Notre Dame Law Review

View PDF

Aaron-Andrew P. Bruhl*

This Article addresses a problem that potentially arises whenever a federal court encounters a state statute.  When interpreting the state statute, should the federal court use the state’s methods of statutory interpretation—the state’s canons of construction, its rules about the use of legislative history, and the like—or should the court instead use federal methods of statutory interpretation?  The question is interesting as a matter of theory, and it is practically significant because different jurisdictions have somewhat different interpretive approaches.  In addressing itself to this problem, the Article makes two contributions.  First, it shows, as a normative matter, that federal courts should generally use state methods, though there should be exceptions in situations in which using state methods would impair federal interests or depart from the state’s own understanding of its methodology’s scope.  Second, the Article shows, through a systematic review, that federal courts are mostly applying state interpretive approaches today, a conclusion that runs contrary to the views of other commentators.

INTRODUCTION

Harry Tompkins was hit by something sticking out of the side of a train.  The landmark case of Erie Railroad Co. v. Tompkins held that the tort law governing his lawsuit against the railroad was state tort law as announced by the relevant state courts, not a “federal general common law” of torts that the federal courts could independently discern and shape.¹   Although the Erie case involved common law, federal courts must apply state statutes too.  To apply statutes means to interpret them.²   Federal interpretation of state statutes is therefore pervasive.

And how does one go about interpreting a state statute?  In many cases, interpretation feels automatic: an automobile is a “vehicle” within the meaning of a drunk-driving statute, and the only rule one needs to know (or intuit) in order to reach that obvious conclusion is that words in statutes normally carry their ordinary meaning.  But statutory interpretation is also a technical activity governed by dozens of canons and presumptions of interpretation (some commonsensical, some not), plus rules about the use of extrinsic sources like legislative history and agency interpretations.³   And just as a railroad’s duty to people walking along its tracks can differ under New York, Pennsylvania, and federal law, the canons and rules of interpretation can differ across jurisdictions as well.⁴ 

Consider a few examples of divergences in interpretive methods:

• Federal courts generally interpret federal statutes with a presumption that the statutes apply domestically only, not extraterritorially.⁵   Some states interpret their state statutes with a similar presumption against extraterritoriality, but others do not.⁶   Whether a state statute applies to conduct outside of the state can therefore depend on whether a court applies the federal presumption or a state nonpresumption.  Which rule should a federal court use when it is applying a state statute from a state that has no such presumption?

• A Virginia statute directs that bill summaries prepared during the legislative process may not be considered as evidence of legislative intent.⁷   The U.S. Court of Appeals for the Fourth Circuit, which encompasses Virginia, believes that legislative history may be considered to discern the intent behind federal statutes, at least if the statutory text is unclear.⁸   May courts in the Fourth Circuit consider bill summaries for Virginia statutes?⁹ 

• Some state courts are experimenting with “corpus linguistics,” which is said to be a more scientific way to determine ordinary meaning than using dictionaries or appealing to judicial intuition.¹⁰   Must federal courts use that method for those states’ statutes, even if competent execution of the technique requires a Ph.D. or other training?

As these examples show, federal courts interpreting state statutes face an Erie-like choice-of-law question, more precisely a choice of interpretive law.  Which interpretive methods should federal courts use?  And what do federal courts actually tend to do in these circumstances—apply the enacting state’s methods or federal methods?  The questions are not only of theoretical interest, for case outcomes can change depending on the governing rules.

Setting out the solution to statutory interpretation’s Erie problem and determining whether the federal courts have reached that solution requires both normative and positive work.  The Article does that work in three parts.

Part I is descriptive and explanatory.  It begins with the existing scholarly accounts of what the courts are doing.  According to most scholars who have studied the matter, the federal courts are either confused about what to do or usually apply federal interpretive principles to state statutes.¹¹   Through a systematic review of the existing caselaw, my research reveals a different reality.  The courts are sometimes inattentive and sometimes make mistakes, but they are not particularly divided over what they should be doing.  What they are doing is, as a general rule, applying state interpretive methods to the interpretation of state statutes.  The lower federal courts are much more careful about this than the Supreme Court.  There are, it is true, certain categories of cases in which the federal courts often apply federal interpretive principles, but, contrary to other scholars’ perceptions, these categories are circumscribed exceptions to the general rule.  And, moreover, it makes sense that these domains would be treated as exceptional.

Part II considers what the federal courts should do when they encounter state statutes.  Happily, the normative prescription largely aligns with the actual practices described in Part I.  Federal courts should generally apply state interpretive principles when interpreting state statutes.  In adopting this prescription, I mostly agree with the majority of the small group of commentators who have considered the matter,¹²  but the account here aims to improve on prior analyses in several respects.  Some of the existing accounts overlook important aspects of the problem, such as countervailing federal interests and what states have told us about the proper scope of their interpretive methods, or they rely too heavily on analogies to doctrinal spaces like contract interpretation at the expense of the relevant first principles.  On the way to reaching the conclusion that state methods should generally apply, Part II considers and rejects the most plausible arguments to the contrary, including that the Constitution itself provides a federal interpretive method, that interpretation is a “procedure” that should be governed by federal law rather than enacting-state law, and that the states themselves do not regard their methods as law applicable in other courts.  Each of those counterarguments has force in some circumstances, and some turn on debatable empirical propositions, but none is generally true.

An important aspect of the Article’s contribution comes in Part III, which sets out exceptions to the general duty to follow state law and elaborates on some special topics.  Most notably, Part III marks out a limited space in which federal interests or federal judicial incapacity mean that federal courts should not follow state rules for interpreting state statutes.  These considerations at least partly justify the limited departures from state methodology observed in Part I.  Moreover, these federal interests could in the future justify broader federal departures from some novel state interpretive practices, such as the corpus-linguistics analysis with which some states are experimenting.¹³ 

I.     THE FEDERAL COURTS’ CURRENT ANSWER

This Part’s task is to describe what the federal courts are currently doing.  I begin with how other scholars have described the situation.  I then undertake my own, more systematic investigation.

A.   The Existing Scholarly Assessment

Abbe Gluck has done more than anyone else to put statutory interpretation’s Erie issues on the scholarly radar screen.  She contends that the federal courts generally should follow state interpretive methodologies when interpreting state statutes.¹⁴   When it comes to describing existing federal practice, she writes that the federal courts have generally failed to do as they should.  The federal courts, she writes, “do not typically apply state methodology to state statutory questions” and indeed “rarely consider state rules of statutory interpretation.”¹⁵  

Most other observers who have considered the matter have reached similar conclusions.  The tome on precedent recently published by Bryan Garner and more than a dozen judges states, for example, that “[a  few (but not all) federal courts” apply state interpretive principles to state statutes.¹⁶ 

There is some dissent from the prevailing scholarly assessment of the federal courts’ approach to state statutes.  Sydney Foster, in a generative early article on methodological stare decisis, wrote that “[f]ederal courts have held that when they are interpreting state statutes, the federal courts must follow state, not federal, statutory interpretation doctrine,” though a footnote acknowledged that “that adherence to that rule may not be uniform.”¹⁷   A leading treatise on statutory interpretation also suggests that state methodology applies, though it is fuzzy on whether its conclusion is descriptive or more aspirational:

When construction of a foreign statute or a statute of another state is at issue, courts must choose the relevant law of statutory construction.  Since the rules, methods, and principles of statutory construction are means to ascertain the substance of statutes, those employed in the enacting state apply most naturally.¹⁸ 

None of these sources provides a systematic demonstration of the federal courts’ practices; they instead cite some cases (quite a few or just a couple, depending on the source) as evidence of the broader pattern they assert to exist.

Alongside the specific claim that federal courts do not apply state interpretive methodology, one often sees the broader claim that federal courts, unlike some states, do not give interpretive methodology precedential effect or that they do not treat it as “law” at all.¹⁹   I think the prevalence of methodological precedent has been underestimated, in large part because commentators focus too much on a few of the U.S. Supreme Court’s seemingly intractable disputes, such the debate over the use of legislative history, to the neglect of other interpretive topics and the rest of the judiciary.²⁰   Those who believe there is little interpretive “law” in the world will find fewer instances in which a choice between federal and state interpretive law would arise.  But if the federal courts frequently state that they are required to use state methods established in state law, that finding would tend to bolster the position that interpretive methodology, or at least much of it, is more or less ordinary law subject to more or less ordinary precedential effect.  That is in fact what we find, as the next Section shows.

B.   The Reality: Federal Courts Routinely Use State Methodology

There are exceptions to any generalization, but, to generalize nonetheless, the federal courts are not confused or conflicted about which interpretive rules apply to state statutes.  They are sometimes inattentive to the choice-of-law question, it is true, but when they address it they almost universally agree that state methods control.  And the more carefully courts think about it, the more likely they are to reach that conclusion.  The conscious departures from state interpretive principles tend to fall into a couple of circumscribed categories, categories that the courts may actually be correct to treat differently.  But those are exceptions to a general rule, not evidence of widespread confusion about the rule.

I employ a few different approaches to reach these conclusions.  None of the approaches is determinative by itself, but if each strategy yields similar results, we can have confidence in the robustness of the result.  I begin with a traditional doctrinal investigation in which one seeks cases clearly affirming or negating the relevant propositions.  But I also try to examine a sample of all the relevant cases to get a sense of the relative prevalence of different positions.

1.   Cases Citing Erie

One test is whether federal courts interpreting a state statute expressly cite Erie or the Erie doctrine and explain that it requires them to apply state interpretive methodology or that is does not.  It turns out that courts relatively rarely cite Erie itself as the relevant authority in this context, but when they do, they virtually always say that Erie requires them to follow state interpretive methods, even if they would otherwise approach the statute differently.²¹   If one looks for cases invoking Erie and determining that it does not require use of state law, one will not find much.  The evidence is so sparse, in fact, that statements to that effect can be written off as simple misunderstanding.  That is how I would describe the Fifth Circuit’s decision in Batterton v. Texas General Land Office, which has been cited as authority for the proposition that federal courts generally do not believe they are required to follow state methods.²²   The case involved federal-question jurisdiction, as the plaintiff claimed that a Texas statute violated the federal constitutional guarantee of due process of law.²³   To resolve the constitutional claim, the court had to interpret the state statute.  In so doing, the court used the Texas courts’ rules of statutory interpretation.²⁴   That, as I have said, is the norm.  The court added, however, that “[its] reference to these [Texas methods] is not mandated by [Erie]”becausejurisdiction was not based on diversity of citizenship.²⁵   If the court meant that the Erie doctrine does not apply except in cases predicated on diversity jurisdiction, that proposition is incorrect under widely accepted law: state law comes into federal court in nondiversity cases, and when it does it is the state authorities’ understanding of their law that applies,²⁶  including their law of interpretation.²⁷   But courts make that same mistake about Erie’s domain in cases that do not involve interpretive methodology.²⁸   It is a big docket out there, and busy courts sometimes just get things wrong, even when faced with things much simpler than the Erie doctrine.  Indeed, later Fifth Circuit cases apply Texas interpretive methods to Texas statutes in nondiversity cases without mentioning Batterton.²⁹ 

In any event, having disclaimed a duty to apply state law under Erie, the Fifth Circuit in Batterton immediately added that “[r]eason dictates . . . that in deriving a meaning from the statute’s words we should approach them exactly as would a Texas court.”³⁰   And, it continued, “It would make little sense for us to proceed to an aberrant construction by refusing to apply state canons, canons which will in all probability govern any authoritative construction which the statute ever receives.”³¹   True!  So having disclaimed an Erie-derived compulsion, the court applies state methods anyway—for Erie-tinged reasons, to boot.  So while Batterton made some imprecise statements about the bounds of the Erie doctrine, any error was harmless.

But, again, cases resolving a choice-of-interpretive-law question usually do not cite Erie or refer to “the Erie doctrine.”  The failure to invoke Erie by name, if a court is otherwise applying state methodology, does not substantially undermine the proposition that the federal courts generally regard themselves as obligated to apply state methods to state statutes.  The vast majority of diversity cases do not cite Erie.³²   Erie is in the judicial muscle memory, so courts can just do it, automatically applying state statutes and common law, without citing Erie or, for that matter, expressly analyzing choice of law at all.³³   When a court feels the need to cite a case for the proposition that state interpretive methodologies apply, an on-point circuit precedent is a more natural choice inasmuch as Erie itself involved common law, not how to interpret a state statute.  Accordingly, most of the cases that say that state methods apply rely not directly on Erie but on circuit law³⁴ —and that is totally fine and normal.

2.   Cases Addressing Whether Federal Courts Must Use State Methods, Without Necessarily Citing Erie

Since explicit citation of Erie is not necessary or even routine in cases applying state law, it is more illuminating to see what the federal courts say, even without invoking Erie, about whether state or federal law governs the interpretation of state statutes.  When courts address that question, they almost always say that state methods apply.

a.   Supreme Court

Although the relevant evidence comes overwhelmingly from the lower federal courts, which is unsurprising given that the Supreme Court has less interaction with state law, it is natural to start with the Supreme Court’s pronouncements.  As it happens, the Supreme Court has made some statements over the years that probably would have been sufficient to settle the typical doctrinal dispute.  Consider these:

• “[T]he weight to be given to the legislative history of an Alabama statute is a matter of Alabama law to be determined by the Supreme Court of Alabama.”³⁵ 

• “The power to determine the meaning of a statute carries with it the power to prescribe its extent and limitations as well as the method by which they shall be determined.”³⁶ 

These statements have generated some citations, but not many.³⁷   That these pronouncements have not propagated through the system very successfully is probably attributable in part to their methodological character.  Most Supreme Court cases are decided because of, and become known and cited for, the question of law that forms the core of the case, not methodological propositions that lead up to its resolution.  As courts become more self-conscious about interpretive methodology as a body of transsubstantive law in its own right, thanks in part to the efforts of legislation scholars and their students, one can predict that statements like those above will gain greater currency.

Although no Supreme Court cases directly disavow the statements above, the Court’s inattention to the Erie question in many cases certainly has not helped to crystalize the law on this question.  A prominent example of a case that did not seek to discern the proper source of its interpretive principles is Stenberg v. Carhart, the case testing the constitutionality of Nebraska’s ban on “partial-birth abortion.”³⁸   Judging the state statute’s constitutionality required first discerning what it prohibited, but the various opinions in Stenberg engaged in interpretation using an unsystematic mélange of authorities.  The majority referred at one point to the level of deference the state attorney general’s construction of the law would receive in the state courts, as if it were a question of state law,³⁹  but at other points the majority used federal interpretive principles without showing that they were equivalent to state rules.⁴⁰   As support for a statement about the value of statutory definitions, the Court cited federal cases plus the general-purpose Sutherland treatise.⁴¹   For good measure, the Court’s interpretation included a dash of deference to the lower federal courts’ interpretations of the state abortion statute (though without describing which methods those courts used).⁴² 

Justice Thomas’s dissent, which disputed the majority’s statutory interpretation at length, did not expressly address choice of law either, but he mostly appeared to apply either federal or assumedly universal principles.  His opinion referred to “the [interpretive] principles that this Court follows in every context but abortion” and how “[w]e interpret statutes.”⁴³   He did at one point acknowledge that the Nebraska courts would give weight to the state attorney general’s narrow interpretation of the statute, but this was offered only as an additional reason to give the statute the interpretation that Justice Thomas thought was correct.⁴⁴ 

As I am about to show, the lower federal courts do not display the inconsistency—frankly, the sloppiness—of Stenberg.  Perhaps that is because the lower federal courts get regular exposure to the statutes of the state(s) in which they sit.  At the Supreme Court, by contrast, questions about the meaning of state law arise relatively less often, no particular state’s law appears very much, and, when state law is at issue, it mostly arises in connection with whether it is preempted by or otherwise inconsistent with federal law.  The reason for the grant of certiorari is generally the federal issue, not the meaning of the state law.⁴⁵ 

b.   Lower Federal Courts

The lower federal courts are very consistent.  They have settled on the proposition that federal courts must, subject to some exceptions mentioned below, interpret state statutes using state methods and canons.  I could provide many, many examples, but the footnote contains published cases from every regional circuit and the D.C. Circuit stating that state statutes should be interpreted using state interpretive methods.⁴⁶   In the words of one district court, “it is well established that a federal court interpreting state legislation must employ that state’s rules of statutory construction.”⁴⁷ 

Most of the cases just cited were decided within the last twenty years or so, because my goal here is to show that the law is clear today, but some courts discerned this rule decades earlier.⁴⁸   A few judges perceived the need to apply state interpretive methodology to state statutes even before Erie, though I do not believe such views were common.⁴⁹   If asked to identify sources of interpretive law, the pre-Erie federal courts might have regarded the question as involving when a presumably universal law of interpretation reflected in leading treatises and leading cases should yield to local methodological peculiarities.⁵⁰   Today, after Erie, it is more natural to perceive a choice between one of fifty sovereign states’ methods on the one hand and a “federal common law of interpretive methodology” on the other.  The systematization and distinctiveness of certain states’ methods, coupled with greater self-consciousness about interpretation, sharpens the choice, making it harder for today’s courts to miss.

Although this Part of the Article is aimed at documenting what the federal courts do, rather than justifying or criticizing their practices (those tasks to be taken up in Parts II and III), it is worth briefly considering what the courts report about why they must apply state interpretive approaches.  Now, they do not tend to say much about why.  That paucity of comment is itself telling, as it may reveal a natural intuition about how to understand a text from a different language community.  If one knows a text is written in Lowland Scots, which looks familiar to a modern American reader but not quite right, one would determine its meaning by using the grammatical rules and lexicon of Lowland Scots, not the rules of its cousin American Standard English.  So too with legal texts written in Utahn or Californian, the implicit logic would run.  That is, some courts see the duty to apply state interpretive methods as flowing obviously from a duty under Erie to construe the meaning of a statute as the state courts would, to mirror the predicted state ruling.⁵¹   In some cases, the federal courts will follow the state’s interpretive methods to a result that they believe the state high court would reach on a proper application of its methods, even if that means departing from nonbinding decisions on the particular substantive question at hand.⁵² 

In the rather rare instances when other justifications are adduced, the courts refer to considerations such as respect for state substantive policies, accuracy in determining the state legislature’s intent, and the “unseemliness” of the two systems yielding different results.⁵³ 

3.   Systematic Study of Federal Cases

The citations in footnote 46 could be multiplied many times over, but even a boatload of supportive citations can prove only so much.  What about unsupportive cases, the ones that use federal interpretive principles?  Do those exist in any significant number?  One wants a sense of the relative proportions.  For that, one needs a more systematic approach, not just a heap of supportive citations.⁵⁴   In this section, I aim for that more systematic demonstration of the federal courts’ behavior, for both the courts of appeals and the district courts.  The following several pages describe the approaches I used and the finding on which they converge.

For the courts of appeals, the first strategy was to search for cases for the years 2000 through 2019 (inclusive) that had both (1) West Key Numbers associated with statutory interpretation and (2) terms aimed at finding diversity cases.⁵⁵   This search yielded 300 cases.  More than half of those were discarded as nonresponsive (such as because they addressed the interpretation of the federal diversity jurisdiction statute or removal statutes or simply applied on-point state holdings without addressing interpretive methods).  Of the remaining cases, the results break down as shown in the “search 1” column below.

The second search, called “search 2” in the chart, located cases from the five-year period 2014 through 2018 that the West Key Numbers categorize as “State or Federal Laws as Rules of Decision; Erie Doctrine” and which also contained search terms aimed at finding statutory cases.⁵⁶   As above, the search terms yield some cases that are false positives, such as cases interpreting only federal statutes or cases in which the federal court made general statements about its duty to follow state law but did not engage in its own first-impression interpretation of a statute.  These were disregarded, leaving the responsive results that are shown in the chart.

I ran the same two searches in the district courts, with the results also shown in the chart.⁵⁷   (Results are on file with the author.)

Table 1: Federal Courts’ Interpretive Methodologies

Position on methodological question	Court of appeals — search 1	Court of appeals — search 2	District court — search 1	District court — search 2
Explicit that state methods control	69 (58.5%)	28 (60.9%)	7 (53.8%)	38 (70.4%))
Cites state methodological authorities	22 (18.6%)	8 (17.4%)	3 (23.1%)	6 (11.1%)
Uncategorized—cites no interpretive authorities	4 (3.4%)	3 (6.5%)	0	1 (1.9%)
Cites a mixture of federal and state methodological authorities	13 (11.0%)	4 (8.7%)	3 (23.1%)	5 (9.3%)
Cites federal methodological authorities	10 (8.5%)	1 (2.2%)	0	3 (5.6%)
Explicit that federal methods control	0	2 (4.3%)	0	1 (1.9%)
Total responsive cases	118	46	13	54

Moving from the top of the chart to the bottom, the rows go from most supportive of using state methods to most supportive of using federal methods.  The two rows in the middle are neutral or ambiguous.  Some cases require judgment calls about which row they belong in, but there is no mistaking the overall message.  The dominant finding is that cases that explicitly address choice of law almost always say that state methods control.

For an example of a case categorized as “explicit that state methods control,” consider the Ninth Circuit’s decision in In re Western States Wholesale Natural Gas Antitrust Litigation.⁵⁸   In addition to interpreting the federal Natural Gas Act, the court had to construe a statute of Wisconsin (a state outside the Ninth Circuit, interestingly).⁵⁹   In doing the latter, the court wrote that “[i]n interpreting a state statute, a federal court applies the relevant state’s rules of statutory construction,” and it then described its understanding of the Wisconsin method.⁶⁰   Earlier in the opinion, in interpreting the federal statute, the court cited federal cases as authority for interpretive propositions like the whole-act rule and the canon against surplusage.⁶¹   The case thus provides a nice example of matching different methods to different statutes.

The group of cases that interpret a state statute by citing mixture of federal and state authorities is ambiguous and of uncertain value.  I would venture that the majority of these cases are just the result of inattention to the existence of a choice-of-law question.  Some others might blend federal and state interpretive sources due to uncertainty about which law applies and a desire to cover all the bases.  Perhaps some judges proceed as if interpretation is general law, neither state nor federal.  There is some practical truth to that view in that most jurisdictions’ principles of interpretation (like their principles of contracts) are mostly similar.

The cases that cite federal interpretive principles divide into two types.  The first and larger group is cases that contain bland statements like “the interpretive enterprise begins with the text” and then cite only federal sources for those propositions.⁶²   I think the most likely explanation for these cases is inattention.  Insurance contracts, trusts, and similar documents jump out as nonfederal because most federal courts encounter so little federal contract law.⁶³   (And yet, even when interpreting a document like an ordinary contract—to which state methodology is universally agreed to apply—federal courts sometimes mix up their interpretive maxims and rules.⁶⁴ )  But federal courts deal with federal statutes every day and know the boilerplate methodological phrases, so the appearance of a state statute in need of interpretation does not register as an anomaly that sends the clerks scurrying to the state reporters or insurance treatises for aid.⁶⁵ 

The second category of cases that cite federal interpretive authorities, and the very few cases that explicitly employ federal principles, involve either the canon of constitutional avoidance or a rule about avoiding inventive interpretations of state law.  I address these cases in more detail in the following Section.

The foregoing analysis did not attempt to determine whether the federal courts correctly discerned the content of state interpretive methodology or applied it correctly.⁶⁶   Here the goal was instead to determine what federal courts take to be their obligations regarding the use of state methods, not how well they discharge them.

C.   Limited Exceptions to the General Use of State Methodology

Although the federal courts generally understand themselves to be bound to apply state interpretive methods to state statutes, there are a few special circumstances in which they intentionally depart from state methods.  The two most important examples involve dynamic interpretations of state law and the doctrine of constitutional avoidance.  The key point, however, is that these are exceptions to a general rule, not the general rule, and they are not even evidence of some widespread confusion about which law applies.  Moreover, although this Section is descriptive and not normative, I will explain later that it is understandable, and sometimes even correct, for the courts to treat these contexts differently than they treat others.

1.   Dynamic Interpretation of State Statutes

As an example of federal courts failing to use state interpretive methods, Gluck cites cases in which federal courts refuse to engage in dynamic updating or otherwise innovate in state law.⁶⁷   The Third Circuit, for example, has referred to a principle that “where ‘two competing yet sensible interpretations’ of state law exist, ‘we should opt for the interpretation that restricts liability, rather than expands it, until the [state supreme court] decides differently.’”⁶⁸   If the best understanding of the state’s interpretive methodology would yield an expansive reading of a statute—despite a “sensible” but less persuasive contrary interpretation—then this principle expressly counsels departure from state methodology.

The phenomenon of federal cases eschewing dynamism in state law does not provide a particularly strong counterexample to the general practice of following state interpretive principles.  To begin with, cases espousing nondynamism in handling another jurisdiction’s law are not limited to or even particularly connected with cases interpreting statutes rather than cases involving common law, nor is the sentiment limited to federal courts encountering state law.⁶⁹   I think a preference for a static approach to state law is questionable as a general matter,⁷⁰  but, agree or not, this dispute is not about statutory interpretation’s supposed departure from generally applicable Erie principles.

Moreover, some of the cases stating this anti-innovation principle are probably best understood not as rejections of using state interpretive methods when they predict novel applications but instead as rulings that the best understanding of state law is that that the state court would not recognize a new claim or other proposition because it is implausible and lacks a toehold in existing jurisprudence.⁷¹   I do not contend that all of the cases fit that description, but it does reduce the number of cases that embrace antidynamism even where it leads to anticipated divergence.

2.   Constitutional Avoidance

The usual rule, albeit stated in different ways, is that courts should interpret statutes so as to avoid serious doubts about their constitutionality.⁷²   When a federal court is interpreting a state statute that presents constitutional difficulties, the question arises whether the federal court should use the federal avoidance canon, the enacting state’s avoidance canon, or perhaps some special avoidance canon just for crossover cases.  The question has outcome-determining potential if the candidate canons meaningfully vary in their content (e.g., how grave must a doubt be to trigger the canon, how reasonable does the doubt-avoiding interpretation have to be).

One does not find a uniform approach among federal courts in crossover avoidance cases.  There are at least four different approaches.

First, there are plenty of federal cases that employ state avoidance canons.⁷³ 

Second, and in contrast, one also finds federal cases that do not apply or seek to discern state avoidance principles but instead apply (or appear to apply) a weakened version of avoidance that is purportedly based on the intersystemic nature of the case.  In particular, there are cases stating that the federal courts have unusually limited leeway to reach avoidance-inspired narrowing constructions of state laws.⁷⁴   One Sixth Circuit case went so far as to say that “[f]ederal courts lack authority and power to give a limiting, narrowing construction to a state statute.”⁷⁵   The court went on to find the challenged part of the state statute unconstitutional, albeit severable from the rest of the statute, rejecting the district court’s formulation of a saving construction.⁷⁶ 

Third, there are federal cases that, without expressly addressing which avoidance doctrine applies, cite only the federal avoidance doctrine in what appears to be its ordinary form.⁷⁷   Without more, it is hard to know what to make of these cases.  They could represent an implicit ruling that the federal canon applies, an assumption that state and federal canons are relevantly equivalent (which would often be correct),⁷⁸  or simple inattention to the choice-of-law question.⁷⁹ 

Fourth, there are cases that cite both federal and state avoidance cases, either in the course of expressing uncertainty about choice of law or without evident explanation.⁸⁰ 

The footnotes attached to the preceding several paragraphs cited some examples of each category, but in this instance it is hard to provide more systematic results as I did earlier.  There is not an easily identifiable dataset because invocations of the avoidance canon often blend into or coincide with other doctrines like the presumption of constitutionality, severability doctrine, and constitutional tests like the void-for-vagueness doctrine and First Amendment overbreadth doctrine.⁸¹   Reluctance to provide narrowing constructions to fend off constitutional challenges to state statutes also blends into, and may be justified by the availability of, procedural doctrines of abstention or certification,⁸²  such that they may not reflect a real choice of interpretive law at all.  I cannot provide a firm percentage, but I can say that the results are mixed and that federal courts have not settled on a clear consensus in favor of state law in this category of cases.  Avoidance doctrine therefore departs from the usual pattern observed above.⁸³ 

Regarding that departure, two comments are in order.  First, intersystemic avoidance is tricky, for it involves consideration of strong federal interests that most encounters with state law do not.  As Section III.A will explain, there is a reasonable, principled basis for federal courts to depart a bit from emulating state courts when it comes to constitutional challenges to state statutes.  The courts may be on to something here, in other words.  Second, this troublesome pocket of doctrine does not disprove the broader pattern.  Indeed, a few courts seem to realize that avoidance scenarios involve an exception to a baseline requirement to use state methods.⁸⁴ 

II.     THE RIGHT ANSWER

We now switch to what the federal courts should do when they encounter state statutes.  This requires grappling with the Erie doctrine and the principles animating it.

Erie analysis is famously difficult.  To start with, “the Erie doctrine” is actually several distinct doctrines with different functions.⁸⁵   And one finds disagreement among courts and scholars over just about every branch: the extent of the federal courts’ power to fashion common law,⁸⁶  the preemptive scope of the Federal Rules of Civil Procedure,⁸⁷  the source of the so-called “twin aims of Erie,”⁸⁸  and many others.  To top it off, the Supreme Court in Erie may have badly misunderstood the Rules of Decision Act, which is hardly a way to get a doctrine off to a good start.⁸⁹ 

Given the state of the Erie doctrine, I will try to provide multiple routes to my ultimate conclusion that federal courts should generally follow state interpretive methods when construing state statutes.  At a few points the argument will make an unusual move, such as giving more attention to the text of the Rules of Decision Act than is typical, but there will be an alternative argumentative path that relies on other grounds.  The reader is welcome to pick and choose and reorganize the chunks of the argument to fit his or her preferred understanding of the Erie doctrine.⁹⁰ 

Nonetheless, the Article needs to unfold in some way, and my organizational plan generally follows the Hanna-inspired approach to organization by beginning with a search for applicable federal positive law and then, if there is no such law, moving to a balancing of state and federal interests.⁹¹   In line with the usual hierarchy of positive sources, I begin with federal constitutional requirements and constraints on the choice of interpretive methodology.  I next consider federal statutes or other enactments that might be thought to control federal interpretation of state statutes.  The search for constitutional and statutory commands yields only a few federal constraints on interpretive methods but no comprehensive federal method for state statutes.  The real action, therefore, comes in the “unguided Erie” analysis, which considers state and federal interests.  This analysis shows that federal courts should generally follow state methods for several partly overlapping reasons: because the state methods are (to quote from the RDA) “rules of decision” that “apply” when state statutes are interpreted,⁹²  because interpretive rules bear on the regulation of primary conduct, because interpretive rules are “outcome determinative” in the relevant way, because interpretive methods are not mere “procedure,” and because the “twin aims of Erie” militate in favor of using state law.

A.   Federal Constitutional Requirements That Bear on Interpretive Methods

The United States Constitution is the supreme law of the land, and so any Erie analysis must begin there.  The Constitution could solve statutory interpretation’s Erie problem by either (1) commanding some specific interpretive method for the federal courts to use even when interpreting state statutes, (2) preempting the use of state methods (even without requiring any particular federal method), or (3) giving each federal judge the freedom to choose an individual methodology free of any other constraint.  We can consider those possibilities in turn.

1.   Some Federal Constitutional Limits

The Constitution imposes some constraints on interpretive approaches in federal and state courts alike.  Holding a séance, interpreting statutes to favor the richer party—those methods have to be excluded whatever the source of the law being interpreted.  And a few interpretive rules might be constitutionally required, such as a core version of the rule of lenity derived from the Due Process Clauses.⁹³ 

Beyond minimal federal constraints like those just mentioned, does the Constitution go further to specify a required interpretive method for federal courts?  The primary locus of debate over constitutional limits on interpretation is the long-running dispute over the use of legislative history.  Some textualists contend that the Constitution implicitly commands their method and, specifically, outlaws certain uses of legislative history.⁹⁴   I am not persuaded on that point, but for the sake of keeping this Article’s Erie argument as neutral as practicable in the broader interpretation wars, the key point is that the most powerful constitutional arguments for limiting the use of legislative history come not from the nature of Article III “judicial power” but instead from Article I or the relationship between Article I and Article III.⁹⁵   That is, reliance on legislative intents and purposes is problematic, according to the textualists, because it circumvents the exclusive, finely wrought Article I, Section 7 process for making law.⁹⁶   Or, to put the argument slightly differently, giving effect to legislative history would allow Congress to delegate its legislative power to subgroups like committees.⁹⁷   If these textualist arguments were correct, they would not apply to the federal courts’ relationship to state law, which may have its own notions about lawmaking functions and separation of powers.⁹⁸ 

It is worth pointing out that some arguments in favor of using legislative history are also based on propositions that do not apply to federal courts’ interpretations of state statutes.  These include arguments that rely on parts of Article I or Congress’s rules of debate.⁹⁹   As above, these federal provisions are not relevant here.

2.   Federal Constitutional Preemption of State Authority

Consider next a different constitutional argument.  It does not posit that the Constitution commands any particular interpretive approach, but it instead posits that the Constitution prohibits the states from having any say over federal courts’ interpretive methods even when state statutes are at issue.  This would be a sort of constitutionally mandated field preemption, in which state authority is excluded from the domain of federal interpretive methodology.  On this view, some kind of federal law is the only choice for filling this space.

One might find some support for this view in the Supreme Court’s statements to the effect that the states have no business regulating how the federal courts operate or, to use the familiar but fraught terminology, no business in regulating federal “procedure.”  As Chief Justice Marshall wrote in Wayman v. Southard, which concerned the method for executing a judgment in a diversity case: “That [the power to regulate federal court procedure] has not an independent existence in the State legislatures, is, we think, one of those political axioms, an attempt to demonstrate which, would be a waste of argument not to be excused.”¹⁰⁰ 

But there are at least two reasons why the Constitution cannot reasonably be read to establish statutory interpretation as a field immune from state regulation.

First, Marshall was discussing what the states had the power to command on their own, not what role state standards might play at the invitation of federal authorities.¹⁰¹   It is unremarkable that Congress and federal rulemakers may and sometimes do direct the federal courts to follow state procedure, and the federal courts themselves do the same—yet none of this means the states have independent, trumping power to regulate federal procedure.¹⁰² 

Second, regardless of any federal invitation, the states’ power to define state-law rights and duties can require other sovereigns—whether the national government or sister states—to handle those claims in ways that differ from the way those sovereigns would handle an otherwise similar homegrown claim.¹⁰³   Matters such as a right to attorneys’ fees or the allocation of the burden of proof often follow a claim into other courts by virtue of the regulatory interests that require the foreign court to entertain the claim itself.¹⁰⁴   This is not a power to meddle in the procedures of other courts but is instead a reflection of the full scope of the claim-creating state’s legitimate regulatory authority.¹⁰⁵   Compared to attorneys’ fees or burdens of proof, methods of statutory interpretation are relatively easier to treat as part of the substance of the state law in which the enacting state has legitimate regulatory interests.  Indeed, although nothing of importance should turn merely on a labeling exercise, interpretive methodology may not be procedural in any relevant sense at all.¹⁰⁶ 

3.   Interpretive Freedom from External Authority?

Another constitutional argument about power over methodology is also worth mentioning.  What if Article III, properly understood, gives every individual federal judge the indefeasible freedom to pick his or her own method (textualism, pragmatism, no grand theory at all), so long as the chosen method abides by other constitutional constraints like due process?¹⁰⁷   If this view of judges’ power is correct, then no authority—not Congress, not the Supreme Court, not the state courts, not the state legislatures—can bind a federal judge to any interpretive method.

I do not believe this individual-interpretive-freedom account of Article III is correct.  If it were correct, it would have far-reaching consequences that go well beyond the Erie issue.  For one thing, it would mean that a Supreme Court possessed of the will would lack the power to direct inferior federal courts to follow a particular methodology or even to follow particular interpretive rules, and neither could a court of appeals establish such a method for its own judges.  Further, if the “judicial power” conferred by state constitutions carries the same meaning, then state courts would not be bound by state legislative directives or directions from superior state courts either.  Given the extraordinary breadth of the individual-freedom view, I feel comfortable skipping over a detailed response here.  I will pause only to say that even if every federal judge has individualized authority to choose an interpretive method for himself or herself, there are good policy reasons—such as those adduced in Section II.F below—for those free-agent judges to choose to follow prevailing state methods when interpreting state statutes.¹⁰⁸ 

*     *     *

In sum, although different readers may find slightly more or less guidance on interpretive methodology in the Constitution, the Constitution does not spell out every detail of an interpretive approach.  Nor does it totally exclude the use of state approaches, either of their own force or at the invitation of federal decisionmakers.  The Erie analysis therefore has to proceed to subconstitutional grounds.

B.   Federal Statutory Interpretive Directives

There are some federal statutes that tell courts how to interpret statutes, but they do not purport to apply to state statutes.  The Dictionary Act, for example, applies to the interpretation of “any Act of Congress.”¹⁰⁹   Congress has also enacted some targeted interpretive directives that apply to the particular federal statutes of which they are a part, but these directives do not apply to the interpretation of federal statutes in general, much less the interpretation of state statutes.¹¹⁰   Appearing to come closer to the mark, the statute implementing the Full Faith and Credit Clause requires federal courts to give state statutes “the same full faith and credit” as they receive in their state of origin,¹¹¹  but this command has heretofore not been understood to say anything useful about methods of interpreting state laws.¹¹² 

The question whether Congress has a dormant constitutional power to enact a statute prescribing a method of statutory interpretation poses some nice constitutional questions that I will only gesture toward here.  There is of course the separation-of-powers worry about legislatures telling courts how to do the job of saying what the law is.¹¹³   Further, a federal interpretive code that applies to state statutes raises federalism concerns too.  Rosenkranz tersely concludes that a federal statute governing the federal courts’ interpretation of state statutes in diversity cases would be unconstitutional, citing Erie,¹¹⁴  but that conclusion, though perhaps ultimately correct, is reached too hastily.  Even in cases that involve primary conduct that lies beyond federal regulatory jurisdiction, Congress still has its powers to create the lower federal courts and to regulate the Supreme Court’s appellate jurisdiction, plus the Necessary and Proper Clause.¹¹⁵   Congress’s interests in the administration of the federal judiciary could be implicated by state-law interpretive methodology in multiple ways, such as where following state methods requires burdensome inquiries or where departures from state methods would invite forum shopping.  So there is a basis for federal interest and, perhaps, future exertion of federal authority to regulate the interpretation even of state statutes.  Even in the absence of such a present exertion of authority, however, the possibility of federal judicial-administrative interests in the interpretation of state statutes should be kept in mind, because those interests may preempt state methods in rare instances, as more fully described in Section III.A below.

Additionally, it is worth noting that the second sentence of the Full Faith and Credit Clause grants authority to Congress to regulate the “[e]ffect” of state laws in other states.¹¹⁶   The power to specify the “effect” of state law in other states arguably extends to authority to regulate methods of interpretation and, together with the Necessary and Proper Clause and congressional power to administer the federal courts, represents an untapped source for future congressional regulation of federal interpretation of state law.¹¹⁷   Again, though, that would be future legislation, not legislation on the books now.

C.   The Rules of Decision Act and the “Relatively Unguided” Erie Analysis

Because there is currently no federal enactment that provides a federal method for interpreting state statutes, we must press on in the search for a resolution of the Erie choice between federal and state interpretive law.  This is the Rules of Decision Act track of the Erie doctrine, or what Hanna calls the “relatively unguided” Erie analysis.¹¹⁸   As the sections below will detail, I believe that this Erie problem should, generally speaking, be resolved in favor of using state interpretive principles.

In the existing literature, the primary approach to statutory interpretation’s Erie question could be described as “doctrinal coherentist.”  This mode of argument looks to how federal courts operate in several adjacent domains of state law, such as the interpretation of contracts and wills, where state interpretive methods are widely understood to apply.¹¹⁹   Coherentism is fine, and I use it myself, but it is not fully satisfying.  For one, the interpretation of contracts, wills, trust instruments, and other privately created documents is not so infused with public concerns as statutory-interpretive methodology is, so they are not perfectly analogous.  Furthermore, there are some other doctrinal spaces, also somewhat pertinent, where the federal courts do not so uniformly follow state methods.  An important example, highlighted by Varsava, is the interpretation of state judicial decisions (such as the rule for constructing holdings out of fractured opinions, factors used in an overruling analysis, and the like).¹²⁰   Another example, from Scoville, is the law of statutory severability, in which a post-Erie consensus in favor of state law has recently been upended in favor of a federal doctrine of severability for state statutes.¹²¹   Finally, although the current doctrine in all of these adjacent domains is, considered as a whole, more supportive than not of treating state statutory-interpretive methods as applicable in federal courts, the existing doctrinal patterns should themselves be subject to scrutiny for their agreement with the underlying principles.

For these reasons, the approach of this Article looks beyond doctrinal analogies and seeks to put the debate on a firmer foundation.  In doing so, I warn that I am going to give more attention to the RDA and its language than one often finds in an Erie analysis.¹²²   Admittedly, it may be that the statute is so obscure, or that the surrounding legal landscape has evolved so much since its enactment, that any attempt at interpreting its text is bootless.¹²³   Still, one reason to consider the RDA’s language is that an increasingly textualist Supreme Court may wish to do so.  Further, even if one regards the RDA as a botch or, at best, merely a declaration of what sensible federal courts would do even without the statute, the key phrases in the RDA provide an organizational structure that can aid exposition of the relevant choice-of-law considerations—whether those considerations are regarded as deriving from the RDA or elsewhere.  I aim to be ecumenical in what follows, not to preach the RDA as the sole source of wisdom.

Originally enacted as section 34 of the Judiciary Act of 1789,¹²⁴  the RDA currently provides that “[t]he laws of the several states, except where the Constitution or treaties of the United States or Acts of Congress otherwise require or provide, shall be regarded as rules of decision in civil actions in the courts of the United States, in cases where they apply.”¹²⁵   Erie famously interpreted the RDA to make obligatory on the federal courts the common law as declared by the judges of each state, such that federal courts must follow state decisions rather than apply their own understandings of a general common law.¹²⁶   Even before Erie, federal courts applied state statutes where applicable.¹²⁷   And when they did so, the federal courts generally applied the statutes as the respective state courts had interpreted them.¹²⁸   That duty of obedience to state interpretations of state statutes is even clearer today: a federal court applying a state statute applies as well the state high court’s on-point constructions of it, no matter whether the federal court would have decided differently as an original matter.¹²⁹   If the state supreme court says its “No Vehicles in the Park” statute prohibits tricycles in the park, the federal court will say the same.

So, then, the RDA’s reference to state “laws” encompasses state statutes, the state courts’ interpretations of their statutes, and (after Erie) the unwritten law as declared by the state high courts on topics both general and local.  It would seem that state interpretive methodology would have to be encompassed within one of those things, which would put it within the RDA’s purview.

Nonetheless, there are a few potential obstacles on the way to the conclusion that the RDA or the doctrines of choice of law and federalism surrounding it require federal courts to follow state interpretive methods.  Those include questions about whether states view their methodology as applicable in other courts or as “law” at all; the possibility that interpretive methodology is not a “rule of decision” but is instead something else, perhaps “procedure”; and the possibility that some other federal interest overrides an obligation to apply state law.  I consider all of these.

D.   “Where They Apply”—State Interests in Federal Interpretation of State Law

The RDA calls for use of state “laws . . . where they apply.”¹³⁰   This phrase is troublesome and, at least on the surface, unilluminating: apply state law where it applies.  The modern reader is inclined to understand it as referring to horizontal choice of law, that is, the choice of which state’s law applies to a question that has already been determined to be governed by the law of some state or another.  The phrase can helpfully illuminate vertical choice of law (i.e., Erie) if we understand the phrase as directing the federal court to consider a state’s position on the scope of its law, including whether its law is meant to apply in federal court.¹³¹   Specifically, the language can direct us to consider whether the state wants its interpretive principles to govern in federal court or instead regards them as applicable only in its own courts.  What the state wants for its law, or does not want, should matter for any doctrine that is rooted, at least in part, on concern for state sovereignty.  Readers who are skeptical of putting weight on the RDA’s “where they apply” language, or the RDA’s language more generally, should still care about state preferences about the scope of their methodology.  State views about the scope of state law are relevant in ordinary horizontal choice-of-law analysis.¹³² 

Some state law is not designed to apply in all courts, not even from the state’s point of view.  For example, a state statute or court rule that provides that “pleadings in the county superior courts shall be on 8½” x 14” paper” would not, by its terms, purport to govern the size of paper used in the federal district court sitting in the state.  And even without the text’s express limitation to a state trial court, state officials would not care what size paper the federal courts use, except perhaps for the very slight interest in making life a bit easier for resident attorneys and their assistants, by saving them from the need to buy two different sizes.  (Leave aside for now any federal interests in paper size, which might involve a balance of the same convenience to the local bar against the interest in national procedural uniformity.)

As against the slight (at most) state interest in the application of the paper-size rule outside of the state courts, consider by contrast that state authorities would strongly desire that their family laws, their laws governing the validity of contracts, and countless other laws apply equally in state and federal court and in sister-state courts too, at least in cases connected in the right way to the state or its people.  The point of those laws—to promote familial welfare, to promote commercial activity, and so on—would be frustrated if their intended benefits and burdens were realized only in the state’s own courts, where only some of the relevant litigation occurs.  As the reader might recognize, the enacting state’s intended scope of its law’s application, in particular whether the law is meant to apply outside its own courts, is one way of usefully rendering the troubled distinction between substance and procedure.¹³³ 

The examples of paper size and contract law mark out easy cases on opposite ends of a spectrum, but where does interpretive methodology fall?  Do the state authorities even mean for the federal courts (and, presumably, sister states too) to apply state interpretive law to the interpretation of the state’s statutes?  Or is it a matter of indifference, like the décor of a courtroom in a distant city?  It is hard to answer with certainty, but the answer is probably mostly that the state has a strong interest in the methods used to interpret its statutes in other courts, in particular in the federal courts that sit within it.  Here it is profitable to separately consider state methodology that stems from state statutory directives on the one hand from methodology that comes from the state courts on the other.

Start with legislatively enacted interpretive directives.  All states have some of these, sometimes quite narrow and facially dull (e.g., the singular shall be construed to include the plural) and sometimes quite broad and theoretical (e.g., interpret so as to effectuate the legislative intent).¹³⁴   What is the scope of application of these directives, as the state sees it?  There are several logical possibilities.  The interpretive directives might be state-statute-regulating-but-forum-independent; that is, they might regulate the interpretation of the state’s statutes in whatever forum those statutes happen to appear (enacting state, federal, sister state, Mars).  Or they might be enacting-state-court-regulating, that is, governing the enacting state’s own courts’ interpretive methods no matter what jurisdiction’s statutes they encounter.¹³⁵   Or they could be purely internally focused, limited to the interpretation of the enacting state’s statutes in the enacting state’s own courts.

We cannot determine the external scope of a state’s interpretive directives by observing the conduct of the state courts, for they do not act outside of their system.  Nonetheless, if we look at the text of the various state directives, which is the most conventionally powerful guide to their meaning,¹³⁶  they are not always explicit about their scope, but where they do address the matter, they are generally state-statute-regulating and apparently forum-independent.  That is, they say things like, “this Code”or “the statutes of this state” should be interpreted in the specified way,¹³⁷  or that a canon popular elsewhere does not apply to “the statutes of this State,”¹³⁸  or they direct how courts (without specifying which ones) should discern the state legislature’s intent.¹³⁹   Although the directives do not explicitly address themselves to the federal courts in particular, such directives are most naturally read as purporting to govern the state statutes wherever they go.  That inference gets a bit stronger when the directives are placed in the part of the state code dealing with the legislature rather than with the courts,¹⁴⁰  though that placement is not necessary to the conclusion.  Although each state would need to be considered individually, it is a fair generalization that state interpretive directives appear to be written to govern the interpretation of state statutes when they appear in federal courts, not just in the enacting state’s courts.

A hope to regulate the meanings of state statutes wherever they go would, moreover, be an entirely sensible aim for a state legislature to have.  The choice between federal and state methodology affects the real-world, outside-of-court conduct of members of the public.¹⁴¹   Incorrect interpretations of state law (incorrect by the state legislature’s lights, that is) could undermine the substantive policies that the legislature aimed to enact in matters touching its citizens and territory.  And, as Justice Harlan warned in Hanna, “debilitating uncertainty in the planning of everyday affairs” is the likely result of having “two conflicting systems of law controlling the primary activity of citizens.”¹⁴² 

The previous paragraph’s contentions about the real-world importance of interpretive methodology are empirical claims, but the soundest view of the world is that they are true.  A prudent attorney advising a client on a course of conduct governed by statutory law would consider not just the statutes on the books but how they had been interpreted and, where there was no settled interpretation, how the courts are likely to interpret them.  Specifically, how the courts of the enacting jurisdiction, above all others, are likely to interpret them.  That forecast requires consideration of, among other factors, interpretive methodology.  At least that is true when there are meaningful differences in interpretive methods, as there are in at least some cases.  Where different methods lead to different results, uncertainty over which method will later be used brings about the difficulties of planning that Harlan feared.  An attorney advising a multinational business on its compliance obligations under a new antidiscrimination statute, for instance, might give different advice depending on the expected rules about whether statutes are presumed not to apply to foreign offices of U.S. companies or how a jurisdiction understands the presumption against retroactivity.¹⁴³ 

When we turn to interpretive regimes and canons generated by the state courts rather than by the legislatures, matters are a bit harder to discern.  Even within their own court systems, some states may have little in the way of established methodology.  As courts become more self-conscious about methodology, the extent of methodological law may increase somewhat within those states where it is already rooted and spread to more states.¹⁴⁴   Where state courts have discernible propositions of methodological law applicable within their own courts, it is probably the case that the state authorities typically intend for their interpretive regimes to follow the statutes into federal court (and sister-state courts, for that matter).¹⁴⁵   Admittedly, it is hard to know for sure, as state courts tend not to have occasion to directly say whether federal courts are supposed to follow the state’s interpretive methods, for the obvious reason that the state judges decide cases only in their own court system.  Perhaps a curious federal court could use the device of certification to ask a state supreme court what it thinks about the scope question!  Nonetheless, even in the absence of solid evidence, we do have reason to think that the state courts want the federal courts to follow state methods.  When answering a certified question about the meaning of a state statute, the state courts answer by using the same interpretive methods they use in other cases.¹⁴⁶   As one Illinois Supreme Court decision on a certified question said: “Proper interpretation of an Illinois statute presents a question of Illinois law.  This court is the final arbiter of such questions.”¹⁴⁷   That statement is not limited to “in Illinois courts,” though we should not read too much into a remark that might not have been fully considered as regards this dimension.  At the least, state courts do not respond to certified questions by inviting the federal courts to use whatever methods they please when state statutes come before them.

Again, it makes sense that state courts (and state legislatures) would want their judicially announced methodology to follow the statutes.  That result serves the state’s legitimate interests in several ways: giving state statutes their correct meaning (as judged by the state, however it determines it) so that they can correctly govern primary conduct, promoting coherent interpretations across the breadth of state law regardless of forum, and providing a clearer interpretive regime against which the state legislature can draft.¹⁴⁸ 

The states’ view on the scope of their methodology is, I concede, an empirical question.  It is conceivable that a state with an interpretive method that is mandatory for its own courts would nonetheless want the federal courts to use whatever method the federal courts think is best.¹⁴⁹   Note as well the possibility that certain aspects of a state’s interpretive method could derive from the state’s forum-linked concerns.  As an example, imagine a state supreme court writing that “legislative history should be eschewed because it is too burdensome for the courts to research and our lower courts tend to handle it inexpertly.”  That reasoning does not apply outside the system, at least not with full force.  Similarly, some aspects of a state’s methodology might build on competencies lacking elsewhere.¹⁵⁰   To use legislative history again as an example, a state’s courts might use it as the most reliable way to achieve the courts’ stated goal of discerning legislative intent, yet the state might want for outsiders, less versed in the ins and outs of state legislative process and politics, to pursue that same goal of legislative intent through whatever means outsiders can most reliably wield.  (The same might be true of a state that uses corpus linguistics; perhaps outsider interpreters would do worse with a corpus than with a dictionary.)

Despite the qualifications in the preceding paragraphs, I feel confident saying that the overwhelming majority of states, the overwhelming majority of the time, will want the federal courts to follow the state courts’ interpretive methods for state statutes.

So, then, to wrap up this point, state interpretive methods do indeed, as a general matter, “apply” in federal court.  That is what state officials want, and they want it for sensible regulatory reasons.  That provides a strong reason—of a state-sovereignty-respecting sort, and even independent of the text of the RDA, for those who eschew it—for federal courts to apply the state’s methods.  This reason need not be determinative, of course, and subsequent Sections will consider federal interests that might militate for or against use of state methods regardless of state interests.  But knowing what the state wants is an important piece of the puzzle.

E.   “Rules of Decision,” “Substance,” and “Procedure”

Another way to resist the conclusion that federal courts must follow state interpretive methodology would be to contend that interpretive methods are not “rules of decision” to which the RDA applies.  They are instead, the argument would go, means of determining the rules of decision.  Or, to put the point in the fraught but familiar terminology, interpretive methodology could be called “procedural” in character, not “substantive.”  And so, according to this contention, interpretation must be governed by federal law, federal common law if necessary.¹⁵¹ 

Labeling exercises are no way to solve an Erie problem, but if intuitionistic categorization is where one wants to go, it seems to me that methods of statutory interpretation come closer to the substance pole than the procedure pole.  Let us say that procedure is the mode through which courts enforce substantive rights and duties, the “how” of the law.  It is true that methods of interpretation address a matter of “how.”  But, crucially, the “how” associated with interpretive methodology concerns how one determines the content of the law, not how a court goes about proving a violation of the law or enforcingwhatever judgment the law and facts yield. ¹⁵²   Interpretive methodology tells us how to decide whether a particular tort statute creates a law of negligence or a law of strict liability as a matter of the content of persons’ rights and duties; procedure (including the law of evidence), on the other hand, tells us how a person files a suit for redress under the theory the statute embodies (if a suit is even worth pursuing under the statute so understood), how a party gathers information from the other side, how the parties prove the facts the law makes relevant, and how a victorious plaintiff executes a money judgment.  In that sense, methodology is substance rather than procedure.

Nonetheless, one might draw some support for the view that an interpretive methodology is not a “rule of decision” by appealing to the old-fashioned notion that foreign law is a matter of fact, not law.  The traditional English view is to treat foreign law as a question of fact to be found according to the forum’s rules of evidence, such as rules about expert testimony and authentication of documents.¹⁵³   That conception of foreign law took root on this side of the Atlantic, and state courts treated the law of their sister states as “foreign” in the relevant sense.¹⁵⁴   If foreign law is regarded as a question of fact subject to proof through witnesses and other evidence, then it makes sense to say, as the Restatement (Second) of Conflict of Laws did, that “[t]he local law of the forum determines how the content of foreign law is to be shown . . . .”¹⁵⁵   This approach would fit with the typical practice of applying forum procedures even when a different jurisdiction’s substantive law governs.

The factual conception of foreign law, the characterization of sister-state law as foreign for this purpose, and the consequent recourse to forum evidence law to find its content—each of these things made some sort of sense in their original context, but they do not offer any useful guidance to a federal court interpreting a state statute today.  For one thing, federal courts never regarded state law as foreign in this sense.¹⁵⁶   And the states too have come to reject the old factual conception of sister-state law, either through legislation or by court decision.¹⁵⁷   Today, a court in California ascertains New York law, California law, and federal law in basically the same way: it reads the parties’ briefs, looks up the relevant statutes and cases on Westlaw, and maybe goes to the library for a treatise.¹⁵⁸   It will not take testimony from a retired New York attorney or require the plaintiff to offer into evidence an authenticated copy of the New York statutes.  Forum interests in regulating modes of proof are accordingly diminished.

Moreover, even if one embraced the old practice according to which determining the content of law was treated like factfinding, the forum’s interests involve matters like admissibility of testimony, authenticity of documents, and modes of proof, such as whether an expert is needed to explicate sources of law and what qualifications the expert must have.  Those interests do not tell the court what portions of foreign law must be proven.  Even a court following the fact conception should not stop at taking the statutory text into evidence but should also take evidence of the foreign interpretations of the text and, when needed, the foreign interpretive methods.¹⁵⁹   The interpretive methods are part of the relevant content of the foreign law that should, on the factual conception, be put into evidence.

Whatever one’s intuitions about what label to put on interpretive methodology, there is a deeper problem with the categorization exercise.  Namely, it is a mistake to assume that “rules of decision” in the RDA must mean something roughly approximating “substance” while excluding procedure and evidence.  The easy association of RDA “rules of decision” with substantive law probably derives at least in large part from the existence of the Federal Rules and the many federal procedural statutes in title 28 and elsewhere, which together provide most of the procedure applied in federal courts.  Thence the familiar shorthand that diversity cases involve state substantive law (per Erie and the RDA) and federal procedure (Hanna and the Rules).¹⁶⁰   But the existence of all this positive law of federal procedure may just show that these other federal directives take precedence over the RDA, as the RDA itself expressly allows,¹⁶¹  not necessarily that the term “rules of decision” is itself confined to some common understanding of substantive law.  Positive federal law like the Federal Rules and title 28 would still cover the same territory if we excised the phrase “rules of decision” from the RDA.  Put differently, to observe the full breadth of “rules of decision,” we would need to see how the RDA expressed itself in a counterfactual world without all of the other federal law that displaces, overlays, or otherwise obscures its command to apply state law.

We can get a slightly clearer view, though one that introduces some distortions of its own, by looking to the past, when there was less federal positive law to get in the way.  (For readers who are not particularly interested in either the RDA’s text or its historical understanding, the history I am about to recount will not be directly relevant but still helps to illuminate the relevant choice-of-law landscape.)  Looking to the past shows that “rules of decision” reaches into the twilight zone between substance and procedure and even reaches some topics that anyone would label procedure.

One of the leading early cases on the distribution of authority over federal-court procedure was Wayman v. Southard, which concerned the method of executing a federal court’s money judgment.¹⁶²   The Supreme Court held that the RDA did not require the application of the state law of execution, but the Court notably did not rely on a substance/procedure dichotomy.  The Court instead reasoned:

This section [i.e., the RDA] . . . has, we believe, been generally considered by gentlemen of the profession, as furnishing a rule to guide the Court in the formation of its judgment; not one for carrying that judgment into execution.  It is “a rule of decision,” and the proceedings after judgment are merely ministerial.  It is, too, “a rule of decision in trials at common law;” a phrase which presents clearly to the mind the idea of litigation in Court, and could never occur to a person intending to describe an execution, or proceedings after judgment, or the effect of those proceedings.¹⁶³ 

The reference to “a rule to guide the Court in the formation of its judgment” sounds much more capacious than “substantive law.”  Indeed, it is broad enough on its face to comfortably embrace interpretive methodology.

Other early cases treated the RDA as requiring the use of state law on topics that could easily be classified as procedural or evidentiary.  As the Court explained in one case, which concerned whether the federal court should require testimony authenticating the signature on a promissory note:

We do not perceive any sufficient reason for so construing [the RDA] as to exclude from its provisions those statutes of the several states which prescribe rules of evidence, in civil cases, in trials at common law.  Indeed, it would be difficult to make the laws of the state, in relation to the rights of property, the rule of decision in the circuit courts; without associating with them the laws of the same state, prescribing the rules of evidence by which the rights of property must be decided.  How could the courts of the United States decide whether property had been legally transferred, unless they resorted to the laws of the state to ascertain by what evidence the transfer must be established?¹⁶⁴ 

Other nineteenth-century cases concerned evidentiary matters like witness competency and privileges, holding that the relevant state laws fell within the RDA.¹⁶⁵   The view that the RDA required federal courts to follow state law extended to what we would now consider a matter of discovery practice, namely whether a court could compel a personal-injury plaintiff to undergo a physical exam.¹⁶⁶ 

Although these cases precede the Federal Rules of Evidence and Civil Procedure, their view of the scope of the RDA was not wholly free of the push and pull exerted by other federal enactments that might obscure or distort.  For in those early days there was other federal positive law that interacted with the RDA in complicated ways.  The 1791 Judiciary Act itself addressed several procedural matters,¹⁶⁷  and this federal law trumped any use of state law otherwise required under the RDA.¹⁶⁸   A greater source of opacity concerning the RDA’s scope comes from the Process Acts, which, using various terminology, borrowed certain existing state practices while also, in some iterations, licensing federal courts to make some of their own rules.¹⁶⁹   Thus, unless an opinion was precise, it is hard to tell whether a decision applying state law to some procedural topic was following the dictates of the RDA or the Process Acts or whether the court was unsure about which one applied.  This is not even to mention proceedings in equity and admiralty, where the RDA did not by its terms apply and where the federal courts, with Congress’s permission, went their own way on certain procedural matters.¹⁷⁰   Because there were multiple sources of procedural law in this messy system, it was not always important to be precise about the scope of the phrase “rules of decision” in the RDA.¹⁷¹   Moreover, even when federal courts did clearly state that the RDA was the reason for applying state law on some apparently procedural matter, their broad interpretation of the RDA may have been meant to compensate for deficiencies in the Process Acts.¹⁷²   

The discussion above concerns the RDA’s application to state law that existed primarily in statutory form, such as evidence codes.  In the world of Swift v. Tyson, the situation was quite muddled when it came to how federal courts responded to unwritten state law on arguably procedural matters.  To simplify, under Swift the federal courts followed state courts on matters of local law but made their own determinations on matters of general common law.¹⁷³   Some state unwritten law on arguably procedural topics was regarded as concerning matters of “general law” subject to the federal courts’ independent judgment.¹⁷⁴   Choice-of-law doctrine, for example, was in those days regarded as a subject for the general law, not a topic on which a particular state court’s view was controlling.¹⁷⁵   But other state procedural practices not embodied in statutes were regarded as matters of “local” law or had even acquired the status of “property,” either of which meant that the federal courts followed the state court’s understandings even under the regime of Swift.¹⁷⁶   Therefore, bringing the matter back to interpretive method, federal courts in the pre-Erie era were open to applying state interpretive modes in the then-rare instances in which they were shown a local interpretive method that differed from the presumably universal prevailing modes of interpretation.¹⁷⁷ 

F.   Outcome Determination, the Twin Aims of Erie, and Federal Interests in Using State Methodology

Another line of argument for employing state interpretive methodology is that it is “outcome determinative” for Erie purposes.  It is reasonably disputable how exactly outcome determination fits into the Erie analysis: as a gloss on “rules of decision,” as a way of distinguishing between “substance” and “procedure,” or something else.  I have already addressed the connection between interpretive methodology and state interests in regulating primary conduct, which is another way of getting at outcome determination.¹⁷⁸   The reader is welcome to put the following remarks about outcome determination into a different box, but here I want to use outcome determination—and its reformulation in terms of the “twin aims of Erie”—to explore the federal interests in using state interpretive methods in federal court.  As this Section will explain, federal courts would have good reasons to use state interpretive methods even if the states themselves were indifferent.

Guaranty Trust Co. v. York¹⁷⁹  is the case most closely associated with outcome determination as an Erie test, though its version was framed in later cases.  Holding that a federal court should apply a state statute of limitations to a state claim, Justice Frankfurter there wrote:

In essence, the intent of [the Erie decision] was to insure that, in all cases where a federal court is exercising jurisdiction solely because of the diversity of citizenship of the parties, the outcome of the litigation in the federal court should be substantially the same . . . as it would be if tried in a State court.¹⁸⁰ 

Now, as one can easily demonstrate, almost any triviality can turn out to be outcome determinative.  If a federal court dismisses a case because the plaintiff filed something on day twenty, based on the plaintiff’s belief that the document was governed by a twenty-one-day state deadline rather than a fifteen-day federal deadline, then this timing rule was outcome determinative.

Providing us a more useful test, Hanna v. Plumer famously reinterpreted outcome determination in terms of the “twin aims of the Erie rule,” those being the discouragement of forum shopping and the avoidance of the inequitable administration of the laws.¹⁸¹   Hanna thereby directs our attention away from the vagaries of a particular case that has gotten tangled up in some federal-state difference and instead prompts us to consider the difference between federal and state law from the ex ante perspective, in particular by asking whether the difference in laws promotes forum shopping or creates inequity based on the accident of citizenship.¹⁸²   The twin aims are federal goals (though not necessarily solely federal), and so achieving them can require the use of state standards even if the state does not care whether its law applies.¹⁸³ 

When a federal court interprets a state statute, achieving the twin aims means not only applying the state law in some abstract sense but applying it as the state courts have interpreted it and, if there is no extant interpretation, would interpret it.  Determining how a state court would interpret a statute prohibiting vehicles in the park may require considering, among other things, whether the state courts value dictionaries (and which particular dictionaries) over legislative history, whether a statute’s title is admissible as a source of meaning, and whether any substantive canons apply.  These methods give legal meaning to marks on a page.  Using the state’s methodology therefore follows from the familiar refrain that the job of the federal court is to “predict” the outcome the state high court would give.¹⁸⁴ 

One might accept the foregoing paragraph in principle but wonder whether mimicking interpretive methods is necessary to accurate prediction in practice.  It often is.  Methods of statutory interpretation have sufficient ex ante effects to count as outcome determining.  Admittedly this is hard to prove, because methods and even “rules” of interpretation are rarely as hard-edged as the difference between a two-year and a five-year statute of limitations.  But, at least for some states and in some respects, there are differences in interpretive approaches across jurisdictions.  Should statutes in derogation of the common law be narrowly construed, as an old canon would have it, or instead liberally to achieve the legislature’s beneficent purposes?¹⁸⁵   Is there a presumption against extraterritorial application of a jurisdiction’s statutes and, if so, how is the presumption overcome?¹⁸⁶   How much weight should be given to a statute’s title?¹⁸⁷   Courts tell us that different interpretive regimes matter, and one can find closely matched pairs of cases that come out differently under contrasting interpretive canons.¹⁸⁸ 

It is instructive here that horizontal choice-of-law decisions can be unpredictable too, yet they are considered outcome determinative for Erie purposes.  The Restatement (Second) generally calls for application of the law of the state with the “most significant relationship” to the case, which in turn requires consideration of a list of factors.¹⁸⁹   Some states use interest analysis, which may require consideration of the purposes and intended scope of state law.¹⁹⁰   Yet the vagaries of choice-of-law doctrine notwithstanding, federal courts must apply forum-state choice-of-law principles.¹⁹¹ 

We can construct parallel cases in the domains of conflict of laws and statutory interpretation.¹⁹²   In the conflict-of-laws scenario, suppose State X law would not allow punitive damages under its wrongful-death statute, while the law of the other interested state, State Y, probably would; suppose further that a state court in State X using State X’s forum-preferring conflicts methodology is more likely to choose its own law rather than Y’s law than is a court applying the principles of the Restatement (Second), which (suppose for this hypothetical) is the approach used by federal courts.  In the parallel scenario involving interpretive methodology, it is clear that the State X wrongful-death statute applies to the case, but suppose the State X statute would probably be interpreted not to allow punitive damages under State X interpretive methods (which include a canon of narrowly construing statutes in derogation of the common law).  Suppose further that the State X statute probably would be interpreted to allow punitive damages under prevailing federal interpretive approaches.  Without Klaxon, a well-counseled plaintiff in the conflicts scenario would have reason to prefer federal court in State X to state court, while a well-counseled defendant would prefer state court.  Likewise, without Erie-for-interpretive-methods, well-counseled plaintiffs and defendants in the interpretive scenario would have conflicting preferences over state vs. federal court.  If the parties are diverse, one gets forum shopping and the inequitable administration of the laws, and there go the twin aims.

G.   Characterization of Methodology as State Law Versus Federal Common Law That Borrows State Content

The Sections above argued for the use of state methodology when federal courts interpret state statutes, and that is the crucial practical point, but a complete analysis needs to acknowledge that there are two different ways federal courts can use state law.¹⁹³   One way is for state law to “apply” in federal court.¹⁹⁴   That is how Erie told federal courts to use state negligence law; the state law applied to the dispute because there was no federal law to govern it.  The other way is for the federal court to determine that a matter is governed by federal law but to “borrow” or “adopt” the content of state law as the federal common-law rule (or to give content to a federal statute), as opposed to fashioning an independent federal standard.¹⁹⁵   When a federal court assesses the preclusive effect of a prior diversity judgment, for example, the question is one of federal law, but the federal law generally adopts the preclusion law prevailing in the rendering state because doing so serves interests of the “twin aims” variety and there is little federal interest, in the run of cases, in crafting different principles of preclusion.¹⁹⁶   In such a case, the law is federal but it looks like state law.

When a federal court uses state interpretive methods for a state statute, as I say they should, is it applying or instead adopting state law?  I believe the better view is that state methodology, like the state statute itself, is genuine state law that applies in federal court.  There are several overlapping considerations that support this characterization.

To begin with, this characterization better aligns with the range of reasons for honoring the content of state law in federal court.  The reasons are not limited to federal interests like avoiding forum shopping.¹⁹⁷   Rather, the reasons are also rooted in state sovereignty.¹⁹⁸   State officials have legitimate authority to make and define their state law (within their proper domain) so as to govern real-world conduct.  They want federal courts to use state methods, and they want this because of those legitimate regulatory interests.¹⁹⁹ 

Treating interpretive methodology as state law also better aligns this corner of the law with federalism doctrines generally and preemption doctrine in particular.  To treat the law of statutory interpretation as federal law despite the state source of the statutes being interpreted would be field preemption.  It would be a domain in which state authority could not operate even without any demonstrated conflict between state and federal law.  Conflict preemption is instead the norm.²⁰⁰ 

My preferred characterization of methodology—genuine state law, not federal common law emulating state law—is consistent with both narrow and broad accounts of federal judicial lawmaking power.  On the narrow view of federal common law, federal courts may create federal common law only where the U.S. Constitution or a federal statute expressly invites it or implicitly requires it.²⁰¹   If one accepts the narrow view, state interpretive methodology would have to apply unless there is some federal directive inviting or requiring, as a general matter, the creation of federal common law for the interpretation of state statutes.  Now, as elaborated upon shortly, there may be federal guardrails that come into play to rule out or preempt certain state methods in rare instances.  To my mind, the best argument for a federal directive arguably inviting the creation of a federal common law of interpretation comes from the view that Article III confers an inherent judicial power over interpretive methods, though this view is usually regarded as vesting each judge with indefeasible power, which detracts from a characterization of the resulting federal methodology as common law as opposed to individual philosophy.²⁰²   The conferral of judicial power would be a most subtle invitation to create a preemptive interpretive methodology over state statutes, but even if it did so, a wise judge would have good reasons, federal reasons, of the “twin aims” type, to borrow relevant state law as the federal method.²⁰³ 

Even if one takes a broader view of the federal courts’ power to create common law, such as Louise Weinberg’s,²⁰⁴  one would still need to identify a national interest that justifies making federal-court interpretation of state statutes a question of federal law.  Notably, although Weinberg discerns only minimal limits on the power of Congress or the federal courts to make federal law, and no such limits coming from Erie itself, she does identify Erie as setting out the limitation that federal courts cannot make state law.²⁰⁵   But to interpret the law is to make it, at least around the edges, and so using federal interpretive principles to interpret what is avowed to be state law is for the federal court to make state law, at least around the edges, which is what Weinberg says the federal courts may not do.

For the reasons above, the better view is that state interpretive methodology applies to federal-court interpretations of state law.  But even if that is wrong, it is important to recognize a feature shared by both the application and the adoption positions.

In particular: whichever characterization one adopts, the federal courts need not apply or adopt every jot and tittle of state interpretive methodology.  On the application view, state law is still subject to occasional preemption as required (including implicitly) by the Constitution or federal statutes.  On the adoption view, the law is federal anyway and its content will mimic the state content only when it serves federal interests to do so.  On either understanding, that is, federal courts could use a blended technique that largely looks like the relevant state’s method but departs from it in those unusual circumstances when the state method either conflicts with some federal directive or impairs federal interests.

This proposition is very familiar in the adoption scenario, as the Supreme Court often emphasizes the conditional and partial nature of the incorporation of state law.²⁰⁶   But blending can also occur in the application scenario.  In the ordinary horizontal choice-of-law context, the Restatement expressly endorses an issue-by-issue approach to choosing governing law.²⁰⁷   Thus, a court handling a car crash might apply the negligence law of the place of the crash, the interfamilial immunities of a state of the parties’ shared domicile, and the contract law of a third state in which a party had signed a waiver of liability—not to mention the forum’s own law of procedure.²⁰⁸   Even more to the point, in vertical choice of law, the rules of contract interpretation, which ordinarily come from state law, become a state-federal blend when federal interests come into play, as when an agreement to arbitrate must be interpreted in light of the national policy favoring arbitration.²⁰⁹  

It is now time to turn to discerning, more specifically, those circumstances when federal law may assert itself in the interpretation of state statutes.

III.     EXCEPTIONS AND SPECIAL CASES

For the reasons stated in Part II, federal courts should generally use state interpretive methods when interpreting state statutes.  That is the basic rule.  Nonetheless, departures from state methods are sometimes appropriate.  This Part explores those exceptional circumstances, and, finally, offers a few elaborations on the basic rule.

A.   Exceptions to the General Rule That State Methodology Applies

1.   Potential Exceptions for Federal Judicial-Administrative Interests

The federal courts and the national government more generally have legitimate interests in the just and efficient operation of the federal courts.  Some of these interests find explicit statement in positive law, but we have already seen that the federal Constitution and enactments do little to guide federal interpretation of state statutes.²¹⁰   Other federal interests in judicial administration derive from implication.  In Byrd v. Blue Ridge, a state practice in which judges decided a particular question of state law was overcome by the federal pro-jury policy emanating from, even where not directly commanded by, the Seventh Amendment.²¹¹   The Court in Byrd was willing to use juries in federal court even though it ran the risk of some divergence in outcomes and attendant forum shopping.²¹²   Similarly, the federal courts have generally chosen a federal judge-made standard for forum non convenienseven in diversity cases, in part because the forum non conveniens factors include court-centered interests in docket control and the administration of foreign law.²¹³ 

It is conceivable that some state interpretive methods could compromise federal interests in judicial administration.  Some states, most famously Utah, have started to use the techniques of corpus linguistics in their statutory interpretation decisions.²¹⁴   Corpus linguistics can act as a complement or alternative to dictionaries as a tool for ascertaining the meaning of words.²¹⁵   If the Utah Supreme Court or legislature establishes corpus linguistics as a mandatory part of its interpretive practice, a practice that is meant to follow its statutes wherever they go, should federal courts applying Utah law follow that method?  Or suppose that Hawaii starts to conduct legislative debates and produce legislative history solely in the Hawaiian language.²¹⁶   And suppose further that Hawaii state courts consider legislative history and regard it as contributing to meaning.  Must the Ninth Circuit continue to use legislative history in Hawaii cases, if it means dealing with what would be a foreign language to nearly all of its judges and clerks?

Despite the presumptive yes to following state methods, even unfamiliar ones, these examples show that there are competing federal interests to consider.  Corpus linguistics requires training and skill if one is to do it competently.²¹⁷   Trying to do it without the skills and resources may be worse, even from the perspective of its practitioners, than not trying at all.  Suppose the Utah Supreme Court, having devoted itself to corpus-linguistics-based textualism, decides that all state judges and law clerks need forty hours of training in corpus analysis (available from consulting outfits) and that each of the justices must hire one law clerk each year who has graduate training in corpus linguistics.²¹⁸   The judges of the Tenth Circuit do not need to follow suit, do they?  Suppose that Brigham Young University starts charging for access to the various corpora it maintains.  The Utah Supreme Court cancels its subscription to some legal periodicals and out-of-state case reporters in order to pay the annual fee.  The Tenth Circuit would need to make similar sacrifices, or increase filing fees by two dollars per appeal, to pay for the corpus.  Must the court do so?

Without pretending to certainty, it is at least plausible that the answer to the questions just posed is that the federal court does not have to emulate the burdensome state method.  There are federal judicial-administrative interests in economy, convenience, and access that are capable of overcoming the usual duty to follow state law.²¹⁹   Indeed, if the state method is hard for outsiders to use, bearing the methodological burdens might not even yield a great result in terms of the accurate determination of state law.  This prospect of federal pain with little state gain provides a reason for the federal court to double-check whether the best understanding of enacting-state law really does call for outsiders to use the burdensome methodology or if instead those aspects are only interpreter-relative means of finding the law, helpful for enacting-state courts but not suitable for others.²²⁰ 

To be clear, the mere fact of some difficulty or unfamiliarity with state methods does not justify ignoring them.  This difficulty is a version of the omnipresent problem of making sense of another system’s legal culture and language.  The law of Louisiana and Puerto Rico derives from the Continental civil law.  In Texas, they call the document that commences a civil suit the “original petition” instead of the “complaint.”²²¹   Strange, right?  But federal courts figure it out.  Federal district judges usually practiced in the forum state, and circuit judges get familiar with the law and lingo of the states within their circuit.  Justice Scalia’s worry in Shady Grove about the difficulties of finding and reading state legislative history has some validity in theory but would not cut it, as a general matter, as a basis for ignoring legislative history if state courts use it.²²²   The Fifth Circuit tries to apply Louisiana’s civil-law methodology in cases governed by Louisiana law,²²³  and that system probably presents more challenges than the usual intersystemic encounter.  A befuddling state-law puzzle, statutory-interpretive or otherwise, might justify a federal court’s decision to certify to the enacting state’s courts, transfer to another federal court with greater familiarity,²²⁴  or, in the truly exceptional case, employ some variety of abstention.  But the existence of some difficulty does not justify a general practice of adhering to federal methods for state statutes.²²⁵ 

If Utah (or another methodologically adventurous state) were upset that the federal courts and sister states were compromising the integrity and uniformity of its law by failing to hire Ph.D.s or were burdening its high court with too many certification requests, it would not be without recourse.  It could create a state-funded corpus linguistics expert to advise other courts or abandon its mandatory burdensome interpretive method.  Being an outlier may have benefits, but it has costs too.

If federal procedural or judicial-administrative interests lead federal courts to depart from state methods, state courts would not have to honor those federal departures when interpreting their state statutes in their own courts.  True, a defining feature of the post-Erie “true” federal common law governing fields like interstate resource disputes or U.S. government contracts is that it is honest-to-goodness supreme federal law that preempts state law and applies in state courts.²²⁶   But there is also plenty of “procedural” federal common law—on topics like remittitur and abstention, for instance—that governs federal courts but does not apply in state courts.²²⁷   Because the occasional use of distinctively federal methodology for state statutes would be driven by federal court-centered concerns, this federal law would not purport to apply in state court.  The situation would be like Byrd, where federal interests required jury decisionmaking on the employee defense in federal court but the South Carolina courts were welcome to assign the question to the judge.²²⁸ 

2.   Potential Exceptions Rooted in the Nature of the Federal Judiciary

Another ground for departing from the usual rule arises from certain limitations of the federal judiciary.  State constitutions and institutional arrangements differ from the federal model, and those differences could license (or demand) important differences in interpretive approach.  The institution of judicial elections presents an obvious example of an institutional difference that complicates standard accounts of judicial restraint and faithful agency.²²⁹   An elected state judiciary might take advantage of its popular pedigree to adopt an interpretive approach that would be problematic for Article III judges.  And, elected or not, state judges are likely to have, by virtue of their geographic, social, and biographical proximity to the legislature, greater understanding of and opportunity for dynamic interbranch dialogue.²³⁰ 

Whether federal judges are legally or practically incapable of following state methods depends, of course, on the nature of state methods.  As things currently stand, some states have engaged in interesting interpretive projects, but they have generally not adopted anything particularly radical.  Nonetheless, there are some doctrines and circumstances in which state methods present particular challenges for federal courts, as the following pages explain.

a.   Constitutional Avoidance as a Blended Canon

Although different cases formulate and apply the avoidance canon somewhat differently, the most common statement of the federal doctrine is to the effect that courts should read statutes so as to avoid serious doubts about a statute’s constitutionality so long as there is a plausible reading of the statute that avoids the difficulty.²³¹   All or nearly all states have some sort of constitutional-avoidance canon for interpreting their own state statutes.²³²   Yet different jurisdictions may formulate the canon differently with respect to matters such as how “serious” a doubt must be to trigger avoidance and how “plausible” (or “reasonably possible” or some similar description) the avoidance interpretation must be to accept it.²³³   Some differences in the formulation of the canon likely involve mere differences in verbiage without any substantial difference in content.  But where there are actual differences in avoidance doctrines across jurisdictions, intersystemic cases present a tricky choice-of-law puzzle.  In particular, when a federal court entertains a constitutional challenge to a state statute, should it apply federal avoidance, state avoidance, or something else?

The answer is complicated because the avoidance canon has multiple justifications, some of which are descriptive and some normative, some of which are legislature-centered and some court-centered.  The canon is sometimes said to reflect the legislature’s desire not to violate higher law or its presumed preference to have some law rather than none.²³⁴   Whatever one thinks of the persuasiveness of those justifications, they depend on which legislature is involved.  Whether any particular legislature likes the avoidance doctrine (and what version of it) is an empirical question.  Federal courts can presume that a state’s version of the avoidance doctrine, as applied in its courts, reflects the answer to that empirical question.  This provides federal courts with one reason to follow state avoidance doctrines when considering challenges to state statutes.

But a federal court cannot stop there, for there are also court-oriented considerations, and they can constrain the federal courts’ use of avoidance from two directions.  On the one hand, federal courts may be required to use some minimally robust form of avoidance doctrine, even if the state has a very weak one or none at all, because doing so restrains the awesome power of judicial review by unelected federal judges.²³⁵   State courts, empowered by different constitutions, do not necessarily have to be so reticent to engage in constitutional adjudication.²³⁶   On the other hand, there may be court-specific restrictions on how aggressively a court may use avoidance, lest a saving construction veer into “rewriting” that is incompatible with the relevant court’s grant of judicial power.²³⁷   Because those are court-oriented justifications and limitations, they depend on the identity of the court rather than the source of the statute.  State courts, with their common-law powers and (in some states) electoral pedigrees, may well enjoy more latitude to “rewrite” than their federal colleagues.  The foreign source of the law may add another layer of reticence in an intersystemic encounter, as a matter of law or at least of practical competence.  Still further variations could arise depending on whether the constitutional challenge to a statute stems from the U.S. Constitution or the state constitution.

Combining all of the above considerations does not yield an obvious answer for the form of avoidance that federal courts should use for state statutes, but it is plausible that the court-focused considerations mean that the enacting state’s canon may not always accompany the state statute into federal court.  To construct a hypothetical, consider a very aggressive state avoidance doctrine according to which the state courts, when faced with any hint of constitutional trouble, should adopt a construction of a statute that avoids that hint of trouble even if the resulting construction is an implausible-but-barely-possible reading of the statute.  Such a rule could be one of those rare rules of interpretation that, even on a latitudinarian view of permissible methods, is incompatible with Article III judicial power or otherwise impermissible for federal courts.²³⁸   And even if a state avoidance canon does not go to the extreme of violating Article III, it may be that some exercises of avoidance involve sufficiently discretionary redrawing of state law that it would be better—from the point of view of institutional competence and accountability—for the state courts to make the choices about how, if at all, to fashion a saving construction.  Examples could be challenges to state statutes analogous to the federal statutes in Zadvydas and Skilling, in which the Supreme Court imposed a limitation on the scope of an otherwise excessively broad or unbounded statute, rather than just choosing one of two meanings for an ambiguous term.²³⁹ 

Federal constraints on aggressive rewriting of state statutes may explain, and possibly justify, the federal courts’ occasional statements, described in subsection I.C.2, to the effect that they have less leeway than state courts to render saving constructions of state statutes.  These federal courts are on to something when they worry about their competence to exercise the discretion that is sometimes involved in creating a limiting construction of a vague or overbroad statute.  Still, a decision to use anything other than the state version of the avoidance canon should be rare.  The federal interests in judicial-interpretive restraint would have to be strong enough to overcome the state legislative interests presumably reflected in the hypothetical state doctrine and overcome the ever-present federal “twin aims” that push toward harmonization.²⁴⁰ 

In what should be the rare instance in which a federal court feels incapable of or inadequate to the task of wielding avoidance as a state court might, the best course will often be to certify to state court to give that court the opportunity to produce a saving construction the federal court does not think it could announce.  In one Second Circuit case, Judge Calabresi suggested that the federal court might not be allowed to use avoidance as aggressively as the state, or at least that it could not reliably decide whether the state court would be able to interpret around the constitutional difficulty.²⁴¹   Whether or not he was correct about the federal court’s ultimate limitations as interpreter of state law, I agree that certifying the question was better than interpreting the state statute in a way that the federal court believed would differ from the state’s interpretation.

b.   Dynamic Interpretation

Federal courts sometimes say, in all kinds of cases, not just statutory cases, that they should not “innovate” in state law or recognize “new” state claims.²⁴²   Such statements could be understood in a few different ways, some more defensible than others.

If the federal court means that a decision recognizing the plaintiff’s “new” claim would be a bolt out of the blue, it sounds like the court is saying that its best Erie guess is that the state court would not recognize the claim.²⁴³   That is fine.  There is no need for federal courts to dress up their guess with solemn language about judicial federalism or judicial restraint.

If instead the federal court has a strong basis to believe the state would expand liability beyond existing fact patterns, or implicitly has done so, but nonetheless refuses to go along, such reticence would likely be unjustified.²⁴⁴   Among other things, it undermines the state’s sovereign interest in having the best view of its law prevail in federal courts.²⁴⁵ 

Finally, statements about reluctance to innovate could accompany genuine uncertainty about the state of state law: whether the state high court will overrule a musty precedent that is out of step with current trends, how it would define the elements of a new tort that is likely to be recognized in some form, etc.  As an outsider, the federal court may feel unable to predict exactly what the state courts will do and feel anxious, given its lack of state democratic bona fides, about marking out a course.  The classic case of Li v. Yellow Cab Company of California furnishes a useful example.²⁴⁶   There the California Supreme Court overruled its longstanding interpretation of the state civil code in order to substitute comparative negligence for contributory negligence.²⁴⁷   Making that same shift might have been imprudent for a federal court lacking intimate familiarity with state tort law, jury practices, and legislative preferences, all of which likely contributed to the state court’s decision.²⁴⁸   And although the state courts could correct a mistaken Erie guess, an error on such a frequently litigated topic could be consequential in the interim.  In these hard cases, it may be advisable to draw on certification or other techniques of avoidance.  But when the merits must be reached in a knife’s-edge case, breaking a tie in favor of stasis rather than innovation is probably more defensible than the contrary presumption.  A federal presumption of stasis fits with the conservative nature of courts generally and recognizes the relatively greater competence (technical and political) of state courts on matters of state law.²⁴⁹   If a presumption in favor of stasis has an effect on dockets, it would probably be to encourage (at the margin) plaintiffs who desire innovation to file in state court, but this effect would likely be approximately cancelled out by countervailing incentives for defendants to remove to federal court.

B.   Special Topics and Applications

The foregoing Sections have established that the federal courts should, as a general matter and subject to certain exceptions, use state interpretive methods when interpreting state statutes.  This Section addresses some additional details and special topics.

1.   State Methods for Federal Statutes in Federal Court?

Reading the RDA to require application of state interpretive principles might be thought to lead to a troubling conclusion regarding the interpretation of federal statutes.  In Campbell v. Haverhill, the Supreme Court held that a state statute of limitations governed a federal patent claim that lacked a congressionally prescribed limitations period, remarking that “the [RDA] itself neither contains nor suggests . . . a distinction” between diversity and other suits.²⁵⁰   And Campbell and the text of the RDA aside, the Erie doctrine is generally thought to require application (or borrowing) of state law in some circumstances in nondiversity cases.²⁵¹   And although a few federal statutes contain their own interpretive directives, and a few generally applicable federal statutes provide presumptive definitions, there is nothing like a generally applicable federal statute directing interpretive methods.²⁵²   Therefore, could it be that my argument requires application of state interpretive principles even for the interpretation of federal statutes in federal court?

No, and we can reach that answer in at least two ways.  The first route starts from the state side.  The puzzled-over last clause of the RDA calls for application of state laws only “where they apply.”²⁵³   Therefore, even setting aside any superseding federal directive, one should consider whether state laws even purport to “apply” to the interpretation of federal statutes in federal court.²⁵⁴   They probably do not.  Recall that state interpretive codes are generally written in a state-statute-oriented way, as governing how the statutes “of this State” are to be applied or how “the General Assembly’s” intent is to be discerned.²⁵⁵ 

Nonetheless, suppose that state judges or legislators want their canons applied as widely as possible, because they think their substantive policies and the canons supporting them are good policies for everyone or because they want to provide simplicity for members of the state bar.  If that is the state officials’ goal, they could not achieve it, and this is the second route to the conclusion that federal courts apply federal interpretive methods to federal statutes.  There is a strong federal interest in the interpretation of federal statutes being both uniform and compatible with federal interests.²⁵⁶   Part of the point of Congress legislating on a topic rather than leaving the topic to the states is to establish its desired policy on the topic and to make that policy uniform across the country.  The interest in geographically uniform interpretations of the statutes leads to an interest in geographically uniform interpretive approaches to the statutes.  Interests in the uniformity and supremacy of federal law also form part of the justification and explanation for the existence of the Supreme Court and its appellate jurisdiction over state high courts and inferior federal courts.²⁵⁷   Now, it is true that we do not have uniformity of interpretive approaches among federal judges, nor do we have uniformity of outcomes.  Circuit splits on interpretations of federal statutes are routine.  But at least we do not currently have to contend with intentional deviations as one might get under a system that licensed fifty different interpretive regimes for one code.

2.   Statutory Interpretation’s Horizontal and Diagonal Erie Problems

Consider a complication that arises when we add multiple states to the mix.  Suppose the federal court sits in State X.  Suppose that a State X state court would, under its choice-of-law rules, apply a statute of State Y as the governing law for a particular case.  Under Klaxon, the federal court in State X is supposed to apply State X’s choice-of-law rules, so the federal court should likewise apply the State Y statute.²⁵⁸   Now, which state’s interpretive rules should the federal court use on the State Y statute?  Let us further suppose that State X courts believe, as some actual courts do, that they should apply State Y law as State Y would, namely by applying State Y’s interpretive methods.²⁵⁹   If State X takes that view, then it is easy to see that the federal court should use State Y interpretive rules in reading the State Y statute.  Doing so both respects State Y’s sovereignty interests and advances the twin aims of Erie by promoting uniformity between the state and federal courts within State X.

Now comes the complication.  Some states, and let us now suppose State X is one of them, say that they can use their own interpretive methods when applying a sister state’s statute.²⁶⁰   What should the federal court sitting in State X do then?  The need to respect the enacting state’s regulatory authority counsels in favor of the federal court applying State Y interpretive rules to the State Y statute.  But doing so in this instance would conflict with the policy of vertical uniformity between state and federal courts in the same state, which policy points toward the federal court acting like the State X courts and applying State X methods to the State Y statute.

One solution to this problem is to require, as a matter of federal law, that State X use State Y methods when interpreting a Y statute.²⁶¹   Existing doctrine does not yet clearly establish such a requirement, but one can find the building blocks for at least a weak version of it in the Supreme Court’s Full Faith and Credit jurisprudence.  That jurisprudence does not render unconstitutional the mere misconstruction of sister-state law, but it does prohibit state courts from misconstructions that “contradict law of the other State that is clearly established and that has been brought to the court’s attention.”²⁶²   It is plausible that the considerations of state sovereignty and regulatory authority that drive the (minimal) duty of fidelity toward sister-state law extend as well to fidelity to sister-state interpretive methods, which give the law its content.²⁶³ 

A federal duty to apply sister-state interpretive methods would be unnecessary if all states took the view that their own state choice-of-law doctrines call for the use of enacting-state methods when construing enacting-state law.  That is likely the correct view for most approaches to choice of law and most divergences in interpretive practices across states.²⁶⁴   To the limited extent state courts have directly considered the matter, many of them adopt this position.²⁶⁵   If state courts apply sister-state methodology, the federal courts can avoid a tough Erie/Klaxon choice between preserving intra-state uniformity and respecting enacting-state regulatory interests.

3.   The District of Columbia and the Territories

The analysis is a bit different when it comes to the District of Columbia and the territories.  The RDA probably does not include the District of Columbia and the territories as “states,” the laws of which federal courts must apply.²⁶⁶   One finds the occasional statement that “federalism” requires adherence to the decisions of the local D.C. courts or the territories,²⁶⁷  which is odd because the federal government is not actually competing with an independent sovereign in those places.  Nonetheless, one should get to the same outcome—follow the local interpretive method—through an interpretation of the organic acts setting up the territorial judicial systems or, if from nowhere else, from sound policy considerations of the “twin aims” variety.²⁶⁸   In sum, federal courts should follow the interpretive methods of D.C. and the territories when their statutes come into federal court.  That general directive is, again, subject to deviations as set out in Section III.A.

CONCLUSION

When it comes to statutory interpretation’s Erie problem, the federal courts are doing well.  They generally should follow state interpretive principles in interpreting state statutes.  And they generally do so, especially the lower federal courts.  Furthermore, the courts are more likely to get to the right answer the more they think about it; that is, most errors are the result of inattention.  This means that the duty to apply state methods is likely to become more consistently observed over time, as courts become increasingly self-conscious about interpretive methodology as a law-governed activity.  All of that is good news.

The duty to follow state methods is not absolute, however, and certain topics can be expected to cause difficulty.  One problem area is the doctrine of constitutional avoidance, which blends federal and state interests.  Another difficulty, though the federal courts have not yet confronted it in an acute way, would arise from a state that adopts an interpretive methodology that is constitutionally permissible but too hard for federal courts to administer.  In subsection III.A.1, I posited the need for expertise in corpus linguistics or another language as potential examples.  Such cases would be challenging because they would pit federal interests in judicial efficiency against respect for state autonomy and the avoidance of forum shopping.

Those are difficult cases, but we should not lose sight of the forest for the trees.  In the more typical case, all of the relevant considerations point toward the same answer: federal courts should apply state interpretive methods when they encounter state statutes.  In sum, the usual answer to this Article’s titular question of how to read a state statute in federal court is, The same way the state court would.

---

© 2022 Aaron-Andrew P. Bruhl.  Individuals and nonprofit institutions may reproduce and distribute copies of this Article in any format at or below cost, for educational purposes, so long as each copy identifies the author, provides a citation to the Notre Dame Law Review, and includes this provision in the copyright notice.

* Rita Anne Rollins Professor of Law, William & Mary Law School.  I received helpful comments from Michael S. Green, Tara Grove, Anita Krishnakumar, Ethan Leib, Zachary Pohlman, Martin Redish, Nina Varsava, Ernest Young, and participants at the Yale Legislation Roundtable, the July 2021 Civil Procedure Workshop, the Georgetown Legislation Colloquium, and a William & Mary roundtable on judicial federalism.  I thank Evan Baines, Brendan Clark, Christopher Collins, Stephanie Cone, Ethan Gaitz, Damian Gallagher, Sarah Morain, Sarah Nagle, Mason Shefa, and Yifei Xu for research assistance.

¹ Erie R.R. Co. v. Tompkins, 304 U.S. 64, 78–79 (1938).

² Marbury v. Madison, 5 U.S. (1 Cranch) 137, 177 (1803); Antonin Scalia & Bryan A. Garner, Reading Law: The Interpretation of Legal Texts 53–55 (2012).

³ See Scalia & Garner, supra note 2, at xxvii (explaining that the law is full of interpretive conventions that make legal interpretation “more than just a linguistic exercise”).  See generally William N. Eskridge Jr., James J. Brudney, Josh Chafetz, Philip P. Frickey & Elizabeth Garrett, Cases and Materials on Legislation and Regulation: Statutes and the Creation of Public Policy 1151–71 (6th ed. 2020) (listing canons of statutory interpretation).

⁴ See, e.g., Powell’s Books, Inc. v. Kroger, 622 F.3d 1202, 1209 (9th Cir. 2010) (explaining that “[t]he Oregon approach [to statutory construction] contrasts with the standard federal statutory construction”); Eskridge et al., supra note 3, at 559–65, 710–12, 1071–74; Judith S. Kaye, State Courts at the Dawn of a New Century: Common Law Courts Reading Statutes and Constitutions, 70 N.Y.U. L. Rev. 1, 19–34 (1995).

⁵ See Morrison v. Nat’l Austl. Bank Ltd., 561 U.S. 247, 261 (2010).

⁶ See William S. Dodge, Presumptions Against Extraterritoriality in State Law, 53 U.C. Davis L. Rev. 1389, 1405–13 (2020) (citing variation across states).

⁷ Va. Code Ann. § 1-247 (2022); Davis ex rel. Woodside Props., LLC v. MKR Dev., LLC, 814 S.E.2d 179, 183 (Va. 2018).

⁸ E.g., Lee v. Norfolk S. Ry. Co., 802 F.3d 626, 631 (4th Cir. 2015).

⁹ Cf. In re Bowman, 582 B.R. 899, 903, 904 n.11 (Bankr. E.D. Va. 2018) (citing bill summaries regarding Virginia statute without acknowledging Virginia’s statutory directive).

¹⁰ See Thomas R. Lee & Stephen C. Mouritsen, Judging Ordinary Meaning, 127 Yale L.J. 788, 796 n.23, 867 (2018) (describing this development).

¹¹ See, e.g., Abbe R. Gluck, Intersystemic Statutory Interpretation: Methodology as “Law” and the Erie Doctrine, 120 Yale L.J. 1898, 1901, 1924, 1931 (2011); see also infra Section I.A (citing commentators taking this view).

¹² See, e.g., Sydney Foster, Should Courts Give Stare Decisis Effect to Statutory Interpretation Methodology?, 96 Geo. L.J. 1863, 1863 (2008); Abbe R. Gluck, The Federal Common Law of Statutory Interpretation: Erie for the Age of Statutes, 54 Wm. & Mary L. Rev. 753, 796 n.157 (2013); Nicholas Quinn Rosenkranz, Federal Rules of Statutory Interpretation, 115 Harv. L. Rev. 2085, 2108, 2152–53 (2002).  But cf. Nina Varsava, Stare Decisis and Intersystemic Adjudication, 97 Notre Dame L. Rev. 1207, 1209–10 (2022) (explaining that, on some jurisprudential theories, federal courts need not follow the state method of interpreting state common law and observing that a similar argument would apply to state methods of statutory interpretation).

¹³ See infra text accompanying notes 214–18 (addressing corpus linguistics).

¹⁴ Gluck, supra note 12, at 791.

¹⁵ Gluck, supra note 12, at 758, 780; see also id. at 759 (stating that “federal courts do not seem to consider Erie as relevant to the choice of statutory interpretation methodology”); id. at 770 (“[F]ederal courts generally do not view themselves obligated by Erie to apply state methodology when interpreting state statutes.”); Gluck, supra note 11, at1901 (“[F]ederal courts do not seem to think of statutory interpretation methodology as ‘law’ in the first place, much less as law subject to Erie.”); id. at 1924 (calling federal courts’ approach “wholly inconsistent”); id. at 1931 (“The Fifth Circuit is apparently the only court expressly and consistently to hold that Erie requires it to use state methodology for state statutes in diversity.”); Abbe R. Gluck, Justice Scalia’s Unfinished Business in Statutory Interpretation: Where Textualism’s Formalism Gave Up, 92 Notre Dame L. Rev. 2053, 2064–65 (2017) (stating that federal courts “do not seek out state interpretive rules when they sit in diversity and otherwise apply all other types of state law”); Abbe R. Gluck, Statutory Interpretation Methodology as “Law”: Oregon’s Path-Breaking Interpretive Framework and Its Lessons for the Nation, 47 Willamette L. Rev. 539, 555 (2011) (“[F]ederal courts do not typically look to state interpretive principles when they interpret state statutes.  Instead, they usually cite only U.S. Supreme Court cases for the interpretive principles that they choose—even though in some cases it is clear that the state supreme court would have employed a different interpretive rule.”).

¹⁶ Bryan A. Garner et al., The Law of Judicial Precedent 613 (2016) (emphasis added); see also Suzanna Sherry, Wrong, Out of Step, and Pernicious: Erie as the Worst Decision of All Time, 39 Pepp. L. Rev. 129, 141 (2011) (stating that “federal courts disagree among themselves about important questions such as whether they should follow state rules of statutory interpretation”); J. Stephen Tagert, Note, To Erie or Not to Erie: Do Federal Courts Follow State Statutory Interpretation Methodologies?, 66 Duke L.J. 211, 232–34 (2016) (concluding that federal courts in the Sixth Circuit generally fail to follow Michigan’s interpretive methods).

¹⁷ Foster, supra note 12, at 1884 & n.119.

¹⁸ 2 Norman J. Singer & J.D. Shambie Singer, Statutes and Statutory Construction § 37:1 (7th ed. 2009).  A subsequent section of the same treatise presents the matter as subject to conflict and even suggests the old-fashioned view that out-of-state law must be proven as if a question of fact.  Id. § 37:5 (“Although there is authority that the rules of statutory construction prevailing under the laws of the forum are followed, the rules of the state in which the statute was enacted should be followed if they have been pleaded and proved.” (footnote omitted)).

¹⁹ Aaron-Andrew P. Bruhl, Eager to Follow: Methodological Precedent in Statutory Interpretation, 99 N.C. L. Rev. 101, 103 n.1 (2020) (citing commentators endorsing this view).

²⁰ See id. at 126–59; see also William Baude & Stephen E. Sachs, The Law of Interpretation, 130 Harv. L. Rev. 1079, 1082–83 (2017) (emphasizing that interpretation is a law-governed activity, not just a linguistic or policy-based activity); cf. Christopher J. Baldacci, Note, The Common Law of Interpretation, 108 Va. L. Rev. 1243, 1243, 1255–75 (2022) (providing evidence for a “soft,” incremental form of methodological precedent).  But cf. Richard M. Re, Permissive Interpretation, 171 U. Pa. L. Rev. (forthcoming 2023) (manuscript at 12–14), https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4184846 (arguing that judges may lawfully choose between three traditional rules, none of which is paramount).

²¹ E.g., Dalfrey v. Boss Hoss Cycles, Inc., 456 F. App’x 329, 332 (5th Cir. 2011); Phelps v. Hamilton, 59 F.3d 1058, 1071 (10th Cir. 1995); Moodie v. Sch. Book Fairs, Inc., 889 F.2d 739, 743–44 (7th Cir. 1989); Presidential Hosp., LLC v. Wyndham Hotel Grp., LLC, 333 F. Supp. 3d 1179, 1229 n.15 (D.N.M. 2018); Davis v. Hollins L., a Pro. Corp., 942 F. Supp. 2d 1004, 1009 (E.D. Cal. 2013).

²² 783 F.2d 1220, 1222 (5th Cir. 1986); see also Gluck, supra note 11, at 1958.

²³ Batterton, 783 F.2d at 1222.

²⁴ Id.

²⁵ Id.

²⁶ 19 Charles Alan Wright, Arthur R. Miller & Edward H. Cooper, Federal Practice and Procedure § 4520 (3d ed. 2016); Zachary D. Clopton, Horizontal Choice of Law in Federal Court, 169 U. Pa. L. Rev. 2193 (2021); Henry J. Friendly, In Praise of Erie—and of the New Federal Common Law, 39 N.Y.U. L. Rev. 383, 408 n.122 (1964).

²⁷ See, e.g., Dennis v. Cnty. Sch. Bd., 582 F. Supp. 536, 540 (W.D. Va. 1984).

²⁸ See Vela v. Alvarez, 507 F. Supp. 887, 889 (S.D. Tex. 1981) (stating that “[c]ourts have frequently asserted that Erie applies only in diversity cases” but calling that view “overly simplistic”).

²⁹ E.g., Cruz v. Abbott, 849 F.3d 594, 599 (5th Cir. 2017) (federal question); Soza v. Hill (In re Soza), 542 F.3d 1060, 1066 (5th Cir. 2008) (bankruptcy).

³⁰ Batterton, 783 F.2d at 1222 (5th Cir. 1986).

³¹ Id.

³² There is no precise way to measure this, but one can get a sense of the proportions by searching for diversity cases and searching for cases citing Erie.  The former set dwarfs the latter.

³³ E.g., Brennan’s Inc. v. Dickie Brennan & Co. Inc., 376 F.3d 356, 366 (5th Cir. 2004) (stating that a contract was interpreted in accordance with state law but not citing Erie or other authorities for the choice of law).

³⁴ See cases cited infra note 46.

³⁵ Exxon Corp. v. Eagerton, 462 U.S. 176, 181 n.3 (1983).

³⁶ Smiley v. Kansas, 196 U.S. 447, 455 (1905); see also Bell v. Morrison, 26 U.S. (1 Pet.) 351, 359–60 (1828) (referring to “the local state tribunals, whose rules of interpretation must be presumed to be founded upon a more just and accurate view of their own jurisprudence, than those of any foreign tribunal, however respectable”).

³⁷ See, e.g., Tippecanoe Beverages, Inc. v. Heineken USA, Inc., 406 F. Supp. 2d 1033, 1036 (N.D. Ind. 2006) (citing Exxon and ruling that “the court applies Indiana’s rules of statutory construction in interpreting the [Indiana statute]”).

³⁸ 530 U.S. 914 (2000).

³⁹ Id. at 940–41.

⁴⁰ See, e.g., id. at 944 (presumption of consistent usage).  At one point the Court referred to the avoidance canon and limitations on its ability to provide a narrowing construction.  Id. at 944–45.  In so doing the Court cited federal authorities, not the Nebraska avoidance canon.  Id.  As explained below, it is possible that the avoidance doctrine really should be a matter of federal law, at least in part.  Infra Section III.A.

⁴¹ Stenberg, 530 U.S. at 942.

⁴² Id. at 940–41.

⁴³ Id. at 983 (Thomas, J., dissenting) (emphasis added).

⁴⁴ Id. at 1004–05, 1005 n.17.

⁴⁵ See Sup. Ct. R. 10.

⁴⁶ Acadia Motors, Inc. v. Ford Motor Co., 44 F.3d 1050, 1055 (1st Cir. 1995) (“Because this issue involves the interpretation of a Maine state statute, we are bound to apply the principles set forth by Maine’s Supreme Judicial Court.”); Morenz v. Wilson-Coker, 415 F.3d 230, 236–37 (2d Cir. 2005) (stating that “we are bound to interpret Connecticut law according to Connecticut’s own interpretive rules” (citing Allstate Ins. Co. v. Serio, 261 F.3d 143, 152 (2d Cir. 2001))); Park Restoration, LLC v. Erie Ins. Exch. (In re Trs. of Conneaut Lake Park, Inc.), 855 F.3d 519, 523 (3d Cir. 2017) (“When interpreting Pennsylvania law, we apply its rules of statutory interpretation.”); Volvo Trademark Holding Aktiebolaget v. Clark Mach. Co., 510 F.3d 474, 482–83 (4th Cir. 2007) (“[A] federal court sitting in diversity is obliged to apply state law principles to resolve [a question of statutory construction], utilizing such principles as enunciated and applied by the state’s highest court.” (citing Volvo Trademark Holding Aktiebolaget v. AIS Constr. Equip. Corp., 416 F. Supp. 2d 404, 410 (W.D.N.C. 2006))); Wright v. Ford Motor Co., 508 F.3d 263, 269 (5th Cir. 2007) (stating that “[w]hen we interpret a Texas statute, we follow the same rules of construction that a Texas court would apply” (citing Int’l Truck & Engine Corp. v. Bray, 372 F.3d 717, 722 (5th Cir. 2004))); Louisville/Jefferson Cnty. Metro Gov’t v. Hotels.com, L.P., 590 F.3d 381, 385 (6th Cir. 2009) (stating that “we are obliged to decide the case as we believe the Kentucky Supreme Court would do . . . [and] will therefore interpret the ordinances using the framework developed by the Kentucky courts” (citing Stalbosky v. Belew, 205 F.3d 890, 893 (6th Cir. 2000))); Doe v. Archdiocese of Milwaukee, 772 F.3d 437, 440–41 (7th Cir. 2014) (“When interpreting a state statute, we apply the same principles of statutory construction that a state court would apply.” (citing Karlin v. Foust, 188 F.3d 446, 457 (7th Cir. 1999))); Gershman v. Am. Cas. Co. of Reading, Pa., 251 F.3d 1159, 1162–63 (8th Cir. 2001) (“[W]e are bound by Missouri’s rules of statutory construction.”); Bass v. Cnty. of Butte, 458 F.3d 978, 981 (9th Cir. 2006) (“[W]e must follow the state’s rules of statutory interpretation.” (citing Planned Parenthood of Idaho, Inc. v. Wasden, 376 F.3d 908, 930 (9th Cir. 2004)); Ward v. Utah, 398 F.3d 1239, 1248 (10th Cir. 2005) (“We interpret state laws according to state rules of statutory construction . . . .” (citing Citizens for Responsible Gov’t State Pol. Action Comm. v. Davidson, 236 F.3d 1174, 1190 (10th Cir. 2000))); Robbins v. Garrison Prop. & Cas. Ins. Co., 809 F.3d 583, 586 (11th Cir. 2015) (“We construe a Florida statute according to Florida’s rules of statutory interpretation, not federal rules, when those rules differ.” (citing Allen v. USAA Cas. Ins. Co., 790 F.3d 1274, 1279 (11th Cir. 2015))); Gatewood v. Fiat, S.p.A., 617 F.2d 820, 824–25 (D.C. Cir. 1980) (relying on official commentary to a uniform act in part because the D.C. Court of Appeals had relied on the commentary to a different section of the act); see also Pritzker v. Yari, 42 F.3d 53, 65–74 (1st Cir. 1994) (applying Puerto Rico interpretive principles).  I did not find a case from the Federal Circuit, probably because that court’s specialized jurisdiction means it rarely addresses state law.

⁴⁷ Phillips v. Chandler, 215 B.R. 684, 688 (E.D. Va. 1997) (emphasis added).

⁴⁸ E.g., Brandau v. J.C. Penney Co., 646 F.2d 128, 131 (4th Cir. 1981) (“Virginia law, of course, controls the interpretation and application of this critical statutory provision.”); Comm’rs of the State Ins. Fund v. United States, 72 F. Supp. 549, 554 (S.D.N.Y. 1947) (“In interpreting a New York statute, this court is bound by the canons of statutory construction applied by the New York courts.”).

⁴⁹ E.g., Knights Templars’ & Masons’ Life Indem. Co. v. Jarman, 104 F. 638, 646 (8th Cir. 1900) (Sanborn, J., dissenting) (stating that “[i]t is a settled rule of construction that in interpreting the statutes of a state the federal courts follow the rules of construction enacted by its legislature or announced by its courts”), aff’d, 187 U.S. 197 (1902).  On the issue Judge Sanborn was addressing, the majority appeared to apply an absurd-results argument without stating whether that was a principle of federal, state, or universal law.  Id. at 642–43 (majority opinion) (referring to the likely intent of “any legislative body in its right mind”).  On appeal, the U.S. Supreme Court also appeared to invoke absurdity (without identifying the rule’s provenance), and it then noted that its construction of the state statute was “fortified” by a state interpretive directive stating that language should generally be interpreted according to its ordinary meaning.  Knights, 187 U.S. at 201–02.  Cf. Ryan Scoville, The New General Common Law of Severability, 91 Tex. L. Rev. 543, 557–69 (2013) (showing that pre-Erie cases sometimes treated severability as a question of state law, a pattern that become cemented between Erie and 2006).

⁵⁰ See Caleb Nelson, A Critical Guide to Erie Railroad Co. v. Tompkins, 54 Wm. & Mary L. Rev. 921, 942 (2013); see also Henry Campbell Black, Handbook on the Construction and Interpretation of the Laws 250–51 (2d ed. 1911) (noting states that differed from the usual rule about the interpretive value of statutory titles).

⁵¹ E.g., Louisville/Jefferson Cnty., 590 F.3d at 385; Moodie v. Sch. Book Fairs, Inc., 889 F.2d 739, 743 (7th Cir. 1989); see also, e.g.,Bass, 458 F.3d at 981.

⁵² E.g., Saiyed v. Council on Am.-Islamic Rels. Action Network, Inc., 346 F. Supp. 3d 88, 96–98 (D.D.C. 2018); N.D. Farm Bureau, Inc. v. Stenehjem, 333 F. Supp. 3d 900, 917 (D.N.D. 2018); Miller UK Ltd. v. Caterpillar Inc., 859 F. Supp. 2d 941, 945–47 (N.D. Ill. 2012).

⁵³ E.g., Phelps v. Hamilton, 59 F.3d 1058, 1071, 1073 (10th Cir. 1995).

⁵⁴ See William Baude, Adam S. Chilton & Anup Malani, Making Doctrinal Work More Rigorous: Lessons from Systematic Reviews, 84 U. Chi. L. Rev. 37 (2017).  Results lists are on file with the author.

⁵⁵ The search string was: adv: (jurisdiction! /s 1332) or (divers! /3 (jurisdiction or citizen!)) or “diversity case” or “diversity suit” and statut! and DA(aft1999 and bef2020) and Key Number Topics 361: III, IV, V, and VI.  For the courts of appeals, I considered only majority opinions and omitted decisions that were vacated for rehearing.

⁵⁶ Specifically, the Erie-related Key Number Topics were 3001 through 3120, and the search terms were “(statut! or legislat!) /s (interpret! or constru! or meaning).”  Note that the search terms for search 2 do not limit themselves to diversity cases.  Note as well that the two different search strategies do not identify entirely distinct sets of cases; a very small number of cases appear in the results for both strategies.

⁵⁷ Key Numbers do not appear in cases that are not collected in West’s reporters.  Given the very clear trend in the cases I examined, it would be surprising to find a contrary trend, on this topic, in the unpublished cases.

⁵⁸ 715 F.3d 716 (9th Cir. 2013), aff’d sub nom. Oneok, Inc. v. Learjet, Inc., 575 U.S. 373 (2015).  The Supreme Court’s decision did not address state-law issues.

⁵⁹ Id. at 729, 746.

⁶⁰ Id. at 746 (citing Lieberman v. Hawkins (In re Lieberman), 245 F.3d 1090, 1092 (9th Cir. 2001)).

⁶¹ Id. at 731, 735.

⁶² See, e.g., United States v. Dedman, 527 F.3d 577, 584–85 (6th Cir. 2008); Campbell v. Wash. Cnty. Tech. Coll., 219 F.3d 3, 6 (1st Cir. 2000).

⁶³ There is a bit of federal contract law.  E.g., Clearfield Tr. Co. v. United States, 318 U.S. 363, 366–67 (1943).  Specialized courts like the Federal Circuit encounter it more regularly than other federal courts.

⁶⁴ E.g., World-Wide Rights Ltd. P’ship v. Combe Inc., 955 F.2d 242, 245–46 (4th Cir. 1992) (citing a federal contract case, a federal statutory case, and a New Jersey contract case in a diversity case that the parties agreed, based on their briefing, was governed by New York law).

⁶⁵ These same features of familiarity and assumed similarity may provide a partial explanation for Varsava’s finding that federal courts tend not to apply state methods for discerning and applying judicial precedent.  Varsava, supra note 12, at 1227.

⁶⁶ Cf. Dodge, supra note 6, at 1401–03, 1403 n.84 (listing several federal cases that attempt to apply a state presumption against extraterritoriality but misunderstand state law).

⁶⁷ Gluck, supra note 11, at 1936–39.

⁶⁸ Travelers Indem. Co. v. Dammann & Co., 594 F.3d 238, 253 (3d Cir. 2010) (quoting Werwinski v. Ford Motor Co., 286 F.3d 661, 680 (3d Cir. 2002)).

⁶⁹ See, e.g., Werwinski, 286 F.3d at 680–81 (applying the principle to both statutory and common-law claims); Insolia v. Philip Morris Inc., 216 F.3d 596, 607 (7th Cir. 2000) (Wisconsin common law); see also Lea Brilmayer & Charles Seidell, Jurisdictional Realism: Where Modern Theories of Choice of Law Went Wrong, and What Can Be Done to Fix Them, 86 U. Chi. L. Rev. 2031, 2066–67 (2019) (describing the reduced judicial freedom associated with applying another jurisdiction’s law compared with local law, and describing the difference as “particularly apparent” with common law); Bradford R. Clark, Ascertaining the Laws of the Several States: Positivism and Judicial Federalism After Erie, 145 U. Pa. L. Rev. 1459, 1535–39 (1997) (discussing the “static” approach as a general matter, with emphasis on common law).  A topic that falls near the line between statutory interpretation and common law is the doctrine of implied private rights of action to enforce state statutes.  Federal courts generally recognize that this is a topic governed by state law, which can diverge from federal law, but they sometimes add that they should add a dash of extra hesitation in light of their position as a federal court.  E.g., Trs. of Bos. Univ. v. ASM Commc’ns, Inc., 33 F. Supp. 2d 66, 75 (D. Mass. 1998).

⁷⁰ See infra sub-subsection III.A.2.b.

⁷¹ E.g., Travelers Indem., 594 F.3d at 248, 250 (stating that “we are convinced” and “we predict” that the state court would not recognize the claim); Insolia, 216 F.3d at 607–08 (stating that there is “little indication that [state] courts would recognize [the claim]” and calling the claim “creative but unlikely”).

⁷² See Eskridge et al., supra note 3, at 677.

⁷³ E.g., Ark. Times LP v. Waldrip, 37 F.4th 1386, 1393 (8th Cir. 2022) (en banc) (citing Booker v. State, 984 S.W.2d 16, 21 (Ark. 1998)); Ctr. for Bio-Ethical Reform, Inc. v. L.A. Cnty. Sheriff Dep’t, 533 F.3d 780, 792 (9th Cir. 2008); Phelps v. Hamilton, 59 F.3d 1058, 1071–73 (10th Cir. 1995); K-S Pharmacies, Inc. v. Am. Home Prods. Corp., 962 F.2d 728, 730–31 (7th Cir. 1992); see also, e.g., In re DNA Ex Post Facto Issues, 561 F.3d 294, 300 (4th Cir. 2009).

⁷⁴ See, e.g., Boos v. Barry, 485 U.S. 312, 330–31 (1988); Allstate Ins. Co. v. Serio, 261 F.3d 143, 152 (2d Cir. 2001); Hill v. City of Houston, 789 F.2d 1103, 1112 (5th Cir. 1986) (en banc), aff’d, 482 U.S. 451 (1987); see also Gluck, supra note 11, at 1948–53 (discussing such cases); id. at 1956 (stating that “federal courts do not seem comfortable applying their own federal avoidance principles to state statutes either”).

⁷⁵ Eubanks v. Wilkinson, 937 F.2d 1118, 1125 (6th Cir. 1991).

⁷⁶ Id. at 1126–28.

⁷⁷ See, e.g., Frisby v. Schultz, 487 U.S. 474, 483 (1988); Connecticut ex rel. Blumenthal v. Crotty, 346 F.3d 84, 93 (2d Cir. 2003).

⁷⁸ E.g., Blake v. Schwartz, 42 P.3d 6, 8–9 (Ariz. Ct. App. 2002) (quoting United States v. Clark, 445 U.S. 23, 27 (1980)) (citing Edward J. DeBartolo Corp. v. Fla. Gulf Coast Bldg. & Constr. Trades Council, 485 U.S. 568, 575 (1988)) (relying on two leading federal avoidance cases); Commonwealth v. Bavusa, 832 A.2d 1042, 1050–51 (Pa. 2003) (quoting Harris v. United States, 536 U.S. 545, 555 (2002)) (using federal definition of avoidance canon).

⁷⁹ It would be understandable for a federal court facing a challenge to a state statute to cite a federal case from the same substantive field of law, overlooking the choice-of-law issues.  E.g., Hopkins v. Jegley, 267 F. Supp. 3d 1024, 1080 (E.D. Ark. 2017) (quoting Gonzales v. Carhart, 550 U.S. 124, 153 (2007)) (citing, in a challenge to a state abortion restriction, a U.S. Supreme Court abortion case that invoked the avoidance canon for a federal statute).

⁸⁰ E.g., 1256 Hertel Ave. Assocs., LLC v. Calloway, 761 F.3d 252, 260 n.5 (2d Cir. 2014) (expressing doubt over whether to apply state or federal avoidance doctrine).  In the days before Erie, mixing citations could reflect a view that interpretation was a matter of general law subject to occasional local departures.  E.g., Grenada Cnty. Supervisors v. Brogden, 112 U.S. 261, 268–69 (1884) (first citing Thomas M. Cooley, A Treaty on the Constitutional Limitations Which Rest upon the Legislative Power of the States of the American Union 255 (7th ed. 1903); then citing Newland v. Marsh, 19 Ill. 376, 384 (1857); then citing People ex rel. Burrows v. Supervisors of Orange Cnty., 17 N.Y. 235, 241 (1858); and then citing Colwell v. May’s Landing Water Power Co., 19 N.J. Eq. 245, 249 (N.J. Ch. 1868)) (citing the Cooley treatise and avoidance cases from various states and then confirming that Mississippi, whose constitution and statute were involved, followed the general rule).

⁸¹ E.g., Eubanks, 937 F.2d at 1125–29 (holding that state statute could not be redrafted to uphold it but that invalid portion could be severed).

⁸² E.g., Arizonans for Off. Eng. v. Arizona, 520 U.S. 43, 76–80 (1997); Lake Carriers’ Assn. v. MacMullan, 406 U.S. 498, 510–12 (1972).

⁸³ Cf. Scoville, supra note 49, at 564–74 (showing that the post-Erie view that severability doctrine was state law was displaced starting in 2006 by Supreme Court cases employing federal principles).

⁸⁴ Planned Parenthood of Idaho, Inc. v. Wasden, 376 F.3d 908, 930–31 (9th Cir. 2004) (“Ordinarily, in construing a state statute, we follow that state’s rules of statutory interpretation . . . . The Supreme Court, however, has instructed us to assume that state courts will endeavor to construe abortion statutes constitutionally . . . . Hence, while under Idaho law the plain-meaning inquiry would almost surely be the end of the matter, we also consider whether, supposing the statute to be in some respect ambiguous, a limiting construction is available.” (citations omitted)).

⁸⁵ John Hart Ely, The Irrepressible Myth of Erie, 87 Harv. L. Rev. 693, 697–98 (1974); Allan Erbsen, Erie’s Four Functions: Reframing Choice of Law in Federal Courts, 89 Notre Dame L. Rev. 579, 583 (2013).

⁸⁶ Compare, e.g., Louise Weinberg, Federal Common Law, 83 Nw. U. L. Rev. 805, 851 (1989) (defending broad view of power), with, e.g., Martin H. Redish, Federal Common Law, Political Legitimacy, and the Interpretative Process: An “Institutionalist” Perspective, 83 Nw. U. L. Rev. 761, 803 (1989) (advancing narrower view of power).

⁸⁷ Compare Shady Grove Orthopedic Assocs., P.A. v. Allstate Ins. Co., 559 U.S. 393, 398–99 (2010) (giving Rule 23 a broad preemptive scope), with id. at 436 (Ginsburg, J., dissenting) (advocating a narrower reading of Rule 23 in order to respect state policies).

⁸⁸ See Michael Steven Green, The Twin Aims of Erie, 88 Notre Dame L. Rev. 1865, 1887–1904 (2013) (describing several potential sources).

⁸⁹ Infra note 123 and accompanying text.

⁹⁰ As one example, Erbsen’s recent proposed reorientation away from Hanna’s distinction between federal positive law and federal common law and towards questions of federal power to create law would probably lead him to move up Section II.G ahead of what currently precedes it.  See Allan Erbsen, A Unified Approach to Erie Analysis for Federal Statutes, Rules, and Common Law, 10 U.C. Irvine L. Rev. 1101, 1103–04, 1106 (2020).

⁹¹ See Hanna v. Plumer, 380 U.S. 460, 465–71 (1965).

⁹² 28 U.S.C. § 1652 (2018).

⁹³ See Julian R. Murphy, Lenity and the Constitution: Could Congress Abrogate the Rule of Lenity?, 56 Harv. J. on Legis. 423, 425 (2019).

⁹⁴ E.g., Off. of Legal Pol’y, U.S. Dep’t of Just., Using and Misusing Legislative History ii–iii, 47–51 (1989); John F. Manning, Textualism as a Nondelegation Doctrine, 97 Colum. L. Rev. 673, 707–25 (1997).

⁹⁵ See Gluck, supra note 12, at 782.

⁹⁶ See In re Sinclair, 870 F.2d 1340, 1343–44 (7th Cir. 1989).

⁹⁷ Manning, supra note 94, at 706–07.

⁹⁸ See Jeffrey A. Pojanowski, Statutes in Common Law Courts, 91 Tex. L. Rev. 479, 540 (2013).

⁹⁹ E.g., James J. Brudney, Canon Shortfalls and the Virtues of Political Branch Interpretive Assets, 98 Calif. L. Rev. 1199, 1217–18 (2010) (Rules of Proceedings Clause and Journal Clause); Victoria F. Nourse, A Decision Theory of Statutory Interpretation: Legislative History by the Rules, 122 Yale L.J. 70, 135 n.284, 149 n.346 (2012) (Rules of Proceedings Clause and cameral rules).

¹⁰⁰ Wayman v. Southard, 23 U.S. (10 Wheat.) 1, 49 (1825).

¹⁰¹ See id. (contrasting what states may accomplish “in virtue of an original inherent power in the State legislatures, independent of any act of Congress” with what Congress might provide).

¹⁰² See, e.g., Fed. R. Civ. P. 64(a), 69(a); Fed. R. Evid. 501; Fullerton v. President, Dirs. & Co of the Bank of U.S., 26 U.S. (1 Pet.) 604, 612–14 (1828)); Wayman, 23 U.S. at 41–44; see also infra Section II.E (discussing the Process Acts, the Conformity Act, and the Rules of Decision Act).

¹⁰³ See infra Section II.D; see also Byrd v. Blue Ridge Rural Elec. Coop., Inc., 356 U.S. 525, 536–37 (1958) (explaining that certain state practices that are arguably procedural can be so “bound up with” state-created rights and obligations that the federal courts must honor them).

¹⁰⁴ See Alyeska Pipeline Serv. Co. v. Wilderness Soc’y, 421 U.S. 240, 259 n.31 (1975) (quoting Sioux Cnty. v. Nat’l Sur. Co., 276 U.S. 238, 243 (1928)); Cities Serv. Oil Co. v. Dunlap, 308 U.S. 208, 212 (1939); Boyd Rosene & Assocs., Inc. v. Kan. Mun. Gas Agency, 174 F.3d 1115, 1125–28 (10th Cir. 1999).

¹⁰⁵ See Michael S. Green, Vertical Power, 48 U.C. Davis L. Rev. 73, 81–89 (2014).

¹⁰⁶ See infra Section II.E (discussing whether interpretive methodology is a “rule of decision” within the meaning of the Rules of Decision Act).

¹⁰⁷ See Jennifer M. Bandy, Note, Interpretive Freedom: A Necessary Component of Article III Judging, 61 Duke L.J. 651, 653 (2011); see also Varsava, supra note 12, at 1251, 1255-56 (arguing that judges who adopt a Dworkinian approach to jurisprudence may determine that the law-supplying jurisdiction has incorrectly discerned its own law of precedent-interpretation).

¹⁰⁸ See Bandy, supra note 107, at 682 (conceding that “good practice” calls for federal courts to consider state methods in making what she regards as their own individual methodological choice).

¹⁰⁹ 1 U.S.C. §§ 1–8 (2018); cf. Fed. R. Civ. P. 44.1 (regulating the sources to be used in determining the law of foreign countries).

¹¹⁰ E.g., Civil Rights Act of 1991, Pub. L. No. 102-166, § 105(b), 105 Stat. 1071, 1075 (codified at 42 U.S.C. § 1981 note (Legislative History for 1991 Amendment)); see Jarrod Shobe, Congressional Rules of Interpretation, 63 Wm. & Mary L. Rev. 1997 (2022).

¹¹¹ 28 U.S.C. § 1738 (2018) (emphasis added).

¹¹² Section 1738’s “same full faith and credit” requirement did not extend to state statutes until the 1948 revision of Title 28.  The extension to state statutes was at best poorly thought out; perhaps it was just a mistake.  David E. Engdahl, The Classic Rule of Faith and Credit, 118 Yale L.J. 1584, 1656–57 (2009); Ralph U. Whitten, The Constitutional Limitations on State Choice of Law: Full Faith and Credit, 12 Mem. St. U. L. Rev. 1, 60–62 (1981).  Perhaps as a result, the statutory command to give state statutes “the same full faith and credit” has mostly been ignored.  E.g., Hughes v. Fetter, 341 U.S. 609, 613 n.16 (1951).  I intend to explore the potential role of the Full Faith and Credit Clause as a source of congressional power over statutory interpretation in future work.

¹¹³ Compare Rosenkranz, supra note 12, at 2088 (taking positive view of congressional power), with Linda D. Jellum, “Which Is to Be Master,” the Judiciary or the Legislature? When Statutory Directives Violate Separation of Powers, 56UCLA L. Rev. 837, 841 (2009), and Gary Lawson, Controlling Precedent: Congressional Regulation of Judicial Decision-Making, 18 Const. Comment. 191, 194 n.16, 210–14 (2001) (both taking negative view on congressional power).

¹¹⁴ Rosenkranz, supra note 12, at 2108 (citing Erie R.R. Co. v. Tompkins, 304 U.S. 64, 78 (1938)).

¹¹⁵ U.S. Const. art. I, § 8, cls. 9, 18; id. art. III, § 1, cl. 1; id. § 2, cl. 2; Wayman v. Southard, 23 U.S. (10 Wheat.) 1, 22 (1825).

¹¹⁶ U.S. Const. art. IV, § 1.

¹¹⁷ Engdahl, supra note 112, at 1657–59; cf. Stephen E. Sachs, Full Faith and Credit in the Early Congress, 95 Va. L. Rev. 1201, 1208–09, 1279 (2009) (making similar observation regarding congressional power to specify the effect of judgments).

¹¹⁸ See Hanna v. Plumer, 380 U.S. 460, 471 (1965).

¹¹⁹ Gluck, supra note 11, at 1968–80; Gluck, supra note 12, at 793–94; see, e.g., Avery v. Hughes, 661 F.3d 690, 693–94 (1st Cir. 2011) (first citing Erie, 304 U.S. at 78; and then citing Eaton v. Penn-Am. Ins. Co., 626 F.3d 113, 114 (1st Cir. 2010)) (contract interpretation).

¹²⁰ See Varsava, supra note 12, at 1220–24.

¹²¹ Scoville, supra note 49, at 557–69.

¹²² See Jerome A. Hoffman, Thinking out Loud About the Myth of Erie: Plus a Good Word for Section 1652, 70 Miss. L.J. 163, 168 (2000) (lamenting the RDA’s status as “candidate for ‘Statute Most Likely to Be Ignored’”).

¹²³ See Weinberg, supra note 86, at 816–18 (contending that the RDA “cannot have any modern meaning”).  Interpretations abound.  Ritz argued that the RDA was a directive to apply American law rather than English law, not a directive to apply the law of particular states; he even argued that the provision was meant to apply only to criminal trials, directing the application of the American common law of crimes until Congress could enact a criminal code.  See Wilfred J. Ritz, Rewriting the History of the Judiciary Act of 1789: Exposing Myths, Challenging Premises, and Using New Evidence 142–48 (Wythe Holt & L.H. LaRue eds., 1990); see also William A. Fletcher, The General Common Law and Section 34 of the Judiciary Act of 1789: The Example of Marine Insurance, 97 Harv. L. Rev. 1513, 1514–15, 1527–28 (1984) (interpreting the RDA as originally understood to apply only to state law on “local” matters, whether statutory or judicial).

¹²⁴ Judiciary Act of 1789, ch. 20, § 34, 1 Stat. 73, 92.

¹²⁵ 28 U.S.C. § 1652 (2018).

¹²⁶ Erie R.R. Co. v. Tompkins, 304 U.S. 64, 71–73, 78 (1938).

¹²⁷ Swift v. Tyson, 41 U.S. (16 Pet.) 1, 18 (1842) (stating that “this Court have uniformly supposed, that the true interpretation of the thirty-fourth section limited its application to state laws strictly local, that is to say, to the positive statutes of the state, and the construction thereof adopted by the local tribunals, and to rights and titles to things having a permanent locality”).

¹²⁸ See id.; see also Elmendorf v. Taylor, 23 U.S. (10 Wheat.) 152, 159–60 (1825) (same, in equity).  But see infra Section III.A (describing situations in which federal courts would depart from state interpretations).

¹²⁹ E.g., Albertson v. Millard, 345 U.S. 242, 244 (1953); Lovely v. Cunningham, 796 F.2d 1, 4 (1st Cir. 1986).  There are some complications here, such as what to do when state case law is unclear or obsolete.  But those complications are familiar in the world of Erie, not distinctive to the realm of interpretive methodology, so I put them to one side.

¹³⁰ 28 U.S.C. § 1652.

¹³¹ This suggestion has the virtue of resonating with a plausible account of what the “where they apply” language meant when it was enacted.  See Swift, 41 U.S. at 18–19 (observing that the New York courts did not consider their law of bills of exchange to reflect local usages but instead to be derived from general principles of the law merchant); Ernest A. Young, A General Defense of Erie Railroad Co. v. Tompkins, 10 J.L., Econ. & Pol’y 17, 35–38 (2013) (arguing that the original understanding of the phrase referred to the states’ power to turn what had been a matter of general law into local, state law).  As I hasten to add in the main text above, we should care about what states are trying to do regardless of what the enactors of the RDA meant.

¹³² E.g., Restatement (Second) of Conflict of L. § 6(2) (Am. L. Inst. 1971); Restatement (Third) of Conflict of L. 1–3 & § 5.01 (Am. L. Inst., Tentative Draft No. 3, 2022); see also Larry Kramer, Rethinking Choice of Law, 90 Colum. L. Rev. 277, 291 (1990) (explaining that the first step in a conflicts analysis is to interpret the candidate laws to determine whether each one applies to the case at hand from its own perspective, such that there is a conflict).

¹³³ See Michael S. Green, The Erie Doctrine: A Flowchart, 52 Akron L. Rev. 215, 217 (2018); Kermit Roosevelt III, Choice of Law in Federal Courts: From Erie and Klaxon to CAFA and Shady Grove, 106 Nw. U. L. Rev. 1, 11–12 (2012).

¹³⁴ See generally Jacob Scott, Codified Canons and the Common Law of Interpretation, 98 Geo. L.J. 341 (2010) (surveying enacted interpretive directives).

¹³⁵ If a state directive purports to regulate its state courts’ interpretation of federal statutes, we would have to consider federal preemption and “reverse Erie.”  The duties of state courts interpreting federal statutes are not taken up here, but for discussions of that issue, see Anthony J. Bellia Jr., State Courts and the Interpretation of Federal Statutes, 59 Vand. L. Rev. 1501 (2006) (discussing early state interpretive practices when dealing with federal statutes); Aaron-Andrew P. Bruhl & Ethan J. Leib, Elected Judges and Statutory Interpretation, 79 U. Chi. L. Rev. 1215, 1272–74 (2012) (noting special concerns stemming from state judicial elections); Gluck, supra note 11, at 1962–68 (surveying cases and highlighting uniformity concerns).

¹³⁶ There is an underlying question here about what methods to use to interpret a state’s interpretive directives.  I thank Nina Varsava for pointing this out.  I finesse the issue here by using textual analysis in this paragraph and more purposive arguments in the next paragraph; it seems to me this is a situation where different plausible default methods would reach the same answer, namely that the state legislature wants the state methods to apply to state statutes in federal court.

¹³⁷ See, e.g., Cal. Civ. Proc. Code § 4 (West 2022); Mich. Comp. Laws § 8.3 (2022); Minn. Stat. § 645.08 (2022); N.C. Gen. Stat. § 12-3 (2022); 1 Pa. Cons. Stat. §§ 1502, 1901 (2022); Tex. Gov’t Code Ann. § 311.002 (West 2021); Va. Code Ann. § 1-202 (2022); Wis. Stat. § 990.001 (2022).

¹³⁸ See Utah Code Ann. § 68-3-2(1) (LexisNexis 2022).

¹³⁹ E.g., 5 Ill. Comp. Stat. 70/1.01 (2022); N.C. Gen. Stat. § 12-3 (2022); 1 Pa. Cons. Stat. § 1922 (2022); see also Unif. Statute and Rule Constr. Act prefatory note (Nat’l Conf. of Comm’rs on Unif. State L. 1995) (stating that the Act “informs courts of a legislature’s expectations as to how its product should be construed” and “will assist drafters in preparing legislation and rules”).

¹⁴⁰ E.g., Or. Rev. Stat. § 174.010 (2022); Tex. Gov’t Code Ann. chs. 311–12 (West 2021); cf. Shady Grove Orthopedic Assocs., P.A. v. Allstate Ins. Co., 559 U.S. 393, 432–35 & n.16 (2010) (Stevens, J., concurring in part and concurring in the judgment) (concluding that a state statute did not serve to define rights and remedies under state law by observing, inter alia, that the state statute governed the conduct of the state’s courts regardless of which substantive law the courts were applying).  The inference is also especially strong when the directive is targeted to the interpretation of a particular code, as the subject-specific targeting makes it harder to imagine that the legislature was motivated by concerns of state judicial administration.  See, e.g.,Cal. Civ. Code § 1798.63 (West 2022) (“The provisions of this chapter shall be liberally construed . . . .”).

¹⁴¹ Cf. Hanna v. Plumer, 380 U.S. 460, 475 (1965) (Harlan, J., concurring) (urging us “to stay close to basic principles by inquiring if the choice of rule would substantially affect those primary decisions respecting human conduct which our constitutional system leaves to state regulation”); see also Section II.F (discussing outcome determination).

¹⁴² Id. at 474.

¹⁴³ See, e.g., Landgraf v. USI Film Prods., 511 U.S. 244, 297 (1994) (Blackmun, J., dissenting) (arguing that the presumption against retroactivity should not apply to legislation increasing the compensatory damages available for previously illegal discrimination); Dodge, supra note 6, at 1401–04 (explaining differing approaches to the presumption against extraterritoriality in different jurisdictions).

¹⁴⁴ See Bruhl, supra note 19, at 159–62 (describing prospects for the growth of methodological precedent).

¹⁴⁵ A complication would arise if a state’s legislature and its state courts disagreed on the Erie question or on the correct interpretive method.  My inclination is to say that the federal courts should treat the state courts as the voice of the state and leave it to the state legislature to bring the state courts into line.  I otherwise abstract away from this difficulty for present purposes.

¹⁴⁶ See, e.g., Simpson v. Cavalry SPV I, LLC, 440 S.W.3d 335, 337 (Ark. 2014) (answering certified question using the court’s “well established” rules of interpretation); Shasta View Irrigation Dist. v. Amoco Chems. Corp., 986 P.2d 536, 539 (Or. 1999) (answering certified question using the state’s then-prevailing PGE interpretive framework).

¹⁴⁷ In re Hernandez, 161 N.E.3d 135, 140 (Ill. 2020).  Notably, this certification arose in a bankruptcy case, not a diversity case.

¹⁴⁸ Regarding this last interest, remember the closer relationship between state judges and state legislators (as compared to their federal counterparts) and the fact that some states have clearly articulated and distinctive methods.  Those things make it more realistic for state legislators to understand and legislate against the background of state interpretive methods.  See Bruhl & Leib, supra note 135, at 1253; see also Grace E. Hart, Comment, Methodological Stare Decisis and Intersystemic Statutory Interpretation in the Choice-of-Law Context, 124 Yale L.J. 1825, 1834–35 (2015) (noting that some state legislative drafting manuals describe the state’s judicially developed interpretive principles).

¹⁴⁹ For example, a jurisdiction may believe that all courts have a duty to make the best constructive Dworkinian interpretation of whatever jurisdiction’s law comes before it, regardless of what the enacting jurisdiction believes its law is or is determined.  See Varsava, supra note 12, at 1253–56 (making this point in the context of intersystemic understandings of stare decisis).  Or consider a court that believes interpretation is a matter of universal law that all courts may independently engage in, regardless of what the enacting state’s courts think about the content of their statutory law.  A potential analogue is the Georgia Supreme Court, which appears to believe, in a sort of holdover of the world of Swift v. Tyson, that one jurisdiction may reach its own independent judgment about the content of another jurisdiction’s common law.  Michael Steven Green, Erie’s Suppressed Premise, 95 Minn. L. Rev. 1111, 1126–27 (2011).  However, it appears that Georgia does not independently interpret other states’ statutes, though Georgia may regard this deference to sister states as a matter of comity rather than duty.  See Coon v. Med. Ctr., Inc., 797 S.E.2d 828, 833–34 (Ga. 2017); Lee v. Lott, 177 S.E. 92, 94–95 (Ga. Ct. App. 1934).

¹⁵⁰ See Aaron-Andrew P. Bruhl, Hierarchy and Heterogeneity: How to Read a Statute in a Lower Court, 97 Cornell L. Rev. 433, 470–84 (2012) (describing interpretation-relevant institutional differences across courts within a judicial system); Varsava, supra note 12, at 1258–59 (describing epistemic difficulties in following another jurisdiction’s law about precedent).

¹⁵¹ Even if interpretive methodology is outside the scope of the RDA, which the rest of this Section denies, the conclusion that the interpretative methodology applied in federal court needs to be federal law does not mean that it has to be uniform federal law, as opposed to federal law that borrows the content of whatever state’s statute is being interpreted at the time.  See infra Section II.G.  There would be good reasons for federal courts to borrow state methods as the federal method when interpreting state statutes.  See infra Section II.F (discussing the “twin aims” of Erie).

¹⁵² See, e.g., Sibbach v. Wilson & Co., 312 U.S. 1, 14 (1941) (defining procedure as “the judicial process for enforcing rights and duties recognized by substantive law and for justly administering remedy and redress for disregard or infraction of them”).

¹⁵³ See 1 Dicey, Morris and Collins on the Conflict of Laws 318, 328 (Lord Collins of Mapesbury, Adrian Briggs, Andrew Dickinson, Jonathan Harris, J.D. McClean, Peter McEleavy, Campbell McLachlan & C.G.J Morse eds., 15th ed. 2012) [hereinafter Dicey].

¹⁵⁴ See Prudential Ins. Co. of Am. v. O’Grady, 396 P.2d 246, 248–49 (Ariz. 1964) (en banc) (acknowledging, but departing from, this approach to sister-state law).

¹⁵⁵ Restatement (Second) of Conflict of L. § 136(2) (Am. L. Inst. 1971).

¹⁵⁶ See Owings v. Hull, 34 U.S. (9 Pet.) 607, 625 (1835).

¹⁵⁷ See, e.g., Prudential Ins. Co., 396 P.2d at 249 (citing Choate v. Ransom, 323 P.2d 700 (Nev. 1958)); N.Y. C.P.L.R. 4511(a), (d) (McKinney 2022).

¹⁵⁸ See Acker v. Ray Angelini, Inc., 259 F. Supp. 3d 305, 309 n.3 (E.D. Pa. 2016) (“[N]otwithstanding what some would say are profound cultural and linguistic differences between ‘Joisey’ and Pennsylvania, all that is required to analyze the issue is access to Westlaw or LexisNexis.”).

¹⁵⁹ J.G. Collier, Conflict of Laws 35 (3d ed. 2001); Dicey, supra note 153, at 328–29; see also A/S Tallinna Laevauhisus v. Estonian State S.S. Line (1947) 80 Lloyd’s List LR 99 (AC) at 108 (Eng. & Wales) (“[E]ven when a proved or agreed translation takes the place of the foreign document, it is still primarily the function of the expert witness to interpret its legal effect, in order to convey to the English Court the meaning and effect which a Court of the foreign country would attribute to it . . . . If [the expert] says that the foreign statute bears a meaning which is patently inconsistent with the words of the English translation, the Court is entitled to reject his construction unless [the expert] goes further and proves some extraneous rule of law, written or unwritten, of the foreign country which compels that apparently forced interpretation.”); Sofie Geeroms, Foreign Law in Civil Litigation: A Comparative and Functional Analysis 183 (2004) (explaining that an English court may apply “English rules of construction” if “no evidence at all is offered that different rules govern the foreign court’s interpretation” of a foreign statute).

¹⁶⁰ E.g., Gasperini v. Ctr. for Humans., Inc., 518 U.S. 415, 426–27 (1996).

¹⁶¹ 28 U.S.C. § 1652 (2018) (providing for the use of state law “except where the Constitution or treaties of the United States or Acts of Congress otherwise require or provide”).  The Federal Rules are not themselves acts of Congress, but they are authorized by the Rules Enabling Act.  See id. § 2072.

¹⁶² Wayman v. Southard, 23 U.S. (10 Wheat.) 1, 2 (1825).

¹⁶³ Id. at 24–25 (emphasis added); see also id. at 26 (stating that “in framing [the RDA], the legislature could not have extended its views beyond the judgment of the Court”).

¹⁶⁴ M’Niel v. Holbrook, 37 U.S. (12 Pet.) 84, 89 (1838) (emphasis added).

¹⁶⁵ E.g.,Conn. Mut. Life Ins. Co. v. Union Tr. Co., 112 U.S. 250, 255 (1884) (stating, in a case concerning doctor-patient privilege, that the RDA “has been uniformly construed as requiring the courts of the Union . . . to observe, as rules of decision, the rules of evidence prescribed by the laws of the States in which such courts are held”); Ryan v. Bindley, 68 U.S. (1 Wall.) 66, 68 (1863) (holding that RDA applies to competency to testify).

¹⁶⁶ Camden & Suburban Ry. Co. v. Stetson, 177 U.S. 172, 174–75, 177 (1900).  The matter of medical exams is now addressed by Federal Rule of Civil Procedure 35, which the Supreme Court upheld as valid under the Rules Enabling Act.  Sibbach v. Wilson & Co., 312 U.S. 1, 9–14 (1941).

¹⁶⁷ See, e.g., Judiciary Act of 1789, ch. 20, § 15, 1 Stat. 73, 82 (compulsory production of documents); id. § 17 (conferring power to “make and establish all necessary rules for the orderly conducting business”); id. § 18 (stays of execution of judgments); id. § 30 (oral testimony, depositions for unavailable parties); id. § 31 (substitution of parties upon death).

¹⁶⁸ See Ex parte Fisk, 113 U.S. 713, 719–23 (1885).

¹⁶⁹ E.g., Act of Sept. 29, 1789, ch. 21, § 2, 1 Stat. 93; Act of May 8, 1792, ch. 35, 1 Stat. 275; Act of May 19, 1828, ch. 68, 4 Stat. 278.  See generally Wayman v. Southard, 23 U.S. (10 Wheat.) 1 (1825) (discussing RDA and various iterations of Process Acts).

¹⁷⁰ See, e.g., 4 Stat. at 278–81 (requiring federal courts to follow forum-state procedure in cases at common law but to follow “the principles, rules, and usages, which belong to” courts of equity and admiralty for those kinds of cases).

¹⁷¹ See Alfred Hill, State Procedural Law in Federal Nondiversity Litigation, 69 Harv. L. Rev. 66, 86 (1955) (“In general the Rules of Decision Act was deemed to make state procedural law applicable in ‘trials at common law’ in the federal courts except where Congress had otherwise provided. . . . [T]he Process Acts and the Conformity Act dealt more specifically than did the Rules of Decision Act with state procedural law, and were therefore deemed to govern in their respective areas of coverage.”); Anthony J. Bellia Jr. & Bradford R. Clark, General Law in Federal Court, 54 Wm. & Mary L. Rev. 655, 672 (2013) (“Although it was not always clear where state forms of proceeding ended and state law rules of decision began, federal courts typically did not have to draw a sharp line between them because Congress directed them to apply both.”).

¹⁷² The Practice Conformity Act of 1872 improved on the earlier Process Acts by requiring federal courts to match the current home-state practice, rather than conforming to the practice as it stood on the date of enactment, as had been the case under the Process Acts.  Practice Conformity Act, ch. 255, § 5, 17 Stat. 196.  There is little reason to desire federal courts to follow outmoded state procedure, as it accomplishes neither nationwide uniformity in federal courts nor vertical uniformity within each state.  That unhappy circumstance may have motivated some pre-1872 decisions that used the Rules of Decision Act rather than recurring to the Process Acts for procedural matters.  See Green, supra note 105, at 122–24; Caleb Nelson, The Persistence of General Law, 106 Colum. L. Rev. 503, 550 n.229 (2006).

¹⁷³ Swift v. Tyson, 41 U.S. (16 Pet.) 1, 18–19 (1842).

¹⁷⁴ Hill, supra note 171, at 83–86.

¹⁷⁵ Dygert v. Vt. Loan & Tr. Co., 94 F. 913, 914–15 (9th Cir. 1899); see also Sampson v. Channell, 110 F.2d 754, 761 (1st Cir. 1940) (opinion of Magruder, J.) (first citing Boseman v. Conn. Gen. Life Ins. Co., 301 U.S. 196 (1937); then citing Citizens Nat’l Bank v. Waugh, 78 F.2d 325 (4th Cir. 1935); and then citing Dygert, 94 F. 913) (describing pre-Erie law); Mark D. Rosen, Choice-of-Law as Non-Constitutional Federal Law, 99 Minn. L. Rev. 1017, 1026–40 (2015) (same).

¹⁷⁶ See Nashua Sav. Bank v. Anglo-Am. Land, Mortg. & Agency Co., 189 U.S. 221, 228 (1903); Hill, supra note 171, at 83–86.

¹⁷⁷ See supra notes 36, 49; see also Myer v. Car Co., 102 U.S. 1, 11–12 (1880) (citing Iowa Const. art. III, § 29) (stating that the general rule that titles can clear up ambiguities in the purview applied with greater force to the state statute at issue because the state constitution required an accurate title); Viterbo v. Friedlander, 120 U.S. 707, 724–26 (1887) (citing Louisiana methodology, though with the view that it accords with generally prevailing principles).

¹⁷⁸ Supra text accompanying notes 141–43.

¹⁷⁹ 326 U.S. 99 (1945).

¹⁸⁰ Id. at 109–10.

¹⁸¹ 380 U.S. 460, 468 (1965).

¹⁸² See id. at 468–69.

¹⁸³ See Green, supra note 133, at 245–46; Roosevelt, supra note 133, at 15.

¹⁸⁴ See King v. Ord. of United Com. Travelers of Am., 333 U.S. 153, 160–61 (1948); Webber v. Butner, 923 F.3d 479, 482 (7th Cir. 2019); Packard v. Provident Nat’l Bank, 994 F.2d 1039, 1046 (3d Cir. 1993) (citing McKenna v. Ortho Pharm. Corp., 622 F.2d 657 (3d Cir. 1980)).

¹⁸⁵ Compare Fuchsgruber v. Custom Accessories, Inc., 628 N.W.2d 833, 841 (Wis. 2001) (employing derogation canon), with Utah Code Ann. § 68-3-2(1) (LexisNexis 2022) (abrogating the derogation canon).

¹⁸⁶ See Dodge, supra note 6, at 1401–04 (citing variation across states).

¹⁸⁷ See 2A Norman J. Singer & J.D. Shambie Singer, Statutes and Statutory Construction §§ 47:3, 47:14 (7th ed. 2007) (discussing the use of titles and section headings).

¹⁸⁸ See Bruhl, supra note 19, at 154–56 (citing examples).  Compare Cohen v. Rubin, 460 A.2d 1046, 1056 (Md. Ct. Spec. App. 1983) (disallowing punitive damages for wrongful-death claim because statutes in derogation of the common law are narrowly construed), with Behrens v. Raleigh Hills Hosp., Inc., 675 P.2d 1179, 1184 (Utah 1983) (reaching contrary decision in part because Utah has abrogated the derogation canon).  See also Presidential Hosp., LLC v. Wyndham Hotel Grp., LLC, 333 F. Supp. 3d 1179, 1229 n.15 (D.N.M. 2018) (explaining that the court would have interpreted a statute according to its plain meaning but for its Erie-derived duty to follow the state’s interpretive approach).

¹⁸⁹ Restatement (Second) of Conflict of L. §§ 6(2), 145, 188 (Am. L. Inst. 1971).

¹⁹⁰ E.g., Kearney v. Salomon Smith Barney, Inc., 137 P.3d 914, 917–18 (Cal. 2006); Babcock v. Jackson, 191 N.E.2d 279, 284 (N.Y. 1963).

¹⁹¹ Klaxon Co. v. Stentor Elec. Mfg. Co., 313 U.S. 487, 496 (1941).

¹⁹² Cf. Larry Kramer, Choice of Law in Complex Litigation, 71 N.Y.U. L. Rev. 547, 569–72 (1996) (drawing similar analogy between interpretation and choice of law in order to show that choice of law is substantive).

¹⁹³ See Kevin M. Clermont, Degrees of Deference: Applying vs. Adopting Another Sovereign’s Law, 103 Cornell L. Rev. 243, 247 (2018); Martha A. Field, Sources of Law: The Scope of Federal Common Law, 99 Harv. L. Rev. 881, 963–77 (1986).

¹⁹⁴ Clermont, supra note 193, at 245.

¹⁹⁵ Id. at 245, 272 n.176.

¹⁹⁶ Semtek Int’l Inc. v. Lockheed Martin Corp., 531 U.S. 497, 508–09 (2001).

¹⁹⁷ See supra Section II.D.

¹⁹⁸ See Green, supra note 133, at 246 (distinguishing between sovereignty and borrowing considerations in Erie analysis).

¹⁹⁹ Supra Sections II.D–F.

²⁰⁰ See Nelson v. Great Lakes Educ. Loan Servs., Inc., 928 F.3d 639, 651 (7th Cir. 2019).

²⁰¹ See Thomas W. Merrill, The Common Law Powers of Federal Courts, 52 U. Chi. L. Rev. 1, 46–47 (1985); Redish, supra note 86, at 766–67.

²⁰² See supra subsection II.A.3 (discussing this individual-philosophy view).

²⁰³ See supra text accompanying note 105.

²⁰⁴ Weinberg, supra note 86, at 809–13.

²⁰⁵ Id. at 811–15.

²⁰⁶ See, e.g., United States v. Little Lake Misere Land Co., 412 U.S. 580, 595–97 (1973) (refusing to borrow a particular state law that damaged federal interests); see also Field, supra note 193, at 970–73 (describing federal-state hybrid law that inserts only so much federal law as federal interests require).

²⁰⁷ Restatement (Second) of Conflict of L. § 145 cmt. d (Am. L. Inst. 1971).

²⁰⁸ See id.; see also, e.g., Ruiz v. Blentech Corp., 89 F.3d 320, 325–26 (7th Cir. 1996) (explaining that the district court should have applied the California law of corporate-successor liability but Illinois tort law).  In the conflicts jurisprudence, this mixing is sometimes called dépeçage.  Id. at 324–26.

²⁰⁹ See, e.g., Lamps Plus, Inc. v. Varela, 139 S. Ct. 1407, 1417–19 (2019).

²¹⁰ Supra Sections II.A–B.

²¹¹ Byrd v. Blue Ridge Rural Elec. Coop., Inc., 356 U.S. 525, 537–38 (1958).

²¹² See id. at 536–40.

²¹³ See 14D Charles Alan Wright, Arthur R. Miller, Edward H. Cooper & Richard D. Freer, Federal Practice and Procedure § 3828.5 & n.10 (4th ed. 2013).

²¹⁴ See, e.g., People v. Harris, 885 N.W.2d 832, 838–39, 838 n.29 (Mich. 2016); Richards v. Cox, 450 P.3d 1074, 1079–81 (Utah 2019).  See generally Neal Goldfarb, A Lawyer’s Introduction to Meaning in the Framework of Corpus Linguistics, 2017 BYU L. Rev. 1359 (describing this technique).

²¹⁵ See Lee & Mouritsen, supra note 10, at 828.

²¹⁶ A more obvious and nonhypothetical example of government business conducted in a language besides English is Puerto Rico.  But as a territory rather than a state, the considerations are potentially different.  See infra subsection III.B.3 (addressing the District of Columbia and the territories).

²¹⁷ See Stefan Th. Gries & Brian G. Slocum, Ordinary Meaning and Corpus Linguistics, 2017 BYU L. Rev. 1417, 1470–71; see also Jonathan H. Choi, Computational Corpus Linguistics 27–35, 66–83 (July 15, 2022) (unpublished manuscript) (on file with author) (explaining that proper use of the technique may require advanced techniques and comparisons among multiple corpora).

²¹⁸ See, e.g., Stephen C. Mouritsen, Kirkland & Ellis, https://www.kirkland.com/lawyers/m/mouritsen-stephen [https://perma.cc/VX9D-YKPK]; Corpus Legal Services, https://www.corpuslegalservices.com/ [https://perma.cc/7RUR-Y9JA].  Justice Thomas Lee recently left the Utah Supreme Court in order to open an appellate boutique and provide consulting in corpus linguistics.  Jessie Yount, Utah Supreme Court Justice to Launch Corpus Linguistics Consultancy and Boutique Upon Retirement, Law.com: The Recorder (June 13, 2022, 12:00 PM), https://www.law.com/therecorder/2022/06/13/utah-supreme-court-justice-to-launch-corpus-linguistics-consultancy-and-boutique-upon-retirement/ [https://perma.cc/ACL8-RDY3].

²¹⁹ See Martin H. Redish & Carter G. Phillips, Erie and the Rules of Decision Act: In Search of the Appropriate Dilemma, 91 Harv. L. Rev. 356, 392 (1977) (writing that avoiding cost and inconvenience to the federal courts is the only federal interest that should be able to “outbalance a truly significant competing state interest”).

²²⁰ See supra notes 149–50 and accompanying text (discussing whether state methods are meant to apply outside of the state courts).

²²¹ Tex. R. Civ. P. 78.

²²² Shady Grove Orthopedic Assocs., P.A. v. Allstate Ins. Co., 559 U.S. 393, 405 (2010).  The statement in question was not part of an argument against using state methods or state legislative history as a general proposition.  Rather, Scalia’s argument concerned whether, for purposes of a Rules Enabling Act analysis, the federal and state directives both purported to govern the same issue.  See id. at 402–06.  Justice Scalia could be right that the best way to apply the Rules Enabling Act is to read federal rules without regard to their potential impairment of state legislative policies (which state legislative history might reveal).  But the objection is overdrawn if it counsels against using state methods generally, on judicial-administrative grounds.

²²³ Apache Deepwater, L.L.C. v. W&T Offshore, Inc., 930 F.3d 647, 654 (5th Cir. 2019); see also Pritzker v. Yari, 42 F.3d 53, 65–74 (1st Cir. 1994) (applying Puerto Rican interpretive principles).

²²⁴ A circumstance in which transfer might be sensible would be when a federal court in State A is interpreting a statute from State B, which has a peculiar interpretive method.  The federal court in State B would presumably be more familiar with that method.  So transfer may be appropriate, at least if State B is one of the states that does not allow certified questions from district courts.  See Gulf Oil Corp. v. Gilbert, 330 U.S. 501, 509 (1947) (noting the value of “having the trial of a diversity case in a forum that is at home with the state law that must govern the case, rather than having a court in some other forum untangle problems in conflict of laws, and in law foreign to itself”).

²²⁵ The Supreme Court has another option available to it, in that it can defer to the lower federal courts on questions of state law.  See Salve Regina Coll. v. Russell, 499 U.S. 225, 235 & n.3 (1991); Bishop v. Wood, 426 U.S. 341, 345–46 (1976).

²²⁶ See Friendly, supra note 26, at 407, 422.

²²⁷ See Amy Coney Barrett, Procedural Common Law, 94 Va. L. Rev. 813, 822–24 (2008); see also Field, supra note 193, at 971–73 (describing circumstances in which federal courts may “reject [an] aberrant state rule, without substituting a generalized federal rule, [which] allows courts sensitively to adjust state and federal interests”).

²²⁸ Byrd v. Blue Ridge Rural Elec. Coop., Inc., 356 U.S. 525, 535–40 (1958).

²²⁹ See generally Bruhl & Leib, supra note 135.

²³⁰ See id. at 1249–54.

²³¹ E.g., Edward J. DeBartolo Corp. v. Fla. Gulf Coast Bldg. & Constr. Trades Council, 485 U.S. 568, 575 (1988).

²³² Gluck, supra note 11.

²³³ See, e.g., Caleb Nelson, Avoiding Constitutional Questions Versus Avoiding Unconstitutionality, 128 Harv. L. Rev. F. 331, 331, 332 & n.12, 335 (2015) (citing a state variation).

²³⁴ See, e.g., Clark v. Martinez, 543 U.S. 371, 381 (2005); Richard A. Posner, Statutory Interpretation—in the Classroom and in the Courtroom, 50 U. Chi. L. Rev. 800, 814–15 (1983).

²³⁵ Cf. Alexander M. Bickel, The Least Dangerous Branch: The Supreme Court at the Bar of Politics 69–71, 180–82 (2d ed. 1986) (discussing techniques of delay and avoidance that may serve to preserve the federal courts’ legitimacy).

²³⁶ See generally Helen Hershkoff, State Courts and the “Passive Virtues”: Rethinking the Judicial Function, 114 Harv. L. Rev. 1833 (2001) (urging sensitivity to institutional differences between state and federal systems in formulating doctrines of justiciability).

²³⁷ Jennings v. Rodriguez, 138 S. Ct. 830, 836 (2018) (distinguishing between permissible avoidance interpretations and forbidden “rewrit[ing]”).

²³⁸ See supra subsection II.A.1 (discussing constitutional constraints on interpretive methods).  But cf. William N. Eskridge, Jr., All About Words: Early Understandings of the “Judicial Power” in Statutory Interpretation, 1776–1806, 101 Colum. L. Rev. 990, 1021–25 (2001) (showing that early American courts exercised an “ameliorative power” to narrow statutes that were unreasonably or unintentionally broad, a cousin to constitutional avoidance).

²³⁹ Skilling v. United States, 561 U.S. 358, 405–11 (2010) (paring down “honest-services” fraud statute to a core of prohibited bribery and kickbacks); Zadvydas v. Davis, 533 U.S. 678, 699, 701 (2001) (reading immigration statute to contain an implicit time limit on detention); cf. Scoville, supra note 49, at 594 (“[T]he severability of a state statute is a matter of state law, but [that rule is subject to] an Article III override in the event that application of the relevant state doctrine would require the completion of a severance that exceeds the limits of federal judicial power.”).

²⁴⁰ See Phelps v. Hamilton, 59 F.3d 1058, 1073 (10th Cir. 1995) (“Should we refuse to interpret the statute in the same manner as would the Kansas courts—i.e. narrowly—we would reach the unseemly result that the constitutionality of the statute would be determined by whether the challenge was brought in federal or state court.”); Gluck, supra note 11, at 1957 (citing the “paradoxical consequence” that federal courts would strike down more state statutes out of “respect” for state decisionmaking).

²⁴¹ Tunick v. Safir, 209 F.3d 67, 75–76 (2d Cir. 2000) (opinion of Calabresi, J.).  This approach draws indirect support from the many cases in which the Supreme Court has required certification or abstention to a obtain construction of a state statute, especially a criminal statute, that the state courts had not previously interpreted.  See, e.g., Virginia v. Am. Booksellers Ass’n, 484 U.S. 383, 395–98 (1988); Bellotti v. Baird, 428 U.S. 132, 143–52 (1976).

²⁴² Supra subsection I.C.1.

²⁴³ See supra note 71 (citing cases that cite the no-dynamism principle but also predict that the novel claim would not be recognized).

²⁴⁴ I agree with Gluck’s normative position on this point.  See Gluck, supra note 11, at 1939; see also Erbsen, supra note 85, at 649 (contending that “Article III . . . allows federal courts to mirror the lawmaking flexibility of state courts”).

²⁴⁵ See Byrd v. Blue Ridge Rural Elec. Coop., Inc., 356 U.S. 525, 535 (1958) (stating that “the federal courts in diversity cases must respect the definition of state-created rights and obligations by the state courts”).

²⁴⁶ 532 P.2d 1226 (Cal. 1975).

²⁴⁷ Id. at 1238–39, 1243.

²⁴⁸ See, e.g., id. at 1240 (citing state jury practices).

²⁴⁹ Cf. Bruhl & Leib, supra note 135, at 1270–71 (describing state court advantages in dynamic lawmaking).  My focus throughout has been on civil cases, but when a federal court is interpreting a state criminal statute (as when determining whether a police search was permissible in a subsequent federal prosecution), background principles of legality and notice could break the tie in favor of narrower criminal liability.  See Wayne A. Logan, Erie and Federal Criminal Courts, 63 Vand. L. Rev. 1243, 1288–90 (2010).

²⁵⁰ 155 U.S. 610, 614–16 (1895).  The Court did observe that federal courts would not have to apply a state limitations period that discriminated against federal claims or failed to give a reasonable chance to sue, and it suggested that the RDA’s “where they apply” caveat might justify that departure from state law.  Id. at 615.  I agree that a federal court can disregard such a problematic limitations period, though I would prefer to describe the decision as implied conflict preemption rather than an interpretation of “where they apply.”

²⁵¹ Supra note 26 and accompanying text.

²⁵² Supra Section II.B.

²⁵³ 28 U.S.C. § 1652 (2018).

²⁵⁴ Supra Section II.D (addressing the intended scope of state interpretive methodology).

²⁵⁵ Id.  But see Mont. Code Ann. § 1-2-101 (2021) (“In the construction of a statute, the office of the judge is . . . .”) (emphasis added).

²⁵⁶ See The Federalist No. 80 (Alexander Hamilton).

²⁵⁷ See Tara Leigh Grove, The Structural Case for Vertical Maximalism, 95 Cornell L. Rev. 1, 33–40 (2009).

²⁵⁸ See Klaxon Co. v. Stentor Elec. Mfg. Co., 313 U.S. 487, 495–97 (1941).  State choice of law is subject only to minimal federal constraints.  A state may, so far as federal law is concerned, choose its own law as long as the state has a slight connection to the case, even where any objective assessment of the relative interests of the affected states would find that the other state’s interests predominate.  Cf. Allstate Ins. Co. v. Hague, 449 U.S. 302, 312–13 (1981).

²⁵⁹ See, e.g., King v. King (In re King), 434 P.3d 502, 510 (Or. Ct. App. 2018) (“When we apply the substantive law of another state, . . . we use that state’s ‘statutory interpretation principles and related case law’ to ascertain how courts of that state would apply its law.” (quoting CACV of Colo., LLC v. Stevens, 274 P.3d 859, 863 (Or. Ct. App. 2012)); cf. Zachary B. Pohlman, State Statutory Interpretation and Horizontal Choice of Law, 70 Kan. L. Rev. 505, 515–24 (2022) (citing inconsistency in judicial practices); Hart, supra note 148, at 1829–32 (same).

²⁶⁰ See, e.g., Weber v. U.S. Sterling Secs., Inc., 924 A.2d 816, 823 n.5 (Conn. 2007) (holding that interpretive methods were “procedural” and so applying forum methods).

²⁶¹ Erbsen and Green endorse such a “horizontal Erie” duty of fidelity to sister state law, though not specifically with regard to statutory interpretation.  See Erbsen, supra note 85 at 655; Michael Steven Green, Horizontal Erie and the Presumption of Forum Law, 109 Mich. L. Rev. 1237, 1262 (2011).  Pohlman argues that the Full Faith and Credit Clause does not require use of enacting-state methodology, but state choice-of-law doctrine generally does.  See Pohlman, supra note 259, at 542–43, 555.

²⁶² Sun Oil Co. v. Wortman, 486 U.S. 717, 731 (1988); see also supra note 112 (discussing the potential role of the Full Faith and Credit Clause in statutory interpretation).  The Court’s caveat about “brought to the court’s attention” means that when a litigant fails to present the sister-state’s interpretive methods, the forum may apply its own methods or assume that the two state’s methods are the same.  Sun Oil, 486 U.S. at 731.  Cf. Dicey, supra note 153, at 332–33 (same under English choice of law).  But see Green, supra note 261, at 1264–66 (arguing that a state court’s duty to correctly apply sister-state law is not owed only to the parties).

²⁶³ Cf. Shady Grove Orthopedic Assocs., P.A. v. Allstate Ins. Co., 559 U.S. 393, 409 (2010) (plurality opinion) (“New York has no power to alter substantive rights and duties created by other sovereigns.”); Roosevelt, supra note 133, at 25 (stating that “states . . . lack power to make other states’ laws”).  I address the possibility of federal constitutional duties on state courts to follow sister-state interpretive methodology in forthcoming work.  Aaron-Andrew P. Bruhl, Improving (and Avoiding) Interstate Interpretive Encounters, 2022 Wis. L. Rev. (forthcoming).

²⁶⁴ See Pohlman, supra note 259; Hart, supra note 148.  A duty to apply sister-state methods to sister-state statutes as a matter of choice of law is subject to exceptions along similar lines to the exceptions afforded federal courts in subsection III.A.1.  A state could have legitimate reasons of judicial administration to want to avoid a complicated, unfamiliar method like corpus linguistics that a sister state practices.  Or consider a textualist state court that believes that its own state constitution or fundamental public policies prohibit it from considering legislative history even when interpreting a different state’s statute.  See Bruhl, supra note 263 (addressing exceptions to the duty to follow sister-state interpretive methodology).

²⁶⁵ See Hart, supra note 148, at 1830–32; see also Restatement (Third) of Conflict of L. § 5.08(b) (Am. L. Inst., Tentative Draft No. 1, 2020) (“Ordinarily, the court should determine foreign law in light of how it is authoritatively interpreted and applied in the foreign state.”).

²⁶⁶ See Waialua Agric. Co. v. Christian, 305 U.S. 91, 109 (1938) (pre-statehood Hawaii).  Still other questions arise in connection with tribal law and the law of foreign countries, but I do not address those here.

²⁶⁷ E.g., Gatewood v. Fiat, S. p. A., 617 F.2d 820, 824 (D.C. Cir. 1980).

²⁶⁸ E.g., Waialua Agric. Co., 305 U.S. at 109 (“arguments of policy”); Edwards v. HOVENSA, LLC, 497 F.3d 355, 360 (3d Cir. 2007) (Virgin Islands Revised Organic Act); Lee v. Flintkote Co., 593 F.2d 1275, 1278 n.14 (D.C. Cir. 1979) (policies behind the District of Columbia Court Reform and Criminal Procedure Act and Erie).