In the world of Internet policy, it is a slur to call something an assertion of extraterritorial jurisdiction. Coverage of, for example, Canada’s recent ruling against Google in the Equustek case suggested that it was somehow a violation of fundamental principles of international law for a state to attempt to regulate Internet activity beyond its borders. (It isn’t.)
In the last few decades, the Supreme Court has had occasion to consider the territorial limits of a number of American laws – from Aramco (Title VII of the Civil Rights Act), Hartford Fire (the Sherman Act), Empagran (the Sherman Act), Kiobel (the Alien Tort Statute), RJR Nabisco v. European Community (RICO), Morrison (the Securities and Exchange Act), and more. If extraterritoriality cases make up a larger portion of the Court’s recent docket, it is not hard to imagine why. Many statutes do not specify a territorial scope, yet much of the activity that Congress hopes to regulate has an international dimension – either because it originates abroad, or its effects are felt abroad, or the parties are foreign, and so on. This is especially true in a globalized world. So it was only a matter of time before the Court heard a case about the extraterritorial reach of statutes that affect the Internet, which is now thoroughly global.
The analysis of extraterritoriality issues, as the Court clarified in RJR Nabisco v. European Community, typically involves at least two questions:
- At the first step, we ask whether the presumption against extraterritoriality has been rebutted—that is, whether the statute gives a clear, affirmative indication that it applies extraterritorially …
- If the statute is not extraterritorial, then at the second step we determine whether the case involves a domestic application of the statute, and we do this by looking to the statute’s “focus.”
In the present case, that means asking: (1) Does the Stored Communications Act apply extraterritorially?, and (2) does a warrant issued under that act operate extraterritorially if it compels a US provider to access its Irish data center? We might imagine four possible outcomes:
|Does the statute apply extraterritorially?|
|Is this an extraterritorial application of the statute?||Yes||DOJ wins||MSFT wins|
|No||DOJ wins||DOJ wins|
Step One: Does the Stored Communications Act Apply Extraterritorially?
The Stored Communications Act simply does not say much about where it is meant to apply. Without some indication from Congress that the statute was meant to apply abroad, the presumption against extraterritoriality would normally prevent it from being applied abroad. This is Microsoft’s argument, which is both straightforward and compelling.
But we might wonder whether the presumption against extraterritoriality makes any sense in this context. Scholars have argued for some time now that the presumption makes less sense in a globalized world; if Congress intends to regulate an inherently global thing, surely it makes at least as much sense that the regulation would have some extraterritorial reach as not. And if that holds true for any topic, it should be true for the regulation of Internet companies whose business models are often premised on connecting users across the world. The motivation for the presumption against extraterritoriality is, in part, an attempt to divine Congressional intent. And in areas like antitrust, courts have assumed, despite the presumption to the contrary, that Congress must have intended to regulate foreign conduct that has an American effect. Perhaps we might draw the same conclusion here: if Congress passed a bill regarding the Internet, without specifying the territorial reach of the statute, why would we assume that Congress intended for its law only to reach American data-centers? We are guessing in either direction –Congress intended for the statute to apply territorially or extraterritorially – and in this particular case it is not obvious that we ought to prefer one guess over another.
Additionally, courts have not relied on the presumption against extraterritoriality perfectly consistently, and where they have used it, they do so in ways that are consistent with prior case law. In Hartford Fire, for example, the Court held that the Sherman Act applied to extraterritorial conduct that had substantial effects in the U.S., even though the Sherman Act says nothing about its territorial reach. One would have thought that the presumption against extraterritoriality would have kicked in – just as it did in American Banana in 1909 – to prevent the Court from applying the Sherman Act to foreign conduct. But Justice Souter, writing for the Court in Hartford Fire, noted:
[I]t is well established by now that the Sherman Act applies to foreign conduct that was meant to produce and did in fact produce some substantial effect in the United States.
How did it get to be “well established”? By a series of decisions from courts that declined to apply the presumption against extraterritoriality.
The Court might be especially inclined not to apply the presumption against extraterritoriality in light of a considerable body of case law – from Bank of Nova Scotia to Marc Rich — suggesting that it is now “well established” that production orders can compel domestic actors to produce evidence they store abroad as long as they have control and custody of that evidence.
This is perhaps unlikely, given how recently and how forcefully the Court insisted on the presumption against extraterritoriality in RJR Nabisco only a few years ago, but it would not be unprecedented.
Step Two: Is This an Extraterritorial Application of the Statute?
The easiest way to resolve this case and produce a sensible policy result is to treat this warrant as operating domestically, more or less mooting the question of whether the SCA applies extraterritorially. The Court’s second-step in its extraterritoriality analysis is to determine the statute’s “focus.” That is a bit of a mystery here, since the SCA has more than one focus: law enforcement access; privacy for users; clarity for providers; among others. But it would be entirely reasonable—and prudent—for the Court to conclude that the focus of the statute is the relationship between U.S. law enforcement and service providers in the U.S., which would suggest that this warrant operates domestically.
Indeed, the bulk of the equities in this case are American: American authorities are investigating a violation of American laws, they have served a production order on an American company in the U.S., and the company can comply with the production order without leaving the U.S. The contested conduct in this case is a search, conducted by American authorities on American soil, of an email account held by an American firm. So it would make good sense for the Court to conclude that this warrant does not constitute an extraterritorial application of the SCA.
One nice thing about this conclusion is that it leaves providers with the option to store their data wherever they want. The SCA governs the production of the warrant which is only good within the United States and which compels Microsoft to comply in the United States, but where Microsoft chooses to store their data—the U.S., Ireland, the moon—has no bearing on the outcome on the case.
What About Ireland’s Sovereignty?
But what about the fact that Ireland might be upset about this “domestic” use of an SCA warrant, which asks Microsoft to dip into the servers they set up in Ireland? Independent of the presumption against extraterritoriality, we might still be worried about interference with another nation’s sovereignty. That was the Court’s concern in Empagran, a case about the extraterritorial reach of the Sherman Act. As the Court noted in RJR Nabisco:
Empagran concerned not the presumption against extraterritoriality per se, but the related rule that we construe statutes to avoid unreasonable interference with other nations’ sovereign authority where possible.
Well, readers, good news: courts have managed foreign affairs concerns for many years and they have a suite of comity doctrines at their disposal to manage these issues. Ireland’s sovereignty interest is especially weak in this case, but were it to be more pronounced, the Court could decide, independent of the statute’s extraterritorial reach, to refrain from enforcing an action that might unnecessarily upset foreign relations. (And for readers interested in this topic generally, I’ve got an article for you: Litigating Data Sovereignty.