In U.S. federal courts, questions about the existence and contours of customary international law (CIL) arise in a variety of cases, both civil and criminal. These have generated a number of interesting debates, including, most prominently, one about the status of CIL in the United States: some contend that custom is self-executing federal common law, while others argue that it’s part of federal law only where Congress has chosen to codify it in a statute, or only where necessary to implement the separation of powers. Commentators on all sides of these debates tend to assume that the customary law that courts are applying is in fact CIL. But what if it’s not? Or, put another way, how do we know that it is?
The established doctrine is that custom arises from general and consistent state practice that is backed by a sense of legal obligation. For the most part, this has been understood to require broad surveys of foreign state practice, plus inquiries into official motives.
Federal courts purport to accept this doctrine, but they don’t follow it. This is no secret to scholars of U.S. foreign relations law, but most of the evidence to date has been anecdotal and impressionistic. As a supplement to prior writing on this issue, consider the following, which reflects an analysis of citation data collected from all reported federal judicial opinions that have analyzed CIL since the Supreme Court’s 2004 decision in Sosa v. Alvarez-Machain:
(1) In ascertaining custom, federal courts have cited U.S. government sources far more frequently than those of all foreign states combined. The map below shows the geographic distribution of those citations, with increasing percentages of the total number of references to official state sources moving from white (0%) to progressively deeper shades of red (100%). The ubiquitous white is no mistake: official U.S. government authorities such as judicial opinions and statutes made up nearly 90% of the total. There are faint hints of color in Western Europe and a couple other places, but that’s about it. Citations to the second most-cited state, the United Kingdom, made up only 3% of the total.
(2) With moderate frequency, federal courts have also relied on academic publications such as law review articles. Yet a clear majority of the cited authors (62%) are American. Citations to works by foreign authors were slightly more common than citations to foreign official sources, but the foreign authors were overwhelmingly from the West. British authors were the second most common (12%), followed by Germans (4%), Swiss (3%), and Italians (3%). This pattern likely reinforces the tendency for federal judges to perceive custom through an American and Western lens.
(3) The utility of judicial reliance on academic publications depends heavily on the quality of the analysis that is contained in the cited books and articles—works that include expansive surveys of state practice and opinio juris are far more likely to yield conclusions that are persuasive under the traditional doctrine than writing that does not. But it turns out that most of the scholarship that courts cited did not examine state practice or opinio juris in any sort of rigorous way and that, while more inclusive than the judicial opinions, the cited publications themselves cited official sources from the United States and other parts of the West far more often than other states. U.S. government sources made up over a third (34%) of all national government sources, with the UK (13%), France (7%), Germany (5%), and Belgium (4%) rounding out the top five. In the map below, notice Brazil, China, India, Japan, and Russia—a number of the largest states by population, more or less major players in the global economy and balance of power, and yet virtually irrelevant to the cited authors.
(4) Like courts, authors tend to reference and rely on other academics in making representations about the content of CIL. In fact, the cited authors relied on other authors more than any other source. Yet these other authors, too, were overwhelmingly from the West: the four states whose scholars were cited most in the cited academic literature were the United States (39%), United Kingdom (18%), Germany (4%), and Italy (4%).
Simply put, citation analysis shows that federal courts have not followed the established doctrine. Rather than conduct inclusive surveys of state practice, they’ve focused overwhelmingly on the United States and, to a lesser degree, other parts of the West. They’ve cited various academic sources, but most of those focus exclusively on the practices of Western states and reflect the views of Western academics who in turn cite most frequently to their Western colleagues. The rest of the world is almost entirely irrelevant.
To be clear, I’m fairly sympathetic toward the courts on this. It’s challenging for judges to research the laws of many foreign states, given language barriers, docket pressures, lackluster databases on foreign law, limited staff, the limited connectivity of much of the developing world, and a tendency for party briefing to focus only on readily available authorities. Moreover, courts do not rely exclusively on national government sources and academic publications; multilateral sources such as treaties and UN resolutions also play a non-trivial role. In terms of citation volume, these have been far less significant than domestic authorities, but their tendency to reflect broader state support probably helps to make up for the under-inclusiveness of the other types of citations.
There are a number of conceivable takeaways, the merits of which will depend in part on one’s preexisting views about a range of related issues, such as the doctrine of sources and the domestic status of CIL, among others. One is that we should understand the law that federal courts are applying for what it is: an almost exclusively Western and highly Americanized version of international law that pays very little attention to the practices and views of most states. Another is that courts and litigants need to do a better job researching CIL. This will be challenging, but may become easier over time with technological developments, and even now a handful of exceptional opinions have shown that courts can, with effort, consider a wider array of sources en route to conclusions about custom. For those who are skeptical that such effort will make a difference, another possibility is that Congress should use its comparatively robust investigative powers to ascertain custom on a wider variety of issues and then codify the results so that federal courts don’t have to research custom themselves. (See, for example, John Coyle’s recent proposal to that effect.) Still another possibility is that U.S. judicial practice reveals fundamental flaws in the traditional doctrine, and that alternative conceptions of custom are in order. A final possible takeaway, and I raise this one respectfully and with an intimate awareness of the limitations of my own work, is that scholars need to be more inclusive in their research on CIL. My impression is that, at least in the United States, it’s common for authors to make representations about custom on the basis of even just a handful of citations to American and British authority and, in the case of the ambitious, a sprinkling of additional references to French and German practice. This strikes me as insufficient. One can hardly expect other states (or scholars) to respect claims of custom that grant them no voice.