Cho Law

The Myth of Choice of Law: Rethinking Conflicts
Author(s): Kermit Roosevelt III

Source: Michigan Law Review, Vol. 97, No. 8 (Aug., 1999), pp. 2448-2538
Published by: The Michigan Law Review Association
Stable URL: https://www.jstor.org/stable/1290191
Accessed: 28-02-2020 01:02 UTC
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms
The Michigan Law Review Association is collaborating with JSTOR to digitize, preserve and
extend access to Michigan Law Review
This content downloaded from 34.192.2.131 on Fri, 28 Feb 2020 01:02:02 UTC
All use subject to https://about.jstor.org/terms
THE MYTH OF CHOICE OF LAW:
RETHINKING CONFLICTS
Kermit Roosevelt III*
TABLE OF CONTENTS
I. INTRODUCTION ..................................... 2449

II. CONFLICTS THEORY ................................ 2454
A. Vested Rights: Joseph Beale .................... 2455
B. Legal Realism: Walter Wheeler Cook .......... 2458
C. Interest Analysis: Brainerd Currie ............. 2461
D. Current Theory ................................ 2466
E. Methods and Objectives: What Interest Analysis
Is, What a Conflicts Theory Should Be......... 246
III. RETHINKING THE THEORY: FROM CHOICE TO
CONFLICT .......................................... 2471
A. Rights and Their Critics ........................ 2471
1. The Failure of Vested Rights /
Territorialism .............................. 2472
2. Salvaging the Concept of Rights.......
B. Governmental Interests and Their Critics ....... 2477
1. The Weakness of Currie's Approach ....... 2477
2. Salvaging the Concept of Interest ........... 2479
C. The Positive Account ........................... 2482
1. Thinking in Terms of Rights ...............
2. Thinking in Terms of Conflict .............
IV. DEFENDING THE THEORY: TWO EXAMPLES ....... 2488
A. The Unseen Conflict: Purely Domestic Cases.. 248
B. The Easy Conflict: State Law vs.
Federal Law ................................... 2493
C. Back to Choice? State Law vs. State Law ..... 2498
V. CONFLICTS AND THE CONSTITUTION ............... 2503
A . Introduction .................................... 2503
B. The Two Clauses ............................... 2511
1. Full Faith and Credit ....................... 2513
* Senior Research Scholar, Yale Law School; Resident Fellow, Informat

Project. A.B. 1993, Harvard; J.D. 1997, Yale.- Ed. I thank Bruce Ackerman,
David Franklin, Allison Moore, Richard Primus, Samuel Rascoff, James W
Stephen F. Williams for their patience and generosity in discussing the ideas
into this article. Of course, none bears any responsibility for the results.
2448
August 1999] Rethinking Conflicts 2449
2. Privileges and Immunities .................. 2516

VI. TOWARD A CONSTITUTIONAL JURISPRUDENCE OF
CONFLICTS ......................................... 2518
A. Rules of Scope and the Constitution: T
M yths .......................................... 2520
1. The Myth of the Unprovided-for Case ..... 2
2. The Myth of the False Conflict ............. 2525
B. Conflicts Rules and the Constitution .......... 2527
1. Conflicts Rules and Law: Full Faith and
Credit ...................................... 2528
2. Conflicts Rules and Domicile: Priv
and Immunities ............................ 2530
C. A Final Wrinkle ................................ 2533
VII. CONCLUSION ....................................... 2534
I. INTRODUCTION
Choice of law is a mess. That much has become a truism. It is a

"dismal swamp,"' a morass of confusion, a body of doctrine "killed
by a realism intended to save it,"2 and now "universally said to be a
disaster."3 One way to demonstrate its tribulations would be to
look at the academic dissensus and the hopelessly underdetermina-
tive Restatement (Second) of Conflict of Laws.4 Another would be
to examine the Supreme Court's abdication of the task of articulat-
ing constitutional constraints on state choice-of-law rules.5 This ar-
ticle will do both. At the outset, though, I want to suggest that one
need look no further than the nomenclature of the subject. I do not
mean the arcane terminology - depeSage, renvoi, retorsion, false
conflicts, comparative impairment, and unprovided-for cases
that falls liltingly from the tongues of conflicts scholars and crushes
listeners into bemusement or horror. I claim instead that the con-
ceptual difficulties of this field can be discerned at the broadest
level of generality, in the dual names of the subject itself: "Choice
of Law" and "Conflict of Laws."6
1. William L. Prosser, Interstate Publication, 51 MICH. L. REV. 959, 971 (1953).

2. Lawrence Lessig, The Zones of Cyberspace, 48 STAN. L. REV. 1403, 1407 (1996).
3. William L. Reynolds, Legal Process and Choice of Law, 56 MD. L. REV. 1371, 1371
(1997).
4. See infra section II.D.
5. See infra section VI.A.
6. Joseph Beale, whose theory of vested rights was received wisdom for the first half of
the twentieth century, considered the question of nomenclature sufficiently important to
merit five sections of his treatise. See 1 JOSEPH H. BEALE, A TREATISE ON THE CONFLICT
OF LAWS ?? 1.15-1.19 (1935). Beale admitted that "conflict of laws" was not "exactly accu-
rate" and commented that "'[t]he only conflict is among the legal authors who are doing this
2450 Michigan Law Review [Vol. 97:2448
The mere existence of multiple monikers should not surprise.

Areas of legal study often go by more than one name. The class
called "Federal Jurisdiction" at one law school might be "Federal
Courts" at another; the same is true for "Corporations" and "Busi-
ness Organizations." Sometimes these names are synonyms; other
times the relation is obvious enough to need no explanation. Fed-
eral courts exercise federal jurisdiction, and the study of one is the
study of the other. Conflicts nomenclature is less transparent. An
ordinary speaker of English might be puzzled to learn that "Choice
of Law" and "Conflict of Laws" denote the same area. When laws
conflict, one might think, the question is not which law should b
chosen but rather which law prevails.7
Legal training teaches us otherwise. When laws conflict, w
learn, courts decide which law to apply. There is almost never
unique "right" answer to the question. More precisely, there is n
right answer that can be articulated without adopting what L
Brilmayer calls the "internal perspective":8 the perspective of
particular forum state. From the subjective perspective of a parti
ular forum, there may be a determinate answer, given by th
choice-of-law rules of that state. But different states will give differ-
ent answers about the same set of facts. If a case has contacts with
a number of different jurisdictions, each may apply its own law if
the case comes to its courts.9 Thus the answer to the question
"what law governs this case?"10 will often vary depending on the
forum in which the suit is brought.
This result may seem natural if we suppose that choice-of-law
rules simply compose part of a state's substantive law." Substan-
work."' He continued, "'[y]et since the expression is consecrated by good use and is simple
we may well make use of it."' Id. ? 1.19 (quoting COMTE DE VAREILLES-SOMMIERES, LA
SYNTHESE DU DROIT INTERNATIONAL PRIVE xviii (1897)). Interestingly, Beale also mentions
Eduardo Cimbali, who argued that the "false designation of the subject" had led scholars
astray. Id. (citing EDUARDO CIMBALI, DI UNA NUOVA CATEGRICA DI CONFLITrI DE LEGGE
(1892)).
7. For precisely this reason Eugene Scoles and Peter Hay find fault with the phrase "Con-
flict of Laws," arguing that because forum choice-of-law rules will select a governing law,
there is no conflict between laws. See EUGENE F. SCOLES & PETER HAY, CONFLICT OF LAWS
? 1.1 (2d ed. 1992).
8. See LEA BRILMAYER, CONFLICT OF LAWS 1 (2d ed. 1995) (emphasis omitted).
9. See Allstate Ins. Co. v. Hague, 449 U.S. 302 (1981) (plurality opinion).
10. One of my claims is that this is a counterproductive way of framing the question. See
infra section IV.B. In fact, I will be changing a fair amount of what I find to be misleading
terminology. At the moment, though, I am simply discussing the current understanding of
conflicts.
11. See Nicholas deBelleville Katzenbach, Conflicts on an Unruly Horse: Reciprocal

Claims and Tolerances in Interstate and International Law, 65 YALE L.J. 1087, 1099 (1956)
(stating that the task of courts always "involves a choice of law"). Beale also held the posi-
tive law differs from state to state, and states will reach different
answers about the legal consequences of the same facts. Thus it is
not surprising that a constellation of facts entitling the victim of a
car accident to recover under the tort law of a state employing a
negligence standard will not allow recovery under the gross negli-
gence standard of another state. Similarly, one might think, it is not
surprising that one state might conclude that the appropriate law is
the law of the state where the accident took place, while another
might look to the law of the victim's domicile.12
This inconsistency arises from, essentially, Brilmayer's "internal
perspective." One of the major goals of this article is to suggest
that this way of viewing the choice-of-law problem is mistaken
not because the alternative that Brilmayer mentions, the "external
perspective," is correct,13 but because the dichotomy itself is false
and the internal perspective fails on its own terms. In fact, I will
argue, conceiving of choice-of-law rules as substantive domestic law
does not legitimize the variance of results across forums. It merely
masks the illegitimacy, hiding the conflict between laws behind the
veil of choice of law, and the veil does not stand up to analysis.
To start seeing this, take a step back. Return to the purely do-
mestic context and imagine a plaintiff who comes to court alleging
that a wrong has been committed against him. He claims that some
tion that choice-of-law rules were part of forum law; in fact, it is a necessary consequence of
his principle that law can operate only within a state. See 1 BEALE, supra note 6, ? 5.4, at 53
("Conflict of Laws is part of the law of the forum. It is quite obvious that since the only law
that can be applicable in a state is the law of that state, no law of a foreign state can have
there the force of law.... The foreign law is a fact in the transaction."). Of course, he did not
conclude that variance was therefore permissible; he found general organizing principles in
the nature of law. Id. ? 4.12, at 46.
I think there is something different about choice-of-law rules, though it turns out to be
essentially that they are so much a matter of federal concern as to be subject to severe consti-
tutional strictures. (On my account, not the Supreme Court's. See infra Part V.)
12. In fact, variance between forums is more likely to take the form of each forum's
looking to its own law. This tendency, which Michael Green calls "lexforism," is the deeply
troubling aspect of disuniformity. See Michael S. Green, Note, Legal Realism, Lex Fori, and
the Choice-of-Law Revolution, 104 YALE L.J. 967, 967 (1995).
13. The external perspective supposes that choice-of-law rules are uniquely determined
by objective principles external to any state's law and hence that each forum must reach the
same conclusion. See BRILMAYER, supra note 8, at 1. The theoretical approach that I advo-
cate will destroy the dichotomy. The Constitution is, in an interesting way, both internal and
external. It is internal in that, by virtue of the Supremacy Clause, it is part of the local law of
each state. See, e.g., Claflin v. Houseman, 93 U.S. 130, 137 (1876). It is external in that it
imposes rules state law cannot change - again, by virtue of the Supremacy Clause. See, e.g.,
Felder v. Casey, 487 U.S. 131, 138 (1988). And what I will suggest is that the Constitution
dictates a mixed perspective - both internal and external. It does not require consistency
across states, so that each state must adopt the same rules (the fully external perspective).
Rather, it requires a sort of consistency within states, a lesser degree of freedom than that
recognized by the internal perspective. This will surely seem cryptic, but a full explanation
must await some ground-laying.
law - here, let us assume local tort law - gives him a right to
relief. The court may disagree with this claim. It may be that the
law gives him no right on these facts, or that the law provides the
defendant with a defense that precludes liability. Either of these
determinations is an appropriate judicial decision. But what if the
court simply refuses to consider his tort claim because, it says, con-
tract law governs the case? This should seem odd.14 The plaintiff
has asserted a right, and surely the court must either recognize that
right or refuse to recognize it. Either he has stated a claim or he
has not. To avoid this dichotomy by invoking a different law seems
at best an oblique rejection of the plaintiff's claim, at worst a deci-
sion based on something other than whether he has an enforceable
right.15 It seems, in short, that the court has made a choice, not
resolved a conflict.16
The substitution of choice for conflict, I will argue, is the funda-

mental error of conflicts jurisprudence.17 It is an attempt to avoid
difficult questions that succeeds only in resolving them sub rosa,
and poorly. Conflicts between rights are a common feature of law-
suits, and in most circumstances, the legal system deals with them as
conflicts: courts look to rules specifying which right shall prevail,
14. I am not objecting to the idea that a court may tell a plaintiff that, although he cannot
recover in tort, he may recover in contract. Cf. BRAINERD CURRIE, Conflict, Crisis and Con-
fusion in New York, in SELECTED ESSAYS ON THE CONFLICT OF LAWS 690, 693-94 (1963)
[hereinafter SELECTED ESSAYS] (labeling as "anachronistic" criticizing a court for "tell[ing] a
litigant that though he cannot recover in contract he may in tort"). The problem I am con-
cerned with arises when the plaintiff pleads in tort and the court refuses to evaluate the tort
claim.
15. Courts do not ordinarily apply law not invoked by the parties. Affirmative defenses,
for example, are waived if not raised at the proper stage. If the court refuses to honor the
law the plaintiff pleads, I will suggest, it must be because that law is legally unavailable. And
that means either that some other law operates to interfere with it, or that the law the plain-
tiff invokes grants him no rights.
16. Of course, choosing one law effectively resolves the conflict between them: it awards
victory to the chosen law. But it does so, we will see, without a consideration of the factors
that should govern a resolution of the conflict. Identifying these factors is, of course, a diffi-
cult task. Common policy considerations include predictability, discouragement of forum
shopping, and the rather amorphous goal of fairness to litigants. These are certainly values
that a system for resolving conflicts between sovereigns should seek to promote. My pre-
scriptions, when they come, will indeed seek to promote them. But they will not rest simply
on those values; instead, they will be drawn from another body of law aimed chiefly at meld-
ing the several states into a federal union: the Constitution. Without the constraints I iden-
tify, states may succumb to the temptation to promote forum interests, slighting the concerns
of other states. Within the constraints, states may do many things to promote the canonical
conflicts values. But interstate discrimination needs to be addressed first; at this moment in
our conflicts jurisprudence, it is the primary evil and the chief distraction from sound con-
flicts rules.
17. Consequently, I will refer to the subject as "conflicts," and continue to refer to
"choice-of-law rules" where appropriate. My ultimate suggestion is that things will be clearer
if we eliminate the idea of "choosing" a "governing" law, but that is a different stage of the
rocket and must drop away later.
and then express their conclusions in such terms.18 In cases where

the conflicting rights originate from different states, however, a dif-
ferent description is employed. Courts speak not of deciding which
right prevails but of choosing which law applies to the case. This
resort to choice-of-law rhetoric is peculiar for two reasons. First, it
is unnecessary. Multistate cases can be described and resolved per-
fectly easily within the vocabulary of conflicts.19 Second, it is de-
scriptively inaccurate. Interest analysis (the choice-of-law
methodology I will consider in the greatest detail) simply does not
select the law that applies to a case.20 The rhetoric of choice per-
sists in part as a conceptual hangover from the early days of con-
flicts theory,21 but it also continues to allure because it makes less
apparent the conflicts that have proven too hard to resolve.22 In-
deed, the Supreme Court swiftly backed away from its initial bold
interventions into state conflicts practices; more recently it seems to
have given up entirely.23 Consequently, there is a temptation to
deny problems we cannot solve, by framing the issue as one of
choice.
This article aims to show that things are not as bad as all that.
Interstate conflicts are a chief concern of the Constitution, and the
Constitution will allow us to deal with them. Conflicts theory has
failed to locate external constraints on state law and has actually
urged states to adopt regimes that are blatantly discriminatory
regimes that, if not explained by parochialism, are in fact self-
contradictory.24 Judicious use of garden-variety antidiscrimination
principles embedded in the Full Faith and Credit and the Privileges
and Immunities Clauses will prevent such favoritism. These consti-
tutional principles do not resolve conflicts by their own force
18. See infra section IV.B.

19. See infra section III.C.2.
20. See infra section IV.C.
21. The territorial approach to conflicts, discussed below, did in fact work by identifying
the law that governed a transaction. Interest analysis retained this vocabulary, even though
the description no longer fit the operation of the theory. See infra section IV.C.
22. When a state finds foreign law inapplicable, it may seem not to have rejected foreign
rights. See CURRIE, Notes on Methods and Objectives in the Conflict of Laws, in SELECTED
ESSAYS, supra note 14, at 177, 181-82. Effectively, of course, it has done so, and perhaps few
will find the rhetoric of choice an effective fig leaf. Whether its potential for obscuring con-
flicts is the real reason for its continued popularity is a psychological question, and my sug-
gestions along these lines are only hypotheses. It does seem to be the case that Currie saw a
difference between applying local law and rejecting foreign rights, see id., and without this
difference, interest analysis's bias toward forum law is obviously problematic. See infra sec-
tion IV.C.
23. See infra section V.A.

they do not dictate unique solutions - but they constrain the

states' resolutions in ways that produce a coherent jurisprudence of
conflicts.25 In order to see how the Constitution works, we need a
theory that frames the issue in terms of conflict, not in terms of
choice.
Part II of the article sets the stage for that theory by briefly
recounting the history of conflicts scholarship and offering a word
on methods and objectives in the conflict of laws. Part III extracts
appropriate building blocks from the rubble of previous theoretical
constructs; it then puts the blocks together, demonstrating in out-
line what the theory should look like. Part IV defends the theory
by examining two situations neglected by conventional conflicts
theory: conflicts within one state's law, and conflicts between state
and federal law. Part V examines the relevance of the Constitution,
and Part VI applies the constitutional principles thereby derived.
II. CONFLICTS THEORY
Articles about conflicts frequently begin with - or are enti

devoted to - a history of the subject.26 The need for another
recapitulation may certainly be questioned. This recounti
though, is not mere intellectual dressage. Because I intend to a
that the correct way of thinking about conflicts may be der
from the historical approaches, it is worthwhile to show both
the essential concepts already exist and how they have been p
vented from uniting into a coherent theory.
Conflicts has a rich history. To begin at the beginning mi
require a return to ancient Egypt and the wrappings of a croc
mummy, which supposedly contain the first recorded choice-o
principles.27 A full account would then consider the theories o
dieval Europe, the early English approach that did away with
problem of foreign transactions via the fiction that all event
curred in London,28 and subsequent developments in the cour
America. The perspective afforded by a thorough historical ex
tion is of significant value, for conflicts revolves around a few
and recurring themes. But the full-dress reenactment has itse
25. See infra Part VI.

26. See BRILMAYER, supra note 8, ? 1.1, at 11-13.
27. See Hessel E. Yntema, The Historic Bases of Private International Law, 2 A
COMP. L. 297, 300-01 (1953).
28. See Friedrich K. Juenger, A Page of History, 35 MERCER L. REV. 419, 436
This led to pleadings asserting, for example, that wrongful acts were committed on the
of Minorca, in the city of London. See id. at 436-37.
ready been done,29 and only a few scenes are necessary to my pro-
ject. Of course, any attempt to sketch the history as mere opening
act for a theoretical venture will inevitably be selectively incom-
plete. What follows is an account that highlights those aspects im-
portant to my project - how the issue of conflict has been
repressed, and how, in later theory, choice has taken its place.
A. Vested Rights: Joseph Beale
Although Joseph Story exerted a profound influence on the

early development of conflicts theory in America,30 my story starts
with Joseph Beale. Beale was the reporter of the First Restatement,
and his three-volume treatise on the conflict of laws is structured as
a commentary to the Restatement.31 The Restatement's task - ra-
tionalizing the law of forty-eight states - was a formidable one.
But Beale did not intend merely a catalogue of judicial decisions;
his quarry was the general common law, of which the decisions of
courts were evidence only.32 Beale's task, as he saw it, was to de-
rive the general common law of conflicts from the raw data of judi-
cial decisions.
In this enterprise he was greatly aided by a few strong principles

concerning the nature of law, rights, and remedies.33 Law, for
Beale, was fundamentally territorial, supreme within a jurisdiction
but generally powerless outside it.34 This principle gives a relatively
easy answer to the question of what law governs a particular occur-
rence. Since local law, and only local law, applies within a given
29. "[E]verything worthy of trying has been tried before, under the same or other labels."
Kurt H. Nadelmann, Marginal Remarks on the New Trends in American Conflicts Law, 28
LAW & CONTEMP. PROBS. 860, 860 (1963). For thorough recapitulations, see, e.g., Juenger,
supra note 28; Yntema, supra note 27.
30. See, e.g., JOSEPH STORY, COMMENTARIES ON THE CONFLICT OF LAWS, FOREIGN AND
DOMESTIC (Melville M. Bigelo ed., Boston, Little, Brown, & Co. 1883). Story extensively
developed the idea of comity as a basis for resolving conflicts. Comity does not govern inter-
state conflicts, however; the Constitution does. Story's work has value for this article primar-
ily because it illuminates the original understanding of some constitutional provisions. Apart
from that, I will largely ignore his contributions.
31. See 1 BEALE, supra note 6, at xv.
32. See id. ? 1.12, at 10. This is, of course, the jurisprudential position of Swift v. Tyson,
41 U.S. (16 Pet.) 1 (1842), which Beale cited approvingly several times. See, e.g., 1 BEALE,
supra note 6, ? 3.3, at 22 & n.1, ? 3.5, at 26, ? 4.6, at 39 & n.l. Erie Railroad v. Tompkins, 304
U.S. 64 (1938), overruled Swift just three years after the publication of Beale's treatise. One
cannot avoid some sympathy for an author whose 2000-page magnum opus, the product of
over twenty years of labor, enjoyed such a brief reign before one of its primary supports was
unceremoniously knocked away. Worse was to follow.
33. Nowadays Beale's first principles appear to be somewhat arbitrary assumptions, but
within the jurisprudential climate of his day, they were fairly unremarkable.
34. See 1 BEALE, supra note 6, ? 4.12, at 45-46, ? 59.2, at 308, ? 61.1, at 311.
jurisdiction, local law must determine the consequence of acts

within that jurisdiction: "If two laws were present at the same time
and in the same place upon the same subject we should also have a
condition of anarchy. By its very nature law must apply to every-
thing and must exclusively apply to everything within the boundary
of its jurisdiction."35 If suit is brought within that jurisdiction,
courts will obviously apply local law. Indeed, Beale denied the abil-
ity of courts to apply any but their own local law - though this
included his general common law.36
The transparent workings of the territorial model become some-
what more turbid when suit is brought in a jurisdiction other than
the one in which the litigated transaction took place. In such cir-
cumstances, territoriality might seem at war with itself: If courts
can apply only local law, but foreign law must determine the conse-
quences of acts in foreign states, how are parties ever to obtain re-
lief in courts of other jurisdictions? Beale's solution to this problem
relied on a somewhat complicated taxonomy of rights, which he
claimed to derive from the "difference made by our law in treating
rights of the different classes with respect to the law creating and
having power over them."37 On his account, law protects interests;
these protected interests he terms primary rights.38 The violation of
a primary right gives rise to a secondary right - a right of redress.39
This right vests at the moment of the violation of the primary right
and thereafter may be considered much like personal property of
the injured party.40 In particular, it may be brought into other fo-
rums and sued upon. Forum courts, in granting relief, are not ap-
plying foreign law but simply recognizing the secondary rights
vested under foreign law.41 To determine whether a right has
35. Id. ? 4.12, at 46.

36. See id. ?? 3.4, 5.4.
37. Id. ? 8A.9, at 66.
38. See id.; see also id. ? 8A.6. Thus, "[p]arties are bound, not by the law, but by obliga-
tions created by the law." Id. ? 3.4, at 25.
39. See id. ? 8A.25.
40. See, e.g., Slater v. Mexican Natl. R.R. Co., 194 U.S. 120, 126 (1904) (Holmes, J.);
Loucks v. Standard Oil Co., 120 N.E. 198, 200 (N.Y. 1918) (Cardozo, J.).
41. Indeed, the granting of what Beale termed a "remedial right" - an actual damages
claim - was in fact a matter of forum law. See 1 BEALE, supra note 6, ? 8A.28, at 85-86.
This allowed the forum to recognize the right to redress while retaining some flexibility in
crafting a remedy - a feature Holmes exploited in Oceanic Steam Navigation Co. v. Mellor,
233 U.S. 718 (1914).
vested, the forum court might need to examine foreign law, but as a
question of fact, not law.42
One aspect of Beale's account is of special interest for my pur-
poses. The aspect is this: given his territorial understanding of law,
there is no such thing as conflict between laws. Each is supreme
within its jurisdiction and does not, by the nature of law, extend
beyond. Because laws operate only territorially, a state's law can-
not create rights from transactions occurring outside its borders.43
Denying the application of foreign law to a transaction occurring
within the forum state, then, is not the denial of a foreign right but
simply a recognition of the nature of law. On this account, laws
cannot even come into contact with each other, much less conflict.44
The task of courts in multistate cases is truly to identify which law
applies, which law creates the parties' rights and obligations.
The elimination of conflicts made Beale's model pleasingly sim-
ple to operate,45 but his approach would ultimately be judged not
for its theoretical niceties but for its real-world results. From this
perspective, hiding difficult questions is not a virtue. Metaphysical
observations about the nature of law do not resolve concrete
problems, and Beale's theoretical purity was purchased at the
of ignoring practical issues. This preference for theory over p
made Beale an easy target for criticism. He suffered so at the
of the realists that his conflicts theory was for quite a whil
missed as an arbitrary metaphysics, based on "jejune notions
'omnipresence' which cannot 'brood' more than three miles f
home."46 But it is more a vessel of reflection, and less a bark
dogma, than such appraisals indicate.47 The internal structure
42. See 1 BEALE, supra note 6, ? 5.4, at 53 ("It is quite obvious that since the only law
can be applicable in a state is the law of that state, no law of a foreign state can hav
the force of law.... The foreign law is a fact in the transaction.").
43. "Since the power of a state is supreme within its own territory, no other st
exercise power there." Id. ? 61.1, at 311.
44. While canvassing objections to the name "conflict of laws," Beale offers a desc
that precisely fits his theory: "The laws of different sovereigns do not contend w
another for the mastery. Each one keeps within its sphere of operation, and only as
power in a foreign country when the law of that country commands or permits it. In p
a conflict is impossible." Id. ? 1.16, at 13 (quotation omitted).
45. Indeed, Beale believed that attempts to resolve conflicts were doomed to f
"Which of the two independent sovereigns should yield is a question not susceptib
solution on which all parties would agree." 3 BEALE, supra note 6, ? 53, at 1929.
46. Katzenbach, supra note 11, at 1096.
47. See BRILMAYER, supra note 8, ? 1.2, at 20 (calling Beale's theory "quaintly
vated" but "well-developed," and rejecting critics' accusation that it was "mindless
tism"). Beale did himself no favors with his vaguely Shakespearean response to acc
of dogmatism: "One cannot deny that most of the statements in this work will be d
Does not the Bar desire dogmatic statements?" 1 BEALE, supra note 6, at xiii.
ally rather elegant, its concepts interacting with a smoothness and

complexity suspiciously reminiscent of celestial spheres, phlogiston,
luminiferous ether, and other refined illusions. And many of
Beale's claims have been taken up more recently by scholars as em-
inent as Ronald Dworkin.48
Admittedly, the practical effects of the theory were somewhat

less pleasing. In order to pick a single jurisdiction where rights
vested - as the territorial principle required - Beale needed to
identify a specific act triggering the rights. Not unreasonably, given
the alternatives, he decided that this should be the last act necessary
to the existence of the cause of action.49 But the rigidity of the last
act doctrine interacted with the intricacies of tort law to produce
results that were undeniably arbitrary and verged on the bizarre.50
The serpent of the practical fatally compromised Beale's conceptual
Eden, and soon enough came the "archangels of doctrinal destruc-
tion":51 the legal realists.
B. Legal Realism: Walter Wheeler Cook
In part of the Preface entitled "Apologia," Beale noted that his

legal principles had been criticized by what he optimistically called
"a current but ephemeral school of legal philosophy"52- namely,
legal realism. The characterization was, if not whistling-past-the-
graveyard bravado, a historic underestimation rivaling that of Louis
XVI.53 Beale struck closer to the mark when he commented that
48. Beale believed that cases had unique right answers and that courts enforce rights that
exist prior to and independent of their decisions. See 1 BEALE, supra note 6, ?? 3.1-3.4. This
cluster has obvious affinities with Dworkin's account in RONALD DWORKIN, LAW'S EMPIRE
(1986). Indeed, elements of Beale's conflicts theory are enjoying a modest resurgence as part
of a conflicts counterrevolution. See Perry Dane, Vested Rights, "Vestedness," and Choice-of-
Law, 96 YALE L.J. 1191 (1987). The present article belongs to the same tradition; like
Dane's, it attempts to resuscitate a few key concepts that Beale got right.
49. See RESTATEMENT (FIRST) OF CONFLICT OF LAWS ? 377 (1934).
50. For example, the victim of a poisoning might travel through many states before the
effects of the poison were felt. Beale, reasoning that no tort exists without an injury, would
look to the law of the state where the poison took effect, since that effect is the last occur-
rence necessary to the vesting of the right. But which state the victim happens to be in when
this occurs has little to do with any policy relevant to conflicts of law; nor, without the strong
territorialist assumption, does it seem to have much to do with the nature of law. I owe this
example to Lea Brilmayer, see BRILMAYER, supra note 8, ? 1.3, at 25-26, who points out
further that "it is no easy matter to determine [as the First Restatement requires] where the
'deleterious substance takes effect upon the body,'" id. (alteration in original) (quoting RE-
STATEMENT (FIRST) OF CONFLICT OF LAWS ? 377 cmt. 2, illus. 2 (1934)).
51. Katzenbach, supra note 11, at 1107.
52. 1 BEALE, supra note 6, at xiii.
53. Louis's diary entry for July 14, 1789, the date of the storming of the Bastille, reads
simply "Rien." ("Nothing."). SIMON SCHAMA, CITIZENS 419 (1989). In fairness to Louis, this
recorded an unsuccessful hunt.
"one who hears the evening bell must hasten his work, if he is to
finish it."54 The ferocity, and the success, of the realist assault on
Beale's verities are well documented in the scholarly literature.
Katzenbach says that the vested rights theory was "brutally mur-
dered" by Walter Wheeler Cook,55 and Brainerd Currie's oft-
quoted evaluation was that Cook "discredited the vested-rights the-
ory as thoroughly as the intellect of one man can ever discredit the
intellectual product of another."56 Cook, for his part, made no se-
cret of his intent to uproot and discard Beale's approach in its en-
tirety: "[U]ntil the intellectual garden is freed of the rank weeds in
question," he wrote, "useful vegetables cannot grow and
flourish."57
The realists directly attacked the idea of vested rights. In part

this was a matter of pointing out practical difficulties with the ap-
proach. The principle that rights vest in the place of the tort seems
easy enough to apply, but in fact it encounters serious difficulties
when the events which make up the tort occur in different jurisdic-
tions.58 As mentioned earlier, Beale accorded decisive importance
to the famous "last act" necessary to the vesting of the secondary
right.59 Unfortunately, the domestic laws of different jurisdictions
might disagree about which act was the last one necessary to the
vesting of a right, producing situations in which each state believed
that rights vested within its territory - or, equally distressing,
within the territory of the other state. To resolve this problem
Beale had invoked the general common law,60 a maneuver that be-
came much less plausible after Erie Railroad v. Tompkins61 rejected
the existence of such an entity.
Still, problems of application were ancillary to Cook's project.
His central attack was aimed at the jurisprudential groundwork of
Beale's theory, his understanding of the nature of law and rights.
54. 1 BEALE, supra note 6, at xii.

55. See Katzenbach, supra note 11, at 1087-88.
56. CURRIE, On the Displacement of the Law of the Forum, in SELECTED ESSAYS, supra
note 14, at 3, 6; see also David F. Cavers, A Critique of the Choice-of-Law Problem, 47 HARV.
L. REV. 173, 175-76 (1933) ("Indeed, one may now wonder how any juristic construct such as
'right' could have been accepted as fundamental in the explanation of any important aspect
of judicial activity."); Juenger, supra note 28, at 435 ("pure sophistry").
57. WALTER WHEELER COOK, THE LOGICAL AND LEGAL BASES OF THE CONFLICT OF
LAWS at ix (1942).
58. See, e.g., id. at 314-18.
59. See supra note 49 and accompanying text.
60. See, e.g., 1 BEALE, supra note 6, ? 3.1-3.6; RESTATEMENT (FIRST) OF CONFLICT OF
LAWS ? 377 cmt. d (1934).
61. 304 U.S. 64 (1938).
Rejecting Beale's conception of "theoretical law" as "the body of

principles worked out by the light of reason and by general usage,
without special reference to the actual law in any particular state,"62
Cook warned that "we must as always guard ourselves against
thinking of our assertion that 'rights' and other legal relations 'exist'
or have been 'enforced' as more than a conventional way of
describing past and predicting future behavior of human beings
judges and other officials."63 He therefore opposed the reification
of rights, arguing, in the words of the ever-quotable Holmes, that
"'a right is only the hypostasis of a prophecy.' 64
Cook's positive program for resolving choice-of-law questions
was not as theoretically well-developed as that of his predecessor
(Joseph Beale) or successor (Brainerd Currie).65 This is under-
standable, given his pragmatic and antimetaphysical bent, but it
means that his importance to this article lies largely in his critique.66
His most notable positive contribution was the "local law theory,"
which asserted that states could apply only their own law. Beale, of
course, agreed with this proposition; it was what necessitated his
distinction between foreign law and the rights that vested under it.
Cook went further, however, arguing that states did not enforce for-
eign rights but rather applied "the rule of decision which the given
foreign state or country would apply, not to this very group of facts
now before the court of the forum, but to a similar but purely do-
mestic group of facts involving for the foreign court no foreign ele-
ment."67 This theory offered a solution to the problem of renvoi
62. 1 BEALE, supra note 6, ? 1.12, at 9.

63. COOK, supra note 57, at 33. One obvious problem with this "predictive" theory of law
is that it fails to explain the thinking of a judge deciding a case, whose attempts to discern the
correct rule of law are surely not attempts to predict his own behavior. Cook points out this
difficulty, then comments that "our discussion at this point does not require further consider-
ation of the matter." Id. at 30 n.52a.
64. Id. at 30 (quoting OLIVER WENDELL HOLMES, Natural Law, in COLLECTED LEGAL
PAPERS 310, 313 (1920)). Of course, Holmes's Supreme Court opinions constituted canonical
applications of the vested rights theory. See, e.g., Western Union Tel. Co. v. Brown, 234 U.S
542, 547 (1914); Slater v. Mexican Natl. R.R. Co., 194 U.S. 120, 126 (1904). Cook offers a
creative reconstruction of Slater as rooted in policy judgment rather than in the vested rights
theory, which he then cautiously attributes to Holmes. See COOK, supra note 57, at 35. It
may be safer to rest with the observation that Holmes contained multitudes.
65. See, e.g., Larry Kramer, Interest Analysis and the Presumption of Forum Law, 56 U.
CHI. L. REV. 1301, 1301 (1989) ("[W]hile [the realist] criticism successfully undercut the intel-
lectual foundation of traditional choice of law theory, a plausible alternative was not pro-
posed until the 1950s.").
66. Cook would probably not have been unhappy to be identified more with his negative
than with his positive analysis; he believed that "[t]he removal of the weeds is ... as construc-
tive in effect as the planting and cultivation of the useful vegetables." COOK, supra note 57,
at ix.
67. Id. at 21.
the interminable ping-pong created when the choice-of-law rules of

two states each directed their courts to look to the other state's law
- but had little other significance.68 The elaboration of a choice-
of-law theory robust enough to direct courts was the work of
Brainerd Currie.
C. Interest Analysis: Brainerd Currie

Currie's important contribution, and perhaps his most signifi-
cant difference with Beale, was to analyze law not as an objectively
existing entity but as a tool of state policy. Beale's theory deter-
mined which law applied by "deduction from territorial postu-
lates,"69 without examining the content of the law. Essentially a
"jurisdiction-selecting" approach, it picked not a particular law but
the sovereign with authority to legislate the consequences of the
transaction.70 In contrast, Currie realized that the first step in
choice of law must be an analysis of the laws contending for appli-
cation. If laws are instruments of state policy, it follows that when
application of a state's law will not advance its policies, the state
would not want its law applied.71
This analysis revealed the arbitrariness of the vested rights ap-
proach. Resolving all choice-of-law questions by territorial princi-
ples results in subordinating the interests of the nonselected state
68. Hessel Yntema calls the local law theory "empty luggage." See Yntema, supra note
27, at 316.
69. See Cavers, supra note 56, at 192-94.
70. See id. at 194.
71. As the text above says, this "instrumental" approach to law is generally considered a
significant difference between Beale and Currie. See, e.g., Lea Brilmayer, Rights, Fairness,
and Choice of Law, 98 YALE L.J. 1277, 1284 (1989). Yet it seems quite easy to assimilat
Currie's insight into Beale's system by reasoning that if the state does not want its law ap
plied, it does not extend its law to cover the transaction. Thus the law attaches no lega
consequences; it creates no rights or liabilities for parties to sue on. (Following Perry Dane,
will call such restrictions on the extension of rights "rules of scope." See Dane, supra note 48,
at 1203-04. I discuss rules of scope at more length in section II.E.) Beale never suggeste
that state laws must always have maximum scope - obviously, he was quite emphatic about
territorial limitations. True, he did not see that limitation as the sovereign's choice, but if
state statute provided that only local citizens could recover for in-state torts, Beale woul
presumably not have maintained that out-of-staters acquired rights thereby. His fascinatio
with the general common law obscures this point but provides no theoretical obstacle. In
truth, it was Currie who tended to disregard states' expressions of intent not to have their
laws applied to cases in which he found them "interested," creating willy-nilly the practica
equivalent of rights. See BRILMAYER, supra note 8, ?? 2.5.2-2.5.4 (noting that Currie ignore
state choice-of-law rules dictating application of another state's law, though these seem lik
expressions of lack of interest). The fact that the language of rights is more characteristic of
Beale than of Currie has been made to bear more jurisprudential weight than it can easily
support. See infra note 166. See generally Green, supra note 12. Now that the clamor has
died down, the realist attack on vested rights looks rather like any other attempt to over
throw an entrenched vocabulary seen as essentially conservative. The rhetoric may be philo
sophical, but the stakes are political.
without even ascertaining that the selected state has an interest that
will be promoted by application of its law.72 Currie correctly sug-
gested that this made little sense, and that the vested rights ap-
proach lumped together quite dissimilar cases precisely because it
determined the applicable law without examining its content.
Currie's approach, by contrast, allowed for distinctions among
cases with multistate contacts. Currie began with a presumption
that the forum would apply its own law.73 If one party suggested
the application of another state's law, the court was to analyze the
substantive law at issue to determine whether the forum or the
other state had an interest in the application of its own law. If o
one state has an interest, the case is what Currie called a "false
conflict." In such cases, it is appropriate to apply the law of th
only interested state. This allows for effectuation of that state's
icies and does no harm to the policies of other states, since they
by definition, not interested. If both states have an interest, t
case is a "true conflict," and more difficult to resolve. Regardi
true conflicts as insoluble, Currie suggested that the forum sho
simply apply its own law.74 A preference for forum law is not
obviously desirable method of resolving true conflicts, and Cur
later suggested that in such cases the court should reexamine t
policies at issue to see if a more moderate reinterpretation mig
eliminate one or the other interest. The last category of ca
72. Readers may note that I have now moved from the question of whether application
a law advances a state's policies to the concept of a "state interest" without much discu
of either. Currie was similarly elliptical, an unfortunate parsimony that has engendered s
stantial confusion. A fairly authoritative source, Currie's student, coauthor, and defe
Herma Kay, has suggested that a state is interested if the policies behind the particular la
issue - discerned by the ordinary process of statutory interpretation - would be prom
by the application of that law. See Herma Hill Kay, A Defense of Currie's Governm
Interest Analysis, 215 RECUEIL DES COURS 9, 50 (1984). This is the conventional wa
determine if a law is intended to apply to a particular case. See, e.g., EEOC v. Arabian
Oil Co., 499 U.S. 244, 248-59 (1991) (discussing statutory interpretation and legislative in
with respect to extraterritorial application of Title VII). Thus, when Currie wrote that
proper role of courts was to promote forum policies, he was saying simply that they sh
apply laws with an eye to legislative intent. And that is simply enforcing the rights confe
by the law - neither an unusual nor an excessively instrumentalist approach. But cf. D
supra note 48, at 1259 (claiming that only a "Decision-Based" (realist) approach would cl
that courts are "primarily charged" with advancing the policies of their states). What mak
interest analysis problematic, I will suggest, is Currie's tendency to construct legislative in
around domiciliary status. See infra section IV.C.
73. Larry Kramer has criticized this presumption as unnecessary. See Kramer, supra
65. I will suggest that it is unconstitutional, joining Dean Ely and others. See, e.g., John H
Ely, Choice of Law and the State's Interest in Protecting Its Own, 23 WM. & MARY L.
173, 180-91 (1981); Douglas Laycock, Equal Citizens of Equal and Territorial States:
Constitutional Foundations of Choice of Law, 92 COLUM. L. REV. 249, 310-12 (1992). I
also suggest that it is basically incoherent, which I think has not been done. See infra sect
IV.C.
74. See CURRIE, supra note 22, at 184.
those in which neither state has an interest, are "unprovided-for

cases." These also proved somewhat embarrassing to interest anal-
ysis.75 Since a lack of interests provides no reason to disturb the
presumption of forum law, Currie suggested that in this category of
cases too, courts should apply the law of their own states.76
The jury is still out on interest analysis. In practice, it proves
quite underdeterminative, given the difficulty in ascertaining the
policy behind a particular law.77 Issues of application aside, Lea
Brilmayer has mounted a more serious attack on the central notion
of governmental interest, suggesting that it is the product of a priori
theorizing rather than conventional interpretation.78 I will have
more to say about her charges later. For present purposes, though,
two observations will suffice.
First, interest analysis makes a very important advance by con-

ceiving of multistate cases as clashes between sovereigns, each at-
tempting to impose its own regulatory scheme in furtherance of its
own policies. It is this perspective that reveals the conflict, which
Beale's analysis hid.
Second, like Beale's vested rights theory, interest analysis avoids
the difficult task of resolving conflicts between laws, though in a
somewhat different way. While Beale's account denies the possibil-
ity of conflict - only one law governs the transaction - interest
analysis admits it: indeed, it is this recognition that allows the dis-
tinction between cases that present conflicts ("true conflicts") and
those that do not ("false conflicts" and "unprovided-for cases").
But having used the concept of conflicts to dispose of cases in which
there are none, interest analysis deals with true conflicts by employ-
ing a technique that suggests they do not need to be resolved. The
technique is what I will call "personal-jurisdiction-style" analysis,
similar to the one courts use to determine whether a state's attempt
to exercise jurisdiction over a defendant violates due process.
75. See, e.g., LEA BRILMAYER, AN INTRODUCTION TO JURISDICTION IN THE AMERICAN

FEDERAL SYSTEM 240 (1986); Aaron Twerski, Neumeier v. Keuhner: Where Are the Em-
peror's Clothes?, 1 HOFSTRA L. REV. 104, 107 (1973).
76. See CURRIE, Survival of Actions: Adjudication versus Automation in the Conflict of
Laws, in SELECTED ESSAYS, supra note 14, at 128, 156, 168.
77. See BRILMAYER, supra note 8, ? 2.1.2, at 61-62.
78. See, e.g., Lea Brilmayer, Interest Analysis and the Myth of Legislative Intent, 78 MICH.
L. REV. 392 (1980) [hereinafter Brilmayer, Interest Analysis]; Lea Brilmayer, Methods and
Objectives in the Conflict of Laws: A Challenge, 35 MERCER L. REV. 555, 555 (1984) [herein-
after Brilmayer, A Challenge] ("[I]nterest analysis is methodologically bankrupt."); id. at 563
("Currie was as metaphysical as Beale."). I discuss Brilmayer's attack in more detail below.
See infra section III.B.2.
The due process test for personal jurisdiction is the familiar min-
imum contacts analysis.79 If the defendant has certain minimum
contacts with the forum, it may exercise jurisdiction. Of course,
other states may also have jurisdiction - indeed, they may have
substantially greater contacts than does the forum - but this has
no effect on the forum's ability to exercise its own.80 The upshot is
that personal-jurisdiction-style analysis does not select a unique
state. It sets a certain baseline - the constitutionally required min-
imum contacts - and any state that meets that baseline require-
ment may exercise jurisdiction.
Interest analysis proceeds in a similar way, using the baseline of
governmental interest. If a state has no interest, its law should not
be applied.81 But if a state does have an interest, there is no basis
on which to prefer any other state. All interested states meet the
baseline requirement; there is no way to choose between them, and
therefore no grounds on which an interest analyst may direct an
interested forum to apply another state's law.82 Thus, just as a state
may exercise personal jurisdiction without derogation of the juris-
diction of other states, it may exercise legislative jurisdiction - ap-
ply its own law - without claiming that its interest in the case is
greater than that of other states.83 Currie made this point quite ex-
plicitly, arguing that "[a] court need never hold the interest of the
foreign state inferior; it can simply apply its own law as such."84
79. See International Shoe Co. v. Washington, 326 U.S. 310 (1945).
80. The forum may of course dismiss the case on forum non conveniens grounds. See
Piper Aircraft Co. v. Reyno, 454 U.S. 235 (1981).
81. The exception is the unprovided-for case, in which no state has an interest. Here
Currie suggested forum law as the only plausible candidate. Larry Kramer suggests to the
contrary that a lack of interest is a lack of interest in granting relief, and that consequently
the plaintiff should lose. See Larry Kramer, The Myth of the Unprovided-For Case, 75 VA. L.
REV. 1045 (1989) [hereinafter Kramer, Myth]; see also Larry Kramer, Rethinking Choice of
Law, 90 COLUM. L. REV. 277, 293-307 (1990) [hereinafter Kramer, Rethinking Choice of
Law]. My analysis will suggest something similar.
82. This is essentially the Supreme Court's constitutional position, though its notion of
interest is even weaker than Currie's. See infra section V.A.
83. Though I will argue that personal and legislative jurisdictions are quite different, they
have similar histories. Both were originally territorial. Compare Pennoyer v. Neff, 95 U.S.
714, 722 (1877) (holding that a state can exercise personal jurisdiction only over people pres-
ent within its borders), with Allgeyer v. Louisiana, 165 U.S. 578 (1897) (rejecting application
of Louisiana law to a contract formed in New York). The Court retreated from territorialism
at about the same time in both contexts. The personal jurisdiction recantation came with
International Shoe Co. v. Washington, 326 U.S. 310 (1945); territoriality gave way with respect
to legislative jurisdiction in Cardillo v. Liberty Mut. Ins. Co., 330 U.S. 469, 476 (1947). But
see Quill v. North Dakota, 504 U.S. 298, 319-20 (1992) (Scalia, J., concurring) (distinguishing
between types of jurisdiction).
84. CURRIE, supra note 22, at 181-82.
I will claim - and this preview is for guidance only - that this
characterization is misleading. Assertions of legislative jurisdiction
involve the rejection of foreign rights; legislative jurisdiction, unlike
personal jurisdiction, is a zero-sum game. We will consequently do
better by abandoning the personal-jurisdiction-style analysis and
thinking instead in terms of a conflict between rights created by
different laws. The classic situation is one in which the plaintiff as-
serts a right derived from the law of one state, and the defendant
counters with a right derived from the law of another. From this
perspective, the fundamental question of conflicts law is simply the
ordinary legal question that arises in every case: whether the plain-
tiff has a right to recover, or whether the defendant's asserted right
blocks the plaintiff's claim. Whatever courts say they are doing, this
is the question that conventional legal thinking implies they answer
when they decide conflicts cases.
This perspective shows the difference between vested rights and
interest analysis in a slightly different light. If we examine the
vested rights theory while thinking in terms of conflicts between
rights created by different laws, we see that they are always re-
solved on the basis of the territorial principle. The right created by
the law where the last necessary act took place prevails. This reso-
lution is arbitrary, in that territorialism does not capture the rele-
vant policy concerns, but it is coherent.85 Interest analysis, by
contrast, denies the conflict in a way that produces incoherencies.86
Currie's prescriptions for conflicts remain plausible only so long as
the conflict is hidden behind the veil of choice, so long as conflicts
are conceived of as giving rise to a choice-of-law question that can
be resolved by personal-jurisdiction-style analysis.87
85. It is coherent in that a conventional legal principle (territorialism) specifies which

right prevails.
86. This is, I realize, a bold claim, and I do not attempt a justification at this point. What
I will show is that interest analysis does not really choose an applicable law, as it claims. See
infra section IV.C. If we examine interest analysis through the lens of conflicts, what
emerges is not a conventional legal rationale for choosing applicable law but simple discrimi-
nation against foreign law and foreign litigants.
87. A more sophisticated version of interest analysis has been developed by Larry
Kramer. See, e.g., Kramer, Rethinking Choice of Law, supra note 81. Because his approach
is in many ways similar to the one I advocate, I will postpone consideration of his work,
noting here only that it escapes many of the faults with which I charge Currie's approach.
D. Current Theory
The only accurate generalization one can make about current

conflicts theory is that consensus is lacking.88 Interest analysis is
the leading scholarly position, and the only doctrine that could
plausibly claim to have generated a school of adherents. It has
been fiercely attacked, however, and the most thoughtful attempts
to develop its insights have been condemned as heresy.89 Other
theories of choice of law, though less popular than interest analysis,
have also been articulated. These approaches, which are not with-
out appeal, urge courts to apply the law favoring the plaintiff,90 the
"better" law,91 or the law whose policies would be more impaired
by rejection.92 Into this chaos came the Second Restatement, syn-
thesizing a wide range of insights into an indigestible stew.93 For
torts, the Second Restatement urges application of the law of the
state with "the most significant relationship" to the action;94 it then
lists a dizzying number of factors with no hint as to their relative
weight.95
More recently, a counterrevolution of sorts appears to be
emerging, marked by the insistence that the concept of rights
should have a greater role to play.96 Finally, there exists also a sub-
stantial body of scholarship insisting that, Supreme Court pro-
nouncements to the contrary notwithstanding, the Constitution has
88. This has been true for a while; more than sixty years ago David Cavers commented
that "the article on a conflict of laws topic which does not deplore a current 'confusion of
authority' is still a rarity." Cavers, supra note 56, at 177.
89. See Louise Weinberg, Against Comity, 80 GEO. L.J. 53 (1991) (criticizing Kramer).
90. See, e.g., RUSSELL J. WEINTRAUB, COMMENTARY ON THE CONFLICT OF LAWS 360 (3d
ed. 1986).
91. See, e.g., ROBERT A. LEFLAR ET AL., AMERICAN CONFLICTS LAW ? 107 (4th ed. 1986).
92. See, e.g., William F. Baxter, Choice of Law and the Federal System, 16 STAN. L. REV. 1
(1963).
93. See, e.g., Kramer, Rethinking Choice of Law, supra note 81, at 321-22 n.149 ("no ex-
planatory power"); Laycock, supra note 73, at 253 ("Trying to be all things to all people, [the
Second Restatement] produced mush."); Joseph William Singer, Real Conflicts, 69 B.U. L.
REV. 1, 77 (1989) ("mystifies rather than clarifies").
94. See RESTATEMENT (SECOND) OF CONFLICT OF LAWS ? 145 (1971).
95. See id. at supra note 94, ?? 6, 145. The Second Restatement may not be as worthless as
it seems. It does at least identify relevant considerations. If states simply used these factors,
sincerely and consistently, to create rules about which law prevailed, we would have a regime
quite like the one I will claim the Constitution imposes. See infra section VI.C.
96. See, e.g., Brilmayer, supra note 71; Dane, supra note 48. It is odd that these scholars
seem to think that interest analysis necessarily opposes the idea of rights. See Kramer, Re-
thinking Choice of Law, supra note 81, at 278. Brilmayer's rights lead to a personal-
jurisdiction-style analysis. See Brilmayer, supra note 71, at 1279. Consequently, I do not
endorse her account.
something to say about choice of law.97 I think it surely does, and

much of this article will be spent showing how two constitutional
clauses, taken seriously, can dramatically change the face of con-
flicts theory.
E. Methods and Objectives: What Interest Analysis Is, What a

Conflicts Theory Should Be
The preceding sections have labeled the objects of their discus-

sion "conflicts theories," but this is not quite accurate. It is some-
what unfair to Currie, and somewhat generous to Beale.
Explaining why this is so requires a look at conflicts analysis from a
more structural perspective. The basic question in a conflicts case, I
have said, is whether the plaintiff has an enforceable right. Later
sections will make the argument for this perspective.98 My aim here
is to give a more fully developed theoretical account.
A plaintiff's claim may fail for two reasons. It may be that the
plaintiff has no right - he might have pleaded a cause of action
that simply does not exist, or failed to allege a necessary element.
But a claim might also fail because the defendant has available a
defense that defeats the plaintiff's right.99 This might be an affirma-
tive defense, such as mutual mistake in a contracts case or consent
to a tort, or it might be something like official immunity.
Deciding whether the plaintiff has an enforceable right thus re-
quires two quite different inquiries. The first is the determination
whether the plaintiff has a right at all. This is a question of the
scope of the right the plaintiff invokes - whether the law he ap-
peals to grants rights to people in his situation. The rules consulted
for this purpose are what I have called "rules of scope," following
Perry Dane. If the plaintiff does have a right, the court must then
perform a similar analysis of scope to determine whether the de-
fendant has a contrary right. Only if the scope inquiry results in the
conclusion that both parties have invoked appropriate rights is
there a conflicts question. In such a case, the court must look to a
97. See, e.g., Ely, supra note 73; Katzenbach, supra note 11, at 1093 ("Among the United
States these are problems ultimately subject to Constitutional prescriptions."); Laycock,
supra note 73.
98. See infra Part IV.
99. Those troubled by the possibility of unenforceable rights might wish to alter the ter-
minology here, perhaps distinguishing "prima facie" rights, which can be defeated, from
"true" rights, which permit recovery. Cf Kramer, Rethinking Choice of Law, supra note 81,
at 293-304. I do not think such semantic reticulation is necessary; where appropriate I will
characterize some rights as "enforceable" without worrying about the implication that some
are not.
different kind of rule to resolve the conflict - to what I will call a

"conflicts rule," which specifies which right shall prevail.
Obviously, then, there are two ways in which a theory might
handle the issues raised by multistate cases. It might eliminate con-
flicts by aggressive use of rules of scope, or it might provide con-
flicts rules to resolve conflicts. Beale's theory is of the first sort.
The reason that it is generous to characterize it as a conflicts theory
is that Beale's approach has no conflicts rules at all. The territorial
principle is a rule of scope - state laws grant rights only with re-
spect to in-state occurrences - and it eliminates conflicts entirely.
Currie's theory is similar, but less extreme. In place of Beale's
territorial rules of scope, Currie uses the concept of governmental
interest to test whether rights exist.100 Currie's rules of scope are
not quite as powerful as Beale's, and consequently some conflicts
do arise with his approach. But Currie has very little in the way of
conflicts rules. He has, in fact, only one such rule - forum law
always prevails. This is, as he readily admits, more a faute de mieux
stopgap than a real attempt to create a conflicts rule.'10
Currie had the misfortune to come up with a very troubling
stopgap, and the obviously discriminatory character of the rule that
forum law always prevails is the source of much of the criticism of
interest analysis. But this criticism - unlike Brilmayer's attack on
the notion of governmental interest - is somewhat misdirected.
Currie conceded the jerry-built quality of his conflicts rule, and in-
deed did not claim to offer resolutions to conflicts; his aim instead
was to show that not all multistate cases featured conflicts. He be-
lieved, in fact, that true conflicts could not be resolved by the body
of law called "conflicts."102 What he hoped for was federal legisla-
tion under the Full Faith and Credit Clause, directed to particular
areas of substantive law.103 It is true, I think, that the preference
100. This is not quite true to Currie's understanding of his theory. If interest analysis
were truly a scope analysis, then unprovided-for cases (where neither state has an interest)
would be cases in which the plaintiff simply had no right. Currie remained blinded by the
idea that the task in a multistate case is to find which law governs, not which right prevails.
He was unwilling to conclude that no law governs, and thus found it necessary to apply forum
law in unprovided-for cases. Larry Kramer has redescribed unprovided-for cases from what I
think is the correct perspective, making the point that a lack of interest implies simply a lack
of rights. See Kramer, Myth, supra note 81, at 1064. He maintains (and I agree) that this is
consistent with Currie's approach.
101. See, e.g., CURRIE, supra note 76, at 169 ("[The resort to forum law] is not an ideal; it
is simply the best that is available.").
102. See CURRIE, Married Women's Contracts: A Study in Conflict-of-Laws Method, in
SELECTED ESSAYS, supra note 14, at 77, 107, 117-21; CURRIE, supra note 22, at 181-83.
103. See, e.g., CURRIE, supra note 22, at 183; CURRIE, supra note 76, at 169-70. It is thus
not quite true that Currie maintained that "[a]ll choice of law decisions should be simply
for forum law is unacceptable and must be rejected: Currie's con-

flicts rule is simply untenable. But the conflicts rule is not an essen-
tial conceptual part of interest analysis. Interest analysis is
fundamentally a scope-based theory, and its conflicts rule can be
discarded without compromising the methodology.
Having made the distinction between rules of scope and con-
flicts rules, we can describe the progress from Beale to Currie quite
simply. Beale eliminated conflicts by territorial rules of scope. Im-
puting these rules to the states would have been implausible, and he
stuck them instead in the twinkling heavens as part of the nature of
law. The realists, rightly, did not believe that the "nature of law"
was much of a constraint on what states did. Without an effective
metaphysical imprimatur, Beale's rules of scope looked silly and
wrongheaded, and Currie set out to show that different rules of
scope were more sensible. The theories are first cousins: each re-
lies on rules of scope to do all of its work. Neither Beale nor Currie
said anything useful about conflicts.104
Because of their close relation, the two theories face common
difficulties.105 The problem with having rules of scope do all the
work is that the scope of state-created rights is first and foremost a
question of legislative intent (or judicial intent, with regard to com-
mon law rights). But if the intent of state officials is the whole
story, then a scope-centered theory is normatively toothless: judi-
cial or legislative statements about the scope of state rights are au-
thoritative, and while the interest analyst may find them misguided,
he cannot claim that they are wrong. To have prescriptive force,
rules of scope must draw on some source external and superior to
the authority of the states.106 Beale relied on the nature of law, but
that gambit is no longer plausible. Currie purported to uncover in-
terests via the conventional process of statutory interpretation, but
this required him to defer to legislative or judicial statements of
substantive decisions about the substantive rules' proper reach." BRILMAYER, supra note 8,
at 108. First, Currie did distinguish between scope analysis and conflicts analysis even within
his theory - though admittedly, since an interested forum will always apply its own law, the
scope analysis ends the matter if it uncovers a forum interest. Second, and more importantly,
Currie realized - and hoped - that federal conflicts rules would displace his rule of prefer-
ence for forum law. This latter point shows quite clearly his awareness of the distinction
between scope decisions and conflicts decisions.
104. Please do not quote this out of context.
105. Brilmayer makes this point well in her extended critique of interest analysis. See
BRILMAYER, supra note 8, ? 2.5. My discussion here is indebted to hers on several points,
most notably the tension between objectivity and subjectivity in the nature of governmental
interests.
106. See id.
scope even if those conflicted with his standard domiciliary-focused

approach to interests. Metaphysics and impotence are the Scylla
and Charybdis of scope-centered theories. Beale's succumbs to the
first while Currie's wavers between the two.107
Modern theory has advanced, of course, and we now have a
wealth of suggestions for conflicts rules.108 Many of these are good;
a number are quite ingenious.109 It is fortunate that scholars have
turned their attention to conflicts rules. One of the arguments this
article will make is that conflicts are more prevalent than Currie's
rules of scope suggest.110 In consequence, conflicts rules are where
the action is.1l
Unfortunately, modern suggestions for conflicts rules share
something with scope-based theories: they are normatively weak.
The rules they offer, if followed by all states, would probably make
all better off; coordination can often be mutually beneficial in mul-
tiparty interactions. The problem lies in achieving coordination in
the absence of external constraints.112 As Larry Kramer has
pointed out, conflicts presents a sort of Prisoner's Dilemma: states
may do better by cooperating, but defection is a danger.113 The
existing suggestions for conflicts rules lack prescriptive force in that,
if states decide instead to pursue narrow or selfish interests, the
scholarship is merely hortatory.
This article will not attempt to prescribe particular conflicts
rules, but will instead show how the Constitution constrains the
states in their crafting of such rules. (It will not construct a building
so much as give a perspicuous view of the foundations of possible
buildings.) Beale and Currie both thought that conflicts were too
107. Currie's waffling is reflected in the alternately objective and subjective nature of
governmental interests. Brilmayer's extended discussion of interest analysis reveals this well,
although I think she goes too far in faulting Currie for not treating state choice-of-law rules
as rules of scope. See infra section III.B.2.
108. See supra section II.D.
109. Baxter's comparative impairment principle, in particular, has the elegance and good
sense that typically prompts scholars to think that we'd have come up with that idea if we'd
thought about it first.
110. See infra section VI.B.
111. More generally, the conflicts problem is that state assertions of legislative jurisdic-
tion overlap. Telling states that their rights do not extend as far as the legislature has said
they do is pointless. In fact, I will argue, the Constitution tends to enlarge, rather than con-
tract, the scope of state-created rights. What a conflicts theory must do is manage the com-
peting claims of authority; it must oversee the conflicts between rights. This is obviously a
matter of conflicts rules, not rules of scope.
112. See generally MANCUR OLSON, THE LOGIC OF COLLECTIVE ACTION (1971).
113. See Kramer, Rethinking Choice of Law, supra note 81, at 339-44. Kramer also sug-
gests that coordination may naturally emerge, see id. at 343-44, although the history of con-
flicts should give optimists pause.
hard to resolve.114 They are hard, and the task of deciding which
interests are more important lies properly within the authority of
the states. The Constitution does not prescribe unique conflicts
rules, but rather restricts the permissible grounds on which states
may assert that their interest in regulating a transaction prevails
over the interest of another state. By so doing, the Constitution
creates a situation in which mutually beneficial coordination is
likely. The proper role of the Constitution has been obscured by
the rhetoric of choice and the concomitant personal-jurisdiction-
style analysis. If we think about the issue in terms of conflicts
which is to say, if we think about it as a conventional legal question
- things become much clearer. But to make the case for this
claim, I need to develop what I have been calling the conventional
legal perspective.
III. RETHINKING THE THEORY: FROM CHOICE TO CONFLICT
The goal of this article is to provide a workable framework fo

resolving conflicts of law, one that looks at them as conflicts a
applies principles appropriate to their resolution. Neither t
vested rights theory nor interest analysis fits the bill because bo
are scope-centered: neither makes a real attempt to deal with co
flicts.115 This sole reliance on rules of scope is both mistaken a
unnecessary. A theoretically sound approach to conflicts can
constructed, and these theories give us the raw materials to do
The framework proposed here will not be built from scratch; no
ing in conflicts is at this point. The concepts I will deploy can
identified quite easily as originating in either vested rights theory o
interest analysis. The aim of this Part is to show what each
proach has to offer, as well as what must be discarded.
A. Rights and Their Critics

Legal realism utterly destroyed Beale's carefully constructed e
ifice. The revolution was necessary; the vested rights theory was
wrong as a legal theory can be. It was wrong, however, primari
because it produced the wrong results, not because of any meta
physical taint. The realist assault went beyond criticism of Beal
results, and in its more ambitious form it seriously overreache
The realists' success in dislodging Beale's rights-based framewor
114. See 3 BEALE, supra note 6, ? 53, at 1929; CURRIE, supra note 102, at 107, 117-2
CURRIE, supra note 22, at 181-83.
115. See supra section II.E.
has led conflicts to its current straits. The rejection of the notion of
rights is responsible for both the esotericism of conflicts and, relat-
edly, its disconnection from ordinary legal discourse in general and
constitutional law in particular.116 Rescuing the rights-based frame-
work requires an evaluation of the legal realist criticisms.
There are two quite different components to the realist attack
on Beale's approach. The first is practical and shows that the terri-
torially based vested rights theory does not work, either positively
or normatively. The second is theoretical and aims to eliminate the
very notion of rights from legal discourse. The following section
considers the first component, which is sound; the next addresses
the second, which is not.
1. The Failure of Vested Rights / Territorialism
From a normative perspective, the most obvious problem with

territorialism is its tendency to produce arbitrary results. Territorial
connecting factors, triggered by the crucial last act, often point in
odd directions. For example, if one resident of state A poisons an-
other state A resident within the borders of state A, common sense
political philosophy does not suggest that state B's law should gov-
ern merely because the victim happens to have crossed into state B
when the poison takes effect. In compensation for this arbitrari-
ness, territorialism is generally supposed to offer predictability.117
However, it turned out to be much less determinative than its pro-
ponents claimed. In part this was a consequence of the need to
characterize actions and their elements. Courts needed to decide
whether the suit sounded in tort or in contract before they could
invoke the appropriate rules. Similarly, since the forum would ap-
ply its own procedural law regardless of whether it applied foreign
substantive law, the characterization of particular issues as substan-
tive or procedural could be dispositive. The related distinction be-
tween rights and remedies also allowed courts some latitude
because under Beale's theory, forum law governed questions of
remedy even when the rights were foreign.118
Underdetermination may not be a critical defect; indeed, terri-
torialism's arbitrariness stemmed from its rigidity, and the elasticity
provided by these "escape hatches" gave judges freedom to reach
116. As Kramer succinctly puts it, "Our entire legal system rests on the concept of vested
rights." Larry Kramer, Return of the Renvoi, 66 N.Y.U. L. REV. 979, 990-91 (1991).
117. See 1 BEALE, supra note 6, ? 1.3; see also SCOLES & HAY, supra note 7, ? 2.6, at 15
n.10.
118. See, e.g., Oceanic Steam Navigation Co. v. Mellor, 233 U.S. 718 (1914).
sensible results. More serious is the problem that territorialism

works oddly, if at all, in the absence of substantive legal uniformity.
The elements of a tort may differ across jurisdictions; similarly, ju-
risdictions may employ different rules to determine where a con-
tract is formed.119 Thus jurisdictions applying territorialist rules
might still disagree about what the essential last act was, and conse-
quently about where it took place. Beale attempted to smooth over
these issues by appealing to general common law to determine the
location of the triggering events, but general common law is no
longer available. Since it is not even clear what results territorial-
ism prescribes, it is hard to maintain that it reaches the right ones as
a normative matter.
Of course, Beale proposed the vested rights theory not as a nor-

mative suggestion but as a positive statement of the law. From this
perspective, the greatest defect of territorialism is that it is not true.
States regularly assert the power to determine the legal conse-
quences of events transpiring outside their geographical bounda-
ries, and sometimes they succeed.120 So too does the federal
government.121 In the face of this widespread disregard, territorial-
ism can be defended as a positive theory only by metaphysical argu-
ments about the nature of law, suggesting that actual practice is
illegitimate, somehow "not law" despite the fact that everyone does
it. But this style of argumentation is no longer convincing, nor
should it be.122 Law is a human practice, not an independent entity
119. For example, one jurisdiction might have adopted the "mailbox rule," providing that
the contract is formed as soon as the recipient of the offer sends acceptance; another might
hold that the contract is formed only upon receipt of the acceptance. See BRILMAYER, supra
note 8, ? 1.3, at 26, ? 1.5.2, at 40-41.
120. See, e.g., Allstate Ins. Co. v. Hague, 449 U.S. 302 (1981) (upholding application of
Minnesota law to accident occurring in Wisconsin); Skiriotes v. Florida 313 U.S. 69 (1941)
(upholding Florida prosecution of Florida resident for actions on high seas); Lea Brilmayer
& Charles Norchi, Federal Extraterritoriality and Fifth Amendment Due Process, 105 HARV.
L. REV. 1217, 1241-42 (1992) ("It is also fairly well established that a state may regulate its
residents, even when they are acting outside the state.").
121. See, e.g., United States v. Alvarez-Machain, 112 S. Ct. 2188 (1992) (upholding kid-
napping of Mexican national to be tried in United States for conspiracy to torture United
States government agents in Mexico). See generally Brilmayer & Norchi, supra note 120, at
1229 (discussing rules on federal extraterritorial criminal prosecution). Even the operation
of federal law within the states would seem to pose problems for Beale's theory of a single,
territorially supreme sovereign. Beale admits no difficulty: he suggests that each state of the
union remains a "separate legal unit," 1 BEALE, supra note 6, ? 2.2 (discussing annexation of
Hawaii), and then explains that federal law is local law everywhere, see id. ? 2.3, at 18 ("It is
perfectly correct to say .. . that the law of each of the states consists of the constitution,
treaties, and statutes of the United States, the constitution and statutes of the particular state,
and the common law of that state."). This analysis allows his theory to operate but, charac-
teristically, suppresses the possibility of conflict between state and federal law.
122. In fact, Beale attempted to defend territorialism as a matter of positive law. "Since
the power of a state is supreme within its own territory, no other state can exercise power
to which practice must conform.123 Consequently, it cannot credi-

bly be attacked for failure to abide by metaphysical principles.
Criticisms of violations of territoriality must be made within the
law, but Beale has no tools with which to make those criticisms.
This should be enough to condemn Beale's version of the vested
rights theory. Territorialism is neither normatively attractive nor
positively accurate. His rules of scope are simply wrong. What we
are left with, then, is a machine missing a vital gear. Without the
territorial principle, it does not run.
2. Salvaging the Concept of Rights
Beale's theory aims to help judges ascertain parties' rights. In

criticizing Beale's results, legal realism left this aspiration undis-
turbed; it merely pointed out that territorialism did a bad job. The
more ambitious aspect of the realist challenge consisted of the re-
jection of the goal itself, the denial of the concept of rights. This
broader attack on received wisdom was part of a widespread reac-
tion against formalism and metaphysics.124 In a classic statement of
the principles of realist jurisprudence, Felix Cohen affiliated himself
with a laundry list of like-minded philosophers - Charles Pierce,
William James, Bertrand Russell, and Rudolf Carnap, to name a
few - and linked the realist movement to similar developments in
physics, mathematics, psychology, and even grammar.125 The com-
mon thread binding these thinkers, what Cohen called the "func-
tional approach," was "an assault upon all dogmas and devices that
cannot be translated into terms of actual experience."126 Cohen
thus demanded, with the pragmatism of William James, that con-
cepts pay their way,127 and asserted, with the verificationism of the
logical positivists, that "[a]ll concepts that cannot be defined in
there," he wrote. 1 BEALE, supra note 6, ? 61.1. He then turned for support to Chief Justice
Marshall's opinion in Rose v. Himely, 8 U.S. (4 Cranch) 241 (1808), but the quotation in fact
suggests that personal and territorial traditions mingle: "It is conceded that the legislation of
every country is territorial; that beyond its own territory, it can only affect its own subjects or
citizens." 8 U.S. (4 Cranch) at 279 (emphasis added). Law has never been purely territorial
in practice.
123. Cf PHILIP BOBBITT, CONSTITUTIONAL INTERPRETATION 24 (1991) ("Law is some-
thing we do, not something we have as a consequence of something we do.").
124. See LAURA KALMAN, LEGAL REALISM AT YALE, 1927-1960, at 14-15 (1986).
125. See Felix S. Cohen, Transcendental Nonsense and the Functional Approach, 35
COLUM. L. REV. 809, 822, 826 (1935).
126. Id. at 822.
127. See WILLIAM JAMES, What Pragmatism Means, in PRAGMATISM AND THE MEANIN
OF TRUTH, 27, 31-32 (1978) ("You must bring out of each word its practical cash-value .
terms of the elements of actual experience are meaningless."128

Armed with these principles, the realists went after the notion of
rights, arguing, for example, that assertions of rights were no more
than predictions of official behavior.129
The attack has been understood in two different ways. First, it
can be seen as a denial that rights exist.130 This argument is deeply
out of tune with the philosophies Cohen invoked, for a denial of the
existence of rights is just as metaphysical as the affirmation it op-
poses.131 It is just as much an attempt to catalog the furniture of the
universe, to provide a description that is not merely useful for par-
ticular purposes but, in the most robust sense of the word, true.
Claiming that rights do not exist independent of their enforcement
(or, equivalently, that law is "made" by judges, rather than
"found") merely embroils law in the sort of ontological quarrel that
has troubled philosophy for centuries - in philosophy of mind, be-
tween behaviorists and mentalists; in philosophy of science, be-
tween realists and antirealists; in epistemology, between realists and
relativists.132
When philosophical disputes have gone on for so long, with so

little in the way of resolution, it is a good bet that there is some-
thing wrong with the terms of the debate.133 Rudolf Carnap's diag-
nosis suggests the problem is that questions at this level of
generality are not ontological at all, but rather practical.134 That is,
within Beale's framework it makes sense to ask whether a particu-
lar party has a right. Asking whether rights exist at all, on the other
128. Cohen, supra note 125, at 826. The logical positivists maintained that the meaning of
a sentence consists of its method of verification. See, e.g., Moritz Schlick, Positivism and
Realism, in LOGICAL POSITIVISM 86-88, 106-07 (A.J. Ayer ed. & David Rynin trans., 1959). It
follows immediately that an unverifiable proposition - one with no empirical consequences
- is meaningless.
129. See, e.g., COOK, supra note 57, at 33; Oliver Wendell Holmes, The Path of the Law,
10 HARV. L. REV. 457, 461 (1897).
130. See, e.g., BRILMAYER, supra note 8, ? 1.5.2, at 37 ("[The realists] believed, in addi-
tion, that there simply were no such things as vested rights."); Dane, supra note 48, at 1225
(commenting that realists "have spent a good deal of ink denying the metaphysical reality of
legal norms or rights 'existing' independent of their enforcement").
131. The logical positivists, at least, were clear that they were not denying the existence of
anything but simply abandoning meaningless discourse. See, e.g., Schlick, supra note 128, at
106-07.
132. Law has, of course, endured this debate, though current theory wisely tends to char-
acterize it as a waste of time. See, e.g., DWORKIN, supra note 48, at 225; Richard H. Fallon, Jr.
& Daniel J. Meltzer, New Law, Non-Retroactivity, and Constitutional Remedies, 104 HARV.
L. REV. 1731, 1764 (1991).
133. See P.M.S. HACKER, WITTGENSTEIN'S PLACE IN TWENTIETH-CENTURY ANALYTIC
PHILOSOPHY 100-03 (1996).
134. See Rudolf Carnap, Empiricism, Semantics, and Ontology, in MEANING AND NECES-
SITY: A STUDY IN SEMANTICS AND MODAL LOGIC 207, 207-08 (1956).
hand, is either trivial or nonsensical within the framework.135 An-

swering that broader question, in a metaphysical sense, requires a
true and singular description of the world, an ideal language among
whose terms the disputed entities will or will not be found. How-
ever, we do not have such a language; what we have are different
sets of linguistic practices adapted to different purposes and circum-
stances. We are dealing not with entities but with ways of talk-
ing.136 Consequently, the decision whether to talk in terms of rights
must be made on practical grounds.137
The better understanding of the realist attack thus takes the ulti-
mate question to be not whether rights exist, but rather whether
they are theoretically useful. From this perspective, the realist
claim is that talking in terms of rights does not advance the ball.
That claim is wrong, and demonstrating its error is the work of this
article.
I obviously cannot argue here for the proposition that thinking

in terms of rights is useful. I can only attempt to show it, and that is
the task of later sections. I can say, however, that realism's attempt
to bring scientific methods to bear on conflicts did not succeed in
producing clarity. What it did succeed in doing was to cut conflicts
loose from the remainder of legal discourse, which pervasively em-
ploys the concept of rights. That should be a prima facie reason to
doubt the realist contribution.138
135. Cf J.L. Austin, The Meaning of a Word, in PHILOSOPHICAL PAPERS 55, 57-58 (3d ed.
1979).
136. Cf. LUDWIG WITTGENSTEIN, PHILOSOPHICAL INVESTIGATIONS ?? 400-01 (1958).
137. Cohen certainly seemed to understand this point. See Cohen, supra note 125, at 835
("A definition of law is useful or useless. It is not true or false ... .").
138. Another reason is the fate of the related movements to which Cohen pointed.
Cook's realist approach to conflicts linked itself quite closely to the logical positivists; he
identified himself explicitly as a "scientific empiricist," COOK, supra note 57, at 46, and took
as his epigraph for Chapter III a restatement of their central principle of verification, attrib-
uting it to the Marxist and pragmatist philosopher Sydney Hook. See id. at 71. Logical posi-
tivism was a dramatic failure; when A.J. Ayer was asked for the chief difficulty in a television
interview, he is reported to have responded, "I suppose its main defect was that it wasn't
true." See Shusha Guppy, Tom Stoppard: The Art of Theater VII, reprinted in TOM STOP-
PARD IN CONVERSATION 177, 187 (Paul Delaney ed., 1994); see also id. (describing criticism
of logical positivism as "attacking a dodo"). For a concise philosophical evaluation of logical
positivism, reaching essentially the same conclusion as Ayer, see HACKER, supra note 133, at
64-65. Both positivists and realists, I suggest, erred by embracing a dogmatic reductivism
rather than a sensitive analysis of the actual use of language. Realism's value lies in its skep-
tical contributions, not in any attempts to create a general theory of law, and it may be better
understood as consisting simply of the former. See JEFFRIE G. MURPHY & JULES L.
COLEMAN, PHILOSOPHY OF LAW 35 (1990).
B. Governmental Interests and Their Critics
What I want to draw from Beale's work, then, is the principle

that the basic task in conflicts is to determine whether the plaintif
has an enforceable right. As discussed above,139 this determination
requires a two-step inquiry. First, the court must analyze the scope
of the laws invoked by plaintiff and defendant to determine
whether a conflict exists. If it finds a conflict, it must employ con
flicts rules to resolve it. The question at this point is how to arrive
at the appropriate scope and conflicts rules. For that enterprise I
will enlist parts of interest analysis, and before coopting the theory
I must evaluate it.
The realists attacked Beale's approach on two levels, faulting

both its results and its theoretical orientation. Interest analysis has
been subjected to the same dual challenge. At the level of result,
the charge is that Currie erred in his creation of rules of scope and
conflicts rules. At the level of theory, put forth most forcefully by
Lea Brilmayer,140 the realist argument is that the governmental in-
terests Currie purported to identify are not part of a realist or func-
tionalist analysis, but rather a metaphysical construct akin to
Beale's vested rights, imported for a similar deus ex machina solu-
tion to conflicts questions. Once again, I will consider the specific
criticisms before moving on to the general ones.
1. The Weakness of Currie's Approach

As a normative matter, Currie's rules are not very attractive.
His examples are complex, but the tendency that emerges is un-
abashedly parochial. Generally speaking, as a matter of scope,
state law grants rights only when to do so favors a local; and when
rights conflict, forum law always wins.141 Neither of these
domiciliary-centered rules seems a good recommendation.
Like Beale, however, Currie advanced his scope analysis as a
positive statement of the law. It is, he claimed, simply interpreta-
tion of the substantive laws at issue.142 The utility of the concept of
governmental interests depends on this claim, and I consider
Brilmayer's challenge to it in the next section. The practical weak-
ness of Currie's approach is more obvious, and it has to do with his
conflicts rule.
139. See supra section II.E.

140. See sources cited supra note 78.
141. See, e.g., BRILMAYER, supra note 8, ? 2.1.2, at 65.
142. See, e.g., CURRIE, supra note 102, at 118.
Currie did not purport to derive his preference for forum law in
true conflicts from analysis of state law. In fact, he advanced it with
some embarrassment, as "not an ideal" but "simply the best avail-
able."143 It is not an attempt to resolve conflicts at all, and it pro-
duces a theory just as impotent as Beale's without the territorial
principle. Vested rights analysis worked only so long as its rules of
scope suppressed the possibility of conflict; without these rules,
conflict appears and cannot be resolved. Interest analysis, while no
longer working explicitly in terms of rights, confronted the same
problem. Currie's rules of scope suggested that some multistate
cases did not present conflicts. Because his rules of scope were
more plausible than Beale's, the approach had some practical value
- identifying false conflicts is generally considered a real contribu-
tion. But it did not even pretend to solve the fundamental problem.
In true conflicts, Currie suggested that courts should apply forum
law, not because it was appropriate according to conflicts principles,
but because such conflicts were at bottom insoluble.144
This is not much of a conflicts theory. If we grant that state

interests are entities discernible by the methods Currie advocates
the theory shows that some cases do not present the basic conflicts
issue: a clash between two sovereigns, each of which demands that
its law be given effect. Where such conflicts do exist, however, in-
terest analysis offers no solution.145
The only reason that this confession of weakness seems anything
less than a confession of failure is that interest analysis employs
what I have called a personal-jurisdiction-style analysis. This ap
proach conceives the task as choosing applicable law, not resolving
conflicts, and it suggests that choosing forum law does not imply
that the interests of other states have been deemed inferior.146 If
choosing forum law is acceptable in true conflicts - Currie calls it
"sensible and clearly constitutional"'47 - then true conflicts do not
create a gaping hole in the heart of the theory. It is for this reason,
I think, that interest analysis is forced to characterize away the con-
flict, to adopt the personal-jurisdiction-style choice-of-law ap-
proach. If the conventional legal perspective reveals that the

144. See CURRIE, supra note 102, at 107, 117.
145. This is not precisely true; after all, Currie suggested that in true conflicts, courts
should apply forum law. But he did not see this as a resolution. See id. at 117-21. Nor should
he have; it is obviously discriminatory and, I will argue, unconstitutional.
146. See CURRIE, supra note 22, at 181-82.
147. CURRIE, supra note 102, at 119.
interest analysis approach to true conflicts is illegitimate - and I

will show that it does, though the demonstration is still a ways off -
the recharacterization is unavailing. In that case, interest analysis
fails because it cannot handle conflicts.
2. Salvaging the Concept of Interest
Just as Beale's rights-based framework can survive without his

territorial rule of scope, Currie's concept of governmental interest
does not depend on his conflicts rule. Nor, of course, does it de-
pend on the precise rules of scope he derived. Currie's admission
that his analyses were only tentative and "subject to modification
on the advice of those who know better"148 - namely state courts
and legislatures - shows that the theory can accommodate a wide
variety of such rules.
Brilmayer's more ambitious attack on the notion of governmen-
tal interest begins with Currie's scope analysis. She rejects the idea
that scope can be determined simply by analysis of the substantive
law and faults Currie for ignoring state choice-of-law rules in his
determination of governmental interests.149 If correct, the criticism
has serious implications; it shows that the key concept of interest is
not something derived from state law but "an externally deter-
mined and objective concept that is imposed on state legislatures
and state judges by scholars."'50 The criticism is not correct, how-
ever: it runs together scope analysis and conflicts analysis. Choice-
of-law rules are not rules of scope, and Currie was right not to defer
to them.
A functional analysis leads to this conclusion, for choice-of-law

rules simply cannot do the work of rules of scope. To the extent
that they might seem to grant or deny rights, they are trumped by
substantive law. First, choice-of-law rules will never affirmatively
produce a right denied by substantive law. If a Connecticut statute
grants rights explicitly only to those injured within the state, a
choice-of-law rule dictating the application of Connecticut law to an
extraterritorial injury will not expand the statute's scope. Second,
choice-of-law rules do not withhold rights affirmatively granted by
substantive law. A state choice-of-law rule codifying the territorial
principle that the law of the place of the tort governs might seem to
indicate a lack of intent to grant rights to a domiciliary injured
148. CURRIE, The Verdict of Quiescent Years, in SELECTED ESSAYS, supra note 14, at 592.
149. See BRILMAYER, supra note 8, ?? 2.5.1-.4.
150. Id. ? 2.5.5, at 110.
outside its borders, but this is not in fact so. Even the territorialists
granted that if a tort occurred in a place with no local law, at least
between two domiciliaries of the same state, that state's law would
determine rights and obligations.151
Instead, choice-of-law rules typically prescribe which rights will
prevail when rights conflict.152 (It is because there are no conflict-
ing rights that the law of common domicile applies in tort cases oc-
curring in lawless lands, despite territorial choice-of-law rules.)
They are, generally speaking, conflicts rules. The confusion over
their nature results presumably from the fact that choice-of-law
rules are drafted to answer the question "what law applies?" This
question is part of the rhetoric of choice, and as I have suggested
before, it prevents us from seeing clearly the structure of conflicts
analysis.153 The question "what law applies?" runs together the is-
sues of whether a party has a right, and whether that right prevails
against a conflicting right. It suggests that a choice-of-law rule is
relevant to both. Thus Larry Kramer, maintaining that "choice of
law is a process of interpreting laws to determine their applicability
on the facts of a particular case," claims that consequently "the fo-
rum can never ignore other states' choice-of-law systems."154 But
if, as I have argued, choice-of-law rules are conflicts rules, not rules
of scope, the conclusion does not follow. A conflicts rule that local
rights will yield to foreign rights on a particular constellation of
facts does not mean that the local rights do not exist.
Of course, the conclusion that choice-of-law rules are conflicts
rules does not salvage Currie's positive analysis; the question re-
mains why he did not defer to them instead of concluding that fo-
rum rights should always prevail over foreign rights. It might seem,
however, to rescue his conception of governmental interests as
things that an analyst can discern by analysis of substantive law. In
fact, there is another element of the challenge that must be faced.
151. See, e.g., American Banana Co. v. United Fruit Co., 213 U.S. 347, 355-56 (1909)
(Holmes, J.) ("No doubt in regions subject to no sovereign ... [civilized nations] may treat
some relations between their citizens as governed by their own law, and keep to some extent
the old notion of personal sovereignty alive."); 1 BEALE, supra note 6, ? 45.2 (discussing
jurisdiction over actions arising on the high seas).
152. A rule looking to another state's law may also reflect an intention to give the same
rights as that state's substantive law would in a domestic case. This is the most natural inter-
pretation of a rule that, for example, provides that the rights of heirs shall be determined by
the law of the testator's domicile. Cf. In re Annesley, 1 Ch. 692 (1926) (Eng.). This is what
Perry Dane calls a "rule of assimilation": it incorporates the terms of foreign law in much the
same way that federal copyright law looks to state law to determine who are the "children"
entitled to share in renewal rights. See De Sylva v. Ballentine, 351 U.S. 570, 580 (1956).
153. See supra Part I.
154. See Kramer, supra note 116, at 1005.
Brilmayer's critique of Currie devotes much space to the claim

that his "governmental interests" are objective, rather than subjec-
tive - that is, that their existence is determined by a priori theo-
rizing, rather than analysis of state law.155 I have suggested that
they are subjective, and Currie, to the extent that he addressed the
issue, said so explicitly.156 Brilmayer's critique of the concept of
governmental interests, however, is actually disjunctive. Like the
realist attack on rights, it maintains not that these interests do not
exist but that the concept is not useful, regardless of the nature of
interests.157 If interests are objective, they are metaphysical fictions
that judges may safely ignore. If they are subjective, then they are
determined, in the absence of legislative specification, by the
judge's own decision. Interest analysts may seek to free judges
from territorialist dogma by showing them a wider range of options,
but that is the extent of their contribution. If the judge decides
upon reflection that state law is nonetheless territorial in scope, the
interest analyst can only carp from the sidelines.158
The horns of the dilemma converge on the same point: the con-
cept of governmental interests cannot direct the resolution of cases.
Interest analysts have no footing for "normative critique of existing
case law."'59 This is true - but it is also true of most current con-
flicts scholarship. Suggestions for conflicts rules or rules of scope
are always mere suggestions; states may decide otherwise. Still, this
hardly shows that the concept is useless; proposals of better law can
be invaluable. My analysis actually aspires to slightly greater nor-
mative bite: it seeks to show the constitutional limits on state con-
flicts practices. From this perspective, Currie's approach has great
utility. The methodology of interest analysis is useful because it
foregrounds the question of what states are attempting to do.
Currie's particular suggestions are especially useful precisely be-
cause they are so parochial. Currie posits a maximally selfish state,
interested only in advantaging its domiciliaries. Starting from this
position and investigating the extent to which the state can achieve
its selfish ends produces a "bad state" view of conflicts, which is
what the field needs at this point. If states cannot be compelled to
155. See, e.g., BRILMAYER, supra note 8, ? 2.5-2.6, at 99-119.

157. Brilmayer actually concludes that Currie's interests are objective and hence denies
their existence. See BRILMAYER, supra note 8, ? 2.5.5, at 115. She does, however, give a
quite complete statement of the weaknesses of subjective interests as well. See id. ? 2.5.3.
158. See id. ? 2.6, at 117-18.
159. See id. ? 2.5.3, at 104.
take the high road, the important question is what barriers exist
along the downward path.
C. The Positive Account
Despite the criticisms of vested rights and interest analysis,

ceptual building blocks remain. Joseph Beale posed the key qu
tion: Does the plaintiff have an enforceable right? And Braine
Currie identified the correct way to begin the inquiry: apply
tools of statutory construction to determine whether the law
plaintiff invokes seeks to give him a right, and whether the law
defendant invokes seeks to restrain that right. This scope ana
disposes of some cases - those in which the plaintiff pleads a
that gives him no right, and those in which the defendant interpo
a law that gives him no defense. On its face, and as developed
Currie, that is as far as it goes. Larry Kramer has taken inter
analysis somewhat further,160 in what I think is essentially the rig
direction, and the positive account I develop here is quite simila
his approach.
The previous sections clarified some foundational points.
idea that rights must vest under one, and only one, law, to be iden
fied by appeal to a priori principles, is indeed misguided; but
idea that analysis must eschew the concept of rights is no bet
Conflicts scholarship has thus conformed (metaphorically
course) to Newton's laws of motion: Beale's misguided metaphy
ics are matched by an equal and opposite error in the realist r
tion. If there is progress in the scholarship, we might hope for
more gradual oscillations of the pendulum between vested rig
and realism. Such is, in essence, the aim of this article: to show
there exists a happy medium. The two approaches are not inco
patible but in fact have much to offer each other.
1. Thinking in Terms of Rights

Both vested rights and interest analysis, I have said, suppress
notion of conflict. The vested rights theory simply denies the p
bility; interest analysis hides it behind the idea of choice. U
mately, both theories effectively resolve conflicts by invo
principles - territorialism and forum-preference - that serve p
cies of conflicts jurisprudence accidentally, if at all. It would be
ter for everyone, I suggest, if we stopped hiding conflicts
started thinking about how they are resolved. I will ultimately
160. See, e.g., Kramer, Rethinking Choice of Law, supra note 81.
gue that the Constitution speaks to this point, but the first step is to
take conflicts out of the shadows in which the choice-of-law per-
spective shrouds them.161
The easy way to do this is to return to the idea of rights. This is
a venture one might hesitate to undertake, given the amount of crit-
icism the concept has absorbed,l62 but I have suggested that the
criticisms are overstated. Nor am I alone in suggesting that a
greater focus on rights would benefit conflicts theory.163 Perry
Dane, in particular, has defended at length what he calls the
"Norm-Based" view of law - essentially, commitment to the rule
of law164- and argued that it implies "vestedness" (the principle
that a party's rights should not depend on the forum).165 While I
sympathize with Dane's project, I do not intend to make a jurispru-
dential argument.166 I suggest instead that a description that oper-
161. Resolving conflicts is hard, and abandoning the personal-jurisdiction-style analysis

will force us to confront some new difficulties. Currie similarly found that his approach faced
problems that the vested rights theory did not - notably, the issue of discrimination between
citizens and noncitizens of a state. His diagnosis was one we should keep in mind:
The fact that these problems come immediately into view when conflicts problems are
approached in this way does not mean that they are generated by the method. Indeed,
their prompt appearance is ground for an inference that they have been present from the
beginning, obscured and suppressed by the traditional conflict-of-laws system.
CURRIE, Unconstitutional Discrimination in the Conflict of Laws: Privileges and Immunities,
in SELECTED ESSAYS, supra note 14, at 445, 448.
162. "Indeed, one may now wonder how any juristic construct such as 'right' could have
been accepted as fundamental in the explanation of any important aspect of judicial activity,"
wrote David Cavers. Cavers, supra note 56, at 175-76; see also Juenger, supra note 28, at 435
(vested rights theory is "sheer sophistry").
163. See generally Brilmayer, supra note 71.
164. See Dane, supra note 48, at 1218.
165. See id. at 1245. I will claim that the Constitution requires a neutrality quite similar
to Dane's vestedness. See infra Part VI.
166. In particular, I do not mean to be endorsing a "deontological" rather than a "conse-
quentialist" approach to conflicts. Dane and Brilmayer, unlike Kramer, become philosoph-
ically self-conscious when they talk of rights. See Dane, supra note 48, at 1218-23;
BRILMAYER, supra note 8, ? 5.2 (describing modern choice of law theory as instrumentalist);
Brilmayer, supra note 71, at 1278 ("There is more at stake than semantics. Choosing to talk
in terms of rights rather than policies or interests represents a fundamental jurisprudential
commitment which is reflected in the way that concrete problems are resolved."). The sug-
gestion that the language of rights implies a deontological rather than consequentialist theory
of law is, however, overstated. Michael Green, for example, has shown that a realist ap-
proach (policy analysis) is in fact compatible with deontological principles. See Green, supra
note 12, at 968-86. A more obvious objection is that the "instrumental" goal that judges seek
to maximize might be vindication of parties' preexisting rights in general. More seriously,
Brilmayer begs the question by assuming that whether application of a law will achieve its
purpose is a different question from whether its application will vindicate preexisting rights.
See BRILMAYER, supra note 8, ? 5.2, at 225; see also Dane, supra note 48, at 1243-44 (sug-
gesting that only the "Decision-Based" view of law would propose that courts, in adjudicat-
ing disputes, are "primarily charged" with advancing the policies of their states). As Larry
Kramer sensibly suggests, these are basically the same question: if allowing a party to appeal
to a particular law will not effectuate the law's purpose, it is quite likely that the law does not
give that party any rights. See Kramer, Myth, supra note 81, at 1064; Kramer, Rethinking
ates in terms of rights is more consistent with ordinary legal

discourse, and that it will prove more useful in developing a juris-
prudence of conflicts. My claim that both vested rights theory and
interest analysis hide real conflicts between laws is not meant as a
metaphysical assertion. It is not a claim that "rights exist independ-
ent of their enforcement" - whatever that means. I describe con-
flicts decisions from a conventional legal perspective because they
are legal decisions. I talk in terms of rights because that is how
lawyers and judges talk, and it is useful to talk in that way.
The description I propose is, essentially, a reworking of interest
analysis that operates in terms of rights and thereby makes explicit
the conflicts that had been hidden by the personal-jurisdiction-style
analysis of "choice of law." I will start with first principles. Thes
need not be accepted, though I hope that they will be uncontrover-
sial enough to arouse little opposition. The ultimate test of the the-
ory should be its utility; whether it seems natural is not as important
as whether it handles conflicts effectively.
Law is an instrument of social organization, designed to allow
society to function and to resolve disagreements without resort to
private violence.167 Consequently, law establishes constraints on
permissible behavior. Transgression of these limits may authorize
the government, if it can prove the proscribed conduct, to impose
civil or criminal sanctions to deter such conduct. It may also au-
thorize other private parties, upon lesser but similar proof, to win a
damages judgment in order to be made whole. This authorization
of recovery may be characterized as the creation of a right: the
Choice of Law, supra note 81, at 291-303. Indeed, this is conventional statutory interpreta-
tion, see Kramer, Rethinking Choice of Law, supra note 81, at 291-303, employing standard
reference to the intent and purpose of the drafters, see, e.g., EEOC v. Arabian Am. Oil Co.
499 U.S. 244, 248-59 (1991).
Conversely, references to parties' rights need not suggest a deontological morality at
work. The language of rights may be used simply because determinate rules (which is wha
rights talk leads to) maximize utility. That is, rights talk might just be rule-utilitarianism
Legal scholars are bafflingly blind to the existence of this approach and tend to think that any
invocation of a rule is deontological. The blindness is the more surprising because the law
and economics movement has devoted much ingenuity and more ink to showing that particu-
lar allocations of rights are or are not efficient, i.e., utility-maximizing according to the
willingness-to-pay metric. See generally RICHARD POSNER, ECONOMIC ANALYSIS OF LAW 12-
17 (1998). Brilmayer does at least mention rule-utilitarianism. See Brilmayer, supra note 71
at 1291 n.53. Perry Dane suggests that commitment to the rule of law implies a commitment
to the idea that a norm has an importance as "a goal in and of itself," and that vindicating
that norm at a cost to its underlying purpose has "a special nobility." Dane, supra note 48, at
1219. But rule-utilitarianism surely is not committed to that proposition; rule-utilitarians
may simply believe that rules are more efficient in general even if their fit is not perfect.
167. Larry Kramer starts from a very similar picture of civil society. See Kramer, Myth,
supra note 81, at 1052.
right to win a judgment upon a showing of the required elements.168

A judgment is a judicial determination that the showing has been
made. It also confers a right, namely the right to the damages
awarded. The next question is what happens to rights created by
one state when they meet the rights of another.
2. Thinking in Terms of Conflict
When governments create rights, other governments may or

may not respect those rights.169 For example, if the law of a state
does not authorize recovery upon a showing that allows a right to
recover in another state, it gives people a right to engage in the
conduct for which the plaintiff seeks damages.170 We then have a
conflict and must decide which right prevails. There is no reason to
describe this as a question of choosing which law applies, and, as we
shall see, this notion of choice does not comport with current con-
flicts methodology.171 Further, speaking in terms of conflicts may
change the way we think about these cases. A conflict is not typi-
cally resolved by a personal-jurisdiction-style analysis that identifies
a number of permissible options, but it is rather a legal question of
whether the plaintiff's or the defendant's claimed right must
168. There is no need, for my purposes, to suppose that these rights vest at the time of
any particular action. A state tort law will give certain people a right to recover damages if
they make the required showing. Whether a tort has been committed will obviously have a
bearing on whether or not the showing can be made, but it need not affect our characteriza-
tion of the right. Beale, on the other hand, did need to identify a unique moment of vesting.
The territorial principle could not operate without a method of determining in which state
the rights vest, and without the last act doctrine, Beale would have been forced to confront
conflicting rights. But my point is exactly that these conflicts exist, and that an analysis that
hides them does us no favors.
169. With judgments, they typically do; a judgment obtained in one state confers rights
enforceable in any state. See, e.g., Matsushita Elec. Indus. Co. v. Epstein, 516 U.S. 367
(1996). There is nothing analytically special about judgments; they are simply one instance of
state-created rights. The difference between rights based on judgments and those based on
legislation or common law is that there will seldom, if ever, be a right the defendant can
assert against enforcement of the judgment. States have not adopted laws giving their citi-
zens rights against foreign judgments. Indeed, Congress has specified that they may not, see
Full Faith and Credit Act, 28 U.S.C. ? 1738 (1994), and the Full Faith and Credit Clause
prevents them from doing so of its own force, see, e.g., Hughes v. Fetter, 341 U.S. 609, 611 n.4
(1951). In consequence, cases in which a party asserts a right derived from a judgment tend
not to feature conflicting rights, and a scope-based analysis will suffice.
170. See generally Wesley Newcombe Hohfeld, Some Fundamental Legal Conceptions as
Applied in Judicial Reasoning, 23 YALE L.J. 16 (1913). To say that one has a right to engage
in certain conduct means (roughly) that the conduct cannot form the basis for recovery or
sanctions; to say that one does not have a right means the conduct may. (This is only roughly
true because in some cases the exercise of constitutional rights may allow recovery of dam-
ages. For example, breaching a personal services contract is constitutionally protected in that
a state may not use its criminal law to compel performance. See Bailey v. Alabama, 219 U.S.
219, 238-44 (1911). But the breaching party will still be liable in contract.)
yield.172 What the law is that resolves this question I must leave for
later.173
I therefore suggest the following analysis. The plaintiff may

plead whatever law he desires. Courts are simply not in the busi-
ness of amending complaints sua sponte. They are in the business
of judging their sufficiency.174 The defendant might argue that the
law pleaded gives the plaintiff no rights at all - that it does not
even purport to give a right to one in the plaintiff's situation.'75
This is the claim that the action falls outside the reach of the plain-
tiff's chosen law - it is an appeal to a rule of scope.
Such would ordinarily be the case, for example, if, in a suit be-
tween two Connecticut domiciliaries over a car accident in New
York, the plaintiff claims a right under Oregon law. It is very u
likely that Oregon intends to give the plaintiff a right in this sit
tion. This is, however, just a question of interpreting Oregon la
Absent some statement to the contrary, it makes sense to presum
that a state's tort laws are intended to apply at most to torts invo
ing its citizens and to torts committed within its borders. Quite p
sibly the intended reach is narrower, though this is more difficult t
ascertain. My point here is simply that if the plaintiff invokes a
that has no application to him, the defendant can defeat the cla
on that basis.
Suppose, however, that there is a generous Oregon statute

claiming to give rights to all persons injured within the United
States. If the plaintiff invokes this statute, the defendant cannot
rely on the rule of scope argument that no right exists. He might,
however, argue that some other law - presumably the law of either
New York or Connecticut176 -gives him a defense against the
right asserted under Oregon law.
Both New York and Connecticut law satisfy the broad condi-
tions set out above for presumptive applicability of law. The de-
fendant is a Connecticut domiciliary, so Connecticut may well
172. Cf. Laycock, supra note 73, at 259.

173. I will claim that although the determination of the prevailing right lies within the
legitimate authority of the states, the Constitution sets out parameters within which state
conflicts rules must be drawn. See infra Part VI.
174. With regard to this point - that what happens in a conflicts case is that the plaintiff
files a complaint alleging violation of some right and the court assesses its sufficiency - I am
in complete agreement with Kramer. See Kramer, Rethinking Choice of Law, supra note 81,
at 282.
175. Kramer characterizes this question as whether the law gives a prima facie right, pre-
sumably to indicate that prima facie rights may not be enforceable. As discussed supra note
99, I think we may speak simply of rights. But of course nothing turns on the terminology.
176. Or possibly federal law. I consider state-federal conflicts in infra section IV.B.
desire to protect him. And the tort occurred in New York, so New
York may well intend to determine the situations under which peo-
ple acting in New York will be forced to compensate those they
injure. To determine whether the defense the defendant invokes is
applicable, we need again to take a closer look at the law.
Let us consider a few possibilities. The defendant may argue
that his driving met the standard of care established by Connecticut
law, and that this absolves him of liability. It is unlikely, however,
that Connecticut intends its domiciliaries to carry with them Con-
necticut's rules of the road. Generally speaking, rules directed to
so-called "primary conduct" - the actions forming the basis for the
lawsuit - should be presumed to have a territorial scope.
Alternatively, the law at issue might not focus on primary con-
duct. It might be a rule that has very little effect on the conduct
forming the basis for the lawsuit, such as a rule providing that tort
claims abate on the death of the tortfeasor.177 Again, both New
York and Connecticut laws meet the test of prima facie applicabil-
ity. But would New York likely intend to prevent one Connecticut
domiciliary from recovering against the estate of another, when this
restriction would have so little impact on their actions within New
York? Probably not; the defendant will have an easier time invok-
ing the abatement rule if it is a provision of Connecticut law.178
Rules that do not focus on primary conduct should generally be
presumed to be intended for domiciliaries.
Let us suppose, finally, that the defense invoked is one that the
state intends to offer to the defendant - it comes from a New York
statute setting the standard of care for its highways. If the Oregon
law at issue purports to hold the defendant to a higher standard of
care, we have what my framework sees as a true conflict. How this
177. A distinction is often drawn between "conduct-regulating" and "loss-allocating"

rules. See, e.g., Schultz v. Boy Scouts of Am., Inc., 480 N.E.2d 679 (N.Y. 1985). Currie draws
a similar line between "rules of conduct" and "rules of decision." See CURRIE, On the Dis-
placement of the Law of the Forum, in SELECTED ESSAYS, supra note 14, at 3, 68-69. I avoid
the terminology because I doubt the distinction is tenable as a general matter. Charitable
immunity, which Schultz treats as loss-allocating, will surely have some effect on how careful
charities are; immunities obviously eliminate the deterrent effects of liability. Still, rules are
directed at particular objects, though they may have broader effects. The question at this
point is simply which parties and transactions a legislature intends its law to cover. An abate-
ment rule will lift liability only from the dead, and death is generally an unimprovable deter-
rent. Thus its effect on primary conduct should be minimal, and its applicability
presumptively domiciliary-focused.
178. It may be, of course, that New York cannot withhold from the defendant defenses
made available to its own domiciliaries. See infra section VI.B.2.
conflict is to be resolved is a different question, and I claim it is a

legal one, to which conventional legal reasoning applies.179
IV. DEFENDING THE THEORY: TWO EXAMPLES
My argument is that a conflicts case presents a convent

legal question. The plaintiff asserts a right, grounded in som
the defendant asserts a contrary right, grounded typically in
ferent law. The court must decide which right prevails, and th
legal question, to be determined by state conflicts rules oper
within constitutional constraints. When a court "chooses" th
plicable law" it is really determining that one right will be r
nized and the other subordinated.180
Given my disavowal of jurisprudential intent, it might be won-
dered how much force the word "really" can muster. Generally
speaking, I think that claims about what "really" happens are argu-
ments in favor of a particular description, to be judged by the utility
of that description. If others prefer a different description, counter-
arguments may be persuasive, but they are never conclusive. Still,
there should be a presumption in favor of the conventional legal
perspective. If conflicts cases depart from this and employ a novel
vocabulary to describe quite ordinary questions, we should ask why.
Moreover, the principle that competing descriptions are essentially
equals does not always hold in law; relativism comes to an end
somewhere. The Constitution eventually takes over, in that if a de-
scription that sees a constitutional violation is sufficiently plausible,
states may not defend their conduct by offering an alternate de-
scription on which there is no violation. That, finally, is what I
claim happens here. The conflicts, or conventional legal, perspec-
tive makes the violation so clear that the Constitution forbids us
from hiding behind the rhetoric of choice of law. This is a familia
point from conflicts jurisprudence; for example, states may not
subordinate federal rights by claiming merely to have applied thei
own law.
179. This assertion may seem so obvious as to be unnecessary. The question is resolved
by the court, and courts resolve legal questions. But in making this claim I am neither flog-
ging the choir nor preaching to a dead horse. Currie's interest analysis seems to suppose that
it is not a legal question. See, e.g., CURRIE, supra note 22, at 182.
180. If the court actually decides that one law governs (to the exclusion of the other), it
has likely awarded victory to a cluster of rights. This is an unfortunate consequence of the
choice-of-law vocabulary. There is little reason to suppose that an intelligent conflicts rule
would treat bodies of law as indivisible units. It might make good sense to look to the law of
the place of a car accident for rules of the road but to another law for other purposes. In fact,
current approaches to conflicts, while claiming to select "the applicable law," do not pick a
law that governs in this sense. See infra section IV.C.
Two examples help substantiate the argument that multistate

conflicts cases raise no distinctive legal issues. Both feature circum-
stances in which a conflict between laws cannot be denied, in which
the court cannot invoke the idea of a "choice of law." Neither, con-
sequently, is typically considered as raising a "choice-of-law" is-
sue.181 This is advantageous for my purposes; it suggests precisely
that "choice of law" is an attempt to avoid conflicts rather than a
sensible method of resolving them.
A. The Unseen Conflict: Purely Domestic Cases

The canonical conflicts cases are, without exception, cases that
have contacts with more than one jurisdiction. From the perspec-
tive that takes conflicts cases to be about choice of law, this makes
sense: if there is no possibility of choosing the law of another juris-
diction, there cannot be a choice-of-law issue. But there are purely
domestic cases in which laws conflict, or seem to.182 These cases,
like the more conventional multistate conflicts cases, require courts
to determine which law prevails, which right will be vindicated. An
examination of such cases is fruitful because it tells us something
about what conflicts analysis is. It lifts the veil of choice-of-law
rhetoric and allows us to see what goes on when courts resolve con-
flicts between laws.
The first sort of domestic conflict arises from transitions be-
tween legal regimes.183 When a state enacts a new statute, creating
new rights or obligations, there will be cases in which the relevant
transactions took place before the enactment of the new statute.
Courts must then decide whether the new rights or obligations will
be recognized in such cases - whether, for example, a defendant
whose conduct met the old standard of care should be held liable
because a new statute imposes a greater duty.
181. The notable exception here is Larry Kramer, who argues that domestic cases and
multistate cases raise similar "choice-of-law" issues. See Kramer, Rethinking Choice of La
supra note 81, at 283.
182. In purely domestic cases, courts must also perform a scope analysis to determine
the law at issue grants the parties rights. See CURRIE, supra note 22, at 184. That is not very
surprising, since scope analysis is just statutory interpretation. The presence of scope analys
in domestic cases might suggest that there are some similarities between conflicts cases a
domestic cases - after all, Beale's theory had nothing more than rules of scope. It might
also suggest the correctness of the initial scope analysis in conflicts cases. More significa
for present purposes is the fact that courts actually employ conflicts rules in domestic cases
They do so rarely, because the detection of a conflict is frequently taken as an indication th
the scope analysis has gone wrong - courts presume that legislatures do not intend to crea
conflicting rights. But the conflicts rules are there if we look for them.
183. Currie also noted the presence of scope analysis in retroactivity jurisprudence. Se
id. Again, my point is slightly different: courts also perform conflicts analysis.
This is, or may be, a conflicts issue; indeed, the Supreme Court
has characterized it as a matter of intertemporal "choice of law."184
The process of decision, however - at least where the new law is
created by statute185- is no complex and murky choice-of-law
calculus. Instead, the court engages in the conventional process of
statutory interpretation to ascertain whether the new law purports
to grant rights to, or impose liability on, the parties.186 This is, of
course, the scope-based first step of interest analysis: determining
whether there is a conflict. Ordinarily, statutes operate only pro-
spectively. The rights they create may not be invoked with respect
to transactions occurring before their enactment, and so there is no
conflict between old and new law. The legislature may also specify,
however, that the new statute is to have retroactive effect. In this
case, there will be a conflict: both the old and the new statute pur-
port to grant rights. Such conflicts are easily resolved: the new law
prevails, unless the Constitution restrains it.187 That is the applica-
tion of a conflicts rule; there is no reason to describe it as a choice
of law.188
Thus it can be seen that at least some purely domestic cases in-
volve conflicts, and the analysis performed in such cases fits com-
fortably within the approach I advocate. Retroactivity cases are not
the only ones. In fact, the potential for a conflict exists in all cases.
This may seem counterintuitive. After all, in purely domestic cases,
once the plaintiff has made a claim that some law entitles him to
relief, it usually does not matter what other laws say. For example,
if the plaintiff makes out a tort claim, it does not matter that the
defendant has available an adequate defense in contract. It will do
him no good to argue that contract law applies to the case, unless
the point is that the plaintiff has actually pleaded in contract and
184. James B. Beam Distilling Co. v. Georgia, 501 U.S. 529, 535 (1991).
185. The issue of the retroactive effect of a judicial decision, rather than a new statute, is
quite murky. Oddly, the analysis now favored by the Court resembles Joseph Beale's vested
rights theory: it hides conflicts between early and later law via the premise that only the law
in effect at the time of the parties' actions can confer rights. See Kermit Roosevelt III, A
Little Theory is a Dangerous Thing: The Myth of Adjudicative Retroactivity, 31 CONN. L.
REV. 1075, 1080 (1999). This is not an encouraging sign, and I have argued that current
retroactivity jurisprudence is the unfortunate legacy of an earlier mistake. See id. at 1087-91.
Matters would be much clearer if the Court simply confronted the existence of conflict and
adopted, as a conflicts rule, the principle that later rights prevail over earlier rights, as it has
done in the statutory context.
186. See, e.g., Landgraf v. U.S.I. Film Prods., 511 U.S. 244 (1994) (engaging in compre-
hensive interpretation of the Civil Rights Act ? 102 to determine whether plaintiff, who ar-
gued for retroactive application, had a cause of action).
187. See Landgraf, 511 U.S. at 280.
188. Legislatures do, of course, choose whether their statutes apply retroactively or not,
but that is a matter of choosing the scope of the rights.
not in tort. The contract defense the state has created does not
interfere with the plaintiff's right to recover in tort; the two sta
created rights do not, so to speak, touch each other. They coexi
perfectly happily; the defense does not apply to the right to recover
nor does it purport to. This is, again, what scope analysis reveal
But such is not always the case. The state may, for examp
have created immunities. The question then is whether the imm
nity applies to the case at hand, whether the defendant's right inter
feres with the plaintiff's. It may or may not; this is a matter of sco
analysis - that is, interpreting the law. Suppose, for example, t
the plaintiff makes out a tort claim, but the defendant claims to be
state officer acting within the course of his duties. The state m
have immunized such officers, or it may have placed a limit on da
ages recoverable in such actions.189 The court must decide wheth
the state immunity may be invoked by the defendant - whether
grants him a right. If it does, the scope analysis has revealed a co
flict. The court must then determine whether the immunity defe
the plaintiff's claim: it must decide whether the immunity prev
over the plaintiff's right to recover in tort.
This is, it should be evident, analytically identical to the issu
that arises when a plaintiff relies on one state's law for his right
recover, and a defendant asserts a defense created by the law
another state. The court must decide if there is a conflict between
the rights asserted and, if so, which prevails. Again, there is no ob
vious reason to describe this as a choice between laws.190
The suggestion that purely domestic cases may involve conflict

of law, like the analysis of the preceding section, bears an obvio
similarity to some of Larry Kramer's work.191 Kramer argues th
all cases involve a choice of law.192 I think this is an important in
sight, and correct, but I would phrase it somewhat differently: n
cases involve a choice of law. Of course, this sounds rather more
like a denial than a rephrasing. Kramer's point, however, is th
cases in which courts perform an explicit "choice-of-law" calculu
189. See generally, e.g., Bott v. DeLand, 922 P.2d 732 (Utah 1996) (discussing Utah Gov
ernmental Immunity Act).
190. This does not mean that it could not be described as a choice of law - though, as
section IV.C shows, it cannot be described as a choice of which law applies. But the fact t
in the domestic context there is no temptation to do so should suggest that something odd i
going on in the realm of conflicts. My technique for revealing that oddity is basically
redescribe conflicts cases from the conventional legal perspective. I will argue that from th
perspective, conventional conflicts analysis is fatally flawed, and that conflicts cases are
scribed as involving choice precisely in order to mask these flaws. See infra section IV.
191. In particular, see Kramer, Rethinking Choice of Law, supra note 81, at 280-83.
192. See id.
do not differ in any fundamental way from those in which they do

not; the same process goes on in determining whether the plaintiff
has a right to relief. I claim that this determination does not require
a court to "choose" which law to "apply." Now, of course, this
claim is a bit hard to square with the practice of courts, and from
the descriptive perspective, Kramer has things rather easier. It
must seem more plausible to suggest that a similar process to the
explicit choice of law goes on in all cases than to suggest that the
explicit choice of law does not occur. I do not - and cannot
deny that courts often characterize their analysis as choosing be-
tween competing laws. I claim that "choice of law" is a misnomer;
it is more accurate to say that when a court "chooses" one state's
law over another's, it has actually rejected a claim of right based on
the nonselected law.193 It has refused to honor a right created by
that law, and thereby determined that rights based in the law it
"selects" prevail in a conflict between rights.194 Courts find no need
to talk of choice in the ordinary case, and there is no need to do so
in multistate cases either.
Before moving on to the second example, I want to draw three

points from the consideration of purely domestic conflicts. First,
there are potential conflicts even where the law is all from one
state. These conflicts are not considered to raise a choice-of-law
issue because they are resolved by rules. And they go away fai
quickly - they do not persist as troubling questions. Again, th
193. It is more accurate because, among other things, it makes clear that a court m
enforce rights created by more than one state, rather than determining that one state
"applies" to the entire case. Because conflicts theory historically tried to determine w
law governs, this possibility appears anomalous and receives the appropriately exotic nam
"d6pe9age." See, e.g., SCOLES & HAY, supra note 7, at 38. More significantly, thinkin
terms of conflicts focuses attention on the conflicts rule that determines which right pre
Interest analysis relies on the choice-of-law vocabulary precisely to deflect attention from
conflicts rules.
194. For a description of this situation we might turn again to Beale, who in the conflict
between law and equity had to confront a situation in which - much as the current situation
under interest analysis - two courts disagreed about which right prevailed.
[In common law jurisdictions] the theory upon which courts proceed is the theory of
separate and independent systems of right. The court of law regards the equitable right
as subordinate to the legal right, while the court of equity takes the opposite view.... It
is clear, however, that there cannot be two separate and distinct laws prevailing in the
same place at the same time; and therefore in fact, whatever may be the theory of the
courts, one of the conflicting rights must be valid and the other invalid.
1 BEALE, supra note 6, ? 4.8, at 41. I do not quote this passage for the correctness of its
conclusion. In a conflict between rights of co-equal sovereigns such as sister states, I see no
reason why there should be what philosophers call a "fact of the matter" about which right
"really" prevails. The Constitution does not resolve conflicts of its own force, and, absent
federal legislation, there is no other superior authority to make the decision. The point is
rather that Beale correctly sees not a choice between laws but a conflict between rights re-
quiring the subordination of one to the other.
because they are resolved by rules. The extent to which statutes

can retroactively alter common law rights might once have been a
difficult question - it was litigated95 - but the rules are now es-
tablished, and they govern.
Second, the conflicts rules for purely domestic cases are well
established: statutes beat common law, and recent statutes beat
earlier ones. Courts follow these rules as legal principles. They do
not suggest that a choice must be made, whereby different results
would be equally legitimate.
Finally, this analysis suggests that resolving a conflict between
two laws does not amount to a determination that one law governs
the transaction to the exclusion of the other. If the state officer
defense applies, it may bar or limit recovery; this does not mean,
however, that the tort cause of action is somehow excluded, that it
is part of a body of law that does not apply. This observation but-
tresses my claim that a conflicts case does not require a court to
identify the law that governs. The question is which right prevails.
Of course, where both contending rights originate from the same
sovereign, it is not clear what it would mean for one body of state
law to apply to the exclusion of the other. This point will be made
somewhat more strongly by the next section.
B. The Easy Conflict: State Law vs. Federal Law
Conflicts between state and federal law are easy: federal law
wins.196 They are so easy that conflicts scholars tend to them give
little attention,197 presumably because they pose no choice-of-law
question.198 My argument, however, is that no case presents a
195. See, e.g., Munn v. Illinois, 94 U.S. 113, 134 (1876) (rejecting the idea of vested rights
in common law rules); see also Landgraf v. U.S.I. Film Prods., 511 U.S. 244, 271 n.25 (1994)
(collecting cases discussing retroactive alteration of property and contract rights).
196. That federal law defeats contradictory state law follows directly from the Supremacy
Clause of Article VI, which provides that the Constitution and federal laws "shall be the
supreme Law of the Land . . . any Thing in the Constitution or Laws of any State to the
Contrary notwithstanding." U.S. CONST. art. VI.
197. By "state-federal conflicts" I mean preemption of state law by federal law. Neither
Brilmayer's treatise nor the Scoles and Hay hornbook discusses preemption as a choice-of-
law issue - presumably for the reason that there is no choice. They do discuss state attempts
to withdraw jurisdiction for federal causes of action, which I see as a somewhat more refined
attempt to subordinate federal rights.
198. It might of course be suggested that there is no issue of choice because federal law is
literally local law everywhere. See, e.g., Claflin v. Houseman, 93 U.S. 130, 137 (1876). Beale
took this tack:
There cannot be two independent laws within a territory, even though that territory be
subject to the legislative jurisdiction of two independent sovereigns. The law of the terri-
tory, resulting from the legislative action of both sovereigns, is a single law. The law of a
single legal unit must be one law, the one and undivided law of that territory.
"choice of law," and for this reason I think that state-federal con-
flicts are importantly illuminating. Federal law wins, and the con-
flict cannot be denied - it cannot be hidden behind the choice-of-
law veil. Consequently, as with domestic conflicts, we gain the op-
portunity to see what is really going on.
There are two different types of federal-state conflicts, depend-
ing on whether it is the plaintiff or the defendant who appeals to
federal law. Where it is the defendant, analysis rather obviously
follows the conventional legal model that I set out above.199 The
court must determine whether the laws invoked grant rights to the
parties invoking them, whether those rights conflict, and which
right prevails - all ordinary legal questions. CSX Transportation,
Inc. v. Easterwood200 illustrates this model.
That case featured a wrongful death suit brought by the widow
of a man killed when a train collided with his truck at a Georgia
crossing. The widow alleged that the railroad was negligent under
Georgia law for failing to maintain adequate warning devices at the
crossing and for operating the train at an excessive speed. The
complaint stated a claim under Georgia law, or at least, no one sug-
gested that it did not. The Court assumed that Georgia law had
standards governing the duties of railroads with respect to train
speeds and the safety of grade crossings, and that the plaintiff had
alleged a violation of those standards.201 Rather than challenge the
sufficiency of the complaint under Georgia law, however, the de-
fendant appealed to federal law, arguing that the Federal Railroad
Safety Act (FRSA) gave it a defense against state law tort claims.202
The Court started with a scope analysis. Whether federal law
preempts state law is a question of congressional intent, to be deter-
mined by an examination of the statute's text, structure, and legisla-
tive history - the ordinary tools of statutory construction.203 The
FRSA preemption clause stated that "laws ... relating to railroad
safety shall be nationally uniform to the extent practicable. A State
1 BEALE, supra note 6, ? 2.4, at 17-18. This seems more like definition than argument, how-
ever, and of course there are obvious senses in which federal law is not local law. Most
patently, it cannot be changed by the will of the state legislature. Equally significant, succes-
sive prosecutions under state and federal law are not barred by the Double Jeopardy Clause
precisely because the laws emanate from different sovereigns. See, e.g., Bartkus v. Illinois,
359 U.S. 121, 121-24 (1959). At any rate, I have already suggested that purely domestic cases
may present conflicts.
199. See supra section III.C.
200. 507 U.S. 658 (1993).
201. See CSX Transportation, 507 U.S. at 666.
may adopt or continue in force any law . . . relating to railroad

safety until such time as the Secretary has adopted a rule, regula-
tion, order, or standard covering the subject matter of such State
requirement."204 Thus the existence of federal regulation provided
a defense to any state-law claim on the same subject. As it turned
out, the Court determined that federal regulation barred the negli-
gence claim based on excessive speed, but not the claim relating to
the warning devices.205
This resolution might seem inconsistent with the suggestion that
federal law provides a defense against state-created rights, or at
least the defense might seem to be of an odd type. In the example I
turned to in section III.C.2, the defendant appealed to New York
law not as a force that displaced the Oregon tort action but rather
as the source for the standard of care. The effect of that maneuver,
if successful, would have been to allow him to argue that he had
met the applicable standard of care, not that no cause of action
could be maintained. The federal preemption in CSX Transporta-
tion, by contrast, did not change the question to whether the train
company had complied with the federal speed limit, but actually
prevented a state tort suit based on excessive speed regardless of
how fast the train was going.206 Thus it might seem that federal law
actually governed the suit.
That would be a misinterpretation. It is one of the axioms of
federal jurisdiction that federal law is interstitial,207 and the claim
that federal law "preempts" state law is not fundamentally different
from the appeal to a defense created by sister-state law. The com-
plete preemption of the excessive speed claim was a consequence of
the statutory language; federal law frequently works only to alter
the standard of care and not to prevent the assertion of state tort
claims.208 Some federal laws are even more permissive, preserving
all state laws that do not proscribe actions permitted, or require
204. See CSX Transportation, 507 U.S. at 662 n.2 (quoting 45 U.S.C. ? 434 (1970)).
205. See CSX Transportation, 507 U.S. at 673, 676.
206. This is not entirely true; it might have been possible for the plaintiff to bring a state-
law claim based on failure to comply with a (federal) statutory speed limit. See CSX Trans-
portation, 507 U.S. at 677 (Thomas, J., concurring and dissenting). Her complaint conceded,
though, that the federal speed limit had not been exceeded. See CSX Transportation, 507
U.S. at 672.
207. See, e.g., RICHARD H. FALLON, JR. ET AL., HART AND WECHSLER'S THE FEDERAL
COURTS AND THE FEDERAL SYSTEM 521-22 (1996).
208. For example, the National Highway Safety Act allows states to enforce laws whose
safety standards are identical to the corresponding federal standard. See 49 U.S.C.
? 30103(b)(2) (1994). It also preserves all claims at common law. See Freightliner Corp. v.
Myrick, 514 U.S. 280, 284 (1995).
actions forbidden, by federal law.209 This is also the result where

Congress has not specified the scope of preemption. Without an
explicit congressional statement, the rules of "implied conflicts pre-
emption" direct that state law is nullified only to the extent that an
actual conflict exists.210 Moreover, appeals to sister-state law do
sometimes prevent a particular cause of action from being main-
tained at all. If the cause of action the plaintiff invokes does not
exist under sister-state law, it may be wiped out just as if it had been
preempted by federal regulation.211
This, then, is what happens when the defendant appeals to fed-
eral law: the court must decide whether Congress intended that the
defense be available. If it did, the defense prevails over the state-
created right. Sometimes this is a relatively easy question: a fed-
eral law explicitly authorizing certain conduct clearly bars the impo-
sition of liability for that conduct under state law.212 Similarly, a
federal law placing a cap on tort damages would create a right avail-
able to defendants in any state tort suit. The difficult questions are
ones of preemption, when the extent of preemption will not always
be obvious. Plaintiffs seeking to circumvent federal defenses can
thus argue that the preemption is narrow and permits the state law
cause of action. All of the action in state-federal conflicts takes
place at the level of scope, because the conflicts rule is clear.
When the party appealing to federal law is the plaintiff, avoid
ance seems more difficult. No defendant could suggest that a stat
law defense vitiates a federal cause of action; this would run direct
counter to the Supremacy Clause. Consequently, a defendant mu
find some way to deny the conflict, to suggest that only state la
applies to the action. Phrasing the issue as choice of law is one wa
to do so - I have suggested that this is currently done with inte
state conflicts. But the characterization as a choice, rather than a
conflict, will be effective only if a choice-of-law rule points aw
from federal law. This requirement created some problems for de
fendants. Our canonical state-federal conflicts cases come from the
territorialist era, and the defendants seeking a territorial choice-of-
209. Title VII expressly preserves state laws that do not conflict. See 42 U.S.C.
? 2000e(7) (1994); Shaw v. Delta Air Lines, 463 U.S. 85, 101 (1983).
210. See Myrick, 514 U.S. at 287. For this reason, I think that "preemption" is a some-
what unfortunate term. What happens is simply that federal rights defeat state rights.
211. In such a case, the law's denial of the cause of action should be seen as granting
parties a right to engage in the conduct at issue.
212. See Hamm v. City of Rock Hill, 379 U.S. 306, 309-11 (1964) (describing availability
of federal defenses to prosecutions for attempts to obtain service in places of public
accommodation).
law rule selecting state law will have great difficulties: federal sov-
ereignty extends through every state, and acts anywhere in the
United States thus create rights under applicable federal laws. For-
tunately for such a defendant, the territorial choice-of-law system
was replete with escape devices that allowed judges to mitigate the
rigidity created by the last act doctrine. A defendant trying to use
choice-of-law analysis to privilege state rights over federal rights
could thus exploit the flexibility of the system to argue for the appli-
cation of state law.
Appeal to these escape devices by defendants is exactly what we

see in state-federal conflicts. In Mondou v. New York, New Haven
& Hartford R.R. Co.,213 the defendant won in state court with the
claim that the federal statute was contrary to the public policy of
the state of Connecticut and could not be enforced in its courts.214
This is, of course, a conventional choice-of-law maneuver. States
have traditionally declined to permit causes of action based on sis-
ter-state law on the grounds that the causes of action offend their
public policy.215
If the choice-of-law perspective were valid - if describing cases
as involving a choice somehow meant there was no conflict216 -
this approach would have been satisfactory. But the Supreme
Court was not fooled; it rejected the suggestion that contrary public
policy could lead to the application of Connecticut law.217 With the
public policy escape hatch closed, defendants turned to others
and the Supreme Court shut them as quickly as they opened.218
Employing a choice-of-law methodology will not allow states effec-
tively to decide that their rights prevail in conflicts with federal
213. 223 U.S. 1 (1912).

214. See Mondou, 223 U.S. at 55-56.
215. See, e.g., Mertz v. Mertz, 3 N.E.2d 597 (N.Y. 1936) (rejecting interspousal tort
claim). Oddly, the invocation of the public policy exception does not seem to correlate with
the repugnance of the foreign law - the New York Court of Appeals used it to reject the
Connecticut law of interspousal tort liability but accepted Hitler's Nuremberg laws. See
Larry Kramer, Same-Sex Marriage, Conflict of Laws, and the Unconstitutional Public Policy
Exception, 106 YALE L.J. 1965,1975 (1997); see also Monrad G. Paulsen & Michael I. Sovern,
"Public Policy" in the Conflict of Laws, 56 COLUM. L. REV. 969, 980-81 (1956).
216. See, e.g., CURRIE, supra note 22, at 181-82; SCOLES & HAY, supra note 7, at 2.
217. See Mondou, 223 U.S. at 57-58.
218. See, e.g., Testa v. Katt, 330 U.S. 386 (1947) (rejecting the argument that the federal
statute was "penal"). Early cases such as Mondou and Testa do, interestingly, stay within the
choice-of-law paradigm. Mondou holds not that public policy must yield to the Supremacy
Clause but rather that because federal policy is local policy everywhere, there is no contrary
local policy. See Mondou, 223 U.S. at 57. Testa similarly relies on the argument that, while
the federal statute might be penal, it is a local penal statute. See Testa, 330 U.S. at 392-94.
More recently, the Court has also considered such cases from the conflicts perspective, noting
that state law must yield to federal rights. See Howlett v. Rose, 496 U.S. 356, 375-81 (1990).
rights. This comports with what I have been arguing: "choice of

law" is a flight from conflicts that redescribes, but does not substan-
tively alter, the underlying issue. This is the main lesson I want to
draw from the consideration of domestic and state-federal conflicts.
Additionally, these cases show that deciding a particular law

prevails in a conflict does not mean that law governs the action.
Consider a federal law that imposes a cap on tort recoveries, pro-
viding every defendant with a resource against state-created rights.
The result, in a case where the plaintiff wins, will be that the federal
defense imposes a limit on the permissible recovery: the state law
right is restricted by the federal defense. This does not mean the
transaction is governed by federal law, however; federal law gives
no right to recovery, and if the entire case were really governed by
federal law, the plaintiff would get nothing. It is not clear why a
similar approach would not make sense in multistate cases - why it
should not be the case that nonconflicting rights from both states
are recognized.219 Saying that the court's task is to "choose" the
law that "applies" obscures this possibility; choice-of-law theory ex-
oticizes the ordinary by calling it "depegage."220
C. Back to Choice? State Law vs. State Law
The preceding section has argued that there is nothing anal

cally special about multistate cases. Purely domestic cases, or s
federal conflicts, may raise the same issues. These latter two
of cases, however, are not generally considered to raise choic
law issues. This is so, I suggest, because the results are clear,
choice-of-law rhetoric will not change them. Attempts
choice-of-law methodology to reach prohibited conflicts resu
the defeat of federal rights by state rights - have been consis
rejected. Choice-of-law rhetoric may have a use, though, if th
flicts results appear unacceptable but the appropriate solution
clear: it may mask the illegitimacy. That, I believe, explains it
tinuing allure in multistate cases. The correct way of implem
constitutional restrictions on state conflicts rules has not yet
discerned, and the rhetoric of choice makes the constituti
219. In fact, Currie suggested that while a state would apply its "rules of decision
accident between two domiciliaries, it would determine negligence by looking to the "r
conduct" (such as speed limits) of the place of the accident. See CURRIE, supra note
68-69.
220. The Second Restatement, to its credit, explicitly contemplates depecage; it ad

a choice of law calculus for each issue in a case. See RESTATEMENT (SECOND) OF CO
OF LAWS ? 145, cmt. d (1971).
problems with the Supreme Court's current laissez-faire approach

less obvious.
Describing the Court's failed attempts and uncovering the ap-

propriate methodology is the task of the subsequent Parts of this
article. The point here is simply that if conflicts results are illegiti-
mate, resort to a choice-of-law description will not save them. That
is the lesson of state-federal conflicts. Given that premise, interest
analysis can be rejected if the conflicts perspective reveals that it
reaches unacceptable results. This section will attempt to show pre-
cisely that. But first I want to make a more ambitious claim: inter-
est analysis is incoherent on its own terms. It simply is not a
method of determining what law applies to a case. The conflicts
perspective is not just more useful; it is the only intelligible choice.
Imagine a married couple with different domiciles. Husband's
domicile (say, Connecticut) has interspousal tort immunity; Wife's
domicile (say, New York) does not. Driving in separate cars, they
collide; both are injured, and both are arguably at fault. Husband
sues Wife in Connecticut, and she files a counterclaim. According
to interest analysis, what law applies to this case?
The result is relatively easy to discern. Connecticut has an inter-
est in affording Husband the protection of interspousal immunity,
so it will. Connecticut law applies to Wife's counterclaim, and she
cannot recover. But Connecticut has no interest in affording Wife
the benefit of that immunity, and it does have an interest in com-
pensating Husband. Husband will recover from Wife, under the
law of New York.221
Thus interest analysis directs application of one state's law to

Husband's claim and another state's law to Wife's counterclaim. (It
parcels out these results so as to favor its domiciliary, but let us
ignore that for the moment.) Yet not only are both these claims
part of the same case, they both arise from the same collision. It is
untrue that one state's law applies to this case, or even to this acci-
dent; consequently, it is untrue that interest analysis allows courts
to choose which law to apply.222
221. This example is analytically similar to the married women's contracts case discussed
by Currie; I have tinkered with it a bit in order to generate a counterclaim arising from the
same transaction. For Currie's similar conclusions, see CURRIE, supra note 102, at 90-91
(finding that a court in a state with married women's disability should apply local law to a
claim by foreign creditor against domestic married woman but should apply foreign law to a
claim by domestic creditor against foreign married woman).
222. The example works with regard to interest analysis, but it should be clear that simi-
lar examples can be generated for any conflicts theory that is not jurisdiction-selecting.
It is true, of course, that interest analysis chooses the law that

applies to a particular claim, but at this point the choice-of-law rhet-
oric is idling. There is really no difference between "choosing" law
claim-by-claim, and accepting or rejecting the rights undergirding
each particular claim. Under even modest analytical pressure, the
choice-of-law perspective seems to collapse into the conflicts
perspective.
A modern interest analyst such as Kramer might have an answer
here. Only one law applies, he might say, and it is forum law; the
law just does not give the same rights to Wife as it does to Hus-
band.223 That account rescues the claim that interest analysis
selects a law to apply, but at some cost. Interest analysis not only
directs that forum law prevail against foreign law, it alters forum
law to disfavor foreigners. That raises the question of discrimina-
tion, which is the main focus of this section.224 I have said that if
the conflicts results of interest analysis are unacceptable, the rheto-
ric of choice will not save them. It is time to consider those results.
Interest analysis is biased in two distinct ways: against foreign
domiciliaries, and against foreign law. I will use two hypotheticals
to highlight these different forms of discrimination. For the first, I
can do no better than an example contrived by Douglas Laycock.225
Laycock asks us to imagine two acquaintances, Mary from Mary-
land and Del from Delaware. They drive together, taking turns be-
hind the wheel, and each is injured in an accident with the other
driving. (I will suppose, though Laycock is not explicit on this
point, that the accidents take place in the same state.) Mary sues
Del in Delaware, and Del files a counterclaim. Delaware has a stat-
ute preventing guests from suing hosts for injuries in auto accidents,
and Maryland does not.
According to interest analysis, Mary's claim will be barred but
Del's will not. Delaware has an interest in applying its guest statute
to protect its domiciliary Del, so it will. But it has no interest in
protecting Mary, and it does have an interest in compensating Del,
so the guest statute will not prevent Del's suit.
What does this mean from the conflicts perspective? Dela-
ware's guest statute grants drivers of automobiles rights against
their guests: it grants them the right not to be held liable for their
guests' injuries. This right is not given to everyone driving in Dela-
223. Cf. Kramer, Myth, supra note 81, at 1054-55.

224. In fact, Kramer finds that selectively granting rights to forum domiciliaries but not
others will sometimes violate the Privileges and Immunities Clause. See id. at 1065-74.
225. See Laycock, supra note 73, at 276.
ware, however - according to the interest analyst's scope-based

approach, the right is not extended to out-of-staters at all. Del can
assert it, but Mary cannot. Local law gives rights to locals but not
to out-of-staters, simply because they are from out of state. This is
discrimination against foreign domiciliaries.
What of the discrimination against foreign law? Suppose that
Al, from Alabama, is driving near his house when he collides with
Georgia resident George. Al sues George in Alabama. Suppose
further that Georgia has a damages cap limiting recovery to
$50,000, while Alabama has a more generous $75,000 cap. Under
interest analysis, the court will apply Alabama law and reject the
Georgia damages cap. Now suppose that Lou, from Louisiana, has
a similar run-in with Al, this time in Louisiana, which has no dam-
ages cap. Lou sues Al in Alabama. Under interest analysis, the
court will again apply Alabama law, giving Al the benefit of the
damages cap.
What has happened here? Conflicts analysis works, in theory,
by examining the contacts that an action has with different jurisdic-
tions. Standard factors for car accident cases, according to Currie,
are the location of the accident, the domicile of the parties, and the
location of the forum.226 Leave aside for the moment the location
of the forum. Al v. George, from the perspective of the Alabama
court, has the following arrangement of contacts: local plaintiff,
foreign defendant, local accident. The result is that local rights pre-
vail. Lou v. Al has the opposite arrangement of contacts: foreign
plaintiff, local defendant, foreign accident. The two suits are what I
will call "mirror-image" cases. If there is any reason, based on
these three contacts, why local rights as to the available damages
should prevail in the Al v. George suit, it can be mustered in favor
of the corresponding foreign rights in Lou v. Al. If the conflicts
rule is neutral between local and foreign rights, each should prevail
in one case.227 But of course under interest analysis, foreign rights
prevail in neither. The location of the forum is dispositive; the fo-
rum applies local law because it is local law. Foreign rights are dis-

227. A common law judge evaluating precedents with no knowledge of interest analysis
might well think that a decision in favor of local rights in Al v. George compelled a decision
favoring foreign rights in Lou v. Al. If there is such a thing as the nature of law, it might be
encapsulated in Cardozo's adage: "It will not do to decide the same question one way be-
tween one set of litigants and the opposite way between another.... If a case was decided
against me yesterday when I was defendant, I shall look for the same judgment today if I am
plaintiff." BENJAMIN N. CARDOZO, THE NATURE OF THE JUDICIAL PROCESS 33-34 (1921).
From the classic common law perspective, interest analysis hardly appears a legal theory at
all.
favored in true conflicts not because of some arrangement of

contacts that can be applied in mirror-image cases, but simply be-
cause they are foreign. Without the principle of rejecting foreign
rights because of their origin, interest analysis is flatly incoherent;
the rules it prescribes are self-contradictory.
Currie suggested that in true conflicts, "[a] court need never
hold the interest of the foreign state inferior; it can simply apply its
own law as such."228 This may be true from the choice-of-law per-
spective: the court has not attempted to weigh interests, and so, a
fortiori, it has not found the foreign interest inferior. But what the
mirror-image Al v. George and Lou v. Al cases show, from the con-
flicts perspective, is that the court has indeed held foreign interests
inferior - not in the sense that it has found the foreign state to be
less interested, but in the sense that it has taken foreign interests
less seriously than Alabama interests. If Alabama can make an ar-
gument that Alabama rights should prevail in Al v. George for any
reason other than that they are local rights, Louisiana can make the
same argument for Louisiana rights in Lou v. Al. If Alabama is
deaf to those arguments in Lou v. Al, when it found them convinc-
ing in Al v. George, it has not given rights created by Louisiana law
the same respect it gives Alabama rights.
Are these forms of discrimination constitutional? State-federal
conflicts are so easy because the federal government is superior to
the states. Interest analysis derives its plausibility from the claim
that the coequal status of the states makes things different. No
state can force another to apply its law; hence an interested state
may always apply its own law to a case in its courts. This makes
some sense from the choice-of-law perspective, using a personal-
jurisdiction-style approach. But the conflicts perspective shows that
assertions of legislative jurisdiction involve the rejection of foreign
rights, and that interest analysis makes these assertions in a discrim-
inatory fashion.
The basic problem with interest analysis is that it does not direct
states to treat each other as equals. It prescribes, quite candidly,
that foreign law should be rejected in conflicts simply because it is
foreign, that scope analysis should withhold rights from out-of-
staters simply because they are not locals. The coequal status of
states is not an explanation of this approach - it is an indictment.
If the principle of state equality has any force at all, it prohibits
228. CURRIE, supra note 22, at 181-82.
interest analysis.229 The next Part argues that even the weakest
reading of the relevant constitutional provisions leads to this con-
clusion, laying the groundwork for Part VI, which shows the consti-
tutional consequences for conflicts more generally.
V. CONFLICTS AND THE CONSTITUTION
Two constitutional provisions will have relevance to this discu

sion: the Full Faith and Credit Clause, and the Privileges and Im
munities Clause.230 I will offer here what I think is a very mode
reading of these Clauses, and then show that even the weak read
has quite dramatic consequences for conflicts theory. Before th
textual exegesis, however, a bit of historical analysis is in order
The Supreme Court has done almost nothing with the Privile
and Immunities Clause in the conflicts arena,231 so there is no his-
tory in need of recapitulation. There is a long series of Full Faith
and Credit decisions, and these are worth revisiting.
A. Introduction
If the current conceptual approach to conflicts has any virtu

is that it spares the courts from confronting a difficult constituti
issue: the extent to which the Constitution requires federal in
ference with state prioritization of interests. Deciding what sor
contacts with a state are important enough to justify a claim of
islative jurisdiction is quintessentially a matter of local concern
is a matter of deciding how important are the various state pol
implicated by different contacts. But these assertions come at
expense of sister-state legislative jurisdiction, and that is quin
sentially a matter of federal concern. If the "choice-of-law" q
tion is conceived of in terms of interstate recognition of righ
and I hope to have shown the correctness of that conception -
becomes immediately apparent that the Constitution has obvio
relevance. The problem is how to accommodate both local and
eral aspects of the issue.232 This is hard, and I think the Supr
Court's abandonment of the conflicts field follows from a reco
229. See, e.g., Ely, supra note 73; Laycock, supra note 73.
230. U.S. CONST. art. IV, ?? 1-2. Additionally, I will make some fleeting reference
Due Process. The Commerce Clause is also important, but not for present purposes.
231. See the sparse discussion in SCOLES & HAY, supra note 7, at 104-07.
232. See Robert H. Jackson, Full Faith and Credit: The Lawyer's Clause of
Constitution, 45 COLUM. L. REV. 1, 28 (1945) ("How to determine when [federal con
tions] require the law of the forum to give way to the law of another state seems to
unsettled question.... The ultimate answer, it seems to me, will have to be based on c
erations of state relations to each other and to the federal system.").
tion of the difficulty and a choice to err on the side of federalism

rather than nationalism.
Many scholars have suggested that the Constitution has a good

deal to say about conflicts of law.233 This is hardly surprising. The
great aim of the Constitution is to knit the discrete sovereignties of
the states into a federal union, and this purpose obviously requires
rules governing the treatment of the laws, and the citizens, of sister
states. If the states related to each other as foreign sovereigns,
these would be political questions, to be answered perhaps by trea-
ties, perhaps by principles of comity. But states do not relate to
each other in that way.234 They cannot make treaties among them-
selves.235 Questions that were political have been made legal
which is to say, constitutional. As Laycock puts it, "How Texas
courts treat the law of a sister state is a matter of law, not comity,
and the choice is no longer voluntary. For this purpose Texas is not
a sovereign state; it surrendered this portion of its sovereignty when
it joined the Union."236
What is surprising, then, is not the suggestion that the
Constitution should supervise state conflicts rules but rather the ex-
tent to which the Supreme Court has ignored the suggestion.
Closer analysis reveals an explanation, though: changed under-
standings of conflicts have created serious federalism challenges to
constitutional supervision, and the current Court is receptive to fed-
eralism concerns.237 The Court did try to develop a doctrine of con-
stitutional conflicts law, at least under the Full Faith and Credit
Clause; its early conflicts cases suggested that the vested rights the-
ory had constitutional force.238 This made sense, from Beale's per-
spective - laws, being territorial, did not conflict. The only
233. Most notably Douglas Laycock, see Laycock, supra note 73, on whose historical ar-
guments this section relies quite heavily. See also Ely, supra note 73; Jackson, supra note 232;
Katzenbach, supra note 11; James R. Pielemeier, Why We Should Worry About Full Faith and
Credit to Laws, 60 S. CAL. L. REV. 1299 (1987).
234. See, e.g., Magnolia Petroleum Co. v. Hunt, 320 U.S. 430, 439 (1943) (stating that the
Full Faith and Credit Clause "altered the status of the several states as independent foreign
sovereignties, each free to ignore rights and obligations created under the laws or established
by the judicial proceedings of the others, by making each an integral part of a single nation").
235. See U.S. CONST. art. I, ? 10.
236. Laycock, supra note 73, at 259; see also Jackson, supra note 232, at 30.
237. See, e.g., Printz v. United States, 521 U.S. 898, 918-22 (1997); Sun Oil Co. v.
Wortman 486 U.S. 717, 727-28 (1988) ("If we abandon the currently applied, traditional no-
tions of [legitimate state legislative jurisdiction] we would embark upon the enterprise of
constitutionalizing choice-of-law rules, with no compass to guide us beyond our own percep-
tions of what seems desirable.").
238. See, e.g., Allgeyer v. Louisiana, 165 U.S. 578, 587-90 (1897); Chicago & Alton R.R. v.
Wiggins Ferry Co., 119 U.S. 615, 622 (1877); Green v. Van Buskirk, 72 U.S. (5 Wall.) 307, 313-
14 (1866); see also Pielemeier, supra note 233, at 1303-04.
question was whether foreign rights would be recognized, and Full

Faith and Credit prescribed that they would. The erosion of the
territorial conception revealed the possibility of conflict between
laws, however, and the Court accordingly began to discuss conflicts
in terms of evaluating the competing interests of the states.239 Ap-
plication of Full Faith and Credit became more difficult, but the
Court still supposed that the Clause resolved conflicts in favor of
the state with superior interests.240
But to tell a state that its interests are inferior is a serious in-
fringement on its power to determine the relative importance of
particular policies. Surely it is for each state to decide which inter-
ests are more important to it. The federal judiciary would under-
standably hesitate before entering so deeply into the internal
workings of state government, and the Full Faith and Credit Clause
gives no guidance as to what makes an interest superior.241 Conse-
quently, the Court quite swiftly abandoned the idea that Full Faith
and Credit determined unique solutions and fell back to a safe dis-
tance from which to oversee state conflicts rules: a baseline test for
legitimate application of forum law.
[T]he full faith and credit clause does not require one state to substi-
tute for its own statute, applicable to persons and events within it, the
conflicting statute of another state, even though that statute is of con-
trolling force in the courts of the state of its enactment with respect to
the same persons and events.242
That at least hewed to a territorial line, invoking the state's in-
terest in regulating "persons and events within it." But territorial-
ism was on the wane, and the Court soon retreated farther, ruling
239. See, e.g., Alaska Packers Assn. v. Industrial Accident Commn., 294 U.S. 532, 547-58
(1935).
240. See Alaska Packers, 294 U.S. at 547-58. I think that at this point the Court was not
far wrong. The Alaska Packers Court seemed to suppose that the Full Faith and Credit
Clause contained the "rational" rules determining which state's interest was superior. That
was a mistake; the setting of priorities is indeed a matter for the states. But this does not
mean that the Clause has nothing to say. It requires that a state base its assertion of legisla-
tive jurisdiction on a claim that its interests are superior; this is the only way to afford foreign
law the appropriate respect while still determining that local rights prevail. It further re-
quires that the interests not be superior merely because another state's interests are weighed
less heavily. See infra section VI.B.1.
241. Justice Jackson noted, "Nowhere has the Court attempted, although faith and credit
opinions have been written by some of its boldest-thinking and clearest-speaking Justices, to
define standards by which 'superior state interests' in the subject matter of conflicting stat-
utes are to be weighed." Jackson, supra note 232, at 16. The reluctance to impose particular
substantive standards, I will suggest, was entirely correct. Full Faith and Credit does not set
out standards by which a court may determine which state's interest is greater; it simply
demands that states respect each other's laws. What I try to show here is how the principle of
respect for sister-state law translates into restrictions on state conflicts rules.
242. Pacific Employers Ins. Co. v. Industrial Accident Commn., 306 U.S. 493, 502 (1939).
that Full Faith and Credit did not require displacement of the fo-
rum state's law if the state had a "substantial connection" to the
action.243 The retreat has by now become a rout. In Allstate Insur
ance Co. v. Hague,244 the Court announced that the only restrictio
the Constitution placed on state conflicts rules was the requiremen
that the choice of law be "neither arbitrary nor fundamentally un
fair."245 Hague was a Wisconsin domiciliary who died when th
motorcycle on which he was a passenger was struck by a car. Th
accident occurred in Wisconsin, and the drivers of both vehicles
were Wisconsin domiciliaries. Neither driver carried valid insur-
ance, but Hague held an insurance policy, issued in Wisconsin, th
offered up to $15,000 for loss incurred in accidents with uninsu
motorists. He owned three cars, and the policy covered each. W
consin law would have limited his recovery to $15,000; Minneso
law, however, allowed the coverage on each car to be "stacke
raising the limit to $45,000. Hague's widow moved to Minnes
after the accident (for what the Court called "bona fide" reasons2
and brought suit there seeking the more favorable terms of Min
sota law, which the Minnesota courts gave her. The Supreme Cou
stated that "if a State has only an insignificant contact with the par
ties and the occurrence or transaction, application of its law is u
constitutional."247 It nonetheless affirmed, finding that the wido
new domicile, and the fact that the decedent had worked in Min
sota (and commuted to work there), created an interest sufficient
justify application of Minnesota law.248
The "neither arbitrary nor fundamentally unfair" language
Allstate is the language of due process, and as a construction of
Due Process Clause, it makes perfect sense. But the Due Proc
Clause is about the rights of individuals, and conflicts cases rai
other issues. Parties may resist application of a particular law o
the grounds that it is unfair, but more often their argument is t
they may not be subjected to the liabilities of one law because a
other law operates to shield them. This is no longer simply a que
tion of individual rights or due process, but one of the respect d
243. See Cardillo v. Liberty Mut. Ins. Co., 330 U.S. 469, 476 (1947).
244. 449 U.S. 302 (1981).
245. Allstate, 449 U.S. at 320.
246. See Allstate, 449 U.S. at 319 n.28.
247. Allstate, 449 U.S. at 310-11.
248. Allstate requires that a state have "a significant contact or significant aggregation
contacts." Allstate, 449 U.S. at 313. This does mean something; the Court has ruled tha
state may not apply its law to suits to which it has no connection. See Phillips Petroleum C
v. Shutts, 472 U.S. 797, 821-22 (1985).
to sister-state law; that is, it is a question of Full Faith and Credit.

Yet Allstate holds that the answer is always the same.249
It is disappointing - though, I have suggested, not incompre-
hensible - that the Court has merged the limitations imposed by
Due Process and those of Full Faith and Credit, because the two
Clauses could not be more different. Apart from the fact that Due
Process governs relations between states and individuals, while Full
Faith and Credit governs interstate relations, there is an important
conceptual difference. Due process analysis sets a minimum thresh-
old; beyond that threshold, there are no restrictions. Consequently,
a due process analysis often leads to the conclusion that a number
of different states' laws may apply. (This is, of course, the personal-
jurisdiction-style analysis whose presence in the conflicts realm I
have been deploring.) Full Faith and Credit, by contrast, demands
that each state accord the greatest degree of respect - full faith
and credit - to the laws of sister states. This may be a baseline
requirement in some sense, but the baseline is set as high as it possi-
bly could be.250 To suppose that such a forceful command results in
the same threshold test as Due Process - in particular, the tooth-
less Allstate test - is to suppose that the Constitution cares very
little about the resolution of conflicts between laws.
That supposition is of course false. Discrimination in choice of
law can easily become discrimination against foreigners - this is
precisely what Currie's interest analysis shows - and discrimina-
tion against sister-state citizens is one of the most obvious threats to
the Union. The Framers were quite clearly aware of this. For evi-
dence we need look no further than the Federalist Papers, where
Alexander Hamilton invoked the "horrid picture of the dissensions
and private wars" that wracked Germany before the creation of an
impartial court to decide questions between members of different
249. See Allstate, 449 U.S. at 308 n.10. This portion of the opinion was only a plurality,
see 449 U.S. at 320-22 (Stevens, J., concurring) (distinguishing between the clauses), but the
Court has shown no subsequent inclination to distinguish between Full Faith and Credit and
Due Process analyses.
250. See Laycock, supra note 73, at 296. Laycock is essentially correct to suggest that
since "full faith and credit" is what state courts give their own laws, the Clause demands
equality of treatment. See 3 JOSEPH STORY, COMMENTARIES ON THE CONSTITUTION OF THE
UNITED STATES ? 1304 (1833) (stating that the Framers' intent was to give foreign laws "full
faith and credit ... so that they cannot be denied, any more than in the state, where they
originated").
sovereignties.251 Indeed the Constitution is replete with provisions

intended to restrain geographical favoritism.252
Although the most textually obvious candidate is the Privileges
and Immunities Clause, which Hamilton "esteemed the basis of the
Union,"253 Full Faith and Credit is no insignificant part of this de-
sign. The Framers most likely supposed that the Full Faith and
Credit Clause prescribed unique answers to choice-of-law ques-
tions, regardless of the forum in which suit was brought, and thus it
was not the source of a personal-jurisdiction-style analysis, which
produces several acceptable, nonunique answers. They thought this
not because they had a stronger reading of Full Faith and Credit
than we do now, but because they drafted it against the backdrop of
a particular understanding of conflicts of law.254
While the precise contours of this understanding are probably
impossible to recover, and different Framers quite likely had differ-
ent understandings,255 they seem to have shared some variant of the
classic territorially oriented theory.256 Story's authoritative treatise,
first published in 1834, comments in its "Introductory Remarks"
that "[i]t is plain that the laws of one country can have no intrinsic
force, propria vigore, except within the territorial limits and juris-
251. See THE FEDERALIST No. 80, at 477 (Alexander Hamilton) (Clinton Rossiter ed.,
1961).
252. The Full Faith and Credit and Privileges and Immunities Clauses are the obvious
examples. Federal diversity jurisdiction is also targeted at interstate discrimination. See, e.g.,
id.; Baxter, supra note 92; Henry J. Friendly, The Historic Basis of Diversity Jurisdiction, 41
HARV. L. REV. 483 (1928); Laycock, supra note 73, at 278-83; Pielemeier, supra note 233, at
1316-22.
253. THE FEDERALIST, supra note 251, at 478; see also Paul v. Virginia, 75 U.S. (8 Wall.)
168, 180 (1868) ("[N]o provision in the Constitution has tended so strongly to constitute the
citizens of the United States one people as this."). Why the Supreme Court's fairly robust
Privileges and Immunities Clause jurisprudence, see, e.g., Supreme Court of New Hampshire
v. Piper, 470 U.S. 274 (1985); Austin v. New Hampshire, 420 U.S. 656 (1975), has exerted so
little influence on its conflicts jurisprudence is a mystery. In Allstate, for example, a Privi-
leges and Immunities argument could have been made that Minnesota courts would surely
not have decided the case the same way if the widow had been a Wisconsin domiciliary, and
hence the courts were awarding to locals benefits withheld from foreigners. See Ely, supra
note 73, at 185-89 (noting tension between Allstate and Austin). It may be that the rhetoric of
choice obscures things here too, by avoiding talk of the rights on which Privileges and Immu-
nities case law focuses.
254. See Laycock, supra note 73, at n.276.

255. See Jackson, supra note 232, at 6.
256. See Laycock, supra note 73, at 289-90. Future Chief Justice John Marshall appar-
ently gave a quite clear statement of the vested rights contract theory in the Virginia ratifying
convention. See id. at 306-07. In 1797, the Supreme Court had quoted Ulrich Huber's (the
Court referring to him as Huberus) territorialist maxims. See Emory v. Grenough, 3 U.S. (3
Dall.) 369, 370 n.* (1797). It was applying a territorial theory to decide torts cases as early as
1842, with no suggestion that it was creating a new approach or rejecting an earlier under-
standing. See Smith v. Condry, 42 U.S. (1 How.) 28, 33 (1842).
diction of that country."257 He repeats the principle as "[t]he first

and most general maxim" of international jurisprudence.258 Nor
was Story innovating in this regard; he derived his general theory in
large part from the Frisian jurist Ulrich Huber, who published in
the late seventeenth century.259 Story claimed that Huber's territo-
rial principles had "been sanctioned both in England and America
by a judicial approbation, as direct and universal as can fairly be
desired for the purpose of giving sanction to it as authority, or as
reasoning.9"260
Under a territorial approach, as explained above, there is no
question of laws conflicting. Only the territorially appropriate law
creates a right. The question, from the Framers' perspective, would
simply have been whether rights created by the law of one state
would be recognized by the courts of another. To answer that ques-
tion, one could hardly draft a more emphatic provision than the
Full Faith and Credit Clause. Rights acquired in one state must be
respected everywhere, regardless of whether other states disagree
with the substantive law creating those rights.261
That, then, is the most plausible original understanding: con-
flicts cases have unique resolutions because the only question is
whether sister-state rights shall be recognized or rejected, and the
Full Faith and Credit Clause requires the former. Indeed, the
Supreme Court's early ventures into conflicts jurisprudence con-
sisted exactly of the constitutionalization of the vested rights the-
ory.262 Things have changed since the Framers' days, however, and
the changes have weakened some of their devices for national
unity. Under the Framers' understanding, parties in federal court
by reason of diversity jurisdiction would have been guaranteed an
unelected, life-tenured federal judge, presumably less prone to pa-
rochialism than an elected state judge. This guarantee has survived.
Those parties would also have been guaranteed the general federal
common law, however, applied by a decisionmaker not bound by
the decisions of state courts, a decisionmaker who would "never
257. STORY, supra note 30, ? 7, at 8.

258. Id. m 18, at 21.
259. See SCOLES & HAY, supra note 7, ? 2.2, at 9, ? 2.4, at 12.
260. STORY, supra note 30, q[ 38, at 36. For a listing of largely territorialist state court
cases roughly contemporaneous with the drafting of the Constitution, see Laycock, supra
note 73, at 307 n.340.
261. See Fauntleroy v. Lum, 210 U.S. 230, 237-38 (1908) (Holmes, J.). The rights that
vested under the territorially appropriate law were not too dissimilar from the rights created
by judgments. See 1 BEALE, supra note 6, ? 8A.10. The significant point here is that there
were no opposing rights.
262. See supra note 238.
immolate truth, justice, and the law because a State tribunal has
erected the altar and decreed the sacrifice."263 With the general
federal common law died one of the antidiscrimination tools of di-
versity jurisdiction.264
Time has similarly enervated Full Faith and Credit. The Clause
was not drafted in an era that saw the possibility of interstate con-
flicts of rights. If the question is only whether rights acquired under
the law of one state shall be respected, when no other rights oppose
them, the Clause provides a clear affirmative answer; but once
there is the possibility of conflicting rights, or doubt about which
law is the law producing the rights, things get harder. What could
Full Faith and Credit mean in this context? The Supreme Court has
suggested that a literal reading would mean that in conflicts cases,
forum law must always yield - that "the statute of each state must
be enforced in the courts of the other, but cannot be in its own"265
- an obviously absurd result. Consequently, where cases feature
competing rights, the Court has said essentially that forum law may
always prevail.266
Douglas Laycock characterizes this approach as embodying the
belief that "the phrase cannot be taken literally, and therefore it
need not be taken seriously at all."267 This is fairly accurate on Lay-
cock's part, and total nonsense on the Court's. Full Faith and
Credit must be taken seriously, and it can be taken literally. We no
longer have the backdrop of territorial rules of scope, but Full Faith
and Credit can do a lot of work without that jurisprudential back-
ground. Demonstrating the power of even a weak reading of the
Clause requires that I give that reading, and it is time now to begin
the textual exegesis.
263. Gelpcke v. City of Dubuque, 68 U.S. (1 Wall.) 175, 206-07 (1863).

264. See Pielemeier, supra note 233, at 1316-19. On diversity as antidiscrimination, see
generally Friendly, supra note 252. Litigants might also have gotten federal conflicts rules, a
hope slain by Klaxon v. Stentor Electric Manufacturing Co., 313 U.S. 487 (1941). I do not
believe that federal conflicts rules are necessary, provided that we pay attention to constitu-
tional restrictions on state conflicts rules. It is troubling that under Klaxon the federal courts
act as ventriloquists' dummies, reproducing the very parochialism and bias their diversity
jurisdiction exists to counter. The Second Circuit's experience with New York law is espe-
cially notable and unfortunate in this regard. See, e.g., Rosenthal v. Warren, 475 F.2d 438 (2d
Cir. 1973); Pearson v. Northeast Airlines, Inc., 309 F.2d 553 (2d Cir. 1962) (en banc). It is the
substantive bias of state choice-of-law rules that is the real problem, though, and if attention
to the Constitution will eradicate it, there is no harm in having federal courts follow state
conflicts law.
265. Alaska Packers Assn. v. Industrial Accident Commn., 294 U.S. 532, 547 (1935).
266. See infra text accompanying notes 276-80.
267. Laycock, supra note 73, at 295.
B. The Two Clauses
The Full Faith and Credit Clause, as discussed above, dicta

that "Full Faith and Credit shall be given in each State to the
Acts, Records, and judicial Proceedings of every other State
The Privileges and Immunities Clause provides that "[t]he Ci
of each State shall be entitled to all Privileges and Immuniti
Citizens in the several States."269
These are clearly both nationalizing Clauses, intended to bind

the several states into a union.270 They are also both clearly antidis-
crimination Clauses. Full Faith and Credit governs discrimination
against foreign law, Privileges and Immunities discrimination
against foreign people. To put the point slightly differently, Full
Faith and Credit determines when parties must be accorded the
rights granted by foreign law; Privileges and Immunities when they
must be accorded rights granted by forum law. The Clauses de-
mand equality of treatment, but what is the cash value of this equal-
ity?271 I do not want my argument to rely on dubious or overstrong
interpretations of the Clauses. The reading I offer is thus a mini-
malist one - the weakest I can come up with.
My reading is the following. The Full Faith and Credit Clause
means that a state may not refuse to recognize rights created by the
268. U.S. CONST. art. IV, ? 1.

269. U.S. CONST. art. IV., ? 2.
270. See, e.g., Hughes v. Fetter, 341 U.S. 609, 612 n.9 (1951) ("[The Full Faith and Credit
Clause] 'altered the status of the several states as independent foreign sovereignties, each
free to ignore rights and obligations created under the laws or established by the judicial
proceedings of the others, by making each an integral part of a single nation."' (quoting
Magnolia Petroleum Co. v. Hunt, 320 U.S. 430, 439 (1943))); Toomer v. Witsell, 334 U.S. 385,
395 (1948) ("The primary purpose of [the Privileges and Immunities Clause] ... was to help
fuse into one Nation a collection of independent, sovereign States.").
271. Different notions of equality compete in conflicts theory. One directs that people
acting in the same jurisdiction be treated the same regardless of where they are from (equal-
ity across domicile, power arranged territorially), the other that people from the same state
be treated the same regardless of where they act (equality across territory, power arranged
personally). See Mark D. Gergen, Equality and the Conflict of Laws, 73 IOWA L. REV. 893,
902 (1988). Gergen suggests there is no clear reason to prefer a territorial arrangement of
state power to a personal one, so that scholars should simply accept "that any approach or
policy will treat people unequally for reasons that may seem arbitrary to some people," id. at
902, but admits that those urging a territorial ordering have "a [constitutionally] stronger
argument," id. at 906. The Constitution indeed seems to have a territorial orientation: at
least, the Privileges and Immunities Clause entirely rejects the idea of ordering power on a
personal basis.
Peter Westen has suggested that equality is an "empty" idea that should be eliminated
from legal discourse. See Peter Westen, The Empty Idea of Equality, 95 HARV. L. REV. 537
(1982). Westen claims that appeals to equality require some underlying notion of a relevant
difference, and that equality arguments can be paraphrased as arguments about the relevance
of the difference. The move is reminiscent of the realist attacks on the concept of rights. I
have doubts about whether the paraphrases do avoid reliance on equality norms, but in any
case I do not think that legal theory would benefit from eliminating the concept.
law of a sister state simply because those rights are foreign. It may
subordinate foreign rights to its own in some circumstances, but it
must have a reason for doing so, and the reason may not be merely
that the foreign rights are foreign.272 Similarly, the Privileges and
Immunities Clause means that a state may not deny to out-of-
staters rights it accords domiciliaries simply because out-of-staters
hail from outside its borders. It may selectively withhold rights in
some circumstances, but it must have a reason, and the reason may
not be merely that those from whom it withholds the rights are for-
eign. Nor may its conflicts rules provide that rights in the hands of
locals have greater force than the same rights in the hands of out-
of-staters simply because of their domicile.
Both of these readings fall comfortably within the current juris-
prudential understanding of the Clauses; both have been explicitly
voiced by the Supreme Court.273 The next section will show that
even such minimalist accounts, taken seriously, have profound im-
plications for conflicts. It remains here only to clear up a few
slightly murky issues about the scope of the Clauses.
272. Full Faith and Credit targets discrimination in favor of local law, but by its own
terms prohibits discrimination against any law solely by reason of its origin. Thus
Massachusetts could not provide, for example, that in conflicts cases it will always apply
Alaska law, because this would discriminate against other laws purely on the basis of their
origin.
273. See Austin v. New Hampshire, 420 U.S. 656, 660-61 (1975); Toomer v. Witsell, 334
U.S. 385, 396 (1948) (stating that the Privileges and Immunities Clause bars "discrimination
against citizens of other States where there is no substantial reason for the discrimination
beyond the mere fact that they are citizens of other States"). With respect to Full Faith and
Credit, the antidiscrimination principle appears clearly in cases involving attempts by states
to close their courts to foreign law. See, e.g., First National Bank of Chicago v. United Air
Lines, 342 U.S. 396 (1952); Hughes v. Fetter, 341 U.S. 609 (1951); see generally Lea Brilmayer
& Stefan Underhill, Congressional Obligation to Provide a Forum for Constitutional Claims:
Discriminatory Jurisdictional Rules and the Conflict of Laws, 69 VA. L. REV. 819, 825-26
(1983) ("The Supreme Court has consistently held, under the full faith and credit clause, that
a state must give the same access to its court system for actions based on another state's law
as it does for similar actions based on its own law."). As discussed in the text, this antidis-
crimination reading is not quite as obvious with respect to rejections of foreign rights based
on conflicting local law, rather than on jurisdictional limitations. Germs of the reading - or
a similar one - did make an appearance in Alaska Packers Assn. v. Industrial Accident Com-
mission of California, 294 U.S. 532, 547-48 (1935), where the Court suggested that Full Faith
and Credit would require recognition of foreign rights whenever a party could show "upon
some rational basis, that of the conflicting interests involved those of the foreign state are
superior to those of the forum." The Court quickly retreated from the suggestion that Full
Faith and Credit contemplated balancing of interests in Pacific Employers Insurance Co. v.
Industrial Accident Commission, 306 U.S. 493 (1939), and buried it entirely in Carroll v.
Lanza, 349 U.S. 408 (1955). I do not suggest that courts should decide which state's interest
is superior; the task is rather to be sure that states do not claim a greater interest because
they hold lightly the interests of other states. See infra text accompanying notes 329-31.
1. Full Faith and Credit
With respect to Full Faith and Credit, the only question ou

standing from the preceding section is whether the Clause appl
to sister-state laws as well as judgments. It does, simply as a mat
of positive law. The historical evidence suggests fairly clearly t
this was the original understanding,274 and even were it otherwi
the Supreme Court has consistently said so.275 The Court has al
read the Clause to embody an antidiscrimination norm in cases
volving attempts by states to close their courts to foreign law.2
The lesson of these cases is that states must allow their courts to
hear actions based on foreign law if they hear similar domestic ac-
tions - they cannot treat foreign actions differently simply because
they are foreign. The antidiscrimination reading is somewhat less
obvious when foreign rights are rejected based on conflicting local
law, rather than on jurisdictional limitations. That is, it seems that
lesser standards apply to a state's decision to substitute local rights
for foreign rights. A number of cases, notably Carroll v. Lanza277
and Nevada v. Hall,278 have held that "the Full Faith and Credit
Clause does not require a State to apply another State's law in vio-
lation of its own legitimate public policy."279 Of course, a state may
not refuse to hear a federal cause of action and instead apply its
own law; that is the lesson of Mondou and Testa. The Supreme
Court thus seems to realize that application of forum law subordi-
nates rights asserted under foreign law. The two lines of cases can
be distinguished by noting that while federal rights may never be
subordinated to state rights, states may sometimes favor their own
rights over those created by other states. Hall and Carroll do not
deny that foreign rights were rejected, but rather suggest that suffi-
cient justification existed.280 Under Hughes and First National
274. See Laycock, supra note 73, at 290-95.

275. See, e.g., Carroll, 349 U.S. at 411 ("A statute is a 'public act' within the meaning of
the Full Faith and Credit Clause.").
277. 349 U.S. 408 (1955).
278. 440 U.S. 410 (1979).
279. Hall, 440 U.S. at 422; see also Carroll, 349 U.S. at 412-14 (distinguishing between
denial of jurisdiction to hear suit based on foreign law and decision according to local law). I
suggest not that Full Faith and Credit requires this, but simply that it requires a reason for
local rights to defeat foreign rights - and that, having announced this reason, a state must
consistently follow it, regardless of whose ox is gored. (Thus while any particular conflicts
decision is likely to be acceptable, certain patterns of decisions will not be.)
280. Both cases featured injuries occurring in the forum state, and a simple territorial rule
explains these results quite easily. In fact, the notion of territorial sovereignty is clear in the
early Full Faith and Credit cases. See, e.g., Pacific Ins. Co. v. Industrial Accident Commn.,
306 U.S. 493, 504-05 (1939) ("Full faith and credit does not here enable one state to legislate
Bank, that justification must be something more than that the rights
are foreign.281 That is all I claim. The antidiscrimination reading is
thus at least implicit in the cases, even where rejection takes the
form of "applying local law" instead of rejecting the cause of action.
The more recent Allstate decision suggests that a state may re-
ject foreign rights in favor of its own on the basis of extraordinarily
slender justifications, which clearly do not embody a consistent pol-
icy.282 Basically this amounts to the principle that one state may
always reject another's law if it disagrees sufficiently to have en-
acted a different law. That is an absurd reading of Full Faith and
Credit; if it commands respect for sister-state law merely when the
states agree, it does nothing at all.283 Allstate is in tension with the
Hughes line of cases, and also with the Court's Privileges and Im-
munities jurisprudence.284 More bluntly, it is wrong.285 Why has
the antidiscrimination principle of Full Faith and Credit emerged
strongly with respect to judgments and jurisdictional limitations,
but remained only implicit with respect to conflicting local law? If
we take the jurisprudence at face value, the answer seems to be that
policy disagreements trump Full Faith and Credit; but this is a
wildly implausible reading, given the Clause's history and aspira-
tions - policy disagreements are obviously one of its chief con-
cerns.286 If it were the case that conflicting local law trumped
another state's law, the Court has never explained why it would not
for the other or to project its laws across state lines so as to preclude the other from prescrib-
ing for itself the legal consequences of acts within it."). The absence of similar territorial
reasoning in Hall (which, oddly enough, quotes the above sentence, see Hall, 440 U.S. at 423-
24) provoked a three-Justice dissent. See Hall, 440 U.S. at 428 (Blackmun, J., dissenting)
(contrasting majority opinion to territorial reasoning of lower court).
282. See Allstate Ins. Co. v. Hague, 449 U.S. 302, 313-20 (1981) (upholding application of
Minnesota law based on decedent's working in Minnesota, widow's after-acquired Minnesota
domicile, and Allstate's "presence" in Minnesota).
283. Cf. Kramer, supra note 215.
284. The tension with Privileges and Immunities arises because it seems unlikely that
Minnesota would have applied its law to benefit a Wisconsin domiciliary in similar straits.
Indeed, if the domiciliary status of one party is the only reason a state has for applying its
law, Due Process will forbid it from extending similar rights to out-of-staters, as Phillips
Petroleum holds. See Phillips Petroleum Co. v. Shutts, 472 U.S. 797, 815-22 (1985). But the
Court's Privileges and Immunities cases suggest that out-of-staters may not be denied the
rights granted locals merely on the basis of their foreign domicile. See supra note 273.
285. It is wrong because, were the contacts reversed, Minnesota would surely not have
held that Wisconsin rights prevailed with respect to an accident between Minnesotans occur-
ring in Minnesota, where the plaintiff moved to Wisconsin after the accident. Consequently,
Minnesota must have rejected the Wisconsin rights because they were foreign and thereby
violated Full Faith and Credit. See infra section VI.B.1.
286. See Wells v. Simonds Abrasive Co., 345 U.S. 514, 521 (1953) (Jackson, J., dissenting)
("The whole purpose and the only need for requiring full faith and credit to foreign law is
that it does differ from that of the forum.").
equally allow rejection of sister-state judgments. A more realistic

explanation is that the Court has simply been unable to come up
with a reading that allows it to handle conflicts between rights.287
The jurisdictional line of cases288 did not raise this issue. When
a state closes its courts to a foreign cause of action, it is simply re-
jecting a foreign right. It is not (superficially) resolving a conflict in
favor of a local right, and therefore condemning the door-closing
practice does not mandate subordination of local rights. Similarly,
interstate enforcement of judgments poses no question of conflict-
ing rights because our approach to judgments still resembles that of
Joseph Beale. The law of the state of rendition determines the ef-
fect of a judgment,289 and when a plaintiff sues to enforce a judg-
ment in another state, he is asking simply that rights created by the
law of the state of rendition be respected. There will seldom be
even a prima facie right that a defendant can invoke against the
enforcement of the judgment.
The situation with rights to recover, rights which have not been
reduced to judgment, used to be much the same: the plaintiff in-
voked the rights created by the territorially appropriate law, and
the defendant had available no contrary rights. But with the death
of vested rights, the picture changed dramatically. Instead of
Beale's no-conflict world, courts confronted cases in which the
plaintiff urged a right created by one law while the defendant ap-
pealed to a defense created by another. (Another way of putting
this is that it became unclear which state's law was the source of the
plaintiff's rights.290) Where rights conflict, or laws contend for ap-
plication, the meaning of Full Faith and Credit is much less clear,
and the temptation for courts is to hide behind a personal-jurisdic-
tion-style analysis that suggests the issue is one of choice, rather
than of conflict. I will later explain how a straightforward applica-
tion to conflicts may fulfil the aims of the Clause.291 What should
be already clear is that it cannot be ignored. It does require that
287. More charitably, the Court's retreat may reflect the realization that constructing a
hierarchy of interests is the legitimate prerogative of the states. But this does not mean that
Full Faith and Credit has no role to play. Its goal is to "guard the new political and economic
union against the disintegrating influence of provincialism in jurisprudence, but without ag-
grandizement of federal power at the expense of the states." Jackson, supra note 232, at 17.
The question is how to balance the federal and local interests, and leaving everything up to
the states is not the answer.

289. See, e.g., Matsushita Elec. Indus. Co. v. Epstein, 516 U.S. 367, 375-59, 386-87 (1996).
290. This is the more popular description. As discussed above, see supra section III.C, I
think it is more useful to talk in terms of conflicting rights.
291. See infra section VI.B.
states give full faith and credit to the laws - both statutory and
common - of sister states.292
2. Privileges and Immunities
Uncertainty also exists with respect to the scope of "privileges

and immunities." The counterpart to the (foolishly) literal reading
of the Full Faith and Credit Clause is the supposition that the Privi-
leges and Immunities Clause requires out-of-staters to be granted
all the rights of locals, including the right to vote and pay in-state
tuition for public schools. The Supreme Court, however, has not
similarly emasculated the Clause in the face of such an absurdly
expansive reading, but has suggested instead that it applies only to
"fundamental" rights.293 Some privileges granted to locals - such
as the elk-hunting license at issue in Baldwin - are not fundamen-
tal. Nor is the Clause absolute in its proscriptions. States may treat
out-of-staters differently if they have a "substantial" reason that is
"substantially advanced" by the discriminatory treatment.294
This restriction, by itself, is not a full explanation. The right to
vote is surely fundamental, but no court has suggested that it is one
of the relevant privileges and immunities. I think a satisfactory an-
swer is that the idea of discrete states presupposes the distinction
between members of the polity and outsiders. If all federal citizens
could vote in all state elections, we would no longer have politically
distinct states. This sort of discrimination is required for the Privi-
leges and Immunities Clause to have meaning; it cannot be a viola-
tion of the Clause. Relatedly, Ely suggests that if outsiders could
vote, they would be able to protect their interests, and there would
be no need for the Clause.295
The more serious question is how privileges and immunities re-
late to conflicts rules. Laycock casts this in terms of "choice-of-law
rules that prefer local litigants" and avoids the "fundamental
rights" limitation by noting that equal treatment in court is surely a
fundamental right.296 This is a tempting argument, but it fails be-
cause it operates at too high a level of generality. All of the dis-
criminations against outsiders that the Supreme Court has
292. See Jackson, supra note 232, at 12; Laycock, supra note 73, at 290-95.
293. See Baldwin v. Fish & Game Commn., 436 U.S. 371, 387-88 (1978).
294. See Supreme Court of New Hampshire v. Piper, 470 U.S. 274, 284 (1985). A mere
desire to benefit locals does not count as a substantial interest.
295. See Ely, supra note 73, at 190.
296. See Laycock, supra note 73, at 265-66. Currie similarly talks about "[t]he right of
access to courts, generally stated." See CURRIE, supra note 161, at 467 n.70.
approved as not implicating fundamental rights do involve the

question of equal treatment in the courts, and they all involve
choice-of-law rules that prefer locals. That is, when the out-of-
stater goes to court to get his elk-hunting permit, he is asking for a
right that local law gives to locals. When the court denies his re-
quest, it is withholding from him rights granted to locals; it is (in a
sense) refusing to apply local law to his claim. This is the paradig-
matic Privileges and Immunities discrimination, analytically identi-
cal to a determination that out-of-staters have no right to recover
for in-state batteries. If we are to give meaning to the "fundamen-
tal rights" restriction, we have to go case-by-case.
There is a sense, however, in which Laycock is right. Earlier I
distinguished between rules of scope, which determine the extent of
state-created rights, and conflicts rules, which prescribe which rights
shall prevail in a conflict. The Privileges and Immunities Clause
(unlike Full Faith and Credit, which applies only to conflicts rules)
governs both sorts of rules. With respect to rules of scope, it de-
mands that states extend to out-of-staters all the rights that they
extend to locals - or rather, all fundamental rights, unless there is
a substantial nondiscriminatory justification. With respect to con-
flicts rules, it demands that out-of-staters asserting rights not be
treated differently from locals asserting the same rights simply be-
cause they are out-of-staters.297 That is, if the state's conflicts rule
provides that a local right will prevail in a particular case when as-
serted by a local, that right must prevail when asserted in the same
case by an out-of-stater, unless there is some nondiscriminatory rea-
son why it should not. Equality of treatment under conflicts rules is
clearly fundamental, and Laycock is correct to say so.
If equality with respect to conflicts rules is to be meaningful,
however, there must be equality with respect to rules of scope. If a
state can limit the scope of a right to locals, then a nondiscrimina-
tory conflicts rule will do nothing to remedy the discrimination.
The question then is whether states, by restricting the scope of
nonfundamental rights, may affect the outcome of the classic con-
flicts cases. It seems unlikely. These cases are usually ones in
which a local would be able to avoid liability as a defendant, or
recover damages as a plaintiff, under local law; and for some reason
297. This is, in a sense, just another way of saying that equal treatment in the courts is a
fundamental right. I think it is somewhat clearer, though, to distinguish between rules of
scope and conflicts rules.
these rights seem to be regarded as fundamental.298 Perhaps there

is something about the adversarial nature of civil litigation that ele-
vates the interests involved, such that the rights an individual pos-
sesses against other individuals are more likely to be deemed
fundamental than his rights to governmental largesse. Or perhaps
the interests protected against infringement by other individuals are
typically more important than the interest in a hunting license.
Neither of these rationales is entirely clear, but I have seen no seri-
ous suggestions that the rights at issue in conflicts cases will fre-
quently not be fundamental for Privileges and Immunities
purposes.299
The minimalist reading I advocate can thus be applied to the
classic conflicts cases. The Supreme Court, which accepts this read-
ing,300 seems to think that it has no serious consequences. The next
section will show, however, that such consequences exist: this read-
ing is in irreconcilable tension with Allstate, with interest analysis,
and with the personal-jurisdiction-style approach to conflicts more
generally.
VI. TOWARD A CONSTITUTIONAL JURISPRUDENCE

OF CONFLICTS
This Part aims to show that even a weak reading of the Consti-
tution imposes real limits on state conflicts rules and rules of
scope.30' The method will be to start with Currie's conception of
governmental interests and then to show how the Constitution
reconfigures his approach. I start with Currie's vision not because it
seems accurate, or even plausible, but because it exemplifies the
excesses the Constitution reins in. Brilmayer is undoubtedly right
that it is dangerous to impute to legislatures policies they seem ex-
plicitly to disavow. States may refrain from pressing their legisla-
298. See Ely, supra note 73, at 182-83 ("Baldwin or no Baldwin, it is not likely to be
suggested that [decisions about liability] implicate rights so unimportant that they can be
dismissed as beyond the coverage of the Privileges and Immunities Clause.").
299. The right of access to courts is acknowledged as fundamental. See McKnett v. St.
Louis & S.F. Ry. Co., 292 U.S. 230, 233 (1934). The refusal to grant the benefits of local law
is not precisely a denial of access, but surely the right of access is meaningless if, having once
gotten into court, the out-of-stater then faces discrimination as to substantive rights.
301. As mentioned earlier, I think that "choice-of-law rules" are misleadingly named.
What they actually do is describe when foreign-created rights will be respected. The appro-
priate way to invoke a foreign-created right is not to sue under forum law and have the forum
court decide which law applies to the claim; it is to sue under foreign law. If the forum court
then decides that local law "applies," it has decided either that the law the plaintiff pleads
grants him no right (a scope decision) or that the foreign right yields to whatever local de-
fense the defendant invokes (a conflicts decision).
tive jurisdiction to the limit; they may also be more solicitous of

sister-state domiciliaries than classic interest analysis supposes.
Currie, of course, recognized both of these facts.302 His aim was to
show what states could do, and what it was appropriate for judges
to do in the absence of legislative guidance.
That is also my aim, though my conclusions are quite different.
The import of this Part is that states cannot do much of what Currie
thought they could. In particular, the Constitution prevents them
from following the discriminatory policies of interest analysis. One
way to express this conclusion would be to say that state policies are
not discriminatory. Because federal interests are local interests
everywhere,303 the antidiscrimination policies of the Privileges and
Immunities and Full Faith and Credit Clauses are the true policies
of the states, whatever their legislatures may in fact desire.304
It will be clearer, however, to talk in terms of state interests ac-
cording to the Currie model and to consider the Constitution as an
external constraint. I will also largely ignore statute-specific rules
of scope. It is important to ask if a particular right is intended to be
granted to those in the particular situation of the party before the
court, but at the level of generality at which this section works, that
inquiry can be sensibly performed only in terms of state interests.
One more introductory point remains. The reason I find the
Constitution so effective in constraining choice-of-law rules is that I
approach "choice of law" from the perspective of conflicts. That is,
I see a "choice-of-law" question as a question of which right will
prevail. State conflicts rules articulate a hierarchy of rights by es-
tablishing factors that determine which right prevails. The
Constitution limits the acceptable factors. In particular, Full Faith
and Credit prevents consideration of the fact that a particular right
is a local one, and Privileges and Immunities similarly prevents the
fact that a party is (or is not) a forum domiciliary from having
weight. Beyond these two restrictions, states may construct what
hierarchies they will. As we shall see, however, the restrictions
suffice.

303. See Mondou v. N.Y., New Haven & Hartford R.R. Co., 223 U.S. 1, 55-57 (1912).
304. See John K. Beach, Uniform Interstate Enforcement of Vested Rights, 27 YALE L.J.
656, 665 (1917) ("Surely the Constitution expresses the real and controlling 'policy' of the
states in this regard."). In a sense, the Constitution thus provides the objective state interests
whose absence Brilmayer believes dooms Currie's theory.
A. Rules of Scope and the Constitution: Two Myths

Currie's analysis is supposed to show that some cases create no
real problems, that they pose no issue of choice of law - namely,
false conflicts and unprovided-for cases. In the last analysis, my as-
piration is much the same. Like Currie, I do not think that conflicts
theory has much to say about true conflicts - situations in which
both states claim priority for the rights created by their law. As
long as the rule that prioritizes local rights is constitutionally sound,
there is no reason why the forum should not apply it. My point is
rather that this will very seldom happen. I do not mean that there
will be few cases in which rights conflict - I actually find more
conflicts than does Currie - but it will be a rare case in which
states disagree on which right should prevail. Explaining why d
greement will be rare requires an analysis of the limitations th
Constitution places on state conflicts rules, but that is the tas
the next section. This section seeks to illustrate the application
the Constitution to the permissible scope of state laws. The iss
are distinct: scope questions are about whether a given law cre
a right or not, conflicts questions are about when one right prev
over another. This section will show that what interest anal
terms false conflicts and unprovided-for cases are actually t
conflicts.305
1. The Myth of the Unprovided-for Case

The title of this subsection is the same as the title of an article b
Larry Kramer.306 It was that article that started me thinking ab
conflicts in the way developed here, and this section unsurprisin
shows his influence. My analysis does, however, differ in some
portant respects, most notably my understanding of the effect of t
Privileges and Immunities Clause.
Recall that Currie found an unprovided-for case when
"[n]either state cares what happens."307 This situation obtains, gen-
erally, when the law of the defendant's domicile permits recovery,
the law of the plaintiff's domicile bars it, and the tort occurs in the
nonrecovery (plaintiff's) state.308 In such a case, the plaintiff's state
305. It is doubtless not obvious how this result will be helpful. The discovery of false
conflicts is generally considered the great achievement of interest analysis; rejecting this in-
sight does not seem like an advance. I will argue, however, that an abundance of conflicts is
not a bad thing. See infra section VI.B.
306. See Kramer, Myth, supra note 81.
308. Lea Brilmayer suggests that only domiciliary factors were generally relevant to
Currie, see BRILMAYER, supra note 8, ? 2.1.2, at 65-66, and hence that an unprovided-for case
has no interest in generating a recovery for the plaintiff; its domes-

tic law does not do so. Nor does the defendant's state have an in-
terest in granting recovery for an out-of-stater, against its own
domiciliary, for a tort that occurred outside its borders. If state in-
terests are thought of as interests in the application of state law, it
seems that the lack of an interested state creates a troubling lacuna:
no state wants its law applied. Interest analysis thus seems to sug-
gest that no law applies to the case, and this is a prospect from
which conventional legal thinking recoils.
Two basic insights drive Kramer's revisionary approach. First,
he conceives of interest analysis as simply a method of determining
when positive law confers rights on the parties.309 That is, he sees
the detection of interests as a matter of rules of scope.310 The de-
termination of interest must thus be made with respect to each
claim of right.311 For example, a decision that California tort law
gives the plaintiff a right to recover does not necessarily mean that
defenses created by California law should be available to the de-
fendant. To say that California law "applies" in this sense is not to
say that the transaction is governed by California law. The case is
not decided as though it were a purely Californian case, as though
arises whenever the plaintiff's home law bars recovery and the defendant's permits it, see id.
? 2.1.2, at 63. This is somewhat of an oversimplification, as she later acknowledges, see id.
? 2.1.2, at 67, and only true with respect to Currie's analysis of married women's contracts,
see CURRIE, supra note 102, at 108. If a tort occurs in the pro-recovery state, Currie found an
interest: the state "may incur responsibility to the person injured in the state." CURRIE,
supra note 76, at 148; see also id. at 157 (constructing table of permutations, finding different
interests based on territorial factors); id. at 149 ("California's interest in the injured plaintiff
is based solely on the fact that he was injured here, but that has been regarded as a substan-
tial basis."); id. at 150-51 ("[T]he fact that the injury occurred in California suggests
though it does not necessarily follow - that California may become very deeply con-
cerned."). In his analysis of married women's contracts, the starting point for Brilmayer's
discussion, Currie in fact found no unprovided-for cases. See CURRIE, supra note 102, at 95
(evaluating effect of application of law of the place of contracting on state interests, and
finding an interest in each permutation).
This results from his rather complex articulation of the interests at stake. It is not unfair
to say that Currie tended to find interests that produced congenial results. Brilmayer sug-
gests that different, equally plausible interests may be constructed, see BRILMAYER, supra
note 8, ? 2.1.2, at 61-62, and this is quite true. To generalize interest analysis sufficiently that
it becomes determinate, rather than retaining enough flexibility to produce whatever result
the judge wishes to reach, it is probably necessary to distort Currie a bit. Tackling Currie on
his own terms is like having a fistfight with a fog. Attributing to him a focus on domiciliary
factors is one way to do so. Kramer produces a slightly more charitable generalization, sup-
posing that states generally have interests in regulating conduct either occurring within their
borders or affecting their domiciliaries. See Kramer, Myth, supra note 81, at 1065.
309. See Kramer, Myth, supra note 81, at 1064.
310. It may be that at this point my reading of Kramer is too strongly colored by my own
perspective; in later work he casts the question in terms of which law applies. See supra note
154 and accompanying text.
311. See Kramer, Myth, supra note 81, at 1051-55. This is of course the procedure fol-
lowed in ordinary cases.
only California law (and the entirety of California law) determines

the rights of the parties. This is the same lesson I urged that we
take from the coexistence of state and federal rights; it is also the
reason I claim that interest analysis does not in fact select the "ap-
plicable law."312
Kramer's second, and related, insight is that a determination
that no state is interested means that no state's law grants any
rights.313 The plaintiff loses; he has failed to state a claim on which
relief may be granted.
What do these insights mean for "unprovided-for" cases? Such
cases occur only when the plaintiff's home law does not permit re-
covery in a purely domestic case. There are two ways in which re-
covery might be restricted. First, on the facts of the plaintiff's case,
the law may create a defense to the cause of action. Kramer be-
lieves that this sort of case is not truly unprovided-for because the
plaintiff's state has no interest in extending the defense to a non-
domiciliary defendant. Therefore, the result in such cases will be
that the plaintiff recovers under his home state's law even though
he could not recover against a codomiciliary.
Kramer has two examples from this category. Having surely
taxed the reader's patience already, I will consider only one, a vari-
ant of Grant v. McAuliffe.314 Both Arizona and California have
wrongful death actions, but Arizona abates its action upon the
death of the tortfeasor. The unprovided-for case arises when an
Arizona plaintiff sues the estate of a California tortfeasor for an
accident that occurred in Arizona. The Arizona defense would or-
dinarily apply, but Arizona has no interest in extending it to a Cali
fornian. Thus, Kramer finds, the plaintiff can recover under
Arizona law.
But this is shockingly discriminatory, and Kramer subsequently

recants. The Privileges and Immunities Clause, he says, will not al-
low states to decompose their "no recovery" rule into rights and
defenses and deny the defenses to out-of-staters. If the plaintiff's
home state does not permit recovery against its own domiciliaries, it
cannot tinker with its law to disadvantage out-of-staters. So in the
Grant variant, the plaintiff loses.
312. See supra section IV.B. One thing Kramer's article thus shows is how far astray the
notion of choosing an applicable law led Currie. Oddly, Kramer at times seems prone to the
same mistake. See supra text accompanying note 154.
313. See Kramer, Myth, supra note 81, at 1062-63.
314. 264 P.2d 944 (Cal. 1953). Grant is the occasion for another of Currie's extended
analyses of possible permutations of contacts. See CURRIE, supra note 76.
The second way in which the plaintiff's home law might "not
permit recovery" is by simply not conferring a right to recover at all
(rather than creating a right but subordinating it to a defense).315
Here, Kramer finds, the "unprovided-for" aspect of the case simply
means that the plaintiff has no right.316 Kramer's example of this
sort of case is Erwin v. Thomas,317 in which a Washington resident
was injured by an Oregon resident in Washington. The victim's
wife sued to recover for loss of consortium, an action recognized by
Oregon but not by Washington law. Oregon law gives no right be-
cause it is not interested in allowing recovery, and Washington law
gives no right because it sees no injury. Thus, once again, the plain-
tiff loses. In sum, unprovided-for cases are simply ones in which the
plaintiff cannot state a claim under his own law.
All this is dead on, as far as it goes - the law of the plaintiff's
state will not help him in an unprovided-for case. But what about
the defendant's home law? The defendant's home law, remember,
permits recovery. Interest analysis says that it confers no right, be-
cause the defendant's home state has no interest in allowing recov-
ery based solely on the fact that the defendant is a local.318 But
perhaps Privileges and Immunities has something to say here as
well, and in fact, Kramer thinks that it does. In his analysis of
Erwin, he initially concludes that the plaintiff loses: neither state's
law gives her a right to recover. When Privileges and Immunities
enters the picture, however, Kramer finds that Oregon cannot jus-
tify withholding the benefits of its law from a nonresident plaintiff
when it would let an Oregonian recover.319 Consequently, the
plaintiff can recover under Oregon law - at least, in an Oregon
court.
315. Distinguishing between these two classes of cases may be difficult. It is not impossi
ble, however, and is sometimes easy. If a state does not permit suit for loss of consortium, it
law clearly confers no right. If it does have a wrongful death action but abates it on the death
of the tortfeasor, its law confers a right but subordinates it to the defense. (These example
are drawn from the facts of Grant and Erwin v. Thomas, 506 P.2d 494 (Or. 1973), which
Kramer considers in Myth, supra note 81, at 1048-56 (Grant), 1060-63 (Erwin).
317. 506 P.2d 494 (Or. 1973).
318. The tort, in these examples, takes place in the plaintiff's state. See, e.g., Erwin, 506
P.2d at 495.
319. See Kramer, Myth, supra note 81, at 1073. This may not be the correct reading of the
Clause; it is at least arguable that it applies only to treatment of out-of-staters with respect
in-state occurrences. See Toomer v. Witsell, 334 U.S. 385, 395 (1948) ("It was designed t
insure to a citizen of State A who ventures into State B the same privileges which the citizen
of State B enjoy."). Whether the Clause is triggered when a State A citizen "ventures int
State B" to litigate an out-of-state transaction is not entirely clear.
If this is so, why should the result be different in the Grant vari-
ation Kramer considers? Why does Privileges and Immunities not
likewise compel California to extend the benefit of its law to the
Arizona plaintiff, entitling him to recover under California law? It
presumably would were the plaintiff a Californian320 - in that case,
both parties would be from California and the case would be a
"false conflict" because Arizona would have no interest.321 Thus
the withholding of the right to recover under California law do
not seem significantly different from the discrimination forbidd
to Oregon.
Kramer's reasoning here is hard for me to discern. It seems tha
the location of the forum makes the difference. Currie's unpro-
vided-for variant of Grant has the suit occurring in California, b
Kramer's Privileges and Immunities analysis considers an Arizon
court.322 He further suggests that courts may deny their own res
dents the benefits of sister-state law without violating the Clau
Thus while an Oregon court cannot deny a Washingtonian the ben
efits of Oregon law, a Washington court could. Similarly, a
Arizona court can deny an Arizonan the benefits of California law
This reasoning is not entirely satisfactory. The Privileges and
Immunities Clause is partly about the permissible actions of stat
courts, but it is also about the permissible content of state laws.
means, in particular, that as a matter of positive law, California
must extend to Arizonans the rights it extends to its own domiciliar-
ies.323 Thus when the Arizona court refuses to allow the plaintiff
rely on California law, it rejects a California right. After we use t
Privileges and Immunities Clause to reconfigure the scope of the
state laws, we find that both states are interested, in the sense th
both attempt to confer rights.
That leads to the main point of this section. The determination
that the Erwin plaintiff has stated a claim under Oregon law is just a
matter of scope analysis. The question remains whether the d
320. Whether it would or not is the crucial question for Privileges and Immunities, an
will consider it in more detail later. See infra part VI.B.1. At this point, we may simply not
that if it followed the prescriptions of interest analysis, it would. As a matter of histor
fact, of course, California did apply its law, see Grant v. McAuliffe, 264 P.2d 944, 949 (C
1953), but I am considering how interest analysis operates.
321. The presence of an Arizona interest will not prove essential to my analysis. With
law directed to primary conduct, Arizona presumably has an interest in deterring dangerou
activity within its borders. The abatement of a tort suit upon the death of the tortfeasor h
only marginal effect on primary conduct, however, so the claim that Arizona has no intere
in applying that rule to two Californians is at least plausible.
322. See Kramer, Myth, supra note 81, at 1073.
323. Subject, of course, to the qualifications noted above. See supra section V.B.2.
fendant has a defense available. Of course he does - Washington

law recognizes no cause of action, and thus privileges the conduct
complained of. Washington obviously extends this right to locals
acting in Washington; following Kramer's analysis, Privileges and
Immunities requires it to extend equally to Oregonians acting there.
So while the plaintiff has a claim under Oregon law, recovery is not
a foregone conclusion - the defendant has a defense under Wash-
ington law. That is as far as rules of scope take us.
This should be a startling conclusion. It may still be that
Kramer gets the results right - though eyebrows will raise at the
suggestion that what should happen in "unprovided-for" cases is
that the plaintiff should win if and only if he sues in the defendant's
home court, a result troublingly similar to the overstrong reading of
Full Faith and Credit. My point here is simply that these cases are
not easy to resolve, even after Kramer's reworking; the outcome is
not as clear as his optimistic assertions.324 He is right that they are
not, in fact, unprovided-for, but he does not go all the way: they are
actually true conflicts.325
2. The Myth of the False Conflict
The preceding section concluded that the Constitution turns

unprovided-for cases into true conflicts. The conclusion of this sec-
tion, with respect to false conflicts, will be the same.
Take as a first example the actual facts of Grant v. McAuliffe: a
collision in Arizona between two California domiciliaries. In this
case, Currie said, "Arizona had no conceivable interest in the appli
cation of Arizona law to the case."326 This is true enough, accord
ing to Currie's construction of interests. But again, the Privileg
and Immunities Clause will change things. What happens when t
California defendant (that is, his estate) invokes the Arizona abate
ment rule? The way to perform this analysis, as intimated above,
to ask whether Arizona would assert an interest if the party aski
for the benefit of her law were a domiciliary. The answer is yes
according to Currie's analysis; with a California plaintiff and an Ar
zona defendant, the case is a true conflict, where both states are

325. Perhaps these should be called "reverse true conflicts" since they feature the un
sual situation in which each state is asserting an interest in disadvantaging its domiciliar
(This is the reason Kramer believes that the plaintiff wins by suing in the defendant's ho
court.)
interested. If Arizona grants this right to its own domiciliaries,

must it not offer it on equal terms to out-of-staters?
Kramer suggests not; he argues that in such a case Arizona may
defer to California's interests by withholding the benefits of Ari-
zona law. Such deference, he claims, will reduce interstate friction,
and thus better serve the aims of the Privileges and Immunities
Clause.327 But interstate friction is not clearly the sole target of the
Clause. By its words, after all, it grants rights to individuals, not to
their states, and the Supreme Court has never suggested that states
may waive the rights of their citizens to the privileges and immuni-
ties of other states' laws. If the Clause aims to promote national
unity, conceptually as well as instrumentally, it seems likely that the
creation of a class of outsiders with fewer rights against each other
than against local citizens offends the principle of equality of indi-
viduals within states.
Admittedly, the facts of Grant v. McAuliffe lend themselves to

the proposition that rights created by the law of the common domi-
cile should have priority, primarily because the law at issue has so
little effect on primary conduct. The idea that a Californian can
invoke such an Arizona rule against another Californian is odd; it is
not clear why Arizona would want to make the law available to
Californians inter se, nor why the Constitution should require it to.
And I do not mean to suggest that the Constitution compels the
application of Arizona law. After all, the California plaintiff can
surely point to a California law that gives him a right to recover, so
the result would be at most a true conflict. Arizona may be able
constitutionally to adopt a conflicts rule deferring to California's
regulation of its domiciliaries, and if it can do that, it can probably
also simply withhold rights with such a marginal relation to primary
conduct. I suspect, though, that the Constitution prevents Arizona
from declaring as a general matter that Californians, inter se, do not
have the rights of Arizonans.
Suppose that instead the law focuses on primary conduct; sup-
pose Arizona has a cause of action that California law lacks - say,
for intentional infliction of emotional distress. There is something
plausibly wrong with a legal regime under which a Californian in
Arizona does not enjoy the protections that Arizonans do, so long
as the tortfeasor is also a Californian. Finally, suppose that the law
is a speed limit - clearly directed to primary conduct. In an acci-
dent between two Californians, it seems intolerably odd that one
would not be able to defend against an allegation of negligence with

the claim that he had complied with the Arizona speed limit, re-
gardless of what California's is.
These examples do not show that situs rights should always de-
feat those created by the law of the common domicile - it makes
sense only in some cases - but rather that d6peSage may be most
sensible. California rights prevail with respect to some issues (such
as the abatement of the cause of action), and Arizona rights with
respect to others (such as the applicable speed limit). This is com-
mon in state-federal conflicts, and we have seen no reason why mul-
tistate conflicts should treat it as anomalous.
The larger point thus is not that all false conflicts are necessarily
true conflicts, but that some must be. Deference to the policies of
the state of common domicile may be a sufficient nondiscriminatory
reason to withhold local rights not affecting primary conduct
though I think this will be quite a small set. But Currie's concep-
tion of the category of false conflicts goes further, because he be-
lieves that states generally have no interest in granting rights to
recover, or defenses against liability, to out-of-staters.328 This is the
discrimination I claim Privileges and Immunities blocks. "False
conflicts" that are false simply because the situs state has no inter-
est in compensating or defending out-of-staters, where it does have
such interests with respect to locals, are made true conflicts by the
Constitution.
B. Conflicts Rules and the Constitution
Thus far I have suggested that, taking the Constitution seriously

and thinking in terms of conflicting rights, unprovided-for cases and
false conflicts are often true conflicts. This is not an auspicious
start; it undoes most of the advances of interest analysis. This sec-
tion will take some steps forward by arguing that the prevalence of
true conflicts is nothing to worry about. Conflicts arise because
rules of scope tend to be broad enough that more than one state
will often create rights with respect to a particular transaction.
Broad rules of scope are not the problem, however; we have seen
that the Constitution works primarily to enlarge, not to constrict,
the scope of state-created rights. The problem lies rather in the
discriminatory manner in which conflicts are resolved. If states are
328. Generalizing about Currie's approach is difficult, and this characterization may not
be entirely fair. Currie did, after all, suggest that a state may have an interest in allowing
recovery to out-of-staters so that they can pay their in-state hospital bills. See CURRIE, supra
note 76, at 145 n.64.
prevented from discriminating, they will usually agree on which

rights should prevail; where they do not, they act within their au-
thority by resolving conflicts according to their own rules.
Whatever theory is invoked, courts resolving these conflicts do
hierarchize rights: they recognize foreign rights or refuse to do so.
Because these are legal decisions, there must be a legal rationale -
a rule specifying which rights prevail and which must yield, what I
have called a conflicts rule. The range of permissible conflicts rules
is not infinite, however; the Constitution imposes constraints.
1. Conflicts Rules and Law: Full Faith and Credit
Full Faith and Credit requires that the conflicts rule a state
adopts not disfavor foreign rights simply because they are foreign.
Whatever the conflicts rule may be, it may not provide that local
rights defeat foreign rights simply by reason of their origin. The
reason must instead be neutral, in that it cannot be conditioned on
the origin of the rights. (I will describe this as the requirement that
the state must assert a greater interest in the suit; the characteriza-
tion is useful but not essential.)
Return to the example of luckless driver Al. Al collides with
George in Alabama and with Lou in Louisiana. In both cases, Al
prefers Alabama law; Alabama, if it is interested in the financial
welfare of its domiciliaries, and certainly if it adheres to interest
analysis, might thus want to draw both cases within its legislative
jurisdiction. But it cannot.
Al v. George and Lou v. Al are what I have called mirror-image
cases, essentially identical (both are car accidents, and the law at
issue in both cases determines the damages available), with the ex-
ception that the relevant contacts have been switched.329 From the
perspective of Alabama, Al v. George has a local plaintiff, a foreign
defendant, and a local accident. Lou v. Al has a foreign plaintiff, a
local defendant, and a foreign accident. Any reason that can be
given in favor of Alabama rights in Al v. George can be given for
Louisiana rights in Lou v. Al, and therefore no conflicts rule that is
329. See supra text accompanying note 226. I use here the contacts that Currie sets out as
potentially relevant, though I omit the location of the forum. See CURRIE, supra note 76, at
141. It should be emphasized that these are not the only contacts a conflicts rule can consider
- states have the freedom to orient their rules around whatever contacts they desire. De-
pending on the contacts deemed relevant, Al v. George and Lou v. Al might not be mirror
images as I have constructed them. For example, it might be relevant whether the parties
know each other and were traveling to the same destination in separate cars. It should be
nonetheless clear that mirror-image cases can be constructed, using whatever contacts are
taken as relevant.
neutral in the required sense can provide that Alabama rights pre-
vail in both cases. Alabama can adopt a conflicts rule privileging
Alabama rights in one case; that amounts to the assertion that its
interest in that case is greater. But this assertion immediately im-
plies that its interest is inferior in the other case, because the other
case is a mirror image.330
Thus Full Faith and Credit will have a real effect on the develop-
ment of state conflicts rules. If states abide by the principle of relin-
quishing the mirror images of cases in which they hold that their
rights prevail, the system will develop appropriately, either by stat-
ute or by common law. But state compliance need not be willing.
If Alabama decides that its rights prevail in Al v. George and then
reaches the same conclusion in Lou v. Al, Lou's Full Faith and
Credit argument has already been made for him. He can demon-
strate "on some rational basis" (to use the short-lived Alaska Pack-
ers formulation) that Alabama's interest is inferior, for Alabama
has said as much. Even if no mirror-image cases exist yet, states are
unlikely to cheat, because an insincere assertion of greater interest
will come back to haunt them when the mirror-image case arises.331
This use of Full Faith and Credit does not require the federal
judiciary to interfere with a state's establishment of priorities.332 If
Alabama's conflicts rule provides that Alabama rights prevail in Al
v. George, a federal court directing that Louisiana rights prevail in
Lou v. Al has not imposed its conception of interests on Alabama
or even deemed Alabama's interest inferior. It has simply listened
to Alabama's analysis of what makes an interest superior and taken
the state at its word. If a state asserts legislative jurisdiction over a
case and its mirror image, it has violated the Full Faith and Credit
Clause. What's sauce for the goose must be sauce for the gander.
330. Cf Jackson, supra note 232, at 25-26 ("'It will not do to decide the same question
one way between one set of litigants and the opposite way between another.... If a case was
decided against me yesterday when I was defendant, I shall look for the same judgment today
if I am plaintiff."') (quoting CARDOZO, supra note 227, at 33 (alteration in original) (quoting
WILLIAM GOLDSMITH MILLER, THE DATA OF JURISPRUDENCE 335 (1903) (internal quotation
marks omitted))).
331. With this particular example, Alabama would probably opt for a rule privileging its
rights in Al v. George and subordinating them in Lou v. Al, primarily because the accident in
Al v. George occurred in Alabama. For the same reason, Louisiana's conflicts rules are likely
to privilege its rights in Lou v. Al. A general preference for territorial sovereignty will lead
to interstate agreement about whose rights should prevail. This is nice, but not essential.
332. I speak of the federal judiciary because state courts have played a role in creating
discriminatory conflicts rules. Of course, the constitutional limits I identify bind state courts
as well.
2. Conflicts Rules and Domicile: Privileges and Immunities
The Privileges and Immunities Clause, I have said, operates at

two levels. As a restriction on rules of scope, it requires states to
extend to out-of-staters the same rights they do to locals. As a re-
striction on conflicts rules, it requires states to grant those rights the
same force. These conditions may seem modest, but they have sub-
stantial effect.
Return to Laycock's example of hapless friends Mary and Del.

They are involved in two accidents together; in each accident a dif-
ferent friend is driving. Assume further that the accidents both oc-
cur in the same state and are substantively identical.333 Del sues
Mary in Delaware, and Mary counterclaims; Delaware has a guest
statute, and Maryland does not. If Delaware is interested simply in
helping its domiciliaries, it will want to use the guest statute to bar
Mary's claim but not Del's. This is precisely what interest analysis
directs. But it violates the Privileges and Immunities Clause.
There are two ways in which Delaware can regulate the effect of
its guest statute: rules of scope, and conflicts rules. Neither offers a
way to distinguish between the two claims. If the scope of the guest
statute extends to Del, it must also to Mary, because the only differ-
ence between the two claims is the domicile of the parties. For the
same reason, if the guest statute in Del's hands defeats the right to
recover, it must also do so in Mary's hands.
This example may suggest a linkage between rules of scope and
conflicts rules, and it bears repeating that they are distinct. Con-
sider again Erwin v. Thomas. An Oregon domiciliary injures a
Washington husband in Washington, and his spouse sues for loss of
consortium, an action existing under Oregon law but not (for wives)
under Washington law. Oregon may not want to grant recovery to
the Washington plaintiff, and I will show that it need not, contra
Kramer,334 but it will probably have to employ a conflicts rule to
achieve this end.
With regard to rules of scope, the Privileges and Immunities

Clause will probably require Oregon to extend the loss-of-
consortium cause of action to the Washington plaintiff because it
would likely grant the right to a similarly situated local. If no other
law interferes, a state will often give its law extraterritorial force for
333. Laycock does not specify these conditions, see Laycock, supra note 73, at 276, but
they are essential to my analysis.
334. See Kramer, Myth, supra note 81, at 1073; see also supra text accompanying note
319.
transactions between two of its domiciliaries.335 Consequently, the

Privileges and Immunities Clause will require the state to grant
rights to similarly situated nondomiciliaries. This is how the Clause
creates the true conflict. But granting the right and prioritizing it
over a competing right are two different things.336 The defendant
can appeal to Washington law for a defense - Washington does not
recognize the cause of action - and to allow this defense to defeat
the Oregon right, the Oregon court need only apply the conflicts
rule that in tort cases, rights (and defenses) created by the law of
the place of the wrong have priority over other rights.337 The Privi-
leges and Immunities Clause, as a restriction on conflicts rules, does
not bar this rule, for the rule does not disfavor rights because the
party invoking them is a foreign domiciliary.
As a last example, let us return to the example of Grant v.
McAuliffe, a collision between two California domiciliaries occur-
ring in Arizona. Interest analysis identifies this as a false conflict,
but we have seen that it is actually a true one, in that Arizona can-
not, as a general matter, withhold the benefits of its local law. The
scope and conflict requirements of the Privileges and Immunities
Clause imply that, if Arizona provides that its rights will defeat for-
eign rights in a suit over a local tort involving one of its domiciliar-
ies, it must also provide that its rights defeat foreign rights in a
similar suit where neither party is a domiciliary. Otherwise, its con-
flicts rule favors rights asserted by a local over rights asserted by a
similarly situated out-of-stater. The result seems to be that, in false
conflicts, the Constitution requires that out-of-staters be able to in-
voke local law. This makes sense in some circumstances; for exam-
ple, it seems natural that a Californian must be able to defend
against an allegation of negligence by showing that he complied
with the rules of the road applicable to Arizonans.
In comparison, the idea that Arizona must provide that its
abatement rule controls a suit between two Californians is harder to
335. Cf American Banana Co. v. United Fruit Co., 213 U.S. 347, 355-56 (1909) (Holmes,
J.) (pointing out that national states - i.e., countries - often seek to apply their own law,
even to acts within other jurisdictions).
336. Again, conflicts rules are not rules of scope. The Privileges and Immunities Clause
operates as both a rule of scope and a constraint on permissible conflicts rules. As a rule of
scope, it requires state laws to extend rights to nondomiciliaries on the same terms as it
extends them to locals. As a conflicts rule constraint, it prevents states from applying rules
that honor those rights helpful to local domiciliaries.
337. For this to work, it must be the case that Washington grants the defendant a right. A
rational attempt to promote state interests would lead it to do so, since otherwise it loses the
ability to control transactions taking place within its borders. That is, if it grants no right to
the Oregon domiciliary, it cannot grant rights to its own domiciliaries in similar cases without
violating the Privileges and Immunities Clause.
swallow; Currie dismisses it as patently absurd.338 The question

must be whether withholding Arizona law serves a substantial non-
discriminatory interest; if it does, Privileges and Immunities will be
no bar. Kramer suggests that deference to California constitutes
such an interest. Although I may not agree as a general matter, the
interest certainly will be clearer where the rule in question has a
minimal effect on primary conduct. This approach tallies with my
intuitions about Grant, but is still not a complete solution, since it
does not explain why California could adopt a conflicts rule provid-
ing that the California right to recovery prevails over the Arizona
abatement defense in such a case. The discrimination here - al-
lowing California rights asserted against a California domiciliary
prevail only when asserted by another Californian - could possi
be justified by similar deference to Arizona. If so, this amounts
the conclusion that this conflicts rule, requiring the law of comm
domicile to prevail in common domicile cases, does not violate Pr
ileges and Immunities - at least where rules with minimal effect
primary conduct are at issue. I do not think the issue is free
doubt, but the conclusion is certainly possible.339
The import of Privileges and Immunities is thus the followi
If a state's rules provide that its rights prevail in a particular ca
those rights must also prevail when the parties' domiciles
switched.340 In many situations, the requirement is more gene
the same rights must prevail regardless of the parties' domicile
338. See CURRIE, supra note 76, at 160-61.

339. If Privileges and Immunities did forbid this "deference," the state of common dom
cile could still regulate. It could attach what legal consequences it wanted to the transac
and re-allocate loss or even impose criminal sanctions - subject to the possible restric
of the Commerce Clause, as Akhil Amar pointed out to me. The significance of domici
precisely that a state can legitimately claim to apply its laws to its domiciliaries' actio
abroad - even if it has to wait until they return home. See Brilmayer, supra note 71, at 12
1303. But see Seth F. Kreimer, The Law of Choice and Choice of Law: Abortion, the Rig
Travel, and Extraterritorial Regulation in American Federalism, 67 N.Y.U. L. REV. 451 (
(arguing that extraterritorial regulation of domiciliaries is unconstitutional). Kreimer's an
ysis seems at odds with conflicts law. If a state can constitutionally apply its law to extrate
torial actions of foreigners who commit torts against its domiciliaries, it would be surprisi
it were less able to regulate its domiciliaries. The question is complex; in some cases (p
haps with regard to gambling) regulation of extraterritorial acts might look too much
protectionism to survive Commerce Clause scrutiny. The majority of scholarly opin
seems to support state authority over domiciliaries' extraterritorial acts. See, e.g., Ger
supra note 271, at 907 n.94; Donald Regan, Siamese Essays: (I) CTS Corp. v. Dyna
Corp. of America and Dormant Commerce Clause Doctrine; (II) Extraterritorial State Le
lation, 85 MICH. L. REV. 1865, 1907 (1987); William Van Alstyne, Closing the Circle of Con
tutional Review from Griswold v. Connecticut to Roe v. Wade: An Outline of a Dec
Merely Overruling Roe, 1989 DUKE L.J. 1677, 1684 n.27.
340. Thus the full contact-switching mirror-image test works only to detect Full Faith
Credit violations. The test for Privileges and Immunities violations is to switch only domi
ary factors and ask if the same rights prevail. If not, there is a violation. Since the Privil
and Immunities constraint thus means that switching domiciliary factors cannot chang
The typical question here is whether a substantial nondiscrimina-

tory interest exists that might justify different treatment of cases in
which both parties are domiciliaries of the same foreign state. As
with Full Faith and Credit, a federal court enforcing this rule does
not infringe on a state's ability to set priorities, except to the extent
that it requires neutrality with respect to domicile. It simply accepts
the state's rule and holds it to its word.
C. A Final Wrinkle
The preceding examples should reveal the workings of t

conflicts-centered understanding, but one thing I have not discu
in detail is the conflicts-oriented counterparts to the categorie
cases that bedeviled Currie's interest analysis: true conflicts
unprovided-for cases. Currie's true conflicts occur whenever b
states grant rights; I have suggested that this category of case
broad but not difficult. In many situations, both states will agre
whose rights should prevail. With the model I set out, howe
cases may also arise in which both states grant rights and
state's conflicts rule provides that its rights should prevail. This
uation, a conflict between conflicts rules, is the real "true confl
And with respect to real true conflicts, I think, Currie's analys
sound. If a state has decided sincerely that its interest is great
than the interest of the other state, without treating the other sta
interests any more lightly than its own,341 there is no reason f
not to follow its own conflicts rule. The principle that the states
coequal sovereigns leads to no other conclusion.
The situation is slightly more complicated with respect to m
analogue to the unprovided-for case, where each state's conflic
rule provides that its rights should yield. The appropriate reso
tion may not be clear. If the other state directs that its rights s
yield, can the forum be blamed for ignoring its own law and "de
ring"? I think it can. If a state's conflicts rule provides tha
rights yield, there is no excuse for not applying that rule. In p
this is a matter of keeping states honest, ensuring that they do
try to game the system and expand their legislative jurisdiction
result, the Full Faith and Credit mirror-image test really comes down to switching the
domiciliary factors.
341. The constitutional restrictions I have identified force each state to do this. A state
will take the interests of other states just as seriously as its own because to claim priority in
case A, it must surrender legislative jurisdiction over mirror-image case B. In case B its
interests are those it has said are the less important in case A. Consequently, discounting
sister-state interests means discounting local interests in mirror-image cases. The interests of
other states are effectively the state's own interests.
part it is just a matter of demanding they apply their laws as written

- the conflicts rule says the rights yield, and yield they must. And
in part it is a matter of Full Faith and Credit. A conflicts rule pro-
viding that local rights yield unless foreign rights also yield imper-
missibly discriminates against foreign law.
VII. CONCLUSION
The approach I have developed allows us to draw some gen

conclusions about what conflicts rules will look like. The Priv
and Immunities Clause destroys the domiciliary-centered con
tion of governmental interests.342 If a state grants rights to its
ciliaries, it must grant them to nondomiciliaries in the same
Its conflicts rules must similarly provide that rights that pre
the hands of domiciliaries will also prevail in the hands
nondomiciliaries. The Privileges and Immunities Clause thus
vents the crudely selective exercise of legislative jurisdict
favor domiciliaries.
The Full Faith and Credit Clause has a similar, though more
subtle effect. A state may surely regulate transactions occurring
within its borders and involving its domiciliaries; that is, it may
adopt conflicts rules providing that in such cases, local rights pre-
vail. But if it tries to draw within its regulatory field other cases
involving its domiciliaries, the mirror-image requirements of Full
Faith and Credit will start to sap its territorial authority. For every
extraterritorial case it claims, it must yield the mirror image, which
will necessarily be a case arising within its borders.
It thus seems likely that territorial factors will play a large role
in conflicts rules,343 but this does not mean a return to the bad old
days of Joseph Beale. Beale's system produced arbitrary results not
because of its territorialist orientation but because of its rigid devo-
342. See, e.g., Ely, supra note 73, at 180; Laycock, supra note 73, at 251 ("[A] state's
interests in enriching local citizens . . . simply should not count."). The significance of a
domiciliary connection, I think, is not so much that it gives states a reason to extend rights as
that it gives them a justification for imposing penalties. See Brilmayer, supra note 71, at
1297-1303. If a state is truly concerned about what its domiciliaries do outside its borders, it
can probably impose criminal sanctions. See supra note 339.
343. This should not be surprising; the Framers were working within the Anglo-American
tradition that saw law as a territorial entity. See supra text accompanying notes 28, 255-60.
Beale notes the two conflicting traditions of "personal" and "territorial" law and pronounces
that "[t]he conception of the common law has always been the conception of a territorial
law." 1 BEALE, supra note 6, ? 5.2, at 52. Even today, federal laws are presumptively territo-
rial in their scope. See EEOC v. Arabian Am. Oil Co., 499 U.S. 244, 248-59 (1991).
tion to the last act doctrine.344 More sophisticated versions of the

territorial approach are possible.345
Nor must territorialism be the whole story. Comparative im-
pairment is a permissible approach, either by itself or as a supple-
ment to territorialist rules; so too is a rule that the law that validates
a contract prevails.346 There is substantial flexibility, and it is im-
possible to predict what rules states would ultimately select. Shared
conceptions of interest (for example, territorial sovereignty), how-
ever, would probably lead to substantive convergence, thereby pro-
ducing the beneficial effect of a reduction in forum shopping. The
constitutional restrictions I have examined essentially require states
to be consistent in their conflicts rules, and therefore prevent states
from favoring their own laws and domiciliaries. States may struc-
ture the rules so as to promote their own interests, but only by de-
ciding which factors are most important. This is clearly a decision
within the legitimate authority of the state - it is a matter of inter-
nal policy.347 Having asserted the importance of these factors,
though, they must yield control over cases in which those factors
point to another state. Consequently, states will probably be sin-
cere in their hierarchization of rights: they will lay claim to those
cases that are most important to them, and cede control of cases
they believe are more important to other states.348 There is no way
for a state to extend illegitimately the reach of its laws. The system
344. See Laycock, supra note 73, at 322.

345. Laycock suggests a territorial approach that looks to the location of the relationship
between the parties. See id. at 323-27. This is an interesting idea, although it seems odd that
if one Californian commits a tort against another in Texas, the parties' rights will differ de-
pending on whether they know each other or not.
346. The basic constitutional concern of conflicts jurisprudence is discrimination against
foreign domiciliaries or, as a lesser and probably derivative matter, against foreign law. The
model I sketch here does not do much to address problems of discrimination in favor of
plaintiffs generally, which may occur if litigation brings benefits to the forum. A state proba-
bly could then adopt the rule that the law favoring the plaintiff prevails. The constitutional
concern is that this would lead to privileging pro-plaintiff substantive law, since plaintiffs pick
where to sue. Similar concerns arise with a preference for the generally adopted rule.
347. Brilmayer puts it this way:
Particularly where the Court is assessing state interests, it should not impose an ideal
definition of interest but only ask whether a reasonable state might think it has an inter-
est under these circumstances. State preferences are likely to differ, in part because of
difference in value choices and in part because of divergent empirical assumptions. That
is what state lawmaking is all about.
BRILMAYER, supra note 8, at 165. I agree with the caveat that the Court should ask whether
a nondiscriminatory state might think it has a greater interest.
348. This surrender of cases in which the conflicts calculus points to another state is pre-
cisely what interest analysis refuses to do, and that is why I believe it is unconstitutional.
cannot be gamed because states are, in a sense, playing against

themselves: each assertion of power implies a retraction.349
The federal judiciary's role in this system would be to oversee
state conflicts rules and invalidate them when they violate one of
the constitutional constraints. Federal courts would not dictate the
substance of conflicts rules or tell states which factors are to be
deemed more important.350 They would ensure only that when a
state asserts that a particular arrangement of factors gives it th
greater interest in a case, it acts consistently and concedes that a
mirror-image array of factors gives another state a greater interes
This could be done simply by surveying other state conflicts dec
sions. The common mistake of the Court and conflicts scholars has
been to focus on individual cases, in which results can almost always
be justified, rather than on patterns of state decisionmaking, which
may reveal discriminatory conflicts rules. The suggested approach
allows states to set their own priorities, but then holds them to their
words. This seems the appropriate role for federal courts imple-
menting antidiscrimination norms: ensuring that when one state re-
jects the claims of another, it does so because it sincerely believes
that its interest in regulation is greater, not because it counts its
interests more heavily than those of sister states.
This approach has applications beyond the field of conflicts of
law, and is, in fact, almost identical to the approach Guido
Calabresi and Allison Moore have proffered as an important,
though neglected, form of judicial review.351 According to this ap-
proach, the defense of fundamental rights is a role appropriately
given to an independent judiciary, but the identification of such
rights - rights society deems important, not necessarily rights de-
349. The retraction does not mean that the state's rights will never be enforced. It means
only that they will yield to sister-state rights, and if there are no opposing rights, there is
nothing to which to yield. It is for this reason that a state's law may govern interactions
between its domiciliaries in places with no local law. See, e.g., American Banana Co. v.
United Fruit Co., 213 U.S. 347, 355-56 (1909) (Holmes, J.). The difference between the ab-
sence of a right and its subordination explains why broad scopes of rights help in keeping
states honest. If rights generally had narrow scope, states could try to game the system by
structuring their conflicts rules so that cases in which their rights yielded tended to be cases in
which no contrary rights existed. That would give them cheap wins in such cases, since their
rights would be enforced without the sacrifice of the claim to authority over mirror-image
cases.
350. On this point I disagree with Laycock, who believes that "[w
applies is a federal question, and each state is obliged to give the same
question." See Laycock, supra note 73, at 301.
351. See Guido Calabresi, The Supreme Court, 1990 Term - Forew
tion and Constitutional Accountability (What the Bork-Brennan Deb
L. REV. 80, 91-103 (1991); Allison Moore, Loving's Legacy: The Ot
Principles, 34 HARV. C.R.-C.L. L. REV. 163, 173-74, 178-82 (1999).
rived from the Constitution - is a task better suited to the demo-

cratically elected legislature.352
Calabresi's approach is an elegant way of taking advantage of
these two institutional competencies. He suggests that states be
permitted to make choices between fundamental values, with courts
reviewing essentially the sincerity of the choice. To subordinate
one important value in the name of another is a legitimate demo-
cratic decision; it is illegitimate only when the disfavored value is
weighed lightly because of who is asserting it.353 To determine this,
Calabresi suggests, the court must first ascertain whether the bur-
den of a challenged law falls upon a class whose interests the legis-
lature may hold more lightly than their own. If so, the court must
then examine the way the state resolves other conflicts between the
competing values, to see if the challenged regulation reflects a hier-
archy of values consistently instantiated in the law. If it does - if
the value subordinated by the particular law at issue also loses
when the burden of its defeat falls on the people well represented in
the legislature - it reflects a permissible choice between values. If
not, it suggests that discrimination is at work.354
The parallels should be clear. Sister state laws and domiciliaries
are always likely candidates for discrimination, for they have no
electoral voice. Federal courts should thus look to other examples
of state conflicts decisionmaking to see if the factors asserted to
make local law prevail in one case succeed when they support for-
eign law or foreign domiciliaries. Constitutional constraints do not
usurp states' abilities to decide which contacts are most important
-the danger analogous to judicial determination of which rights
are fundamental - but they do require that states' claims to au-
thority over particular cases be based on a consistent hierarchy of
contacts and not on discrimination against foreign law or citizens.
Preventing this discrimination is all that the Constitution does.
A domestic conflicts theory may in fact need more, but this is the
most important thing, and probably the only realistic goal for the-
ory, rather than federal legislation. If states comply with their con-
stitutional obligations, laying claim only to those cases to which
they sincerely believe they have a superior claim, there is no theo-
retical basis on which to fault them. There may be irresoluble dis-
agreements between states even on this approach, and these are the
352. See Calabresi, supra note 351, at 91.

353. See id. at 91-93.
354. See id.
real true conflicts. Federal legislation may reso

set out the substantive rights that always prevail,
conflicts rules by which states must abide.355
reveals what federal judges should do in the
legislation.
355. Congress's power to legislate under Article I is of course limited, see City of Boerne
v. Flores, 117 S. Ct. 2157, 2162 (1997); LAURENCE H. TRIBE, AMERICAN CONSTITUTIONAL
LAW ? 5-1, at 297 (2d ed. 1988), and it might be hard to displace local tort laws. Congress can
also legislate pursuant to the Full Faith and Credit Clause, which permits it to specify con-
flicts rules. U.S. CONST. art. IV, ? 1; see TRIBE, supra, ? 5-2, at 298.

Cho Law

Uploaded by

Copyright:

Available Formats

Cho Law

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

Cho Law

Uploaded by

Copyright:

Available Formats

The Myth of Choice of Law: Rethinking Conflicts

Author(s): Kermit Roosevelt III

Kermit Roosevelt III*

I. INTRODUCTION ..................................... 2449

* Senior Research Scholar, Yale Law School; Resident Fellow, Informat

2. Privileges and Immunities .................. 2516

Choice of law is a mess. That much has become a truism. It is a

1. William L. Prosser, Interstate Publication, 51 MICH. L. REV. 959, 971 (1953).

The mere existence of multiple monikers should not surprise.

11. See Nicholas deBelleville Katzenbach, Conflicts on an Unruly Horse: Reciprocal

The substitution of choice for conflict, I will argue, is the funda-

and then express their conclusions in such terms.18 In cases where

18. See infra section IV.B.

23. See infra section V.A.

they do not dictate unique solutions - but they constrain the

II. CONFLICTS THEORY

Articles about conflicts frequently begin with - or are enti

25. See infra Part VI.

A. Vested Rights: Joseph Beale

Although Joseph Story exerted a profound influence on the

In this enterprise he was greatly aided by a few strong principles

jurisdiction, local law must determine the consequence of acts

35. Id. ? 4.12, at 46.

ally rather elegant, its concepts interacting with a smoothness and

Admittedly, the practical effects of the theory were somewhat

B. Legal Realism: Walter Wheeler Cook

In part of the Preface entitled "Apologia," Beale noted that his

The realists directly attacked the idea of vested rights. In part

54. 1 BEALE, supra note 6, at xii.

Rejecting Beale's conception of "theoretical law" as "the body of

62. 1 BEALE, supra note 6, ? 1.12, at 9.

the interminable ping-pong created when the choice-of-law rules of

C. Interest Analysis: Brainerd Currie

74. See CURRIE, supra note 22, at 184.

those in which neither state has an interest, are "unprovided-for

First, interest analysis makes a very important advance by con-

75. See, e.g., LEA BRILMAYER, AN INTRODUCTION TO JURISDICTION IN THE AMERICAN

85. It is coherent in that a conventional legal principle (territorialism) specifies which

The only accurate generalization one can make about current

something to say about choice of law.97 I think it surely does, and

E. Methods and Objectives: What Interest Analysis Is, What a

The preceding sections have labeled the objects of their discus-

different kind of rule to resolve the conflict - to what I will call a

for forum law is unacceptable and must be rejected: Currie's con-

106. See id.

scope even if those conflicted with his standard domiciliary-focused

III. RETHINKING THE THEORY: FROM CHOICE TO CONFLICT

The goal of this article is to provide a workable framework fo

A. Rights and Their Critics

1. The Failure of Vested Rights / Territorialism

From a normative perspective, the most obvious problem with

sensible results. More serious is the problem that territorialism

Of course, Beale proposed the vested rights theory not as a nor-

to which practice must conform.123 Consequently, it cannot credi-

2. Salvaging the Concept of Rights

Beale's theory aims to help judges ascertain parties' rights. In

terms of the elements of actual experience are meaningless."128

When philosophical disputes have gone on for so long, with so

hand, is either trivial or nonsensical within the framework.135 An-