Finding Customary International Law

I.     Introduction

Questions about customary international law (“CIL”) arise in a wide variety of federal litigation. For example, the Alien Tort Statute (“ATS”) provides the district courts with jurisdiction over “any civil action by an alien for a tort only, committed in violation of the law of nations or a treaty of the United States.”1 In Sosa v. Alvarez-Machain, the Supreme Court held that the statute’s reference to the “law of nations”—the old term for CIL—empowers federal courts to recognize common law causes of action for violations of international norms that are “accepted by the civilized world” and defined with adequate specificity.2 Plaintiffs have since relied on the ATS to bring tort claims against multinational corporations, foreign and U.S. officials, governments, and private individuals for torture, extrajudicial killing, crimes against humanity, and war crimes, among other acts.3 In many of these cases, a central question is whether custom prohibits the alleged conduct.

As another example, the Foreign Sovereign Immunities Act (“FSIA”) provides that foreign states are immune from the jurisdiction of federal and state courts unless a statutory exception to immunity applies.4 One such exception strips immunity in certain cases “in which rights in property taken in violation of international law are in issue.”5 To decide whether this language applies, federal courts must often identify the circumstances in which CIL prohibits governmental takings of private property.6

Similar issues arise in criminal cases. One federal statute establishes a penalty of life in prison for anyone who, “on the high seas, commits the crime of piracy as defined by the law of nations, and is afterwards brought into or found in the United States.”7 In applying this text, courts have looked to CIL to resolve issues such as whether piracy encompasses acts of aiding and abetting,8 and whether it includes acts of violence that do not accompany a robbery.9 Elsewhere, judicial efforts to discern custom are often necessary to apply the so-called Charming Betsy canon, which requires a court to interpret an ambiguous statute in a way that avoids conflict with international law.10

Finally, questions about CIL arise in constitutional interpretation. In United States v. Bellaizac-Hurtado, for example, the government prosecuted several foreign nationals under the Maritime Drug Law Enforcement Act (“MDLEA”) for engaging in drug trafficking in the territorial waters of Panama.11 In response, the defendants argued that the statutory prohibition is unconstitutional—while Article I grants Congress the power to “define and punish . . . Offences against the Law of Nations,”12 and while that power confers authority to criminalize conduct that CIL prohibits,13 custom does not prohibit drug trafficking.14 Persuaded by this argument, the court invalidated the statute as applied and vacated the convictions.15

As a matter of international law and federal common law, it is clear what courts should do when faced with these kinds of issues. The traditional doctrine is that a rule of CIL arises from a “general and consistent practice of states followed by them from a sense of legal obligation.”16 A correct analysis would thus seek to answer two questions: first, is there a general and consistent state practice on the given issue? And second, to the extent such a practice exists, have states engaged in it out of a sense of international legal obligation (otherwise known as opinio juris)? With respect to the first question, the predominant view is that a practice cannot be “general” unless, at a minimum, a “large share of [the] affected states” have engaged in it,17 that the requirement of consistency necessitates steady adherence even if not “absolutely rigorous conformity,”18 and that the occurrence of a practice over a substantial period of time supports, but is not strictly necessary for the development of a norm.19 It is also widely accepted that state practice can take diverse forms, including diplomatic acts, statutes, judicial decisions, official statements, and regulations, among others.20 The purpose of the second question is to ensure that measures undertaken as a matter of mere courtesy, habit, or policy do not automatically acquire the status of law.21 A critical consequence of these rules is that questions about CIL are substantially empirical; one cannot identify a norm without first investigating the practices and views of the various states that comprise the international order. This limitation reflects the positivist notion that custom’s legitimacy depends upon state consent.22

But while the basic rules for ascertaining CIL are fairly clear,23 adhering to them presents a serious challenge for the federal judiciary. In part, the difficulty is one of limited resources, both institutional and human: courts have easy access to cases, statutes, and other types of domestic authority for questions about U.S. law, but the same is not true with respect to international custom. Court filings tend not to reflect exhaustive analyses.24 There is no complete database—electronic or otherwise—of state practice and opinio juris.25 And any judge bold enough to attempt research into original sources would immediately confront linguistic, cultural, and historical barriers that could challenge even a seasoned comparativist.

Say, for instance, litigants raise a question about the scope of diplomatic immunity for special envoys under CIL. An examination of the official practices of China and Russia—two of the most influential states in the contemporary order—would seem to constitute a necessary part of any thorough analysis. Yet where does one go to identify those practices? Foreign materials are a daunting option. Even if a judge manages to collect them, she would also have to arrange for translation and then learn enough about the Chinese and Russian legal systems to appreciate the sources’ meaning and significance in context. Moreover, even if the judge overcomes these hurdles and correctly identifies the foreign state practice, she would still face the additional task of determining whether China and Russia pursued that practice out of a sense of international legal obligation rather than for some other purpose. These difficulties only grow as the inquiry shifts to the practices of smaller and less connected countries in the developing world. The result is substantial epistemic uncertainty.

This Article seeks to answer a simple question: faced with an imperative to ascertain customary international law, and with the various institutional and other limitations that interfere with rigorous analysis of state practice and opinio juris, how do federal courts proceed? In other words, how do courts go about finding custom?

The answer will carry real significance, both for litigants and U.S. foreign relations. As the preceding examples illustrate, the contours of CIL can dictate whether victims are entitled to compensation from perpetrators of mass atrocities, decide foreign sovereign immunity, and determine the ambit and constitutionality of criminal statutes. Moreover, the process by which federal courts ascertain custom may either contribute to or undermine U.S. foreign relations and the international rule of law. For instance, if courts respond to resource limitations by engaging in variant and unpredictable research strategies or relying on untrustworthy secondary sources, then it may be difficult for parties to anticipate the content of custom on their own, and U.S. courts might adopt conclusions that are at odds with prevailing global norms. Further, analysis that is under-inclusive with respect to the number of states considered might weaken the legitimacy of custom by exposing it as a set of rules over which many states have little to no de facto influence.

Given the stakes, it is no surprise that the judicial application of CIL has attracted a lot of attention over the years. Scholars have scrutinized various dimensions of the topic, including judges’ tendency to think about international law through the lens of constitutional law26 and the risk of norm fragmentation that accompanies the application of CIL in national fora with divergent legal traditions and political cultures.27 Scholars have also surveyed the Supreme Court’s application of CIL from the founding of the United States to the close of the twentieth century,28 examined the use of CIL in statutory interpretation,29 collected anecdotal evidence of judicial methods,30 and elaborated on the various problems that attend judicial efforts to apply custom.31 What is missing is a systematic, empirical account of contemporary methods of proof.

In an effort to improve understanding of the process by which federal courts ascertain custom, I conducted a study of the citations contained in all reported federal cases in which a court has drawn a legal conclusion about the existence or contours of a customary international norm since the Supreme Court’s decision in Sosa v. Alvarez-Machain.32 The cases span a little more than a decade and include over 250 separate conclusions on a wide variety of issues from district and appellate courts throughout the country. Given Sosa’s admonition for courts to scrutinize custom-based claims,33 this collection should include the federal judiciary’s most rigorous analyses of CIL. In evaluating the results, I hope to identify trends and problems in the judicial method, and generate ideas on how to mitigate current challenges.

This project draws heavy inspiration from the work of Stephen Choi and Mitu Gulati, who recently examined how international courts go about finding custom.34 Based on a collection of judgments from the International Court of Justice and its predecessor, plus a handful of decisions from other international tribunals, Choi and Gulati report on the kinds of materials cited in analyses of customary international law, the types of evidence that are most likely to yield affirmative conclusions about the existence of a norm, the relationships between norm categories and citation patterns, and the effects of variables such as time period and opinion status, whether majority or dissent.35 Their work supplies valuable insights. They find, for example, that “treaties are the most important type of evidence in CIL determinations,”36 that UN materials and domestic statutes “also play a big role,”37 and that tribunals rely on academic sources somewhat infrequently.38 What distinguishes my study is principally the target: U.S. courts rather than international tribunals. I anticipate that a comparison of results will further clarify the features of CIL analysis in these different settings.

The Article proceeds as follows. Part I explains the research methodology. Part II provides context for the results by reporting on the kinds of cases, courts, and legal questions that are generating precedent on customary international law in the United States. Part III reveals the key results, including data on the types and nationalities of sources that federal courts have considered in reaching conclusions about custom. These data show three clear patterns: parochialism, multilateralism, and “delegation.”39 First, judicial analyses are parochial in the sense that they rest primarily on U.S. government sources, particularly federal case law. Official sources from other advanced democracies, such as the United Kingdom and France, also play a role, but only on occasion, and citations to the legal authorities of all other governments are virtually nonexistent. Second, judicial analyses are multilateral in the sense that courts often cite important treaties and the work of international organizations. Third, analyses frequently delegate the task of ascertaining CIL to scholars by relying in part on academic writing in books and law reviews, particularly from American authors. As a doctrinal matter, this particular maneuver would be unproblematic if the academic sources reflected meticulous and inclusive studies of state practice and opinio juris, but they often do not. Conducting a separate study of citations contained in the academic sources to which courts have cited, I find that much of the academic writing on which courts have relied exhibits the same tendencies as the judicial opinions. Most of the cited scholars did not attempt to survey state practice and opinio juris, and even when they did they focused principally on the West. To this extent, contemporary CIL retains the under-inclusive and overwhelmingly occidental genealogy of the historical law of nations.40 Part IV explores the implications of these findings, and Part V offers ideas to help courts grapple more effectively with the challenge of finding custom.

The analysis proceeds on certain normative assumptions. In particular, I assume the traditional test for CIL is a good one, that legal practice and formal doctrine should align, and that a globally inclusive approach to the identification of custom is desirable. One might fairly question some of these positions, but I view such debates as beyond the scope of the present analysis. My primary purpose is description.

II.     Methodology

I performed a study of the volume and types of citations that federal judges include in their opinions in the course of reaching conclusions about customary international law. A key assumption behind this project—that citations matter—underlies a substantial collection of citation-count studies41 and likely holds true for a couple of reasons. One is that citations provide at least partial evidence of the internal reasoning that drives judicial decision-making—judges typically do not cite sources they never considered, and typically do cite sources that mattered most in their reasoning. The other is that citations influence subsequent litigation, in part because they confer on the chosen sources an authority that increases the likelihood of their future use, and in part because they signal permissible research and argument strategies for other parties and judges. Citation patterns thus offer valuable insight into past and future case outcomes.

The study proceeded in several steps. I started by performing a Boolean search42 of all reported federal43 opinions that mention customary international law from July 2015 back to the date of the Supreme Court’s 2004 decision in Sosa v. Alvarez-Machain,44 yielding a list of over 700 cases. I chose this time period in part because it generated a little more than a decade of precedent—enough to understand how contemporary courts ascertain custom, but not so much as to be insurmountably voluminous. I also chose this period because, in theory, most CIL analysis in the post-Sosa era should be qualitatively different than that of earlier periods. Sosa’s principal lesson is that federal courts must carefully scrutinize CIL claims—a court needs to ensure that the alleged norm is not only “accepted by the civilized world,” but also defined with adequate specificity.45 One would expect courts operating under this new instruction to be more exacting in their evaluation of custom. Of course, CIL questions also arise outside of the ATS context with which Sosa was concerned, but ATS cases give rise to the majority of CIL analysis, and courts have cited the decision as instructive in other types of litigation, both civil and criminal.46 Moreover, Sosa aside, the advancement of international connectivity over the past decade should have made the task of collecting evidence of CIL easier in recent years than at any other point in U.S. history.47 If the federal courts do not engage in thorough analyses of foreign state practice and opinio juris even now, we might reasonably assume that earlier opinions exhibited the same tendency to a greater degree.48

With the initial search results assembled, I proceeded to subtract any irrelevant or invalidated hits. I discarded all opinions that mentioned CIL without analyzing a customary norm. I also discarded opinions that reached a conclusion about CIL solely on the basis of binding precedent, the logic being that such decisions reveal little about how judges ascertain custom when forced to undertake their own research and analysis. Finally, I discarded all opinions that have been vacated, reversed due to flawed CIL reasoning, abrogated, or withdrawn. Approximately 120 opinions remained.49

The next step was to develop a coding strategy for what was left. My approach was to identify each conclusion about a customary norm, and then categorize by source type and nationality each citation made en route to that conclusion. This line of attack reflected a number of strategic choices. First, I coded and aggregated citations separately for each CIL conclusion within an opinion rather than treat each opinion as an undifferentiated whole, on the assumption that the additional granularity would be more revealing. Imagine, for example, that an opinion contains two separate conclusions—one concerning an alleged norm prohibiting material support for terrorism and another on child labor. A study that treated the whole opinion as a single unit would not only gloss over potentially significant differences in the quality of treatment for each issue, it would also complicate the task of identifying any norm-specific trends in the case law.

Second, by focusing on citations, I gave no credit for non-cited sources even if an opinion implied that a judge had considered them. As a result, the data understate the volume and variety of evidence that courts examined.50 But the choice is defensible. As a practical matter, it was difficult to identify sources with precision when courts referred to them only vaguely, and guessing would have created a reliability problem. Cited sources tend to be more influential to courts and parties for the reasons stated above.51 And the problem of references to non-cited materials was relatively uncommon. It is safe to say that the results reflect a clear majority of the sources that directly informed the judicial analysis, and probably all of the most influential sources.

Third, I coded for citations made en route to each CIL conclusion, rather than citations that judges relied upon to justify each conclusion. This choice means that I counted not only supporting citations, but also citations to sources that courts found unpersuasive or otherwise distinguished or rejected. A narrower focus might also yield intriguing results, particularly for future litigants who want to know which types of authority courts find most persuasive. My approach, however, offers a more reliable commentary on the quality and capacity of judicial analysis—limiting the inquiry to sources that judges relied upon could substantially underreport the rigor of any opinion that conscientiously rejected proffered evidence of CIL or its absence.

Fourth, I counted a citation only once regardless of how many times a judge might have repeated it in relation to a particular CIL conclusion. This practice differs from that of other citation studies52 and makes it difficult to identify the particular sources that were most influential in any given decision. At the same time, it is sufficient for a project focused on the evaluation of source types and nationalities. It is consistent, moreover, with the notion that the quality of an analysis substantially depends on the extent to which it has identified evidence of “a general and consistent practice of states”53—i.e., the idea that the diversity of sources matters at least as much as the varying depths of their treatment.

As the last component of the coding strategy, I excluded citations that appeared unrelated to the specific norms in question. For example, the results do not include citations made in the course of general explanations of how CIL works and is found, such as the Restatement (Third)’s discussion on the types of state practice on which custom might rest.54 Nor do they include citations made to explain the relationship between CIL and domestic law, such as citations to the Sosa opinion.55

Having settled upon this approach, the final step was to organize and record the results. I identified four categories of citations and divided them into 15 subcategories:

A) Foreign Citations: (1) direct evidence of foreign state practice, such as foreign judicial decisions, statutes, constitutional provisions, and the like, plus the name of the originating state;

B) Delegation Citations: (2) law review articles, books, and treatises; (3) amicus briefs and expert declarations; and (4) NGO reports, plus the nationality56 of the author(s) for each citation;

C) Multilateral Citations: (5) multilateral57 treaties that the United States has ratified;58 (6) multilateral treaties that the United States has not ratified; (7) UN Security Council resolutions; (8) UN General Assembly resolutions; (9) opinions from international courts; (10) UN reports; and (11) official documents of any kind from any other international organization;

D) Domestic Citations: (12) federal statutes; (13) federal and state judicial opinions; (14) statements and reports from Congress; and (15) statements, reports, and regulations from the Executive Branch.

I recognize that the application of these rules entails an element of judgment that may complicate efforts to replicate the results. This is a common problem in citation-count studies.59 To minimize it and ensure internal consistency, I separately coded the opinions twice and harmonized the results. Moreover, most coding decisions did not present close questions, so other researchers are likely to find the same basic trends even if they disagree with some of the decisions I have made.

III.     Context

Questions about custom tend to arise in certain types of cases. To identify those types, I collected information for each CIL conclusion on the name of the authoring court, publication year, case type, topic, defendant type, and geographic location of the alleged conduct. I also noted whether the judge held in favor of or against the existence of the norm in question. This information provides context for the citation data that follows and illuminates the stakes, both for the parties and U.S. foreign relations.

A number of contextual features stand out. First, judges are generally amenable to finding custom: 76% of the analyses yielded an affirmative conclusion about the existence of a norm. The percentage in criminal cases (83%) was slightly higher than it was in civil cases (75%), and within the latter grouping the numbers ranged from 100% for decisions under the FSIA to 73% under the ATS and 71% for habeas. The variation likely reflects the types of norms that these cases implicated. Criminal cases often raised questions about norms on extraterritorial jurisdiction that are, for the most part, domestically uncontroversial.60 Civil litigation, by contrast, typically involved claims based on human rights norms, some of which are contested.61 Yet the habeas and ATS percentages are still quite high and suggest that courts are not hostile toward CIL per se even post-Sosa.62

Second, there is an uneven distribution of CIL conclusions among the circuits, as demonstrated in Figure 1. Jurisdictions containing a number of the largest and perhaps the most cosmopolitan urban centers—specifically, the Second, Ninth, and D.C. Circuits—generated the largest volume of conclusions, with the Second Circuit being the clear leader among the group. On the other end of the spectrum, the Sixth, Eighth, and Tenth Circuits hardly ever addressed questions about custom. If this pattern also extends back into the pre-Sosa cases, we might reasonably anticipate differences in the comfort and skill with which judges in the various circuits go about ascertaining CIL.


Figure 1. Post-Sosa CIL Conclusions by Circuit



Third, there is an uneven distribution of conclusions among the various categories of cases. As shown in Figure 2, a clear majority (68%) occurred in ATS litigation, followed by all non-MDLEA criminal cases (11%), FSIA cases (6%), residual civil cases (6%), habeas (5%), and MDLEA cases (4%). This distribution highlights the significance of the recent decision in Kiobel v. Royal Dutch Petroleum, where the Supreme Court dismissed an ATS claim for alleged violations of CIL that occurred abroad, on the reasoning that the statute provides jurisdiction only for “claims [that] touch and concern the territory of the United States . . . with sufficient force to displace the presumption against [the] extraterritorial application” of U.S. law.63 By all accounts, Kiobel dealt a serious blow to many who might hope to use ATS jurisdiction as a path to redress,64 and in the wake of the decision lower courts have dismissed a significant number of claims concerning foreign conduct.65 Because Kiobel’s focus on territorial nexus does not require an analysis of CIL, the long-term effect of the decision is likely to be a significant reduction in the application of custom in federal court. Depending in part on one’s view of the results below,66 this may or may not be a positive development.


Figure 2. Post-Sosa CIL Conclusions by Case Type



Other aspects of the context are also important. Figure 3 shows the topics that generated the largest number of conclusions, with aiding and abetting liability; prescriptive jurisdiction; torture; cruel, inhuman, and degrading treatment; and corporate liability constituting the five that arose with the greatest frequency. Figure 4 identifies the most common categories of defendants in these cases—corporations, foreign nationals, U.S. nationals, foreign officials, and foreign governments. Figure 5 in turn reveals the geography of the acts that gave rise to the collected cases. A relatively large number of conclusions implicated conduct that occurred in international waters—a reflection of piracy cases, both civil and criminal, and convictions under the MDLEA, which federal authorities have relied upon to combat drug trafficking. Otherwise, CIL conclusions typically concerned the legality of acts that occurred at least partly in the developing world.


Figure 3. Post-Sosa CIL Conclusions by Issue Volume



Figure 4. Post-Sosa CIL Conclusions by Defendant Type



Figure 5. Post-Sosa CIL Conclusions by Site of Conduct



Together, these figures confirm a common impression about CIL litigation in federal courts—namely, these are typically not domestic disputes with merely coincidental international elements, but rather international disputes that happen to be in domestic court. Thus the stakes: because the cases often involve foreign defendants and at least an element of extraterritorial conduct, the quality of the analysis will often carry foreign policy consequences. And because many of the parties affected are either criminal defendants or alleged victims of serious human rights abuses, there is a real imperative for careful research and reasoning. The next task is to identify how judges proceed in the face of that imperative.

IV.     Results

This Part reports the study results.67 I begin by discussing findings concerning the quality of CIL analysis in the courts and then proceed to identify three salient trends of parochialism, multilateralism, and delegation. For each of these, I provide data and hypothesize on contributing influences.

The data support a number of conclusions about the quality of judicial analysis, as measured by citation count. First, quality varied by jurisdiction. In general, courts ascertained CIL on the basis of relatively few authorities of any kind. The cases cited an average of slightly less than eight (7.8) and a median of five different sources per conclusion. Yet the numbers for some courts were noticeably higher. Appellate courts averaged nearly ten (9.9) sources per conclusion—quite a bit more than the district courts (6.5). Moreover, the appellate courts that handled the largest volume of CIL questions had a tendency to engage in the most thorough analyses, as shown below in Figure 6. This particular result is consistent with the possibility that courts familiar with CIL claims are more willing and able to perform a full, thorough analysis, although it might also indicate a tendency among lawyers who practice before certain courts to provide more substantial briefing on these issues.


Figure 6. Average Number of Sources Cited Per CIL Conclusion in Federal Appellate Opinions



Second, quality of analysis varied moderately among the different types of cases. As shown below, opinions in ATS cases exhibited the highest average number of sources cited per conclusion (8.4). Criminal cases had the second-highest average (7.3), followed by FSIA (6.6), habeas (6.2), and other civil cases (4.5). Yet criminal and FSIA cases had the highest medians of 6.5 and 5.5, respectively. Observers might reasonably react to these numbers in different ways. On the one hand, the high average for the ATS, particularly in comparison to other civil cases, is consistent with Sosa’s instruction for lower courts to scrutinize ATS claims,68 and the high median in criminal cases is comforting if one believes in a need for particularly careful research and analysis where personal liberty is at stake. On the other hand, some might have expected an even bigger difference between the criminal and civil cases. A significant number of opinions from the criminal context were comparable to or even less rigorous than those of the civil variety.69


Figure 7. Sources Cited Per CIL Conclusion by Case Type



B.     Parochialism

Moving from citation volume to type, the most conspicuous pattern is that of parochialism: the opinions focused extensively on American practices and perspectives.70 Among the four categories, domestic citations were by far the most common, making up nearly half (48%) of all citations in the data set.71 As shown in Figure 8, multilateral (28%), delegation (19%), and foreign (4%) citations trailed far behind. Figures 9 and 10 accentuate the point by reporting the number of citations to official government sources from the most-cited countries. These show that an overwhelming 89% of national government sources were attributable to the U.S. government, followed by the British (3%), Russian (2%), and French (2%) governments. Citations to government sources from all countries in Africa, Asia, Central America, the rest of Europe, Oceania, and South America made up a combined total of only 32% of foreign government sources and 4% of all national government sources.72 And the focus on U.S. authority was even more pronounced than these numbers suggest—a clear majority of the direct citations to British, French, and Russian state practice concerned actions undertaken jointly with the United States in the immediate aftermath of World War II, such as the enactments of the Allied Control Council.73 If one were to remove those from the tabulation, citations to official sources from all foreign governments combined would constitute only 5% of national government sources. From this perspective, U.S. government authorities exerted overwhelming and nearly exclusive influence on federal judicial analysis of CIL.


Figure 8. Post-Sosa Citations by Category74



Figure 9. Citations to National Government Sources by Country



Figure 10. Citations to Foreign Government Sources by Country



Within the national government sources, certain types of authority appeared more often than others. By far the most frequently cited U.S. authority was non-binding federal judicial precedent, which made up 85% of all domestic citations.75 References to executive branch reports and statements (8%), federal statutes (5%), and legislative reports and statements (2%) were relatively uncommon, and references to the practices of the fifty states were non-existent. In contrast, treaty participation was the most commonly used direct evidence of foreign state practice (64%), followed by foreign judicial decisions (18%), constitutional provisions (9%), treaty reservations (4%), statutes (4%), and diplomatic statements (1%). Figures 11 and 12 display these results.


Figure 11. U.S. Government Sources by Type



Figure 12. Foreign Government Sources by Type



Courts also used foreign and U.S. government sources in different ways. The former usually operated as evidence of state practice.76 In contrast, courts often employed domestic authorities not as evidence of state practice or opinio juris per se, but instead as persuasive evidence regarding the propriety of a particular conclusion about a contested norm.77 To put the distinction in the language of international law, courts were more likely to apply U.S. government sources—mostly case law, but occasionally statutes and executive actions as well78—as a “subsidiary means for the determination” of custom.79 This is intriguing in part because it is hard to reconcile with the traditional doctrine of sources, which identifies “judicial decisions” as the only government source that can serve as a subsidiary means.80 But it is also noteworthy because the tendency to use domestic authorities in this way substantially empowers them.81 Rather than view a federal case or statute as a single point in a global constellation of national practices, courts often relied on domestic authorities as reliable and independently sufficient evidence of an international norm.

Sarei v. Rio Tinto PLC provides an extreme but not uncommon example. There, residents of Papua New Guinea sued Rio Tinto for violating an alleged norm against environmental degradation in the course of mining operations on the island of Bougainville,82 but the court dismissed the claim as “based on international law norms that have not, as yet, achieved the status of matters of ‘universal concern.’”83 The court did not cite any foreign or international sources for support.84 Instead, it relied exclusively upon decisions from the Second and Fifth Circuits and a collection of precedent from other district courts.85 Implicitly, these cases were deemed to reflect not simply the practice or position of the United States, but evidence of a general and consistent global practice backed by opinio juris.

A number of factors may explain the case law’s inward orientation. Perhaps the most likely and significant is the vexing logistical challenge of locating, collecting, and evaluating direct evidence of the practices and views of numerous foreign states. This challenge would have two consequences: First, it would make it difficult for the parties—particularly those who are cost-sensitive—to supply the courts with relevant foreign authorities in their briefing.86 Second, it would limit courts’ ability to undertake independent research. On this view, courts and parties focus on U.S. government sources simply because it is the convenient thing to do; domestic authorities are readily available and require no translation or knowledge of foreign legal systems.

But logistics is probably not the only explanation. Some courts might cite fewer foreign official sources due to discomfort with the idea that foreign governments influence the content of substantive law in the United States.87 Often lacking in international legal training,88 American lawyers and judges might not understand what CIL entails.89 A strong national orientation might also reflect judicial habit; perhaps judges cite to federal precedent more frequently than other sources because that is an ingrained element of legal analysis. Or lower courts might simply be following the example of Sosa, where the majority did not cite any foreign government sources in the course of concluding that CIL fails to prohibit executive detention in excess of a state’s domestic legal authority.90

Social psychology might also play an intriguing role. Researchers have demonstrated that individuals have a tendency to assume that others share their opinions, values, and habits, and that this bias generates a perception of consensus even when consensus does not exist.91 Known as the false-consensus effect, the phenomenon occurs in a variety of social settings and has several empirically grounded explanations. One is selective exposure: because people tend to associate with individuals who are similar to themselves, they are more likely to encounter agreement than disagreement, and they extrapolate from that unrepresentative sample of experiences to anticipate the views of others.92 Another explanation emphasizes that the false-consensus effect can be logical insofar as people attribute their behavior to situational forces that others also encounter.93 Finally, some have found that the effect occurs because we derive functional value from perceiving similarity between self and others.94 A perception of similarity “may bolster perceived social support, validate the correctness or appropriateness of a position, maintain self-esteem, maintain or restore cognitive balance, or reduce tension associated with anticipated social interaction.”95

As applied to judicial analyses of CIL, the false-consensus effect may contribute to parochialism by excusing the failure to rigorously examine foreign state practice and opinio juris. The subconscious projection of familiar, American views onto an external world, in other words, would substantially obviate the need to canvas the foreign by assimilating it to the domestic. This might occur for any number of reasons. Judges, for one, are steeped in national legal culture and interact primarily with American lawyers who not only share their professional socialization, but also actively appeal to, and thus reinforce, the culture’s dominant sensibilities in order to prevail.96 This is helpful for the maintenance of domestic legal order, but it also amounts to a form of selective exposure that probably encourages the imputation of universality to what may be American idiosyncrasy. Standard rhetoric about America’s hegemonic position in global affairs probably contributes as well; it does not take much to go from recognition of the fact of American dominance to the conclusion that evidence of U.S. state practices and views are both necessary and sufficient to discern the content of custom.97 Moreover, judges might subconsciously derive value from perceiving global consensus. It is comforting to think the rest of the world shares our norms.

C.     Multilateralism

The second most significant citation pattern is multilateralism. As evidence of state practice and opinio juris, courts also cited significant treaties, international tribunals, and the work of various international organizations. Such sources made up 28% of all citations in the data set—less than domestic (48%), but more than delegation (19%) and foreign citations (4%).98 Within this collection, references to decisions by international tribunals were the most common at 31% of the total. Multilateral treaties that the United States has ratified were also common (26%), followed by treaties that the United States has not ratified (19%), UN Security Council Resolutions (9%), UN Reports (9%), UN General Assembly Resolutions (3%), and publications from other international organizations (3%), as shown below in Figure 13.


Figure 13. Citations to Multilateral Sources by Type



A relatively small number of sources generated the bulk of the citations for some of these subcategories. The International Criminal Tribunal for the Former Yugoslavia and the International Criminal Tribunal for Rwanda published a majority of the cited international judicial decisions.99 The Universal Declaration of Human Rights was the reason for most citations to UN General Assembly Resolutions.100 And the statutes for the international criminal tribunals were behind most citations to UN Security Council Resolutions.101

The data on multilateral treaties are also significant. Although ratification is an important route to treaty formation,102 the results show that a treaty’s ratification status in the United States did not carry much significance. As depicted in Figure 14, courts cited treaties to which the United States is a party only slightly more often than those to which it is not, and the cited academic works cited unratified treaties more frequently than ratified treaties. Some of these, moreover, were treaties to which the United States is also not a signatory103—the Rome Statute104 and European Convention on Human Rights105 were among the most-cited, as shown in Figure 15. This practice has the effect of partially circumventing the formal ratification process;106 while the United States is not bound to the unratified treaties as such, the courts treat as binding the CIL that these agreements may have played a substantial role in creating.107


Figure 14. Average Number of Citations to Ratified & Unratified Treaties



Also of note is the way in which the opinions discussed unratified treaties. Courts typically did not consider whether a decision against ratification might constitute a form of persistent objection to the norm in question.108 Nor did they typically consider whether treating an agreement as evidence of CIL would be appropriate as a matter of international law. Of the opinions that cited an unratified treaty, hardly any referred to the International Court of Justice’s famous decision in the North Sea Continental Shelf Cases, which set forth guidelines for deciding whether a treaty has helped to generate custom.109 While the executive branch has affirmed that some of these agreements codify CIL,110 and while judicial deference to the executive branch has been common in cases that implicate foreign affairs,111 it was also typical for courts not to look to the executive for guidance on an agreement’s CIL implications.112 And no courts inquired into treaty compliance. The typical approach seems to have assumed either that states comply113 or that the expressed intention to comply is sufficient to contribute materially to the formation of a norm.

Finally, the multilateral sources exhibited varying degrees of attenuation to state practice and opinio juris. In many cases, the link to one or the other was fairly clear, such as when courts cited resolutions of the UN Security Council114 or General Assembly.115 But in other cases it was not. It was not altogether uncommon,116 for example, for courts to cite the work of organizations composed of individuals rather than states, such as the Office of the Secretary General of the United Nations.117 This practice is significant because, while some dissent,118 an influential view holds that international organizations cannot contribute to the formation of CIL insofar as their actions are not attributable to states,119 and for the most part the executive branch seems to agree.120

The commonality of judicial citation to multilateral sources might have a number of explanations. First, and perhaps most importantly, these authorities are readily available; it is easy for parties and courts to access UN materials and other similar sources online.121 Second, courts might view international treaties and organizations as particularly powerful evidence of opinio juris, given the number of states they involve and the clarity with which they are capable of addressing contested issues. Finally, lower courts may be following the lead of Sosa, which discussed multilateral sources more than any other category en route to rejecting the appellant’s claim.122


Figure 15. Total Number of Citations to Unratified Treaties



D.     Delegation

Lastly, the results demonstrate a pattern of delegation. Courts frequently considered books, treatises, law review articles, amicus briefs, and expert declarations in the course of adjudicating questions about international norms. While this practice certainly occurs elsewhere as well,123 it appears to be far more common in the present context—delegation citations comprised nearly 20% of all citations in the data set. Yet the utility of this practice largely depends on the nature of the underlying sources; careful and inclusive academic analyses of U.S. and foreign practice and opinio juris would minimize the need for judicial consideration of official state sources, while incomplete, nationalistic or otherwise problematic studies could easily operate as a source of misunderstanding and error. In this Section, I analyze delegation citations from a multi-generational perspective by considering both delegation sources that judges cited (“Generation n”) and sources cited within the delegation sources (“Generation n–1”). Evaluating the two offers a sense for the utility of delegation and the proximate genealogy of custom in domestic court.

1.     Generation n

These citations displayed a number of characteristics. First, most referred to sources that appeared reliable, including many books and articles by influential scholars. Only 7% referred to student notes. In this regard, the cases were generally consistent with the Statute of the International Court of Justice, which provides that the teachings of only the “most highly qualified publicists of the various nations” can serve as subsidiary means for the determination of international law.124

Second, the citations were substantially concentrated in the sense that approximately one-third referred to only a handful of publications. Atop the list was the Restatement (Third) of the Foreign Relations Law of the United States, which alone accounted for 18% of the delegation cites—more than all sources from foreign authors, and more than any other individual source. Also influential, although much less so, were M. Cherif Bassiouni (3%), Eugene Kontorovich (3%), co-editors Robert Jennings and Arthur Watts (3%), and Antonio Cassese (3%), all of whom were cited in a number of cases and for a range of propositions. To this extent, one might reasonably assume that delegation has been a successful strategy. Indeed, given the serious logistical challenges to research in original sources, reliance on respected academics who have time to canvas state practice and opinio juris could be the single best method for ascertaining CIL.

But there are reasons for concern. For one, delegation citations also exhibited a strong national and Western orientation. A clear majority (61%) referred to works by American authors.125 Otherwise, the most common author nationalities were British (13%), German (4%), Swiss (3%), Italian (3%), Canadian (2%), Egyptian (2%), Dutch (1%), Ukrainian (1%), and Australian (1%), as shown below in Figure 16. These are similar to the numbers on government sources insofar as they document the outsized influence of the United States and, to a lesser degree, Western Europe. U.S. courts never cited a single work by a Chinese, Japanese, Indonesian, Pakistani, or Russian author. Only 9% of the delegation citations had non-Western authors, and many of those individuals have spent substantial time in the West.126


Figure 16. Post-Sosa Delegation Citations by Author Nationality



These sources probably contributed to the nationalist lens through which courts understood custom. As Anthea Roberts and others have shown,127 there is evidence that communities of international legal scholars in different states tend to approach international law with distinctive understandings and ideological orientations. Although the causation is complex, such variation probably has a lot to do with socialization.128 Individuals operating in the legal academic culture within a particular state typically graduate from the same law schools, acquire a standard set of professional experiences before entering academia, read each other’s work, and attend the same conferences and workshops, all while interacting infrequently with foreign colleagues who, meanwhile, experience their own professional socialization in distinctive foreign communities. Moreover, even those who remain incompletely socialized must still work within the predominant stylistic, doctrinal, and theoretical paradigms in order to appeal to their national colleagues. To report that courts cite most frequently to American authors, then, is to suggest that the courts often obtain their understanding of CIL in substantial part from individuals who have, quite plausibly, conveyed distinctly American views.129

A second reason for concern is that, although framed as relevant to questions about CIL, many delegation citations did not refer to academic sources for propositions about state practice or opinio juris. As an example, consider the Seventh Circuit’s decision in Flomo v. Firestone National Rubber Co., where the question was whether CIL provides for corporate liability for violations of international norms.130 Rejecting precedent from the Second Circuit as unpersuasive, the panel referred to a variety of federal case law, multilateral treaties, and books and law review articles en route to concluding that corporate actors can be held liable.131 The number and diversity of sources suggests a reliable conclusion, but at least half of the academic publications had no direct bearing on CIL. A citation to a book chapter by Robert Cryer established simply that “[t]he first international tribunal was the Nuremberg IMT [International Military Tribunal] which sat between 1945 and 1946 to prosecute high-ranking Nazis.”132 A citation to an article by John Coffee offered a rationale for the criminal punishment of corporations under domestic law.133 And so forth. Courts, in other words, relied on academics for a variety of propositions, some of which were quite tangential to the core task of identifying a norm. By my count, nearly half (45%) of the delegation citations did not even purport to identify or reflect state practice and opinio juris. To this extent, the academic sources were less helpful than one might imagine.

2.     Generation n–1

Despite the foregoing issues, we might nevertheless credit delegation as a useful strategy if the portion of cited academics who made representations about CIL did so on the basis of rigorous and inclusive studies of state practice and opinio juris. To determine whether they did, I conducted a separate analysis of the sources cited within the cited academic sources.

This segment of the study generally followed the same parameters I applied to the case law. Specifically, I identified the academic works that both contained and were cited for a proposition about contemporary CIL, and then coded for the citations the author made with regard to that proposition. Thus, I excluded books and articles cited for other purposes, such as insights on the historical law of nations, discussions about current events, or purely theoretical or admittedly aspirational arguments about contested norms, and within the remaining works I disregarded any citation unrelated to the CIL propositions for which courts cited them. To generate a standard measure of quality, I also excluded sources in which the author did not attempt to follow contemporary citation conventions for legal scholarship. And, as in my survey of case law, I focused only on sources published since Sosa. This cut-off date helps to focus the inquiry on how authors and citing courts implement current understandings of CIL134 and provides a more distinctive sense for the state of recent scholarship. It also guarantees that courts and the authors they cited have analyzed CIL in the same geopolitical context. This latter point is important because the number of states has risen sharply since World War II as a result of decolonization135 and ethnic conflict.136 Works published in prior decades, particularly pre-1980, would understandably cite to the practices of a smaller number of states for the simple reason that fewer states existed.137

Operating within this framework, I coded and analyzed the citations contained within a collection of more than seventy books, articles, and amicus briefs. These citations support several conclusions. First, as measured by citation count, the academic sources seemed to offer comparatively rigorous analyses (see Figures 17 and 18). On average, the books and articles cited more than twice as many different sources per conclusion (16.5) as the judicial opinions (7.8), and the amicus briefs cited more than four times as many (35). These sources also had higher medians: 28 sources per conclusion in the briefs and seven in the books and articles, compared to only five in the judicial opinions. The academic sources, moreover, cited direct evidence of state practice and opinio juris with greater frequency—while the courts averaged only one foreign authority for every three analyses (.33), the books and articles averaged three per analysis (3), and the amicus briefs averaged nearly eight (7.9). From this perspective, delegation appears to be a useful strategy.


Figure 17. Quality of Analysis by Volume of Cited Sources



Figure 18. Quality of Analysis by Volume of Citations to Foreign Government Sources



Yet there are again reasons for concern. Like the judicial opinions, most books and articles did not attempt to survey direct evidence of state practice and opinio juris. Instead, they focused primarily on the work of other academics—41% of their citations referred to other scholarly publications. Citations to the various forms of official authority, whether domestic, foreign, or multilateral, were less common, as shown in Figures 19 and 20.


Figure 19. Citations in Cited Books & Articles, by Category



The amicus briefs present a slightly different, but nevertheless consistent, picture. Like the books and articles, a clear majority of them did not attempt to survey state practice and opinio juris. But two did, and the sample size of only 23 briefs was small enough to allow these to substantially affect the results. The first brief, by Philip Alston, canvassed an impressive array of foreign evidence in support of the proposition that CIL prohibits summary execution.138 The second, by Stefan Talmon, marshaled evidence from a significant number of governments across Africa, Asia, and Latin America to argue that there is insufficient agreement on the meaning of “terrorism” under CIL to meet the test set forth in Sosa.139 Without counting these works, the briefs cited most frequently to academic books and articles (25%), international judicial decisions (16%), treaties to which the United States is not a party (10%), U.S. judicial decisions (10%), and treaties to which the United States is a party (7%). Upon counting the filings by Alston and Talmon, however, the numbers shift to show that amicus briefs cited most frequently to foreign state practice and opinio juris (20%), books and articles (18%), international judicial decisions (10%), UN reports (10%), and U.S. judicial decisions (9%). These results are in Figure 21.


Figure 20. Citations in Cited Books & Articles, by Subcategory



Figure 21. Citations in Cited Amicus Briefs, by Subcategory140



The citations in Generation n–1 also exhibited a second characteristic: the academic publications were substantially parochial in terms of which governments they cited. Official U.S. sources such as federal judicial opinions and statutes comprised 36% of all citations to national government sources in the cited books and articles and 41% in the cited amicus briefs. By comparison, the second-most-cited state was the United Kingdom, which comprised 14% of the citations to national government sources. Figures 22 and 23 illustrate these results. Comparing these numbers to the other data above, it becomes apparent that the international law scholarship to which judges referred typically exhibited a moderated but nevertheless significant degree of the same inward orientation as the judicial opinions.


Figure 22. Citations to National Government Sources in Cited Books & Articles



Figure 23. Citations to National Government Sources in Cited Amicus Briefs



The citations to national government sources were also under-inclusive in the sense that books and articles focused overwhelmingly on the advanced democracies of the West. In this scholarship, 83% of the national government sources documented official acts or statements of Australia, Canada, New Zealand, the United States, or Western Europe. Outside of exemplary works by Anthony Aust141 and co-authors Jean-Marie Henckaerts and Louise Doswald-Beck,142 authors virtually never cited any government sources from any states in Africa, Asia, or Latin America. To put the point more bluntly, most of the cited scholars thought it possible to identify CIL without even a brief glance at a majority of states.143

Amicus briefs exhibited a similar, but somewhat less exclusive, focus on the West. In this subcategory, the authors referred to official sources from the American, British, Dutch, French, German, and Italian governments, but all of these combined were only a little over half of all citations to official state sources at 52%. The remaining 48% comprised a long list of one or two citations to official sources from non-Western governments all over the world, from Afghanistan to Zimbabwe. This cosmopolitanism, however, was primarily a result of the two outlier briefs from Alston144 and Talmon.145 The rest focused exclusively on American and other Western government sources, if they used national government sources at all.146

Some forms of national authority appeared more frequently than others within the academic writing. With respect to U.S. government sources, authors referred most frequently to federal judicial decisions (82%), followed by executive branch sources (10%), statutes (5%), the Constitution (1%), and congressional reports (1%). Judicial decisions were also the most common among foreign government sources (61%), followed by statutes (24%), treaty participation (10%), and executive or administrative measures (5%). Figures 24 and 25 display these results.


Figure 24. U.S. Government Sources in Cited Books & Articles



Figure 25. Foreign Government Sources in Cited Books & Articles



Finally, the cited academic sources were substantially parochial in terms of the types of scholarship to which they referred. In a familiar pattern, a large percentage (40%) of the cited books and articles had American authors, with British (15%), German (5%), Italian (4%), and Sri Lankan (4%) authors rounding out the top five. The amicus briefs were similar. Figures 26 and 27 capture these results and show that the academic analyses were only moderately more cosmopolitan than the delegation citations in the judicial opinions, 61% of which had American authors.147


Figure 26. Nationality of Authors Cited in Cited Books & Articles



Figure 27. Nationality of Authors Cited in Cited Amicus Briefs



The condition of the cited scholarship may reflect a wide variety of influences. First, like judges, legal academics might be unable to overcome the logistical challenge of surveying foreign state practices and views. Second, the scholarship might reflect ideological considerations: some might intentionally deemphasize state practice on the belief that a focus on the current order is unduly apologetic of state power,148 particularly regarding matters such as human rights, while others might focus on the United States and other similar countries on the view that Western norms are superior to competing paradigms from other parts of the world. Third, the false-consensus effect might play a role. Attending conferences and sharing papers with other like-minded individuals from a small handful of states, some may view certain norms as uncontroversial and established even while a more global perspective could suggest otherwise.

It is possible to reject at least one of these, however: the ideological explanation appears to be unpersuasive insofar as the phenomenon of parochialism extends beyond analyses of human rights norms. Even upon limiting the data to scholarship on issues such as extraterritorial jurisdiction and the definition of piracy, the cited authors still exhibited a marked tendency not to survey state practice or opinio juris, and to focus on the United States and Western Europe when conducting such surveys. In non-human rights cases, citations to U.S. government sources were as common (14%) as citations to sources from all foreign governments combined (14%), as shown in the figures below. Moreover, a majority (55%) of the foreign citations referred to sources from the governments of Australia, New Zealand, and a small number of European states. And of the cited authors, 35% were American and 87% from the West. It is clear, in other words, that parochialism is a cross-topical phenomenon.


Figure 28. Citations in Cited Books & Articles on Issues Other than Human Rights, by Source Type



Figure 29. Citations to Foreign Government Sources in Cited Books & Articles on Issues Other than Human Rights 



Figure 30. Nationality of Authors Cited in Cited Books & Articles on Issues Other than Human Rights



V.     Implications

To recap, contemporary analyses of CIL in federal court reveal several noteworthy patterns. First, although the numbers varied by jurisdiction and case-type, courts reached conclusions about custom on the basis of relatively few sources. The typical opinion reflected an extremely limited sample of state practice and opinio juris, rather than an expansive global survey. Second, the limited sample typically had a decidedly parochial orientation. Courts frequently cited official U.S. sources, occasionally cited sources from other advanced democracies of the West, and almost never cited official sources from any other governments. Third, although less common than the use of U.S. government sources, courts often cited multilateral authorities, such as treaties and international judicial decisions. Finally, courts partially delegated the task of ascertaining CIL by considering, with unusual frequency, academic sources such as law review articles. Nearly half of the cited academic works, however, did not purport to identify CIL, and most of those that did made representations about CIL without conducting broad surveys of state practice and opinio juris. Most were written by American academics who cited most frequently to other academics, most of whom were Americans or others from the West. To the limited extent that these individuals considered direct evidence of state practice and opinio juris, they focused on the United States and, to a much lesser degree, other similar countries of the West. The practices of most states and the views of most foreign scholars were simply never considered. These findings have a number of important domestic and international implications for courts, practitioners, and scholars.

A.     Disuse of the Traditional Doctrine

First, and perhaps most obviously, federal courts have not followed the traditional doctrine. The typical opinion neither demands evidence that a state practice is truly “general and consistent,” nor assesses the motivation for that practice, nor pays any particular attention to its duration. CIL, according to actual federal practice, depends more on the laws and policies of the United States than those of the international community as a whole. Moreover, in relying primarily upon non-binding federal precedent to identify norms, U.S. courts act in tension with the international doctrine that the decisions of national courts are, at most, “subsidiary means for the determination of rules of law.”149 The typical approach has elevated a subsidiary means into the principal means.

This is not to say that the focus on domestic authorities necessarily yields conclusions that are erroneous under the traditional doctrine. The political branches have substantial resources with which to investigate thoroughly and report on the content of CIL.150 As the most influential state in the contemporary order, U.S. practice may ultimately constitute a fairly reliable indicator of global norms.151 The sociological phenomenon of cross-national institutional isomorphism suggests that the practices of one state might serve as reliable clues about the behavior of others.152 And the focus on domestic authority was not exclusive; courts also used multilateral and academic sources.

Yet the common method would seem to carry a non-trivial risk of error. Many U.S. government sources do not reflect rigorous examinations of state practice and opinio juris. The study results suggest that non-binding federal judicial precedents—by far the most common official source—often exhibit the same parochialism as the citing opinions.153 Moreover, while the political branches are capable of conducting more cosmopolitan and exhaustive inquiries, there are no studies on how Congress and the executive branch go about ascertaining custom, and, in any event, courts rarely cited U.S. government authorities other than judicial opinions. References to executive sources comprised only 4% of all citations, while references to statutes (2%) and other legislative sources (1%) were even less common.154 Academic publications were also of little help; most had Western authors who chose not to survey state practice and opinio juris in any sort of inclusive way. If courts have reached conclusions about CIL that are correct under the traditional doctrine, they have likely done so either coincidentally, with critical assistance from multilateral sources, or on the basis of an intergenerational game of telephone, whereby judges rely on the potentially distortionary say-so of domestic precedent that in turn relied on representations in older domestic precedent.

Many scholars and practitioners are of course comfortable with the idea that CIL has certain definite features—a norm of diplomatic immunity, a prohibition on piracy, an allowance for extraterritorial jurisdiction, and the absence of a general rule against environmental degradation, to name just a few. And for the most part, the judicial decisions were consistent with those ideas. Courts, for example, consistently concluded that CIL permits governments to exercise extraterritorial jurisdiction based on the protective principle155 and to expropriate the property of their own nationals.156 For this reason, one might defend the case law on the ground that it reliably captures the predominant sense of the U.S. legal community, methodological issues notwithstanding.

But that prevailing sense is largely a matter of faith. Strictly speaking, one cannot know that any rule qualifies as CIL under the traditional doctrine without evaluating state practice and opinio juris, and yet, few have conducted such an inquiry with respect to even a single rule at a particular moment in history, much less the entire corpus of norms that American judges and practitioners have come to accept as such. The problem of the false-consensus effect cautions against such faith and suggests that some norms might appear obviously correct simply because of our own insularity.157 Moreover, many of the cases raised questions that lack an obvious resolution even on principles of faith. Some claims required courts to decide not simply the existence of a norm but its precise features. In United States v. Ali, for example, the issue was not whether CIL prohibits piracy, a topic on which few in the American legal community would disagree, but whether that prohibition includes a ban on acts of aiding and abetting piracy,158 a topic on which the dominant sensibility provides little insight. Other claims concerned norms that are contested even within the United States. Courts have been divided, for instance, on whether CIL prohibits cruel, inhuman, and degrading treatment;159 provides for corporate liability;160 and requires exhaustion of local remedies.161 On these issues, at the very least, the choice not to thoroughly examine state practice and opinio juris creates a risk of error. In some cases, courts might decline to recognize a norm that qualifies for the status of custom, while in others they might recognize a norm that fails to qualify.

This risk also has an intertemporal dimension. As the product of a constantly changing aggregation of what states do and say, CIL has a capacity to evolve organically, including within short periods of time.162 The result is that even where a federal court has correctly identified the applicable norm at a particular moment of adjudication, other courts’ reliance on that decision as precedent might, over the long run, underplay the possibility that the norm has since evolved, and that small distinctions in international and domestic application can accrete into major differences. This is the possibility of a normative Galapagos effect—an idiosyncratic local evolution decoupled from the broader international environment.

To be sure, many contend that there is more than one permissible approach to the identification of CIL. The so-called traditional method, and the one I have emphasized so far, operates by an inductive process that identifies custom on the basis of broad surveys of state practice over time.163 But a competing “modern” method operates through a process of deduction, whereby norms can be derived from official statements such as General Assembly resolutions and fundamental principles, rather than the practice of states.164 If federal courts are simply following this latter approach, then their tendency to ignore foreign state practice may still yield doctrinally justifiable conclusions.

The problem with this view is that it lacks support in the case law. To the extent that judges talked about methods, they repeatedly expressed a commitment to the inductive approach.165 This was true even in areas such as human rights,166 where the deductive method appears to be more influential in legal scholarship and international practice.167 Moreover, it is doubtful that even the modern approach justifies parochialism. It is one thing to question the importance of state practice, but quite another to say that the United States and its Western allies are the primary or even only relevant expositors of global norms.

The common method in the federal courts may also operate in tension with the Supreme Court’s Sosa decision, where the majority held that the ATS empowers federal courts to recognize common law causes of action only for violations of international norms that are “accepted by the civilized world and defined with [adequate] specificity.”168 Foreign relations was one of the principal justifications for this restraint: attempts to craft remedies for the violation of new norms “should be undertaken, if at all, with great caution,” due to the “risks of adverse foreign policy consequences.”169 Yet in looking primarily to domestic authority and hardly ever to foreign government sources, especially those of non-European states, federal courts often fail to ensure that the precise features of the norms they recognize are in fact accepted as CIL by the civilized world. Particularly in an increasingly multipolar order, the idea that “civilization” exists only in the West is plainly antiquated even as a purely descriptive matter.170 To identify norms on the basis of such a focus seems to risk precisely the foreign policy consequences that Sosa aims to avoid.

The tendency not to follow the traditional doctrine might also have other adverse effects abroad. Melissa Waters has argued that foreign courts are less likely to cite American precedents that have not attempted to engage seriously with foreign and international sources,171 and at least one Canadian Supreme Court Justice seems to agree.172 If other foreign judges share this view, then parochialism may very well decrease the global influence of U.S. precedent. Equally significant, the tendency to focus on U.S. sources might strain relations with affected states. Some governments have complained that ATS cases amount to judicial imperialism,173 but such objections might justifiably intensify where federal courts not only purport to decide the legality of foreign government acts, but also rely upon a heavily domesticated version of CIL in doing so.

Some might argue in response that a strong national orientation is precisely the point, especially in human rights cases. Courts and litigants, in other words, hold no illusion that all governments share American views about CIL, but purposefully characterize those views as international law anyway in order to propagate and strengthen the norms in question. I suspect that few would admit to this sort of rhetorical maneuver, but either way, its underlying assumption that American views lack sufficient global resonance to permit safe engagement with a broad array of foreign legal authorities seems to overcorrect for the opposite assumption that American views are universal. While more inclusive research and citation practices might jeopardize some norms by revealing a lack of common understanding, those practices might just as easily vindicate dominant domestic perceptions by clarifying the existence of broad global consensus, and in turn strengthen the legitimacy with which U.S. courts promote those norms that qualify.

At the same time, parochialism is not without potential benefits. For example, the common practice probably makes CIL more palatable in the domestic setting. Curtis Bradley and Jack Goldsmith have argued that the application of CIL as federal common law “is in tension with basic notions of American representative democracy” because custom is not “generated by U.S. lawmaking processes,” but instead “derived from the views and practices of the international community,” which is “‘neither representative of the American political community nor responsive to it.’”174 Whatever the merits of this position as a purely theoretical matter, the evidence discussed above takes some of the wind from its sails. The courts’ primary reliance upon U.S. government sources, American authors, and authorities from multilateral institutions within which the United States exercises significant influence means that CIL in federal court is far more an American creation than the traditional doctrine would have it. Moreover, the clear preponderance of Western European sources among the foreign citations shows that foreign influence is largely confined to states that more or less share basic American values. CIL in federal court is decidedly not dependent upon the practices and positions of ideologically hostile or culturally dissimilar states. Thus, if the domestic application of custom is undemocratic, it is so in much the same way that the application of garden-variety federal common law is undemocratic; the issue is not so much foreign influence as the unelected character of the judicial expositor.

B.     The Centrality of Courts

A second implication is one of confirmation. Although much of the evidence to date has been anecdotal and impressionistic, scholars have noted that municipal courts hold an increasingly important role in the development of CIL, not only because questions about custom have become more common in domestic litigation,175 but also because courts frequently look to foreign case law in deciding those questions.176 My findings are largely consistent with these observations. To identify the contours of contested norms, federal courts considered non-binding precedent from other federal courts more than any other source, foreign judicial decisions more than any foreign government acts beside treaty participation, and the decisions of international tribunals more than any other multilateral source. Moreover, the cited academic publications followed a similar pattern by citing to domestic, foreign, and international judicial opinions more than almost any other authorities.

As a matter of U.S. foreign relations law, this development stands in contrast to the federal judiciary’s traditional timidity in cases implicating foreign affairs.177 Congress and the executive branch, after all, were virtually absent from both the opinions and the cited academic sources even while citations to case law were ubiquitous. CIL is an area in which courts appear perfectly willing to resolve questions without input from the political branches, institutional limitations notwithstanding.178

 C.     The Importance of Access

Third, the data suggest that CIL as we understand it hinges, to a significant degree, not upon the general and consistent practice and opinio juris of states, but upon the vagaries of a fundamentally unprincipled system of access to information about practice and opinio juris. There is no single database that organizes the multifarious types of authority on which CIL may hinge. Westlaw and Lexis provide extensive resources on U.S. law, substantial databases of treaties and ICJ precedent, and even some foreign legal materials.179 International organizations such as the United Nations provide online access to official reports and resolutions.180 And many states provide searchable, online databases of their public laws.181 But not all of these resources are in English, finding them can require some ingenuity, and providers generally do not coordinate with one another. The result is a patchwork system with a subtle, if largely unintentional politics—states will possess power to influence foreign perceptions of CIL only insofar as their laws and policies are available to foreign legal audiences in influential languages.

The importance of access in turn raises questions for information providers. Why, for example, does the standard Westlaw account for American academics provide access to Korean but not Japanese cases? And what is the justification for offering a database on EU law but not the laws of China and Russia? One can guess at market- or translation-based explanations, among others. But it is fair to question whether these should suffice. Entities like Westlaw and Lexis exert enormous power over how the legal profession understands CIL simply by choosing whether to make available and how to organize foreign and international authorities. Surely it is no coincidence that the cases comprising the data set included far more citations to the laws of states that are available in the popular databases.

D.     Comparisons to Other Courts

Other implications are comparative. As discussed earlier, Choi and Gulati recently conducted a study to identify tendencies in the process by which international tribunals ascertain custom.182 Their primary conclusion is that these “courts do not come anywhere close to engaging in the type of analysis the officially stated two-part rule . . . sets up,” but instead “generally engage[] in a forward-looking or aspirational exercise” that discounts historical practice.183 According to their findings, treaties were the most commonly cited materials at nearly 33% of all sources in the data set, followed by international tribunal cases (16%), domestic cases (8%), actions by states (8%), academic articles and treatises (7%), reports from international interest groups such as the Red Cross and the International Law Association (7%), domestic statutes (7%), UN resolutions (5%), UN reports (5%), statements by state officials (3%), and party agreements (2%).184 On the basis of these numbers, Choi and Gulati conclude inter alia that international courts examine statements of national government officials only rarely, that international treaties are the most influential type of evidence, and that academic sources are somewhat unimportant.185

While the Choi and Gulati study covered a longer period of time and employed a slightly different coding strategy,186 their results suggest important points of similarity and contrast. On one hand, neither international nor U.S courts follow the traditional doctrine; judges tend to focus primarily on other types of evidence. But the types of evidence on which they focus differ in several ways. International courts, for example, cite more frequently to other international courts, while U.S. courts cite more frequently to other U.S. courts. International courts also cite more to multilateral treaties, but less to academic sources. And while Choi and Gulati’s study does not document trends in source nationality, it is probably safe to assume that international tribunals do not rely as heavily on U.S. legal authorities. Figure 31 illustrates these differences by showing, for each type of court, the percentage of the cited sources belonging to each source subcategory.187

The study results also make it possible to compare contemporary and earlier methods in the domestic context. Examining sources the Supreme Court utilized to identify CIL from 1861 to 1900, David Bederman found that while the Court’s rhetoric emphasized the importance of state practice, “direct evidence of the actual practice of states” was “notably absent” from the decisions.188 This sounds quite familiar. Yet there are important differences as well. Rather than federal judicial opinions, the “writing of international law scholars or ‘publicists’” was “far and away the most common” source in the earlier cases.189 And somewhat surprisingly, the academic sources may have been less parochial than they are today, even if comparably Western. Bederman reports that the Court showed “little favoritism” toward influential American commentators and “seemed comfortable citing French, Spanish, and Italian sources (through competent English versions), occasionally rendering their own translations.”190 This point calls into question the view that municipal courts’ application of CIL has become less parochial in recent years.191 And given that Western Europe represented a far larger portion of the recognized community of states in the post-Civil War period, it is not implausible that the earlier approach enabled the Court to ascertain custom with greater accuracy than the federal courts of today.


Figure 31. Comparison of Citation Patterns



E.     A New Approach to Custom?

While contemporary judicial methods are problematic under the traditional doctrine, the data also raise the possibility that the more fundamental problem is the doctrine itself. The evidence that courts simply do not follow the two-part test is increasingly abundant. As discussed above, Choi and Gulati found that international tribunals do not apply the test, and in a recent article Stefan Talmon reached a similar conclusion, observing that “in a majority of cases the [International Court of Justice] has not examined the practice and opinio juris of states but, instead, has simply asserted the rules that it applies.”192 David Bederman’s research suggests a similar tendency in U.S. Supreme Court cases,193 and as Jack Goldsmith and Eric Posner have argued, even The Paquette Habana, perhaps the Court’s most famous and influential decision on CIL, is riddled with evidentiary deficiencies.194 Combining this evidence with the data revealed above, it is worth considering: Of what value is a doctrine that no courts and few scholars appear to follow? Is there something to the traditional approach, or is it just a fig leaf for national and Western bias?

Some have responded to the evidence of methodological and other difficulties by arguing for a reconceptualization of custom. Curtis Bradley, for example, recently offered a “common law account,” whereby international courts develop CIL by looking to past practice while “necessarily mak[ing] choices about how to describe the practice, which baselines to apply in evaluating it, and whether and when to extend or analogize it to new situations.”195 Others have advocated rational choice models196 or doctrinal tweaks that place greater emphasis on either state practice or opinio juris.197

These are important proposals with deep and varied implications, and for their advocates, the data discussed above might bolster the argument for reforming or abandoning the traditional test. Yet there are at least a couple of reasons for caution. First, the two-part test has value insofar as it formally honors state consent, without geographic or cultural limitation. Any replacement that aspires for global acceptance would presumably need to share that characteristic, one result of which would be to hold parochialism in a state of continued illegitimacy. It is not clear, in other words, that contemporary methods would fare much better even under a new concept of CIL. Indeed, I am unaware of any proposal that aims to justify or apologize for an overwhelmingly Western focus. Even under Bradley’s proposal, which acknowledges substantial judicial discretion,198 courts would presumably still need to conduct at least moderately inclusive global surveys. Second, it may be too early to give up on the traditional doctrine. While courts have had a difficult time adhering to it, judges are not operating in a static technological or informational environment. Surely it is easier to research foreign law now than at any prior point in world history, given economic development, the expansion of international connectivity, the proliferation of legal research databases, and the increasing sophistication of automated translation. Many anticipate, moreover, that these developments will continue.199 It would be unfortunate if courts and scholars gave up on the longstanding doctrine just as technology rendered it practical. Indeed, one might argue that the problem with the traditional doctrine is simply its futurism; it presumes a level of access to foreign information that technology is only beginning to provide.

VI.     Recommendations

There is no easy way to find custom in accordance with the established doctrine, but courts, parties, and scholars can undertake a variety of measures to mitigate the challenge. In offering these, I operate on the assumption that the two-part test will continue to operate as the recognized doctrine in both international law and federal common law.

First, both sides of the bench can become familiar with and emulate the work of the most impressive judicial opinions to date. In the post-Sosa era, several decisions stand out as particularly well-researched, including Judge Leval’s concurrence in Kiobel v. Royal Dutch Petroleum Co.,200 Judge Katzmann’s concurrence in Khulumani v. Barclay National Bank Ltd.,201 Judge Bybee’s dissent in Sarei v. Rio Tinto, PLC,202 and Judge Davis’s decision in United States v. Hasan.203 These opinions may or may not reach the right conclusions on the merits, but they are noteworthy for the rigor with which their authors considered a large and diverse set of materials—the citation patterns were less nationalistic than most, included an array of multilateral and academic authorities, and even considered some foreign government sources.204 By reviewing these decisions, judges and litigants, particularly those who are unfamiliar with CIL, can develop a sense for the best practices of the federal judiciary.

Second, American legal scholars can follow the lead of some of their European counterparts and commit to study state practice and opinio juris in a more inclusive fashion.205 Courts need access to research that focuses not just on the familiar American, British, and other European sources that are already widely available, but also sources from non-Western states. This particular type of research is not en vogue in the United States, but academic writers are clearly in the best position to provide it. Unlike judges and litigants, they tend to have expertise in international or comparative law, and operate free from the docket pressures that severely limit the potential for exhaustive inquiry. Moreover, this kind of research is easier now than it has ever been. The Library of Congress, for example, has an extensive collection of links to underutilized government databases on the law of virtually every foreign state.206 Particularly given the increasing ease of comparative research, the tendency for current scholarship to ignore the numerical majority of international society seems hard to justify.

Third, courts can make use of Rule 44.1 of the Federal Rules of Civil Procedure in cases where the briefing fails to provide sufficient evidence of state practice and opinio juris. This rule provides that in determining foreign law, a court “may consider any relevant material or source, including testimony, whether or not submitted by a party or admissible under the Federal Rules of Evidence.”207 The rule’s standard application focuses on pure questions of foreign law,208 but CIL is derivative of foreign law insofar as the traditional doctrine holds that state practice can take the form of foreign judicial decisions, statutes, constitutions, regulations, and a wide range of other similar authorities.209 As the Restatement (Third) acknowledges, “questions of foreign law and of international law are often intermingled.”210 Under the proposed application, courts would not utilize Rule 44.1 to identify CIL per se, but rather to facilitate the predicate identification of foreign state practices that may or may not amount to CIL. This application would enable a court to “engage in its own research and consider any relevant material thus found.”211 Alternatively, where courts decide against this option, they might respond to deficient briefing not by concluding definitively that a norm is absent, but rather that the norm’s proponent has failed to provide sufficient supporting evidence.

Fourth, legal database companies and other information-providers should consider doing more to improve the quality and availability of their offerings. The ideal would be a single, searchable database that contains comprehensive and translated collections of the laws of all states. A slightly less helpful but, for now, more realistic alternative would allow users to search simultaneously the international law digests of all states that publish them. The development of these resources will make it easier for courts, litigants, and scholars to conduct inclusive analyses and in doing so strengthen the legitimacy of U.S. precedent abroad.

Finally, the data raise questions for additional research. For example, is parochialism equally common abroad? If U.S. courts are global outliers and other states adhere more closely to the traditional doctrine, then it will be that much harder to justify the methods that tend to appear in American cases. If, on the other hand, other courts exhibit similar patterns of national orientation, inattention to foreign government sources, limited multilateralism, and reliance upon academic publications that only rarely examine state practice and opinio juris, then we might fairly question whether the traditional doctrine has any bearing in reality.

Relatedly, do the courts of other states engage in similarly high rates of citation to academic sources, and if so, who are the cited authors? If foreign courts tend to focus on the potentially distinctive scholarship of their own nationals, then judicial decisions will be more likely to reach conflicting conclusions about contested norms and, in doing so, contribute to norm fragmentation. If, on the other hand, foreign courts also rely disproportionately on academics from the West, then the prospects for fragmentation may drop as courts around the world operate under the soft power of American and European scholarship.

Still another question is whether the pattern of multilateralism makes up for parochialism and delegation. If multilateral treaties, UN resolutions and reports, international judicial decisions, and other similar materials accurately reflect the practice and opinio juris of many states, then there will be far less reason to care about the problems associated with the other kinds of sources. But if the multilateral authorities fail to capture state practice or exaggerate the influence of the West, then they will do far less to improve the quality of the analysis.

VII.     Conclusion

International and federal common law have long held that CIL depends upon the existence of both general and consistent state practice and opinio juris. Yet citation patterns suggest that federal courts do not follow this doctrine. Courts depend heavily on portrayals of CIL in other U.S. government sources, rarely consider direct evidence of foreign state practice, focus almost exclusively on the advanced democracies of the West even when they do look abroad, and cite to Western academics who exhibit a similar tendency to focus on the laws and policies of the West. In short, federal courts are not applying customary international law in any strict doctrinal sense. The norms they have framed as CIL are domestic and Western norms that simply may or may not align with the practices and views of a clear majority of states. To a significant degree, this is probably a product of resource limitations over which courts have little control. Yet there are a number of steps that judges, litigants, and scholars can take to mitigate the current challenges. Judges and parties can acquaint themselves with best practices, scholars can expand the scope of their research to consider non-Western sources, courts can make use of Rule 44.1 to supplement deficient briefing, and information providers can expand their offerings in foreign law. Over time, such measures will help courts to adhere more closely to the established doctrine and enhance the legitimacy of the law that is international custom.

  1. [1]. 28 U.S.C. § 1350 (2012).

  2. [2]. Sosa v. Alvarez-Machain, 542 U.S. 692, 725 (2004).

  3. [3]. See, e.g., John Roe I v. Bridgestone Corp., 492 F. Supp. 2d 988 (S.D. Ind. 2006) (forced labor); Presbyterian Church of Sudan v. Talisman Energy, Inc., 453 F. Supp. 2d 633 (S.D.N.Y. 2006) (war crimes).

  4. [4]. 28 U.S.C. § 1604 (2012).

  5. [5]. Id. § 1605(a)(3).

  6. [6]. See, e.g., LaLoup v. United States, 29 F. Supp. 3d 530 (E.D. Pa. 2014); Agudas Chasidei Chabad of United States v. Russian Fed’n, 466 F. Supp. 2d 6 (D.D.C. 2006).

  7. [7]. 18 U.S.C. § 1651 (2012).

  8. [8]. See United States v. Ali, 982 F. Supp. 2d 85, 86 (D.D.C. 2013).

  9. [9]. See United States v. Hasan, 747 F. Supp. 2d 599, 603, 640–41 (E.D. Va. 2010).

  10. [10]. See Murray v. Schooner Charming Betsy, 6 U.S. (2 Cranch) 64, 118 (1804). See generally Curtis A. Bradley, The Charming Betsy Canon and Separation of Powers: Rethinking the Interpretive Role of International Law, 86 Geo. L.J. 479 (1998) (examining the canon’s origins and justifications).

  11. [11]. United States v. Bellaizac–Hurtado, 700 F.3d 1245, 1247 (11th Cir. 2012).

  12. [12]. U.S. Const. art. I, § 8, cl. 10.

  13. [13]. See Eugene Kontorovich, Discretion, Delegation, and Defining in the Constitution’s Law of Nations Clause, 106 Nw. U. L. Rev. 1675, 1682 (2012) (“The originalist evidence strongly supports the view that Congress can define only offenses that already exist in international law.”).

  14. [14]. Bellaizac–Hurtado, 700 F.3d at 1248­–49.

  15. [15]. Id. at 1258. For other examples, see Zivotofsky v. Kerry, 135 S. Ct. 2076, 2084 (2015) (relying in part on CIL to define the scope of the President’s recognition power), and Brown v. United States, 12 U.S. (8 Cranch) 110 (1814) (using the law of nations in interpreting the Captures Clause).

  16. [16]. Restatement (Third) of the Foreign Relations Law of the United States § 102 (Am. Law Inst. 1987); see also Bellaizac–Hurtado, 700 F.3d at 1252 (“We agree with our sister circuits that customary international law is determined by examining state practice and opinio juris....”).

  17. [17]. Andrew T. Guzman, Saving Customary International Law, 27 Mich. J. Int’l L. 115, 150 (2005); see also James Crawford, Brownlie’s Principles of Public International Law 24 (8th ed. 2012) (suggesting that CIL requires “substantial uniformity” in state practice).

  18. [18]. Military and Paramilitary Activities in and Against Nicaragua (Nicar. v. U.S.), Merits, 1986 I.C.J. 14, 98 ¶ 186 (June 27).

  19. [19]. Cf. Int’l Law Comm’n, Rep. of the Work of Its Sixty-Sixth Session, U.N. Doc. A/69/10, ¶ 138 n.829 (2014) (“Provided that [a] practice is sufficiently general and consistent, no particular duration is required.”).

  20. [20]. Restatement (Third) of the Foreign Relations Law of the United States § 102 cmt. b (Am. Law Inst. 1987) (explaining that state action can take the form of “diplomatic acts and instructions as well as public measures and other governmental acts and official statements of policy, whether they are unilateral or undertaken in cooperation with other states”).

  21. [21]. Id. § 102 cmt. c.

  22. [22]. Birgit Schlütter, Developments in Customary International Law 16–17 (2010).

  23. [23]. To be sure, many other aspects of CIL, such as the relative importance of state practice and opinio juris, are contested. See generally J. Patrick Kelly, The Twilight of Customary International Law, 40 Va. J. Int’l L. 449 (2000) (critiquing CIL as conceptually flawed and illegitimate).

  24. [24]. The evidence of this is anecdotal but seemingly widespread. See infra note 85 (providing examples).

  25. [25]. Cf. Matthew J. Wilson, Demystifying the Determination of Foreign Law in U.S. Courts: Opening the Door to a Greater Global Understanding, 46 Wake Forest L. Rev. 887, 897–98 (2011) (discussing limited research options as a problem for judges who must apply foreign law).

  26. [26]. See Andrea Bianchi, International Law and US Courts: The Myth of Lohengrin Revisited, 15 Eur. J. Int’l L. 751, 751 (2004).

  27. [27]. See André Nollkaemper, National Courts and the International Rule of Law 217–43 (2011).

  28. [28]. See David J. Bederman, Customary International Law in the Supreme Court, 1861–1900, in International Law in the U.S. Supreme Court: Continuity and Change 89–123 (David L. Sloss et al. eds., 2011); William S. Dodge, Customary International Law in the Supreme Court,
    , in id. at 353–79; Michael D. Ramsey, Customary International Law in the Supreme Court,
    , in id. at 225–56.

  29. [29]. See generally Bart M.J. Szewczyk, Customary International Law and Statutory Interpretation: An Empirical Analysis of Federal Court Decisions, 82 Geo. Wash. L. Rev. 1118 (2014).

  30. [30]. A. Mark Weisburd, American Judges and International Law, 36 Vand. J. Transnat’l L. 1475, 1494–1500 (2003) (examining the application of jus cogens norms in a collection of nineteen federal appellate decisions).

  31. [31]. See, e.g., Daniel Abebe, Not Just Doctrine: The True Motivation for Federal Incorporation and International Human Rights Litigation, 29 Mich. J. Int’l L. 1, 45 (2007); John F. Coyle, The Case for Writing International Law into the U.S. Code, 56 B.C. L. Rev. 433, 460­–62 (2015); Julian Ku & John Yoo, Beyond Formalism in Foreign Affairs: A Functional Approach to the Alien Tort Statute, 2004 Sup. Ct. Rev. 153, 181–89; Michael D. Ramsey, The Empirical Dilemma of International Law, 41 San Diego L. Rev. 1243, 1246–51 (2004); Phillip R. Trimble, A Revisionist View of Customary International Law, 33 UCLA L. Rev. 665, 680–84 (1986); Weisburd, supra note 30, at 1478–84.

  32. [32]. Sosa v. Alvarez-Machain, 542 U.S. 692 (2004).

  33. [33]. Id. at 725–31.

  34. [34]. See generally Stephen J. Choi & Mitu Gulati, Customary International Law: How Do Courts Do it?, in Custom’s Future: International Law in a Changing World 117 (Curtis A. Bradley ed., 2016).

  35. [35]. Id. at 131–46.

  36. [36]. Id. at 132.

  37. [37]. Id. at 133.

  38. [38]. See id.

  39. [39]. Choi and Gulati coined this usage. See id. at 127.

  40. [40]. Cf. Antony Anghie, Imperialism, Sovereignty and the Making of International Law (2005) (exploring the colonial origins and features of public international law).

  41. [41]. See Stephen J. Choi & G. Mitu Gulati, Ranking Judges According to Citation Bias (As a Means to Reduce Bias), 82 Notre Dame L. Rev. 1279, 1283–86 (2007) (summarizing this work).

  42. [42]. I used the following search terms in Westlaw’s “Federal Cases” database: [“customary international law” or “law of nations” or “opinio juris” or “jus cogens” or “international custom” or “customary norm” or “customary international norm” or “international custom” or “international norm” or “international practice”]. The search was last performed on July 15, 2015.

  43. [43]. I began with the goal of also scrutinizing how state courts find CIL, and performed an identical search in Westlaw’s “All State Cases” database. This search produced an initial list of 140 cases, but hardly any of them analyzed CIL. A clear majority contained nothing more than summary rejections of the argument that the death penalty violates international law. See, e.g., People v. Adams, 336 P.3d 1223, 1252 (Cal. 2014). Remarkably, only two state court opinions since Sosa—a period of more than a decade—contain even a brief discussion of custom. See Comm’r of Corr. v. Coleman, 38 A.3d 84, 108–12 (Conn. 2012) (concluding that force-feeding an inmate for reasons of medical necessity does not violate CIL); In re Guardianship of Mark C.H., 906 N.Y.S.2d 419, 433 n.49 (N.Y. Sup. Ct. 2010) (noting that the Vienna Convention on the Law of Treaties codifies pre-existing CIL). This shows that, as a practical matter, judicial efforts to ascertain CIL are essentially exclusive to the federal courts in contemporary litigation. Cf. generally Julian G. Ku, Customary International Law in State Courts, 42 Va. J. Int’l L. 265 (2001) (discussing the historical role of state courts in the development of CIL).

  44. [44]. Sosa v. Alvarez-Machain, 542 U.S. 692 (2004).

  45. [45]. Id. at 725.

  46. [46]. See, e.g., Abelesz v. Magyar Nemzeti Bank, 692 F.3d 661, 676 (7th Cir. 2012) (FSIA); Hamdan v. United States, 696 F.3d 1238, 1250 n.10 (D.C. Cir. 2012) (Military Commissions Act); United States v. Hasan, 747 F. Supp. 2d 599, 626–27 (E.D. Va. 2010) (piracy prosecution).

  47. [47]. See Anthea Roberts, Comparative International Law? The Role of National Courts in Creating and Enforcing International Law, 60 Int’l & Comp. L.Q. 57, 58 (2011) (“Internet and publishing developments are... making domestic judgments on international law matters more accessible.”).

  48. [48]. Cf. Bederman, supra note 28, at 109 (“[N]otably absent in the post–Civil War Supreme Court’s handling of customary international law is any direct evidence of the actual practice of states.”).

  49. [49]. The reduction from roughly 700 to only 120 was primarily a product of the over-inclusiveness of the initial search terms. For example, nearly 60 of the 700 were opinions that cited to precedent involving a company named International Custom Products. None of these opinions had anything to do with CIL, but they turned up in the search results because of the company’s name.

  50. [50]. Cf. David Zaring, The Use of Foreign Decisions by Federal Courts: An Empirical Analysis, 3 J. Empirical Legal Stud. 297, 310 (2006) (“Underinclusivity is... likely because judges are conservative citers, in that they may rely on, but not cite to, foreign authority that they find persuasive.”).

  51. [51]. See supra text accompanying note 41.

  52. [52]. See, e.g., Zaring, supra note 50, at 311 (counting each time an opinion cited a foreign court).

  53. [53]. Restatement (Third) of the Foreign Relations Law of the United States
    § 102(2) (Am. Law Inst. 1987).

  54. [54]. See id. § 102 cmt. b.

  55. [55]. Sosa v. Alvarez-Machain, 542 U.S. 692 (2004).

  56. [56]. I defined nationality as the country in which the author received his or her first law degree, which was typically a J.D. or an LLB. I did this on the assumption that the site of an academic’s legal education exerts substantial influence on his or her legal perspectives. See Ryan M. Scoville, International Law in National Schools, 92 Ind. L.J. (forthcoming 2016) (arguing that distinctive national approaches to international legal education socialize law graduates to view international law in different ways). In addition, I counted the nationality of co-authors and thus coded some sources as having multiple nationalities.

  57. [57]. I coded treaties as multilateral if they had more than five parties, and as the foreign state practice of each individual party if they had five or fewer. The line is somewhat arbitrary, but enabled me to credit individual states where their influence was most significant.

  58. [58]. The numbers reported throughout this Article operate on the premise that U.S. treaties are multilateral sources, given the large number of states that often join, but coding them as domestic sources would also be sensible in light of the Supremacy Clause. See Carlos Manuel Vázquez, Treaties as Law of the Land: The Supremacy Clause and the Judicial Enforcement of Treaties, 122 Harv. L. Rev. 599, 601–02 (2008) (“[T]he Supremacy Clause declares treaties to be the ‘[S]upreme Law of the Land,’ just as the Constitution and federal statutes are, and instructs judges to give them effect.”).

  59. [59]. See, e.g., Zaring, supra note 50, at 311 (mentioning potential reader disagreements with some of his coding decisions).

  60. [60]. See, e.g., United States v. Hijazi, 845 F. Supp. 2d 874, 883–85 (C.D. Ill. 2011) (discussing norms on extraterritorial prescriptive jurisdiction).

  61. [61]. See, e.g., Flomo v. Firestone Nat. Rubber Co., 643 F.3d 1013, 1021–24 (7th Cir. 2011) (concluding that CIL does not prohibit child labor); Sarei v. Rio Tinto PLC, 650 F. Supp. 2d 1004 (C.D. Cal. 2009) (rejecting an ATS claim based on environmental destruction).

  62. [62]. Cf. Cortelyou C. Kenney, Measuring Transnational Human Rights, 84 Fordham L. Rev. 1053, 1078–83 (2015) (finding that federal opinions “do not reflect a hostility toward or an unwillingness to engage international law qua international law” in the context of human rights litigation).

  63. [63]. Kiobel v. Royal Dutch Petroleum Co., 133 S. Ct. 1659, 1669 (2013).

  64. [64]. See Ernest A. Young, Universal Jurisdiction, the Alien Tort Statute, and Transnational Public-Law Litigation After Kiobel, 64 Duke L.J. 1023, 1062–65 (2015) (discussing how Kiobel “limit[ed] the ATS even more sharply”); see also Kenny, supra note 62, at 1109 (finding that, post-Kiobel, the presumption against extraterritoriality has been “by far the most prominent ground invoked in dismissing” claims under the Alien Tort Statute and the Torture Victim Protection Act).

  65. [65]. See, e.g., Chowdhury v. Worldtel Bangl. Holding, Ltd., 746 F.3d 42, 48–50 (2d Cir. 2014) (dismissing a claim for torture that allegedly occurred in Bangladesh); Mujica v. AirScan Inc., 771 F.3d 580, 591–96 (9th Cir. 2014) (dismissing claims based on conduct that allegedly occurred in Colombia).

  66. [66]. See infra Part IV.

  67. [67]. To access the dataset, see Ryan M. Scoville, Customary International Law in U.S. Courts (2015),

  68. [68]. Sosa v. Alvarez-Machain, 542 U.S. 692, 725–31 (2004).

  69. [69]. See, e.g., United States v. Ahmed, 94 F. Supp. 3d 394, 415 n.8 (E.D.N.Y. 2015) (citing one opinion from a federal district court for the proposition that the Article I Offenses Clause provides constitutional authority for Congress to proscribe material support for terrorism).

  70. [70]. See, e.g., United States v. Campbell, 798 F. Supp. 2d 293, 308–09 (D.D.C. 2011) (citing exclusively to non-binding precedents from other federal courts in identifying the contours of the protective principle under CIL); In re XE Servs. Alien Tort Litig., 665 F. Supp. 2d 569,
    582–85 (E.D. Va. 2009) (citing to non-binding precedent from other federal courts, U.S. treaties, a federal statute, and a report from the executive branch in concluding that war crimes do not require state action under CIL). One scholar has described this practice as “methodological provincialism.” See Patrick M. McFadden, Provincialism in United States Courts, 81 Cornell L. Rev. 4, 14–15 (1995).

  71. [71]. If one were to code ratified treaties as domestic sources rather than multilateral sources in light of the Supremacy Clause, the percentage of all citations falling into the category of domestic citations would rise to 55%, while the percentage falling into the category of multilateral citations would fall to 21%.

  72. [72]. See, e.g., Sosa, 542 U.S. at 762 (Breyer, J., concurring) (citing an Israeli judicial opinion); United States v. Beyle, 782 F.3d 159, 168 n.2 (4th Cir. 2015) (citing a Somalian presidential proclamation and submission to the Commission on the Limits of the Continental Shelf); Kiobel v. Royal Dutch Petroleum Co., 642 F.3d 268, 270 (2d Cir. 2011) (Jacobs, J., concurring) (citing statements from the South African government); Nuru v. Gonzales, 404 F.3d 1207, 1222–23 n.11 (9th Cir. 2005) (citing provisions in the Israeli penal code and the constitutions of Argentina, Brazil, Eritrea, Ethiopia, Iran, Japan, and Thailand); In re Chiquita Brands Int’l, Inc. Alien Tort Statute & S’holder Derivative Litig., 792 F. Supp. 2d 1301, 1319 & n.14–17 (S.D. Fla. 2011) (citing treaty declarations by Bahrain, Brazil, China, Egypt, Guatemala, Syria, Thailand, Venezuela, and Vietnam); WesternGeco L.L.C. v. Ion Geophysical Corp., 776 F. Supp. 2d 342, 368–70 (S.D. Tex. 2011) (citing an Australian patent statute and a South African judicial opinion); United States v. Hasan, 747 F. Supp. 2d 599, 618 (E.D. Va. 2010) (citing a Kenyan criminal statute and judicial opinion); Almog v. Arab Bank, PLC, 471 F. Supp. 2d 257, 282 n.28 (E.D.N.Y. 2007) (citing treaty declarations by Egypt, Jordan, Pakistan, and Syria); Presbyterian Church of Sudan v. Talisman Energy, Inc., 374 F. Supp. 2d 331, 336 (S.D.N.Y. 2005) (citing an opinion from the Federal Court of Australia).

  73. [73]. See, e.g., Abdullahi v. Pfizer, Inc., 562 F.3d 163, 200 (2d Cir. 2009) (Wesley, J., dissenting) (citing enactments of the Allied Control Council); Bowoto v. Chevron Corp., No. C 99-02506 SI, 2006 WL 2455752, at *8 (N.D. Cal. Aug. 22, 2006) (same).

  74. [74]. The percentage of the total volume of citations associated with each subcategory was as follows: federal judicial opinions (41%); academic articles, books, and treatises (19%); opinions from international courts (8%); U.S. treaties (7%); treaties not ratified by the United States (5%); foreign state practice (4%); sources from the executive branch (4%); resolutions of the U.N. Security Council (3%); U.S. statutes (2%); U.N. reports (2%); amicus briefs (1%); sources from the legislative branch (1%); resolutions of the U.N. General Assembly (<1%); sources from other international organizations (<1%); and NGO materials (<1%). As explained earlier, if one were to code ratified treaties as domestic sources rather than multilateral sources in light of the Supremacy Clause, the percentage of all citations falling into the category of domestic citations would rise to 55%, while the percentage falling into the category of multilateral citations would fall to 21%. See supra note 71.

  75. [75]. Cf. Szewczyk, supra note 29, at 1133 (reporting that, in statutory interpretation cases, “[t]he predominant source [of CIL] was federal case law, which was referenced in all instances but one and was the sole source in nearly forty percent of the cases”).

  76. [76]. See, e.g., Nuru, 404 F.3d at 1222 & n.11 (9th Cir. 2005) (concluding that “torture is illegal under the law of virtually every country in the world” and citing the laws of Argentina, Brazil, Eritrea, Ethiopia, Iran, Israel, Japan, Russia, and Thailand as support).

  77. [77]. See, e.g., Best Med. Belg., Inc. v. Kingdom of Belgium, 913 F. Supp. 2d 230, 239–40, 242 (E.D. Va. 2012) (citing a series of nonbinding federal precedents for the proposition that government expropriation of the property of its own nationals is not a violation of international law, without discussing the requirements of state practice and opinio juris).

  78. [78]. See United States v. Hasan, 747 F. Supp. 2d 599, 634–35 (E.D. Va. 2010) (concluding that the UN Convention on the Law of the Sea represents CIL and citing presidential statements as support); Bowoto v. Chevron Corp., 557 F. Supp. 2d 1080, 1091 (N.D. Cal. 2008) (citing the Torture Victim Protection Act’s legislative history for the proposition that torture violates CIL).

  79. [79]. Statute of the International Court of Justice, art. 38(1)(d), June 26, 1945, 59 Stat. 1055, T.S. No. 993.

  80. [80]. Id. In this sense, there is a tension between the international law doctrine of sources and the traditional tendency for federal courts to look to the political branches whenever possible in cases that implicate foreign affairs.

  81. [81]. Cf. Roberts, supra note 47, at 63 (“[N]ational court decisions are often given greater weight than the practice of a single State would suggest because they are treated as a subsidiary means of identifying international law rather than as State practice per se.”); see also Roger O’Keefe, Domestic Courts as Agents of Development of the International Law of Jurisdiction, 26 Leiden J. Int’l L. 541, 557 (2013) (making a similar point).

  82. [82]. Sarei v. Rio Tinto PLC, 650 F. Supp. 2d 1004, 1006–07, 1024–26 (C.D. Cal. 2009).

  83. [83]. Id. at 1025.

  84. [84]. See id. at 1024–25.

  85. [85]. Id. at 1024–26.

  86. [86]. Anecdotal evidence suggests that the briefing tends to focus on U.S. government sources. See, e.g., Brief of the Appellant Yimmi Bellaizac-Hurtado, United States v. Bellaizac-Hurtado, 700 F.3d 1245 (11th Cir. 2012) (No. 14049-HH) (citing no direct evidence of foreign state practice or opinio juris); Brief for the United States, United States v. Bellaizac-Hurtado, 700 F.3d 1245 (11th Cir. 2012) (No. 14049-HH) (same); Reply Brief of the Appellant Yimmi Bellaizac-Hurtado, United States v. Bellaizac-Hurtado, 700 F.3d 1245 (11th Cir. 2012) (No. 14049-HH) (same).

  87. [87]. Cf. Ralph G. Steinhardt, Kiobel and the Weakening of Precedent: A Long Walk for a Short Drink, 107 Am. J. Int’l L. 841, 845 (2013) (arguing that the Supreme Court has a “continuing, seemingly visceral resistance to treating modern international law in both treaty and customary form as law of the United States”).

  88. [88]. See Coyle, supra note 31, at 468 (reporting that no state bar exam tests international law and that the Federal Judicial Center’s training program for new federal judges does not cover international law); United States, PIL Map (2015), (reporting that only 4% of U.S. law schools require students to complete a course on international law).

  89. [89]. See Wilson, supra note 25, at 897 (discussing how “American legal education fails to systemically equip future judges and attorneys to conduct research on foreign law,” and how “judges and their law clerks may receive only limited exposure to international law or transnational legal matters during their law school studies”); see also Flomo v. Firestone Nat. Rubber Co., 643 F.3d 1013, 1015 (7th Cir. 2011) (“[T]he relevant evidence of customary international law is widely dispersed and generally unfamiliar to lawyers and judges.”).

  90. [90]. See Sosa v. Alvarez-Machain, 542 U.S. 692, 732–38 (2004). But see id. at 762 (Breyer, J., concurring) (citing a decision from the Supreme Court of Israel on the principal of universal jurisdiction).

  91. [91]. See generally Gary Marks & Norman Miller, Ten Years of Research on the False-Consensus Effect: An Empirical and Theoretical Review, 102 Psychol. Bull. 72 (1987) (discussing the accumulated evidence on the effect); Lee Ross et al., The “False Consensus Effect”: An Egocentric Bias in Social Perception and Attribution Processes, 13 J. Experimental Soc. Psychol. 279 (1977) (identifying the effect).

  92. [92]. Steven J. Sherman et al., The False Consensus Effect in Estimates of Smoking Prevalence: Underlying Mechanisms, 9 Personality & Soc. Psychol. Bull. 197, 200–05 (1983).

  93. [93]. See generally Thomas Gilovich et al., Causal Focus and Estimates of Consensus: An Examination of the False-Consensus Effect, 45 J. Personality & Soc. Psychol. 550 (1983).

  94. [94]. Marks & Miller, supra note 91, at 72–73.

  95. [95]. Id.

  96. [96]. Coyle, supra note 31, at 447–50 (arguing that this socialization influences how federal judges perceive international law).

  97. [97]. Cf. Harold Hongju Koh, Is International Law Really State Law?, 111 Harv. L. Rev. 1824, 1853–54 (1998) (“[I]nsofar as customary international law rules arise from traditional State practice, the United States has been, for most of this century, the world’s primary maker of and participant in this practice.”).

  98. [98]. As indicated earlier, these numbers will be slightly different if one codes on the premise that ratified treaties are domestic sources rather than multilateral sources. The former would rise to 55%, while the latter would fall to 21%. See supra note 71.

  99. [99]. See, e.g., Khulumani v. Barclay Nat’l Bank Ltd., 504 F.3d 254, 270 (2d Cir. 2007) (Katzmann, J., concurring).

  100. [100]. See, e.g., Doe v. Qi, 349 F. Supp. 2d 1258, 1308 (N.D. Cal. 2004).

  101. [101]. See, e.g., United States v. Al Bahlul, 820 F. Supp. 2d 1141 (U.S. Ct. Mil. Comm’n Rev. 2011).

  102. [102]. See Vienna Convention on the Law of Treaties art. 14, May 23, 1969, 1155 U.N.T.S. 331.

  103. [103]. See, e.g., Al-Quraishi v. Nakhla, 728 F. Supp. 2d 702, 756 (D. Md. 2010) (citing the African Charter on Human and Peoples’ Rights and the European Convention for the Protection of Human Rights and Fundamental Freedoms).

  104. [104]. Rome Statute of the International Criminal Court, July 17, 1998, 2187 U.N.T.S. 90. President Clinton signed the Statute in 2000. See William J. Clinton, Statement on the Rome Treaty on the International Criminal Court 4 (2000), President Bush, however, unsigned the treaty in 2002. See Letter from John R. Bolton, Under Sec’y of State for Arms Control & Int’l Sec., to Kofi Annan, U.N. Sec’y Gen. (May 6, 2002),

  105. [105]. Convention for the Protection of Human Rights and Fundamental Freedoms as Amended by Protocols No. 11 and No. 14, Nov. 4, 1950, 213 U.N.T.S. 221.

  106. [106]. See U.S. Const. art. II, § 2, cl. 2.

  107. [107]. This is also a practice for which there is historical precedent. See The Peterhoff, 72 U.S. 28, 50 (1866) (citing the 1856 Declaration of Paris Respecting Maritime Law as evidence of CIL); see also Bederman, supra note 28, at 105 (“[T]here appears to be a substantial basis for inferring from the Court’s post–Civil War jurisprudence that customary international law rules could be derived from treaty provisions, even those that the United States had not expressly adhered to.”).

  108. [108]. Cf. Curtis A. Bradley & Mitu Gulati, Withdrawing from International Custom, 120 Yale L.J. 202, 233–39 (2010) (discussing the nature and origins of the persistent objector doctrine).

  109. [109]. See North Sea Continental Shelf (W. Ger./Den.; W. Ger./Neth.), 1969 I.C.J. Rep. 3,
    ¶¶ 60–81 (Judgment of Feb. 20) (identifying the clarity and substance of the treaty text, the number of ratifications, the time since entry into force, and the uniformity of subsequent state practice as factors relevant to whether the treaty has created CIL). Only two cases cited this decision and discussed its significance. See Kiobel v. Royal Dutch Petroleum Co., 621 F.3d 111, 139 (2d Cir. 2010); United States v. Hasan, 747 F. Supp. 2d 599, 633 (E.D. Va. 2010); see also Igartúa-De La Rosa v. United States, 417 F.3d 145, 176 (1st Cir. 2005) (Torruella, J., dissenting) (citing the Restatement (Third), which in turn cited the ICJ’s decision).

  110. [110]. See, e.g., Hasan, 747 F. Supp. 2d at 634–35 (citing executive branch statements affirming that the UN Convention on the Law of the Sea codifies CIL in many respects).

  111. [111]. See Harold Hongju Koh, Why the President (Almost) Always Wins in Foreign Affairs: Lessons of the Iran-Contra Affair, 97 Yale L.J. 1255, 1305–17 (1988). But see generally Harlan Grant Cohen, Formalism and Distrust: Foreign Affairs Law in the Roberts Court, 83 Geo. Wash. L. Rev. 380 (2015) (arguing that the Roberts Court has become less deferential to the executive on matters of foreign relations law).

  112. [112]. See, e.g., Aldana v. Del Monte Fresh Produce, N.A., Inc., 452 F.3d 1284, 1285–86 (11th Cir. 2006) (Barkett, J., dissenting) (discussing various international agreements without reference to the views of the executive branch on their implications for CIL).

  113. [113]. This assumption is questionable. See Oona A. Hathaway, Do Human Rights Treaties Make a Difference?, 111 Yale L.J. 1935, 1978 (2002) (reporting evidence that noncompliance with human rights treaties is “rampant,” and that “countries with poor human rights ratings are sometimes more likely to have ratified the relevant treaties than are countries with better ratings”).

  114. [114]. See, e.g., Kiobel, 621 F.3d at 136.

  115. [115]. See, e.g., Doe v. Qi, 349 F. Supp. 2d 1258, 1321 (N.D. Cal. 2004).

  116. [116]. Approximately 11% of the cases cited one or more reports from a UN-affiliated organization, some of which were comprised of individuals rather than states.

  117. [117]. See, e.g., Khulumani v. Barclay Nat’l Bank Ltd., 504 F.3d 254, 274 (2d Cir. 2007) (Katzmann, J., concurring).

  118. [118]. See Isabelle R. Gunning, Modernizing Customary International Law: The Challenge of Human Rights, 31 Va. J. Int’l L. 211, 221–27 (1991) (arguing that international organizations and their agencies “should... hav[e] the ability to create custom” even when they do not represent states as members).

  119. [119]. Sir Michael Wood, International Organizations and Customary International Law, 48 Vand. J. Transnat’l L. 609, 618 (2015); see also Restatement (Third) of the Foreign Relations Law of the United States § 102(2) (Am. Law Inst. 1987) (“Customary international law results from a general and consistent practice of states followed by them from a sense of legal obligation.”) (emphasis added).

  120. [120]. See Stephen Townley, Counsellor for Legal Affairs, U.S. Mission to the United Nations, Remarks at the 69th General Assembly Sixth Committee on Agenda Item 78: Report of the International Law Commission on the Work of its 66th Session (Nov. 5, 2014) (transcript available at (“While we are not persuaded as yet that the practice of international organizations is of general relevance to the identification of custom, we are open to the possibility that there may be circumstances in which some activities of international organizations may contribute to the formation of customary international law.”).

  121. [121]. See, e.g., United Nations Treaty Collection, (last visited May 7, 2016).

  122. [122]. Sosa v. Alvarez-Machain, 542 U.S. 692, 732–38 (2004).

  123. [123]. See David L. Schwartz & Lee Petherbridge, The Use of Legal Scholarship by the Federal Courts of Appeals: An Empirical Study, 96 Cornell L. Rev. 1345, 1359 (2011) (finding that 7.6% of reported federal appellate decisions from 1950 to 2008 cited at least one law review article).

  124. [124]. Statute of the International Court of Justice, art. 38(1)(d), June 26, 1945, 59 Stat. 1055, T.S. No. 993.

  125. [125]. This figure includes citations to the Restatement (Third), which was written with the assistance of an international advisory panel but otherwise reflects the work of reporters and advisors who were mostly American. See Restatement (Third) of the Foreign Relations Law of the United States at V–VII (Am. Law Inst. 1987).

  126. [126]. For example, given my coding rules, I designated works by Ved Nanda as having an Indian author because he obtained his initial legal education at Delhi University, but he has spent more time in American legal academia than I have. See Faculty Profile: Ved P. Nanda, U. Denver Sturm C. L., (last visited May 7, 2016). Presumably his work also reflects the socializing effects of this professional experience.

  127. [127]. See generally Anthea Roberts, Is International Law International? (forthcoming 2016) (comparing international law casebooks and academics from Australia, China, France, Russia, the United Kingdom, and the United States); see also Joost Pauwelyn, Europe, America and the ‘Unity’ of International Law, in The Europeanisation of International Law: The Status of International Law in the EU and Its Member States 205–25 (Jan Wouters et al. eds., 2008) (discussing differences between American and European approaches to international law); Scoville, supra note 56, at Part I (documenting different national approaches to the study of international law). For an historical example of these differences and why they matter, see generally Isabel V. Hull, A Scrap of Paper: Breaking and Making International Law During the Great War (2014) (exploring how some of the key states in World War I approached and understood international humanitarian law in different ways, with significant consequences on the battlefield).

  128. [128]. Cf. Scoville, supra note 56, at Parts II–III (arguing that law schools are capable of socializing students to adopt views about international law that law faculties have chosen to privilege, and that cross-national variation in the content of privileged views might explain varying rates of state compliance with international norms). Influences other than socialization, such as the structure of the contemporary international order, might also play a role. See William W. Burke-White, Power Shifts in International Law: Structural Realignment and Substantive Pluralism, 56 Harv. Int’l L.J. 1, 25–47 (2015) (arguing that structural features of the current international system promote pluralism in international law).

  129. [129]. Cf. Roberts, supra note 47, at 76 (suggesting that the Restatement (Third) “represents a particular US perspective on international law”).

  130. [130]. See generally Flomo v. Firestone Nat. Rubber Co., 643 F.3d 1013 (7th Cir. 2011).

  131. [131]. Id. at 1017–21.

  132. [132]. Id. at 1017–18 (quoting Robert Cryer, International Criminal Law, in International Law 752, 770–71 (Malcolm D. Evans ed., 3d ed. 2010)).

  133. [133]. Id. at 1018 (citing John C. Coffee, Jr., “No Soul to Damn: No Body to Kick”: An Unscandalized Inquiry into the Problem of Corporate Punishment, 79 Mich. L. Rev. 386, 447–48 (1981)).

  134. [134]. See generally Schlütter, supra note 22 (surveying past and recent developments in the theory and practice of CIL).

  135. [135]. See Trust and Non-Self-Governing Territories (1945–1999), United Nations, (last visited May 7, 2016) (listing the states that have acquired independence since 1945).

  136. [136]. Cf. Yahya Sadowski, Ethnic Conflict, Foreign Pol’y, Summer 1998, at 12, 12 (“Ethnic conflicts have consistently formed the vast majority of wars ever since the epoch of decolonization began to sweep the developing countries after 1945.”).

  137. [137]. See Trust and Non-Self-Governing Territories, supra note 135.

  138. [138]. Expert Declaration of Professor Philip Alston, United Nations Special Rapporteur on Extrajudicial, Summary or Arbitrary Executions at 44, Wiwa v. Royal Dutch Petroleum Co., 626 F. Supp. 2d 377 (S.D.N.Y. 2009) (Nos. 96 Civ. 8386 (KMW) (HBP), 01 Civ. 1909 (KMW) (HBP)). This was by far the most thorough survey of state practice that I came across in the course of my research.

  139. [139]. Declaration of Stefan Talmon at 6, Almog v. Arab Bank, PLC, 471 F. Supp. 2d 257 (E.D.N.Y. 2007) (Nos. 04–CV–5564(NG)(VVP) & 05–CV–0388(NG)(VVP)). Interestingly, the district court brushed aside this evidence in concluding that there is a CIL prohibition on “terrorism” that meets the Sosa test. See Almog, 471 F. Supp. 2d at 270, 276–84.

  140. [140]. Appellate courts did not cite the vast majority of the amicus briefs that were filed. It is unclear whether there was a material difference between the cited and non-cited briefs.

  141. [141]. See Anthony Aust, Modern Treaty Law and Practice 324–26 (3d ed. 2013).

  142. [142]. See Jean-Marie Henckaerts & Louise Doswald-Beck, Customary International Humanitarian Law: Volume I: Rules 3–5 (2005).

  143. [143]. This practice appears to predate the period under review. See Kelly, supra note 23, at 472 (“The practices and attitudes of Japan, China, and the many nations of Africa, Asia, and Latin America are virtually ignored in the Western literature.”).

  144. [144]. See Alston, supra note 138.

  145. [145]. See Talmon, supra note 139.

  146. [146]. Citations for the other briefs are located in the “Amici” tab of the downloadable data set. See Scoville, supra note 67.

  147. [147]. See supra Figure 16.

  148. [148]. Cf. generally Martti Koskenniemi, From Apology to Utopia: The Structure of International Legal Argument (2005) (suggesting that a focus on state practice amounts to an “apology” for power).

  149. [149]. Statute of the International Court of Justice, art. 38(1)(d), June 26, 1945, 59 Stat. 1055, T.S. No. 993. The effect of this doctrine is to establish that the decisions of municipal courts, much like academic publications, can help one to identify CIL but cannot operate as a source of CIL. See Armin von Bogdandy & Ingo Venzke, Beyond Dispute: International Judicial Institutions as Lawmakers, 12 German L.J. 979, 992–93 (2011).

  150. [150]. Cf. Watkins v. United States, 354 U.S. 178, 187 (1957) (describing Congress’s expansive powers of investigation).

  151. [151]. Cf. Koh, supra note 97, at 1853–54 (“[I]nsofar as customary international law rules arise from traditional State practice, the United States has been, for most of this century, the world’s primary maker of and participant in this practice.”).

  152. [152]. Cf. David Strang & John W. Meyer, Institutional Conditions for Diffusion, 22 Theory & Soc’y 487, 491–92 (1993) (theorizing on how modernity contributes to institutional isomorphism on a global scale); see also Martha Finnemore, International Organizations as Teachers of Norms: The United Nations Educational, Scientific, and Cultural Organization and Science Policy, 47 Int’l Org. 565 (1993) (identifying international organizations as a source of isomorphism among national governments).

  153. [153]. The evidence for this is anecdotal. See, e.g., William v. AES Corp., 28 F. Supp. 3d 553, 566 (E.D. Va. 2014) (rejecting a claim that the provision of substandard electrical supply constitutes cruel, inhuman, or degrading treatment under CIL, based on citations to Bowoto v. Chevron Corp., 557 F. Supp. 2d 1080, 1092 (N.D. Cal. 2008); Doe v. Qi, 349 F. Supp. 2d 1258, 1325 (N.D. Cal. 2004); and Xuncax v. Gramajo, 886 F. Supp. 162, 187 (D. Mass. 1995)); Bowoto, 557 F. Supp. 2d at 1092 (concluding that the definition of torture under CIL does not require acts to be undertaken while the victim is in custody and pursuant to official state policy and citing a federal statute, a ratified treaty, and two non-binding federal precedents for support); Doe, 349 F. Supp. 2d at 1321–25 (citing non-binding federal precedent and an affidavit from international law scholars in concluding that the plaintiff had stated a claim for cruel, inhuman, or degrading treatment under CIL); Xuncax, 886 F. Supp. at 187 (citing a case from the European Court of Human Rights in concluding that certain acts constituted cruel, inhuman, or degrading treatment).

  154. [154]. See supra note 74.

  155. [155]. See, e.g., United States v. Bravo, 489 F.3d 1, 7–8 (1st Cir. 2007).

  156. [156]. See, e.g., Best Med. Belgium, Inc. v. Kingdom of Belgium, 913 F. Supp. 2d 230, 239 (E.D. Va. 2012).

  157. [157]. To be clear, this is not to say that CIL should not accord with predominant Western views, or even that it does not. The point is strictly epistemological—regardless of what one might hope CIL to be, few undertake the effort to know what it is in fact.

  158. [158]. See United States v. Ali, 982 F. Supp. 2d 85, 86 (D.D.C. 2013).

  159. [159]. Compare Doe v. Qi, 349 F. Supp. 2d 1258, 1296 (N.D. Cal. 2004) (concluding that CIL prohibits cruel, inhuman and degrading treatment), with Aldana v. Del Monte Fresh Produce, N.A., Inc., 416 F.3d 1242, 1247 (11th Cir. 2005) (reaching the opposite conclusion).

  160. [160]. Compare Flomo v. Firestone Nat. Rubber Co., 643 F.3d 1013, 1021 (7th Cir. 2011) (concluding that CIL provides for corporate liability), with Kiobel v. Royal Dutch Petroleum Co., 621 F.3d 111, 145 (2d Cir. 2010) (concluding otherwise).

  161. [161]. Compare Abelesz v. Magyar Nemzeti Bank, 692 F.3d 661, 679 (7th Cir. 2012) (holding that CIL requires exhaustion of local remedies), with Sarei v. Rio Tinto, PLC, 487 F.3d 1193, 1219–20 (9th Cir. 2007) (holding otherwise).

  162. [162]. See Bin Cheng, Custom: The Future of General State Practice in a Divided World, in The Structure and Process of International Law: Essays in Legal Philosophy Doctrine and Theory 513, 532 (R. St. J. MacDonald & Douglas M. Johnston eds., 1983) (proposing a concept of instant custom); Michael P. Scharf, Accelerated Formation of Customary International Law, 20 ILSA J. Int’l & Comp. L. 305, 329–40 (2014) (providing examples of rapid developments in CIL and characterizing them as “Grotian Moments”).

  163. [163]. Anthea Elizabeth Roberts, Traditional and Modern Approaches to Customary International Law: A Reconciliation, 95 Am. J. Int’l L. 757, 758 (2001).

  164. [164]. Id.

  165. [165]. See, e.g., Aziz v. Alcolac, Inc., 658 F.3d 388, 399 (4th Cir. 2011); Al-Bihani v. Obama, 619 F.3d 1, 17–18 (D.C. Cir. 2010); Kiobel, 621 F.3d at 131; Garcia v. Chapman, 911 F. Supp. 2d 1222, 1233 (S.D. Fla. 2012); United States v. Hasan, 747 F. Supp. 2d 599, 631 (E.D. Va. 2010).

  166. [166]. See Kiobel, 621 F.3d at 120; Garcia, 911 F. Supp. 2d at 1230.

  167. [167]. Roberts, supra note 163, at 759, 764–66 (citing scholars who argue that modern custom is an important source of human rights norms and offering a theoretical justification for reliance on the deductive process in that context).

  168. [168]. Sosa v. Alvarez-Machain, 542 U.S. 692, 725 (2004).

  169. [169]. Id. at 728.

  170. [170]. See generally Fareed Zakaria, The Post-American World (2008).

  171. [171]. Melissa A. Waters, Mediating Norms and Identity: The Role of Transnational Judicial Dialogue in Creating and Enforcing International Law, 93 Geo. L.J. 487, 557 (2005) (suggesting that the purported development of CIL through under-inclusive analyses of state practice and opinio juris “greatly decreases the likelihood that U.S. norms will influence the development of international legal norms on a given issue”).

  172. [172]. See Claire L’Heureux-Dubé, The Importance of Dialogue: Globalization and the International Impact of the Rehnquist Court, 34 Tulsa L.J. 15, 38 (1998) (“[T]he most useful judgments for courts looking to comparative sources are those that use comparative materials themselves, and situate their judgments in the context of international debates and discussions. Decisions which look only inward, which see only the situation in the place where they are rendered, have less relevance to those outside that jurisdiction....”).

  173. [173]. See Mark Sherman, Justices Weigh Foreigners’ Suits vs. Companies, New Haven Reg. (Feb. 26, 2012, 12:01 AM), (quoting South African President Thabo Mbeki as saying that an ATS suit against multinational companies that had been associated with apartheid amounted to “judicial imperialism”); see also Brief of the Governments of the Kingdom of the Netherlands and the United Kingdom of Great Britain and Northern Ireland as Amici Curiae in Support of Neither Party at 29, Kiobel v. Royal Dutch Petroleum Co., 133 S. Ct. 1659 (2013) (No. 10-1491) (arguing that the exercise of jurisdiction over ATS claims that lack a sufficient nexus to U.S. territory is “entirely inappropriate” and citing “ample evidence” that such jurisdiction generates international disputes); Brief of the Federal Republic of Germany as Amicus Curiae in Support of Respondents at 10, Kiobel, 133 S. Ct. 1659 (No. 10-1491) (characterizing the “unreasonable extraterritorial application of the ATS” as “unacceptable”); John B. Bellinger III, Enforcing Human Rights in U.S. Courts and Abroad: The Alien Tort Statute and Other Approaches, 42 Vand. J. Transnat’l L. 1, 2, 8 (2009) (reporting that “the ATS has given rise to friction, sometimes considerable, in our relations with foreign governments,” and that “we are regarded as something of a rogue actor” in the way we use the statute).

  174. [174]. Curtis A. Bradley & Jack L. Goldsmith, Customary International Law as Federal Common Law: A Critique of the Modern Position, 110 Harv. L. Rev. 815, 857 (1997) (quoting Phillip R. Trimble, A Revisionist View of Customary International Law, 33 UCLA L. Rev. 665, 721 (1986)).

  175. [175]. See Antonios Tzanakopoulos & Christian J. Tams, Introduction: Domestic Courts as Agents of Development of International Law, 26 Leiden J. Int’l L. 531, 533–34 (2013).

  176. [176]. See Roberts, supra note 47, at 58 (discussing these trends).

  177. [177]. See Koh, supra note 111, at 1305–17.

  178. [178]. See Waters, supra note 171, at 520 (“Courts participating in [transnational judicial dialogue] are becoming increasingly less deferential to executive branch prerogatives in foreign policy....”).

  179. [179]. The standard Westlaw account for U.S. academics offers searchable databases of cases from the courts of Australia, Canada, the European Union, Hong Kong, and Korea, plus treatises on the laws of a number of other jurisdictions. See International Materials, Thomson Reuters Westlaw, (last visited May 8, 2016).

  180. [180]. See Annual Reports, United Nations Security Council,
    documents/reports/ (last visited May 8, 2016).

  181. [181]. See infra note 205 (linking to the Library of Congress collection of foreign law databases).

  182. [182]. See supra notes 34–39 and accompanying text

  183. [183]. Choi & Gulati, supra note 34, at 146–47.

  184. [184]. Id. at 132. I use Choi and Gulati’s labels in listing these source categories and refer the reader to their paper for clarification on the definitions they have attached to them. See id. at 128–31.

  185. [185]. See id. at 131–46.

  186. [186]. The cases they considered ranged from recent opinions of the International Court of Justice to decisions issued by the Permanent Court of International Justice, which existed from 1922 to 1946. See id. at 126; Permanent Court of International Justice, Int’l Ct. Just., (last visited May 8, 2016). As for coding strategies, Choi and Gulati created separate categories for “domestic cases,” “actions by states,” “domestic statutes,” and “statements by state officials.” Choi & Gulati, supra note 34, at 132. In contrast, I created separate categories for different types of U.S. authority, but categorized all citations to foreign government sources as citations to “foreign state practice.” See supra text accompanying notes 56–59.

  187. [187]. This Figure leaves out subcategories of sources that do not have clear equivalents in the Choi and Gulati study. For example, the Figure omits my numbers on citations to federal judicial opinions, federal statutes, and executive branch sources because Choi and Gulati did not separately code for those authorities.

  188. [188]. Bederman, supra note 28, at 109.

  189. [189]. Id. at 105.

  190. [190]. Id. at 106.

  191. [191]. Cf. Yuval Shany, National Courts as International Actors: Jurisdictional Implications, (July 29, 2007),
    13810 (suggesting that a “certain quantitative and qualitative change is taking place in the manner international law is being applied by national courts: more international law is applied by more national courts in a more consequential (and less parochial) way”).

  192. [192]. Stefan Talmon, Determining Customary International Law: The ICJ’s Methodology Between Induction, Deduction and Assertion, 26 Eur. J. Int’l L. 417, 441 (2015).

  193. [193]. See generally Bederman, supra note 28.

  194. [194]. See Jack L. Goldsmith & Eric A. Posner, The Limits of International Law 66–78 (2005) (scrutinizing the evidentiary basis for the Court’s decision in The Paquette Habana, which purported to identify a rule of CIL that excluded enemy coastal fishing vessels from rights of capture).

  195. [195]. Curtis A. Bradley, Customary International Law Adjudication as Common Law Adjudication, in Curtis A. Bradley, Custom’s Future: International Law in a Changing World 34, 49 (2016). Although it would have to operate within the limits of Sosa and associated constitutional understandings, it is conceivable that this account might also apply to federal courts.

  196. [196]. See generally Guzman, supra note 17 (arguing for a rational choice lens for CIL).

  197. [197]. See generally Brian D. Lepard, Customary International Law: A New Theory With Practical Applications (2010) (arguing for a doctrine that focuses exclusively on opinio juris).

  198. [198]. See Choi & Gulati, supra note 34.

  199. [199]. See Nataly Kelly, Ray Kurzweil on Translation Technology, Huffington Post (June 13, 2011, 3:44 PM), (“Well-known inventor, author, and futurist Ray Kurzweil [predicts that computers] will reach human levels of translation quality by the year 2029.”); Digital Life in 2025: Experts Predict the Internet will Become “Like Electricity”—Less Visible, yet More Deeply Embedded in People’s Lives for Good and Ill, Pew Res. Ctr. 6, 28–29 (Mar. 11, 2014),
    Internet_Predictions_031114.pdf (surveying experts who anticipate that the Internet will evolve to feature even greater global connectivity and facilitate broader access to foreign information).

  200. [200]. See generally Kiobel v. Royal Dutch Petroleum Co., 621 F.3d 111, 149–82 (2d Cir. 2010) (Leval, J., concurring).

  201. [201]. See generally Khulumani v. Barclay Nat’l Bank Ltd., 504 F.3d 254, 264–84 (2d Cir. 2007) (Katzmann, J., concurring).

  202. [202]. See generally Sarei v. Rio Tinto, PLC, 487 F.3d 1193, 1224–46 (9th Cir. 2007) (Bybee, J., dissenting).

  203. [203]. See generally United States v. Hasan, 747 F. Supp. 2d 599 (E.D. Va. 2010).

  204. [204]. An excellent model from an international tribunal is the ICTY Appeals Chamber’s decision in Prosecutor v. Galic, Case No. IT-98-29-A, Judgment (Int’l Crim. Trib. for the Former Yugoslavia Nov. 30, 2006) (citing multiple treaties and dozens of foreign legal authorities from a diverse set of countries en route to concluding that the prohibition of terror against a civilian population has the status of customary international law).

  205. [205]. See, e.g., Alston, supra note 138 (surveying state practice in support of the view that CIL prohibits summary execution); Talmon, supra note 139 (collecting evidence of foreign state practice and opinio juris in arguing that the meaning of “terrorism” is unsettled in CIL); Henckaerts & Doswald-Beck, supra note 142, at 3–5 (surveying foreign state practices and statements in support of the proposition that CIL requires combatants to distinguish between themselves and civilians in conducting military operations).

  206. [206]. Nations, Libr. Congress, (last visited May 8, 2016). The databases are mostly available only in native languages, but online translation options are readily available and improving in quality.

  207. [207]. Fed. R. Civ. P. 44.1.

  208. [208]. See, e.g., Sunstar, Inc. v. Alberto-Culver Co., 586 F.3d 487, 495 (7th Cir. 2009) (applying 44.1 to decide a question of Japanese law, and explaining that the Rule “permits foreign law to be proved by testimony or affidavits of experts,” and “also permits judges to consult other sources of foreign law, such as articles, treatises, and judicial opinions”).

  209. [209]. See Restatement (Third) of the Foreign Relations Law of the United States
    § 102 reporters’ notes 1 (Am. Law Inst. 1987).

  210. [210]. Id. § 113 reporters’ notes 1.

  211. [211]. Fed. R. Civ. P. 44.1, Notes of Advisory Committee on Rules–1966.


Assistant Professor of Law, Marquette University Law School; J.D., Stanford Law School. 

Sincere thanks to Bruce Boyden, John Coyle, Jean Galbraith, Jack Goldsmith, Nadelle Grossman, Mitu Gulati, Vada Lindsey, Dave Moore, Michael O’Hear, Chad Oldfather, David Papke, Andrea Schneider, and the participants in the 2015 Inland Northwest Scholars’ Workshop for helpful comments on earlier drafts, and to Alex DeGuire and Ami Regele for excellent research assistance.