiefly Medical Assistance
and MinnesotaCare; (2) to serve as a marketplace that facilitates the comparison
and purchase of certain commercial health insurance products, known as
qualified health plans; and (3) to determine consumers eligibility for federal tax
credits and cost-sharing reductions that lower the premiums and out-of-pocket
expenses associated with qualified health plans.1 In this chapter, we critically
examine MNsures reported enrollment, analyze the characteristics of enrollees,
and discuss what is known about the extent to which MNsure enrollees were
previously uninsured.
MNsure met its overall enrollment target for the first enrollment
period, but this target was seriously flawed due to a Department
of Human Services error that significantly underestimated
Medical Assistance enrollment.
Survey results showed that 28 percent of individuals who

enrolled in commercial insurance through MNsure were
uninsured immediately before they enrolled.
MNsures data reporting capabilities are weak, limiting its ability

to produce information for management and decision-making
purposes.
TOTAL ENROLLMENT
Before discussing MNsures total number of enrollments, it is important to
discuss how MNsure counts enrollments for reporting purposes. We think there
are several distinctions worth making.
First, what it means to be enrolled by MNsure depends on whether one obtains
public or commercial insurance. For public programs, MNsure considers a
person to be enrolled from the point at which the person is deemed eligible for
that program. For qualified health plans (that is, commercial insurance), MNsure
considers individuals to be enrolled when they select an insurance product.
MNsure also sells health insurance for employees of small businesses and stand-alone dental
insurance. These types of insurance are not discussed in this chapter.
60
Second, the definition of enrollment in a qualified health plan is different in 2015

than it was in 2014. In 2014, individuals were given several options for payment
methods, and they had to select a methodin addition to selecting an insurance
productto be considered enrolled. In 2015, individuals no longer have a choice
of payment method, so they are considered enrolled as soon as they select an
insurance product.
Third, MNsures definition of enrolled is different from an insurers definition
of enrolled. An insurer considers people who selected qualified health plans
through MNsure to be enrolled only when they pay their first months premiums.
MNsures enrollment data (unlike the enrollment data of the insurers) do not take
into account whether payments have been made.
Between October 1, 2013, and November 11, 2014, MNsure reported that it processed about
371,000 enrollments in health insurance for 2014 coverage.
The total number of reported enrollments included 234,751 enrollments in
Medical Assistance; 80,387 enrollments in MinnesotaCare; and 55,900
enrollments in qualified health plans.
MNsures publicly reported total of enrollments is a cumulative measure; it does
not reflect any attrition that occurred. Although qualified health plan enrollees
are supposed to notify MNsure if they terminate their insurance coverage, such
communication does not always occur. In our September 2014 survey of a
sample of enrollees in qualified health plans, 13 percent said that they had
terminated or never received their coverage by the time of our survey.2 This is
generally consistent with a statement made by MNsures chief executive officer
at a MNsure Board meeting in September 2014. He said that the actual number
of MNsure-enrolled qualified health plan enrollees as of that month was about
85 percent of MNsures reported cumulative enrollment figure, according to data
from the insurers.3
The attrition rate for Medical Assistance appears to be much lower than the
attrition rate for commercial products. Data from the Minnesota Department of
Human Services indicate that there was an average of 242 disenrollments of
MNsure-processed Medical Assistance enrollees per month between February
and August 2014.4 This would suggest an attrition rate of about 0.9 percent of
See the Introduction for a description of the survey. Some MNsure enrollees may never have
actually obtained insurance coverage, although we were unable to determine the extent of this. In
our survey of enrollees, 8 respondents (out of 281) indicated in written comments that they had
never received insurance coverage through MNsure. Because respondents did not provide this
information in response to a direct question about ever receiving insurance coverage from MNsure,
there may have been others we surveyed who never obtained coverage.
3
Thus, about 47,500 persons would have been enrolled in qualified health plans as of Fall 2014,
compared with the cumulative 55,900 qualified health plan enrollments reported by MNsure.
A disenrollment was defined for these purposes as someone who was eligible for Medical
Assistance in the previous month but not eligible for Medical Assistance in the current month (for
example, eligible in January 2014 but not in February 2014).
ENROLLMENT
61
MNsures reported Medical Assistance enrollments over the course of MNsures

first year.5
Access to Management Information

In the early stages of our evaluation, we talked with staff from MNsure and the
Office of MN.IT Services about how we could obtain enrollment data for
analyses we wanted to conduct. We learned that the ability of staff to access and
extract data from the MNsure system was limited.
The MNsure enrollment system lacks good reporting capabilities, making it difficult for
MNsure to extract data needed for management and decision-making purposes.
During the first year of enrollment, MNsures enrollment system contained
various personal identifierssuch as Social Security numbers and something
called the MNsure IDbut no single identifier provided a consistent, reliable
basis for tracking consumers end-to-end in the enrollment process. The data
were stored in nearly 3,000 tables, making it very difficult for analysts to know
where to look to find any particular data field. Staff were still learning the
system components and at times felt completely dependent on the information
technology vendors to understand the system.
Deficiencies in MNsures data reporting capabilities created additional work for
insurers with respect to federal reporting of enrollees receipt of tax credits. The
federal Centers for Medicare and Medicaid Services (CMS) requires aggregated
monthly reporting of the total premiums owed and the total tax credit amount
selected by all individuals enrolled in each insurance product sold through
MNsure. CMS uses the reports to make payments to insurers for those enrollees
receiving tax credits. CMS requires that this reporting be done either by MNsure
or by the insurers, but it must be done by only one or the other. MNsure
generated reports for CMS in December 2013 and January 2014. However,
MNsure found these reports very challenging to produce. As a result, the
subsidy reporting responsibility was shifted to the insurerswith their assent
making this task one of several that insurers have had to complete that they
originally expected to be done by MNsure.
Staff from the Office of MN.IT Services have been unable to generate
repeatable reports from MNsures enrollment data, making the data retrieval
process time intensive. One solution to this problem would be to develop an
electronic data warehouse, a storage system that facilitates efficient and
intuitive data retrieval. Information technology staff told us that, in many
systems, there is an advantage to building a data warehouse after the system is
actually operating. In mid-2014, a MNsure official told us that it was MNsures
intention to construct such a warehouse over a three- to five-year period.
However, we were also told that the eligibility software that MNsure selected for
MNsure reports that there were 196,027 cumulative Medical Assistance enrollments as of
September 8, 2014. There were 1,695 MNsure-processed Medical Assistance disenrollments by
the end of August 2014 not accounted for in the cumulative enrollment number.
62
its enrollment system is dependent on having a data warehouse to translate the

data and make it accessible, which suggests there should be some urgency in
addressing that issue.
RECOMMENDATION
MNsure should develop ways to improve its access to the applicant and enrollee data it
collectsfor the purpose of assessing MNsure performance, generating management reports,
and responding to public inquiries.
We recognize that development of a data warehouse has not been MNsures top
priority, given the need to address the basic functionality of the MNsure
enrollment process. But, for management and accountability purposes, it is
important for MNsure to more readily access the information it collects.
Until late 2014, MNsure did not have what information technology experts call a
system of record in its enrollment data that would have provided
comprehensive and authoritative documentation of consumers ultimate
enrollment choices. In late 2014, the Office of MN.IT Services developed a
system of record within the MNsure enrollment system. As of early January
2015, however, MNsure was not able to transmit electronic enrollment records of
acceptable quality to insurers in the expected format (known as an 834EDI file).
Such files normally require little manual processing by insurers, but the files they
received from MNsure were delayed by weeks and contained what insurers
believed to be obvious errors and duplicated enrollment data. To ensure
accuracy, insurers had to rely on supplemental data provided by MNsure in an
alternate format (known as an 834ST file), which required extensive manual
processing. The recent development of a system of record within MNsures
enrollment system should help MNsure to track the records associated with
individual consumers who have applied or enrolled through MNsure. We
suggest that MNsure develop additional improvements in its ability to extract and
analyze applicant and enrollee data.
Duplicate records in MNsures data have made it more difficult for MNsure to accurately
report on enrollment.
While analyzing MNsures enrollment data, we saw many instances of duplicate
records. Some appeared to reflect real duplications of individual enrollments,
while others did not. For example, some individuals went through the
application and enrollment process more than once because of technical problems
they encountered. Also, some individuals submitted more than one application in
an effort to amend information submitted in previous applications. As a result,
some individuals actually had multiple records within MNsures data.
Information on dependents (spouses and children) sometimes appeared in
MNsure records as duplicates of the person who submitted the application. In
such cases, MNsure had to manually enroll the dependents because the
enrollment system had lost their enrollment information. Because of uncertainty
ENROLLMENT
63
about the nature of duplicate records in MNsures data, we had to make certain
assumptions when analyzing enrollee characteristics.6
Some of the duplicate records hindered MNsures ability to publicly report on its
enrollments. For instance, some duplicate enrollments indicated that an
individual had selected more than one insurance product, perhaps on the same
date. Thus, for purposes of reporting the products individuals purchased,
MNsure had to make an effort to deduce which product the person ultimately
selected. MNsure often used payment data to make this determination, but
sometimes even these data did not enable MNsure to determine this conclusively.
In these cases, MNsure relied on communications with the insurers or the
consumers to determine which product the individuals actually purchased.
ENROLLEE CHARACTERISTICS
We used enrollee data obtained from the Office of MN.IT Services, MNsure, and
the Department of Human Services to examine some characteristics of enrollees,
such as where they lived, whether they received subsidies, and their age and
gender. We did not examine the racial and ethnic characteristics of enrollees
because enrollees were not required to provide race/ethnicity data and often did
not.7 Thus, although MNsure set enrollment goals for racial/ethnic groups and
developed strategies to achieve these goals, MNsure cannot (and we could not)
evaluate MNsures success in this area.8 Exhibit 4.1 shows those characteristics
that we could reliably measure for most enrollees based on the available data.
Enrollments by Region
For each region of Minnesota, we examined the number of MNsure enrollees
(ages 18 to 64) as a percentage of the number of uninsured people (ages 18 to 64)
If an individual had multiple enrollment records, we looked only at the record corresponding to
the latest date of enrollment. If there were multiple enrollments on the latest date of enrollment, we
chose any one of that persons records to examine, but treated any information that differed across
records from that same date as unknown. For example, if someone enrolled with more than one
insurer on the same last date, we treated that persons insurer as unknown.
7
The application questions on race and ethnicity allowed respondents to write in a response. Those
responses included many that were not amenable to standard categorization. The written responses
further revealed that many people experienced website difficulties with the race question.
Specifically, some users commented that the box for Japanese was already checked for them, and
those users who wished to uncheck the box were not able to do so.
As discussed below, the overall enrollment projection was seriously flawed. Any derivative
projectionssuch as those for race and region of the statewere likewise flawed.
64
Exhibit 4.1: Characteristics of MNsure Enrollees

(October 1, 2013, to June 30, 2014)
Attribute
Gender
Female
Male
Age (years)
0-17
18-24
25-34
35-44
45-54
55-64
65+
Region (Rating Area)a
South East (1)
North East (2)
South Central (3)
South West (4)
Mid Central (5)
West Central (6)
North Central (7)
Twin Cities (8)
North West (9)
Metal Level
Platinum
Gold
Silver
Bronze
Catastrophic
Unknown
Insurer
Blue Cross Blue Shield
HealthPartners
Medica
PreferredOne
UCare
Unknown
Total Enrollment
Qualified Health Plan
Medical Assistance
MinnesotaCare
51.3%
48.7
100.0%
51.3%
48.7
100.0%
53.3%
46.7
100.0%
9.7%
5.8
19.1
14.9
19.3
31.0
0.3
100.0%
35.7%
13.7
20.8
10.8
10.3
8.6
0.1
100.0%
2.4%
11.5
27.4
17.0
18.6
20.7
2.4
100.0%
6.5%
6.0
3.8
1.7
3.4
3.6
8.0
66.0
1.0
100.0%
7.2%
5.9
4.5
2.0
3.8
4.1
9.0
62.3
1.3
100.0%
6.7%
6.3
4.0
1.8
3.6
3.9
8.9
63.6
1.2
100.0%
26.8%
12.1
34.2
25.3
0.9
0.8
100.0%
22.5%
12.0
4.6
58.9
1.4
0.6
100.0%
54,026
152,671
56,654
NOTES: Our total enrollments may differ from what MNsure had reported as of July 10, 2014, due to
differences in when the data were extracted and in how we removed duplicate enrollment records.
Duplicate records are instances when two or more observations in the data have the same identifying
information. Unknown categories represent individuals for whom the data contained duplicate
records from the same (most recent) enrollment date that did not all report the same metal level or
insurer.
a
The counties in each rating area are shown in Appendix C.
SOURCE: Office of the Legislative Auditor, analysis of MNsure enrollment data provided by the
Office of MN.IT Services.
ENROLLMENT
65
in that region. 9 Appendix C shows a list of counties in each region. Our analysis
was not intended to indicate how many uninsured people enrolled in MNsure.
Rather, we used the number of uninsured people as a rough measure of the
relative size of MNsures potential customer pool in each region. Results are
shown in Exhibit 4.2.
Exhibit 4.2: Enrollment of Nonelderly Adults as a Share of the Uninsured

Population by Region (October 1, 2013, to June 30, 2014)
Region/Rating Area
South East (1)
North East (2)
South Central (3)
South West (4)
Mid Central (5)
West Central (6)
North Central (7)
Twin Cities (8)
North West (9)
Margin of Error
Uninsurance
Rate,
Overall
Uninsurance
Rate,
Ages 18-64
Overall
Enrollmenta
Qualified
Health Plan
b
Enrollment
Medical
Assistance
c
Enrollment
7.9%
8.6
7.9
8.7
8.6
8.0
10.3
8.5
9.6
10.7%
11.4
10.1
12.3
11.4
10.9
13.9
11.0
12.9
48.7%
56.9
51.2
42.3
50.3
52.5
50.3
54.7
36.4
20.2%
24.8
21.5
16.6
17.4
19.1
18.3
23.3
11.0
58.2%
63.1
54.5
49.3
66.0
67.5
61.8
66.5
52.3
53.8%
68.9
61.3
50.4
59.9
67.3
63.1
65.1
45.0
0.1 - 0.6
0.2 - 0.8
0.9 - 2.7
0.6 - 1.7
2.2 - 7.1
3.0 - 9.3
MinnesotaCare
Enrollmentd
NOTES: These calculations were derived from aggregated, county-level data, making it difficult to estimate a precise standard error. We
show the approximate range of margins of error for each column to indicate that there is some degree of uncertainty around the
estimates.
a
Overall enrollment is expressed as the ratio of the number of people in the region ages 18 to 64 enrolled in any plan or program through
MNsure to the number of uninsured people ages 18 to 64 in that region.
Qualified health plan enrollment is expressed as the ratio of the number of people in the region, ages 18 to 64, enrolled in a qualified
health plan through MNsure to the number of uninsured people of any age in that region with incomes at least 200 percent of the federal
poverty level. For a single adult with no dependent children, 200 percent of the federal poverty level was equal to $23,340 in annual
income in 2014.
Medical Assistance enrollment is expressed as the ratio of the number of people in the region, ages 18 to 64, enrolled in Medical
Assistance through MNsure to the number of uninsured people of any age in that region with incomes between 0 and 138 percent of the
federal poverty level. For a single adult with no dependent children, 0 to 138 percent of the federal poverty level was equal to $0 to
$16,105 in annual income in 2014.
MinnesotaCare enrollment is expressed as the ratio of the number of people in the region, ages 18 to 64, enrolled in MinnesotaCare
through MNsure to the number of uninsured people of any age in that region with incomes between 139 and 200 percent of the federal
poverty level. For a single adult with no dependent children, 139 to 200 percent of the federal poverty level was equal to $16,106 to
$23,340 in annual income in 2014.
SOURCES: Office of the Legislative Auditor, analysis of MNsure enrollment data and uninsurance rates from the 2009-2013 American
Community Survey.
The regions are the nine rating areas as defined by the departments of Commerce and Health.
This analysis included all public program and qualified health plan enrollees who were ages 18 to
64 as of June 30, 2014. We used county-level data on the uninsured population from the
2009-2013 American Community Survey (ACS); we aggregated those data into regional-level data.
The ACS is conducted annually by the U.S. Census Bureau and is representative of the civilian,
noninstitutionalized U.S. population. Use of the five-year pooled sample achieves a sample size
that permits estimation of the number of uninsured people for all Minnesota counties.
66
The number of MNsure enrollees relative to the size of MNsures potential pool of customers
was lowest in northwest and southwest Minnesota.
For example, the number of MNsure enrollees in northeast Minnesota was
57 percent of the number of uninsured people in that part of the state. In contrast,
the number of MNsure enrollees in northwest Minnesota was about 36 percent
of the number of uninsured people in that part of the state. Similarly, the number
of Medical Assistance enrollees in the Twin Cities metropolitan area was
67 percent of the number of uninsured people with incomes below 138 percent of
the federal poverty level in that part of the state. In contrast, the number of
Medical Assistance enrollees in northwest Minnesota was about 52 percent of the
number of uninsured people with incomes below 138 percent of the federal poverty
level in that part of the state, and in southwest Minnesota it was about 49 percent.
Receipt of Subsidies
We also examined the extent to which qualified health plan enrollees obtained
subsidies through MNsure. As described in Chapter 1, people who enrolled
through MNsure in a qualified health plan may have been eligible for (1) an
advanced premium tax credit (APTC) or (2) a cost-sharing reduction (CSR).
These subsidies help to lower premium costs or out-of-pocket health care
expenses.10
Qualified health plan enrollees can be divided into two categories: those who
requested that MNsure check their eligibility for subsidies, and those who did
not. We estimate that the former group, which includes both people who are
determined eligible for a subsidy and those who are determined ineligible for a
subsidy, constituted approximately 82 percent of qualified health plan enrollees.
We estimate that 18 percent of qualified health plan enrollees chose not to check
their eligibility for subsidies while enrolling.11 The extent to which these persons
would have qualified for subsidies had they completed the eligibility
determination process is unknown.
Our analysis found that at least 41 percent of qualified health plan enrollees received the
advanced premium tax credit, and 13 percent received a cost-sharing reduction.
MNsure reported that 45 percent of current qualified health plan enrollees
received the advanced premium tax credit and 14 percent received cost-sharing
reductions, as of August 2014. This was based on data MNsure obtained from
health insurers, and it reflected persons enrolled at the time the data were
collected. We made our own estimate, using data on individuals who enrolled in
a qualified health plan at any time from October 2013 through June 2014. Our
10
The availability of tax credits was one of the most-cited reasons by survey respondents when
asked why they purchased health insurance through MNsure rather than directly from an insurer.
That response was given by 37 percent of survey respondents.
11
Consumers had the option to answer additional application questions to determine their eligibility
for subsidies and public insurance programs, or not.
ENROLLMENT
67
analysis indicated that 41 percent of individuals who were ever enrolled in a

qualified health plan received an APTC, and 13 percent received a CSR. It is
worth noting that some enrollees qualify for both an APTC and a CSR.
Both MNsures and our estimates were based on those who qualified for a tax
credit greater than $0 and who chose to take at least some of that tax credit as a
monthly reduction in premiums. Neither estimate included those enrollees who
qualified for the tax credit but chose to get a refund on their taxes rather than
apply the credit to monthly premiums. MNsures enrollment data did not allow
us to determine how often this occurred. As mentioned earlier in this chapter,
MNsure depends on health insurers to accurately report APTC eligibility and
amounts to the federal Centers for Medicare and Medicaid Services, but the
insurers do not have data on persons who will be receiving their tax credit as a
tax refund.12 Therefore, both MNsures and our estimates of tax credit usage are
likely to be undercounts.
MNsures enrollment data do not allow for a complete determination of all
individuals who were eligible for a tax credit. In the MNsure data we examined,
it was only possible to see the amount of tax credit that a consumer selected to
receive as a monthly reduction in premiums, which was not necessarily the full
amount that the person was qualified to receive.13 All qualified health plan
enrollees who did not select an amount greater than $0 appeared in the data as
having selected a tax credit of $0. Therefore, the following groups were not
differentiated in the data: (1) those who chose not to have their eligibility for a
tax credit determined; (2) those who were found ineligible for the tax credit; and
(3) those who were found eligible for the tax credit but chose to defer it as a tax
refund.14 Exhibit 4.3 shows the median selected tax credit by age, gender, and
other characteristics among those enrollees who selected a tax credit greater than
$0.
ENROLLMENT PROJECTIONS
It was challenging for MNsure to project participation in the health insurance
exchange, given that the state had no prior experience running an exchange.
MNsure contracted with experts to help it project participation levels, and it
refined projections over time. Most of MNsures enrollment projections were for
2016; our analysis focused on MNsures projections for 2014.
12
The insurers do not need such data because the consumer is paying the entirety of his or her
premium.
13
When consumers were deemed eligible to receive a tax credit, they were shown the maximum
amount of tax credit for which they qualified. They then had to select how much of that tax credit
they wanted to apply toward their monthly premiums; they could choose any amount from $0 up to
the maximum.
14
Another group that is shown in the data as having a selected tax credit of $0 is persons who met
the eligibility requirements for the tax credit but received a maximum credit of $0 because the
benchmark plan in their service area costs less than their required contribution. The benchmark
plan is the second-lowest-cost silver plan in a given service area.
68
Exhibit 4.3: Median Monthly Amount that MNsure

Enrollees Selected for Advanced Premium Tax Credit
(October 1, 2013, to June 30, 2014)
Head of Household Attribute
Median Advanced
Premium Tax Credit
Gender
Female
Male
$146
154
Age (years)
0-17
18-24
25-34
35-44
45-54
55-64
65+
$100
69
41
57
118
199
213
Region (Rating Area)a

South East (1)
North East (2)
South Central (3)
South West (4)
Mid Central (5)
West Central (6)
North Central (7)
Twin Cities (8)
North West (9)
$344
195
201
201
166
154
161
116
138
Metal Levelb
Platinum
Gold
Silver
Bronze
$137
146
184
122
Insurer
Blue Cross Blue Shield
HealthPartners
Medica
PreferredOne
UCare
$208
117
256
132
111
Overall Median
$154
NOTES: The data in this exhibit are representative of qualified health plan enrollees who were
identified in their MNsure application as head of household. The median amounts presented here
refer only to heads of household who qualified for a tax credit and who chose to accept at least some
of that tax credit as a monthly reduction in premiums. (N=14,505)
a
The counties in each rating area are shown in Appendix C.
MNsures data showed that the median amount of a tax credit selected by people who enrolled in
catastrophic plans was $10. However, people who enrolled in a catastrophic plan were not eligible
for subsidies. It is unclear why the data showed that some of these people received subsidies.
SOURCE: Office of the Legislative Auditor, analysis of MNsure enrollment data provided by the
Office of MN.IT Services.
ENROLLMENT
69
During the first open enrollment period, MNsure enrolled more persons than it had
projected, but only because the projection for Medical Assistance enrollment was a flawed
underestimate.
MNsure staff presented 2014 enrollment projections to the MNsure Board in
October 2013. The projections were broken out by those that would occur during
the first open enrollment period (ending March 31, 2014) and those that would
occur during the remainder of calendar year 2014. In Exhibit 4.4, we compare
the projected and actual enrollments for the first open enrollment period.
Exhibit 4.4: Projected and Actual Enrollments for the

First Open Enrollment Period (October 1, 2013, to
March 31, 2014)
Type of Enrollee
Qualified Health Plans
SHOP (small business)
MinnesotaCare
Medical Assistance
TOTAL
Projection
Actual Enrollment
69,904
8,925
44,084
12,240
135,153
47,902
726
37,985
99,531
186,144
NOTES: Projections were presented to the MNsure Board on October 16, 2013. Actual enrollment
reports enrollments as of April 13, 2014. The figures for SHOP represent individuals, not businesses.
SOURCES: MNsure staff documents presented to MNsure Board, specifically 2014 Enrollment
Goals (St. Paul, October 16, 2013) and Enrollments by Program, April 13, 2014 (St. Paul, April 16,
2014).
The actual MNsure enrollment in Medical Assistance (MA) far exceeded

MNsures projected MA enrollment. That projection, along with the
MinnesotaCare projection, was prepared by the Department of Human Services
(DHSs) Reports and Forecasts Division in September 2013. In response to our
inquiry about the MA projection, DHS said that its projection was an error; it did
not include all of the new MA cases DHS intended to include. We think this
error should have been obvious since the projection of 12,240 new MA enrollees
over a six-month period was much lower than past enrollment numbers. In just a
single month (August 2013), DHS enrolled 23,178 new people in MA, nearly
twice as many as it projected for all of MNsures first open enrollment period.
DHS said it noticed its error in the spring of 2014. MNsures total enrollment
goal of 135,153 was never revised; MNsure continued to use the original
projection as a benchmark for its enrollment success. If a more realistic
projection of MA enrollment had been included in MNsures projections (at least
100,000), actual total enrollments through MNsure for the first open enrollment
period would have fallen far short of the overall projection, since other types of
enrollment were low.
MNsure projected large increases in its number of MA and MinnesotaCare
enrollees for the remainder of 2014 beyond open enrollment. Those projections
envisioned a large-scale eligibility redetermination and conversion of hundreds
of thousands of existing public program enrollees from DHSs existing
70
enrollment systems to MNsures system. However, as discussed in Chapter 2,

concerns over the MNsure systems lack of functionality led DHS to scrap that
original plan. Only a relatively small number of conversions occurred by the end
of 2014, with the rest now planned to be moved into either the MNsure
enrollment system or the existing DHS system (MAXIS) by the end of 2015.15
For the remainder of 2014 beyond open enrollment, MNsure projected to enroll
an additional 32,896 persons in qualified health plans, for a total of 102,800
qualified health plan enrollees. As of November 11, 2014, 55,900 persons had
enrolled in qualified health plans, or 54 percent of the 2014 target. In late 2014,
MNsure lowered its qualified health plan enrollment goals for the second open
enrollment period. MNsure now bases its fiscal year 2015 budget on a target of
67,000 qualified health plan enrollees, including 37,000 renewals of coverage
year 2014 enrollees and 30,000 new enrollees.
Gruber-Gorman Reports
In 2011, the Minnesota Department of Commerce hired Jonathan Gruber,
professor of economics at the Massachusetts Institute of Technology, and the
firm Gorman Actuarial to estimate the impact of the ACA and a state-based
health insurance exchange on Minnesota.16 Together, the amount of the contracts
was $560,000.
The Gruber-Gorman estimates of the Affordable Care Acts impacts on Minnesota were for
2016, making it premature to fully assess their accuracy.
Legislators and legislative staff have posed some questions about the accuracy of
the Gruber-Gorman projections. Appendix E compares the 2013 Gruber-Gorman
projections for 2016 with what is known as of early 2015. In some cases, the
projections appear to be quite different from what has actually occurred.
However, we think that a fair assessment of the projections cannot be made until
the full period of the projections has passed.
MNSURES IMPACT ON THE UNINSURED POPULATION

The State Health Access Data Assistance Center (SHADAC), a health policy
research center at the University of Minnesota, released a report in June 2014 that
examined aggregate shifts in health insurance coverage among Minnesotans
15
In November 2014, DHS started processing Medical Assistance enrollee renewals in MAXIS,
based on modified adjusted gross income. The Centers for Medicare and Medicaid Services has
approved this type of eligibility determination process for Minnesota and some other states.
16
Professor Gruber worked as a consultant for Massachusetts when that state implemented its
health reforms in 2006. Those reforms became a basis for the Affordable Care Act, and Professor
Gruber served as a consultant to the federal government during the crafting of that legislation.
ENROLLMENT
71
occurring between September 30, 2013, and May 1, 2014.17 The analysis used a
methodology that has been employed by the Minnesota Department of Health
since the early 1990s of taking the total state population and subtracting from it
the number of people known to have various types of insurance. After all types
of insurance have been taken into account, the remaining share of the population
is assumed to be uninsured.
The number of uninsured Minnesotans fell significantly after MNsure opened for business,
but the impact of MNsure on this reduction is unclear.
SHADAC determined that the number of uninsured Minnesotans decreased by
180,520 over the seven-month period of analysis. This represented a
40.6 percent decline in the uninsurance rate, which fell from 8.2 percent to
4.9 percent of the states population. As of May 1, 2014, SHADAC estimated
that 264,480 Minnesotans remained uninsured.
Most of the drop in the uninsurance rate occurred because Medical Assistance
enrollment grew by about one-third, from 11.5 percent of the states population
in September 2013 to 15.3 percent in May 2014.18 As discussed in Chapter 1, the
expansion of Medical Assistance, Minnesotas Medicaid program, was a state
policy decision authorized by the ACA. Gains in Medical Assistance enrollment
may have been due to the expanded eligibility criteria for the program, increased
efforts through MNsure to enroll individuals in public programs, the possibility
of a tax penalty for not complying with the ACAs mandate to purchase health
insurance, or a combination of these factors. The Minnesota Department of
Health (MDH) estimated that 67 percent of the states 2013 uninsured population
was eligible for a public health insurance program under the criteria that were in
effect at the time.19
MDH and SHADAC conduct a statewide survey every two years to determine,
among other things, the health insurance coverage status of Minnesotas
population. That survey, known as the Minnesota Health Access Survey, was
last conducted between August and November 2013, during the weeks leading up
to and the first few weeks of MNsures first open enrollment period. To get a
sense of how the distribution of insurance coverage had changed for some groups
since MNsure opened, MDH and SHADAC conducted a follow-up survey,
known as the Minnesota Health Insurance Transitions Study (MN-HITS), from
17
Julie Sonier, Elizabeth Lukanen, and Lynn Blewett, Early Impacts of the Affordable Care Act on
Health Insurance Coverage in Minnesota (Minneapolis: State Health Access Data Assistance
Center, University of Minnesota, June 2014).
18
Enrollment in MinnesotaCare fell by 43 percent during this time period. Combined with the
increased Medical Assistance enrollment, there was a net increase of 20.6 percent in the share of
the population enrolled in a state public health insurance program.
19
Health Insurance Coverage in Minnesota: Results from the 2013 Minnesota Health Access
Survey (St. Paul: Health Economics Program, Minnesota Department of Health, May 2014).
72
August to October 2014.20 The MN-HITS survey re-contacted persons who

participated in the 2013 Minnesota Health Access Survey and had reported at that
time that they were uninsured.21
The 2013 Minnesota Health Access Survey found that 8.2 percent of the states
populationabout 445,000 peoplewere uninsured. According to preliminary
findings from the MN-HITS, 50 percent of that group (about 222,500 people)
have since gained insurance coverage; the other 50 percent remain uninsured.
Among those who gained coverage, 53 percent (about 118,000 people) enrolled
in public coverage, including state programs, Medicare, and Veterans Assistance/
military coverage; 25 percent (about 55,000 people) enrolled in group coverage
through an employer; and 22 percent (about 49,000 people) enrolled in individual
market coverage. Among those who gained coverage, 44 percent (about 98,000
people) reported doing so through MNsure. The MN-HITS survey asked
respondents the reasons they gained (or lost) insurance coverage since the 2013
Minnesota Health Access Survey. Among those who had gained coverage,
38 percent reported that they had done so for ACA-related reasons. Such reasons
included complying with the laws mandate for individuals to have health
insurance and that MNsure made it easy to sign up for insurance coverage.
The MN-HITS was not able to determine the characteristics of all of the
remaining uninsured population in Minnesota, nor could it determine the types of
insurance previously held by the population who signed up for coverage through
MNsure. Answers to those questions will only be available through a large
population survey, such as the next Minnesota Health Access Survey, which is
scheduled to be conducted in the fall of 2015 with results available in early 2016.
In the meantime, we attempted to determine the prior insurance status of people
who gained coverage through MNsure by conducting a smaller-scale survey of
MNsure enrollees. Our survey sample was representative of adults ages 18 to 64,
who were listed in their MNsure applications as heads of households, and who
enrolled in qualified health plans between October 1, 2013, and June 30, 2014.22
That target population consists of 37,205 members. We distributed the survey in
September 2014 to a random sample, stratified by region of the state, of 1,000
20
Primary funding for the MN-HITS was provided by the Robert Wood Johnson Foundations
State Health Reform Assistance Network, with additional financial support provided by MDH. We
present preliminary findings here, with additional findings expected to be forthcoming in future
publications from MDH and SHADAC.
21
The MN-HITS also recontacted people who reported that they were covered by individual market
insurance or by the Minnesota Comprehensive Health Association (the states high-risk pool) at the
time of the 2013 Minnesota Health Access Survey. We do not report any results for those
populations. Only respondents who were ages 0 to 63 at the time of the 2013 Minnesota Health
Access Survey were recontacted for the MN-HITS so as not to include those who had aged into
Medicare.
22
We chose to limit the sample to nonelderly adults to eliminate, on one end of the age spectrum,
the need for proxy reporting of childrens responses by an adult in the household, and to avoid, on
the other end, sampling people who are now very likely to be enrolled in Medicare. We assumed
the person listed in the application as head of household or a self purchaser (rather than spouse
or dependent) was likely to be the person who had filled out the households application,
meaning that the person could report his or her direct experience. Moreover, by limiting the survey
to only one member of a household, we were able to reduce the response burden on survey
participants that would have resulted from having each person in the household respond separately.
ENROLLMENT
73
members of the target population. The survey had a response rate of

29 percent.23
According to our survey, just over one-quarter of
qualified health plan enrollees were uninsured
immediately before enrolling in coverage through
MNsure.
Survey Respondents Insurance Coverage

Immediately Before Purchasing a Qualified
Health Plan Through MNsure
Uninsured (28 percent)

Insurance purchased directly from an
insurer (27 percent)
Employer-sponsored insurance, including
coverage through a labor union and
continuation coverage through the
Consolidated Omnibus Budget
Reconciliation Act of 1985 (COBRA)
(24 percent)
Dependent coverage through a parent,
spouse, or domestic partner (10 percent)
Public insurance, including Medical
Assistance and MinnesotaCare
(5 percent)
Minnesota Comprehensive Health
Association (5 percent)
Did not know (1 percent)
Our survey found that 28 percent of

qualified health plan heads of
households were uninsured
immediately before enrolling in
insurance through MNsure.24 In a
similar national study that looked at

persons who enrolled in individual
market insurance through an

exchange, the percentage of
previously uninsured persons was

57 percent.25 The box on this page
shows the distribution of the types

of insurance coverage reported by our survey respondents.26
Many newly insured qualified health plan enrollees had been uninsured for two years or
more.
Our survey found that 23 percent had been uninsured for at least two but less
than five years; 29 percent had been uninsured for five years or more. The
remainder had been uninsured for less than two years.
23
There were 281 valid responses to the survey out of 975 valid mailings. To produce
representative estimates, survey results have been weighted to reflect both the probability of the
respondents selection into the sample, which differed by region, and the probability of responding
to the survey, based on age and gender.
24
Following the definition used by the National Center for Health Statistics, uninsured includes
both those who reported that they had no insurance at all and those who reported having only a
single-service plan, such as dental or hospitalization insurance. The uninsured category would also
have included anyone who reported being enrolled in the Indian Health Service, but no respondents
reported that type of coverage.
25
Liz Hamel, Mira Rao, Larry Levitt, Gary Claxton, Cynthia Cox, Karen Pollitz, and Mollyann
Brodie, Survey of Non-Group Health Insurance Enrollees (Menlo Park, CA: The Henry J. Kaiser
Family Foundation, June 19, 2014).
26
These results should be interpreted with caution. Research shows that consumers self-reports of
health insurance status are most reliable when stating whether or not they have any health
insurance. Their reports about what type of insurance they have are considerably less reliable.
Coverage through Medicaid, in particular, tends to be underreported in population surveys.
74
COMPARISON WITH OTHER STATES

Researchers have attempted to make state-to-state comparisons on the
performance of exchanges, including states cost per enrollee and their success in
enrolling the populations meant to be attracted to the exchanges. This section
reviews some of that research but shows why it is particularly difficult to
compare Minnesotas experience with that of other states.
The availability of MinnesotaCare hinders the ability to make meaningful comparisons
between MNsures enrollment performance and that of other states exchanges.
Below, we briefly comment on two previous studies.
Angoff Study
This analysis found that Minnesota ranked 37th out of 51 states (including the
District of Columbia) on exchange cost per enrollee.27 At the time of the study,
Minnesota had enrolled 48,495 people in qualified health plans. The study
assumed that Minnesota had spent $155 million, which was the total amount of
federal grants Minnesota had received for the exchange at the time, resulting in a
per-enrollee cost of $3,197. The median cost per enrollee for all states was
$1,715.28
However, Minnesotas enrollment figures as presented in this study are not
comparable to those of other states. MinnesotaCare enrollees would, in any other
state, be eligible to enroll in subsidized qualified health plans. Similarly, many
children who are covered by Medicaid in Minnesota would be eligible to enroll
in subsidized qualified health plans in some other states. (Some states have less
generous Medicaid eligibility rules for children than Minnesota.) About 38,000
people had enrolled in MinnesotaCare through MNsure at the time of this
analysis; adding this number to Minnesotas qualified health plan enrollment, the
per-enrollee cost is reduced to approximately $1,792. That revised calculation
would move Minnesota from 37th to 27th on exchange cost per enrollee out of all
states plus the District of Columbia, and the 5th lowest cost per enrollee out of
states operating their own exchanges in 2014.29
Moreover, this studys figures on the total amount spent on each exchange are
somewhat misleading. First, Minnesota had spent only a portion of its federal
grants at the time of the analysis, and the same may have been true of other
27
Jay Angoff, Cost-per-Enrollee in Each States Exchange (Washington, DC, May 7, 2014),
http://kaiserhealthnews.files.wordpress.com/2014/05/5-7-14-exchanges-report.pdf, accessed
December 29, 2014.
28
This analysis reported that the state with the lowest cost per enrollee was Florida ($76 per
enrollee), and the state with the highest was Hawaii ($23,899 per enrollee).
29
It is not feasible to do a similar correction based on Medicaid eligibility levels for children, as
any such correction would need to be applied to all states, and we did not have data on Medicaid
enrollment by age for each state.
ENROLLMENT
75
states. Second, Minnesotas spending helped enroll individuals in both public

and commercial insurance, while federal expenditures assigned to states that used
the federal health insurance exchange generally focused on enrolling individuals
in commercial insurance.
Urban Institute Study

An analysis by the Urban Institute looked at how many individuals each state
enrolled relative to the institutes estimate of the potential pool of exchange
enrollees in that state.30 Using a simulation model and American Community
Survey data, the Urban Institute projected that 75,000 people would enroll in
qualified health plans through Minnesotas exchange for the 2014 coverage year.
As of data available in April 2014, Minnesota had enrolled 48,495 people, or
64.7 percent of the projected enrollment. By that measure, Minnesotas
percentage of projected enrollment ranked 45th lowest out of 51 states (including
the District of Columbia). However, similar to the previous analysis,
Minnesotas number of enrollees was an undercount because of the existence of
MinnesotaCare. As of April 2014, MinnesotaCare had enrolled approximately
38,000 people. If they are added to the qualified health plan enrollees at the time,
Minnesota enrolled 115 percent of the total projected by the Urban Institute.
Based on that revised calculation, Minnesota ranked 17th overall among states in
the percentage of its projected pool of exchange consumers that were enrolled.
30
Linda J. Blumberg, John Holahan, Genevieve M. Kenney, Matthew Buettgens, Nathaniel

Anderson, Hannah Recht, and Stephen Zuckerman, Measuring Marketplace Enrollment Relative to
Enrollment Projections: Update (Washington, DC: Urban Institute, May 2014).
Chapter 5: User Experiences

T
he purpose of MNsure, as declared by the MNsure Board, is to ensure that

every Minnesota resident and small business, regardless of health status, can
easily find, choose, and purchase a health insurance product that they value and
does not consume a disproportionate share of their income.1 In this chapter, we
look at consumers experiences with MNsure. We also examine the experience
of some key MNsure stakeholdersspecifically, health insurers and counties.
Consumers and the people who helped them enroll encountered

numerous technical problems during MNsures first year of
enrollment.
Individuals who enrolled through MNsure generally reported

more satisfaction than dissatisfaction with the products they
purchased.
Problems with MNsures enrollment system had a significant

impact on the ability of insurers and counties to manage
individuals cases.
USER EXPECTATIONS
Before discussing users actual experiences with the MNsure website, it is
important to consider what expectations users may have had. To some extent,
these expectations reflected the way state officials described the exchange to the
media, in their publications, and on the exchange website.
MNsure created unrealistic expectations about the experience that users of the health
insurance exchange would have.
On a number of occasions, exchange officials likened the process of purchasing
health insurance through a state exchange to the process of making purchases at
popular consumer websites. For example:
1
2
In a 2011 press release, the exchanges executive director compared the

exchange to a website much like Travelocity or Expedia.com.2
MNsure Board Policy #01 (Charter and Bylaws), Article 1, Section 1.2.
MNsure, Minnesota Receives $1-Million Federal Grant to Plan Health Insurance Exchange
(February 25, 2011), https://www.mnsure.org/news-room/news/news-detail.jsp?id=486-53893,
accessed October 6, 2014.
78
In a newspaper opinion article, the Department of Commerce

commissioner said: Envision for a moment a consumer-friendly
website much like Orbitz.com (http://Orbitz.com) or Expedia.com
(http://Expedia.com) where Minnesotans can shop for affordable health
coverage.3
A Minnesota television stations story included the following comment

from MNsures executive director: Hopefully the end product to the
consumer will look [like] something as consumer friendly as searching
for an airline ticket on Travelocity or using something like Zappos to buy
shoes. Were hoping that its that simple for consumers.4
According to the exchanges marketing and communications plan, the

exchange is a user-friendly website like Priceline.com or Orbitz.com, an
easier way for Minnesotans to purchase private health insurance or
determine eligibility for public programs like Medical Assistance.5
MNsures homepages have used the terms easy and simple to

describe the enrollment process. For example, MNsures webpages have
said the exchange makes it easy to find and easy to compare
insurance, and that the exchange is easy-to-use.
Comparisons with other consumer websites may have helped people understand
that an exchange would provide an online shopping experience. However,
buying insurance is a more complex transaction than buying many other
products. For example, insurance purchases may require a consumer to provide
income and family information to determine whether the consumer qualifies for
public programs or tax credits. In addition, insurance products have many
variablesthe extent of coverage, the breadth of the provider networks available,
the quality of health care provided, premium costs, and out-of-pocket costs. As
we discuss below, the actual user experience was not as simple as MNsure
suggested it would be.6
3
Mike Rothman, Get Started on Minnesotas Own Health-Insurance Exchange, Pioneer Press,
April 27, 2011, http://www.twincities.com/alllistings/ci_17933791?source=rss, accessed
October 14, 2014.
4
MNsure Website Underway in St. Paul, April 25, 2013, http://www.kttc.com/story/22083332

/2013/04/25/mnsure, accessed October 9, 2014.
Minnesota Department of Commerce, 2012 Integrated Marketing Communications Plan (St. Paul,
August 2012), 2.
Problems with the MNsure website contributed to the complexity of the enrollment process. As
one MNsure-certified assister told us: People came into our office extremely exasperated when
trying to do it themselves at home. The advertisement [that] you can easily complete an application
at home is completely false! Browser incompatibilities, web server down, programming
incomplete made it a complete mess.
USER EXPERIENCES
79
CONSUMER ENROLLMENT PROCESS

We used several approaches to assess
Consumer Assisters We Surveyed
the consumer enrollment process
Brokers (or insurance agents) sell
during MNsures first open
insurance products on behalf of
enrollment period. First, we
insurance companies.
conducted surveys of individuals who
Navigators work for community
organizations and provide general
provided assistance to consumers.
enrollment assistance.
Specifically, we surveyed
Certified application counselors
representative, random samples of
typically work for hospitals and provide
MNsure-certified navigators, brokers,
enrollment assistance to patients.
and certified application counselors.
The box above shows the consumer assisters we surveyed, and Chapters 1 and 6
discuss these assisters further.7
Second, we surveyed a representative random sample of heads of households
who purchased a qualified health plan through MNsure.8 The experience of
successful enrollees may not represent the experience of others who visited the
MNsure website (and perhaps even created accounts) but did not enroll.
However, due to the limitations of MNsures data systems, we were unable to
obtain information on all individuals who created accounts but did not enroll.
Third, we obtained access to a test version of the MNsure enrollment website
from the first open enrollment period. This allowed us to view the content of the
website firsthand. Because this was a test version and not a live version of the
website, we did not fully experience the websites technical functionalityfor
example, whether heavy traffic at the site caused slowdowns, errors, or other
problems.
Complexity
As noted earlier, MNsure created expectations that its enrollment website would
be relatively simple to use. In fact, the website proved to be challenging for
many consumers.
Not counting members of our samples whose contact information was incorrect, we sent survey
invitations by e-mail to 313 navigators, 292 certified application counselors, and 547 brokers. We
received responses from 222 navigators (71 percent), 167 certified application counselors
(57 percent), and 295 brokers (54 percent). The survey results reported in this evaluation only
include those of respondents who said they did, in fact, work with MNsure applicants during the
first open enrollment period. Each sample of respondents was large enough that we can be
95 percent confident that the true percentage of the population who would have selected a
particular response was within 5 or 6 percentage points of the survey respondents answers,
depending on the question.
Not counting members of our sample whose contact information was incorrect, we sent survey
invitations by U.S. mail to 975 MNsure enrollees. We received responses from 281 enrollees
(29 percent). The sample of respondents was large enough that we can be 95 percent confident that
the true percentage of the population who would have selected a particular response to a survey
question was within 6 percentage points of the survey respondents answers.
80
MNsures enrollment website was difficult for some people to navigate and the enrollment
process was often lengthy.
The enrollment process often took a considerable amount of time. In our survey
of consumers who purchased qualified health plans, 58 percent reported that they
spent at least four hours on the enrollment process.9 In addition, majorities of
consumer assisters said the enrollment process took what they considered to be
an unacceptable amount of the consumers time. Specifically, 52 percent of
navigators, 56 percent of certified application counselors, and 86 percent of
brokers said the process was unacceptably long.
People who helped individuals enroll through MNsure generally said the process
was not an easy one. In our surveys of MNsure-certified assisters, 62 percent of
navigators said the process was not simple; 74 percent of certified application
counselors and 96 percent of insurance brokers said likewise.
Among the consumers we surveyed who enrolled in a qualified health plan,
64 percent said the MNsure website was not easy to use. Consumer difficulties
are illustrated in the following comment from a MNsure enrollee:
The insurance we ended up with, after MONTHS of effort, is a
product that we are glad to have. However the process and
experience we had with MNsure was HORRIBLE! My wife and
I both have extensive experience with computer software and
applications. MNsure was not ready for deployment, and we
spent over 150 hours trying to apply and trying to resolve
multiple problems with our application. We were without
insurance for two months because of the problems we
encountered.
The complexity and length of the enrollment process did not necessarily mean
that consumers struggled to understand the application questions. In our surveys
of consumer assisters, a majority of navigators and certified application
counselors said that, in general, MNsures application questions were written in
language that was easy for consumers to understand; most brokers disagreed.10
Consumer assisters we surveyed generally reported spending less than this amount of time, on
average, with the people they helped. Our surveys of assisters asked for the average amount of
time it took to complete the online portion of the enrollment process; in contrast, our survey of
consumers asked for the total time spent online, on the phone, or with an assister. The median
response for navigators and certified application counselors was one to two hours; we did not ask
this question of brokers.
10
The percentage of assisters who said the application questions were written in language that was
easy for consumers to understand was 58 percent for navigators, 67 percent for certified application
counselors, and 38 percent for brokers. The percentage that disagreed was 41 percent for
navigators, 30 percent for certified application counselors, and 60 percent for brokers.
USER EXPERIENCES
81
Technical Issues
One reason that the enrollment process was time-consuming and complex is that
the MNsure website did not work as well as intended.
Consumers and MNsure-certified assisters experienced numerous technical problems with
the MNsure website during the first open enrollment period.
Even individuals who successfully enrolled in a qualified health plan had
difficulties with the MNsure website. Among people we surveyed who enrolled
through MNsure in a qualified health plan, 75 percent reported having
significant technical problems with the website.11
Assisters who helped consumers enroll also experienced many technical
problems. When asked whether MNsures enrollment website was generally
free of technical problems and glitches during the first open enrollment period,
only 7 percent of navigators said it was, as did 8 percent of certified application
counselors, and 1 percent of brokers. When asked how often they experienced
technical problems that caused significant delays, a majority of brokers (62
percent) said this happened in 76 to 100 percent of the applications with which
they assisted. Navigators and certified application counselors reported somewhat
better experiences but 50 and 56 percent of respondents, respectively, said they
had significant technical problems with at least half of the applications they
handled.
The MNsure enrollment website was not equally compatible with all web
browsers. For a while, MNsures customer service staff who answered phone
and e-mail questions directed consumers to complete the first part of their
application in one browser and then switch to a different browser to complete the
application. MNsure had some information on its website about which browsers
to use when enrolling, but this guidance was not particularly easy to locate.
Once applicants accessed the online enrollment system, they encountered a
variety of technical problems. For example, one person who purchased insurance
through MNsure said:
Im very thankful that MNsure exists. It allowed me to start my
own company and insure my family with a much better plan than
I thought was possible for a reasonable price. However, the
entire website was a disaster top to bottom. Disjointed
navigation, basic website features found everywhere online were
missing or broken, slow, error prone, different user experience
across sections of the websiteincredible that such a bad
product could come from such a huge expenditure.
11
On the survey, 75 percent of respondents disagreed or strongly disagreed with the following
statement: During the enrollment process, I did not encounter any significant technical problems
with the MNsure website.
82
Another MNsure enrollee said:

The website for obtaining [information] about subsidies and for
enrollment was poorly designed. As a former software engineer
with [graphical user interface] experience, it was painfully clear
to me that (1) the software/system was not adequately designed
or tested, (2) the servers could not handle the load of inquiries,
(3) the user interface was cumbersome, (4) data had to be entered
twice: first for qualifying for subsidies and then again for
enrollment.
As noted in Chapter 2, many MNsure applications got stuck inside the
enrollment system. It sometimes took weeks or months for MNsure to find these
applications, resolve the remaining issues, and get the individuals enrolled.
Users encountered a variety of other technical problems, such as: inability to
access accounts if usernames contained characters such as @ or #; error
messages for reasons that were unclear; and online screens that froze.
Website Content
Consumers who tried to enroll in insurance through MNsure expected a website
without technical problems; they also expected the website to have clear,
understandable guidance. For people purchasing a qualified health plan through
MNsure, the process consisted of multiple steps:
Create a MNsure account and have your identify verified.
Provide information that will be used to determine eligibility for tax

credits or a public health care program.12
Shop on the MNsure website for an insurance product.
The MNsure website provided consumers with the option of shopping

anonymouslythat is, the ability to look at health plan options and prices prior
to creating a MNsure account and starting the enrollment process. In our view,
this option was not very well publicized during the first open enrollment period,
perhaps causing some users to go through the more tedious steps in the
enrollment process before fully understanding what options MNsure might offer
them.13
Below, we discuss issues related to website content.
12
Individuals were not required to request an eligibility determination for subsidies or public health
care programs, but most people who enrolled did.
13
Also, MNsures anonymous shopping process was more complex than those of other exchanges,
according to a website that helps consumers find health insurance. This website said that among
single web site exchanges with anonymous health plan comparisons, testers found Minnesotas
exchange, MNsure, had the most steps at 18. It noted that more steps increase the risk of web
site visitors abandoning the shopping process. See Healthpocket, Healthpocket Compares State
Health Insurance Exchanges, October 17, 2013, http://www.healthpocket.com/healthcare-research
/infostat/ranking-state-health-insurance-exchanges#.VH3D5smZiSo, accessed December 1, 2014.
USER EXPERIENCES
83
Account Creation
For people seeking to purchase a qualified health plan, the first real step in the
enrollment processcreating a MNsure accountprovided ample opportunities
for consumer confusion or frustration. The initial webpage was titled
Introduction to Identity Proofing and Obtaining an Account. The process of
authenticating the identity of a MNsure applicant is an essential one; however,
the term identity proofing was not defined on this page, and it was jargon that
many users might not immediately understand. Furthermore, the MNsure
website did not provide clear assistance or instructions for persons with limited
English proficiency who were trying to create an account.
When creating an account, users were asked to provide personally identifying
information, including name, address, contact information, birthdate, and Social
Security number. Users then encountered a visual security test (known as a
CAPTCHA) that is intended to ensure that the account is being created by a
person rather than a computer.14 This type of test is used by many websites, but
it can be frustrating even for users who have encountered them before.
To complete the account creation process, users were asked to provide answers to
several security questions and a shared secret. The security questions were
intended to help MNsure verify the identity of someone trying to access a users
account at a later date. A person creating a MNsure account was given eight
options for security questions (such as What city would you like to retire to?)
and had to provide responses to five. Some of the security questions were
vaguely worded or prompted responses that might have been hard for a user to
recall at a later date.15 A user was then asked to provide a shared secret to
MNsure, with no guidance about the nature or topic area of the secretperhaps
making the shared secret hard for the consumer to remember later.16 Also,
while these types of security provisions are used by many websites that store
personal information, the requirement for MNsure users to provide a shared
secret plus responses to five security questions may have taxed a consumers
patience.
Some individuals who successfully completed the account creation process later
encountered problems if they needed to get their password reset or did not recall
their responses to the security questions. For example, a MNsure-certified
consumer assister was told by MNsure customer service staff in late 2014 that
14
CAPTCHA stands for Completely Automated Public Turing test to tell Computers and Humans
Apart. The user is typically shown a distorted image of letters and numbers and is asked to type in
what he or she sees. Sometimes the distortion makes it difficult for users to provide an accurate
response. CAPTCHAs have also been criticized for providing a difficult challenge for some people
with disabilities.
15
For example, a question asking for the city in which someone would like to retire might be
difficult for a young adult to answer. A question that asked the consumer to specify your
grandfathers occupation did not specify which grandfather it referenced. A question asking for
the most memorable date in your life may have been intended to solicit a calendar date, but it did
not specify the proper format for entering the date.
16
MNsure uses the shared secret to help verify user identity during phone contacts with MNsure
contact center staff.
84
there was no recourse if consumers did not recall the answers to security
questions.
Application Questions
A person seeking to enroll in either a subsidized qualified health plan or in a
publicly funded health care program was required to complete a lengthy series of
application questions. Some of these questions were clear and well explained. In
our view, other questions were potentially confusing. Exhibit 5.1 shows
examples.
The application process did not make sufficient accommodations for consumers
with limited English skills. A question about whether the applicant needed an
interpreter occurred well into the application process. Also, MNsures website
provided a sortable directory of in-person assisters for the first open enrollment
period, but website users could not sort this directory by language. This may
have made it harder for some limited English speakers to find an assister who
spoke their language at a nearby location.
MNsure had potentially helpful advice on its website, but it was not always easy
to find. The website had separate Get Help, Learn More, Frequently Asked
Questions, and glossary sections. It was not always clear which of these
sections contained answers to particular questions.
Shopping for Insurance Products

Individuals who wished to purchase a qualified health plan could choose from
dozens of products available through MNsure.
MNsure provided consumers with potentially useful online shopping tools, although
consumer assisters and MNsure officials said these tools need further improvement.
On the MNsure website, consumers had the opportunity to narrow the health
insurance products under consideration by specifying certain variables that
matter most to me. These variables included things such as the type of
insurance product, whether the product included a particular clinic or hospital,
whether the insurance covered wellness programs, and the level of the
deductible. These shopping options enabled useful comparisons.
USER EXPERIENCES
85
Exhibit 5.1: Examples of Potentially Confusing

Application Questions from the First Open Enrollment
Period
The application had separate paths for persons applying for health insurance with
discounts and without discounts. However, the term discounts was not
adequately explained. Users may have been confused about whether discounts
referred to tax credits, public programs, or specially priced insurance products.
When checking for discounts, consumers were asked some questions without being
clearly told the reason for the questions and what bearing their answers may have for
eligibility determinations. For example: Is anyone getting services from the Center
for Victims of Torture?
Some questions asked for information but did not clearly specify the format in which
the consumer should replyfor example, whether Social Security numbers should
include dashes, or the proper format for reporting birthdates. Consumers had to click
on a separate help button to find the proper format.
Regarding income, the applicant was prompted to enter Amount in one question,
and then Frequency in the next. This was potentially confusing. Frequency was
intended to refer to the periodsuch as Annualfor which the Amount question
was answered, not the frequency of the consumers pay periods.
The application asked the consumer to report current taxable income. It urged
consumers to check their tax forms to see how this was defined or to visit the IRS
website. Consumer assisters told us there was considerable confusion over this
question.
One question in the application was a run-on sentence, which should have been
corrected during the editing process: Does [the primary applicant] pay for certain
things that can be deducted on an income tax return, telling us about them could
make the cost of health insurance a little lower.
The application contained the following statement: I know Ill be asked to cooperate
with the agency that collects medical support from an absent parent. If I think that
cooperating to collect medical support will harm me or my children, I can tell the
agency and wont have to cooperate. Below this statement, there was a box that
simply said Medical Support. The box had to be checked to proceed with the
application, but the application did not clearly indicate what a marked box indicated.
There were several other checkboxes that had similar problems.
If consumers appeared to be eligible for Medicaid or the Childrens Health Insurance
Program (CHIP), the application asked them to answer a number of questions to
ensure that [they got] the right services. (Medicaid and CHIP were not explained.)
Some of these questions were very personal. For example, consumers were asked
whether any members of the household have a communicable disease (HIV, AIDS,
tuberculosis, hepatitis, etc.)? or Have any of these people been determined as being
seriously and persistently mentally ill or as being severely emotionally disturbed? This
part of the application had no further explanation about why this information was being
collected, whether this information was relevant to the MNsure application, how
respondents would be directed to the right services, or with whom this information
would be shared for the purposes of getting people to the right services.
In our survey of consumers who purchased a qualified health plan, we asked about
the value of MNsures online shopping tools. We focused on responses from
individuals who had previous experience purchasing health insurance. As shown
in Exhibit 5.2, those who said MNsure made the shopping experience easier than
what they had previously experienced outnumbered those who said MNsure made
it harder (48 percent vs. 31 percent). Likewise, among persons with previous
experience shopping for insurance, 62 percent said MNsure made it easier to find
insurance that fit their budget, and 19 percent said MNsure made it harder.
86
Exhibit 5.2: Perceptions of MNsure Enrollees Who

Had Prior Experience Buying Insurance, 2014
Percentage of Respondents Who Said:
62%
48%
31%
19%
MNsure
made
MNsure made
MNsuremade
MNsuremade
shopping for
shopping for
theshopping
and
comparing theshopping
and comparing
experience
experience
insurance
insurance
harder
easiereasier
harder
MNsure made
MNsure made MNsuremadeit
MNsuremadeit
it harder to
it easier to
easiertofind hardertofind
find insurance find insurance
insurancethat
insurancethat
that fits my
that
fits my
fitmybudget
fitmybudget
budget
budget
NOTES: The survey was conducted in September to November 2014. Respondents who offered no
opinion or said that MNsure made no difference are not shown. Overall, the sample of respondents
was large enough that we can be 95 percent confident that the true percentage of the population who
would have selected a particular response to a survey question was within 6 percentage points of the
survey respondents answers.
SOURCE: Office of the Legislative Auditor, survey of heads of households who enrolled in qualified
health plans (N=205).
We also asked assisters about the adequacy of online tools MNsure provided for
making comparisons among qualified health plans. As shown in Exhibit 5.3,
navigators and certified application counselors were generally more favorable in
their assessments than brokers. For example, 45 percent of navigators,
39 percent of certified application counselors, and 26 percent of brokers said
MNsure provided useful tools for helping consumers select the best available
insurance for a given price. In our discussions with MNsure board members and
administrators, several told us they would like to see improvements in the
consumer shopping experience on the MNsure website. However, much of
leaderships focus during the websites first year was on addressing the sites
technical problems.
USER EXPERIENCES
87
Exhibit 5.3: Satisfaction of Consumer Assisters with

Tools MNsure Provided to Consumers, 2014
Percentage of Survey Respondents Who Said MNsure Provided Useful Tools to:
70%
60%
50%
Compare the prices of

health insurance products
40%
Compare the features of

health insurance products
30%
20%
Select the best available

insurance for a given price
10%
0%
Navigators
Certified
Application
Counselors
Brokers
NOTES: The surveys were conducted from July through October 2014. The sample of respondents
would have selected a particular response to a survey question was within 5.0 to 6.6 percentage
points of the survey respondents answers. The percentages of assisters who offered no opinion on
these questions ranged from 17 to 21 percent for navigators, 25 to 27 percent for certified application
counselors, and 4 to 5 percent for brokers.
SOURCE: Office of the Legislative Auditor, surveys of navigators (N=203), certified application
counselors (N=124 to 125), and brokers (N=252 to 253).
Most consumers we surveyed said that the MNsure website provided adequate information
on the insurance products for sale, while consumer assisters had mixed opinions.
We surveyed a sample of individuals who purchased qualified health plans. As
shown in Exhibit 5.4, 54 percent of respondents said MNsures online
information on health insurance products was easy to understand. Likewise, our
surveys of MNsure-certified consumer assisters asked whether MNsure provided
information about health insurance products in language that was easy for
consumers to understand. The percentage who said MNsure did so varied among
the assister groups surveyed50 percent of navigators, 61 percent of certified
assistance counselors, and 30 percent of brokers.17
17
The percentages of respondents who said that MNsure did not provide information on health
insurance products in understandable language were 44 percent for navigators, 34 percent for
certified application counselors, and 66 percent for brokers.
88
Exhibit 5.4: User Perceptions of Information MNsure

Provided on Health Insurance Products, 2014
Percentage of Survey Respondents Who Said MNsures Information about Health
Insurance Products was:
Easy to understand
Not easy to understand

66%
61%
54%
50%
42%
44%
34%
Enrollees
Navigators
Counselors
30%
Brokers
NOTES: Survey respondents who offered no opinion are not shown in the graph. The surveys were
conducted in July through November 2014. The samples of respondents were large enough that we
can be 95 percent confident that the true percentage of the population who would have selected a
particular response to a survey question was within 5.3 to 6.3 percentage points of the survey
respondents answers.
SOURCE: Office of the Legislative Auditor, surveys of heads of households who enrolled in qualified
health plans (N=268) and surveys of MNsure-certified navigators (N=203), certified application
counselors (N=128), and insurance brokers (N=256).
In our view, some descriptions of insurance products on the MNsure website

were potentially confusing to consumers. For example, many of PreferredOnes
products had non-embedded deductibles.18 This term was not defined in
PreferredOnes product information or in MNsures online glossary, and even
some health care experts we talked with were unfamiliar with the term.
Exhibit 5.5 shows the opinions of consumer assisters about the online
information MNsure provided on various aspects of insurance products.
Specifically, the exhibit shows the percentage of assisters who said the MNsure
websites information was often, almost always, or always sufficient.
Assisters expressed mixed satisfaction with the information that MNsures
website provided. Assisters expressed the least satisfaction with information on
provider networks; only about one-fourth of the assisters said this information
18
Insurance plans with embedded deductibles include individual deductibles and a family
deductible. Having the individual deductibles embedded within the family deductibles allows for
each member of the family to have expenses covered before the entire amount of the family
deductible is reached. Insurance plans with a non-embedded deductible require the policy holder to
incur expenses equaling the entire amount of the family deductible before the insurer will pay for
any medical bills.
USER EXPERIENCES
89
Exhibit 5.5: Consumer Assister Satisfaction with

Information Provided by MNsure on Specific
Elements of Insurance Products, 2014
Topic
Percentage of Respondents Who Said MNsures

Website Always, Almost Always, or Often
Provided Sufficient Information on the Topic:a
Certified
Application
Navigators
Counselors
Brokers
Scope of the insurance coverage

Provider networks
Deductibles
Copayments
Out-of-pocket maximums
Tax credits and cost-sharing
reductions
The final price of insurance for a
qualified health plan after discounts
46%
35
57
47
46
33%
27
42
36
37
31%
23
61
50
47
47
39
29
50
34
37
NOTES: The surveys were conducted from July through October 2014. The samples of respondents
were large enough that we can be 95 percent confident that the true percentage of the populations
who would have selected a particular response to a survey question was within 4.9 to 6.6 percentage
points of the survey respondents answers.
a
Respondents had a choice of Always or almost always, Often, Sometimes, Rarely or never, or
No opinion. This exhibit shows the total percentage that provided one of the first two responses.
SOURCE: Office of the Legislative Auditor, surveys of navigators (N=202 to 203), certified
application counselors (N=124 to 126), and brokers (N=249 to 253).
was sufficient. Assisters expressed the most satisfaction with information on

health plan deductibles.19
We also asked consumers why they bought insurance through MNsure rather
than directly from an insurer; Exhibit 5.6 shows the responses to our survey. The
most common reasons cited by respondents were that MNsure was less
expensive, provided access to tax credits not available elsewhere, or allowed
consumers to compare various insurance products.
Consumer Notifications
Consumers who shop online expect businesses to provide prompt confirmation
that their purchases were successfully processed. In the case of MNsure,
consumers expected to quickly learn what benefits they were eligible to receive,
whether their applications were processed, and whether they were successfully
enrolled in a health insurance plan or public program.
19
Between 25 and 32 percent of certified application counselors offered no opinion to these

questions, compared with 15 to 21 percent for navigators and 4 to 8 percent for brokers.
90
Exhibit 5.6: Enrollees Reasons for Buying Insurance

through MNsure Rather than Directly from an Insurer
Percentage of Respondents Who Cited These Reasons:
39.4%
37.4%
31.0%
14.6%
12.0%
7.8%
Did not
Did not
know direct think they
purchase
would
was an
qualify for
option
direct
purchase
7.6%
MNsure
Availability
MNsure
Referred to
MNsure
was less
of tax
was easier MNsure by facilitated
expensive
credits
to use
someone comparison
than direct
through
than direct
of insurance
purchase
MNsure
purchase
products
NOTES: Respondents could select more than one option. Respondents who responded No
opinion or who wrote in another response are not shown. The sample of respondents was large
enough that we can be 95 percent confident that the true percentage of the population who would
have selected a particular response to a survey question was within 6 percentage points of the survey
SOURCE: Office of the Legislative Auditor, September to November 2014 survey of heads of
households who enrolled in a qualified health plan (N=281).
In many cases, MNsure customers received inadequate information about (1) the status of
their application for insurance or (2) their eligibility for public programs or tax credits.
When people make online purchases, they usually get confirmation of their
purchase at the end of the process and receive a confirmation e-mail. But, for the
first open enrollment period, and even into the second one, the MNsure website
did not provide a confirmation page to individuals at the end of the application,
stating that they had successfully enrolled in public or commercial plans. In
addition, MNsure did not send consumers a confirmation e-mail. Our survey of
insurance brokers asked whether the MNsure website provided sufficient
confirmation that the enrollment process had been completed; two-thirds said it
did not.20 As one broker suggested to us, Brokers should be able to print online
applications with a confirmation number showing the application successfully
was completed and went through the system.
20
Two other groups of assisters we surveyed were less critical than brokers of MNsures online
confirmations. Still, 32 to 40 percent of navigators and certified application counselors said the
confirmations were not sufficient (depending on the type of enrollees we asked about).
USER EXPERIENCES
91
MNsure also created a process for communicating with applicants by mail.

Specifically, applicants were to receive letters that informed them of their
eligibility for tax credits or public health care programs. Also, applicants for
public programs whose applications were incomplete were to be informed by
letter that their applications were pending. (Applications were incomplete if
information regarding identification, income, citizenship, tribal membership, or
other issues could not be verified by state or federal sources.) The Minnesota
Department of Human Services (DHS) was supposed to mail consumers letters
on behalf of MNsure. But, for several months in 2014, these processes for
written notifications did not work as intended.
First, from January to July 2014, DHS did not provide written notification to
individuals whose applications for public programs were pending. DHS found
errors in the notices being sent to individuals, and DHS and MNsure suspended
the notification process on January 8, 2014. Later that month, DHS determined
that the errors had been corrected. However, due to miscommunication between
DHS and a vendor, immediate resumption of these mailings did not occur. This
problem came to the attention of top DHS officials in June 2014, and mailing of
the pending notices resumed in July 2014.21 According to DHS, more than
14,000 MNsure applications had pending issues as of June 2014 but had not
received a DHS notification.22
The impact of this problem was that many MNsure applicants were not told for
weeks or months that they did not have insurance coverage, nor were they told
what steps to take to get it. During this period, some of these individuals may
have been alerted by county staff that their applications were not yet complete.23
Second, during this same time period, MNsure applicants did not receive written
notification of their eligibility for tax credits or public programs. Initially,
MNsure and DHS halted the mailing of these letters in January 2014 because the
letters had incorrect dates of eligibility. This took several weeks to fix.
Exchange officials then identified additional changes that were needed in the
content of the letters. Mailings of letters to individuals regarding their eligibility
for tax credits or public programs resumed in July 2014.
21
In early 2014, a MNsure vendor (EngagePoint) told DHS that it was sending notices of pending
applications to DHS to be mailed to consumers. Months later, however, a DHS official said the
vendors internal review showed that notices may not have been printed since January 7, 2014.
Staff in some counties told us that, on occasions prior to June 2014, they expressed concern to DHS
about the apparent lack of notices being sent to program applicants. High-level DHS officials said
they were unaware that pending notices were not being mailed until June 2014.
22
In addition to 14,467 cases pending as of June 2014 that had not yet been notified by DHS, there
may have been other individuals whose pending issues were resolved during the first six months of
2014, thus negating the need for DHS notification of pending status in June 2014.
23
DHS sent counties lists of pending cases on three occasions during the first half of 2014 and
asked counties to follow up with these individuals to obtain the necessary information.
92
CONSUMER SATISFACTION WITH PRODUCTS

PURCHASED
It is important for consumers to have a relatively trouble-free experience when
enrolling in health insurance through MNsure, but it is also important for
consumers to find products they like. Our survey of a sample of consumers who
purchased qualified health plans provided an early picture of consumers views.
Most of these consumers probably bought their plans during the first open
enrollment period, which ended March 31, 2014; we conducted our survey in
September 2014. It may take some consumers longer than a few months to use
their health insurance and determine what they think of it.
Persons who enrolled through MNsure in the first open enrollment period generally reported
more satisfaction than dissatisfaction with the products they purchased.
In our survey of people who purchased a qualified health plan through MNsure,
67 percent said they would choose the same product again, if given the choice;
24 percent said they would not. It is worth noting that most people who
purchased a qualified health plan during MNsures first open enrollment period
did not have the opportunity to buy the same product through MNsure in the
second open enrollment period. The insurer (PreferredOne) that accounted for
about 59 percent of the qualified health plans sold during MNsures first open
enrollment period decided not to offer products through MNsure in the second
year. People who had purchased a PreferredOne product the first year could only
re-enroll in a PreferredOne product outside of MNsure. In doing so, the
consumer would not have been eligible to receive tax credits available only by
purchasing insurance through MNsure.
Our consumer survey also asked how likely the respondent would be to
recommend MNsure to friends or family looking to purchase health insurance.
Using a scale from 0 (not at all likely) to 10 (extremely likely), each survey
recipient was asked to select a response that reflected the likelihood of making
such a recommendation. On one hand, we found that more respondents chose
options 6 through 10 than options 0 through 4; this suggests that, on balance,
there was a level of satisfaction with what they had purchased. On the other
hand, when this question format is used in consumer research, it is often analyzed
with the assumption that only persons responding 9 or 10 are satisfied to the
point of being promoters of the product.24 Using this alternative method of
analysis, MNsures customers included more of what researchers call
detractors than promoters.
We also asked respondents to identify any specific benefits or negative impacts
MNsure has had on them. Exhibit 5.7 shows the responses. The percentage of
24
See a discussion of the method in Frederick Reichheld, The One Number You Need to Grow,
Harvard Business Review 81, n. 12 (December 2003), 46-54. Under this approach, ratings of 9 and
10 are considered promoters of a product, and ratings of 0 through 6 are considered detractors.
Persons rating a product as 7 or 8 are considered passively satisfied.
USER EXPERIENCES
93
survey respondents who cited any of the benefits shown in the exhibit
(63 percent) outnumbered those who cited any negative impacts (42 percent).
Interestingly, the most common benefit mentioned by the enrollees was lower
costs (cited by 43 percent of respondents), while the most common negative
effect mentioned was higher costs (cited by 26 percent of respondents).
Exhibit 5.7: Enrollee Perceptions of Positive and

Negative Effects of MNsure, 2014
Benefits from MNsure that Enrollees Said They Received

Lowered my health insurance costs
Increased my choices for insurance
Improved my insurance coverage
Improved my access to the health care providers I want to see
Improved my access to medication
Total who cited at least one of the benefits above
Negative Effects of MNsure that Enrollees Said They Experienced

Increased my health insurance costs
Reduced my health insurance coverage
Reduced my choices for insurance
Reduced my access to the health care providers I want to see
Reduced my access to medication
Total who cited at least one of the negative effects above
Percentage of
Surveyed Enrollees
Who Cited this
Benefit
43%
25
23
11
8
63%
Percentage of
Surveyed Enrollees
Who Cited this
Negative Effect
26%
16
11
9
6
42%
NOTES: The survey was conducted in September to November 2014. The sample of respondents
survey respondents answers.
health plans (N=281).
In addition, we asked people who bought qualified health plans through MNsure
about specific aspects of the product they purchased. As shown in Exhibit 5.8, at
least two-thirds of surveyed enrollees said (in response to separate questions) that
the product they purchased through MNsure had met their expectations regarding
choices of health care providers, policies regarding deductibles, policies
regarding copayments, and coverage of products and services.
For enrollees who had health insurance immediately before enrolling in MNsure, survey
respondents reported mixed views on their MNsure product compared with their previous
insuranceparticularly regarding premiums and out-of-pocket costs.
94
Exhibit 5.8: MNsure Enrollees Views of Certain

Aspects of the Insurance They Purchased, 2014
Features of Insurance Purchased
by Enrollees
Choice of health care providers
Deductibles
Copayments
Coverage of products and services

0%
20%
40%
60%
80%
100%
Percentage of Survey Respondents Who Said

the Features of the Insurance They Purchased:
Met their expectations
Did not meet their expectations
No opinion
NOTES: In our survey of MNsure enrollees, we asked whether the insurance enrollees purchased
through MNsure met their expectations. For example, we asked them if the insurance they
purchased offered the choice of health care providers or the deductibles that they expected. The
survey was conducted in September to November 2014. The sample of respondents was large
enough that we can be 95 percent confident that the true percentage of the population who would
have selected a particular response to a survey question was within 6 percentage points of the survey
health plans (N=276 to 277).
Exhibit 5.9 shows how people who purchased insurance through MNsure
compared this insurance with what they had previously. Forty-eight percent of
consumers who had been insured immediately prior to MNsure reported that their
premiums (after tax credits) were lower through MNsure than they were
previously; 32 percent said their MNsure premiums were higher. In addition, 43
percent of previously insured people said they paid higher out-of-pocket costs
with the insurance they bought through MNsure than with the insurance they had
previously. Most people reported no change in the options their new insurance
offered for primary care doctors.
USER EXPERIENCES
95
Exhibit 5.9: Enrollees Comparison of Insurance

Purchased Through MNsure with Insurance They Had
Previously, 2014
Insurance Characteristic
Percentage of Respondents Who Said the

Insurance They Purchased through MNsure was
Better Than, About the Same, or Worse
Than the Insurance They Had Previously:
Better
Same
Worse
Coverage for health care services and

products
Out-of-pocket costs
Options for primary care doctors
Premiums (after any tax credits)
16%
30
9
48
48%
23
69
14
30%
43
15
32
NOTES: The survey was conducted in September to November 2014. The sample of respondents
survey respondents answers. Respondents who said dont know are not shown in this table. This
survey question was not asked of individuals who said they did not have insurance immediately prior
to purchasing insurance through MNsure.
SOURCE: Office of the Legislative Auditor, survey of heads of households who purchased a quality
health plan through MNsure (N=202 to 204).
IMPACTS ON KEY STAKEHOLDERS

MNsure was developed to help consumers enroll in insurance. A variety of
stakeholders have assisted in this process. In this section, we focus on the
experience of two groups: (1) health insurance companies whose products were
sold through MNsure and (2) counties that administer the public health care
programs in which most MNsure enrollees participate.
Health insurers and counties were frustrated by the impact of MNsures problems on their
workloads, business practices, and ability to serve consumers.
To help us better understand the experience of these stakeholders, we used
interviews and surveys to solicit input from representatives of all five health
insurers that participated in MNsure in its first year. In addition, we obtained
input from officials in 12 county human services agencies.
Health Insurers
Once a person signs up for health insurance through MNsure, MNsure notifies
the insurer of the enrollment. This enables the insurer (also known as a carrier
96
or health plan) to correspond with the enrollee, collect premiums, and

eventually pay the enrollees claims.25
MNsure had insufficient processes for transmitting information on enrollees to insurers
during MNsures first open enrollment period.
Originally, MNsure planned to send insurers specially formatted electronic
enrollment files with information on new enrollees. But shortly before
October 1, 2013, MNsure informed insurers that the files would not be available
right away, and that some of them would not be available in the format the
insurers were expecting. About two months into the open enrollment period,
MNsure sent the first files to insurers. Many of these files were created manually
rather than automatically by MNsures system, not formatted in the way the
insurers expected, and did not include all information the insurers needed. Later,
MNsure began sending separate files that contained the missing information.
Insurers said they devoted considerable time to manually entering and matching
these files, due to (1) the unexpected format in which MNsure sent many of the
files, (2) the differing dates on which insurers received separate reports on the
same individuals, and (3) questions about the accuracy of the data MNsure
provided. One carrier official summarized the difficulties as follows:
The inability to automate files from MNsure to health plans was
the greatest difficulty during open enrollment. This inability led
to numerous manual workarounds, and every time a new
workaround is necessary the possibility of errors in the
information being shared increases. The overall inability to
receive accurate, timely enrollment files from MNsure delayed
processing on our end, delayed the delivery of membership
materials [to consumers], resulted in inaccurate enrollments, and
necessitated extensive resources to track and verify information.
Insurers cited a variety of other problems:
MNsure collected initial premium payments from some consumers

during part of the first open enrollment period, and MNsure has been
collecting ongoing premium payments from small businesses that enroll
through MNsure.26 However, MNsure was unable to send payment files
to insurers in the format expected by insurers, so insurers had to enter
payment information manually into their information systems.
25
This notification of the insurer is known in the insurance industry as an 834 EDI transmission.
It is not a new process; it is used by employers, insurance brokers, health insurance exchanges, and
governments to provide insurance companies with descriptive information on individuals that have
purchased or enrolled in policies.
26
During the latter part of the first open enrollment period, insurers assumed responsibility from
MNsure for collecting initial premium payments from individuals. One insurer said MNsure
assessed its limited abilities to handle payments and made the right decision to give this
responsibility to insurers that have long experience with customer billing.
USER EXPERIENCES
97
Contrary to original expectations, MNsures system cannot identify

which enrollees are still covered by the plans they purchased, and which
enrollees terminated their coverage mid-year. Consequently, MNsure
has relied on insurers to provide it with up-to-date information on
enrollees so it could inform individuals of the need to re-enroll.
Insurers have produced monthly reports for the federal government that
indicate the number of individuals enrolled through MNsure receiving
premium tax credits or cost-sharing reductions. MNsure originally
intended to prepare this report.
Insurers spent considerable time working with MNsure to address

applications that were stuck in the MNsure system. Technical
problems left some consumers frustrated and without coverage for a
period of time.27
Insurers collected documents from external sources to verify individuals

eligibility to enroll outside of open enrollment periods (due to life
events, such as changes in residence, income, or family size) because
MNsure was not doing these verifications.
Many individuals submitted paper applications, but insurers said

MNsures processes for handling this type of application were
inadequate.28 They said this resulted in enrollment delays.
When consumers who purchased insurance through MNsure in 2014

logged into their online accounts to renew their coverage for 2015,
MNsure planned to provide them with information about their current
insurance policies and rates. This feature was meant to allow consumers
to use their existing policy as a point of reference as they shopped for a
new policy. MNsure was unable to do this, and insurers thought this
placed customers at a disadvantage as they shopped for their 2015
coverage.
Overall, the insurers did not think the MNsure enrollment process for their
customers in public health care programs was an improvement over the process
used previously. Additionally, insurers participating in MNsure that also sold
commercial products outside MNsure said that MNsures enrollment process for
commercial customers was more difficult. One insurer told us that it spent at
least $500,000 to modify its information systems in response to changes MNsure
made in its original plans.
27
Some insurers told us that consumers were given retroactive coverage for the period when their
applications were stuck, but in the meantime enrollees may have postponed health care visits or
worried about the status of their coverage.
28
Examples of problems cited with MNsures process for handling paper applications were: lack
of systems for tracking which applications had been received or the status of these applications; no
option on the paper applications for applicants to select the health plan they wanted; and no process
for applicants to verbally authorize an application when missing application information was
provided by phone.
98
Counties
Counties are the front lines for Minnesotas public health care insurance
programs. The state sets overall health care policy for these programs, operates
statewide information and eligibility determination systems, and oversees
program implementation, while the counties assist with the enrollment process,
manage individual cases, and troubleshoot issues that arise for enrollees. A
recent estimate said that the 2011 county cost of administering health care
programs on behalf of the state was about $100 million dollars.29
Due to weaknesses in MNsures enrollment system, counties have not been able to
effectively manage the MNsure cases for which they are responsible.
Perhaps the largest county frustration has been the absence of a well-functioning
caseworker portal into the MNsure enrollment system. The portal was
supposed to provide a special point of entry for county staff, allowing them to
view the status of MNsure applications and enrollments so they could actively
manage these cases. However, county officials told us the portal does not show
staff how many Medical Assistance cases are on their caseloads at a given time.
In addition, the MNsure caseworker portal provides information to counties on
tasks that need to be completed (such as closing the case of a person who has
moved out of state), but it only allows county staff to view 500 tasks at a time.
This may be a small fraction of a large countys total case-related tasks.30
Furthermore, county staff have had limited ability to address the tasks identified
in the caseworker portal, due to problems with MNsures technology. For
example, county staff could not add a newborn to a MNsure-enrolled case until
mid-2014. Over time, DHS has developed manual workarounds for counties so
staff can perform tasks they were initially unable to perform through the portal.
But the time it now takes for a county to add a newborn to a MNsure case is 60 to
90 minutes; such a task required only a few minutes in the case management
system (MAXIS) that DHS previously used to manage Medical Assistance cases.
There continue to be many tasks in the MNsure caseworker portal that county
staff are not authorized to complete. For example, as of late 2014, when MNsure
enrollees died or moved, or when their incomes rose to a level above eligibility
thresholds, county staff could not close the enrollees cases.
Other concerns we heard from county staff about their experiences as users of the
MNsure system included the following:
29
30
The MNsure system has not given counties the ability to determine how
many MNsure applications involving their residents were pending at a
given time. (As noted earlier, a pending application is one that is
incomplete and requires additional information from the applicant.)
Because of this limitation in the system, DHS offered to send counties
Fiscal note prepared for H.F. 5-8E, 2013 Leg., 88th Sess. (MN), 20.
For example, as of mid-2014, Hennepin County had over 20,000 tasks to perform in order to
complete enrollment for that countys MNsure applicants.
USER EXPERIENCES
99
lists of their pending cases monthly. But county staff told us they only
received these lists once or twice in all of 2014.
Contrary to MNsures original intentions, applicants deemed eligible for

public health care programs could not select their health plan through the
MNsure online enrollment system. Thus, counties had to mail
information packets to enrollees for this purpose.
In mid-November 2014, counties assumed responsibility from DHS for

administering all applications for public health care programs submitted
in paper form. County staff expressed concerns that this new
responsibilitycombined with larger caseloads, workarounds delegated
to counties by MNsure, and plans to transition all people on public
programs to MNsure enrollment during 2015will stretch county
resources thin. Counties hired additional staff to help with MNsure cases
and received enhanced federal funding to help cover these costs, but
some believe that local property taxes will finance part of the increased
workload.31
County staff expressed concern that they had minimal training for using
the MNsure enrollment system and little participation in the systems
testing.
Some county staff expressed optimism that the new enrollment system, when
fixed, may be an improvement over previous processes. Counties like the fact
that some people enrolling in public programs through MNsure have been able to
complete the entire process on their own, without county help. But, on balance,
county staff we talked with were more negative than positive about MNsures
functionality so far. County staff work on the front lines of the enrollment
process, answering consumers questions and confirming their enrollment in
public programs. Because of limited access to the MNsure enrollment system
and inability to fix certain types of problems, it has been hard for county staff to
do their jobs.
31
At a November 2014 meeting of the MNsure Legislative Oversight Committee, the Director of
the Anoka County Economic Assistance Department testified that his countys administrative costs
for people in public health care programs rose $1 million in 2014 and is expected to increase
another $1 million in 2015.
Chapter 6: Operations
F
ederal rules require that states provide resources to help consumers make
informed decisions when obtaining health insurance through an exchange,
including: application assistance, a toll-free phone number, and outreach and
education.1 In this chapter, we discuss the design, implementation, and
performance of these resources. We also examine other operational issues from
the first year of enrollment, including advertising and data security.
Many consumers were referred back and forth between brokers

and navigators, due to differences in the roles and compensation
practices for these assisters.
MNsures contact center failed to provide adequate customer

service during the first open enrollment period.
MNsure has somebut not completeability to analyze who has

accessed private data on enrollees.
CONSUMER ASSISTERS
MNsure provides consumers with application and enrollment assistance through
its partnerships with external organizations and individuals. Exhibit 6.1 defines
MNsures three types of application assistersnavigators, certified
application counselors, and brokersand highlights some of the differences
between them.2
Navigators, which are typically employed by community-based organizations,
help consumers enroll in both public health insurance programs and commercial
products offered through MNsure. When consumers need help with an
application, they locate navigators through MNsures online assister directory.
MNsures contact center also refers consumers to navigators for help. Navigators
must help any MNsure user that requests their assistance and provide information
in a fair, accurate, and impartial manner.3 In order to maintain their impartiality,
navigators may not receive any compensation from an insurance company.4
45 CFR, sec. 155.205 (2014); and 45 CFR, sec. 155.225(a) (2014).
Throughout this report, we use the term consumer assisters to refer to all three types of
application assisters: navigators, certified application counselors, and brokers.
45 CFR, sec. 155.210(e)(2) (2014); and Minnesota Rules, 7700.0040, subp. 1B, posted July 11,
2013.
45 CFR, sec. 155.210(d) (2014); and Minnesota Rules, 7700.0030, subp. 1B, posted July 11,
2013.
102
Navigators must help consumers understand the differences between their health
insurance options, but they may not offer advice about which commercial
product to select.5
Exhibit 6.1: Overview of MNsure Consumer Assisters

Navigatorsa
Typical employer
Role
Per-enrollment
compensation
Certification
requirements
Must help anyone

who requests it
Listed in online
assister directory
Receive referrals
from contact center
Community organizations and

counties
Providing MNsure application
assistance and clarifying the
distinctions among public and
commercial health coverage
options, including describing the
benefits, terms, or conditions of
commercial plansc
MNsure pays $70 per commercial
enrollment; DHS pays $25 for
Medical Assistance and $70 for
MinnesotaCare enrollments
20 hours of training; passage of
exams and background check;
works for an organization contracted
by the state; compliance with
conflict of interest, privacy, and
security standards
Yes
Counselors
Brokersb
Hospitals or clinics
Insurance agencies
Providing MNsure
application assistance
to clients as part of
regular job duties
Selling insurance plans offered

on MNsure through MNsures
online application, and offering
advice about the benefits,
terms, or conditions of
commercial plansd
None
Same as navigators
Insurers pay brokers per

enrollment if the broker is
appointed to sell plans on
behalf of the companye
2 hours of training; passage of
an exam and background
check; verification of broker
license; compliance with
privacy and security standards
No
No
Yes
No
Yes
Yes
No
Yes
Throughout this report, we use the term navigator to refer to both in-person assisters and navigators. County workers may be certified
as in-person assisters and be paid for enrolling consumers in commercial plans, but they may not serve as navigators or be paid for
enrolling consumers in Medical Assistance or MinnesotaCare. Counties workers that serve as in-person assisters are not listed in
MNsures Assister Directory.
b
In this report, we use the term brokers to refer to agents and insurance producers licensed under Minnesota Statutes 2014, 60K.
45 CFR, sec. 155.210(e) (2014); Minnesota Rules, 7700.0040, posted July 11, 2013; and MNsure and the Minnesota Department of
Commerce, Consumer Assistance Partner Bulletin 2013-1 (St. Paul, 2013).
d
Minnesota Statutes 2014, 62V.05, subd. 3; Minnesota Statutes 2014, 60K.32; and Minnesota Statutes 2014, 60K.31, subds. 6, 12, 14,
and 15.
e
According to the Henry J. Kaiser Family Foundation, in 2010, brokers in Minnesota earned an average of $97 per enrollment or renewal
that they facilitated, or 3.9 percent of premiums in the individual market.
SOURCES: Office of the Legislative Auditor; 45 CFR, sec. 155.210 (2014); 45 CFR, sec. 155.215 (2014); 45 CFR, sec. 155.220 (2014);
45 CFR, sec. 155.225 (2014); and Minnesota Rules, chapter 7700, posted July 11, 2013, and September 26, 2013.
45 CFR, sec. 155.210(e)(2-3) (2014), requires navigators to clarif[y] the distinctions among
health care options, including [commercial plans] and to facilitate selection of a [commercial
plan]. According to Minnesota Statutes 2014, 60K.32, and Minnesota Statutes 2014, 60K.31,
subd. 15, individuals must be licensed as brokers in order to urg[e] a person to apply for a
particular kind of insurance from a particular company. Also, see MNsure and the Minnesota
Department of Commerce, Consumer Assistance Bulletin 2013-1 (St. Paul, 2013).
OPERATIONS
103
MNsures Navigator Program grew out of the Department of Human Services

(DHSs) Minnesota Community Application Agent Program, which was
established in 2007.6 Organizations that participated in the Minnesota
Community Application Agent Program were paid $25 for each person they
enrolled in a public health insurance program.7 In 2013, MNsure contracted with
a number of these organizations, as well as others, to provide similar services
under its new Navigator Program.8 During the first year of the exchange,
MNsure and DHS jointly ran the Navigator Program. MNsure used federal grant
funds to pay organizations $70 for each consumer enrolled in a commercial
product through MNsure, while DHS paid $25 for enrollments in Medical
Assistance and $70 for enrollments in MinnesotaCare.9 During MNsures first
year, navigators assisted about 12 percent of MNsures Medical Assistance
enrollees and MinnesotaCare enrollees, and 13 percent of MNsures commercial
plan enrollees.10
Federal rules also required states to develop a program for Certified Application
Counselors (CACs).11 CACs are often hospital or clinic employees who, as part
of their daily job duties, assist clients with health insurance applications. They
are not paid by MNsure or insurers for their services.12 Unlike navigators, CACs
are not listed in MNsures assister directory and are not obligated to help any
consumer that contacts them.
Licensed health insurance brokers also may be certified by MNsure to sell
insurance offered through the exchange.13 In 2010, insurers paid brokers $97, on
During the first year of operations, MNsures Navigator Program operated under the policies of
the Minnesota Community Application Agent Program, Minnesota Statutes 2014, 256.962.
Minnesota Statutes 2014, 62V.05, subd. 4, required the MNsure Board to establish policies for the
Navigator Program that would have become effective January 1, 2015. However, as of early
February 2015, the board had not adopted formal policies for the Navigator Program.
DHS stopped contracting with organizations under the Minnesota Community Application Agent
Program during the first year of the MNsure Navigator Program.
Federal rules prohibit the exchange from using federal grant funds to pay navigators for enrolling
consumers in insurance through MNsure but allow the exchange to use them to pay in-person
assisters. As a result, MNsure operates parallel navigator and in-person assister programs that are
identical except for their funding mechanisms. Internally, MNsure uses the term in-person
assisters to refer to organizations paid by MNsure using federal funds for enrolling consumers in
qualified health plans, and the term navigator to refer to organizations paid by DHS for enrolling
consumers in public programs. Most organizations serve as both navigators and in-person assisters.
Counties, however, are only paid for in-person assister services. Publicly, MNsure uses the term
navigators to refer to both types of assisters, and our report follows this practice, too.
10
Because of problems with MNsures enrollment and assister data, we could not independently
verify the exact number of consumers enrolled with the help of navigators. These figures represent
our best estimates given the data made available to us.
11
45 CFR, sec. 155.225(a) (2014).
12
45 CFR, sec. 155.225(g)(2) (2014); and Minnesota Rules, 7700.0020, subp. 5, posted July 11,
2013.
13
45 CFR, sec. 155.220(d) (2014); and Minnesota Statutes 2014, 62V.05, subd. 3(a). In this report,
we use the term brokers to refer to insurance agents or insurance producers licensed under
Minnesota Statutes 2014, 60K.
104
average, for each person they enrolled in a commercial product in Minnesota.14

In Minnesota, brokers must be appointed (authorized) by an insurance
company before selling that insurers products.15 For example, a broker who has
been appointed by HealthPartners and Blue Cross Blue Shield of Minnesota, but
not Medica, cannot sell a Medica plan to a consumer. When helping consumers
purchase insurance through MNsure, brokers must disclose which insurers pay
them to sell their products.16 Like navigators, brokers are listed in MNsures
online assister directory, and MNsures contact center staff refer callers to
brokers for help enrolling in a commercial product. Unlike navigators, brokers
are not obligated to help everyone who requests it and they are usually not paid
for enrolling consumers in public health insurance programs.
For the 2014 enrollment year, MNsure expected to spend 8 percent of its budget,
or around $10 million, on its consumer assister programs. This included
expenditures on per-enrollment payments to navigator organizations, grants to
organizations for outreach and application assistance, and staffing for program
administration.
Training and Certification

All of MNsures consumer assisters must be certified by MNsure before they can
help consumers enroll in insurance using the exchange.17 For MNsures first two
open enrollment periods, navigator and CAC certification required 20 hours of
training, passage of exams and a background check, and compliance with conflict
of interest, security, and privacy standards.18 Brokers, on the other hand, only
had to complete two hours of training. Like navigators and CACs, they also had
to comply with MNsures privacy and security standards.
MNsure did not provide sufficient training for its consumer assisters in the first year of
enrollment.
We surveyed a random sample of consumer assisters about their experiences with
the exchange during the first year.19 Fifty-three percent of navigators and
72 percent of brokers did not think MNsure provided sufficient training. CACs
were more satisfied, with almost two-thirds approving of the training. Assisters
were particularly concerned that they received no hands-on training with
14
According to data compiled by the Henry J. Kaiser Family Foundation, brokers in Minnesota
earned, on average, $97 per enrollment or renewal, or 3.9 percent of premiums in the individual
market in 2010.
15
Minnesota Statutes 2014, 60K.49, subd. 2.
16
Minnesota Statutes 2014, 62V.05, subd. 3(e).
17
Minnesota Rules, 7700.0060, posted July 11, 2013.
18
Organizations also had to sign a contract with the state before their employees could be certified
as navigators.
19
We received responses from 222 navigators (71 percent response rate), 167 CACs (57 percent
response rate), and 295 brokers (54 percent response rate).
OPERATIONS
105
MNsures online application. As one navigator said, The first time I saw the
application was with my first applicant.
Many assisters said they wanted more training on how to answer MNsures
application questions related to household income. Consumers trying to find out
if they qualified for public health insurance programs, tax credits, or cost-sharing
reductions had to estimate their taxable income for the coming year and disclose
it on MNsures application. But it can be challenging for individuals in certain
professions or life circumstances to estimate taxable income. For example,
farmers may have negative income over the course of a year, and seasonal
workers may not know how much they will make over the coming months.
Assisters said they were not adequately prepared to help consumers with these
types of questions. One certified application counselor said, We are not
accountants that know all tax deductions [available] for consumers.
MNsure had limited success getting assisters certified by the October 1, 2013,
launch date. On the first day of open enrollment, MNsure estimated that around
450 brokers and no navigators were certified.20 This meant that application
assistance was not available for many consumers when the exchange first
opened.
MNsure did not have an adequate method for tracking assister certification in its first year of
operation, and this hindered its ability to make accurate payments.
State rules require MNsure to maintain documentation of certification training for
navigators and CACs.21 MNsure staff provided us with the records they kept of
individuals that had been certified. Upon inspection, we found the records to be
incomplete. They did not properly document which individuals had successfully
finished various portions of the certification process and by what date. We also
could not distinguish individuals who had been certified and later dropped out of
the program from individuals who were never certified; both types of individuals
were labeled inactive in the records. Because MNsure did not have an
adequate method for tracking assister certification, we could not conclusively
determine the total number of assisters who were certified by MNsure during its
first year of enrollment.
Weak recordkeeping also made it difficult for MNsure to ensure that it made
accurate payments to navigator organizations. For example, MNsure intended to
give each navigator a unique identifier so it could (1) track enrollments that were
facilitated by navigators and (2) make payments to navigator organizations. But
MNsure assigned some identifiers to navigators that it had already allocated to
other navigators. This made it hard for MNsure to determine which navigator
should get credit for an enrollment. Also, at least two organizations were paid
20
These numbers are based on internal MNsure communication that could not be verified
independently by our office.
21
Minnesota Rules, 7700.0050, subp. 1, posted July 11, 2013.
106
over $6,000 for services provided by navigators that were not certified.22
MNsure was aware of a number of these issues, plus some others. The agency
intends to reconcile any over- or under-payments that were made to navigator
organizations, but staff told us this process might not be complete for a while.23
Navigator and Broker Roles

MNsure users options for getting enrollment assistance during the first year of
the exchange were, at times, confusing and inefficient. Because brokers only get
paid for enrolling consumers in commercial products, it may have been
reasonable to expect them to primarily work with consumers shopping for
commercial productsand, similarly, for navigators to primarily work with
consumers enrolling in public programs. But MNsures consumer assister
programs did not function this discretely in practice.
Many consumers were referred back and forth between brokers and navigators.
Some consumers who were eligible for public programs originally contacted
brokers for assistance and were subsequently referred to navigators. Likewise,
navigators referred consumers to brokers when the consumers needed help
choosing a commercial product. Poor customer service resulted from consumers
being referred back and forth between navigators and brokers. As one broker
said, There needs to be a way to identify the best way to help these people.
Sending them to two or three places before getting help is inefficient and they
should not have to deal with this.
One reason consumers were referred back and forth between navigators and
brokers was because navigators could not provide start-to finish application
assistance for all consumers, regardless of the type of coverage they chose.
Navigators could only provide limited assistance to consumers enrolling in commercial
insurance products.
Navigators are required by federal rules to clarify the distinctions among
commercial health insurance options for consumers.24 But, according to state
law, only licensed insurance brokers can urge a consumer to choose a particular
product.25 The difference between helping someone understand commercial
insurance options and giving them advice on specific products can be a fine
distinction. In 2013, MNsure and the Department of Commerce issued guidance
22
Staff told us they discovered that dozens of navigators helped consumers enroll before they
were certified. They said these navigators will not be paid for these enrollments.
23
When calculating payments owed to navigator organizations, MNsure compared its records of
assister certification with reports that were generated by its online enrollment system which
identified which consumers used assisters. Staff told us that problems with the online enrollment
system also contributed to the payment errors.
24
45 CFR, sec. 155.210(e) (2014).
25
Minnesota Statutes 2014, 60K.32, and 60K.31, subd. 15.
OPERATIONS
107
to clarify the kind of assistance navigators are allowed to offer consumers. The
guidance said that navigators could describe the benefits, terms, or conditions of
commercial health products, but they could not recommend a specific product or
make a decision on behalf of a consumeronly licensed brokers could do that.
If consumers who had sought help from a navigator wanted help with product
selection, they had to schedule another in-person visit with a broker.
The distinction between the help navigators could and could not give consumers
enrolling in commercial products was difficult for some consumers, contact
center staff, and assisters to grasp. As one navigator explained:
I send people to brokers when they need help in comparing and
choosing plans. Sometimes I get the idea that people are told by
those they call at MNsure that navigators are supposed to be able
to help them. I cant. I dont know insurance, and I dont want
to be a broker, I want to be a navigator. Sometimes folks are a
bit angry about this because theyve been told to contact a
navigator for all aspects of getting insurance.
Consumers who met with a navigator may have needed to visit a broker at a later
date. For example, some consumers did not know before they met with a
navigator and entered information into MNsures application whether they would
qualify for a public program. (If they did not qualify, they had to schedule a
separate meeting with a broker if they wanted help choosing a commercial
product.) Also, in some households, certain members of the household were
eligible for public programs while others were not. If households with mixed
eligibility wanted help enrolling in a public program and help choosing a
commercial product, they had to visit both a navigator and a broker.
Similarly, if consumers initially chose to meet with a broker instead of a
navigator, they could not be sure they would receive all the help they needed in
one visit. MNsure-certified brokers were required by federal rules to take
consumers through the application questions related to eligibility for public
programs, tax credits, and cost-sharing reductions before they could sell the
consumers a product.26 So, if consumers were deemed eligible for public
programs, some brokers referred them to navigators to complete the enrollment
process.
Some brokers helped consumers enroll in public programs, although they received no
compensation for doing so.
This was one of the most frequently cited concerns among brokers who
submitted written comments to our survey. One broker said, I think I assisted
about 15 to 20 individuals during the open enrollment period, none of which
[purchased] a [commercial insurance product]. I worked for hours without
making a dime. Another said:
26
45 CFR, sec. 155.220(c)(1) (2014). Consumers completing MNsure applications on their own
had the option of skipping these eligibility questions and simply enrolling in an unsubsidized
commercial health insurance product; brokers helping consumers did not have this option.
108
Clients came into the agency not knowing until we were into the
application process that they would qualify for public assistance.
I did the work, I provided the information, and I received no
compensation.
One reason brokers helped enroll so many people in public programs was
because it was the contact centers policy to refer callers whose incomes were on
the edge of public program eligibility to brokers, rather than navigators, when
they needed application assistance.
Also, brokers were not compensated for their assistance in cases where consumers
purchased commercial products through MNsure that the brokers were not authorized to
sell.
According to state law, brokers must be appointed (authorized) by an insurance
company before they are allowed to sell the companys products, for which they
are paid a commission.27 In the first year of the exchange, five insurance
companies sold products through MNsure. Neither state law nor MNsure
required each insurance company selling products through the exchange to
appoint each MNsure-certified broker. Some consumers who were receiving
help from a broker decided to purchase products their particular brokers were not
appointed to sell. As a result, some brokers did not earn a commission for their
services. For example, of the 295 MNsure-certified brokers we surveyed, only
34 percent were appointed with PreferredOne, the company that sold a majority
of the products during MNsures first year.28
Some brokers had the impression they would automatically be appointed by all of
the insurance companies selling products on MNsure once the brokers received
their MNsure certification. As one broker said:
I am extremely disappointed that we were not compensated by
the companies that we were not appointed with. [It] was not
made clear that we would only be paid if the consumer chose a
product through our appointed companies. I uploaded and
helped over two dozen families and was compensated for one of
them.
Overall, during the first year of the exchange, brokers were frustrated when they
were not compensated for assistance they gave to some public program and
commercial product enrollees. Some brokers told us they did not plan to seek
MNsure certification in the second open enrollment periodas one said, We
lost money trying to assist people. According to MNsures count, the number of
certified brokers dropped from about 2,300 in April 2014 to about 500 at the start
27
28
Minnesota Statutes 2014, 60K.49, subd. 2.
Seventy-eight percent of MNsure brokers we surveyed were appointed to sell products for Blue
Cross Blue Shield of Minnesota, 63 percent for Medica, 40 percent for HealthPartners, and 13
percent for UCare.
OPERATIONS
109
of the second open enrollment period.29 It is unclear what impact this decline
will have on MNsure enrollments. But, if MNsure determines that this smaller
number of brokers does not provide consumers with enough access to application
assistance and insurance advice, MNsure and DHS should consider whether to
pursue changes.
RECOMMENDATION
MNsure and DHS should ensure that brokers are fairly compensated for enrolling consumers in
insurance through MNsure.
A first option would be for DHS to use existing state law to pay MNsure-certified
brokers for public program enrollments. Under the Minnesota Community
Application Agent Program, brokers are eligible to receive $25 from DHS for
each person they enroll in Medical Assistance or MinnesotaCare.30 This program
is still authorized in state law, although DHS no longer operates it now that the
Navigator Program has taken its place. We were told that few brokers took
advantage of reimbursement under the Minnesota Community Application
Program. Perhaps brokers were not widely aware of this compensation option; if
so, DHS or MNsure could take steps to better publicize it to brokers. If the
compensation rate offered through this program is too low to attract brokers,
DHS or MNsure could seek an increase through legislative action.31
Second, the MNsure Board could consider requiring each insurer that sells plans
through MNsure to appoint each MNsure-certified broker to sell its products.
Currently, each insurer decides whether or not it will appoint a given broker to
sell its products. If the board decides to pursue this option, it could make this
requirement part of its approval process for insurers selling plans through
MNsure, or it could ask the Legislature for a change in state law.
Application Access for Assisters

When designing Minnesotas health insurance exchange, state officials intended
to build a navigator-broker portal that would allow assisters to view and edit
consumers applications on their behalf. Because MNsures navigator-broker
portal did not function during the first year of the exchange, assisters were only
29
According to MNsures reports, the number of brokers rose during the second open enrollment
period, reaching around 820 in January 2015.
30
31
This compensation rate is lower than others offered to brokers. In 2010, Minnesota insurers paid
brokers $97, on average, for a years enrollment or renewal in the individual market, according to
data we obtained from the Henry J. Kaiser Family Foundation. Under the Minnesota
Comprehensive Insurance Act (the former program for people who could not find insurance
elsewhere), brokers received $50 per enrollee from insurers. Also, Minnesota insurers paid brokers
from $120 to $410 per Medicare enrollee, according to data published in 2013 by the Centers for
Medicare and Medicaid Services.
110
able to view an account if the consumer they were assisting was physically
present to log in through MNsures website.32
Assisters and MNsure staff told us that many problems resulted from the lack of a functional
navigator-broker portal.
Because assisters could not use the navigator-broker portal, it was difficult for
them to provide follow-up services. Follow-up is required, for example, when
the enrollment system does not produce an immediate determination of public
program eligibility. In such cases, assisters often had to schedule follow-up inperson appointments. The largest share of navigators and CACs said it took one
to two hours, on average, for them to complete the online portion of enrollment
for consumers enrolling in Medical Assistance or MinnesotaCare.33 Forty
percent of navigators said it took more than two hours, on average, to enroll
consumers in commercial health plans.34 But as one navigator explained, the
application did not account for all of their time; they also spent a lot of time
following up on consumers applications:
My major gripe is that navigators were not warned that there
would be considerable follow-up required. I work with nonEnglish speaking people and we continue to spend time calling
[various help lines] trying to get status updates and figure out
why things are delayed. Navigators become the clients primary
contact, which shifts a lot of work to us. I would say the
applications themselves only account for a third [of] the time
navigators put [in] trying to get the application processed.
At times, assisters had to schedule multiple appointments to finish an enrollment, which
was a burden for some consumers.
Sometimes a single appointment was not sufficient to complete a MNsure
application. For example, some assisters encountered long contact center wait
times, problems connecting with the Federal Data Services Hub (which verified
consumers identities), or problems with the MNsure website. One navigator
described the toll this took on consumers, saying, We are working with clients
who have low income, most often with transportation barriers. These clients
cannot afford additional visits to our office. Another said:
32
As of February 2015, the navigator-broker portal was still not functional.
33
Fifty-one percent of navigators and 53 percent of CACs said it took, on average, one to two hours
to complete the online portion of enrollment in Medical Assistance through MNsure. Fifty-three
percent of navigators and 56 percent of CACs said it took this amount of time for MinnesotaCare
enrollments. Also, 50 percent of navigators and 51 percent of CACs said it took one to two hours
for enrollments in qualified health plans.
34
Navigators said commercial plan enrollments generally took longer than public program
enrollments. Sixteen percent of navigators said it took more than two hours for Medical Assistance
enrollments and 18 percent said it took more than two hours for MinnesotaCare enrollments
through MNsure.
OPERATIONS
111
As a navigator working in a rural area, most of my consumers

[have to] travel long distances to meet with me. It was very
frustrating that some people had to schedule multiple
appointments to get the application complete.
Some parts of Minnesota had relatively few navigators, which increased the
enrollment challenges in those areas. According to MNsures enrollment data,
navigators did not successfully enroll a consumer in 30 of Minnesotas 87
counties. In another 23 counties, fewer than one navigator per 1,000 uninsured
residents successfully enrolled a consumer.
Consumer Assister Grants

In its first year of operation, MNsure awarded $4.75 million in competitive grants
to 41 organizations, as seen in Appendix F. The grants had two purposes. First,
they supported navigator organizations that provided application assistance to
consumers.35 For example, they helped organizations purchase computers, hire
navigators, and pay for navigators to travel around communities. Second, the
grants supported organizations outreach efforts to targeted populations,
including minorities, rural populations, and small employers. Some
organizations received funds for outreach activities only and did not provide
application assistance.
Soon after MNsure announced the grant winners, members of the public and
Legislature criticized the agency for not awarding funds to any of the primarily
African-American organizations that had applied, despite the fact that African
Americans are disproportionately likely to be uninsured.36 In response to the
criticism, MNsure reallocated funds and awarded an additional $833,000 to 12
organizations from its original pool of applicants. The organizations that were
awarded funds in the second round represented several demographic groups that
were initially underrepresented.
Many MNsure grantees failed to reach their outreach or enrollment goals.
Each of the grantees contracts, which ran through September 2014, included
enrollments goals developed by the organizations and approved by MNsure.37
Some grantees performed well in terms of their enrollment goals, as shown in
Appendix F. For example, Hmong American Partnership enrolled nearly 2,200
35
Organizations could be certified as navigator organizations without receiving grant funds; most
did not receive grant funds during MNsures first year.
36
According to MNsures assessment of the proposals, 6 of the 29 original grantees intended to

serve African Americans in their outreach or enrollment activities.
37
Three grantees received funds for outreach activities only; they did not establish enrollment goals
in their contracts with MNsure. None of the contracts for the first round of grantees was finalized
until after open enrollment started in October 2013; contracts for the second round were not
completed until late November 2013.
112
people in insurance using MNsure, which was 211 percent of its goal.38 Portico
Healthnet enrolled by far the most people for a single organization (6,162).
Some organizations, like Minnesota Community Action Partnership, enrolled a
large number of people (2,786), but still fell far short of their goalsperhaps
indicating the challenge organizations had setting goals during the first year of
the Navigator Program. Grantees enrolled about 70 percent of the consumers
that received assistance from a navigator between October 2013 and September
2014.
Some grantees simply did not enroll many people. For example, the National
Association of Mental Illness (NAMI Minnesota), the Confederation of Somali
Community in Minnesota, Springboard for the Arts, and the Minnesota
Chippewa Tribe together received more than $400,000 but together enrolled only
175 people, according to MNsure records. Each of these organizations enrolled
fewer than 10 percent of its individual goal. For the second year of open
enrollment, MNsure awarded another $4.6 million in grants to 28 organizations.
Several organizations that performed poorly during the first year were awarded
funds again for the second year, as seen in the first table in Appendix F.
Grantees contracts also included outreach goals. Grantees used a variety of
methods to achieve their outreach goals. For example, Springboard for the Arts
produced a series of informational videos which were distributed to its members.
Other organizations staffed booths at community events, gave presentations to
organizations, visited food shelves, sent out mailings, wrote articles for
newsletters, and performed door-to-door canvassing. Grantees submitted
monthly and final reports listing the outreach events they held, the number of
people reached, and the number of people they helped to enroll.
MNsure provided insufficient oversight of its consumer assister grants.
Even though 30 grantees failed to meet their outreach or enrollment goals,
MNsure did not withhold any grant payments for performance reasons.39 Staff
told us that when MNsure made final payments to grantees at the end of 2014,
they did not take into account grantees contracted enrollment goals or their
enrollment performance. Staff said they did consider organizations outreach
performance when making final grantee payments. But even grantees that did
not meet their outreach goals received their full award by the end of the grant
period.
In addition, the grantee data related to outreach activities were self-reported and,
during the first year of operations, MNsure did not verify the occurrence of the
outreach events listed in grantees reports, the actual number of people reached
38
As noted earlier in this chapter, problems with MNsures enrollment and assister records
prevented us from independently verifying the exact amounts earned by navigator organizations.
The figures cited in this section represent our best estimates.
39
MNsure and one grantee (Small Business Minnesota) mutually agreed to terminate the grantees
contract in June 2014 because problems with the functionality of MNsures small business
application would have made it difficult for the organization to fulfill its grant duties. Small
Business Minnesota returned all of the funds it was awarded.
OPERATIONS
113
by the events, or the quality of such efforts. Some of the grantees self-reported
data, which is shown in Appendix F, seemed implausible. For instance,
Minnesota Community Action Partnership, whose contract included 15 partners,
estimated that it reached 2.8 million people through its outreach eventswhich is
over half of the states population. This figure also represented 28,000 percent of
the organizations contracted outreach goal. Women Venture, which did not
have any grant partners, reported that it reached 674,000 people, or 48,000
percent of its outreach goal.
We think MNsure should have provided more oversight of grantees outreach
activities. MNsure staff told us they acknowledged this shortcoming and for the
second year of operations, MNsure hired three additional staff to oversee grant
activities and provide grantees with greater support services. We also think
MNsure should have considered grantees performance in relation to both their
contracted outreach goals and their contracted enrollment goals when evaluating
grantees final performance.
CONTACT CENTER
According to federal rules, MNsure must provide a toll-free call center to address
the needs of consumers requesting assistance.40 In this section, we review the
development and performance of the MNsure contact center and the various
problems it has faced since opening on September 3, 2013. We also review how
the contact center functions within the states existing network of customer
service resources.
Customer Service Standards

In the first year of operation, MNsure staff established two primary customer
service goals and tracked the contact centers performance related to those goals.
First, the contact center sought to answer more than 85 percent of calls in less
than two minutes.
MNsures contact center failed to answer calls within an acceptable amount of time.
Exhibit 6.2 shows the average wait time callers experienced each month, from
October 2013 to August 2014. As shown in the exhibit, the contact centers
average wait times were extremely high in November (23 minutes), December
(60 minutes), and January (40 minutes). Wait times declined in the spring, and
even fell below the contact centers two minute goal in May; but they rose again
in the summer months, nearing 25 minutes in August. Across MNsures first
open enrollment period (October 2013 through March 2014), the average amount
40
45 CFR, sec. 155.205(a) (2014).
114
Exhibit 6.2: Average Amount of Time Consumers

Spent Waiting on Hold, October 2013 to August 2014
Minutes
70
60
50
40
30
20
10
Goal
0
Oct.
2013
Nov.
Dec.
Jan.
2014
Feb.
Mar.
Apr.
May
June
July
Aug.
NOTES: These figures include wait times for both MNsures contact center agents and contracted
agents. They do not include the amount of time consumers spent waiting on hold after their calls
were answered. The contact centers goal was to answer 85 percent of calls within 2 minutes.
Contracted customer service agents were supposed to answer 85 percent of calls within 30 seconds,
according to the vendors contract with MNsure.
of time consumers waited on hold was 24 minutes. Some consumers had

especially long waits. For example:
On December 20, 2013, the average wait time was 1 hour and
49 minutes, the longest average over the course of a single day.
The longest wait for an individual caller, which occurred on

December 18, 2013, was 2 hours and 39 minutes.
On 78 percent of the
Assister Responses to the Following Survey
days the contact
Statement: Wait times to speak with a call center
representative were reasonable.
center was open
between October 1,
Navigators
CACs
Brokers
2013, and August 29,
Always or almost always
2%
6%
3%
2014, the average
Often
7
10
2
daily wait time
Sometimes
35
30
14
exceeded MNsures
Rarely or never
52
47
77
two-minute goal. Not
surprisingly, MNsure users were overwhelmingly dissatisfied with wait times. A
majority of brokers and navigators we surveyed said wait times were rarely or
never reasonable.
OPERATIONS
115
MNsures second customer service goal was to have less than 5 percent of callers
hang up before speaking to a customer service agent. MNsure failed to achieve
this goal for abandoned calls.
Many people abandoned their calls before speaking to a customer service representative.
With such long wait times, many people hung up before speaking to a contact
center agent. Between October 2013 and August 2014, about one-third of callers
abandoned their calls. Exhibit 6.3 shows the percent of calls abandoned each
month, compared with MNsures goal during that time period. Only in May
2014 did the call abandonment rate fall below 5 percent. In January 2014,
35,000 calls (60 percent) were abandoned. The highest rate of abandoned calls
for a single day occurred on January 3, 2014, when 81 percent of calls were
abandoned. On the last day of open enrollment (March 31, 2014), 4,063 calls
were abandonedthe most calls for a single day.41
Exhibit 6.3: Percentage of Phone Calls to the MNsure

Contact Center Abandoned by Callers, October 2013
to August 2014
70%
60%
50%
40%
30%
20%
10%
Goal
0%
Oct.
2013
Nov.
Dec.
Jan.
2014
Feb.
Mar.
Apr.
May
June
July
Aug.
NOTES: MNsures goal was to have less than 5 percent of callers hang up before their calls were
answered.
41
The data we reviewed did not capture the number of calls that were disconnected because the
contact center was too overloaded to handle them.
116
Design
A number of factors contributed to the contact centers customer service failures
during its first year of operation. One factor was that state officials did not
accurately predict the number of staff the contact center would need when
designing MNsures customer service strategy.
In February 2013, the Department of Human Services (DHS) assigned three staff
(which it called the Customer Service Team) to develop a customer service
strategy for the exchange that could be ready by September 2013.42 One of the
teams guiding principles was to design a customer service system that
provided the same level of service to all consumers, whether they purchased
commercial health insurance products through MNsure or enrolled in public
programs.43 The team reviewed existing state and county call centers and
technologies and found systemic problems that often result in poor service for
program participants. In response, they recommended that MNsure build a
state-operated call center that could serve as a front door and one-stop shop
for all of the exchanges health insurance options. In this model, they
recommended that MNsure and DHS collaborate to provide seamless service
between the contact center and existing customer service resources at DHS.
The Customer Service Team
Initial Plan for Contact Center Staffing, 2013
also designed the contact
Tier I
centers staffing structure.
15
basic contact center agents
They recommended that the
contact center transfer a
Tier II (specialists)
portion of its calls using an
2 Tribal enrollment specialists
1 Medical Assistance enrollment specialist
interactive voice response
1 Tax credits, brokers, and small employers specialist
(IVR), which routes calls
1 Public programs and childrens benefits specialist
based on a consumers
1 Complaints, grievances, and appeals specialist
selection from a menu of
1 Pharmacy and carrier network specialist
options. The rest of the calls
Total
would be handled or
22 contact center agents
transferred by customer
service agents, separated into two tiers. According to the teams design, Tier I
agents were supposed to answer basic questions and refer callers to consumer
assisters or other call centers. Tier II specialists were supposed to handle more
complicated cases and complaints and be trained in specific areas, like tax credits
or childrens benefits.
The team recommended that MNsure hire a total of 22 customer service agents,
with up to 7 specialists. They based this number on: historical call data from
42
According to their report, MNsure Customer Service Recommendations Focused on Contact

Center (St. Paul, April 2013), the three staff assigned to the Customer Service Team were
temporarily reassigned from the Minnesota Department of Human Services Continuing Care
Administration and had experience implementing and managing the Senior LinkAge Line and the
Minnesotahelp NetworkTM, which are Minnesotas aging and disability resource centers.
43
These guiding principles were laid out in Minnesota Health Insurance Blueprint Application:
2.0 Consumer and Stakeholder Engagement and Support (St. Paul, November 2012), 2. MNsure
Customer Service Recommendations Focused on Contact Center (St. Paul, April 2013), 3.
OPERATIONS
117
DHS call centers; assumptions about MNsures usage rates; and standard
forecasting methods, which assumed calls would be answered in two minutes or
less, on average. Overall, the team projected that the call center would receive
around 1,060 calls per day during the open enrollment period.
State officials underestimated contact center call volumes and staffing needs during the first
year of the exchange.
In June 2013, MNsure hired a contact center manager to implement the Customer
Service Teams plans and recommendations. Early on, contact center staff were
concerned about the accuracy of the teams call volume and staffing projections.
Staff noted, for example, that MNsures contact center was slated to handle calls
from a wider variety of programs, with more complex cases, than the DHS call
centers the Customer Service Team used as points of comparison. The teams
plan also did not consider how staffing might be affected if the exchange
experienced any technical problems, which, by this point, MNsure staff knew
was a possibility. When contact center staff expressed their concerns to
MNsures leadership, they were given authority to hire five additional customer
service agents. But contact center staff told us that even if the online application
had worked perfectly, MNsure probably would have needed more than the
number of customer service agents it initially hired.
Exhibit 6.4 shows the call volume and staffing levels the Customer Service Team
projected for MNsures first open enrollment period and for the remainder of the
year, compared with actual levels. It also shows the contact centers actual
customer service performance (in terms of wait times and abandoned calls)
compared with the Customer Service Teams projections. As shown in the
exhibit, the contact centers actual call volumes and staffing needs were much
higher than the Customer Service Team projected. Over the course of open
enrollment, for example, MNsure received more than twice as many calls per
month, on average, as it anticipated. Actual wait times and call abandonment
rates were also much higher than the team predicted.
MNsure waited until well into the first open enrollment period to contract with a vendor to
provide extra customer service staff.
The Customer Service Team developed its staffing recommendations under the
assumption that MNsure would hire a vendor to provide extra agents during the
initial open enrollment period, and as needed thereafter. MNsure staff
considered one vendors proposal before October 2013, but they deemed it too
expensive. The agency did not hire a vendor to provide extra customer service
staff until February 2014, less than two months before the end of the first open
enrollment period.44
44
The vendors customer service agents began training on February 5 and answering calls on
February 13. However, MNsures contract with the vendor was not officially signed until
February 21, 2014. All of these agents were trained at the Tier I level and answered calls at a
separate site.
118
Exhibit 6.4: Projected and Actual Contact Center

Staffing and Performance, October 2013 to August
2014
(Open Enrollment Period)
Oct. 2013 to Mar. 2014
Projected
Actual
Average monthly call volume
Average number of agents
Average number of minutes on hold
Average percentage of calls
abandoned
27,560
22
0.5
8.8%
Apr. to Aug. 2014

Projected
Actual
58,143
a
80
24.3
37.5%
3,782
6
0.6
9.3%
37,269
76a
11.3
25.8%
NOTES: MNsure projected the contact center would receive 1,060 calls per day during open
enrollment and 163 calls per day during the remainder of the year. Between October 1, 2013, and
March 31, 2014, the contact center was open on 156 days; between April 1 and August 29, 2014, the
contact center was open on 116 days. We did not include September 2013 data in these figures
since the exchange had not yet opened.
a
This figure includes both MNsure contact center agents and vendor agents.
SOURCES: Projection figures by MNsure Customer Service Recommendations Focused on Contact

Center (St. Paul, April 2013). Actual 2014 data was provided by MNsure with analysis by the Office
of the Legislative Auditor.
The month before the back-up vendor was put in place, callers were waiting on
hold an average of 40 minutes, and nearly 60 percent gave up before ever talking
to a customer service agent. As seen in Exhibit 6.5, the new vendor added over
100 agents at the end of February and in March 2014, and fewer in April, May,
and June. Although wait times and abandon rates greatly improved in these
months, as seen in Exhibits 6.2 and 6.3, the contact center still failed to achieve
its customer service goals for most of this period.
Knowledge and Training

The quality of the contact centers staff and training also contributed to MNsures
customer service failures during the first year of operation.
The contact centers training was insufficient, and most callers were dissatisfied with the
quality of information they received from customer service agents.
When MNsures contact center manager was hired in June 2013, staff had three
months to develop curriculum and hire and train customer service agents. Staff
told us the recruitment process was rushed, resulting in a limited number of
qualified candidates. They also said the contact centers training was inadequate.
This was in part because contact center agents, like MNsures consumer assisters,
did not have access to MNsures online application during their training. Most
OPERATIONS
119
Exhibit 6.5: Contact Center Staffing Levels,

September 2013 to August 2014
Month
September 2013
October
November
December
January 2014
February
March
April
May
June
July
August
Estimated
Number of
Staff Needed
6
22
22
22
22
22
22
6
6
6
6
6
Actual Number of Staff

(Monthly Average)
MNsure
Vendor
Total
27
30
29
42
58
54
58
55
48
45
43
42
108a
102
69
49
31
-
27
30
29
42
58
162
160
124
97
76
43
42
NOTE: MNsures contact center opened on September 3, 2013, in advance of the first day of the
exchange, October 1, 2013.
a
Vendor staff did not begin taking calls until February 13, 2014.
SOURCES: Estimated Number of Staff Needed was taken from Customer Service
Recommendations focused on Contact Center (St. Paul, April 2013). Actual staffing data were
provided by MNsure.
contact center agents accessed MNsures system for the first time when the
exchange opened on October 1, 2013.45
In our surveys of
Assister Responses to the Following Survey
consumers and
Statement: Call center representatives were
knowledgeable.
consumer assisters,
we asked them about
Navigators
CACs
Brokers
their experiences with Always or almost always
8%
9%
2%
MNsures over-theOften
27
25
13
46
phone help. While
Sometimes
49
40
48
Rarely or never
11
19
31
most navigators and
brokers we surveyed
said customer service representatives were courteous, a majority said they were
sometimes or rarely or never knowledgeable.47 As one certified application
counselor said, Call center representatives [were] polite but initially only able to
45
As of mid-September 2014, contact center agents still did not have access to a training version of
MNsures online application.
46
As we discuss later in this chapter, some navigators and certified application counselors did not
understand the difference between the MNsure contact center and the Assister Resource Center (a
phone support line for navigators and CACs). So, in our surveys, we asked them to evaluate their
experiences with any of the phone support they received. We asked brokers to evaluate their
experiences with the Broker Line at the MNsure contact center.
47
Seventy-one percent of navigators, 70 percent of CACs, and 63 percent of brokers said customer
service representatives always or almost always or often provided courteous service.
120
offer empathy rather than answers. The majority of consumer assisters also said
customer service representatives sometimes or rarely or never resolved their
questions or concerns the first time they called.48
In addition, more consumers rated the help
Surveyed Consumers Ratings of
they received from the contact center as
the Help They Received from the
poor or very poor than good or very Contact Center
good.49 Many people also told us they
Very good
6%
Good
12
received conflicting information from
Fair
14
various customer service agents. As one
Poor
11
consumer described, [Each representative]
Very poor
25
told me a different answer and often passed
Did not use/no opinion
31
me along with no direct answers or, worse
yet, inaccurate information. Contact center agents struggled to provide callers
with helpful information, at least in part, because MNsure leadership did not
provide agents with talking points or solutions related to callers technical
problems.
Technical Calls
During the first year of operations, MNsure did not have an adequate strategy for handling
callers technical questions.
When consumers began experiencing major technical problems with MNsures
online application in the fall of 2013, the contact center became inundated with
calls. Exhibit 6.6 shows the volume of calls MNsure expected compared with the
actual number of calls it received from October 2013 to August 2014. When call
volume skyrocketed, the contact centers two-tiered staffing structure dissolved
and all available staff began handling as many calls as possible.
Even if the contact center had more agents, they would not have been able to
resolve callers technical questions because none of the Tier II specialists were
trained to handle them. In fact, MNsure had no available technical staff to which
contact center agents could refer technical calls. As a result, contact center
agents with no technical training were left helping consumers with technical
problemswhich, from October 2013 to April 2014, was the contact centers
second most common type of call.
48
Sixty-one percent of CACs, 67 percent of navigators, and 85 percent of brokers we surveyed said
call center representatives sometimes or rarely or never resolved their concerns so they did not
have to make a second call.
49
Thirty-one percent of consumers said they did not use the contact center or had no opinion about
it in their survey responses.
OPERATIONS
121
Exhibit 6.6: Contact Center Call Volume, October

2013 to August 2014
Number of Calls
140,000
120,000
100,000
80,000
Actual
60,000
Projected
40,000
20,000
0
Oct. Nov. Dec. Jan. Feb. Mar. Apr. May June July Aug.
2013
2014
SOURCES: Projection figures by MNsure Customer Service Recommendations Focused on Contact

Center (St. Paul, April 2013), 11. Actual 2014 data was provided by MNsure with analysis by the
Office of the Legislative Auditor.
Contact center staff were not satisfied with the temporary strategies MNsure
leadership employed over the course of the first year to deal with these types of
calls.50 Starting in November 2013, contact center agents could notify a technical
team of DHS, Office of MN.IT Services, and Maximus staff about callers
technical issues.51 If a consumer called in after receiving an error message or
getting stuck, contact center agents recorded the issue (which they called
creating a ticket) and sent the ticket to the team. But contact center staff said
this arrangement did not work well because the technical team did not directly
take calls. In addition, when the team resolved issues, they did not inform the
contact center or the callers. Contact center staff said they struggled throughout
the first open enrollment period to get information from technical staff about
fixes that had been developed. At one point, contact center agents were
instructed to simply tell callers to call back in two weeks in the hopes that their
technical issues had been resolved.
By December 2013, MNsure relocated some technical team members to the
contact center to help track technical issues and improve communication about
solutions that were available. But the technical team members still did not
50
MNsure hosted a temporary conference call (called a bridge) that contact center leadership
could call during the first week of the exchange to get answers to technical questions. When the
exchange opened, this was the only support system the contact center had for handling callers
technical questions.
51
Maximus was one of the firms in charge of building MNsures technical infrastructure.
122
handle calls directly, and contact center staff said this was not very effective. By
April 2014, the technical team left the contact center, and customer service
agents remained without adequate technical support during the summer of 2014.
Contact center staff told us that establishing an adequate mechanism for handling
callers technical issues was their biggest priority for MNsures second year.
Shortly before the second open enrollment period began, MNsure made available
two e-mail addresses the contact center leadership could use to get answers about
callers technical issues. But contact center staff told us in January 2015 that this
strategy was still insufficientdescribing it as a last minute Band-Aid again this
year.
Coordination with Other Resources

Another reason customer service was inadequate during MNsures first year was
callers did not receive seamless referrals from the contact center or experience
consistent customer service from the states various customer service resources.
State officials established several guiding principles for the state to follow
when developing the exchanges customer service approach.52 The guiding
principles included:
Simplify[ing] the consumer experience by providing seamless

customer service and referrals between the contact center and other
customer service resources.
Ensur[ing] that all consumers have the same service experience,

whether they seek coverage in a public program or a commercial health
insurance plan.
MNsures contact center was not well coordinated with other customer service resources.
Minnesotas health insurance exchange is connected to several public programs
as well as private health insurers, so the contact centers referral protocols are
complex. For example, the contact center refers calls about Medical Assistance
to DHSs Minnesota Health Care Programs Member Help Desk, calls about
MinnesotaCare to the DHS MinnesotaCare Operations call center, and calls
about specific health insurance products to insurers.53 It also refers consumers to
hundreds of individual brokers and navigators and to many other state and county
resourcesall of which adhere to different hours, policies, and customer service
52
Minnesota Health Insurance Exchange Blueprint Application Documentation, 2.0 Consumer and
Stakeholder Engagement and Support: Contact Center (St. Paul, November 2012), 2; and MNsure
Customer Service Recommendations Focused on Contact Center (St. Paul, April 2013), 3.
53
Contact center agents transfer and refer callers to many different customer service resources,
including: all of the health and dental plans sold on MNsure; health care providers; hundreds of
brokers, CACs, and navigators; the County Line (which refers callers to county human service
agencies); tribal agencies; the Disability Linkage Line; the Senior LinkAge Line; LinkVet; the
Assister Resource Center; the Minnesota Health Care Programs Member Help Desk;
MinnesotaCare Operations; the Provider Help Desk; the Managed Care Ombudsman; the MMIS
Resource Center; the Minnesota Family Planning Program; and the State Medical Review Team.
OPERATIONS
123
standards. This makes it challenging for the contact center to provide seamless
referrals and to ensure that all callers received the same level of service.
Many consumer assisters we surveyed said there was not enough coordination
between the contact center and other customer service resources. As one broker
explained, The disconnect between the carriers, MNsure, and the county was
very painful. Similarly, a navigator said, DHS would tell us one thing and
MNsure another. For example, when consumers went through the application
process, were determined eligible for a DHS-administered public program, and
then got stuck in the online system, DHS often referred them back to MNsures
contact center for help. But MNsure staff told us that once consumers were
deemed eligible for a public program, they were sent to DHS. Thus, there was a
lack of clarity about which agency was responsible for handling these issues.
The majority of navigators and brokers said customer service representatives
only sometimes or rarely or never provided appropriate referrals when they
did not know the answer to a question.54 Some navigators and brokers said call
center representatives from various agencies referred consumers to them when
representatives did not know the answer to a question. One navigator said, Call
centers often just passed the buck. [They] would refer clients to navigators when
they should have answered clients questions.
One source of the confusion between MNsures contact center and other call
centers was that each had different degrees of access to consumers information.
For example, although DHS and MNsure have an agreement to share consumers
data, DHS had notas of January 2015granted contact center agents access to
the DHS public program information systems that contained the data. This
access would make it easier for MNsure contact center staff to identify and
resolve some callers issues without transferring them to another call center.55
Counties also had a different degree of access to consumers case information.
As one certified application counselor explained, The countycant see what
MNsure sees [on an application].
RECOMMENDATION
MNsure should improve its referral guidance for customer service staff in its contact center and
in DHS and county call centers.
Over the course of the contact centers first year, staff developed dozens of
protocols diagramming when, where, and in what manner calls should be
transferred to and from the contact center. They also developed some guidance
about the kinds of calls other call centers could handle. But, judging from
concerns we heard from users, these efforts were insufficient.
54
Forty-nine percent of certified application counselors, 57 percent of navigators, and 75 percent of
brokers said customer service representatives sometimes or rarely or never made appropriate
referrals when consumers asked questions for which they did not have answers.
55
Also, when MNsure contact center staff have called DHS for information on specific cases, DHS
has not provided this information verbally.
124
Phone Support for Assisters

MNsures consumer assisters often called the contact center when they needed
help assisting a consumer with an application, but they also got help from
MNsures assister support linesthe Broker Line and the Assister Resource
Center (ARC).56 Originally, only one of the contact centers Tier II specialists
was trained to assist brokers, but this eventually increased to about five agents.
ARC was a separate call center dedicated to answering calls from navigators and
CACs. During the first year of MNsure operations, ARC was located within
DHS and run jointly by MNsure and DHS.57 By July 2014, ARC customer
service staff were handling around 175 calls per day. A number of navigators
and CACs we surveyed said they were more satisfied with the help they received
from the Assister Resource Center than with the contact center.
To access ARC or the Broker Line, consumer assisters called MNsures toll-free
number and selected an unlisted (or hidden) option from the menu. However,
some assisters did not dial in correctly and waited in the contact centers long
lines rather than being directed to ARC or the Broker Line. Others were not
aware that MNsure provided specialized phone support for them.58
For the second year of operations, MNsure reorganized the way it provided
phone support to its consumer assisters. The agency installed dedicated phone
lines for ARC and the Broker Line, so assisters no longer have to dial into
MNsures main line. This may eliminate some of the confusion assisters
experienced. MNsure also trained more staff to support the Broker Line and
relocated ARC to the MNsure contact center.59 This move allowed MNsure to
align the hours of operation for the two groups of customer service
representatives. It may also make training more uniform for representatives
helping consumer assisters and make customer service more consistent.
Changes
The contact center failed to provide adequate customer service to callers during
the first year of the exchange, but MNsure made changes to address some of its
problems for the second year. For example, by November 15, 2014, the first day
of the second open enrollment period, MNsure had a total of 239 customer
service agents taking callsa dramatic increase from the 27 agents that started in
the fall of 2013. A majority of the customer service agents answering calls in the
56
The Broker Line was staffed by MNsure contact center agents and was a part of the contact
center.
57
Before the exchange was implemented, the Assister Resource Center supported the Department
of Human Services Minnesota Community Application Agent Program, the precursor to MNsures
Navigator Program.
58
Also, some assisters unwittingly dialed into the Broker Line or ARC and incorrectly thought they
were speaking to regular contact center agents.
59
As of January 2015, MNsure considered ARCs relocation to the contact center a pilot project.
OPERATIONS
125
second open enrollment period were employed by a vendor.60 According to

MNsure staff, the average wait time in November 2015 was 3 minutes and 30
seconds; in December it was 5 minutes and 45 seconds.61 We did not review in
detail the impact of MNsures changes on the contact center services provided to
consumers and assisters during MNsures second enrollment period.
ADVERTISING
MNsure was created to help enroll uninsured people and, more generally, to help
Minnesotans make informed choices about health insurance. Consequently,
MNsure leaders tried to find ways to convey their messages to a wide array of
Minnesotans. During planning for the exchange, a work group offered advice on
possible approaches:
The Outreach Work Group acknowledges the value of utilizing
all marketing tactics to ensure an effective marketing campaign
across the entire audience. The group strongly feels a larger
effort should be expended on grassroots outreach through
organizations that already serve their community rather than
mass advertising.62
Earlier, we discussed MNsures consumer assistance grants, which provided
some of the grassroots outreach referenced above.
In addition, exchange officials entered into other contracts for marketing and
outreach in preparation for the first open enrollment period. The main contracts
are shown in Exhibit 6.7. For example, the exchange spent more than $200,000
to research, test, and develop a brand. As stated in law, a purpose of the
exchange is to establish and modify as necessary a name and brand for [the
exchange] based on market studies that show maximum effectiveness in
attracting the uninsured and motivating them to take action.63 Exchange
officials hired a vendor that developed options for the exchanges name and
designed possible logos.64
60
According to MNsure staff, about 75 MNsure contact center agents and 164 vendor agents were
taking calls when the second open enrollment period began on November 15, 2014. At that time,
the contact center also had about 23 permanent staff performing work other than answering calls.
61
In the second open enrollment period, the contact centers automated voice recordings no longer
provided callers with estimated wait times, in contrast to contact center practice during much of the
first period. Also, the contact centers prerecorded messages in the second period gave some
inaccurate information. For example, on December 13, 2014, a recorded call center message said
that over 3,000 assisters were available to help consumers. But on that date, according to
MNsures online assister directory, MNsure only had 1,335 certified assisters.
62
Outreach, Communications and Marketing Work Group, Report to the Health Insurance
Exchange Advisory Task Force (St. Paul, December 11, 2012), 29.
63
64
The 2013 law that created the exchange called it the Minnesota Insurance Marketplace. The
Legislature adopted the MNsure name into law later in the 2013 session.
126
Exhibit 6.7: MNsures Key Marketing and Outreach

Contracts for First Open Enrollment Period
Vendor
Contract Dates
Salter Mitchell
April 2012-July 2012
Himle Rapp
October 2012December 2012

November 2012October 2013
April 2013-March 2014
Haberman
BBDO
SideDish
Total
January 2014-March 2014
Key Focus
Research on potential
audiences and messages
Public relations plan
Branding the exchange
Outreach, especially the
advertising campaign
Testimonial radio/TV ads
Amount
$ 165,901
34,000
210,000
1,592,047
112,900
$2,114,848
SOURCE: Office of the Legislative Auditor, based on review of MNsure documents.
The largest contract shown in Exhibit 6.7 (about $1.6 million) was primarily for
development and purchase of advertising.65 Shortly before open enrollment
began in October 2013, MNsure began publicizing itself through television,
radio, and other ads featuring state icons Paul Bunyan and Babe the Blue Ox.
The ads showed Paul sustaining injuries, with the message: Minnesota: Land
of 10,000 Reasons to Get Health Insurance. Because MNsure was completely
new, the ads were intended to create a general awareness of MNsure. The
original plan was to run the Paul and Babe ads through the entire open enrollment
period (ending in March 2014). But, in late 2013, MNsure management decided
to discontinue the ads. Some testimonial adsfeaturing consumers who had
enrolled in MNsureaired during the final two months of open enrollment.
MNsures advertising campaign during the first open enrollment period helped to increase
awareness of the exchange, but the campaigns effectiveness was undermined by MNsures
technical problems.
The contractor that developed the original advertising campaign hired a firm to
conduct an evaluation of the campaign; this evaluation did not examine the
testimonial ads. The evaluation was based on surveys given to samples of
Minnesotans before and after the ad campaign. The evaluation showed that
75 percent of respondents were aware of MNsure after the ad campaign
(compared with 20 percent before the campaign), and more than half the
respondents said they had seen a MNsure ad multiple times. But, the evaluation
65
The original BBDO contract (for $666,590) was increased to about $1.6 million, but this
amendment occurred after the additional work was completed. This resulted in an audit finding in:
Office of the Legislative Auditor, Minnesota Health Insurance Exchange: MNsure, Internal
11-13. Also, one component of the BBDO contract was a $70,000 subcontract for an outreach
plan. This plan was issued in May 2013, but many of its recommended strategies were not used in
the first year of enrollment.
OPERATIONS
127
said, The enormous gains in awareness did not translate into intention to enroll
in MNsure.66 This suggested that the ads probably did not have much impact on
enrollment.
The Paul and Babe ads were memorable, but they provided little information
about MNsure. The evaluation of the ads reported that the commercials lacked
the educational value.67 In addition, only 28 percent of survey respondents who
were shown a Paul and Babe commercial said they could relate to the ad.68
The evaluation said that the ad campaign also showed an increase in negative
impressions of MNsure. It said that problems with the websites functionality
and related media coverage contributed to the negative impressions.
DATA SECURITY
Information that MNsure collects, creates, or maintains regarding individuals
who apply for health coverage is classified by law as not public.69 During the
application process, individuals may be required to provide sensitive information,
such as birthdates, Social Security numbers, and income. An applicant may also
be asked to provide information about which, if any, family members have
communicable diseases, terminal illnesses, or mental illnesses. Because there is
considerable legislative interest in the protection of these data, we examined
some issues related to data security.
In 2013, as the Legislature discussed bills to establish MNsure, legislators asked
many questions about the security and privacy of data that would be provided to
the exchange. MNsures 2013 enabling legislation contained a section that
specified how MNsure should handle the data it collected.70 For example, the
law makes MNsure subject to the Minnesota Government Data Practices Act,
which defines various types of data and specifies who has access to each type of
data.71 State law says MNsure may share data with other state or federal agencies
to help verify individual identities, determine eligibility, process enrollments,
process premiums, or investigate fraud, but it must have data-sharing agreements
with those agencies before this occurs.72
The Affordable Care Act required the U.S. Department of Health and Human
Services to develop standards and protocols to ensure security and privacy when
individuals enroll in federal and state health care programs. The department
66
Persuadable Research Corporation, Evaluating Advertising Campaign: MNsure (Overland Park,

KS, March 2014), 7.
67
Ibid., 20.
68
Ibid.
69
70
Minnesota Statutes 2014, 62V.06.
71
Ibid., subd. 1.
72
Ibid., subd. 5(a). State law also authorizes MNsure to share information with nongovernmental
entities for these purposes, provided that MNsure has contracts with these entities that comply with
the states government data practices provisions.
128
developed Minimum Acceptable Risk Standards for [Health Insurance]

Exchanges, commonly referred to as MARS-E. These standards provide a
starting point for security guidance that exchanges can use in implementing and
operating their [information technology] systems in support of the Affordable
Care Act.73
State information technology experts consider MNsures data security structure to be
sophisticated and effective, even though the system does not yet meet all of the federal
governments security requirements.
An independent contractor issued a security assessment of MNsure in late August
2013, about one month before MNsures first open enrollment began.74 That
report assessed 282 security controls and found that 145 (51 percent) had been
fully implemented. The remaining controls were either partially implemented or
planned. The report identified 40 system risks, and it categorized none as high
risks, 7 as moderate risks, and 33 as low risks. The report concluded that
MNsures system was substantially compliant with the MARS-E standards.
Minnesotas Office of MN.IT Services helped to build and implement security
practices for Minnesotas exchange. Shortly before open enrollment began, an
official from that office testified to legislators that the low and moderate risks
identified by the August 2013 external review had been addressed.75 When asked
by a legislator whether there were any security issues that should cause MNsure
to delay open enrollment, the official said: Thats a difficult question to
answer. We still have some final security work to do. At this point in time,
we dont see a list of those show-stopper issues from a security perspective.
The federal government gave Minnesotas health insurance exchange authority
to connect to the Federal Data Services Hub on September 27, 2013. Curiously,
a federal document from the following day said that the Centers for Medicare and
Medicaid Services chief information security officer rates the risk associated
with the Minnesota MNsure Application [Authority to Connect] as High.76 An
Office of MN.IT Services official told us his agency never saw this federal
document at the time, and he does not understand the basis for its conclusion.
73
U.S. Department of Health and Human Services, Centers for Medicare and Medicaid Services,
Minimum Acceptable Risk Standards for ExchangesExchange Reference Architecture
Supplement, Version 1.0 (Washington, DC, August 1, 2012), 2.
74
WIPFLi, MNsure Security Assessment Report, August 29, 2013. The report is not a public
document.
75
76
MNsure Legislative Oversight Committee, September 24, 2013.
Jane Kim, Center for Medicare and Medicaid Services, [Chief Information Security Officer]
Reviewer Overall Comments & Recommendations, September 28, 2013. The document said this
was based on various considerations, including 110 security controls not documented or
incomplete.
OPERATIONS
129
A second independent security assessment of MNsures system was issued in

June 2014.77 It said that 82 percent of the security controls it assessed had been
fully implemented. The assessment identified 3 security issues it categorized as
high risks; there were also 9 moderate risks and 23 low risks. We asked the
Office of MN.IT Services for an update on the status of the issues identified in
this security audit. As of October 2014, seven moderate and eight low risk issues
remained, according to the office.
Despite these outstanding issues, officials from the Office of MN.IT Services
told us that MNsures security controls may be the best that Minnesota state
government has ever built. They said there is room for improvement, and they
expressed confidence that officials from MNsure, the Department of Human
Services, and the Office of MN.IT Services will jointly give attention to the
remaining security issues.
To our knowledge, there have been no instances of large-scale security breaches
involving data supplied to MNsure by applicants or enrollees.78 There have been
a few instances in which personal information submitted to MNsure has been
unintentionally shared with others. For example, a consumer asked MNsure to
reset an account password but gave MNsure an incorrect application number; the
consumer used the incorrect application number and new password to view an
online application and recognized that it contained information about another
person. MNsure disabled the account.
Ability to Detect Improper Data Access

State law says: Only individuals with explicit authorization from the [MNsure
Board] may enter, update, or access not public data collected, created, or
maintained by MNsure.79 The law also says: All queries and responses, and
all actions in which data are entered, updated, accessed, or shared or
disseminated outside of MNsure, must be recorded in a data audit trail.80
MNsure has somebut not completeability to identify inappropriate access to individuals
not-public records.
77
WIPFLi, MNsure Security Assessment Report, June 6, 2014. The report is not a public
document. The report said: Due to the limited nature of this assessment, and the fact that it is a
point in time assessment, we cannot provide an opinion as to the overall security posture of the
MNsure system and organizational environment (p. 2). This assessment excluded review
of (1) security controls that had been implemented before the first review and did not have
moderate or high risks at that time; and (2) security controls deemed not critical.
78
However, there was an incident in September 2013 in which a MNsure employee inadvertently
sent not-public information about a large number of insurance brokers to a broker. See Office of
the Legislative Auditor, MNsure: An Unauthorized Disclosure of Private Data, Special Review
(St. Paul, November 7, 2013).
79
Minnesota Statutes 2014, 62V.06, subd. 8(a).
80
Ibid.
130
At a September 2013 legislative hearing, MNsure was asked about the status of
its audit trail capabilities.81 A MNsure official described how MNsure
intended to analyze a large volume of data for the purpose of tracking every
single view or modification to any piece of personally identifiable data within
[the] system. The official seemed to indicate that this capability was still under
development, so a legislator asked if there was a timeline for putting this in place.
The MNsure official responded, in part: The audit trail is fully functional within
the system.
To better understand the functionality of MNsures audit trail system, we talked
with Office of MN.IT Services staff. We learned that, prior to October 1, 2013,
the plan to install a system-wide, integrated analytics system to help MNsure
monitor access to data and identify possible privacy breaches was postponed
indefinitely. We were told in late 2014 that action on this system has been
postponed until at least 2015.
Without such a system, MNsure still has the ability to review who has accessed
not-public data, but in limited ways. For example, if an individual wants to know
who has accessed his or her private data, MNsure can look at its records to
determine this. However, MNsure does not currently have the ability to do
large-scale reviews of its entire enrollee database to identify red flags that
might indicate inappropriate access.
In our view, MNsures response to a 2013 legislative question on this topic did
not provide a sufficiently clear picture of MNsures audit trail capabilities. We
have no reason to think that MNsures response was intentionally misleading.
Nevertheless, we think it is important to clarify that while MNsure has
established the data audit trail required by state law, MNsures ability to
identify and investigate possible cases of inappropriate access is, at this time,
somewhat limited.
81
MNsure Legislative Oversight Committee, September 24, 2013.
List of Recommendations
The Legislature should amend Minnesota Statutes 2014, chapter 62V, to

ensure that MNsures future information technology work is subject to
oversight from the Office of MN.IT Services. (p. 36)
The MNsure Board should ensure implementation of its policies or, if

necessary, revise policies that are not realistic to implement. (p. 50)
The MNsure Board should:
Adopt consumer assister compensation rates annually and ensure that

changes in the rates are published in a timely manner in the State
Register; and
Adopt navigator, call center, and customer service policies in 2015.

(p. 51)
The Legislature should establish in state law a structure for governing

MNsures online enrollment system. (p. 52)
The Legislature should amend state law to give the governor, rather than the
MNsure Board, authority to appoint the MNsure chief executive officer.
(p. 55)
The Legislature should consider whether to retain the MNsure Board as a

governing body or to make it advisory. (p. 55)
MNsure should develop ways to improve its access to the applicant and
enrollee data it collectsfor the purpose of assessing MNsure performance,
generating management reports, and responding to public inquiries. (p. 62)
MNsure and DHS should ensure that brokers are fairly compensated for
enrolling consumers in insurance through MNsure. (p. 109)
MNsure should improve its referral guidance for customer service staff in its
contact center and in DHS and county call centers. (p. 123)
Appendix A: Glossary of Terms

Actuarial Value: The estimated percentage of total expenditures for essential
health benefits (see below) that would be covered by an insurance product for a
standard population of insured individuals. For example, an insurance product
with an actuarial value of 70 percent would cover 70 percent of the costs for
essential health benefits that are provided to individuals covered by that
insurance. An insurance products metal level (see below) indicates the
actuarial value of that product.
Advanced Premium Tax Credit (APTC): A type of federal tax credit available
to people who purchase health insurance through an exchange, depending on
their income and household size. The credit may be claimed by eligible people
when they file income taxes. Alternatively, qualified individuals may use the tax
credit right away to lower their monthly health insurance premium costs; in this
case, the U.S. Treasury Department sends monthly payments to the persons
insurer to cover all or part of the premium.
Adverse Selection: The tendency for people to avoid buying insurance until
they think they will benefit from it. As a result of adverse selection, people who
are sicker tend to be more likely to obtain insurance coverage than healthier
people.
Affordable Care Act (ACA): The common name for the federal legislation
signed into law in March 2010 that aimed to reduce the number of people without
insurance, increase the quality and affordability of coverage, and reduce the costs
of health care. The full title of the act is the Patient Protection and Affordable
Care Act.
Assister Resource Center (ARC): A MNsure call centerseparate from
MNsures contact centerdedicated to answering questions from MNsurecertified navigators and certified application counselors.
Basic Health Program (BHP): A federal option under the Affordable Care Act
that allows states to receive federal assistance if they offer insurance to certain
low-income people with incomes too high to be eligible for Medicaid. For states
implementing a Basic Health Program, the federal government pays 95 percent of
the advanced premium tax credit and cost-sharing reductions for which
participating individuals would have qualified if they had enrolled through a
health insurance exchange in a qualified health plan. In 2015, Minnesota is the
only state offering a Basic Health Program. Minnesotas Basic Health Program
is called MinnesotaCare.
Broker: A state-licensed seller of health insurance who helps individuals or
businesses purchase insurance. Brokers (also called insurance agents or
producers) are licensed and regulated by the Minnesota Department of
Commerce. Brokers receive payments from a health insurer for enrolling an
individual in that insurers products, assuming the insurer has authorized the
134
broker to sell insurance on its behalf. Brokers must be certified by MNsure to

provide assistance to individuals seeking to obtain insurance through MNsure.
Certified Application Counselor (CAC): A persontypically a hospital
employeewho helps individuals applying for insurance through MNsure.
CACs do not receive compensation for this assistance from MNsure or insurers.
Childrens Health Insurance Program (CHIP): A federal program that
provides health insurance coverage to uninsured children under age two from
low-income families and to uninsured, low-income, pregnant women.
Coinsurance: A type of out-of-pocket cost required by some insurance
products. Specifically, health insurance products may require that an insured
individual who is making an insurance claim pay a percentage of the bill that
remains after any required deductible amount has been paid.
Consumer Assister: A general term that we use in this report to refer to several
categories of people who provide application and enrollment assistance to
individuals seeking to enroll through MNsure. As used here, a consumer assister
may be a navigator, certified application counselor, or broker; all of these terms
are separately defined in this glossary.
Contact Center: A customer service call center operated by MNsure. Federal
law requires health insurance exchanges to have toll-free phone lines that
individuals can call to get help when enrolling in insurance through the
exchange.
Copayment: A type of out-of-pocket cost usually paid at the time an insured
person receives a service. For example, an insurance product may require that a
covered individual pay $20 each time the individual has a health care
appointment.
Cost-Sharing Reduction (CSR): A discount available to certain individuals
who buy health insurance through an exchange. The Affordable Care Act
authorizes cost-sharing reductions for people with incomes between 139 and
250 percent of the federal poverty level who purchase a silver plan (see the
definition of metal level below). In Minnesota, only individuals with incomes
between 201 and 250 percent of the federal poverty level are eligible for
cost-sharing reductions, due to Minnesotas implementation of a Basic Health
Program for people with incomes between 139 and 200 percent of the federal
poverty level. With a cost-sharing reduction, a persons health insurance product
has lower out-of-pocket health care expensescopayments, deductibles, and
coinsurancethan it would otherwise have. A silver plan would generally cover
70 percent of the costs of covered individuals essential health benefits, but a
cost-sharing reduction effectively increases this coverage to 73 percent.
Deductible: The amount people pay for health care services before their insurer
begins to pay. For example, if someone buys an individual insurance product
with a $2,000 deductible, insurance will generally not cover the persons claims
until the bills submitted for health care services in that year exceed the amount of
the deductible.
APPENDIX A
135
Essential Health Benefits: A comprehensive set of services that qualified

health plans (see below) must offer under the Affordable Care Act. The plans
must offer services in ten categories, including doctor visits, hospitalization,
mental health services, prescription drugs, and others. In Minnesota, the small
group insurance plan (that is, an insurance product sold to businesses with 2 to
50 full-time-equivalent employees) with the largest enrollment statewide
provides the benchmark for evaluating the benefits in other plans. This means
that all qualified health plans in Minnesota must offer benefits similar to those
offered by the benchmark plan.
Exchange: A resource where individuals, families, and small businesses can
learn about their health coverage options, compare health insurance products, and
enroll in coverage. Exchanges are sometimes called health insurance
marketplaces. In this report, we use the term exchanges to refer to
exchanges authorized by the Affordable Care Act, although privately operated
exchanges may have similar features. Individuals only qualify for federal
advanced premium tax credits or cost-sharing reductions when they purchase
insurance through an ACA-authorized exchange.
Individual Market (or Nongroup Insurance): Insurance that is sold to
individual buyers, rather than to groups or employers. In Minnesota, such
insurance may be purchased directly from an insurer or through MNsure.
Insurer (or Health Plan): A company that provides insurance coverage to
individuals or businesses.
Large-Group Insurance: Health insurance that is provided to individuals or
families through a group health plan by an employer with more than a certain
number of employees. The Affordable Care Act defines large-group as 50 or
more full-time-equivalent employees.
Life Event: A circumstance that allows a person to enroll in insurance or make a
change to their existing insurance coverage outside of an open enrollment period.
A qualifying life event might include a change of residence, a change in income,
or a change in family size (for example, through a birth, death, marriage, divorce,
or adoption).
Medical Assistance (MA)/Medicaid: A programfunded roughly equally by
federal and state dollarsthat pays for health insurance for low-income people.
The federal program is called Medicaid, and Minnesotas version of the Medicaid
program is called Medical Assistance.
Metal Level: A designationbronze, silver, gold, or platinumthat is given to
qualified health plans (see below) to help consumers compare plans offering
similar coverage. Metal levels reflect the percentage of total essential health
benefits costs (see above) the plan will cover, based on actuarial assessments.
For example, a bronze plan would cover 60 percent of the expected value of the
essential health benefits; a silver plan would cover 70 percent; a gold plan would
cover 80 percent; and a platinum plan would cover 90 percent.
136
MinnesotaCare: A Minnesota program, created in 1992, that provides health

insurance to individuals who do not qualify for Medical Assistance. Starting in
2015, MinnesotaCare serves as Minnesotas Basic Health Program under the
Affordable Care Act. MinnesotaCare mostly provides coverage to adults with
modified adjusted gross incomes between 139 and 200 percent of the federal
poverty level. Most MinnesotaCare recipients pay a sliding-scale premium,
ranging from $4 to $50 monthly.
MNsure: Minnesotas health insurance exchange, as authorized in state law by
the 2013 Legislature. MNsure is the name of the state agency, governed by a
board, that has directed the development of the exchange, pursuant to the federal
Affordable Care Act. MNsure is not an insurance company. However,
individuals and businesses may purchase and enroll in commercial insurance or
public health care programs through MNsureor, more accurately, through
MNsures online enrollment and eligibility determination system. People
purchasing commercial insurance may access through MNsure any Affordable
Care Act tax credits or subsidies for which they qualify.
Modified Adjusted Gross Income (MAGI): A definition of income that is used
to determine eligibility for most recipients of Medical Assistance,
MinnesotaCare, advanced premium tax credits, and cost-sharing reductions. As
defined for federal tax purposes, modified adjusted gross income is a households
adjusted gross income plus any tax-exempt Social Security, interest, or foreign
income.
Navigator: A type of consumer assister, typically employed by a communitybased organization, that helps people enroll in public health insurance programs
or commercial products offered through MNsure. Navigators must help any
MNsure user that requests their assistance and provide information in a fair,
accurate, and impartial manner. Navigators receive state compensation for each
person they help enroll through MNsure; they do not receive compensation from
insurance companies. Navigators must help consumers understand the
differences between their health insurance options, but they may not advise
people about which specific commercial product to select.
Network: The group of health care providers that an insured person is
authorized to use, under the terms of the insurance they have obtained.
Open Enrollment: The period of time when individuals may obtain or change
commercial insurance through MNsure. For example, individuals purchasing
commercial insurance for calendar year 2015 could enroll or re-enroll through
MNsure between November 15, 2014, and February 15, 2015. During other parts
of the year, individuals may only make changes in their commercial insurance
coverage if they have a qualifying life event (see above) or are members of an
American Indian tribe.
Out-of-Pocket Costs: Costs for health care services that are the responsibility of
the insured person. These costs may include deductible amounts, coinsurance,
and copayments.
APPENDIX A
137
Portal: A point of access to a health insurance exchanges online enrollment

system. MNsure was intended to have separate portals for use by
navigators/brokers, health insurers, county and state human services caseworkers,
and MNsure financial staff. However, these portals were not all fully functional
during MNsures first two enrollment years.
Premium: The monthly amount paid by an individual to an insurance company
for providing health care coverage.
Premium Withhold: The portion of paid premiums that MNsure uses to cover
part of its operating expenses. Under state law, MNsure is authorized to retain up
to 3.5 percent of premiums paid by individuals who have enrolled in qualified
health plans through MNsure.
Qualified Health Plan: Insurance products that have been certified, in
accordance with the Affordable Care Act, to be sold through health insurance
exchanges. Insurers may apply to MNsure to have their commercial insurance
products sold through the state exchange. The Affordable Care Act requires
qualified health plans to offer essential health benefits (see above) and have
designated metal levels (see above). Minnesota law also requires certain
products sold in the small-group and nongroup markets outside of MNsure to be
certified as qualified health plans.
Small-Group Insurance: Health insurance that is provided to individuals or
families through a group health plan by an employer with fewer than a certain
number of employees. The Affordable Care Act defines small-group as fewer
than 50 full-time-equivalent employees.
Appendix B: Key Events in

MNsures Development and
Implementation
T
he following pages show selected events in the development and

implementation of MNsure. This timeline is not intended to be
comprehensive, but it provides context for understanding MNsures relatively
short history.
140
Timeline of Key Events
Governor
Pawlenty
prohibits
ACA-related
planning
Affordable
Care Act
signed into
law
2010
JAN
Governor Dayton
authorizes state
to seek ACA
planning grants
2011
JUL
Minnesota
receives first
ACA planning
grant
State
issues RFP
seeking
vendors for
technical
prototypes
Department of
Commerce
establishes
Minnesota
Health
Insurance
Exchange
Advisory Task
Force
JAN
JUL
Exchange task
force issues initial
recommendations
Commerce
contracts with
vendors to
develop
exchanges
technical
infrastructure
2012
Governor
Dayton
elected
Legislatively
mandated
task force
says
Minnesota
well
positioned to
develop an
exchange
JAN
(continued on next page)
JUL
Governor
Dayton directs
Commerce to
build
exchange
Governor
Dayton
announces
plan to move
exchange to
different
agency (MMB)
Minnesota
submits its
blueprint for
exchange to
federal officials
Minnesotas
exchange is
conditionally
approved by
federal officials
APPENDIX B
141
Timeline of Key Events (continued)

Exchange
officials begin
process of
amending
vendor contracts,
reducing the role
of Maximus
2013
MNsureenabling
legislation
signed into
law by
Governor
Dayton
Governor
Dayton
appoints
MNsure Board
JAN
MNsure board
assumes its
full authority
from MMB
2014
Launch of
MNsure
website
JUL
MNsure hires
Deloitte as lead
contact with its
technology
vendors
MNsure hires
Optum for
external
assessment
MNsure
receives
authority to
connect to
federal data
hub
Second open
enrollment
period begins
First open
enrollment
period
ends
JAN
JUL
MNsure
executive
director
resigns;
replacement
named
Appendix C: Health Insurance

Rating Areas in Minnesota
Rating Area
Number
Region
South East
North East
South Central
South West
Mid Central
West Central
North Central
Twin Cities
North West
2
7
6
5
4
8
3
Counties
Dodge, Fillmore, Freeborn, Goodhue, Houston,
Mower, Olmsted, Steele, Wabasha, Winona
Carlton, Cook, Itasca, Koochiching, Lake,
Lake of the Woods, Saint Louis
Blue Earth, Faribault, Le Sueur, Martin, Nicollet,
Rice, Waseca, Watonwan
Brown, Cottonwood, Jackson, Lincoln, Murray,
Nobles, Pipestone, Redwood, Rock
Big Stone, Chippewa, Kandiyohi, Lac qui Parle,
Lyon, McLeod, Meeker, Renville, Sibley, Swift,
Yellow Medicine
Becker, Clay, Douglas, Grant, Otter Tail, Pope,
Stevens, Traverse, Wilkin
Aitkin, Beltrami, Cass, Chisago, Crow Wing,
Hubbard, Isanti, Kanabec, Mille Lacs, Morrison,
Pine, Roseau, Todd, Wadena
Anoka, Benton, Carver, Dakota, Hennepin,
Ramsey, Scott, Sherburne, Stearns, Washington,
Wright
Clearwater, Kittson, Mahnomen, Marshall, Norman,
Pennington, Polk, Red Lake
SOURCE: Minnesota Department of Commerce.
The federal Public Health Service Act, as amended by the Affordable Care Act,
requires states to establish geographic divisions known as rating areas.1 A
health insurance enrollees rating area of residence is one of four factors that may
be considered by health insurers when varying the premium to be paid by the
enrollee for any given insurance product sold in the individual or small-group
market, whether through MNsure or not.2 To be presumed adequate by the U.S.
Department of Health and Human Services (HHS), the number of rating areas in
a state must not exceed one more than the number of metropolitan statistical
areas in that state.3 State law requires that geographic rating areas contain no
fewer than seven counties that create a contiguous region.4 Minnesotas rating
areas were defined by the states departments of Commerce and Health.
42 U.S. Code, sec. 300gg (2013).
The other factors that may affect the premium are the age of the enrollee, whether the enrollee
uses tobacco products, and whether the insurance product covers an individual or a family.
Metropolitan statistical areas (MSAs) are defined by the U.S. Office of Management and Budget
as having at least one urbanized area of 50,000 or more inhabitants. Portions of Minnesota are
included in eight MSAs, some of which have their principal cities in North Dakota or Wisconsin.
Therefore, Minnesota could have a maximum of nine rating areas.
Minnesota Statutes 2014, 62A.65, subd. 3(b).
Appendix D: 2014 Federal Poverty

Guidelines
For the 48 Contiguous States and the District of Columbia
Family/Household Size
100%
138%
200%
1
2
3
4
5
6
7
a
8
$11,670
15,730
19,790
23,850
27,910
31,970
36,030
40,090
$16,105
21,707
27,310
32,913
38,516
44,119
49,721
55,324
$23,340
31,460
39,580
47,700
55,820
63,940
72,060
80,180
250%
$ 29,175
39,325
49,475
59,625
69,775
79,925
90,075
100,225
275%
$ 32,093
43,258
54,423
65,588
76,753
87,918
99,083
110,248
NOTE: The 2014 Federal Poverty Guidelines are used to make eligibility determinations for coverage year 2015.
a
For families/households with more than 8 persons, add $4,060 to the 100 percent column for each additional person.
SOURCE: Poverty Guidelines, 79 Fed. Reg. 3,593 (2014).
400%
$ 46,680
62,920
79,160
95,400
111,640
127,880
144,120
160,360
Appendix E: Current Assessment

of Gruber-Gorman Projections
I
n 2011, the Minnesota Department of Commerce hired Jonathan Gruber,

professor of economics at the Massachusetts Institute of Technology, and
Gorman Actuarial to estimate the impact of the Affordable Care Act (ACA) and
a state-based health insurance exchange on Minnesota. The initial GruberGorman economic and actuarial analysis was released in April 2012. A second
Gruber-Gorman report updated the earlier analysis and was released in February
2013, when the enabling legislation for Minnesotas exchange was moving
through the Legislature. Both reports made projections of the ACAs impacts for
2016.
The February 2013 report consistently assumed that public health insurance
would be available for children up to 275 percent of the federal poverty
guidelines (about $65,588 for a family of four) and sometimes assumed that
Minnesota would adopt a Basic Health Program. This report made four key
projections regarding the impact of the ACA. Below, we compare the 2013
Gruber-Gorman projections for 2016 with what is known as of early 2015.
Projection 1
The 2013 Gruber-Gorman report projected that if MinnesotaCare were
implemented as the Basic Health Program, the number of uninsured people in
Minnesota under the ACA would drop by between 298,000 and 340,000 people
(up to 68 percent) in 2016 from the projected uninsurance rate without the ACA
(10.9 percent). This would mean that there would be as few as 159,000
uninsured Minnesotans in 2016.
In fact, a 2014 study estimated that the number of uninsured people in Minnesota
fell by 180,520 people (40 percent) from the Fall 2013 uninsurance rate of
8.2 percent.1 As of May 1, 2014, there were an estimated 264,480 uninsured
Minnesotans. Thus, as of mid-2014, Minnesotas number of uninsured people
was greater than the number that Gruber-Gorman projected for 2016.
Projection 2
The Gruber-Gorman report projected that the individual insurance market would
cover 530,000 individuals in 2016. It said there would be little change in the
number of people covered by employer-sponsored insurance. Projection 2 did
not contemplate the availability of MinnesotaCare as the Basic Health Program.
Julie Sonier, Elizabeth Lukanen, and Lynn Blewett, Early Impacts of the Affordable Care Act on
Health Insurance Coverage in Minnesota (Minneapolis: State Health Access Data Assistance
Center, University of Minnesota, June 2014).
148
In fact, the individual insurance market covered 324,510 individuals as of May 1,

2014.2 As projected, between September 30, 2013, and May 1, 2014, there was
little change in the number of people covered by employer-sponsored insurance.
Projection 3
The 2013 Gruber-Gorman report projected that Minnesotas health insurance
exchange would enroll 1.3 million people in 2016. The report estimated that
qualified health plans would account for 35 percent of total enrollments through
the exchange, assuming that the state retained MinnesotaCare as the Basic Health
Program. In fact, the exchange enrolled approximately 371,000 people as of
November 11, 2014. About 15 percent of these enrollments were in qualified
health plans.
Projection 4
The Gruber-Gorman report projected that overall premium costs in the individual
marketinside and outside the exchangewould fall by 34 percent on average
after accounting for subsidies, relative to what premiums in that market would
have been in 2016 had the ACA not been implemented. The report estimated
that 70 percent of the individual market would experience no change or a
decrease in premiums. Projection 4 did not contemplate the availability of
MinnesotaCare as the Basic Health Program.
In fact, an analysis by Yale University economist Amanda Kowalski,
supplemented by our own research, showed that, as of mid-2014, premium costs
in the overall individual market in Minnesota averaged 5 percent higher,
accounting for tax credits received through MNsure, than what premiums in that
market would have been in mid-2014 if the Affordable Care Act had not been
implemented and there were no changes in plan characteristics.3
Specifically, the Kowalski study projected that, without the ACA, average
premiums in the individual market in Minnesota would have been $230 per
month if trends since 2008 had persisted into the second quarter of 2014. Actual
average premiums in Minnesotas overall individual marketboth inside and
outside of MNsureas of mid-2014 were $256 per month, not accounting for
subsidies. This represented an increase of 11 percent over Kowalskis estimate
of average premiums without the ACA.
We extended Kowalskis analysis using enrollment data from the Office of
MN.IT Services to account for the average advanced premium tax credit taken by
qualified health plan enrollees who received it, which amounted to around $203.
That tax credit was only taken by about 7 percent of the overall individual
market. After weighting to account for that, the average monthly premium paid
in the overall individual market was $242 as of mid-2014, a 5 percent increase
2
3
Ibid., 6.
Amanda Kowalski, The Early Impact of the Affordable Care Act State-by-State (Washington, DC:
Economic Studies at Brookings, September 2014).
APPENDIX E
149
over what Kowalski estimated individual market premiums would have been
without the ACA.
It is not currently known what percentage of the overall individual market
experienced no change or a decrease in premiums. Our survey of MNsure
qualified health plan enrollees found that 69 percent of those who had previously
purchased insurance directly from an insurer said that their MNsure premiums,
after subsidies, are better than or about the same as their previous insurance
premiums.
Appendix F: MNsure Grants to

Promote Enrollment and
Consumer Outreach, October 2013
to September 2014
A
s discussed in Chapter 6, MNsure awarded $4.75 million in competitive

grant funds to 41 community organizations in the fall of 2013. The grants
supported organizations that provided application and enrollment assistance to
consumers via navigators, and organizations that conducted outreach and
education about MNsure to targeted populations across the state.
The following tables show the performance of MNsure grantees in enrolling
individuals and holding outreach events. The grantees contracts established
goals in each of these areas. We analyzed MNsure enrollment data to estimate
each grantees achievement of its enrollment goal during MNsures first year.
For goals related to the number of outreach events and people reached through
these events, the only performance data available were those self-reported by
grantees. The outreach data were not verified by MNsure or the Office of the
Legislative Auditor.
MNsure and one grantee (Small Business Minnesota) mutually agreed to
terminate the grantees contract in June 2014 because problems with the
functionality of MNsures small business application would have made it
difficult for the organization to fulfill its grant duties. Small Business Minnesota
returned all of the funds it was awarded.
152
MINNESOTA STATE HEALTH EXCHANGE (MNsure)
Exhibit F.1: Number of People that MNsure Grantees Enrolled through

MNsure
Grantee
Arrowhead Economic Opportunity Agencya
C.A.R.E. Clinic
Central MN Jobs and Training Services
Centro Cultural Chicano
Children's Dental Services
Communicating for Americaa
a
Community Resource Connections
Comunidades Latinas Unidas en Servicio (CLUES)a
Confederation of Somali Community in MN
Dakota County
Family Tree Clinic
Generations Healthcare Initiativesa
a
Health Access MN
HealthFinders Collaborative
a
Hmong American Partnership
International Institute of MN
McDonough Organization with Respect and Equality for
People (MORE)
Minneapolis Urban Leaguea
MN Adult and Teen Challenge
MN AIDS Projecta
MN Chippewa Tribea
MN Community Action Partnership
MN Council of Churches
MN Recovery Connection
a
National Alliance on Mental Illness
NorthPoint Health and Wellness Centera
a
Pillsbury United Communities
Planned Parenthood MN, ND, SDa
a
Portico Healthnet
ResourceWest
a
Somali Health Solutions
Southside Community Health Servicesa
Springboard for the Arts
Stairstep Foundationa
Sub-Saharan African Youth and Family Services in MN
West Side Community Health Servicesa
a
Western Community Action
Women's Health Center of Duluth, P.A.
Total
Grant Award
Number of
People Enrolled
People Enrolled,
as a Percentage
of Enrollment
Goal in Contract
$128,000
20,000
118,836
47,293
35,000
38,836
141,600
99,120
50,000
190,367
30,000
230,369
326,606
31,662
153,552
128,560
864
142
1,080
124
18
221
337
350
47
0
142
2,024
1,729
516
2,196
1,419
229%
29
63
41
18
44
65
101
3
0
142
174
29
172
211
71
2,407
100,000
56,640
192,576
192,573
424,150
159,830
45,000
125,000
138,052
100,000
150,333
377,592
45,974
149,421
94,400
50,032
100,000
25,000
75,000
257,183
21,385
75
118
379
649
66
2,786
240
166
31
1,992
375
572
6,162
133
1,721
946
31
173
83
493
1,183
191
29,774
33
59
24
43
7
70
75
66
6
203
94
11
103
76
143
20
9
58
55
66
59
76
58%
NOTES: This exhibit only includes the 28 grantees that set enrollment goals in their contracts with the state. The enrollment figures
presented here represent our best estimates of the number of consumers successfully enrolled with the help of a navigator, based on our
analysis of MNsures payment reports. Because of MNsures weak recordkeeping and problems with MNsure's enrollment system, we
could not independently verify the exact number of navigator-aided enrollments, as explained in Chapter 6. Enrollments facilitated by
navigator organizations that did not receive grant funds were not included in this table. Enrollments facilitated by subcontractor
organizations were included in the primary grantees total enrollments.
a
Grantee received additional funds for outreach and enrollment activities in the 2014-15 enrollment year through another round of
MNsure grant awards.
SOURCES: MNsures grant contracts and Office of the Legislative Auditors analysis of MNsures enrollment payment reports.
APPENDIX F
153
Exhibit F.2: Extent of MNsure-Related Outreach Reported by MNsure

Grantees
Grantee
Accountability MN
Arrowhead Economic Opportunity Agency
C.A.R.E. Clinic
Central MN Jobs and Training Services
Centro Cultural Chicano
Children's Dental Services
Communicating for America
Community Resource Connections
Comunidades Latinas Unidas en Servicio
Confederation of Somali Community in MN
Dakota County
Family Tree Clinic
Generations Healthcare Initiatives
Health Access MN
HealthFinders Collaborative
Hmong American Partnership
International Institute of MN
Minneapolis Urban League
MN Adult and Teen Challenge
MN AIDS Project
MN Chippewa Tribe
MN Community Action Partnership
MN Council of Churches
MN Recovery Connection
National Alliance on Mental Illness
NorthPoint Health and Wellness Center
Pillsbury United Communities
Planned Parenthood MN, ND, SD
Portico Healthnet
ResourceWest
Somali Health Solutions
Southside Community Health Services
Springboard for the Arts
Stairstep Foundation
Sub-Saharan African Youth and Family
Services in MN
West Side Community Health Services
Western Community Action
Women Venture
Women's Health Center of Duluth, P.A.
Reported
Outreach
Events, as a
Percentage
of Goal in
Contract
Grant
Award
Number of
Outreach
Events, as
Reported
by Grantee
$ 41,337
128,000
20,000
118,836
47,293
35,000
38,836
141,600
99,120
50,000
190,367
30,000
230,369
326,606
31,662
153,552
128,560
100,000
56,640
192,576
192,573
424,150
159,830
45,000
125,000
138,052
100,000
150,333
377,592
45,974
149,421
94,400
50,032
100,000
526
114
52
150
93
42
26
22
56
15
23
550
204
294
60
161
349
42
119
208
125
1369
158
68
76
129
109
560
413
45
140
121
108
25
1,052%
146
325
313
2,325
420
130
92
193
125
192
4,583
1,569
74
120
847
241
700
66
114
179
1,521
93
1,700
117
33
95
1,120
25,000
75,000
257,183
45,746
21,385
63
65
950
24
19
252
346
280
242
360
417
a
7,917
104
127
Number of
People
Reached,
as Reported
by Grantee
24,195
333,398
42,043
6,071
4,478
1,562
5,000
1,322
5,656
3,500
63,228
5,000
97,721
11,034
1,500
15,528
4,460
37,036
1,187
115,571
5,489
2,769,505
3,287
10,219
71,437
1,849
28,370
37,755
30,814
465
51,000
2,272
46,102
200,000
2,143
10,212
9,869
673,647
876
Reported
People
Reached,
as a
Percentage
of Goal in
Contract
144%
1,334
210
467
110
312
100
26
377
58
315
100
244
46
107
2,724
1,025
7,407
37
770
253
27,695
1,027
204
1,587
66
142
95
a
116
173
45
115
4,000
11
101
164
48,118
219
NOTES: The performance data listed in this table were self-reported by grantees at the end of the grant period and were not
independently verified by MNsure or the Office of the Legislative Auditor. MNsure and its grantees defined an outreach or education
event broadly. For example, grantees considered radio or Internet-based ads and direct mailings to be outreach events. Also, the
McDonough Organization with Respect and Equality for People (MORE) is not included in this exhibit because it neither set outreach
goals with the state nor reported on any outreach activities at the end of the grant period.
a
The grantees contract did not include a goal in this area.
SOURCES: Office of the Legislative Auditors review of MNsures grant contracts and grantees final reports to MNsure.
February 9, 2015
James Nobles
Legislative Auditor
Centennial Office Building, Room 140
658 Cedar Street
St. Paul, Minnesota 55101
Dear Mr. Nobles:
Thank you for conducting a program evaluation of MNsure. We appreciate the professionalism shown
by you and your staff throughout this process. As you know, we welcome your review and view it as
part of an ongoing process of improvement for the organization. MNsures response to the specific
findings is attached.
First, we measure success thus far in several distinct ways. Since October 1, 2013, the uninsured rate
in Minnesota has dropped by 40 percent to less than five percent. Now, 95 percent of Minnesotans
have comprehensive, affordable health insurance coverage.
Second, MNsure has been instrumental in the enrollment of hundreds of thousands of Minnesotans in
comprehensive, affordable health coverage. Many of those who enrolled had previously not had
health insurance coverage. MNsure has added competition to the insurance market and transparency
to price comparison which, combined, drive down costs for Minnesotans. This is evidence of success.
Furthermore,MNsure has made dramatic improvements to the consumer experience in less than 24
months of operation. MNsure has completed its second open enrollment period and in contrast to year
one
Consumers are enrolling through the website with relative ease.
Call volume is high and call wait times are on average less than five minutes.
A robust statewide network of navigators, brokers and other assisters is in place to help
consumers enroll.
Consumers are saving money. Minnesotans who enrolled in qualified health plans saved over
$30 million as a result of tax credits on health insurance plans sold through MNsure.
We have a stong, multi-agency project management team and decision-making process in
place to set priorities.
We have a deep commitment to transparency and accountability.
We are listening, and our partners and stakeholders are informed and engaged with us as we
continue to grow and improve.
I appreciate your frequent mention of the hard work of the MNsure team in your report. Our entire
teamincluding our partners at MN.IT Services and the Department of Human Servicesis
extraordinarily dedicated to the mission and success of MNsure.
We continue to take our responsibility to be an accountable and transparent organization extremely

seriously. We have been working as an organization since early 2014 to proactively identify and make
improvements to all areas of MNsure.
Reviews and audits such as this one are important tools for us to improve. In the interest of
transparency and accountability, we will continue to make necessary adjustments to the organization.
while maintaining our focus on improving the consumer experience.
Again, thank you for the work that you and your staff have done on this review.
Sincerely,
Scott Leitz
Chief Executive Officer
Attachments
OLA Summary Conclusion: In its first year of operations, MNsures failures outweighed
its achievements.
MNsure Response
To the extent this statement was intended to be the overall conclusion of this report, MNsure
strongly disagrees with this conclusion. Since October 1, 2013, the uninsured rate in
Minnesota has dropped by 40 percent with 95 percent of Minnesotans having
comprehensive, affordable health insurance coverage. To date, the MNsure IT system has
been instrumental in enrolling hundreds of thousands of Minnesotans in affordable health
coverage. Many of those enrolled did not previously have coverage. Minnesotans who
enrolled in qualified health plans through MNsure saved over $30 million as a result of tax
credits on health insurance plans sold through the Exchange.
For further context, all of this was accomplished as MNsure worked to stand up, from
scratch, a brand new state agency-developing and improving all of the business functions
necessary to support its mission outlined in Minnesota Statutes, Chapter 62V.
Significant improvements have been made to the enrollment process and customer service.
Minnesotans have received and continue to receive considerable benefits from the
existence of MNsure. For all of these reasons, MNsure cannot agree with the statement that
its failures outweighed its achievements.
MNsure General Comments

1. The use of the term MNsure in this report
While MNsure appreciates that the definition of the term MNsure is included in Appendix A:
Glossary of Terms, and that, in certain sections of the report the phrase MNsure enrollment
system or MNsures enrollment system is used, this definition does not provide sufficient
clarity regarding the difference between MNsure (the state agency) and MNsure (the IT
system). MNsure is concerned that instead of resolving any confusion regarding this
differentiation, this confusion will only be perpetuated amongst the public at large as they
read this report. It is MNsures understanding that the Office of the Legislative Auditor views
certain statutory provisions in Minnesota Statutes, Chapter 62V to require MNsure to solely
operate the MNsure IT system, and thus, that MNsure (the state agency) is solely
responsible for allowing other agencies to use this system and responsible for providing them
with output from the systems various functions (eligibility determinations, billing, etc.).
MNsure does not agree that Chapter 62V contains this requirement, and believes this view
misconstrues the legal framework that governs use of the MNsure IT system. In an effort to
further explain, the following paragraphs provide some additional background on the
development of the legal framework for the MNsure IT system.
As noted in this report, early on in the planning and development of the MNsure IT system
there was keen interest on the part of the state agencies to assuring that this system would
provide a single, streamlined access point for individuals to receive eligibility determinations
for and enroll in public health care programs and private health coverage. In fact, the Patient
Protection and Affordable Care Act (ACA) itself, as well as federal regulations and
guidance, have identified this as an important role for Exchanges.
However, the early, proposed regulation and high-level guidance relating to how Exchanges
and State Medicaid agencies should interact on the responsibilities and authorities for
eligibility determinations only appeared to provide a binary approach: 1) an Exchange
performs an assessment of Medicaid eligibility and hands-off to the State Medicaid
Agency for the final determination; or 2) an Exchange directly makes the Medicaid eligibility
determination. While the proposed regulation identified an option for a combination of these
approaches, of particular note was that neither the term assessment nor directly were
defined. As legislation for the creation of MNsure as a separate state agency was preceding,
there was not clarity on the scope of regulatory requirements for Exchanges and State
Medicaid agencies regarding specific authority and responsibilities for eligibility
determination, and thus, the legislation was drafted to allow for flexibility to meet these
possible requirements while retaining specific provisions to identify this commitment to
providing consumers a single, streamlined application process to access public programs
and private health insurance coverage.
Subsequent to the enactment of MNsures enabling legislation, MNsure and Minnesota
Department of Human Services (DHS) had many joint communications with federal officials
from both the Consumer Information and Insurance Oversight (CCIIO) and Centers for
2
Medicare and Medicaid Service (CMS) in which the agencies presented Minnesotas
approach to Medicaid eligibility determinations going through the MNsure IT system.
Specifically, the agencies described the approach that the MNsure IT system would
simultaneously serve as the automated eligibility system for both the DHS (the state Medicaid
agency) and MNsure (the Exchange), and thus, would not fit neatly into the binary framework
suggested by the regulations and guidance which presumed one automated eligibility system
maintained by a State Medicaid agency and another separate automated eligibility
system maintained by an Exchange. Both MNsure and DHS explained that the approach
being built into the MNsure IT system most likely complies with the assess and hand-off
option for Medicaid eligibility, but the mechanics of the assessment and hand-off are
unique in that the system doing the assessment and the hand-off is the same system that
is being handed-off to.
Federal officials acknowledged that this approach appropriately met the regulatory
requirements, and that even though MNsure and DHS would be working very closely and that
the work of each agency could have benefits that accrue to the other (i.e. MNsure contact
center staff would field calls from public program enrollees requesting status updates) the
approach identified by the two agencies did not require a delegation of authority from DHS to
MNsure to conduct eligibility determinations because it was clear that DHS was continuing to
make the eligibility determinations for Medicaid, albeit on a shared IT system.
We believe that readers of this report would have a more accurate understanding of the
unique relationship of MNsure (the state agency), DHS, and the MNsure IT system if those
entities and terms were used in their correct context. In some cases an attempt has been
made to use the term MNsure enrollment system but the term is undefined and used
inconsistently throughout the report.
2. Scope of the time period being reviewed
MNsure appreciates that some mention was made of the time period being evaluated, but
believes that this reference was insufficient to give readers proper context for the sweeping
statements or findings made in this report. No mention is made of several key improvements
put in place for the second open enrollment period, even though information on these
improvements has been publicly disclosed and discussed. For example, functionality and
experience testing occurred prior to the release of new system functionality for the open
enrollment period; robust technical information and training was available to staff in advance
of the open enrollment period; public website was updated to make specific information
easier to access; and information about assisters and enrollment events was updated and
provided in a searchable format.
Chapter 1: Background
OLA Key Finding 1: Minnesotas health insurance exchange (MNsure) has some
important differences from other executive branch state agencies.
MNsure Response
MNsure generally agrees with this finding, but it is more accurate to characterize MNsure as
being a state agency subject to all the same requirements other state agencies with some
key exemptions and some additional requirements not applicable to any other state
agencies.
OLA Key Finding 2: In contrast to practices in many other states, Minnesotas exchange
provides a single website at which individuals eligibility for tax
credits and public health care programs can be determined.
MNsure Response
As has been explained above in this response, it is important to distinguish between the
roles played by MNsure (the state agency) and DHS and to clarify the co-development and
co-ownership of the MNsure IT system by the two agencies. MNsure (the state agency) has
no legal authority to determine eligibility for public programs. However, the two agencies
share an automated eligibility system in the MNsure IT system.
Chapter 2: Exchange Development and Implementation

OLA Key Finding 1: Minnesotas efforts to meet an ambitious federal deadline were
hindered by late federal rules, delays in passing state legislation,
and problems with vendor selection and performance.
MNsure Response
Aggressive timelines, delays in receiving federal regulations and guidance, delays in
enactment of enabling legislation creating MNsure (the state agency), and delays in the
selection of vendors significantly complicated the development of the MNsure IT system.
In addition to these factors identified in the report, it is important to note that at the time of
vendor selection, Minnesota was competing with both the federal government and other
state-based exchanges for a very limited pool of vendors with ability to perform an IT system
build of this complexity, in the limited amount of time available. The project leadership and
other agencies involved in the contract negotiations had to carefully balance pushing
prospective vendors to agree to the best price for the State of Minnesota, and at the same
time moving quickly to assure available vendors did not become separately engaged in
exchange projects with other entities and moving quickly to assure the project could begin.
It is also important to note that significant changes in federal requirements in January 2013
(commonly referred to as the fed 70) forced a reevaluation of the projects ability to meet
4
federal certification requirements for go-live on October 1, 2013. In response to these

requirements, leadership at the exchange (then under direction of the Commissioner of
MMB), DHS, and MN.IT services came to the conclusion that a reorganization of the then
existing contractual roles was critical to meeting the milestones necessary to achieve go-live
on October 1, 2013.
OLA Key Finding 2: Serious technical deficiencies plagued MNsures enrollment system
throughout its first year of operations.
MNsure Response
The limited functionality of the MNsure IT system on October 1, 2013, resulted in a
frustrating experience for many consumers who attempted to use the system. However, as
the impact of the system limitations became evident to the MNsure Board, it authorized
MNsure staff to quickly identify and implement solutions. There was an immediate focus on
identifying system issues, bringing appropriate resources to bear to address the issues, and
implementing fixes to address the issues.
In early 2014, MNsure leadership commissioned an end-to-end review of the MNsure IT
system by Optum Health. Based on the recommendations of the review, MNsure hired
Deloitte Consulting LLP as the lead vendor to assist the State in managing and
implementing MNsure IT system improvements. State operations and IT staff from MNsure,
DHS and MN.IT Services have worked tirelessly over the past year to make improvements
to provide consumers a better experience. As a result, the MNsure IT system is more stable
and is operating in a consumer-friendly manner for the 2015 open enrollment period.
OLA Key Finding 3: MNsure did too little testing of the technology it developed, and it
did not make sufficient use of state governments information
technology experts.
MNsure Response
It is generally true that too little testing was performed on the MNsure IT system prior to go
live in October1, 2013, but this should also be understood in the context of a non-negotiable
federal deadline of October 1, 2013, by which Minnesota was required to have the system
go live. As the MNsure IT system project has continued, improving the management of the
release schedule which includes allocating sufficient time for testing- has been a focus of
the project and significant improvements have been made. Initially vendor staff and
consultants were brought on board to expand the quality assurance effort. Over the past few
months, MN.IT Services has hired a state quality assurance manager and a team of quality
assurance staff to coordinate quality assurance on the project.
MNsure takes no position on the opinion of certain MN.IT Services officials that state IT
experts were not sufficiently involved in the development and implementation of the MNsure
IT system. However, MNsure does want to address the suggestion that this report provides
5
that MN.IT Services staff had no involvement in the development and implementation of the
MNsure IT System. To suggest this directly contradicts the facts of this project. While MN.IT
Services project management and oversight expertise may not have been utilized in the
early stages of this project, MN.IT Services staff have been deeply engaged in this project
and MNsures current coordination with MN.IT Services is robust. MNsure, MN.IT Services
and DHS are all focused on improving the functionality of the MNsure IT system.
OLA Key Finding 4: Because of technical problems with MNsures online enrollment
system, many Medical Assistance recipients did not receive timely
reviews of their eligibility.
MNsure Response
MNsure takes no position and defers to DHS (the state Medicaid agency) on this finding.
However, it is important to note that the appropriate reference in this finding should be to
MNsure IT system and not MNsure (the state agency).
OLA Recommendation: The Legislature should amend Minnesota Statutes chapter 62V to
ensure that MNsures future information technology work is
subject to oversight from the Office of MN.IT Services.
MNsure Response
MNsure takes no position on this recommendation.
Chapter 3: Governance
OLA Key Finding 1: The MNsure Board had little influence over the exchange operations
prior to the launch of the MNsure enrollment website.
MNsure Response
The MNsure enabling legislation directed an initial appointment of MNsure Board members
but did not provide those Board members with the authority to direct the actions of MNsure
until certain preconditions were met. These preconditions were fulfilled in late summer of
2013, and for this reason, MNsure generally agrees with this finding.
OLA Key Finding 2: MNsure staff withheld key information from the board and other
state officials during 2013.
MNsure Response
Since early 2014, MNsure leadership has placed a great emphasis on keeping the MNsure
Board and other key stakeholders informed of key developments related to agency
operations and the development of the MNsure IT system.
OLA Key Finding 3: MNsure leadership has not implemented some internal policies and
statutory requirements.
OLA Recommendation 1: The MNsure Board should ensure implementation of its
policies or, if necessary, revise policies that are not realistic to
implement.
OLA Recommendation 2: The MNsure Board should:
Adopt consumer assister compensation rates annually and
ensure that changes in the rates are published in the State
Register;
Adopt navigator, call center, and customer services policies
in 2015.
MNsure Response
As described in the report and further elaborated in this document, the Board and MNsure
staff took steps to address technical and operational issues during the first year of
operations as they became evident. For this reason, MNsure generally agrees with the
finding that it has not implemented some internal policies and statutory requirements, but
MNsure disputes the finding that certain policies and procedures have not been established
as required under Minnesota Statutes, section 62V.05, subdivision 4, paragraph a. The
business operational units with responsibility for these functions have implemented policies
and procedures to govern their day-to-day management of these functions as authorized by
the MNsure Boards Delegation of Authority Policy, and the statute places no further
requirements upon the establishment of these policies and procedures. Thus, the claim that
the policies and procedures have not been implemented is inaccurate. MNsure has
implemented this statutory requirement.
OLA Recommendation 3: The Legislature should amend state law to give the governor,
rather than the MNsure Board, authority to appoint the MNsure
chief executive officer.
MNsure Response
OLA Key Finding 4: The multi-agency governing structure for MNsures online
enrollment system lacks formal authority.
OLA Recommendation: The Legislature should establish in state law a structure for
governing MNsures online enrollment system.
MNsure Response
MNsure strongly supports the existing interagency governance structure for the MNsure IT
system. MNsure takes no position on the establishment by statute of a formal governance
structure for the MNsure IT system.
Chapter 4: Enrollment
OLA Key Finding 1: MNsure met its overall enrollment target for the first enrollment
period, but this target was seriously flawed due to a Department of
Human Services error that significantly underestimated Medical
Assistance enrollment.
MNsure Response
MNsure agrees that its overall enrollment target was accurate, but that the actual mix of
public and private consumers differed from the estimates.
OLA Key Finding 2: Survey results showed that 28 percent of individuals who enrolled in
commercial insurance through MNsure were uninsured immediately
before they enrolled.
MNsure Response
MNsure lacks the information to comment on the validity or accuracy of the survey results
presented in this finding, but can comment that the survey results appear to be generally
consistent with other third-party surveys showing a significant decrease in the rate of
Minnesotans without health insurance coverage.
OLA Key Finding 3: MNsures data reporting capabilities are weak, limiting its ability to
produce information for management and decision-making
purposes.
MNsure Response
MNsure, along with other state agencies, are focusing significant efforts on the development
and improvement of reporting capabilities. MNsure, DHS, and MN.IT Services are working to
develop and implement a data warehouse for the MNsure IT system and MNsure has
appointed a reporting manager within the Policy and Plan Management team to coordinate
the development and implementation of reporting at MNsure. MNsure generally agrees with
this finding.
8
OLA Recommendation: MNsure should develop ways to improve its access to the
applicant and enrollee data it collectsfor the purpose of
assessing MNsure performance, generating management
reports, and responding to public inquiries.
MNsure Response
In the past few months, MNsure has worked with MN.IT Services to develop and implement
an Enrollment System of Record (ESOR) that will greatly improve MNsures ability to provide
reports to management, insurance carriers, and regulators. Additionally, as described
above, MNsure is collaborating with DHS and MN.IT Services on the development and
implementation of a data warehouse for the MNsure IT System to address this issue.
MNsure generally agrees with this finding.
Chapter 5: User Experiences

OLA Key Finding 1: Consumers and the people who helped them enroll encountered
numerous technical problems during MNsures first year of
enrollment.
MNsure Response
As the impact of the IT system limitations became evident to the MNsure Board, it
authorized MNsure staff to quickly identify and implement solutions to the crisis. There was
an immediate focus on working with vendors to bring additional functionality online on an
accelerated schedule.
In the spring of 2014, MNsure leadership commissioned an end-to-end review of the
MNsure IT system by Optum Health. Based on the recommendations of the review, MNsure
hired Deloitte Consulting LLP as a lead vendor to assist the State in managing and
implementing MNsure IT system improvements. Operations and IT staff have worked
tirelessly over the past year to make improvements to provide consumers a better
experience. As a result, the MNsure IT system is more stable and is operating in a more
consumer-friendly manner for the 2015 open enrollment period.
In addition, beginning in October 2013, MNsure increased staffing levels in its contact
center, beginning with a total head count of about 38 staff at the beginning of October 2013
and for the first open enrollment, peaking at about 75 staff at the end of March 2014. As
noted in your report, MNsure has engaged vendors to provide overflow services. These
vendors have provided access to over 100 additional staff, and including up to
approximately 200 additional staff during certain peak periods. With all of this said, MNsure
generally agrees that technical problems significantly impacted user experience in the first
year of MNsures operation.
9
OLA Key Finding 2: Individuals who enrolled through MNsure generally reported more
satisfaction than dissatisfaction with the products they purchased.
MNsure Response
MNsure lacks the information to comment on the validity or accuracy of the consumer
survey results presented in this finding but can comment that the survey results appear to
be generally consistent with other third-party surveys showing consumer satisfaction.
OLA Key Finding 3: Problems with MNsures enrollment system had a significant impact
on the ability of insurers and counties to manage individuals cases.
MNsure Response
With respect to insurers, MNsure has continued to work on improving the process of
enrollment data transmission. These efforts have included the construction of an Enrollment
System of Record (ESOR) that will better meet MNsures reporting needs, including the
submission of enrollment data to insurers and regulators. In addition, it is anticipated that the
current development and implementation of a data warehouse for the MNsure IT System will
assist in addressing this issue. MNsure generally agrees with this finding.
10
Chapter 6: Operations
OLA Key Finding 1: Many consumers were referred back and forth between brokers and
navigators, due to differences in the roles and compensation
practices for these assisters.
MNsure Response
MNsure has improved and continues to improve the experience through new referral
processes, enrollment tools, and communications with navigators, brokers, and counties.
Our experience has been that most brokers will assist consumers regardless of whether the
consumer obtains public or private insurance. Likewise, our experience has been that most
consumers who sought assistance from navigators, were able to complete enrollment
without consulting a broker. State law prohibits navigators from providing advice on the
selection of insurance products and under federal law brokers cannot be navigators.
Navigators may guide the consumer through the selection process and assist consumers in
using the selection tools and information available through MNsure. Consumer questions on
the merits of one insurance product over another must be referred to a licensed insurance
broker. Due to this nuanced distinction, consumers occasionally must work with multiple
assisters. For these reasons, MNsure does not dispute that some referrals occurred that
were disruptive to a customer service experience, but it disagrees with the suggestion that
these were common or widespread.
OLA Key Finding 2: MNsures contact center failed to provide adequate customer
service during the first open enrollment period.
MNsure Response
Call volume in the first open enrollment period was higher than expected. Many of the calls
were in relation to technical issues. MNsure contact center staff rose to this unexpected
challenge.
Since January 2014, MNsure has built out its customer service operations and now provides
a number of ways consumers can receive assistance. For many customers, the MNsure
Contact Center toll-free line is their first stop and their issue is usually resolved in the first
call. The number of calls to the MNsure Contact Center has dramatically increased, while
the average hold times have remained low. For example, for the week of December 14,
2014 (a high volume week because of a deadline), the MNsure Contact Center received
35,598 calls with the average hold times remaining under 10 minutes.
11
OLA Key Finding 3: MNsure has some -but not completeability to analyze who has
accessed private data on enrollees.
MNsure Response
MNsure is providing no response to this finding because doing so would likely result in the
disclosure of security information as defined by Minnesota Statutes, section 13.37,
subdivision 1, paragraph a, and MNsure is prohibited from making such a disclosure under
Minnesota law.
OLA Recommendation 1: MNsure and DHS should ensure that brokers are fairly
compensated for enrolling consumers in insurance through
MNsure.
MNsure Response
OLA Recommendation 2: MNsure should improve its referral guidance for customer
service staff in its contact center and in DHS and county call
centers.
MNsure Response
Contact center staff are constantly updating their reference materials and are working with
contact centers at DHS to improve transfers of calls related to public programs.
12
Forthcoming OLA Evaluations

Managed Care Organizations Administrative Expenses
Mineral Taxation
Minnesota Board of Nursing: Complaint Resolution Process
Minnesota Film and TV Board
Government Operations (continued)

Fiscal Notes, February 2012
Capitol Complex Security, May 2009
County Veterans Service Offices, January 2008
Recent OLA Evaluations
Health
Minnesota Health Insurance Exchange (MNsure),
February 2015
Financial Management of Health Care Programs,
February 2008
Nursing Home Inspections, February 2005
Agriculture
Agricultural Commodity Councils, March 2014
Green Acres and Agricultural Land Preservation
Programs, February 2008
Pesticide Regulation, March 2006
Criminal Justice
Health Services in State Correctional Facilities, February
2014
Law Enforcements Use of State Databases, February 2013
Public Defender System, February 2010
MINNCOR Industries, February 2009
Substance Abuse Treatment, February 2006
Education, K-12, and Preschool
Special Education, February 2013
K-12 Online Learning, September 2011
Alternative Education Programs, February 2010
Q Comp: Quality Compensation for Teachers,
February 2009
Charter Schools, June 2008
Education, Postsecondary
Preventive Maintenance for University of Minnesota
Buildings, June 2012
MnSCU System Office, February 2010
MnSCU Occupational Programs, March 2009
Energy
Renewable Energy Development Fund, October 2010
Biofuel Policies and Programs, April 2009
Energy Conservation Improvement Program, January 2005
Environment and Natural Resources
Recycling and Waste Reduction, February 2015
DNR Forest Management, August 2014
Sustainable Forest Incentive Program, November 2013
Conservation Easements, February 2013
Environmental Review and Permitting, March 2011
Natural Resource Land, March 2010
Watershed Management, January 2007
Government Operations
Councils on Asian-Pacific Minnesotans, Black Minnesotans,
Chicano/Latino People, and Indian Affairs, March 2014
Helping Communities Recover from Natural Disasters,
March 2012
Human Services
Medical Assistance Payment Rates for Dental Services,
March 2013
State-Operated Human Services, February 2013
Child Protection Screening, February 2012
Civil Commitment of Sex Offenders, March 2011
Medical Nonemergency Transportation, February 2011
Personal Care Assistance, January 2009
Housing and Local Government
Consolidation of Local Governments, April 2012
Jobs, Training, and Labor
State Protections for Meatpacking Workers, 2015
State Employee Union Fair Share Fee Calculations,
July 2013
Workforce Programs, February 2010
E-Verify, June 2009
Oversight of Workers Compensation, February 2009
JOBZ Program, February 2008
Misclassification of Employees as Independent Contractors,
November 2007
Miscellaneous
The Legacy Amendment, November 2011
Public Libraries, March 2010
Economic Impact of Immigrants, May 2006
Liquor Regulation, March 2006
Gambling Regulation and Oversight, January 2005
Transportation
MnDOT Selection of Pavement Surface for Road
Preservation, March 2014
MnDOT Noise Barriers, October 2013
Governance of Transit in the Twin Cities Region,
January 2011
State Highways and Bridges, February 2008
OLA reports are available at www.auditor.leg.state.mn.us or by calling 651-296-4708.

Mnsure PDF

Uploaded by

Copyright:

Available Formats

Mnsure PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Mnsure PDF

Uploaded by

Copyright:

Available Formats

OLA

OFFICE OF THE LEGISLATIVE AUDITOR

Program Evaluation Division

The Program Evaluation Division was created

James Nobles, Legislative Auditor

Topics for evaluations are approved by the

OFFICE OF THE LEGISLATIVE AUDITOR

2. EXCHANGE DEVELOPMENT AND IMPLEMENTATION

APPENDIX A: Glossary of Terms

MINNESOTA HEALTH INSURANCE EXCHANGE (MNsure)

RECENT PROGRAM EVALUATIONS

MINNESOTA HEALTH INSURANCE EXCHANGE (MNsure)

In its first year of

Minnesota is 1 of 13 states that are

About 28 percent of MNsure

During its first year, MNsure failed

MNsure implemented its enrollment

MNsure staff withheld information

MNsure reported that it met its

Many people who bought insurance

The Legislature should amend state

The Legislature should amend

MNsure and DHS should ensure

MNsure should improve its ability

MINNESOTA HEALTH INSURANCE EXCHANGE (MNsure)

federal governments rules that

began. The tests were limited in

A 2014 analysis by University of

MINNESOTA HEALTH INSURANCE EXCHANGE (MNsure)

insurance they had immediately prior to

The Legislature should reconsider

During MNsures first year, many

MNsure is governed by a seven-person

MNsure provided inadequate

The Governor appointed MNsures

he Affordable Care Act of 2010 significantly changed the federal

To what extent has Minnesotas state-based health insurance

What are the characteristics of MNsure enrollees? Has the exchange

Was the exchange adequately planned and implemented? Has

Has MNsures governing board exercised sufficient oversight of the

To address these questions, we solicited input through interviews and written

MINNESOTAS HEALTH INSURANCE EXCHANGE (MNsure)

We analyzed a variety of documents, including statutes, contracts, planning

ccording to a recent analysis, 366,000 Minnesotansor 7 percent of

KEY FINDINGS IN THIS CHAPTER

Minnesotas state health insurance exchange (MNsure) has some

In contrast to practices in many other states, Minnesotas exchange

FEDERAL HEALTH CARE REFORM

MINNESOTA HEALTH INSURANCE EXCHANGE (MNsure)

program covered specific categories

Minnesotans Sources of Health

SOURCE: Julie Sonier, Elizabeth Lukanen, and

A resource where individuals, families, and small businesses

Exhibit 1.1: Types of State Health Insurance

State-Based Exchange (14)

California, Colorado, Connecticut, District of Columbia,

State-Federal Partnership (5)

Arkansas, Delaware, Illinois, Iowa, New Hampshire

Federal Exchange (32)