The Supreme Court Takes Up Sovereign Immunity from Criminal Prosecutions

File:Halkbank.jpg” by Vikiçizer

is licensed under CC BY-SA 3.0

On the first day of the October 2022 Term, the Supreme Court granted certiorari in Türkiye Halk Bankasi A.S. v. United States. The case, put simply, asks whether the U.S. government can bring criminal prosecutions against foreign companies owned by foreign sovereigns. The United States has charged Halkbank, in which Turkey’s sovereign wealth fund has a seventy-five percent interest, with sanctions evasion, bank fraud, and money laundering. Halkbank argues that the Foreign Sovereign Immunities Act (FSIA) bars this prosecution and all like it. The Second Circuit rejected that claim, and the Solicitor General told the Court that the case wasn’t certworthy. Nonetheless, a grant ensued.

Given the profusion of state-owned companies in the world economy, especially (though not only) in the financial sector, a ruling in favor of Halkbank would have enormous consequences. Chimène Keitner reviewed the issue on this blog shortly after Halkbank signaled its intent to seek the Court’s review. She recognized the importance of the question and defended U.S. criminal jurisdiction against the FSIA argument. At the time, very few people (by which I mean, not me) thought the Court would take the bait. Yet it has, leaving us to puzzle out why and what may happen next.

The Appeal of Literalism

The “why” question rests on speculation, but a ready answer is at hand. Section 1604 of Title 28, the heart of the FSIA, provides in full: “Subject to existing international agreements to which the United States is a party at the time of enactment of this Act a foreign state shall be immune from the jurisdiction of the courts of the United States and of the States except as provided in sections 1605 to 1607 of this chapter.” If “the jurisdiction” comprises criminal jurisdiction, then the immunity claim becomes a slam dunk. The exceptions to immunity in sections 1605 to 1607 seem to apply only to civil suits. Section 1605(a)(2), the commercial activity exception most relevant to prosecutions for financial crimes, bestows the exception on an “action,” a term that encompasses civil litigation and not criminal cases. In addition, Section 1330 of Title 28, the provision that establishes district court jurisdiction over cases against foreign states that fall within one of the exceptions to immunity, allows only civil suits. Finally, the definitions provision of the FSIA makes clear that a separate legal person, such as a corporation, majority owned by a state and not incorporated in the United States, counts as a foreign state for the Act’s purposes. (More on this last point, however, below.) A court inclined to literalism (what it might call “plain meaning”) would stop there: Jurisdiction include criminal prosecutions, Halkbank falls within the class of persons on which immunity is bestowed, and no statutory exception to immunity applies.

For what it’s worth, the Restatement (Fourth) of the Foreign Relations Law of the United States, in Reporters’ Note 4 to Section 451, recognized the issue and treated it as an open question. That tome, in which I had a hand, observes that a handful of lower courts had concluded (incorrectly as the Supreme Court later held in RJR Nabisco, Inc. v. European Community) that criminal and civil jurisdiction under the Racketeering Influence and Corrupt Organizations Act (RICO) are co-extensive, and that civil RICO suits against foreign state-owned firms are barred because (by hypothesis) the FSIA bars criminal cases. The Restatement further notes, however, that no court has ever invoked the FSIA in an actual criminal case, and that such cases do exist.

Is literalism called for here? Chimène and I conclude that it is not, but we get there by somewhat different paths. We agree that no one in 1976 thought that the FSIA was meant to address criminal proceedings. Section 1330 suggests an exclusive focus on civil suits, not a purposeful choice to bar criminal prosecutions. Not only the exceptions themselves, but the special procedural rules provided by Section 1608, indicate that the Act was thought to address only civil, and not criminal, jurisdiction. Finally, Title 18, where most of federal criminal law can be found, gives the district courts original jurisdiction of “all offenses against the laws of the United States.” The same literalist seeker of plain meaning that treats “the jurisdiction” as comprising criminal might also understand “all” as meaning all.

A possible rejoinder is that the provisions of Title 28 regarding the subject-matter jurisdiction of the federal district courts in civil cases also contain the modifier “all,” yet the authors of the FSIA found it necessary to add Section 1330. Of course, this provision does extra work. It not only creates subject-matter jurisdiction in the district courts, even in instances where the other subject-matter provisions of Title 28 do not apply, but also deems in personam jurisdiction to exist. Still, the absence of a counterpart provision in Title 18 might imply that no criminal cases were permitted. The authors of the FSIA may not have had criminal jurisdiction in mind, but their failure to expressly address the issue leads, the argument would go, to complete immunity in criminal cases.

How to Interpret “Jurisdiction”

The key question is whether “the jurisdiction” of U.S. courts, as used in Section 1604, means only civil jurisdiction or all possible jurisdiction. The literalist would say that Congress could have inserted “civil” into the provision, and its failure to do so speaks volumes. The comprehensive language in Title 18 would be dismissed as tautological: If immunity exists, there is no offense against the laws of the United States. The structuralist would respond that everything about the FSIA, not just its legislative history, indicates a purpose of regulating only civil jurisdiction.

An important hint comes from Republic of Argentina v. NML Capital, Ltd. That case addresses the power of a federal court to order a foreign state to submit to a discovery order. Justice Scalia, writing for a seven-justice majority, argued that the FSIA confers two kinds of immunity, namely from jurisdiction to decide a case and from jurisdiction to attach and execute property. Because the FSIA imposes no other limitation on judicial power, Scalia argued, it does not limit the jurisdiction involved in seeking discovery against a third party. The Court apparently believed that the word “jurisdiction” did not mean all exercises of judicial power.

The Irrelevance of the FSIA’s Exceptions

Some argue that if the FSIA applies to criminal cases, its exceptions should also apply. In practice, the United States prosecutes only business-related corporate crimes with a nexus to the United States, as the Halkbank case exemplifies. These prosecutions fit comfortably with the FSIA’s “commercial activity” exception to immunity, even if a prosecution is not an “action.” The problem with that move, however, is that the Supreme Court clearly rejected the exception-by-analogy argument in Samantar v. Yousuf. There the Court read the FSIA literally (as well as structurally) to conclude that the statute does not regard foreign officials as foreign states. Accordingly, a foreign official cannot claim the immunity given to states, and civil plaintiffs cannot invoke the exceptions to immunity to get federal court subject-matter and in personam jurisdiction as well as the FSIA’s special rules for service of process.

Yet several courts of appeals have taken the exemption-by-analogy approach. Indeed, the Second Circuit did this, citing the Fourth Restatement’s description of the commercial-activity exception as support for denying immunity to Halkbank. In its opposition brief, the government makes this argument as an alternative to the position that the FSIA does not apply to criminal cases. I read Samantar, however, as closing that door.

If the FSIA Does Not Apply, What Does?

This doesn’t mean, however, that Samantar is irrelevant to the criminal context. It observed that, even though the FSIA does not apply to foreign officials, cases against these persons are still governed by non-statutory law that could accord immunity. The Court left open where to look for that immunity – the law that existed at the time of adoption of the FSIA, which remains in effect because the FSIA did not displace it, or something else, such as customary international law. With a nod to Samantar, I would argue that a fair reading of the FSIA would treat it as governing only civil suits. If the FSIA provides no immunity in criminal cases, then any immunity that does exist must rest on some other body of law.

What might be an alternative source of corporate immunity from criminal liability? Pre-FSIA law presents few if any examples of prosecutions of state-owned firms, so we cannot state with great confidence what that law allowed. Still, there is no evidence of which I am aware that the United States considered itself unable to prosecute state-owned firms. The Trading With the Enemy Act, adopted in 1917 and in force when the FSIA was passed in 1976 as well as today, contemplates non-civil judicial proceedings against state-owned firms, including forfeiture of their assets. The Court might conclude that pre-1976 law extended the general discretion of prosecutors to choose when and where to enforce the criminal laws of the United States to crimes by state-owned entities, and that this remains the rule today.

Some people, including Ingrid (Wuerth) Brunk, have argued that unlimited executive discretion to control judicial outcomes has become obsolete and should not be read into any immunities that are adjacent to, but not covered by, the FSIA. I have some sympathy for the argument with respect to civil suits, where the government does not exercise any other control over the proceedings. But in criminal cases, prosecutorial discretion is at the core of the matter. There executive discretion is not an anomaly, but integral.

Another possibility is that the authority bestowed on U.S. prosecutors through Title 18 should be read in light of international law. A U.S. canon of interpretation instructs courts, when faced with statutory ambiguity, to choose the interpretation that saves the United States from violating international law. But what does international law have to say about criminal prosecutions of state-owned firms?

 There is, to my knowledge, no practice of criminal prosecution in domestic courts of states themselves. The closest analogy are sanctions, such as confiscations of property. In the United States, these measures are governed mostly by the Trading With the Enemy Act, which applies only to countries against which the United States has declared war, and, since adoption of the Patriot Act in 2001, the International Emergency Economic Powers Act, but only in cases of armed attacks in the absence of a declaration of war.

There is a common view that international law forbids criminal prosecution of foreign sovereigns in domestic courts, mainly because we don’t see any attempts to do this. Such prosecutions seem inconsistent with the general principle of sovereign equality, which objects to subordination of one sovereign to another. International law, however, has almost nothing to say about prosecution of state-owned firms. In much of the world only real or “physical” people are subject to criminal sanctions, which makes it difficult to detect a consistent practice one way or the other.

Countries like the United States, which recognizes the concept of corporate criminal liability, over the last few decades have consistently prosecuted state-owned corporations. Other states sometimes have argued that these cases involved extraterritorial application of domestic law that goes beyond what the international law of prescriptive jurisdiction permits. I do not know, however, of any instance where a foreign state asserted international-law sovereign immunity as a defense against a U.S. criminal prosecution of a company it owned. For what it’s worth, Canada and the United Kingdom, which also recognize corporate criminal liability, exclude state-owned corporations from their immunity statutes altogether, excepting in the case of the United Kingdom when a firm exercises sovereign functions. Halkbank does not assert an immunity based on international law. Rather, it argues only that the FSIA bars prosecution.

Strong claims about international law often come back to haunt those who make them. It seems to me, however, that there is no clear and consistent practice by states that indicates the existence of a binding legal norm forbidding the criminal prosecution of state-owned businesses. If I am right, then the canon of statutory construction that strives to avoid violation of international law poses no impediment to prosecutions of state-owned firms.

Halkbank’s briefs in support of its petition, as well as an amici brief filed on behalf of Azerbaijan and Pakistan, characterize the case not as a prosecution of a state-owned firm, but of the sovereign itself. This proceeding is, its reply brief asserts, “our nation’s first criminal trial of a foreign sovereign.” Such rhetoric, inviting us to imagine Erdogan in the dock, obscures what the case is about, a prosecution of a foreign corporation doing business in the United States that happens to be owned by a foreign state. International law treats the distinction between a state and its property as significant, indeed conclusive. Moreover, the brief’s rhetoric distracts from a critical point: If the government had chosen to enforce exactly the same laws by way of a civil proceeding, as federal law permits, the FSIA clearly would allow the suit to go to trial under the commercial activity exception, and nothing in international law would stand in the way.

The Direct Ownership Wrinkle

A technical wrinkle further complicates Halkbank’s claim of immunity. Its brief states that it is seventy-five-percent owned by the Turkish Wealth Fund, which it describes as “part of and owned by the Turkish state.” Does “owned by” mean that the Wealth Fund has a separate legal personality from the state? The issue is critical, because Dole Food Co. v. Patrickson held that the FSIA does not apply to second-tier subsidiaries, that is corporations owned in part by state-owned corporations. The government apparently never raised this point in the lower courts, or if it did the courts did not address it. As the prevailing party, however, the government is free to raise new arguments before the Supreme Court, if only to argue for a remand so that the lower courts may make appropriate findings of fact.


Not only does existing law point against immunity, but the consequences of ruling otherwise seem astounding. State-owned firms play a major role in the financial industry. China’s banks are one significant example. Pierre Verdier’s important book, Global Banks on Trial, documents how ambitious U.S. prosecutions exactly like that against Halkbank have brought much needed discipline to international finance. This would not work if state-owned banks enjoyed a free pass. The U.S. government has also successfully prosecuted state-owned firms for violations of the Foreign Corrupt Practices Act. Should we instead let state-owned firms that tap into U.S. capital markets build their businesses through bribery? Also consider safety and environmental enforcement, administrative as well as criminal, against state-owned carriers, whether airplanes or ships. And what about sanctions regimes, now urgent in the case of Russian banks and other state-owned firms such as Gazprom and Rosneft? Do we really need a new act of Congress before the United States can resume law enforcement, including the imposition of sanctions, against state entities?

The grant of certiorari in Halkbank is surprising as well as interesting. A reversal and the complete immunization of foreign state-owned firms from U.S. laws would be an earthquake.