Territoriality in American Criminal Law

It is a bedrock principle of American criminal law that the authority to try and punish someone for a crime arises from the crime’s connection to a particular place. Thus, we assume that a person who commits a crime in some location—say, Philadelphia—can be arrested by Philadelphia police for conduct deemed criminal by the Pennsylvania legislature, prosecuted in a Philadelphia court, and punished in a Pennsylvania prison. The idea that criminal law is tied to geography in this way is called the territoriality principle. This idea is so familiar that it usually goes unstated.

This Article foregrounds and questions the territoriality principle. Drawing on a broad and eclectic set of sources, it argues that domestic criminal law is less territorial than conventional wisdom holds. Although the territoriality principle is central to criminal law ideology, territorialism is a norm in decline. In reality, over the past century, new doctrines and enforcement practices have unmoored criminal law from geographic boundaries. The result is a criminal legal system in which borders are negotiable and honored in the breach.

Scholars have largely overlooked the deterritorialization of domestic criminal law, but the decline of the territoriality principle has striking implications. It undermines constitutional doctrines and academic theories built on the classic account of criminal law. It upsets foundational conceptual distinctions that structure public law. And it raises normative questions about just how far criminal laws should reach. This Article grapples with those questions and argues that borders are an underenforced constraint on the police power.


Introduction

Twenty years ago, Dustin Higgs was sentenced to death in a federal court in Maryland.1Higgs v. United States, 711 F. Supp. 2d 479, 493 (D. Md. 2010). Federal law required Higgs to be executed “in the manner prescribed” by the state where he was tried.2United States v. Higgs, 141 S. Ct. 645, 647–48 (2021) (mem.) (Sotomayor, J., dissenting) (quoting 18 U.S.C. § 3596(a)). But while Higgs was on death row, Maryland abolished the death penalty. So prosecutors asked the court to designate Indiana, a state that permits capital punishment, as the site of Higgs’s execution.3Id. at 648; see also David Cole, A Rush to Execute, N.Y. Rev. Books (Feb. 25, 2021), https://www.nybooks.com/articles/2021/02/25/trump-supreme-court-execution-spree [perma.cc/EXD2-TXYZ]. The trial court refused, but the Supreme Court intervened.4See Higgs, 141 S. Ct. at 645 (lifting the stay on Higgs’s execution). The United States government put Dustin Higgs to death during an unprecedented “spree of executions,” four days before President Trump left office.5Id. at 647 (Sotomayor, J., dissenting) (“[After Higgs’s execution,] the Federal Government will have executed more than three times as many people in the last six months than it had in the previous six decades.”); see also Hailey Fuchs, U.S. Executes Dustin Higgs for Role in 3 1996 Murders, N.Y. Times (Jan. 16, 2021), https://www.nytimes.com/2021/01/16/us/politics/dustin-higgs-executed.html [perma.cc/68S7-4XWJ].

Eleven months later, it became a crime to mail abortion pills to Texas.6See S. 4, 87th Leg., 2d. Spec. Sess. § 5(b)(1) (Tex. 2021). Under a new Texas law, a doctor in Vermont who prescribes two pills online faces years in a Texas jail.7Id.; see also Anne Branigin, In Texas, Anyone Who Mails Abortion Pills Can Now Be Sent to Jail, The Lily (Dec. 7, 2021), https://www.thelily.com/in-texas-anyone-who-mails-abortion-pills-can-now-be-sent-to-jail [perma.cc/637D-4ATV]; Letter from Jerry McGinty, Director, Legis. Budget Bd. to Hon. Lois W. Kolkhorst, Chair, Senate Comm. Health & Hum. Servs. (Aug. 8, 2021), https://capitol.texas.gov/tlodocs/872/impactstmts/html/SB00004IB.htm [perma.cc‌/GZ2S-NQGL]. A Texas court can issue a warrant for that doctor’s arrest, and the governor could seek the physician’s extradition.8See Tex. Code Crim. Proc. Ann. art. 51.13, § 22 (West 2015). The current governor, Greg Abbott, hailed the statute as “a celebration of Texas values.”9Branigin, supra note 7. A lobbyist behind the bill described it as an effort to ensure that Texas prosecutors can reach people “outside of the state’s strict limits.”10See Ashley Lopez, Prescribing Abortion Pills Online or Mailing Them in Texas Can Now Land You in Jail, NPR (Dec. 6. 2021, 1:42 PM), https://www.npr.org/sections/health-shots/2021/12/06/1060160624/prescribing-abortion-pills-online-or-mailing-them-in-texas-can-now-land-you-in-j [perma.cc/QML9-ZJ2D].

These stories seem to have little in common. But both challenge the intuition that criminal law is tied to a particular place. In the first case, the government selected Indiana as the right site to impose a punishment handed down in Maryland. In the second, the Texas legislature criminalized conduct across the country. These cases feel strange—legal, perhaps, but unconventional. To those trained in Anglo-American criminal law, it is odd to think that a Texas criminal statute can reach Vermont and that a crime committed in Maryland can result in an execution in Indiana, hundreds of miles and several states away.

These cases are unsettling because they conflict with a basic assumption about how domestic criminal law works. It is a bedrock principle of American criminal law that the authority to try and punish someone for a crime arises from the crime’s connection to territory. Thus, we assume that a person who commits a crime in some place—say, Philadelphia—can be arrested by Philadelphia police for conduct deemed criminal by the Pennsylvania legislature, prosecuted in a Philadelphia court before a jury of Philadelphia residents, and punished in a Pennsylvania prison. As a corollary, we also tend to assume that Pennsylvania criminal law stops at the state’s borders and that Ohio courts cannot apply Pennsylvania criminal law.11There are conceptual and technical differences between legislative jurisdiction to criminalize conduct, judicial jurisdiction to try criminal cases, constitutional venue requirements, choice of law rules, and criminal law enforcement norms. These aspects of criminal jurisdiction need not run together: the geographic scope of the criminal law is a separate question from which government’s law applies in a given criminal case; a legislature can have power a court lacks; and so on. But as this Article explains, these distinctions are underappreciated and undertheorized because the territoriality principle is so entrenched in Anglo-American criminal law.

The idea that criminal law is bound to geography in this way is called the territoriality principle.12See R.A. Duff, The Realm of Criminal Law 103–04 (2018); Lucia Zedner, Is the Criminal Law Only for Citizens? A Problem at the Borders of Punishment, in The Borders of Punishment 40, 46 (Katja Franko Aas & Mary Bosworth eds., 2013). This idea is deeply embedded in American legal thought. The Model Penal Code (MPC) cites the territoriality principle as a “maxim of American jurisprudence” that differentiates criminal from civil law.13 . Model Penal Code § 1.03 note (Am. L. Inst., Proposed Official Draft 1962). Philosophers identify territoriality as one of the central tenets of criminal law in common law regimes.14Duff, supra note 12, at 106; see also Zedner, supra note 12, at 46; Lindsay Farmer, Territorial Jurisdiction and Criminalization, 63 U. Toronto L.J. 225, 230 n.14 (2013); Kimberly Kessler Ferzan, The Reach of the Realm, 14 Crim. L. & Phil. 335, 336 (2020). Conflicts experts presume that criminal laws are territorial and thus “obviously” fall outside their field.15Herma Hill Kay, Larry Kramer, Kermit Roosevelt & David L. Franklin, Conflict of Laws 84–85 (10th ed. 2018). Constitutional scholars describe the territorial approach to criminal law as a “very long tradition” in American jurisprudence,16Rick Hills, Will (Should) Congress Use Its FF&C “Effects” Power to Regulate Post-Roe “Abortion Tourism”?, PrawfsBlawg (Dec. 6, 2021, 3:14 PM), https://prawfsblawg.blogs.com/‌prawfsblawg/2021/12/will-should-congress-use-its-ffc-effects-power-to-regulate-post-roe-abortion-tourism.html [perma.cc/4F2L-KMGZ] (discussing “the very long tradition of states[] using the law of the place of alleged offense to govern crimes”). and the Supreme Court treats the territoriality of domestic criminal law as self-evident.17See, e.g., Bigelow v. Virginia, 421 U.S. 809, 823–24 (1975) (stating that Virginia’s legislature “obviously” could not criminalize abortion in New York or “prosecute [its residents] for going there”). These sources rarely specify why criminal law is or must be territorial,18Philosophers of criminal law are an exception. See supra note 14; see also Gideon Yaffe, Punishing Non-citizens, 14 Crim. L. & Phil. 347, 350 (2020). but they invoke territorialism as a foundational and distinctive feature of domestic criminal law. Everyone else simply assumes the territoriality principle. In practice, the idea that domestic criminal law is territorial is so familiar that it usually goes unstated.

This Article foregrounds and questions the territoriality principle. Drawing on a broad and eclectic set of sources—from penal codes and constitutional cases to private contracts, police manuals, interviews, and materials obtained through open records requests—it argues that domestic criminal law is less territorial than conventional wisdom holds. Territorialism is central to criminal law ideology and a good deal of legal doctrine. But the territoriality principle is a norm in decline.

In reality, over the past century, new laws and enforcement practices have unmoored criminal law from geographic boundaries. This phenomenon has unfolded at every phase of the criminal legal process, from criminalization to punishment. Over time, legislatures have redefined crimes to stretch across territorial borders; courts have expanded their own criminal jurisdiction and recognized states’ authority to prosecute extraterritorial acts; police have pooled their resources and formed interjurisdictional task forces; and prison superintendents have agreed to trade prisoners between states.19See infra Part II. The result is a legal system in which state criminal law exceeds state borders. In some cases, criminal law need not be territorial at all.

Legal scholars have largely overlooked these developments. But the decline of the territoriality principle has striking implications. As a matter of doctrine, the shift away from territorialism upsets legal rules built on the classic account of criminal law. The claim that domestic criminal law is territorial runs throughout substantive criminal law and constitutional jurisprudence. The territoriality principle shapes federalism and criminal procedure doctrine and surfaces in everything from habeas and abstention cases to Commerce Clause disputes.20See infra Section I.D. The erosion of territorialism makes these lines of doctrine feel anachronistic and strained.

At a more practical level, the evolution of criminal jurisdiction creates basic good governance concerns. As this Article explains, the territoriality principle receded over the course of the twentieth century as police and prison officials began to enter private agreements—contracts, compacts, memoranda of understanding—that extend the reach of criminal laws. These agreements effectively rewrite the borders of domestic criminal law. Their proliferation raises pressing questions about transparency in criminal law enforcement.

The decline of territorialism also has broad theoretical implications. Although it is rarely discussed, the territoriality principle plays a critical role in defining and legitimating criminal law. In treatises and legal opinions, territoriality is understood as a unique feature of domestic criminal law—a background principle that distinguishes the field from both civil and international law.21See infra Section I.A. On the standard account, domestic criminal law takes a more territorial approach to jurisdiction than international criminal law, and criminal law’s commitment to territorialism explains why choice of law exists in only the civil domain.22See infra Sections I.A, III.A. In theory, territorialism also helps to justify sanctions imposed in the name of “the people.”23See infra Section III.B (exploring the relationship between territorialism and the claim that criminal law is a form of public law in which prosecutors represent “the people”).

If territorialism is receding, these claims make less sense. If, as this Article shows, domestic criminal law contains nonterritorial theories of jurisdiction and the administration of criminal law blurs territorial boundaries, then criminal law is not so different from other fields. And if criminal law is not distinctive, it is difficult to explain some of the more peculiar aspects of our criminal legal system. Without territorialism, for example, it is hard to understand the prohibition on choice of criminal law and the presumption that state criminal law stops at state lines. It also becomes trickier (even more impossible than it already is) to defend penal sanctions. To echo Professor Richard Ford, if criminal jurisdiction is not territorial, “all that is solid melts into air.”24Richard T. Ford, Law’s Territory (A History of Jurisdiction), 97 Mich. L. Rev. 843, 851 (1999); cf. William Shakespeare, The Tempest act 4, sc. 1; Karl Marx & Friedrich Engels, The Communist Manifesto (Samuel Moore trans., Penguin Classics 2002) (1848).

That insight, in turn, opens up some interesting questions about whether territorialism is good for criminal law. This Article’s basic thesis is that the borders of domestic criminal law are up for grabs. Questions of criminal jurisdiction are much more contested and flexible than many scholars and critics of American criminal justice seem to have realized. Recognizing this fact reveals some provocative possibilities for criminal justice reform, such as redistricting police forces, regionalizing criminal lawmaking, and permitting criminal defendants to forum shop. The erosion of territorialism has the potential to transform the design and administration of criminal law.

Yet, at the same time, the story outlined below should give reformers pause. The decline of the territoriality principle has not been a boon for criminal defendants. On the contrary, one of this Article’s core lessons is that the erosion of territorialism has empowered police, prosecutors, and prison wardens, who have benefited from growing flexibility about the outer boundaries of their authority. The history of criminal jurisdiction demonstrates that, although they are fraught and arbitrary, borders are a crucial restraint on the police power.25By police power, I mean the amorphous and supposedly inherent power of a government—in the United States, state governments—to secure social order, including by using actual police to enforce criminal laws. For a history of the term, see Markus Dirk Dubber, The Police Power (2005). See also Bond v. United States, 572 U.S. 844, 854 (2014) (contrasting states’ “police power” and the federal government’s limited, enumerated powers). Those concerned about abuse of power in the criminal justice system might then want to revive territorialism rather than embrace its decline. Animated by that concern, this Article makes a case for more territorialism in American criminal law.

This argument proceeds in three Parts. Part I defines the territoriality principle and traces its history. At first, it is unclear why so many authoritative sources seem to believe that domestic criminal law has to be territorial. Part I sets out to answer that question. It locates the origins of the territoriality principle in substantive criminal law, constitutional law, and the institutional design of the criminal justice system.

Part II documents territorialism’s decline over the course of the twentieth century. Bringing together a wide variety of sources, Part II depicts a criminal justice system that has been deterritorialized in two respects. First, courts have embraced nonterritorial conceptions of criminal jurisdiction—holding, for example, that states can predicate their criminal laws on citizenship rather than presence on state soil.26See, e.g., Skiriotes v. Florida, 313 U.S. 69, 73–77 (1941); see also infra Section II.A. Second, the administration of criminal law has become more centralized and collaborative. Together, these trends have produced a legal regime that is less territorial than one might expect, and, where it remains tied to geography, is much less wedded to state borders than one steeped in doctrine and rhetoric about American criminal law would imagine.

Part III explores the implications of this account. It begins by explaining how the deterritorialization of criminal law undermines doctrines and conceptual distinctions that separate criminal law from other fields. Part III then turns to the normative questions implicit in the erosion of the territoriality principle. In the end, the Article offers a defense of territorial jurisdiction.

Its real aim, though, is to demonstrate that the territoriality principle is worth debating. At its heart, this is a piece about the importance of criminal jurisdiction. Scholars and students of American law tend to think that the study of jurisdiction is technical and dry, or as Professor William Prosser put it in 1953, “a dismal swamp” inhabited by “eccentric professors who theorize about mysterious matters in a strange and incomprehensible jargon.”27Lea Brilmayer, Jack Goldsmith, Erin O’Hara O’Connor & Carlos M. Vázquez, Conflict of Laws, at xxiv (8th ed. 2020) (quoting William L. Prosser, Interstate Publication, 51 Mich. L. Rev. 959, 971 (1953)). Perhaps it is unsurprising, then, that there has been little scholarship on territoriality in domestic criminal law.28There is a very small body of early- and mid-twentieth century writing on territoriality in criminal law. See, e.g., Albert Levitt, Jurisdiction over Crimes (pt. 1), 16 J. Am. Inst. Crim. L. & Criminology 316 (1925) [hereinafter Levitt I]; Albert Levitt, Jurisdiction over Crimes (pt. 2), 16 J. Am. Inst. Crim. L. & Criminology 495 (1926) [hereinafter Levitt II]; Daniel L. Rotenberg, Extraterritorial Legislative Jurisdiction and the State Criminal Law, 38 Tex. L. Rev. 763 (1960); Robert A. Leflar, Conflict of Laws: Choice of Law in Criminal Cases, 25 Case W. Rsrv. L. Rev. 44 (1974); Wendell Berge, Criminal Jurisdiction and the Territorial Principle, 30 Mich. L. Rev. 238 (1931). There is also a larger and more recent body of literature on territoriality in adjacent fields like conflict of laws and immigration law, and in the last decade philosophers of criminal law (particularly in Europe) have grown increasingly interested in the scope of legislative criminal jurisdiction. See infra Sections I.A, II, III.B. But scholarship on this topic remains rare and almost none since the 1970s has examined how the territoriality principle works (or fails to work) in the United States. As Professor Markus Dubber observed in 2013, “Despite the recent upsurge of interest in criminal jurisdiction in the international sphere, domestic criminal jurisdiction remains understudied. This is a shame . . . .” Markus D. Dubber, Criminal Jurisdiction and Conceptions of Penality in Comparative Perspective, 63 U. Toronto L.J. 247, 247 (2013). American criminal law textbooks omit jurisdiction, and the leading article on the topic remains a piece published in 1926.29Levitt II, supra note 28; see Leflar, supra note 28, at 44; see also, e.g., Sanford H. Kadish, Stephen J. Schulhofer & Rachel E. Barkow, Criminal Law and Its Processes (10th ed. 2017). In the meantime, criminal law scholars have turned their attention toward the political incentives and institutional dynamics that make the country’s criminal justice system so harsh and indefensible.30This literature is vast. For two examples, see Jonathan Simon, Governing Through Crime (2007), and William J. Stuntz, The Collapse of American Criminal Justice (2011).

But the territoriality principle persists in doctrine, theory, and ideology about criminal law, and it plays a pivotal role in justifying the power to punish. In other areas of law, academics are having lively debates about the merits and drawbacks of territorialism. Scholars of constitutional law, federal courts, civil procedure, immigration law, international law, and conflicts have all asked whether the law they study is (or should be) based on a territorial theory of state power.31 See, e.g., Kal Raustiala, Does the Constitution Follow the Flag? (2009) (constitutional law); Stephen E. Sachs, The Unlimited Jurisdiction of the Federal Courts, 106 Va. L. Rev. 1703 (2020) (federal courts); Linda Bosniak, Being Here: Ethical Territoriality and the Rights of Immigrants, 8 Theoretical Inquiries L. 389 (2007) (immigration law); Cedric Ryngaert, Territorial Jurisdiction over Cross-Frontier Offences: Revisiting a Classic Problem of International Criminal Law, 9 Int’l Crim. L. Rev. 187 (2009) (international law); Michael H. Hoffheimer, The Case Against Neo-Territorialism, 95 Tul. L. Rev. 1305, 1306 (2021) (civil procedure); Douglas Laycock, Equal Citizens of Equal and Territorial States: The Constitutional Foundations of Choice of Law, 92 Colum. L. Rev. 249, 317 (1992) (conflicts). Criminal law scholars have said less about this question. Yet the shift away from territorialism is no less evident in criminal law, and it may even be more destabilizing to the field.

Moreover, this seemingly obscure topic has serious stakes. The decline of the territoriality principle is why Dustin Higgs could be executed in Indiana.32See United States v. Higgs, 141 S. Ct. 645, 647–48 (2021) (mem.) (Sotomayor, J., dissenting). It is why Texas can criminalize prescriptions written in Vermont.33See Skiriotes v. Florida, 313 U.S. 69, 77 (1941); Higgs, 141 S. Ct. at 648–49 (Sotomayor, J., dissenting). There are of course more technical legal explanations for these cases. (Higgs’s case, for example, turned on interpretation of the Federal Death Penalty Act, 18 U.S.C. § 3596(a)). But as this Article explains, the statutes and doctrines that permit the sort of displacement involved in extending Texas criminal law to cover conduct in Vermont or redesignating a federal sentence imposed in Maryland to permit an execution in Indiana developed from the deterritorialization of criminal law. To come into existence, these statutes and doctrines required a basic shift in how the American legal system approaches criminal jurisdiction. This Article documents that shift. It is also—to use a looming example—why it is less clear than one might think whether Missouri can ban its residents from getting abortions outside the state now that the Supreme Court has overturned Roe v. Wade.34Dobbs v. Jackson Women’s Health Org., 142 S. Ct. 2228 (2022) (overruling Roe v. Wade, 410 U.S. 113 (1973)). To be clear, there are compelling—in my view, winning—arguments that states cannot prohibit out-of-state abortions, whether or not criminal jurisdiction is territorial. Legal scholars have argued that extraterritorial regulation of abortion is impermissible under a standard conflicts analysis. See infra note 295. Scholars have also explored constitutional limits on such laws. See infra note 148. The point here is simply that the decline of territorialism has unsettled a theory of criminal jurisdiction that would provide an easy resolution to questions about the ambit of state criminal law. The legal system’s waning commitment to territorialism raises disquieting questions about the limits of criminal law.

I. The Origins of the Territoriality Principle

The territoriality principle is so entrenched in domestic criminal law as to feel unremarkable. It is uncontroversial to think that a crime committed in Maine will be policed, prosecuted, and punished in Maine. This Part asks why this approach to criminal law seems straightforward—how, in other words, the conventional wisdom came to be.

Before beginning, it is helpful to introduce my key term. As I explain below, the territoriality principle is a shorthand for two ideas: first, that a government’s power to define and punish crime arises from its authority over a bounded geographic region; and second, that criminal laws apply to any person within that region. The territoriality principle is thus a claim about both the source and the subjects of criminal law. In a territorial criminal law regime, governments with legally defined borders have authority to criminalize conduct within those borders. That authority extends to anyone who is physically present, regardless of her personal traits or place of residence.

Accordingly, it undermines the territoriality principle for state legislatures to criminalize out-of-state conduct. It would also offend the territoriality principle to apply criminal laws only to citizens or residents and not, for instance, to noncitizens and visitors. In the first example, the government extends the reach of criminal law beyond its borders. In the second, the government predicates criminal law on legal status rather than presence on state soil.35Territorial borders are themselves sources of legal status. (For example, citizenship status flows from the territorial concept of a bordered nation.) See generally Ford, supra note 24, at 858. My assertion is not that territorialism is status-neutral but rather that there is a basic distinction between territorial legal systems, in which law attaches by virtue of physical presence, and nonterritorial regimes, in which law attaches by virtue of some other trait (such as membership in a group) and follows a person as she travels. Part III explores this distinction in more detail.

This Article is interested in both kinds of violations of the territoriality norm because both upend traditional assumptions about how domestic criminal law works. My basic claim is that criminal law has become less territorial over time—more expansive, more flexible, and more receptive to forms of jurisdiction based on identity rather than geography. To understand why this development is notable, it helps to begin by asking why territorialism has so much purchase in the classic account of criminal law.

A. Criminal Law

Territorial borders are built into the definitions of American criminal law. In the first weeks of law school, students learn that a crime is an act that happens in a given place as a result of voluntary bodily movement.36See, e.g., Kadish et al. supra note 29, at 221–27. This axiom, known as the voluntary act requirement, embeds a “where” question in the foundation of criminal law. To know that a crime occurred, one must decide where the salient conduct—the actus reus—took place. This is why courts describe venue as “a necessary, if often subtle, element of every criminal statute.”37United States v. Lewis, 768 F.3d 1086, 1089 (10th Cir. 2014). Appellate courts disagree about whether venue must be proven beyond a reasonable doubt. Compare United States v. Lopez, 880 F.3d 974, 982 (8th Cir. 2018) (“Proof of venue is an essential element of the Government’s case . . . .” (quoting United States v. Netz, 758 F.2d 1308, 1312 (8th Cir. 1985)), with United States v. Romans, 823 F.3d 299, 309 (5th Cir. 2016) (“This Circuit has not treated territorial jurisdiction and venue as ‘essential elements’ in the sense that proof beyond a reasonable doubt is required.” (quoting United States v. White, 611 F.2d 531, 536 (5th Cir. 1980)). But the basic point—that the site of a crime is part of its definition, necessary to prosecution in any jurisdiction—is consistent across circuits. At its core, crime is an idea tied to a location.38See Michael Hirst, Jurisdiction and the Ambit of the Criminal Law 2 (2003) (“[I]ssues that criminal lawyers generally classify as ‘matters of jurisdiction’ . . . as if they were merely concerned with the powers or competence of the court, should properly be considered as elements relating to the actus reus itself.”).

Traditionally, the relevant location of a crime is a legally defined territory whose political body has deemed the action criminal. Stating this principle out loud makes it sound complicated, but the idea is familiar. When I commit an assault in Battery Park, what matters is that I am in New York because states define and prosecute assaults. (One could make the same point about federal crimes or city ordinances. The observation is simply that the ambit of domestic criminal law is territorial, and territories have formal legal boundaries.) In other words, when defining crimes, American criminal law generally cares about where you are rather than who you are. The relevant fact is that you are on New York land, not that you are an American citizen, a New York citizen, a senior citizen, or a person who identifies as female, Black, or Jewish. And because of the way criminal law has developed in the United States, the operative “where” is most often the state. Thus, the Model Penal Code provides that criminal law typically extends to any offense committed “within the [s]tate.”39 . Model Penal Code § 1.03 (Am. L. Inst., Proposed Official Draft 1962); see infra Section II.A (exploring exceptions to “strict territoriality”).

The territorial scope of criminal law is supposed to be one of the things that makes it distinctive. In a section of the Model Penal Code called “territorial applicability,” the MPC’s reporters explain that criminal law is fundamentally different from civil law because in the criminal context, “jurisdiction and choice of law . . . are merged.”40 . Model Penal Code § 1.03 explanatory note (Am. L. Inst., Proposed Official Draft 1962). Again, the principle sounds tricky but feels simple. In a New York criminal court, New York criminal law applies. Whereas in civil actions a court can apply a sister state’s rules, in criminal court the forum state’s law governs.41Farmer, supra note 14, at 230 n.14 (“[W]hile the civil courts will on occasion seek to apply foreign law, criminal courts can only ever apply domestic law, the law of the territory.”). A New York judge cannot choose to apply Connecticut criminal law to your case even if you live in Connecticut or planned your crime there. No, the MPC tells us, “it has long been a maxim of American jurisprudence that a state will not enforce the penal laws of another state.”42 . Model Penal Code § 1.03 (Am. L. Inst., Proposed Official Draft 1962); see also Kay et al., supra note 15, at 84 (tracing this maxim to “Chief Justice Marshall’s unsupported dictum” about the law of nations in The Antelope, 23 U.S. (10 Wheat.) 66, 123 (1825)); see infra note 278 (discussing The Antelope’s importation into domestic criminal law). So if you committed your crime in New York, you are prosecuted there. And if you committed a crime in Connecticut and drove to New York, you must go back to Connecticut to be prosecuted in the right place.

Conflicts scholars call the prohibition on interstate enforcement of criminal laws “the public law taboo.”43William S. Dodge, Breaking the Public Law Taboo, 43 Harv. Int’l L.J. 161 (2002). As Dodge notes, the “phrase is Andreas Lowenfeld’s.” Id. at 161 n.4; see also Andreas F. Lowenfeld, Public Law in the International Arena: Conflict of Laws, International Law, and Some Suggestions for Their Interaction, 163 Collected Courses Hague Acad. Int’l L. 311, 322–26 (1979); Kay et al., supra note 15, at 84–85 (discussing the scope of the “so-called penal law exception” and stating in passing that “criminal laws obviously qualify” and are therefore unenforceable outside the jurisdiction that enacted the law). When discussing the public law taboo, conflicts scholars usually refer to “public law” generally, not to choice of criminal law in particular. My invocation of the taboo is an effort to apply ideas from a field focused on civil litigation to criminal law theory. Robert Leflar made a similar move in 1974. See Leflar, supra note 28, at 44 (“Conflict of laws treatises and casebooks usually omit material on choice of law in the criminal law field almost altogether. They take up conflicts problems as though they arose only in civil litigation.”). This blind spot in conflicts scholarship is a testament to the territoriality principle’s hold on criminal law. The intuition driving this taboo is that there is something special about public law, and “penal” law in particular, that makes it offensive and illegitimate to apply a criminal law outside the place it was enacted. It is not entirely clear why penal law is so special—why, for instance, it is more problematic for a state to impose another state’s criminal rules than its civil rules. The answer seems to lie in deeply held (if not especially concrete) beliefs about criminal law’s relationship to concepts like sovereignty and democracy. Later Sections of this Article elaborate and question those beliefs.44See infra Section III.B. See generally Dodge, supra note 43, at 164 (challenging the rationales for the public law taboo). For now, the critical point is that criminal law is supposed to be territorial in ways that civil law is not.

To be more precise, domestic criminal law is supposed to be territorial. When it comes to international law, scholars and courts have long recognized several bases for criminal jurisdiction, including forms of extraterritorial jurisdiction in which nations can criminalize conduct by and against their citizens abroad.45See Cedric Ryngaert, Jurisdiction in International Law 104–10 (2d ed. 2015) (discussing active and passive personality jurisdiction); Florian Jeßberger, A Short History of Jurisdiction in Transnational Criminal Law, in Histories of Transnational Criminal Law 261 (Neil Boister, Sabine Gless & Florian Jeßberger eds., 2021). For international law scholars, it is not unusual in the least to think that criminal law could be pegged to a personal trait like citizenship status rather than to the place the crime was committed. That idea is also familiar to those who prosecute federal crimes such as terrorism and human trafficking, which often involve extraterritorial conduct and out-of-country defendants.46See, e.g., Michael Farbiarz, Accuracy and Adjudication: The Promise of Extraterritorial Due Process, 116 Colum. L. Rev. 625, 625 (2016) (discussing federal terrorism and drug prosecutions). Scholars of Roman and German law would also find membership-based jurisdiction familiar. See, e.g., Dubber, supra note 28, at 268 (German criminal law); Rollin M. Perkins, The Territorial Principle in Criminal Law, 22 Hastings L.J. 1155 (1971) (Roman criminal law). Territoriality is one option among many for international law theorists and national security lawyers.

But when it comes to the criminal law that governs conduct in the United States—the criminal law under which tens of millions of state prosecutions proceed each year—territorialism is entrenched and rarely questioned. As the Introduction noted, criminal law textbooks omit jurisdiction.47See, e.g., Kadish et al., supra note 29. When the subject is discussed, territoriality is understood to be the prevailing rule. In 1909, the Supreme Court announced without fanfare that an act “done within the territorial limits of [one state] . . . cannot be prosecuted and punished by [another state].”48Nielsen v. Oregon, 212 U.S. 315, 321 (1909). A century later, criminal law philosophers describe territoriality as the principle “we currently observe,”49Ferzan, supra note 14, at 336 (comparing extraterritorial bases for international criminal jurisdiction to “the territorial boundaries we currently observe” in domestic criminal law); see, e.g., Duff, supra note 12, at 103 (noting, in a discussion of criminal law, that “ambit and jurisdiction are typically defined in territorial terms”). a central tenet in “classic[] accounts [of] domestic criminal law,”50Zedner, supra note 12, at 46. and a norm that distinguishes criminal from international law, a field in which ideas about jurisdiction are more varied and flexible.51See supra notes 45–46 and accompanying text. A commitment to the territoriality principle is thus thought to be what makes domestic criminal law both domestic and criminal.

To be clear, the territoriality principle is neither a settled doctrine nor an absolute rule. Cases and treatises on criminal law never quite specify what sort of law territoriality is—a doctrine, a practice, or simply an assumption. (This Article is one attempt to solve that mystery.) There are also recognized exceptions to the territoriality principle in state criminal law. The Model Penal Code acknowledges that states can criminalize out-of-state conduct that affects their “legitimate interests,” and the MPC’s reporters take pains to note that the Code sets “forth a number of alternative bases for jurisdiction, thus rejecting the old common law doctrines of strict territoriality.”52 . Model Penal Code § 1.03(1)(f) (Am. L. Inst., Proposed Official Draft 1962). But most of those exceptions are territoriality adjacent; the MPC imagines, for example, criminalizing out-of-state conspiracies that involve an overt act “within the State.”53Id. § 1.03(1)(c). These are, moreover, exceptions to a baseline rule that domestic criminal law is territorial. That rule guides the MPC and explains the strong presumption against extraterritoriality when interpreting state criminal statutes.54See William S. Dodge, Presumptions Against Extraterritoriality in State Law, 53 U.C. Davis L. Rev. 1389, 1391, 1412 (2020) (noting that “[s]tate rules on extraterritoriality . . . differ from the federal presumption” and citing the California Supreme Court’s assumption that a state dram shop law could have no extraterritorial effect because it was “a criminal statute”).

The idea, in short, is that paradigmatic criminal law—the stuff taught in first-year criminal law classrooms—is territorial. Much of this Article is devoted to challenging that idea, to exploring when state criminal law fails to be territorial and how the deterritorialization of domestic law undermines claims about the distinctiveness of criminal law and its special relationship to democracy. These critiques make more sense and land harder if one pauses at the outset to observe just how ingrained territorialism is in Anglo-American criminal law, and state criminal law in particular. On the conventional account, the relationship between criminal law and territory is uncomplicated. Extraterritorial criminal law is exceptional and technical, the sort of material one learns in upper-level courses and specialist subjects like international or national security law. And territorialism is the characteristic that makes criminal law identifiable as a form of domestic (rather than international), criminal (rather than civil), and public (rather than private) law.

B. Constitutional Law

One finds a similar commitment to the territoriality principle in American constitutional law. The federal Constitution contains several provisions that appear to require criminal law to be territorial. The two most obvious are the Venue Clause in Article III, which fixes the place of criminal trials “in the State where the said Crimes shall have been committed,”55 . U.S. Const. art. III, § 2, cl. 3. This rule applies except in cases of impeachment or when a crime is “not committed within any State.” Id. and the Vicinage Clause of the Sixth Amendment, which requires juries to be drawn from “the State and district wherein the crime shall have been committed.”56 . Id. amend. VI. There are distinctions between these two provisions: venue is the location of a criminal trial; vicinage is the catchment area for the jury.57Drew L. Kershen, Vicinage (pt. 1), 29 Okla. L. Rev. 801, 805 (1976) [hereinafter Kershen I]; see also Drew L. Kershen, Vicinage (pt. 2), 30 Okla. L. Rev. 1, 150 (1977) [hereinafter Kershen II]. Kershen’s article was published across two volumes of the Oklahoma Law Review. See also Scott Kafker, Comment, The Right to Venue and the Right to an Impartial Jury: Resolving the Conflict in the Federal Constitution, 52 U. Chi. L. Rev. 729, 729 n.2 (1985). But the clauses are related conceptually and historically, and both are responsible for the strong strain of territorialism that runs through American criminal law.

The Venue Clause was a response to British laws that permitted colonists to be transported to England—“3,000 miles to an alien environment”—for criminal trials.58Kershen I, supra note 57, at 806, 809. In a magisterial article published in 1976, Professor Drew Kershen explained that the Constitution’s authors adopted the Venue Clause with two ideas in mind. First, the Founders believed that limiting venue would protect defendants because the “legal and moral support” necessary to mount a defense was most readily available close to home.59Id. at 808. Second, and more fundamentally, the drafters espoused a common law theory of criminal jurisdiction in which courts may “enforce criminal laws only with respect to crimes committed within the territory assigned to the court.”60Id. at 811 (citing debates at the Constitutional Convention).

Under this common law theory, venue (the place of a criminal trial) and jurisdiction (the power to hear a criminal case) were one and the same. Courts had no authority to enforce the criminal laws of another place, so of course criminal trials would be held within a court’s territory. This understanding of criminal law—in which venue and jurisdiction are merged—is what early Americans meant when they described criminal law as “local.”61Id. at 811 & n.27. To say that criminal law was local was not merely to recount a fact about criminal adjudication. Calling criminal law local was a claim about the source and limits of judicial power.

The Founders expected that the twin goals of the Venue Clause—protecting defendants and constitutionalizing territorial jurisdiction—would align. Because people usually committed crimes near their homes,62See id. (noting, based on a review of constitutional debates, that the Venue Clause’s drafters expected that the site of a “crime and the residence of the accused would ordinarily coincide”). the Constitution’s authors assumed that tying venue to the site of a crime would have the incidental benefit of protecting criminal defendants. But when that assumption faltered, the Founders chose territoriality as the touchstone of the Venue Clause.63Id. Rather than linking venue to a defendant’s residence, the drafters tied venue to the crime’s location. The theory behind this choice was that the government’s authority to enforce criminal laws arises from a power of self-determination that starts and stops at territorial borders—that is, from sovereignty.64Id.; see also infra Section III.B.1 (defining and critiquing the concept of sovereignty). From this perspective, the choice between tying venue to a defendant’s residence and predicating venue on a crime’s situs was no choice at all. Venue had to track the crime because the government’s power to enforce criminal law flowed from territory, not from facts about the defendant.65Kershen II, supra note 57, at 150 (“[A]n accused’s place of residence is irrelevant under the Constitution.”).

The Venue Clause thus constitutionalized territorialism. The Vicinage Clause did too, though in slightly different ways. The Sixth Amendment’s vicinage requirement began as a proposed amendment to the text of the Constitution, which James Madison introduced to the House in 1789.661 Annals of Cong. 435 (1789) (Joseph Gales ed., 1834). Madison’s proposal would have replaced Article III’s Venue Clause with language requiring trials before a jury “of the vicinage” in the county where the crime occurred, except when insurrectionists controlled that county (in which case the trial could be moved nearby).67Kershen I, supra note 57, at 820; Kershen II, supra note 57, at 129 (quoting Madison’s proposal). This proposal morphed several times during the First Congress and ultimately landed in the Bill of Rights rather than Article III.68Kershen I, supra note 57, at 821–22 (explaining that before the proposed amendment to the Venue Clause went to the Senate, the House voted “to submit the propositions to the Senate as a Bill of Rights, rather than as amendments which would be inserted at the appropriate place in the body of the Federal Constitution”). At a pivotal moment in the process, the Senate rejected the term “vicinage” on the ground that it was “either too vague or too strict”69Id. at 822 (quoting Letter from James Madison to Edmund Pendleton (Sept. 23, 1789), in 5 The Writings of James Madison 424 & n.1 (Gaillard Hunt ed., 1904)).—too vague because it did not refer to a “particular geographical territory recognized as a political or governmental unit,”70Id. at 823. and too strict because it might require trials to be held in rebellious counties where juries would not convict.71Id. at 824 (quoting Power of the Judiciary (June 20, 1788), in 5 The Writings of James Madison 216, 224 (Gaillard Hunt ed., 1904).); see also Ramos v. Louisiana, 140 S. Ct. 1390, 1400 n.40 (2020) (“In private writings, Madison . . . explain[ed] some of the Senate’s objections with his original phrasing of the vicinage requirement.”).

The dispute over the word “vicinage” is illuminating. Supporters of the vicinage provision liked the idea that vicinage could mean “neighborhood” or “community,” loose concepts that captured a preference for localism and a belief that juries were “the conscience of the community” and had a substantive role to play in interpreting criminal laws.72Kershen I, supra note 57, at 823, 833–36. Opponents of the word “vicinage”—the Federalists—were worried about national uniformity and local rebellions, so they wanted a clear term that would not require “neighborhood” juries in every case.73Id. at 823–24. The final language of the Sixth Amendment was a negotiated settlement between these camps. The Amendment requires juries to be drawn from the “State and district wherein the crime shall have been committed, which district shall have been previously ascertained by law.”74 . U.S. Const. amend. VI. Section 29 of the Judiciary Act of 1789, which implemented the Sixth Amendment and established the first judicial districts, created an exception for capital cases: it required trials for crimes punishable by death to be in the county where the crime was committed unless “great inconvenience” would occur, in which case the trial could be moved. However, at least twelve jurors still had to “be summoned from thence.” Judiciary Act of 1789, ch. 20, § 29, 1 Stat. 73, 88. This provision preserved the narrower “right to trial by a jury of the vicinage as at common law” for capital crimes. Kershen I, supra note 57, at 854; see also Charles Warren, New Light on the History of the Federal Judiciary Act of 1789, 37 Harv. L. Rev. 49, 105 (1923). But it became dead letter—courts routinely found “great inconvenience”—and was repealed in the 1862 Amendments to the Judiciary Act. Kershen II, supra note 57, at 55–61.

This rule retains and entrenches a basic form of territorialism: it makes the site of a crime the threshold issue when determining the proper criminal procedure. But the Sixth Amendment calls for juries from “states” and “districts” rather than the “vicinage.”75See U.S. Const. amend VI. The final vicinage provision is thus not really a vicinage provision. Instead, it summons juries from formal political units with legally defined boundaries, and it refers to districts “ascertained by law,” a phrase that empowers the legislature to determine the jury’s geographic origins within a state. In this respect, the Sixth Amendment was a significant grant of discretion to Congress.76Kershen II, supra note 57, at 46. Note, however, that the borders of federal judicial districts coincided with state borders at the time the Sixth Amendment was enacted, so at first the Amendment simply required what the Venue Clause already did—namely, trial within the state. Id. The Amendment also represented the triumph of legal notions of jurisdiction—and Anglo-American legal definitions of territory77See generally Gregory Ablavsky, Federal Ground (2021).—over looser and more longstanding conceptions of political community.

One could go on at some length here. The subsequent history of the venue and vicinage provisions is interesting, not least because it involves protracted disputes about what happens to criminal jurisdiction78Kershen II, supra note 57, at 3–6 (collecting cases on whether the subdivision of judicial districts limited federal courts’ criminal jurisdiction to events that occurred within a particular division, a question the Supreme Court answered in the negative in three cases decided in the 1890s). As Part II explains, these late-nineteenth-century cases created a new distinction between venue and jurisdiction. and federal juries79See id. at 46–50, 67–68 (tracing the development of vicinage doctrine, which—in broad strokes—affirms judicial discretion to summon the jury from anywhere the court chooses so long as it is within the judicial district). when Congress tinkers with the boundaries of judicial districts. Part II explores some of that history. Most notably, it examines the emergence in the twentieth century of the idea that jurisdiction and venue are distinct such that the location of a criminal trial is a waivable right possessed by the defendant rather than a structural requirement for the legitimacy of criminal adjudication.80See infra Part II. As Part II explains, the fracturing of jurisdiction and venue marks an important moment in the deterritorialization of American criminal law.81Id.

But at this stage, that history is a detour from the core point: the commitment to territorialism in American criminal law can be traced to the U.S. Constitution. The creators of the federal Constitution treated criminal law as territorial when they debated and drafted the rules on criminal trials. As a result, courts have understood it to be a problem of constitutional significance when a crime is tried outside the place it was committed.82See Kershen II, supra note 57, at 3–8 (collecting cases on the constitutionality of criminal proceedings held outside the original judicial district where a crime occurred). One might cite any constitutional case alleging improper venue here too. The claim here is not that it is unconstitutional to try crimes extraterritorially. The whole purpose of this Article is to show that the legality of extraterritorial criminal law has shifted over time. Rather, the observation is that we tend to think domestic criminal law is territorial because the Constitution’s drafters assumed it had to be.

State constitutions contain the same background assumption. The Sixth Amendment vicinage right is one of the few that is not incorporated,83See, e.g., Caudill v. Scott, 857 F.2d 344, 345 (6th Cir. 1988) (“[T]he term ‘district’ as used in the Sixth Amendment ha[s] never been defined to apply to states . . . .”); Zicarelli v. Dietz, 633 F.2d 312, 325–26 (3d Cir. 1980) (“[W]e conclude that the provision of the Sixth Amendment providing for the right to have a jury from a district ‘previously ascertained by law’ applies only to federal criminal trials, and not to state criminal trials.”); see also Akhil Reed Amar, The Bill of Rights 275 (1998) (arguing against “mechanical incorporation” of the vicinage requirement). But see Steven A. Engel, The Public’s Vicinage Right: A Constitutional Argument, 75 N.Y.U. L. Rev. 1658, 1706–07 (2000) (“The Supreme Court has not determined the extent to which the Sixth Amendment Vicinage Clause is incorporated . . . . The arguments that the Fourteenth Amendment does not incorporate the Vicinage Clause are particularly unconvincing.”). but most state constitutions have Sixth Amendment analogues that build territoriality into the criminal legal process.84Leflar, supra note 28, at 46. Leflar draws on and cites Albert Levitt’s 1925 article, Jurisdiction over Crimes, which catalogued state venue and vicinage requirements, “[f]ew if any of [which] have changed.” Id.; see also Levitt I, supra note 28, at 331–35. Montana, for example, requires criminal trials to be held in “the county or district” where the crime occurred, before a jury from the same area.85 . Mont. Const. art. II, § 24; see also Levitt I, supra note 28, at 331 n.46 (counting eight states with identical clauses and more with similar provisions). Massachusetts limits criminal prosecutions to “the vicinity where [the facts] happen.”86 . Mass. Const. pt. I, art. XIII. In Louisiana, criminal trials must be in the “parish where the offense or an element of the offense occurred.”87 . La. Const. art. 1, § 16. There are distinctions between these provisions, but they all reflect and reinforce a territorial understanding of criminal law. When searching for the origins of the territoriality principle, constitutional rules on criminal trials prove to be a critical source.

Territorialism appears in other parts of the federal Constitution too. The Extradition Clause obliges states to “deliver[] up” alleged criminals who “flee from Justice.”88 . U.S. Const. art. IV, § 2, cl. 2; see also 18 U.S.C. § 3182 (implementing the Extradition Clause); Unif. Crim. Extradition Act (Unif. L. Comm’n 1936) (model extradition statute). The UCEA has been adopted by every state except Mississippi and South Carolina. Utt v. Warden, 427 A.2d 1092, 1096 n.6 (Md. Ct. Spec. App. 1981). This Clause, which immediately prompted disputes about the rendition of fugitive slaves,89See Ariela Gross & David R. Upham, Article IV, Section 2: Movement of Persons Throughout the Union, Nat’l Const. Ctr., https://constitutioncenter.org/interactive-constitution/interpretation/article-iv/clauses/37 [perma.cc/87DS-4STD]. requires interstate cooperation in criminal law enforcement. More to the point, it presumes that cooperation means sending fugitives back rather than trying them in the state where they are found. One could imagine a legal regime in which states authorized each other to enforce their criminal codes—so, for example, Georgia would say, “Massachusetts, try that defendant for me,” rather than, “Send him back to my criminal courts.” But that is not how the U.S. Constitution works. Because early Americans had sharp disagreements over the criminal laws used to regulate slavery,90Id. (describing Extradition Clause controversies beginning the year after the Constitution’s adoption, “all [of which] had to do with slavery”). and because the Founders assumed the common law conception of criminal jurisdiction,91See supra notes 60–61. the Constitution provides for extradition. As a result, when people “flee from Justice,”92 . U.S. Const. art. IV, § 2, cl. 2. we ship them back to the place where justice is supposed to happen.

American courts have also read territorialism into the Due Process Clause.93 . Id. amend. XIV, § 1. In cases involving crimes that cross state lines—for example, a cross-border shooting or a theft in one state where stolen goods wind up in another—courts have suggested that the Due Process Clause imposes an outer limit on extraterritorial prosecutions.94See, e.g., State v. Rimmer, 877 N.W.2d 652, 665–66 (Iowa 2016) (“We agree with the New Jersey Supreme Court that ‘[t]he extraterritorial application of state criminal law is subject to due process analysis’ under the Fourteenth Amendment.” (alteration in original) (quoting State v. Sumulikoski, 110 A.3d 856, 866 (N.J. 2015))); State v. Randle, 647 N.W.2d 324, 329 n.4 (Wis. Ct. App. 2002) (“Territorial jurisdiction is part of the due process restrictions on the power of a court . . . .”); see also Model Penal Code § 1.03 (Am. L. Inst., Proposed Official Draft 1962) (noting that “due process” limits the permissible bases for criminal jurisdiction). As one New York court described the doctrine, the Fourteenth Amendment prohibits states from “overreaching” when they “exercise [criminal] jurisdiction over out-of-state events.”95People v. Puig, 378 N.Y.S.2d 925, 934 (Sup. Ct. 1976). Part II explains why this doctrine is not especially restrictive in practice,96See infra Part II. but here the invocation of the Constitution is what matters. In due process cases, as in venue and vicinage cases, courts believe it to be a constitutional problem when state criminal law exceeds state borders.

Together, these examples portray a constitutional order wedded to territorialism. Indeed, they suggest that territorial criminal law was a critical part of the project of constituting the United States. The rules on extradition and criminal procedure in American constitutions are not just rules of the road for efficient trials. These rules are how states announced their borders and enacted their power to govern within them. (You can really tell you have a state once its police arrest and imprison you.) Early Americans used criminal law to instantiate state power and to navigate interstate relations, including disputes over slavery. From this perspective, the territoriality principle is bound up with the history of American federalism and reflects foundational ideas about what it means to be a union of sovereign states.

On a more prosaic level, the territoriality principle clearly has roots in constitutional law. The previous Section explored how territoriality is ingrained in substantive criminal law, in concepts like the actus reus requirement and the prohibition on interstate enforcement of “penal” laws. The observations in that Section apply broadly to criminal law in common law regimes.97See Duff, supra note 12, at 103–04 (describing the “Territorial Principle” in English criminal law); Hirst, supra note 38, at 1 (exploring the “territorial and extraterritorial ambit of English criminal law”); Dubber, supra note 28, at 248 (comparing German “jurisdictional concepts” to “jurisdictional norms in common law countries”). As Dubber notes, “domestic criminal jurisdiction remains understudied.” Id. at 247. To the extent that modern scholars have explored the topic, their discussion tends to focus on territoriality in common law regimes rather than on American territorialism in particular. This approach highlights the pervasiveness of the territoriality principle in criminal law, but it can obscure the ways in which territorialism is built into American constitutions and tied up with ideas about American federalism. Here, we might add that territorialism is also distinctively American. The country’s federal and state constitutions contain clauses mandating territorial criminal procedures, limiting extraterritorial criminal jurisdiction, and requiring states to facilitate territorial criminal law enforcement. These provisions are one key reason we tend to think criminal law must be tied to a particular place.

C. Institutional Design

Territorialism is also built into the institutional design of the criminal legal system. The Federal Rules of Criminal Procedure require crimes to be prosecuted “in a district where the offense was committed” unless narrow exceptions apply.98 . Fed. R. Crim. P. 18. For a history of this rule, detailing its origins in Section 53 of the Judicial Code of 1911, see Kershen II, supra note 57, at 15–20. See also infra Part II (exploring how revisions to criminal procedure rules in the latter half of the twentieth century liberalized and deterritorialized the venue requirement). State criminal procedure rules mirror this requirement, usually mandating prosecution in the county or district where a crime occurs.99With exceptions for quirky scenarios like crimes committed on a county line. See, e.g., Ala. Code §§ 15-2-2, 15-2-7 (LexisNexis 2018); Alaska R. Crim. P. 18(b); Ariz. Rev. Stat. Ann. § 13-109 (2020); Cal. Penal Code § 777 (West 2020); Colo. Rev. Stat. § 18-1-202 (2021); Fla. Stat. Ann. § 910.03(1) (West 2014); Ga. Code Ann. § 17-2-2 (2020); Idaho Code § 19-2120 (2017). This list could go on and would include nearly every state, with variations in the precise language used to limit venue to some territorially defined area. This summary glosses over some important variation—for instance, states use different words to limit venue, which, as the “vicinage” debate suggests,100See supra Section I.B. reflects competing views about the values protected by criminal procedure. Technically, state venue rules fall along a spectrum from more to less strictly territorial. But they all embrace a baseline territorialism in which criminal procedure hinges on the site of a crime.

Criminal law enforcement is organized around territory as well. Typically, police are clumped into territorial units, such as the New York or Los Angeles Police Departments. (The Bergen County Sheriff’s Office, the New Jersey State Police, and the Federal Bureau of Investigation also work as examples here. All of these law enforcement agencies are defined by geography and empowered to act within some legally bounded area.) The same is true of state prisons and county jails, which have names like Ohio State Penitentiary and Cook County Jail and receive funding through county and state corrections budgets.101See, e.g., Maggie West, Ohio Dep’t of Rehab. & Corr., Greenbook: LBO Analysis of Enacted Budget 1 (2019), https://www.lsc.ohio.gov/documents/budget/133/MainOperating/greenbook/DRC.PDF [perma.cc/D9WX-HMVB] (outlining the Ohio state prison budget). Each of these law enforcement units purports to act on behalf of a certain place.

The United States also has a complex interstate extradition apparatus, which developed from the Extradition Clause.102 . U.S. Const. art. IV, § 2, cl. 2. Before a person can be sent to face criminal charges in another state, he must go through an elaborate legal process that involves written notices, communication between state governors, approval from secretaries of state, and hearings before an Article III court.103For example, a chart on Pennsylvania’s government website depicts the state’s twenty-two-step extradition process. Extradition and Interstate Rendition, Pa. Off. of Gen. Couns., https://www.ogc.pa.gov/Extradition/Pages/default.aspx [perma.cc/5PA5-S35G]. In Pennsylvania, for example, interstate extradition requires two warrants, three hearings, and twenty-two distinct steps.104Id. Defendants have a federal constitutional right to this process, which they can sue to enforce.105See, e.g., Crumley v. Snead, 620 F.2d 481, 483 (5th Cir. 1980) (“Almost 100 years ago . . . the Supreme Court recognized that individuals have a federal right to challenge their extradition by writ of habeas corpus. . . . Any denial of this right gives rise to a cause of action under 42 U.S.C. § 1983 . . . .”).

All of which is to say there is a considerable edifice built up around the idea that criminal law is territorial. American governments have organized their law enforcement bureaucracies, allocated resources, and recognized enforceable rights to effectuate the presumption that criminal law has territorial boundaries. Put simply, territorialism is shot through the entire system. In addition to being a foundational premise of substantive and constitutional law, the territoriality principle is a basic feature of how criminal law is operationalized in the United States.

And then there is the practice of punishing noncitizens, which is perhaps the clearest example of territorialism in American criminal law. The United States has always subjected noncitizens within its borders to its criminal laws. Foreign nationals have been prosecuted in state and federal criminal courts—and imprisoned in state and federal prisons—for as long as those institutions have existed.106See Emma Kaufman, Segregation by Citizenship, 132 Harv. L. Rev. 1379, 1388–94 (2019). The assumption behind this practice is that criminal law attaches as soon as a person enters the country whether or not she is a legal member of the polity. This assumption is usually implicit, though when courts need a citation they turn to the discussion of states’ police power in Mayor of New York v. Miln, in which the Supreme Court observed that “[t]he right to punish, or to prevent crime, does in no degree depend upon the citizenship of the party who is obnoxious to the law.”10736 U.S. (11 Pet.) 102, 140 (1837), overruled in part by Edwards v. California, 314 U.S. 160 (1941); see also, e.g., In re Manuel P., 263 Cal. Rptr. 447, 458–59 (Ct. App. 1989) (citing Miln in a juvenile delinquency case involving an undocumented minor from Tijuana). Rather, the Miln Court explained, a state has criminal jurisdiction “over all persons . . . within its territorial limits.”108Miln, 36 U.S. at 139. Thus, the “alien who shall just have set his foot upon the soil of the state, is just as subject to the operation of the law, as one who is a native citizen.”109Id. at 140.

The idea that a person who is not a citizen can nonetheless be subjected to criminal law because the right to punish arises from “soil” is the essence of the territoriality principle. In a regime based on this principle, criminal law is egalitarian (because it is blind to personal traits like citizenship status) and state-building (because it depends on and reaffirms state borders). Later Parts of this Article discuss these features of territorial criminal law and consider whether they are desirable. The goal here is simply to observe that territorialism is embedded in the institutional design of American criminal justice. Though it often goes unremarked, the territoriality principle shapes everything from the organization and funding of police departments to the criminal prosecution of hundreds of thousands of noncitizens each year.110The number of noncitizens subject to criminal prosecution across all fifty states in a given year is difficult to determine because of imprecise data collection and inconsistent definitions of the term noncitizen. But to offer a rough sense of scale, California prosecutors filed between 4.5 and 8.2 million criminal cases each year from 2010 to 2020, and in 2019, somewhere between 14 and 19 percent of the state’s population was foreign born. See Jud. Council of Cal., 2020 Court Statistics Report 83 (2020), https://www.courts.ca.gov/documents/2020-Court-Statistics-Report.pdf [perma.cc/T9WM-G5D7] (counting caseloads); Kristin F. Butcher & Anne Morrison Piehl, Crime, Corrections, and California, Cal. Counts, Feb. 2008, at 1–2 (counting noncitizen prisoners); Joseph Hayes, Justin Goss, Heather Harris & Alexandria Gumbs, California’s Prison Population, Pub. Pol’y Inst. of Cal. (July 2019), https://www.ppic.org/publication/californias-prison-population [perma.cc/Y4SY-MJ8L] (same). In the federal system, prosecutors filed 57,822 criminal cases in 2020, and noncitizens made up approximately 16 percent of the federal prison population. See U.S. Dep’t of Justice, United States Attorneys’ Annual Statistical Report 4 (2020), https://www.justice.gov/usao/page/file/1390446/download [perma.cc/7QLQ-YAW3] (counting caseloads); E. Ann Carson, Bureau of Just. Stat., U.S. Dep’t of Just., Prisoners in 2020—Statistical Tables (2021), https://bjs.ojp.gov/content/pub/pdf/p20st.pdf [perma.cc/47HA-36J6]. These figures suggest that huge numbers of noncitizens are subjected to state and federal criminal laws each year.

D. Rhetoric

Territoriality is also a prominent theme in constitutional case law. Constitutional opinions are full of rhetoric about the inherently local nature of criminal law. In the Rehnquist era, the Supreme Court invoked the localness of criminal law to restrict congressional power under the Commerce Clause.111United States v. Morrison, 529 U.S. 598, 617–18 (2000) (“The Constitution requires a distinction between what is truly national and what is truly local. . . . [W]e can think of no better example of the police power . . . than the suppression of violent crime . . . .”); United States v. Lopez, 514 U.S. 549, 561 n.3 (1995); see also United States v. Lamont, 330 F.3d 1249, 1252 (9th Cir. 2003) (referring to crime as “truly local” and noting that the “Supreme Court has recently spoken with unusual force regarding the need to reserve to the states the exercise of the police power in traditional criminal cases”); cf. Bond v. United States, 572 U.S. 844, 858 (2014) (“Perhaps the clearest example of traditional state authority is the punishment of local criminal activity.”). In abstention and habeas cases, the Court has cited the local nature of criminal law to explain why federal courts ought not police state criminal proceedings.112On abstention, see, for example, Younger v. Harris, 401 U.S. 37, 43–44 (1971) (announcing the doctrine), and In re Gruntz, 202 F.3d 1074, 1084, 1087 (9th Cir. 2000) (in a bankruptcy proceeding, citing Younger for the “fundamental policy against federal interference with state criminal prosecutions” and stating that “[t]he right to formulate and enforce penal sanctions is an important aspect of [state] sovereignty” (first alteration in original) (first quoting Younger, 401 U.S. at 46; then quoting Kelly v. Robinson, 479 U.S. 36, 47 (1986))). On habeas, see, for example, McCleskey v. Zant, 499 U.S. 467, 491 (1991) (asserting that “[re]examination of state convictions on federal habeas ‘frustrate[s] . . . the States’ sovereign power to punish offenders’ ” (quoting Murray v. Carrier, 477 U.S. 478, 487 (1986))), and Engle v. Isaac, 456 U.S. 107, 128 (1982) (identifying the Great Writ’s costs “on our federal system,” including incursion on states’ “primary authority for defining and enforcing the criminal law”). Claims about criminal law’s localness also surface in Eighth113See, e.g., Harmelin v. Michigan, 501 U.S. 957, 999–1000 (1991) (Kennedy, J., concurring) (“[M]arked divergences both in underlying theories of sentencing and in the length of prescribed prison terms are . . . inevitable [and] often beneficial . . . . [D]iffering attitudes and perceptions of local conditions may yield different, yet rational, conclusions regarding the appropriate length of prison terms for particular crimes.”); Cocio v. Bramlett, 872 F.2d 889, 889, 891 (9th Cir. 1989) (upholding as consistent with the Eighth Amendment a life sentence without parole for a drunk driving accident that resulted in one death, asserting “the right of a state under its police power to determine the proper prison sentence that should be imposed within its borders”). and Fourteenth114See, e.g., Medina v. California, 505 U.S. 437, 445–46, 452 (1992) (rejecting a due process challenge to a California law that placed the burden to prove incompetence on the defendant on the ground that “the criminal process is grounded in centuries of common-law tradition”). Amendment cases, in which the Supreme Court has upheld harsh sentences and prosecution-friendly procedural rules on the ground that criminal law is local and “belongs exclusively”115Knapp v. Schweitzer, 357 U.S. 371, 376 (1958). to states.

There is enormous nuance in these lines of doctrine. Each forms its own field of study. But grouping them together demonstrates just how much work rhetoric about the localness of criminal law is doing in constitutional jurisprudence. The idea that criminal law is local—naturally, fundamentally, traditionally, essentially—runs throughout constitutional thought. This idea has been cited to support a state-centric conception of American federalism, a restrictive view of criminal defendants’ rights, and limits on the Article III docket (which in turn maintain the prestige of federal courts).116See supra notes 111–114. On the relationship between Article III jurisdiction and federal court prestige, see Judith Resnik, Trial as Error, Jurisdiction as Injury: Transforming the Meaning of Article III, 113 Harv. L. Rev. 924, 930 (2000). The localness of criminal law is, in other words, part of the foundation on which American constitutionalism is built.

Courts are not especially clear about what it means for criminal law to be inherently local. Usually, the claim simply seems to mean that the federal government should let states manage an issue. But as the previous Section explained, the assertion that criminal law is local has a rich history: it refers to an eighteenth-century conception of criminal jurisdiction in which the power to define crimes and enforce criminal laws was limited to a particular territory.117Kershen I, supra note 57, at 811 (citing Justin Miller, Handbook of Criminal Law §§ 178–182 (1934)). At the founding, “crimes [were] considered ‘local’ in nature, i.e., local to the territory of the enacting sovereign and local to the territory of the enforcing court.”118Id. Modern references to the “essential” and “traditional” localness of criminal law call forth this early American understanding of criminal jurisdiction, which was premised on the territoriality principle. Thus, when courts say criminal law is local, they are in an important sense saying that criminal law is territorial.

This is yet another way in which territorialism lurks beneath American law. The constitutional cases described above rarely include discussions of the proper scope of criminal jurisdiction. These are cases about the Commerce Clause, due process, and debt collection.119See supra notes 111–114. But they rely on rhetoric about the territoriality of criminal law and indeed gain some of their force from the looseness of that rhetoric. As we will see in Part II, the claim that criminal law is quintessentially local becomes less persuasive the more one examines it. But as a truism, this claim functions powerfully to support a slew of different constitutional doctrines. The territoriality principle is prevalent and potent in the background of constitutional law as an unquestioned “maxim of American jurisprudence.”120 . Model Penal Code § 1.03 explanatory note (Am. L. Inst., Proposed Official Draft 1962). Its presence in so many corners of constitutional doctrine helps to explain why it seems so strange to think that New York could criminalize gambling in Nevada or authorize Missouri to prosecute violations of the New York criminal code.

II. The Erosion of the Territoriality Principle

After reading Part I, the territoriality principle should seem central to domestic criminal law, perhaps even required by the Constitution. Many aspects of the American legal system, from the basic definition of crime to the organization of criminal justice institutions, work together to make criminal law feel naturally and necessarily territorial. Yet, over the past century, criminal law’s commitment to the territoriality principle has faltered.

This Part traces the trend away from territoriality in domestic criminal law. It depicts a very different criminal justice system than we saw in Part I. The legal regime outlined above was defined by geographic borders and was particularly committed to state lines. The one below is flexible, negotiable, and remarkably unconcerned about territorial limits on the police power.

To trace this system’s emergence, this Part proceeds chronologically through the criminal legal process, from criminalization to policing, prosecution, and then punishment. In some ways, this approach is atypical. To the extent that legal academics discuss criminal jurisdiction, they tend to focus exclusively on criminalization and specifically on whether legislatures can criminalize extraterritorial acts.121See, e.g., Duff, supra note 12; The Boundaries of the Criminal Law (R.A. Duff et al. eds., 2010); The Constitution of the Criminal Law (R.A. Duff et al. eds., 2013); Vincent Chiao, Criminal Law in the Age of the Administrative State (2019); Dubber, supra note 28; Ferzan, supra note 14; Zedner, supra note 12. This Part departs from that norm by examining both criminalization and criminal law enforcement. It thus explores jurisdiction not just in the narrow terms of legislative intent but—to borrow the Supreme Court’s phrase—in “its popular sense of authority to apply the law to the acts of men.”122Nielsen v. Oregon, 212 U.S. 315, 320 (1909).

I take this approach because “the criminal law” is more than the set of liability rules created by legislatures. One of the core insights of criminal law scholarship of the past thirty years is that, given the extraordinary scope of the criminal code, the real substance of criminal law is made by law enforcers—by the police and prosecutors who decide which offenses are actually criminal.123William J. Stuntz, The Pathological Politics of Criminal Law, 100 Mich. L. Rev. 505, 506 (2001) (“[P]rosecutors . . . are the criminal justice system’s real lawmakers.”). From this perspective, it misses the point to study only criminalization. To know if criminal law is territorial, one has to ask not only what conduct legislatures decide to criminalize but also which crimes courts claim the power to adjudicate, when police can cross borders, where criminal prosecutions happen, and when punishment is tied to the place a crime occurred. The answer, it turns out, is that territorialism has eroded at every turn.

A. Criminalization

Criminal law begins when an authoritative legal body, typically a legislature,124See Kadish et al., supra note 29, at 162 (stating that “nearly all American jurisdictions . . . have now abolished [by statute] the common-law doctrine that courts can create new crimes,” though “[t]he doctrine still survives in a few states” and the Supreme Court “has never held it unconstitutional for state judges to create new common-law crimes”). decides that some conduct is criminal. At this stage of the criminal legal process, the key question is how far a criminal law reaches. Theorists call this the question of criminal law’s “geographical ambit.”125Farmer, supra note 14, at 231; see also Hirst, supra note 38, at 11. (So to review: venue is where a criminal trial happens, vicinage is where the jury comes from, and ambit is where the substantive criminal law applies.) As Part I noted, the traditional view is that criminal statutes reach offenses within some defined territory, like New Jersey or Pennsylvania.126See supra notes 39–42 (discussing the MPC provision on “territorial applicability” and the presumption against extraterritorial application of state criminal law). But legislatures and courts have expanded criminal law beyond these boundaries in several ways.

First, they have redefined crimes to stretch across multiple places. Take the continuing offense doctrine, under which crimes such as fraud and possession of contraband continue to occur over time “and, possibly, in a number of different places”127State v. Allah, 750 S.E.2d 903, 909 (N.C. Ct. App. 2013). as a person moves around. Or consider inchoate offenses like attempt and conspiracy, whose definitions and relationship to the actus reus requirement are notoriously fluid.128See George P. Fletcher, Basic Concepts of Criminal Law 191–93 (1998) (discussing the origins and modern doctrine of conspiracy); Kadish et al., supra note 29, at 651–56 (exploring the relationship between the actus reus requirement and attempt liability). Because continuing offenses travel with a person and the precise location of an inchoate offense is difficult to pin down, these crimes have an attenuated relationship to geography. (This problem is exacerbated when coconspirators can be charged for each other’s actions and tried wherever an accomplice acted.)129 . Kadish et al., supra note 29, at 745, 774 (describing “continuing controversy over Pinkerton”); see also United States v. Cabrales, 524 U.S. 1, 8 (1998) (analyzing Hyde v. United States, 225 U.S. 347 (1912), in which “[a]lthough none of the defendants had entered the District [of Columbia] as part of the conspiracy, venue was nevertheless appropriate . . . based on the overt acts of a co-conspirator there”). Note that the discussion of ambit is bleeding into venue here. See infra Section II.C (addressing venue). See generally Farmer, supra note 14, at 231–32 (arguing that the distinction between ambit and venue is “artificial and technical” but nonetheless useful when exploring the geographic scope of criminal law). Ongoing and inchoate crimes make up a significant portion of the modern criminal docket.130State-level data collection and reporting varies, but for a rough sense of prevalence: in 2020, a continuing or inchoate offense was the top charge in 15 percent of New York criminal arraignments, and 26.5 percent of New York state prisoners had a continuing or inchoate offense as their top criminal charge. Div. of Tech. & Ct. Rsch., OCA-STAT Act Report, NYCourts, https://ww2.nycourts.gov/oca-stat-act-31371 (June 15, 2021) (providing arraignment data); Inmates Under Custody: Beginning 2008, N.Y. State: OpenNY, https://data.ny.gov/w/55zc-sp6m/caer-yrtv?cur=65n9o7AxVKY&from=rsP3XmFxvCz (June 15, 2021) (providing data on New York’s prison population). Because these data list only the top charge in an arraignment or conviction, they likely undercount continuing and inchoate offenses. Often, they can be prosecuted in several different places.

Continuing and inchoate crimes are territorial insofar as a court must determine where they occurred before a criminal prosecution can proceed. Asking where a conspiracy or an ongoing theft took place is a territorial question. But in modern criminal law, the answer to that question is often a legal fiction based on artificial extensions of time and concepts like constructive presence.131See Larry Kramer, Comment, Jurisdiction over Interstate Felony Murder, 50 U. Chi. L. Rev. 1431, 1435 (1983). The approach to territoriality when conceptualizing these crimes is a far cry from the eighteenth-century understanding of criminal jurisdiction described in Part I.132See supra Section I.B. Ongoing and inchoate offenses are at best quasi-territorial—which is why, when legislatures invented continuing offenses in the early 1900s, defense attorneys argued that they violated the Venue Clause and the Sixth Amendment.133Kershen II, supra note 57, at 39–40 (citing cases challenging the Elkins Act of 1903, in which “Congress introduced the concept of a continuing offense”). That argument resurfaced in the 1970s, when critics protested that the very “idea of a continuing offense must be declared unconstitutional.”134Id. at 159. This objection has never been successful, but its persistence illustrates just how much ongoing crimes undermine the territoriality principle.

Courts have also relaxed territorial rules by expanding their own jurisdiction over out-of-state conduct. Think here about a Texas criminal case in which prosecutors charged someone who forged a deed for Texas property while in Louisiana135Hanks v. State, 13 Tex. Ct. App. 289, 290–91 (1882). or a California court adjudicating the criminal prosecution of a Colorado-based doctor who prescribed pills that someone later ingested in California.136Hageseth v. Superior Ct., 59 Cal. Rptr. 3d 385, 400–01 (Ct. App. 2007). In these sorts of cases, criminal prosecutions are based on the in-state effects of out-of-state activity. American courts first began to recognize this species of criminal jurisdiction in the 1860s,137See Simpson v. State, 17 S.E. 984, 985–86 (Ga. 1893) (finding Georgia jurisdiction over a defendant who fired a fatal shot from across the border in South Carolina); Commonwealth v. Macloon, 101 Mass. 1, 6 (1869) (observing “[t]he general principle, that a man who does a criminal act in one county or state may be held liable for its continuous operation in another”); Commonwealth v. Smith, 93 Mass. (11 Allen) 243, 259 (1865) (establishing Massachusetts criminal jurisdiction over perjury committed out of state). The Supreme Court then cited these state cases in Strassheim v. Daily, 221 U.S. 280, 285 (1911). and the Supreme Court upheld effects-based jurisdiction in 1911.138Strassheim, 221 U.S. at 285. In embracing the effects doctrine, Justice Holmes explained that a “civilized world” required a flexible approach to the territoriality principle, so that even a criminal who had “never had set foot in the State” could be punished if his conduct caused in-state harm.139Id. at 284–85 (“If a jury should believe the evidence . . . the usage of the civilized world would warrant Michigan in punishing him, although he had never set foot in the State until after the fraud was complete.”). To put this precedent into some historical context: at the turn of the twentieth century, as the federal government began to take shape and new forms of transportation enabled mass mobility,140See generally Sarah A. Seo, Policing the Open Road (2019). courts grew bolder about their power over offenses committed outside the state.

The Constitution was no barrier to this development. As Part I explained, the Due Process Clause is supposed to prevent states from “overreaching” when they exercise criminal jurisdiction over extraterritorial activity.141See supra Section I.B (citing People v. Puig, 378 N.Y.S.2d 925, 935 (Sup. Ct. 1976), in which a New York court observed that criminal jurisdiction under the relevant statute was limited to “out-of-state offenses which by their nature produce palpably harmful consequences which are of necessity local and peculiarly injurious to the rights of [the] state or its citizens,” not out-of-state “offenses pertaining to the general community welfare”). But one can search in vain for a meaningful due process limit on criminal prosecution of out-of-state conduct. Even the New York court that boldly denounced jurisdictional “overreaching” was eventually overturned.142People v. Kassebaum, 696 N.Y.S.2d 23, 24 (App. Div. 1999) (rejecting the jurisdictional limitation in Puig), aff’d, 744 N.E.2d 694 (N.Y. 2001). In practice, any effect inside a state—harm to a state resident, a piece of state property, and so on—can ground a state criminal prosecution.143For one case grounding criminal jurisdiction on the in-state effects of extraterritorial activity, see infra notes 144–147 and accompanying text. For more, see, for example, In re Vasquez, 705 N.E.2d 606, 610–12 (Mass. 1999); State v. Kane, 625 A.2d 1361, 1363 (R.I. 1993); Roberts v. State, 619 S.W.2d 161, 164 (Tex. Crim. App. 1981); and State v. Doyen, 676 A.2d 345, 349 (Vt. 1996). Again, this approach to criminal law is basically territorial in that courts must conjure up some connection to state soil to proceed. But the invention and expansion of effects-based jurisdiction loosened nineteenth-century assumptions about the limits of judicial power over criminal cases.

Courts have also developed nonterritorial theories of criminal jurisdiction. The most significant case in this category is Skiriotes v. Florida, a 1941 Supreme Court precedent.144313 U.S. 69, 79 (1941). Skiriotes arose when a Florida resident named Lambiris Skiriotes went sponge fishing in the Gulf of Mexico.145Brief of Appellant at 7, Skiriotes, 313 U.S. 69 (No. 658). The Sheriff of Pinellas County arrested Skiriotes under a state law that banned sponge fishing in Florida; he was then prosecuted and convicted under state law.146Id. at 7, 10. The Supreme Court upheld the conviction on the ground that Florida could “govern the conduct of its citizens upon the high seas with respect to matters in which the State has a legitimate interest”—specifically, Florida’s “interest in the proper maintenance of [a] sponge fishery.”147Skiriotes, 313 U.S. at 75–77. The Court thus concluded that Florida could predicate its criminal laws on state citizenship rather than state territory.

Skiriotes represents a remarkable shift away from the territoriality principle. The Supreme Court has never clarified the case’s outer boundaries, and because the defendant was in international waters, it is not clear what would happen if Florida criminalized its residents’ behavior in Georgia or Alabama.148Conflicts scholars considered these questions in the 1990s after the Supreme Court granted certiorari in Planned Parenthood of Southeastern Pennsylvania v. Casey, 505 U.S. 833 (1992). The prospect that abortion “would be remitted entirely to the political process” prompted academics to examine whether states could prohibit their residents from traveling to other states to obtain abortions. Seth F. Kreimer, “But Whoever Treasures Freedom . . .”: The Right to Travel and Extraterritorial Abortions, 91 Mich. L. Rev. 907, 907 (1993); see also Lea Brilmayer, Interstate Preemption: The Right to Travel, the Right to Life, and the Right to Die, 91 Mich. L. Rev. 873, 880 (1993). More recently, scholars have explored a similar question with respect to marijuana laws. See, e.g., Mark D. Rosen, Marijuana, State Extraterritoriality, and Congress, 58 B.C. L. Rev. 1013, 1015 (2017). While scholars have opined that such a statute could implicate “multiple complicated doctrines”—including due process, the Sixth Amendment, the Dormant Commerce Clause, and the right to travel149Rosen, supra note 148, at 1015 (listing doctrines implicated by extraterritorial state regulation, though not focusing on criminal statutes); cf. Sachs, supra note 31, at 1734 (discussing civil personal jurisdiction).Skiriotes established that criminal law can attach by virtue of citizenship rather than soil. If there is a limit on the ambit of criminal law, it will have to come from as-yet-unannounced constitutional rules, not a requirement that criminal law must be territorial.
Skiriotes is the most prominent example of deterritorialized domestic criminal law. But the basic proposition in the case is that state criminal law can proceed without a clear territorial hook. This precept guides more run-of-the-mill cases too. Sometimes, for example, courts ground criminal jurisdiction on a state’s general interests in some industry or value rather than on the effects of a crime. In 2017, Florida invoked its interest in the cruise ship industry to try an attempted sexual assault between two noncitizens at sea.150Paul v. State, 233 So. 3d 1181, 1183 (Fla. Dist. Ct. App. 2017). Alaska has succeeded with similar prosecutions undertaken in the name of the state’s “tourism industry.”151State v. Jack, 125 P.3d 311, 322 (Alaska 2005) (justifying the criminal prosecution of a sexual assault on an offshore ferry on the ground that “ferries are important to the tourism industry . . . and [Alaska’s] economy.”). These sorts of cases extend the theory of effects-based criminal jurisdiction to include abstract harms to a state’s economic and political order. At that point, territoriality has more or less disappeared.

Courts have also extended the scope of state criminal law by relaxing the rules around extradition. Part I outlined the development of an elaborate legal apparatus to facilitate interstate extradition.152See supra Section I.C. Initially, state governments used that apparatus to regulate and denounce each other’s criminal legal systems. In 1860, Ohio’s governor refused to extradite a man who helped to free a slave in Kentucky.153Kentucky v. Dennison, 65 U.S. (23 How.) 66, 66–67 (1861), overruled by Puerto Rico v. Branstad, 483 U.S. 219, 224 (1987) (overruling Dennison’s holding that the federal court could not compel compliance with the Extradition Clause). In the 1930s, New Jersey’s governor declined to extradite Robert Elliott Burns after reading his memoir, I Am a Fugitive from a Georgia Chain Gang!.154Kenyon Bunch & Richard J. Hardy, Continuity or Change in Interstate Extradition? Assessing Puerto Rico v. Branstad, Publius, Winter 1991, at 51, 55. In 1975, the governor of Tennessee ignored Oklahoma’s request to extradite country singer Faron Young, who was “wanted on a morals charge.”155Id. at 55 & n. 22 (citing this and other “historical examples of governors refusing to grant extradition”); see also Extradition Refused, Longview News-J., Dec. 4, 1975, at 7–8. Until the 1970s, state courts policed extradition too, denying requests when they felt a sister state had failed to show probable cause.156Michigan v. Doran, 439 U.S. 282, 289–90 (1978) (overruling the Michigan Supreme Court’s holding that Arizona’s probable cause determination was deficient); see also Wellington v. South Dakota, 413 F. Supp. 151, 154 (D.S.D. 1976) (denying extradition on the ground that Minnesota had failed to establish probable cause “to the satisfaction of the State court”); Ierardi v. Gunter, 528 F.2d 929, 932 (1st Cir. 1976) (drawing the same conclusion in a case involving extradition to Florida from Massachusetts). The Supreme Court put a stop to that practice in 1978 when it held that state courts may not opine on the sufficiency of extradition requests.157Doran, 439 U.S. at 289. Instead, the Court emphasized, judicial approval of extradition is “summary and mandatory.”158Id. at 288–89. The Supreme Court’s mandatory approach to extradition is particularly important because the Court has “interpreted the crimes for which a person is subject to extradition very broadly, to include every offense punishable by the law of the [requesting] state.” Gross & Upham, supra note 89.

This development in extradition doctrine undermined the territoriality principle. As Part I pointed out, extradition is quintessentially territorial; it is the practice of sending fugitives back to the “right” state to be tried. Extradition is criminal law’s substitute for choice of law and in that respect is the lynchpin of territorialism. But making extradition mandatory weakens the territoriality principle, because it means that the borders of state criminal law matter less. Imagine two countries: one that refuses to extradite a criminal and one where extradition is routine. Territorial borders are more meaningful in the first country—the one where you are safe once you touch its soil. The scope of criminal law feels more territorial there too, since the offended country’s laws cannot reach you.

In other words, denying extradition is one way to make clear that criminal law has geographic limits. By requiring “summary and mandatory” extradition between states, the Supreme Court has ensured that Arizona’s criminal laws reach New Hampshire, and vice versa. The Court made this choice explicitly. In Michigan v. Doran, a leading extradition case, Chief Justice Burger explained that mandatory extradition “foster[s] national unity” by preventing any state from becoming “a sanctuary for fugitives” and “ ‘balkaniz[ing]’ the administration of criminal justice.”159Doran, 439 U.S. at 287–88. Writing for the Court, Chief Justice Burger articulated a national vision of criminal law: “In the administration of justice, no less than in trade and commerce,” unity is “served by de-emphasizing state lines.”160Id. at 288. Modern extradition doctrine thus extends the reach of state criminal laws to the edges of the nation. As a practical matter, summary extradition means that it is not particularly difficult to prosecute someone who flees across a state border.161This doctrine may become a flashpoint now that the Supreme Court overturned Roe v. Wade. Dobbs v. Jackson Women’s Health Org., 142 S. Ct. 2228 (2022); see, e.g., David S. Cohen, Greer Donley & Rachel Rebouché, The New Abortion Battleground, 123 Colum. L. Rev. (forthcoming 2023) (arguing that abortion-supportive states should exempt abortion providers from state extradition law).

There are important technical distinctions between the examples in this Section. Conceptually, there is a difference between a legislature’s power to deem conduct criminal (legislative jurisdiction) and a court’s power to adjudicate criminal cases (judicial jurisdiction). The former power is what the term “criminalization” calls to mind, while the exercise of judicial jurisdiction is rarely described as an instance of criminalization.162But see Adventure Commc’ns, Inc. v. Ky. Registry of Election Fin., 191 F.3d 429, 435–36 (4th Cir. 1999) (noting that although legislative and judicial jurisdiction are distinct, “the concepts are closely related” in that both are governed “by the due process clause” and the standards for evaluating each kind of jurisdiction “substantial[ly] overlap”). Both types of jurisdiction are different from questions about the functional reach of Maine or Texas criminal law given modern extradition rules.

In a course on criminal jurisdiction, one would parse these distinctions. But here the salient observation is that a range of doctrinal shifts have come together to extend the reach of domestic criminal law. When legislatures define crimes to cover extraterritorial conduct and courts exert jurisdiction over out-of-state activities, both acts expand state criminal law beyond its traditional boundaries. The result is a corpus of criminal law that is detached from territorial borders—not wholly, but more than one might expect.

B. Policing

The deterritorialization of American criminal law is even more pronounced in law enforcement. As Part I noted, police are typically organized into territorial units. Police officers wear badges engraved with city insignia163Ray Rivera, The Officer Is Real; The Badge May Be an Imposter, N.Y. Times (Nov. 30, 2009), https://www.nytimes.com/2009/12/01/nyregion/01badge.html [perma.cc/8QQM-EZ8P]. and swear oaths to state constitutions.164See, e.g., N.Y. Civ. Serv. Law § 62 (McKinney 2011); see also Personnel Service Bulletin from William J. Diamond (Mar. 21, 1997), https://www1.nyc.gov/assets/dcas/downloads/pdf/reports/100_6.pdf [perma.cc/4YRT-EXRV]. Yet, in practice, cooperative agreements and permissive doctrines mean policing is only loosely tied to boundary lines.

Sometimes the deterritorialization of policing is formal, as when police departments sign agreements to provide “mutual aid.”165See, e.g., Mutual Police Assistance Compact Signed by Stonington, Conn. (July 14, 2017) (on file with the Michigan Law Review). I obtained this agreement and many other police policy documents through a series of open records requests under state law. Many states permit local law enforcement agencies to enter partnerships under which they can police each other’s territory.166See, e.g., Fla. Stat. Ann. § 23.1225(1)(a) (West 2021); Mass. Ann. Laws ch. 40, § 4J (LexisNexis 2018); N.H. Rev. Stat. Ann. § 48:11-a (2012); W. Va. Code Ann. § 15-10-4(a) (LexisNexis 2019). See generally John M. Baines, Robert Heggestad, Dale A. Kelley & Charles P. White, Mutual Aid Planning 8–9 (1973) (providing a national overview of mutual aid statutes). These mutual aid agreements, which are essentially private contracts between police departments, create standing commitments to share resources and jointly enforce criminal laws.167See e.g., Ball v. City of Coral Gables, 301 F. App’x 865, 867 (11th Cir. 2008); Commonwealth v. Pike, 41 N.E.3d 332, 2015 WL 8285092, at *2 (Mass. App. Ct. 2015) (unpublished table decision) (reading a mutual aid agreement to grant a police officer “the authority to effectuate an extraterritorial stop”). Police departments also enter into memoranda of understanding (MOUs) and more informal agreements to combine their powers.168See, e.g., Memorandum of Agreement Between the Town of Glastonbury and the Town of East Hampton for Public Safety Dispatch Services (May 17, 2016) [hereinafter East Hampton MOU] (on file with the Michigan Law Review). On the practice of “pooling” of statutory powers through interagency cooperation, see Daphna Renan, Pooling Powers, 115 Colum. L. Rev. 211 (2015). These contracts expand the footprint of local police departments—or, as one 1988 white paper put it, mutual aid agreements “knit jurisdictions together.”169 . Advisory Comm’n on Intergovernmental Rels., Metropolitan Organization: The St. Louis Case, at iii (1988), https://library.unt.edu/gpo/acir/Reports/information/m-158.pdf [perma.cc/G9RA-8X7A].

The standard mutual aid agreement merges the staff and resources of two small, neighboring police departments within a single state.170See, e.g., East Hampton MOU, supra note 168; Non-Emergency Interagency Agreement Between Darien, Greenwich, New Canaan & Stamford Police Departments (Nov. 30, 2012) (on file with the Michigan Law Review). But mutual aid compacts can extend “local” policing much further. In some states, municipal police can contract to provide services outside the state. In Nebraska, for example, “[a]ny city or village” can authorize its police to assist another state’s police force.171 . Neb. Rev. Stat. § 18-1706 (2021) (“Any city or village may by resolution authorize its . . . police department or any portion thereof to provide . . . police[] and emergency service outside of the limits of such city or village either within or without the state.”). Local police departments can also join multijurisdictional task forces that span state boundaries.172See, e.g., Four Arrests Linked to 24 Robberies of Elderly, ABC7 (Mar. 21, 2008) [hereinafter Four Arrests], https://abc7chicago.com/archive/6035045 [perma.cc/E39U-Y5Z5] (describing a task force between Chicago and Indiana police); see also David W. Hayeslip & Malcolm L. Russell-Einhorn, Evaluation of Multi-Jurisdictional Task Forces Project 10 (2002), https://www.ojp.gov/pdffiles1/nij/grants/200904.pdf [perma.cc/2HTM-N289] (describing a federal grant program created in 1988 to encourage multijurisdictional task forces “to target gangs, illegal firearms, specific crimes and other cross-jurisdictional crime-related problems”). These task forces are designed to overcome territorial jurisdictional limits. They are a mechanism to blur legal boundaries so that “city and suburban cops in Illinois and Indiana” can “work[] the case together.”173Four Arrests, supra note 172 (explaining that a Chicago-area task force developed after police “noticed that our . . . criminals utilized our borders to commit crimes . . . in Chicago and flee to the neighboring suburbs and into Indiana”).

At a high level of generality, cooperative policing is nothing new. American police departments have always shared equipment and “backup forces,” particularly during emergencies.174 . Baines et al., supra note 166, at 1 (tracing “the origins and development of law enforcement mutual aid in the United States” (cleaned up)). But police sharing practices were interlocal and “strictly informal” until the 1930s, when increased mobility prompted a movement to expand and institutionalize interjurisdictional policing.175Id. In 1934, citing concerns about “the interstate nature of crime and the growing complexity of law enforcement,” Congress passed the Crime Prevention Compact Act (CPCA).176Id.; Crime Prevention Compact Act, ch. 406, 48 Stat. 909, 909 (1934) (codified as amended at 18 U.S.C. § 420). That statute authorized states to enter compacts “for cooperative effort and mutual assistance in the prevention of crime and in the enforcement of their respective criminal laws and policies.”177Crime Prevention Compact Act, 48 Stat. at 909. The CPCA encouraged states to establish “such agencies, joint or otherwise, as they may deem desirable for making [cooperative policing] effective.”178Id.

A flurry of state laws followed. In the wake of the CPCA, states passed statutes enabling cooperative policing and establishing interjurisdictional criminal task forces.179In order of passage, see, for example, Act of June 30, 1936, No. 94, § 20, 1936 La. Acts 280, 285 (codified as amended at La. Stat. Ann. § 40:1391 (2016)); Act of Apr. 17, 1946, ch. 834, 1946 N.Y. Laws 1601, 1603 (codified in relevant part at N.Y. Gen. Mun. Law §§ 209-f to -g (McKinney 2016)); Act of June 11, 1947, No. 198, 1947 Ohio Laws 288 (codified as amended at Ohio Rev. Code Ann. § 737.04 (LexisNexis 2015)); and Interlocal Co-operation Act, ch. 100, 1957 Kan. Sess. Laws 255 (codified as amended at Kan. Stat. Ann. § 12-2904 (2021)). States also began to enter regional compacts that provided for joint policing and created centralized “criminal intelligence bureau[s]” to gather and disseminate information about crime.180See, e.g., Act Adopting the New England State Police Compact, No. 315, 1967 Conn. Pub. Acts 377 (codified as amended at Conn. Gen. Stat. § 29-162 (2021)) (forming the “New England State Police Compact”); Act of June 14, 1971, No. 159, § 3, 1971 La. Acts 467, 468 (repealed 2022) (forming the Southern State Police Compact). This cooperative model flourished during World War II as local law enforcement agencies coordinated their training programs and emergency plans for “civil defense.”181 . Baines et al., supra note 166, at 2 (“During World War II the Mutual Aid concept expanded . . . . Contingency plans were developed and training commenced . . . to enable law enforcement agencies to better cope with possible enemy attack or invasion.”).

Cooperative policing then expanded in the 1960s, when police turned to mutual aid agreements to manage civil unrest, Vietnam protests, and “rock festivals.”182Id. Between 1960 and 1972, twenty-five states and the District of Columbia passed laws enabling interjurisdictional policing.183In chronological order, see Act of Mar. 31, 1953, ch. 144, § 9, 1953 Alaska Sess. Laws ch. 144, § 9 (codified as amended at Alaska Stat. § 19.65.090 (2020)); Act of Apr. 3, 1962, ch. 623, § 15.1-131, 1962 Va. Acts 960, 991 (codified as amended at Va. Code Ann. § 15.2-1724 (Supp. 2021)); Act of Apr. 15, 1963, ch. 207, § 1, 1963 Colo. Sess. Laws 207, 207 (codified as amended at Colo. Rev. Stat. § 29-5-103 (2021)); Act of Apr. 28, 1965, ch. 382, 1965 Mass. Acts 204 (codified as amended at Mass. Gen. Laws ch. 41, § 99 (2020)); Act of May 17, 1965, ch. 106, 1965 R.I. Acts & Resolves 381 (codified at 42 R.I. Gen. Laws § 42-37-1 (2007)); Act of Sept. 3, 1965, ch. 435, 1965 Me. Laws 591 (codified at Me. Rev. Stat. Ann. tit. 25, § 1665 (2007)); Act of Jan. 1, 1966, ch. 722, 1965 Tex. Gen. Laws 362, 362–63 (codified as amended at Tex. Code Crim. Proc. Ann. art. 14.03 (West Supp. 2022)); Act of June 21, 1967, ch. 846, 1967 N.C. Sess. Laws 1090 (codified as amended at N.C. Gen. Stat. § 160A-288 (2021)); Act of July 10, 1967, No. 236, 1967 Mich. Pub. Acts 350 (codified as amended at Mich. Comp. Laws Serv. §§ 123.811–.814 (LexisNexis 2021)); Act of Aug. 8, 1967, ch. 105, 1967 Wis. Sess. Laws 445 (codified as amended at Wis. Stat. § 66.0313 (2022)); Act Concerning Mutual Police Assistance Among Municipalities, No. 198, 1967 Conn. Pub. Acts 255 (codified at Conn. Gen. Stat. § 7-277a (2021)); Act Adopting the New England State Police Compact, No. 315, 1967 Conn. Pub. Acts 377 (codified at Conn. Gen. State § 29-162 (2021)); Act of Mar. 15, 1968, No. 288, 1967 Vt. Acts & Resolves 153 (codified at Vt. Stat. Ann. tit. 20, §§ 1951–1959, 1971–1972 (2021)); Act of Mar. 21, 1968, ch. 68, 1968 Ky. Acts 193 (codified as amended at Ky. Rev. Stat. Ann. § 65.255 (LexisNexis 2014)); Act of Apr. 13, 1968, No. 1054, 1968 S.C. Acts 2521 (codified as amended at S.C. Code Ann. § 5-7-120 (2004)); Act of Aug. 5, 1968, No. 2157, 1968 Ill. Laws 26 (codified in relevant part at 65 Ill. Comp. Stat. Ann. 5 / 1-4-8 (West 2020)); Act of Aug. 12, 1968, ch. 1222, 1968 Cal. Stat. 2302 (codified as amended at Cal. Penal Code § 830.1 (West 2020)); Act of Oct. 17, 1968, Pub. L. No. 90-587, 82 Stat. 1150 (codified as amended at D.C. Code § 2-209.01 (2022)); Act of Apr. 17, 1969, ch. 177, 1969 Okla. Sess. Laws 227 (codified at Okla. Stat. Ann. tit. 11, § 34-103(a) (West 2022)); Act of May 14, 1969, ch. 596, 1969 Md. Laws 1398 (codified as amended at Md. Code Ann., Crim. Proc. § 2-105 (LexisNexis 2018)); Act of June 11, 1969, ch. 224, 1969 N.H. Laws 170 (codified at N.H. Rev. Stat. Ann. § 106-D:1 (2013)); Florida Mutual Aid Act, ch. 69-112, 1969 Fla. Laws 589 (codified as amended at Fla. Stat. Ann. §§ 23.12–.127 (West 2014)); Police Mutual Aid Agreement Act, ch. 433, 57 Del. Laws 1222 (1970) (codified as amended at Del. Code Ann. tit. 11, §§ 1941–1947 (2015)); Act of Apr. 12, 1971, ch. 51, § 8, 1971 Ariz. Sess. Laws 119, 125–38 (codified as amended at Ariz. Rev. Stat. Ann. § 26-309 (Supp. 2021)); Act of June 9, 1971, ch. 197, 1971 N.J. Laws 773, 787–89 (codified in relevant part at N.J. Stat. Ann. § 40A:14-156 (West 2019)); and Mutual Aid Act, ch. 153, § 3, 1971 N.M. Laws 440, 440 (codified at N.M. Stat. Ann. § 29-8-3 (2022)). By the 1980s, federal commissions were studying how best to coordinate local police forces,184See, e.g., Advisory Comm’n on Intergovernmental Rels., supra note 169, at 61–63. and the federal government was awarding grants to encourage multijurisdictional task forces.185 . Hayeslip & Russell-Einhorn, supra note 172, at 10; see also High Intensity Drug Trafficking Areas, Off. of Nat’l Drug Control Pol’y (2021), https://www.hidtaprogram.org/summary.php [perma.cc/EL3F-FN9K] (describing a program created by Congress in 1988 to “coordinate[] . . . Federal, State, Local, and Tribal law enforcement agencies . . . to address regional drug threats”). Federal funds for these local partnerships ballooned after September 11, 2001, as police associations argued that only an “interjurisdictional enforcement approach” could “address the threats of international and domestic terrorism.”186 . Bureau of Just. Assistance, U.S. Dep’t of Just., Mutual Aid: Multijurisdictional Partnerships for Meeting Regional Threats, at vii–1 (2005), https://www.ojp.gov/pdffiles1/bja/210679.pdf [perma.cc/YH6Z-JMQH].

The rise of cooperative policing deserves its own separate study. The birth and expansion of mutual aid agreements is a fascinating case of subfederal coordination and bureaucratic state-building. This model of policing provides a counternarrative to the dominant account of twentieth-century criminal law, which emphasizes the proliferation of federal criminal laws and the emergence of a federal enforcement bureaucracy that (the story goes) displaced state and local police.187See Daniel C. Richman & Sarah Seo, Driving Toward Autonomy? The FBI in the Federal System, 1908–1960, at 3 (Univ. of Iowa, Legal Stud. Rsch. Paper No. 2019-22, 2019), https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3415103 [perma.cc/C6XS-6EP6] (identifying and critiquing this federalization narrative); see also Rachel E. Barkow, Federalism and Criminal Law: What the Feds Can Learn from the States, 109 Mich. L. Rev. 519, 523–29 (2011) (summarizing “the existing debate over the federalization of crime”). Mutual aid agreements complicate this federalization story. They show that criminal law expanded during the twentieth century not just because the federal government grew but also because—in the shadow of a growing federal government, and perhaps in resistance to it—states and localities formed regional unions that detached criminal law from territorial borders and upended nineteenth-century ideas about criminal jurisdiction.

This history is nuanced, but even the quick version illustrates that modern policing is significantly less tied to jurisdictional boundaries than the phrase “local police” suggests. Over the course of the twentieth century, contracts to provide emergency aid became broader agreements to combine both the resources and the legal authority of local police forces. Today, mutual aid agreements are routine.

Sometimes, moreover, agreements are not even necessary. Although compacts and task forces are particularly clear examples of interjurisdictional policing, local law enforcement is deterritorialized in other ways as well. Some jurisdictions, for example, empower nominally local police to act anywhere in the state.188E.g., Cal. Penal Code § 830.1 (West 2020) (outlining when the authority of municipal police officers “extends to any place in the state”); Conn. Gen. Stat. § 7-281 (2021) (authorizing local police to execute warrants “in any part of the state”); see also Armendariz v. State, 123 S.W.3d 401, 404–05 (Tex. Crim. App. 2003) (construing Texas’s criminal code, which authorizes local “peace officers” to make arrests when they observe illegal conduct, to permit an arrest outside “the Odessa police officers’ geographic jurisdiction”). In California, a peace officer’s authority extends to any offense he witnesses;189 . Cal. Penal Code § 830.1(a)(3) (West 2020). in Connecticut, “active members of any . . . police force in a town, city, or borough” can execute warrants statewide.190 . Conn. Gen. Stat. § 7-281 (2021). Texas permits out-of-state police officers to enter its territory to make arrests for felonies committed elsewhere.191 . Tex. Code Crim. Proc. Ann. art. 14.051 (West 2015). Tennessee has a “special enforcement unit” focused on Medicaid fraud, whose police routinely arrest people “in Alabama, Mississippi, and Georgia.”192Amy Yurkanin, The TennCare Trap: How One State’s War on Medicaid Fraud Ensnares Working Moms in Alabama, AL.com (Apr. 28, 2021, 2:31 PM), https://www.al.com/news/2021/04/the-tennessee-trap-how-one-states-war-on-medicaid-fraud-ensnares-working-moms-in-alabama.html [perma.cc/9DAJ-7S97]. These practices abrade territorial jurisdictional boundaries.

Most states also have “citizen’s arrest” statutes that allow private citizens to apprehend alleged criminals.193See Citizen’s Arrest Laws by State, Sols. Inst., https://solutions-institute.org/tools/citizens-arrest-laws-by-state [perma.cc/NU3V-4MS6]. Courts have relied on these statutes, which are controversial relics of the Reconstruction era,194See Kemp Hails Passage of Legislation to Repeal Citizen’s Arrest Law in Georgia, WRDW (Apr. 1, 2021, 11:45 AM), https://www.wrdw.com/2021/04/01/kemp-hails-passage-of-legislation-to-repeal-citizens-arrest-law-in-georgia [perma.cc/93PB-BU6E] (describing the history of Georgia’s citizen’s arrest statute, which was amended after the murder of Ahmaud Arbery). to uphold arrests made by police officers outside of their territorial jurisdiction.195See, e.g., State ex rel. State v. Gustke, 516 S.E.2d 283, 290–91 (W. Va. 1999) (upholding such an arrest and citing cases from eighteen states in which courts had done the same); People v. Lahr, 566 N.E.2d 12, 13 (Ill. App. Ct. 1991) (“An extensive line of cases in Illinois has upheld the validity of extraterritorial arrests made by police officers who, lacking official authority, were found to have been authorized to make ‘citizen’s arrests.’ ”). Some states even relax “the requirements for a citizen’s arrest” when the citizen in question is a police officer—a doctrine that turns police into sort of roving super-citizens.196E.g., Commonwealth v. Claiborne, 667 N.E.2d 873, 876 (Mass. 1996). Courts use “hot pursuit” doctrines to a similar effect.197See, e.g., Commonwealth v. Peters, 965 A.2d 222, 225 (Pa. 2009) (“[W]e note that the [hot pursuit statute] is to be construed liberally to give effect to its purposes.”). These doctrines create exceptions to “the territorial limits of . . . [police] jurisdiction” and expand the scope of criminal law enforcement.198Id. at 223 (citing 42 Pa. Stat. and Cons. Stat. Ann. § 8953(a)). They do so deliberately. In adopting a wide interpretation of the state’s hot pursuit law in 1991, the Pennsylvania Supreme Court explained that “constructing impenetrable jurisdictional walls benefit[s] only the criminals hidden in their shadows.”199Commonwealth v. Merchant, 595 A.2d 1135, 1139 (Pa. 1991). This dictum has become something of a maxim in hot pursuit cases. See Peters, 965 A.2d at 225 (citing Merchant, 595 A.2d at 1139)); Commonwealth v. Henry, 943 A.2d 967 (Pa. Super. Ct. 2008) (citing Merchant, 595 A.2d at 1139)).

One could go on and on with gripping stories of extraterritorial policing—of late-night arrests outside city limits200Gustke, 516 S.E.2d at 286. and NYPD raids in Newark.201Adam Goldman & Matt Apuzzo, NYPD Built Secret Files on NJ, Long Island Mosques, NBC4 N.Y. (Feb. 22, 2012, 10:13 a.m.), https://www.nbcnewyork.com/news/local/muslim-surveillance-nypd-newark-cory-booker-mosque/1972322 [perma.cc/XK2S-39U4]. Together, these examples demonstrate that the borders of police jurisdiction are negotiable. Territorial jurisdiction is not meaningless; the technical limits of police authority still matter to police, to defendants, and to some courts. But mutual aid agreements, task forces, compacts, citizen’s arrest statutes, and hot pursuit doctrines are all legal devices that detach policing from geography. In practice, the insignia on police uniforms belie a nuanced, interjurisdictional reality in which “local” and “state” police have more than local and state power.

C. Prosecution

It is no surprise that police cross borders. Policing is, after all, an infamously discretionary activity.202See John Paul Stevens, Our ‘Broken System’ of Criminal Justice, N.Y. Rev. Books (Nov. 10, 2011), https://www.nybooks.com/articles/2011/11/10/our-broken-system-criminal-justice [perma.cc/6YTH-5X49] (reviewing Stuntz, supra note 30) (discussing the expansion of police discretion). But one would expect prosecution to be different. Given the Constitution’s focus on criminal trials—the Venue Clause, the Vicinage Clause, all the debate over criminal trials at the founding203See Coffin v. United States, 156 U.S. 432, 453 (1895) (“The principle that there is a presumption of innocence in favor of the accused is the undoubted law, axiomatic and elementary, and its enforcement lies at the foundation of the administration of our criminal law.”); see also supra Section I.B (describing state and federal constitutional limitations on the location of criminal trials).—it would seem as if prosecution must have strict geographic boundaries.204See supra Section I.B. Even at this phase of the criminal legal process, though, territorialism has receded.

Several historical developments drove the deterritorialization of prosecution. First, in the late nineteenth century, courts began to distinguish jurisdiction from venue.205Kershen II, supra note 57, at 3–6 (providing a detailed account of this history). As Part I explained, early American legislators and courts adopted a common law theory of criminal jurisdiction in which the power to hear a criminal case and the location of a criminal trial were coextensive because crime was “inherently local.”206See supra Section I.B. That conception of criminal jurisdiction started to splinter in the 1870s as Congress began to separate federal judicial districts into smaller divisions.207Kershen II, supra note 57, at 4. The subdivision of judicial districts raised a new question: Could federal trial courts adjudicate criminal cases involving crimes in another division within the same district? The Supreme Court answered that question in the affirmative in a trio of cases between 1892 and 1897.208Id. at 5 (discussing Logan v. United States, 144 U.S. 263 (1892), Post v. United States, 161 U.S. 583 (1896), and Rosencrans v. United States, 165 U.S. 257 (1897)). Those cases began to carve out a distinction between the power to conduct criminal proceedings and the proper place for a criminal trial.

That distinction deepened between 1925 and 1931, when appellate courts—first the Fifth Circuit,209Silverberg v. United States, 4 F.2d 908, 909 (5th Cir. 1925). then the D.C. Circuit210Hagner v. United States, 54 F.2d 446, 448 (D.C. Cir. 1931), aff’d, 285 U.S. 427 (1932).—held that venue was a criminal procedure right that could be waived. Once venue was a right that belonged to criminal defendants rather than a limit on judicial power or a requirement for the legitimacy of criminal trials, the separation of jurisdiction and venue was “complete.”211Kershen II, supra note 57, at 9. Defendants could agree to move trials (or simply fail to make a timely venue objection),212See, e.g., People v. Simon, 25 P.3d 598 (Cal. 2001) (holding that a criminal defendant had forfeited his venue challenge). and so long as the court could find a hook for its jurisdiction, the criminal prosecution could proceed.

Recall, moreover, that the Supreme Court expanded the bases for criminal jurisdiction to include the effects of a crime in 1911.213See supra note 138 and accompanying text (discussing Strassheim v. Daily, 221 U.S. 280, 281, 285 (1911)). As the previous Section explained, courts began to adopt a bolder and more flexible approach to judicial power over out-of-state crime during the progressive era. Alongside that doctrinal shift, venue’s transformation into a criminal procedure right meant that criminal trials could be prosecuted in a wider range of places. By the middle of the twentieth century, venue was a right that defendants could forfeit or waive. Jurisdiction was a separate matter, satisfied so long as a crime caused harm or implicated an interest in the place where the court sat.214See supra Section II.A (describing the emergence of effects-based criminal jurisdiction in the early twentieth century).

This new approach to criminal jurisdiction enabled a second development: the spread of special venue laws. Once courts had abandoned strict territorial ideas about their power in criminal cases, legislatures began to pass more laws dislocating criminal prosecution.215There are a handful of specific venue statutes that predate the 1920s, but this legal device really gained traction after the 1930s. See Charles Doyle, Cong. Rsch. Serv., RL33223, Venue: A Legal Analysis of Where a Federal Crime May Be Tried 3–16 (2018). These laws take several forms. The most common is the “specific” venue statute, which designates the place of trial for a certain type of offense. Specific venue statutes can provide for venue in a particular court or in multiple places—for example, anywhere a stolen item is taken, carried, or found.216See, e.g., 18 U.S.C. § 3238; see also infra note 218 (citing state law sources). Their close cousin, the “buffer” statute, allows prosecutors to choose between neighboring counties when crimes occur in multiple jurisdictions or near a county boundary line.217See Brian C. Kalt, Crossing Eight Mile: Juries of the Vicinage and County-Line Criminal Buffer Statutes, 80 Wash. L. Rev. 271, 274, 277 (2005) (surveying these statutes and arguing that “buffer statutes should be eliminated”). State laws and procedural rules can also allow prosecutions to proceed where a person is arrested or resides.218See, e.g., Tenn. R. Crim. P. 18(d)(2) (“An offense committed wholly outside Tennessee may be prosecuted in any Tennessee county in which the offender is found.”); Tex. Code Crim. Proc. Ann. art. 13.01 (West 2015) (“Offenses committed wholly or in part outside this State . . . may be prosecuted in any county where the offender is found . . . .”); Utah Code. Ann. § 76-1-202(g) (2017) (providing for trials in the county where a defendant resides, is apprehended, or to which he is extradited in cases where the site of the crime is unclear); Ariz. Rev. Stat. Ann. § 13-109(C) (2020) (same); Or. Rev. Stat. § 131.325 (2021) (same); Va. Code Ann. § 19.2-244(b) (2015) (same). These laws are exceptional; as Part I explained, venue is usually proper in the county where the crime occurs.219See supra Section I.B. But special venue statutes demonstrate that, in modern criminal law, legislatures can move trials where they wish. (Courts participate here too, by treating venue defects as harmless.)220See, e.g., People v. Houthoofd, 790 N.W.2d 315, 330 (Mich. 2010) (“[L]ack of proper venue is subject to a harmless error analysis.”); State v. Blankenship, 170 S.W.3d 676, 682 (Tex. Ct. App. 2005) (“[F]ailure to prove venue does not negate the guilt of the defendant.”).

Prosecution also grew more coordinated and centralized during the twentieth century. Like police, prosecutors can join task forces to investigate and try offenses that span formal boundary lines. In New York City, for example, a “Special Narcotics Prosecutor” appointed by the district attorneys from the city’s five boroughs can prosecute drug and related crimes across the city.221About Us, Off. of the Special Narcotics Prosecutor for N.Y.C., https://www.snpnyc.org/about-us [perma.cc/6XGK-NJE5]. The state created this interjurisdictional office in 1971 “to address [the] free flow of narcotics across county lines.”222Id. Illinois began to establish similar “multijurisdictional drug prosecution units” in the early 1980s.223 . Joan E. Jacoby, Carl B. Hammond, Edward C. Ratledge & Stephen W. Ward, Ill. Crim. Just. Info. Auth., Evaluation of Illinois’ Multijurisdictional Drug Prosecution Programs (1999), http://icjia.state.il.us/assets/pdf/researchreports/EvalMjdrug.pdf [perma.cc/Q7TR-29J8] (cleaned up). California uses an analogous model for identity theft crimes, which can be combined with “all associated offenses,” consolidated, and tried in the county of the prosecutor’s choice.224Howard A. Wise & Joe Williams, Jr., Using Enhanced Jurisdictional Laws to Prosecute Multi-county Identity Thefts, 36 CDAA Prosecutor’s Brief 27 (2010).

It is important not to overstate the trend toward centralized prosecution. Criminal prosecution of state law remains one of the most localized and varied parts of the American justice system.225See Barkow, supra note 187, at 550–70. States routinely delegate prosecutorial power to local actors, who often act without centralized oversight.226Id. The impression here should not be one of a highly consolidated legal regime. But there is no question that technological developments and relaxed conceptions of criminal jurisdiction encouraged cooperation between prosecutors and spurred the creation of interjurisdictional prosecution offices.227See Terence Dunworth, Information Technology and the Criminal Justice System: An Historical Overview, in Information Technology and the Criminal Justice System 3, 5, 19 (April Pattavina ed., 2005). In some cases, the boundaries of state criminal jurisdiction also expanded more formally. In 1942, for example, South Carolina abandoned a system in which criminal prosecutors had countywide jurisdiction in favor of its current system, which gives prosecutors multicountry jurisdiction. See S.C. Code § 3869 (1912); S.C. Code Ann. § 1-251.1 (Supp. 1960); S.C. Code Ann. § 1-7-310 (Supp. 2022). As with special venue statutes, these innovations depart from the territoriality norm and permit a measure of forum shopping in criminal prosecution.228See Kalt, supra note 217, at 274 (denouncing buffer statutes for “reward[ing] forum shopping by prosecutors”).

Finally, and perhaps most important, the twentieth century witnessed the rise of the plea bargain. The exponential growth of negotiated pleas is one of the most documented and lamented features of the American criminal justice system.229See, e.g., George Fisher, Plea Bargaining’s Triumph (2003); Daniel Epps, Adversarial Asymmetry in the Criminal Process, 91 N.Y.U. L. Rev. 762 (2016); Jennifer L. Mnookin, Uncertain Bargains: The Rise of Plea Bargaining in America, 57 Stan. L. Rev. 1721 (2005) (reviewing Fisher, supra). Today, plea bargains account for almost all criminal convictions in the United States.230Stephen B. Bright, The Failure to Achieve Fairness: Race and Poverty Continue to Influence Who Dies, 11 U. Pa. J. Const. L. 23, 24 (2008) (“[T]he overwhelming majority of criminal cases—90% to 95%—are resolved with plea bargains.”). Professor John Langbein famously observed that this system “recapitulate[s] much of the doctrinal folly of the law of torture.”231John H. Langbein, Torture and Plea Bargaining, 46 U. Chi. L. Rev. 3, 17 (1978). Plea bargaining also displaces the jury, which transforms vicinage into a rather weak criminal procedure right. Ironically, given the Founders’ fixation on juries, the Constitution requires territoriality at the least important moment of the modern criminal process.

These four developments—the splintering of jurisdiction and venue, the proliferation of special venue rules, the creation of multijurisdictional prosecution units, and the plea-bargaining revolution—mean that even prosecution is not as territorial as one would think. In the modern criminal justice system, a court’s power to hear a case is different than where it may be heard. Legislatures can designate special places for criminal trials. Prosecutors can forum shop. Venue defects are harmless errors. Plea bargains have largely undermined the relevance of the territorial limits on trials and juries. The cumulative result is a criminal justice system in which even prosecution, the lodestar of constitutional criminal procedure, is less territorial than it used to be.

D. Punishment

Finally, there is punishment. The last phase of the criminal legal process is the least tied to territorial boundaries. Take imprisonment, the most well-known criminal sanction. American law does not require people to be imprisoned in the jurisdiction whose laws they violated, nor do prisoners need to be held near the site of their crime.232See Emma Kaufman, The Prisoner Trade, 133 Harv. L. Rev. 1815, 1823, 1856 (2020). Prison officials may send federal prisoners across the country, no matter where they were convicted. State prisoners can also be transferred, sometimes thousands of miles, to serve their sentence in another state’s prisons.233Id. at 1821, 1857. And noncitizens convicted in American courts can be repatriated to their country of origin—so a Massachusetts or Texas prison sentence can lead to prison time in Australia or Germany.234Emma Kaufman, Extraterritorial Punishment, 20 New Crim. L. Rev. 66, 67 (2017).

My previous research has explored the history of these prisoner transfer systems. As that work explains, the prison was originally conceived in the late eighteenth century as a territorial alternative to transportation punishment, the reigning noncapital sanction at the time.235Kaufman, supra note 232, at 1823, 1856. Early Americans believed imprisonment was a local and therefore more democratic alternative to transportation, a “monarchical” sanction used by “kings and despots.”236Id. at 1823. (One can hear echoes of how the term “local” was used in early debates about criminal jurisdiction here.)237See supra Section I.B; infra Section III.B (connecting early American debates over the “inherently local” nature of criminal law to a theory of criminal jurisdiction in which territorialism protects both sovereignty and democracy). The first American prison systems were built on this “positive republican theory of crime,” in which territorial punishment connected prisons to the polity that enacted criminal laws and promised to reintegrate prisoners into that polity upon release.238See Rebecca M. McLennan, The Crisis of Imprisonment 19 (2008).

This philosophy of punishment ebbed with the rise of the administrative state and the professionalization of a prison bureaucracy.239Kaufman, supra note 232, at 1826–27. As prison systems expanded in the 1940s and 1950s, state prison officials began to sign contracts to share prisoners and prison bed space.240Id. at 1828. Those contracts turned into regional compacts and eventually a national prison network in which states can send their prisoners across the country.241Id. at 1828–29. The Supreme Court upheld interstate punishment in 1983 when it ruled that prisoners have no due process right to in-state confinement, even if prison transfers “involve[] long distances and an ocean crossing.”242Olim v. Wakinekona, 461 U.S. 238, 247 (1983). Thus, between 1930 and 1990, it became both normal and legal to punish prisoners far from the site of their crime.

International prisoner repatriation proceeded along a similar timeline. In the mid-1970s, after high-profile news reports and congressional hearings on the treatment of Americans imprisoned abroad, the United States entered its first treaty for international prisoner transfers, a bilateral agreement with Mexico.243Kaufman, supra note 234, at 70. A series of other treaties followed, and by 1985 the United States was a party to agreements providing for repatriation to and from sixty-seven countries.244Id. at 71. That number has since grown to more than one hundred.245Id. Under the terms of these treaties, the United States can repatriate both federal and state prisoners to serve prison sentences in their countries of origin. The United States can also receive and “resentence” Americans convicted abroad,246American citizens repatriated to the United States are resentenced by the United States Parole Commission “as though [they] were convicted in a United States district court of a similar offense.” 18 U.S.C. § 4106A(b)(1)(A). This process undermines the conventional story about the separation of powers in which judges impose prison sentences. See Kaufman, supra note 234, at 74. though in practice a treaty regime born for that purpose is used mainly to export noncitizens out of the country.247Kaufman, supra note 234, at 78–79 (describing how repatriation “transformed into a vehicle for prisoner exportation” in the mid-1990s).

Interestingly, the systems for both interstate and international prison transfers are underused.248This is a descriptive claim about government actors’ uptake of legal transfer regimes, not a normative assertion that more prisoners should be transferred. See id. at 67–68 (explaining that repatriation is “remarkably rare” neither because prisoners wish to stay in the United States nor because the law prevents transfers, but rather because prison bureaucrats deny transfer applications on the ground that prisoners are either “too American” or committed a crime too serious to license punishment abroad); Kaufman, supra note 232, at 1842, 1861 (noting that the overall number of interstate prisoner transfers is small relative to the total American prison population and connecting this trend to political and financial incentives to keep prisoners in state). Prisoners are outsourced less than the law would permit—a testament to the durability of the territoriality norm.249Resistance to prison transfers reflects both the lingering territoriality norm in American criminal law and entrenched political dynamics—including bureaucratic turf wars and concerns about prisoners’ proximity to family—that lead to in-state punishment. See Kaufman, supra note 232, at 1863. These political dynamics keep imprisonment loosely territorial even as American courts have given up on the territoriality principle as a necessary basis for punishment. But the emergence of prisoner transfer systems represents a notable shift away from the belief that punishment has to be local to be lawful or democratic. Over the course of the twentieth century, courts and prison bureaucrats concluded that punishment need not be connected to the place where criminal law is made.

By now, the arc of this narrative should feel predictable. The history of American imprisonment parallels the deterritorialization of policing and prosecution, with glimmers of interjurisdictional cooperation in the early twentieth century giving way to more systematic coordination by the 1970s. In each of these contexts, increased mobility and state capacity put pressure on the territoriality principle and encouraged more creative—which is to say extraterritorial—forms of criminal law enforcement.

Once again, moreover, the Due Process Clause was no real impediment to the expansion of criminal law.250See supra notes 142–143 and accompanying text. The Supreme Court has never policed the location of prisons, and the Court explicitly declined to constitutionalize prison placement in the early 1980s.251See Olim v. Wakinekona, 461 U.S. 238, 247–48 (1983). As a result, constitutional law imposes very few limits on where prisoners may be held. American prison officials can move prisoners all over the country (in some cases all over the world), and the site of a crime is barely related to where punishment occurs. Instead, personal facts about a prisoner—his sentence length, disciplinary record, family’s location, citizenship status—determine where he will be confined.252See Kaufman, supra note 232; Kaufman, supra note 106. This is a legal regime that has long since given up on the idea that punishment must be local to be legitimate.

E. Post-Territorial Criminal Law

In the end, then, every stage of criminal law has been deterritorialized. This transformation has been uneven. Some parts of criminal procedure, such as jury trials, remain more territorial than others. If asked to rank the system by its commitment to borders, the right response is probably that criminalization is the most territorial phase of American criminal law, followed by prosecution, policing, and then punishment. To be even more granular: legislative jurisdiction is slightly more territorial than judicial jurisdiction; both are a bit more territorial than prosecution; and prosecution has clearer geographic limits than either policing or punishment, which are deeply discretionary. But these are differences in degree rather than kind. At every point in the process of making and applying domestic criminal law, the territoriality principle has declined.

This observation should be surprising. In thinking about when American criminal law is territorial, one might expect to discover a disconnect between criminalization and criminal law enforcement—that is, a system in which legislatures define crimes with territorial borders in mind and courts understand their jurisdiction in territorial terms, but police and prison officials go rogue when implementing legal rules. On this account, the deterritorialization of criminal law would be an enforcement pathology. This story would be consistent with the view that criminal law “on the books” always looks different than criminal law “in action,” particularly in a country where courts afford law enforcers so much latitude.253Issa Kohler-Hausmann, Jumping Bunnies and Legal Rules: The Organizational Sociologist and the Legal Scholar Should Be Friends, in The New Criminal Justice Thinking 246, 246–47 (Sharon Dolovich & Alexandra Natapoff eds., 2017).

Alternatively, one could imagine a criminal legal system that was territorial before conviction but unconcerned about the location of punishment. In such a system, policing and prosecution would be territorial, but imprisonment would not be. This system would line up with the prominent (though not uncontested) theory that criminal procedure rights exist to protect the innocent.254 . Akhil Reed Amar, The Constitution and Criminal Procedure 154 (1997) (“[The] commonsensical point, I submit, is the essence of our Constitution’s rules about criminal procedure, and so I shall repeat it: the Constitution seeks to protect the innocent.”). But see Louis Michael Seidman, Akhil Amar and the (Premature?) Demise of Criminal Procedure Liberalism, 107 Yale L.J. 2281, 2282 (1998) (reviewing Amar, supra) (critiquing Amar’s argument for “a general reorientation of criminal procedure toward factual guilt and innocence”). On this account, territorial policing and prosecution would be a mechanism for providing fair and accurate criminal adjudication, and deterritorialized punishment would simply reflect the Constitution’s fixation on innocence. This version of criminal law would also accord with the basic thesis of the sociology of punishment, which is that “criminal law and punishment are distinct social and cultural practices.”255 . Lindsay Farmer, Making the Modern Criminal Law 21 (2016) (citing David Garland for this proposition); see also David Garland, Punishment and Modern Society 19 (1990) (“Like architecture or diet or clothing or table manners, punishment has an instrumental purpose, but also a culture style and an historical tradition . . . .”). For both the constitutional theorist and the sociologist, it would not be especially remarkable to discover that punishment looks different from pre-conviction criminal justice.

Given the stories we tell about criminal law, it would make sense if deterritorialization were the product of unruly officers or a general disregard for the convicted. If these molds fit, the conclusion would be that criminal law is territorial, notwithstanding some predictable nuance. Yet neither of these two models captures the current legal system. In reality, criminal law is more disjointed: it is sometimes territorial, sometimes predicated on alternative ideas like citizenship or abstract state interests, and most territorial at the parts of the criminal process that matter the least. It strains credulity to call this an essentially territorial legal regime. In fact, American criminal law is split between territorialism and a different theory of state power in which criminal law applies to a state’s members and interests wherever they go.

Recall, for instance, that after Skiriotes state criminal laws can follow citizens outside state territory; that the separation of venue and jurisdiction enables both prosecutors and defendants to move criminal trials; that local police forces can pool their powers across state lines; and that state prison officials can trade prisoners and rent prison beds in another state.256See supra Section II.A (discussing Skiriotes v. Florida, 313 U.S. 69 (1941)); Section II.B (interjurisdictional policing); Section II.C (venue’s transformation into a waivable right and prosecutorial forum-shopping); Section II.D (interstate punishment). These are formal, legal instances of deterritorialized criminal law, not aberrations or pathologies so much as evidence of a post-territorial legal regime that has been warped over time by new rules and enforcement practices. Over the course of the twentieth century, tectonic social shifts—increased mobility, the birth of a criminal enforcement bureaucracy, the rise of interstate compacts, a turn to private contracting—changed the character of criminal justice and made territorialism seem cumbersome and antiquated. These social developments have produced a legal system that is home to several competing ideas about the source and outer limits of the power to enact criminal law.

III. The Stakes of Territorialism

We are left with a conflicted criminal justice system. On one hand, the territoriality principle is entrenched in constitutional law and institutional design. On the other hand, the borders of criminal law have faded in clarity and import. Over time, both the substance and the enforcement of domestic criminal law have become less territorial: more regional, more discretionary, and more flexible about permissible forms of jurisdiction. Our state-based criminal justice system is decreasingly tethered to state soil. And criminal courts have long since abandoned nineteenth-century ideas about the limits of their authority.

These developments lead to some concrete conclusions. As the Introduction noted, the emergence of the modern criminal legal system—with its contractual, overlapping boundaries—embarrasses constitutional doctrines premised on the inherent localness of criminal law. The doctrines discussed in Part I look outdated and contrived in light of Part II. Meanwhile, the rise of task forces and ad hoc enforcement agreements demands more transparency about the reach of criminal laws. People subject to law enforcement ought to know which laws govern them, which police can detain them, what practices their taxes are funding, and which criminal prosecutions can be brought in their name.

This Part focuses, though, on the deeper implications of territoriality’s decline. It explores two observations. First, the evolution of criminal jurisdiction makes it more difficult to explain why criminal law is distinctive and thus why we have special rules for criminal law administration. The more one examines the territoriality principle, the more one wonders why Massachusetts can’t enforce Nevada’s criminal code. Second, the erosion of territorialism raises normative questions about whether borders—and, in particular, state borders—are good for criminal law. This Part considers that question and concludes with a cautious call to revive the territoriality principle as a strategy for criminal justice reform.

A. Distinguishing Criminal Law

Part II described a legal regime shaped by private agreements—contracts, compacts, MOUs—that alter the boundaries of criminal law. This is not the criminal legal system of ideal theory, in which the power to enact criminal law depends on the consent of a polity with clearly defined, stable borders.257See R.A. Duff, Punishment, Communication, and Community 36–39 (Michael Tonry & Norval Morris, eds., 2001) (discussing liberal theories of punishment); Farmer, supra note 255, at 27 (critiquing ideal theories of criminal law); Nicola Lacey, Approaching or Re-thinking the Realm of Criminal Law?, 14 Crim. L. & Phil. 307, 313 (2020) (discussing the place of ideal theory in philosophy of criminal law). See generally Don Herzog, Happy Slaves (1989). Instead, we live in a system of private ordering in which government officials negotiate the scope of criminal laws, often behind closed doors.258In this respect, American criminal law looks more like a body of private contract law than a subfield of public law. Cf. Fletcher, supra note 128, at 190–91 (“[A]s a general matter, Americans are a bit easy about the conceptual differences between private law and criminal law.”). Fletcher distinguishes this American approach from German criminal law, in which private law notions such as vicarious liability are “anathema to the criminal law.” Id. at 191. This system developed over the course of the twentieth century as mobility and technology undermined territorial rules, encouraged more relaxed conceptions of criminal jurisdiction, and demanded flexible institutional arrangements.

In many respects, this system’s birth is a familiar story. The account of criminal law outlined in Part II lines up with the standard narrative of legal development in other disciplines, such as civil procedure and conflict of laws. In those fields, it is not at all shocking to say: “This body of law changed around the middle of the twentieth century, as social pressure to modernize and privatize made territorial boundaries seem antiquated and generated new, less formalist theories of law.” This statement broadly describes the trajectory of personal jurisdiction jurisprudence259Danielle Keats Citron, Minimum Contacts in a Borderless World: Voice over Internet Protocol and the Coming Implosion of Personal Jurisdiction Theory, 39 U.C. Davis L. Rev. 1481, 1504–07 (2006) (describing how “[t]he twentieth century’s sea change in transportation and communication technologies” resulted in International Shoe); see Barbara Allen Babcock, Toni M. Massaro, Norman W. Spaulding & Myriam Gilles, Civil Procedure: Cases and Problems 113 (7th ed. 2021) (“International Shoe . . . cut [personal jurisdiction] analysis loose from the mooring of territoriality and set it upon the uncharted sea of ‘minimum contacts.’ ”). Recent Supreme Court decisions on personal jurisdiction have led scholars to identify—and in most but not all cases, decry—a return to territorialism in civil procedure jurisprudence. See Hoffheimer, supra note 31; cf. Stephen E. Sachs, Pennoyer Was Right, 95 Tex. L. Rev. 1249 (2017). and the rejection of territorial approaches to choice of law in the 1950s.260See Brilmayer et al., supra note 27, at xxiv (tracing the decline of territorialism in choice of law theory); Patrick J. Borchers, The Choice-of-Law Revolution: An Empirical Study, 49 Wash. & Lee L. Rev. 357, 360–61 (1992) (same); Laycock, supra note 31, at 317 (defending territorialism).

Indeed, the transformations of civil and criminal law occurred along strikingly similar timelines. The Supreme Court decided International Shoe Co. v. Washington261326 U.S. 310 (1945).—the case credited with unmooring civil jurisdiction from territoriality262See sources cited supra note 259.—just four years after Skiriotes,263Skiriotes v. Florida, 313 U.S. 69 (1941). the sponge-fishing case that recast criminal law. This timeline suggests that the Court was engineering dramatic shifts in both civil and criminal jurisdiction in the middle of the twentieth century. Yet only the civil story has received extended treatment in legal scholarship. Scholars who study procedure and conflicts—that is, academics who study the civil side of American law—widely recognize and routinely debate the decline of territorialism.264For the conflicts debate, see sources cited supra note 260. For a summary of the civil procedure debate, see Hoffheimer, supra note 31. Criminal law scholars, by contrast, write very little about territoriality.

But as Part I made clear, territorialism is meant to be one of the field’s core attributes. Everyone from the drafters of the Model Penal Code to constitutional theorists identifies the territoriality principle as a “maxim of . . . jurisprudence” that distinguishes domestic criminal law.265 . Model Penal Code § 1.03 explanatory note (Am. L. Inst., Proposed Official Draft 1962); see supra notes 12–18 (collecting sources that make this claim). And the Supreme Court treats the territoriality of criminal law as an implicit and longstanding rule.266See, e.g., Bigelow v. Virginia, 421 U.S. 809, 822–25 (1975). In Bigelow v. Virginia, a First Amendment case that invalidated the conviction of a Virginia newspaper editor who printed an advertisement for a New York abortion clinic, the Court deemed it “obvious” that Virginia could not proscribe abortions in New York or prosecute its residents for going there.267Id. at 823–24 (1975). Interestingly, Justices Rehnquist and White dissented, arguing that the majority’s “rigid and unthinking territorial limitation” on Virginia’s power flew in the face of precedents like Strassheim and Skiriotes and had no clear “constitutional source.” Id. at 834 n.2 (Rehnquist, J., dissenting). As this dissent suggests, the Supreme Court’s presumption that criminal law is territorial “is quite at war” with the cases explored in Part II. Id. For this obvious proposition, the Court cited Huntington v. Attrill, an 1892 case in which it had observed that penal laws are territorial because crimes offend “the public justice of the state.”268146 U.S. 657, 674 (1892); see infra note 278 (discussing Huntington); see also Note, The Extraterritorial Operation of Penal Statutes, 1 Va. L. Rev. 390 (1914).

In all of these accounts, territorialism is what makes criminal law identifiable as a distinct genre of public law. Territorialism is also what makes domestic criminal law real law. As Professors Jack Goldsmith and Daryl Levinson have observed, American legal scholars tend to treat domestic law as real and concrete, by contrast to the purportedly unstable and unenforceable world of international law, where legal rules inevitably devolve into power relations.269Jack Goldsmith & Daryl Levinson, Law for States: International Law, Constitutional Law, Public Law, 122 Harv. L. Rev. 1791, 1792–93 (2009). In this framework, domestic law is taken to be the “paradigm” version of law, while international law is a “lesser species of law—if it qualifies as law at all.”270Id. at 1793. To extend Goldsmith and Levinson’s argument, domestic criminal law is supposed to be the paradigm sort of law, with clear rules, boundaries, and sources of authority. It belongs in the category of “real” law because of its dedication to the territoriality principle, which, as I noted at the outset, is a shorthand for the idea that a government’s power to enact criminal law stems from and stops at its borders.

The territoriality principle is thus central to how scholars conceptualize criminal law. It is what divides domestic from international criminal law. It is what distinguishes criminal from civil law. It is what makes domestic criminal law into the sort of content that belongs in a first-year legal curriculum rather than some upper-level seminar on the limits of sovereignty. But as Part II demonstrated, domestic criminal law is much less territorial than it seems at first pass. Modern criminal law is riddled with exceptions to the territoriality principle: criminal laws that follow Florida citizens into the Gulf of Mexico; prosecutorial forum shopping; regional police task forces; a prison system that pays no attention to the place a person is convicted.271See supra Part II. The list goes on, and as these exceptions stack up, domestic criminal law starts to seem much less committed to the principle that allegedly defines it.

My claim is not that American criminal law is completely or even mostly deterritorialized. Criminal law is often territorial. Rather, the point is that even “normal” criminal law is less tied to territory than one would expect given that territoriality is supposed to be what makes this body of law normal. In reality, states and localities cooperate with each other, domestic criminal law is split between competing theories of jurisdiction, and the criminal justice system is so thoroughly saturated with discretion that territorial borders can be ignored by both law enforcers and courts. As a result, state criminal law is just as unsettled and prone to power dynamics as “lesser” bodies of law,272Goldsmith & Levinson, supra note 269, at 1792–93. and its justification and scope are just as contested. When all is said and done, domestic criminal law is not all that territorial—which means it is not all that distinctive, either.

It would take a separate article to explore the full implications of the observation that criminal law is not distinctive. The big, underlying thesis here is that the principle that divides criminal law from civil law, and domestic law from international law, has eroded over the last century such that these conceptual divisions no longer withstand scrutiny. This thesis is unnerving given how much rides on the distinction between criminal and civil law. Our basic legal infrastructure (not to mention the entire constitutional rights regime) is predicated on the idea that criminal and civil law are meaningfully different. The harshness of criminal sanctions is also supposed to be justified by the distinctiveness of criminal law. The decline of territorialism puts real pressure on these claims.

Waning territorialism also prompts questions about why criminal law operates differently than other fields. For instance, if domestic and international criminal law are not especially distinct, why can’t states (like nations) criminalize their residents’ out-of-state behavior? What is wrong with a Texas statute that prevents a Texas resident from gambling in Las Vegas? What stops Alabama from criminalizing its citizens’ conduct in California? To those immersed in Anglo-American criminal law and American constitutionalism, these sorts of statutes feel intuitively problematic—imperial, dismissive of tradition, and threatening to national unity. That intuition may be right and worth solidifying in legal doctrine. But as the law stands, it is not clear whether or why these statutes would be unlawful, aside from the “long-held maxim” that criminal law doesn’t work that way. That maxim looks precarious after Part II. It is hard to argue that state criminal law must be territorial when state legislatures, courts, police, prosecutors, and prison officials all routinely act beyond state boundary lines.

In a similar vein, one might ask why courts continue to cite and endorse the prohibition on interstate enforcement of penal laws. If criminal and civil law are not particularly distinct, why is there no choice of criminal law? Why can’t California authorize Texas to apply its criminal code? What would be wrong with a legal system where Ohio courts imposed Pennsylvania’s criminal law? And what, exactly, is the difference between that system and one where Ohio can police and imprison Pennsylvania’s criminals?

The goal is not to answer these questions so much as to bring foundational puzzles about criminal law into view by pointing out that the territoriality principle, which is supposed to be defining, has receded. The territoriality principle undergirds the intellectual framework of public law. It separates disciplines like international and domestic criminal law, and it drives unusual features of the criminal legal system, such as the practice of extradition, the presumption against extraterritorial state criminal laws, and the prohibition on shared enforcement of criminal codes. All of these practices look stranger and less defensible once one recognizes that borders are no longer sacrosanct, not even in state criminal law.

B. Taming Criminal Law

The question that follows is whether borders are desirable. The decline of the territoriality principle raises the possibility of a more flexible, less territorial criminal justice system. Part II demonstrated that government officials have long been expanding and redrawing the boundaries of criminal jurisdiction. That history suggests that one could get even more creative about the scope of criminal laws. Or perhaps the deterritorialization of criminal law is an unwelcome development. To know how to think about this issue, one has to start by asking what purpose borders serve in American criminal law.

1. The Failure of Borders

There were three justifications for territorialism lurking in the doctrines surveyed in Part I. First, the territoriality principle was supposed to protect criminal defendants. This theory appeared in the Model Penal Code, which explains that geographic limits on criminal prosecution “ensure . . . []fairness to the defendant.”273 . Model Penal Code § 1.03 explanatory note (Am. L. Inst., Proposed Official Draft 1962). The defendant-protective understanding of territorialism also featured in early constitutional debates over venue and vicinage.274See supra Section I.B. There, the claim was that territorialism would benefit defendants by keeping trials close to home, which would guarantee access to counsel and sympathetic juries.275See supra Section I.B. The basic belief animating this theory of the territoriality principle is that borders temper the harshness of criminal law. Territorialism limits state power, which protects those subject to coercive state force.

Second, the territoriality principle was supposed to instantiate sovereignty. By marking out the place where governments can enact and enforce their criminal laws, borders are meant to preserve a right to self-determination and nonencroachment, which in turn ensures comity between (allegedly distinct and independent) states. We saw this version of territorialism at work in the interstate extradition process276See supra Section I.B. and in the constitutional cases that relied on claims about criminal law’s localness to limit federal power over states.277See supra Section I.D (discussing Commerce Clause, abstention, and habeas cases that rely on the “inherent localness” of criminal law to limit congressional power over states and federal judicial power over state criminal courts). In those examples, keeping criminal law within territorial boundaries was a way to show that state governments are real and deserve respect. The basic belief animating this conception of territorialism is that the power to enforce criminal law is what makes a state a state.

Third, the territoriality principle was supposed to legitimate criminal law. In theory, the reason it is problematic for Texas courts to apply Florida criminal law, or for Massachusetts police to arrest a Vermonter in Vermont, is that criminal law is authorized by a democratic process in a certain state. Accordingly, the theory goes, it is illegitimate to apply the criminal laws of that state to people who had no say in their creation. This is the understanding of territoriality that explains the penal law taboo278See supra Section I.A (describing the prohibition against interstate enforcement of penal laws). This “incontrovertible maxim” comes from Chief Justice Marshall’s statement about the law of nations in The Antelope, the first case in which the Supreme Court considered (and upheld) the international slave trade. The Antelope, 23 U.S. (10 Wheat.) 66, 122–23 (1825) (“[I]t is almost superfluous to say in this Court [that] . . . [t]he Courts of no country execute the penal laws of another.”); see also Huntington v. Attrill, 146 U.S. 657, 666 (1892) (considering “the true scope and meaning of [this] fundamental maxim of international law”). As noted above, the MPC adopts this maxim and converts it into a principle of domestic criminal law. See Model Penal Code § 1.03 explanatory note (Am. L. Inst., Proposed Official Draft 1962) (“[I]t has long been a maxim of American jurisprudence that a state will not enforce the penal laws of another state.”). The Supreme Court has followed suit. See Nelson v. George, 399 U.S. 224, 228–29 (1970) (citing Huntington for the proposition that “the Full Faith and Credit Clause does not require that sister States”—here California and North Carolina—enforce each other’s “foreign penal judgment[s]”). and the rule that in criminal law, “jurisdiction and choice of law . . . are merged.”279 . Model Penal Code § 1.03 explanatory note (Am. L. Inst., Proposed Official Draft 1962). It is also the theory behind the claim that crime is an offense against “the public justice of the [s]tate”280Huntington, 146 U.S. at 674. rather than any private citizen—which is why criminal prosecutors act on behalf of “the people” and criminal cases have captions like State v. Guthrie.281461 S.E.2d 163 (W. Va. 1995); see Jocelyn Simonson, The Place of “the People” in Criminal Procedure, 119 Colum. L. Rev. 249, 250–51 (2019). The basic belief animating this conception of territorialism is that criminal law is an especially public sort of law, which requires an especially tight connection between the law and the people. Borders provide that link.

To summarize, the territoriality principle is supposed to protect defendants, safeguard sovereignty, and sustain a connection between criminal law and democracy. The problem is that territorialism is not doing particularly well on any of these fronts.

As Part II showed, borders rarely protect criminal defendants. In modern criminal law, police and prison officials pool their powers across boundary lines; courts decline to enforce territorial jurisdictional limits because they “benefit[] only the criminals hidden in their shadows”;282Commonwealth v. Merchant, 595 A.2d 1135, 1139 (Pa. 1991). and plea bargaining has diminished the relevance of geographic restrictions on criminal adjudication. Moreover, as Part I noted, territorialism was never meant to protect defendants, at least not primarily. Despite the Model Penal Code’s insistence that the territoriality principle exists to provide fairness to defendants,283 . Model Penal Code § 1.03 explanatory note (Am. L. Inst., Proposed Official Draft 1962). it was actually a particular, now-outdated theory of criminal jurisdiction that motivated the Venue Clause.284See supra Section I.B. The Founders could have required trials to be held close to a defendant’s home; instead, they tied venue to the site of the crime. This history reveals a simple but critical lesson: territorialism might help defendants, but it does so only incidentally and only when courts are willing to enforce borders against the government. In a system where officials can contract around territorial rules and courts treat them as needless formalisms, the borders of criminal law fail to protect those accused of crimes.

Borders seem to bear a closer relationship to sovereignty. In the current criminal justice system, states get to make their own criminal laws; state criminal courts refuse to apply sister states’ “foreign” criminal codes; and the Supreme Court cites variation among state criminal laws and procedures as evidence that states enjoy self-determination. These examples suggest that borders are fairly effective surrogates for the concept of sovereignty. But the idea that states are distinct and autonomous would not disappear if criminal law were less territorial. Arguably, sovereignty is enhanced when a state can criminalize out-of-state acts and contract with other jurisdictions to enforce its criminal code. (No one thinks Germany is not a country because its criminal code travels with its citizens, and its police support INTERPOL.)285See Dubber, supra note 28 (discussing German criminal jurisdiction). Surely one of the main lessons of Part II is that states have expanded their power by ignoring and relaxing territorial rules. From this perspective, territorialism inhibits sovereignty.

The issue here is that “sovereignty” is a slippery term. State sovereignty is not a thing that exists before police arrest people in the government’s name. It is an idea that is constituted by those arrests. Jurisdiction flows from the exercise of power, not the other way around.286See Ford, supra note 24, at 847. See generally Ablavsky, supra note 77; Lauren Benton, A Search for Sovereignty (2010); Lisa Ford, Settler Sovereignty (2010). So it is slightly strange to say that sovereignty demands strict territorial borders. Moreover, it is never quite clear what sovereignty means. Professor Don Herzog recently denounced the concept as “obsolete, confused, and pernicious,” a word that either means “actor with jurisdiction” or “state,” or functions as a “vacuous adjective suitable for trotting out on formal occasions.”287 . Don Herzog, Sovereignty, RIP, at ix–xii (2020); see also Gregory Ablavsky, Empire States: The Coming of Dual Federalism, 128 Yale L.J. 1792, 1797 & n.11 (2019) (collecting the literature “lament[ing] sovereignty as hopelessly imprecise”). Others have tried to redefine sovereignty to keep up with “modern day reality” in which governments cooperate and bargain with each other.288See Bridget A. Fahey, Federalism by Contract, 129 Yale L.J. 2326, 2411–12 (2020) (collecting this literature). Some academics think the idea of state sovereignty is outmoded; others believe the concept still has value.289See Gillian E. Metzger, The States as National Agents, 59 St. Louis U. L.J. 1071, 1073 (2015) (“I don’t think we can really get rid of the concepts of state autonomy and state sovereignty quite as much as [Professor Heather] Gerken wishes.”). One need not resolve this ongoing debate to appreciate that sovereignty and soil are not the same thing. Borders often stand in for the idea that a government has a coherent identity and legal authority. But the mere existence of extraterritorial statutes and choice of law rules does not decimate the state.

Finally, with respect to democracy, is it not at all clear why criminal law has to be territorial to reflect the will of the people. The best version of the democracy-based argument for territorialism is the public law theory outlined above: the criminal code is the pronouncement of a polity, which is defined by legal borders. Territorialism keeps laws contained within those borders and thus forges a link between criminal law and “the people.” There is something attractive about this theory. The idea that crime is a public harm runs deep in American legal culture, so it makes sense to try to line criminal law up with the geographic unit that represents the public.

But it takes an idealized vision of the polity to get this theory off the ground. As recent debates over community policing and “democratizing” criminal law have shown,290See John Rappaport, Some Doubts About “Democratizing” Criminal Justice, 87 U. Chi. L. Rev. 711 (2020) (summarizing the democratization literature). legally defined government units may not be the best representation of collective will about how criminal law should work. Certainly, state lines do not always capture the relevant boundaries of a political community. In any event, even if state borders do reflect “the public,” it remains unclear why it would be undemocratic for democratically elected state legislatures to allow their neighbors to apply their criminal laws. Nor is it apparent why a citizenship-based system of criminal law would be undemocratic—after all, one could argue that citizenship is a better marker of membership in a polity than physical presence on a piece of land.291E.g., Ferzan, supra note 14, at 336.

Ultimately, then, the territoriality principle is not a particularly good proxy for the values it is meant to serve. Territorialism is supposed to protect defendants, sovereignty, and democracy. But borders fail to shield criminal defendants in practice. States don’t lose their autonomy when they cooperate. And it takes a series of artificial assumptions to conclude that democracy requires criminal law to be territorial.

These observations reveal some new possibilities for criminal justice reform. If territorialism is declining and the theoretical defenses of the territoriality principle are weak, one might wonder whether we should develop an even less territorial criminal legal regime. Imagine, for a moment, a version of criminal law built around a capacious conception of legislative jurisdiction and a robust conflicts jurisprudence in which courts evaluated a variety of factors—facts about the crime, the defendant, and the policy interests at stake—to determine which criminal law to apply in a given case. Consider a criminal justice system in which Rhode Island judges interpreted Indiana’s criminal laws and applied Indiana’s sentencing rules. Or a system with even looser venue requirements, in which criminal defendants could move to combine multiple cases in a single proceeding or transfer their cases to a “foreign” court in the name of convenience. These reforms, which amount to permitting a degree of forum shopping by criminal defendants, might provide a check on excessive prosecutorial discretion, which scholars widely agree is one of the American criminal justice system’s main defects.292See, e.g., Rachel Elise Barkow, Prisoners of Politics 143 (2019) (“Meaningful institutional reform must begin with changing the way prosecutors operate.”).

To take the idea even further, imagine regionalizing criminal lawmaking to better reflect the groups most affected by criminal law enforcement. If the borders of criminal law are constructed and negotiable, why not have “the criminal law of Chicagoland”? Or the criminal law of Chicago’s South Side? Lest one protest that these proposals are outlandish, recall from Part II that Indiana and Illinois have created police task forces for the express purpose of overcoming territorial jurisdictional limits, and states already join compacts to harmonize their criminal laws.293See supra Section II.B. (One might also note that, given uneven policing, we effectively do have different criminal law in different neighborhoods.)294See generally James Forman, Jr., Locking Up Our Own (2017); Elizabeth Hinton, From the War on Poverty to the War on Crime (2016). It is worth asking why redistricting criminal law is unthinkable when legislatures use legal fictions to criminalize extraterritorial acts, courts can adjudicate cases involving out-of-state conduct, police can ignore territorial boundaries, and state prison officials can outsource their prisoners.

These are different sorts of proposals. Some reconsider which borders to use in criminal law, while others reject territorialism altogether. The preceding pages have shown both that domestic criminal law contains nonterritorial theories of jurisdiction and that the American criminal justice system is less state-based than doctrine and dogma would suggest. In other words, criminal law need not be territorial, and, where it is, it need not rely on state lines. One could build from either insight to advance criminal justice reform.

2. The Promise of Borders

But perhaps it is unwise to hasten territoriality’s decline. There are some real virtues to a criminal justice system with clear geographic boundaries. Think, for instance, about the upsides of a legal regime built around state borders. Allowing state governments to make their own criminal laws—and then requiring those laws to stay within state lines—constrains criminal law imperialism. In a territorial legal system, Texas cannot prevent its residents from getting abortions in New York.295Bigelow v. Virginia, 421 U.S. 809, 824 (1975). A restriction on out-of-state abortion should be impermissible in any event. Scholars have argued persuasively that a standard conflicts analysis leads to the conclusion that states cannot prevent out-of-state abortions. E.g., Brilmayer, supra note 148; Kreimer, supra note 148. Others have explored a range of constitutional doctrines that might protect the right to choice. See supra note 149. The point here is simply that criminalizing extraterritorial abortion would be impermissible, without question, in a territorial criminal law regime. Given a right to travel between states and the means to do so,296Shapiro v. Thompson, 394 U.S. 618 (1969). this approach to criminal law protects individual liberty. It also protects New Yorkers insofar as they have different policy views on abortion than Texans and have chosen to live in a state whose laws enshrine those views. To use the language of public choice, the territoriality principle minimizes the negative externalities of a government’s decisions about how to define and police crime.

For the same reasons, territorialism promotes policy variation, particularly around whether certain conduct should be criminal at all. Although there is widespread agreement that some acts are crimes—murder is the classic example here—that consensus disappears when it comes to contested topics like low-level drug crimes and immigration offenses. There is even wider debate about the appropriate punishments for these crimes.297See, e.g., Forman, supra note 294; Gabriel J. Chin, Illegal Entry as Crime, Deportation as Punishment: Immigration Status and the Criminal Process, 58 UCLA L. Rev. 1417 (2011). The territoriality principle means that Kansas and Colorado can disagree about marijuana, and a Kansan can go to Colorado to try out a different criminal code.

On some level, these are just the standard reasons to like federalism.298See, e.g., Heather K. Gerken, Our Federalism(s), 53 Wm. & Mary L. Rev. 1549, 1566–67 (2012). These arguments are not unique to criminal law, but they are especially compelling in a field where legal rules have moral overtones, policy disagreements are sharp, and sanctions are harsh. We should be most concerned about promoting policy variation and the ability to vote with one’s feet when the stakes of law are at their apex.

There are also more fundamental reasons to like territorialism in criminal law, whether or not it tracks state lines. As Part I noted, predicating criminal law on territory is a relatively egalitarian way to decide who is subject to criminal liability.299See Ford, supra note 24, at 890–91 (noting that the American conception of territorial jurisdiction was designed to promote egalitarianism and was in this respect central to the project of “political liberalism”). This Article began by observing that the key jurisdictional question in a territorial legal regime is where something happened rather than who committed or suffered a harm. As a result, territorial criminal law is agnostic about identity and status. Questions like whether an alleged criminal is a U.S. citizen or a legal resident are irrelevant in a territorial criminal law regime. And potential victims are protected by virtue of their presence, no matter when they arrived or how long they plan to stay.

Of course, territorialism is only relatively egalitarian. Territory is not apolitical; borders are laced with the history of colonialism and oppression.300See, e.g., Ford, supra note 286; Ablavsky, supra note 287, at 1824–27, 1856–61; Ford, supra note 24, at 868–72, 888–95. To the extent that a criminal legal regime relies on borders, it reifies that history. But there is no uncontested way to define membership for the purpose of criminal law, and we need some way to decide when criminal law applies and whom it protects. The most common alternative to physical presence is citizenship status.301See, e.g., Ryngaert, supra note 45, at 104–10; Dubber, supra note 28, at 250; Perkins, supra note 46. That alternative—a regime in which the criminal code applies to citizens and travels with them—is far more status conscious and far less egalitarian than a system built around the territoriality principle, and it raises thorny questions about the government’s authority to subject noncitizens to criminal laws.302See Yaffe, supra note 18; Zedner, supra note 12. In short, in a society where citizenship status is racialized and everyone deserves the law’s protection, geography may be the least problematic metric for criminal law.

Finally, and most hopefully, territorialism could be a tool to curb abuses in the criminal justice system. Part II did not offer much cause for optimism about borders; it portrayed a legal system in which the territoriality principle rarely constrains the police power. But one has to wonder if territorialism could be more effective. In tracing the decline of the territoriality principle, Part II offered a roadmap for using borders to reform criminal law. Consider all the ways territoriality has been undermined and all the ways it might be revived. Courts could reinvigorate the Due Process Clause to prohibit extraterritorial arrests and restrict extraterritorial punishment.303See Kaufman, supra note 234 (defending a presumption that states should confine the prisoners they convict). They could read the Extradition Clause to permit states to police each other’s criminal justice systems. They could use any number of different constitutional provisions to narrow the conditions under which states can criminalize out-of-state conduct. At the very least, government officials could interpret the due process guarantee to require transparency and public deliberation about the permissible reach of domestic criminal laws.

These reforms would restore some balance to the territorialism in American criminal law. At present, the territoriality principle is asymmetric: state governments invoke territorialism to justify their power to make and enforce criminal laws, but the principle does little to temper that power. Criminal law would be more fair—more coherent and restrained—if the territoriality principle worked both ways. Imagine if the territoriality principle not only authorized Tennessee to make criminal laws but also prohibited Tennessee police from arresting people in Georgia. Consider a version of territorialism that permitted Hawaii to punish people but then prevented Hawaii from outsourcing its prisoners to Kentucky. In this account of criminal law, territorialism licenses and limits state power. Borders are both the fount of government authority and a restraint on officials’ discretion. Academics frequently argue that criminal law enforcers have too much discretion.304See, e.g., Alexandra Natapoff, Criminal Municipal Courts, 134 Harv. L. Rev. 964, 1041–43 (2021) (“[I]nformality poses special dangers to criminal adjudication. . . . [F]ormalism and the legality principle with all their imperfections still play an outsized role in regulating substantive justice . . . .”). Perhaps one problem is that the territoriality principle is underenforced.

The intuition driving these suggestions is that formalism is beneficial in criminal law. These days, it can be strange to call for more formalism in any area of law. The realist turn in the American legal academy over the past half-century has made formalist arguments seem old-fashioned and slightly tone-deaf.305See Emma Kaufman, The New Legal Liberalism, 86 U. Chi. L. Rev. 187, 198–203 (2019) (reviewing Justin Driver, The Schoolhouse Gate (2018)) (discussing the rise and impact of legal realism). In other branches of public law, including constitutional and administrative law, formalism is often critiqued as insincere and undue.306See, e.g., Gillian E. Metzger, The Supreme Court, 2016 Term—Foreword: 1930s Redux: The Administrative State Under Siege, 131 Harv. L. Rev. 1, 36 n.215 (2017). See generally Cass R. Sunstein, Must Formalism Be Defended Empirically?, 66 U. Chi. L. Rev. 636 (1999). But informality is uniquely destructive in the criminal legal system.307See Natapoff, supra note 304, at 1041–42. American criminal justice is dominated by ad hoc arrangements, loose administration, and unfettered discretion. In this context, formalism—even stilted and “hopelessly arbitrary”308Ford, supra note 24, at 850 (“The boundaries that define territorial jurisdictions are a legal paradox because they are both absolutely compelling and hopelessly arbitrary.”). constraints like state lines—is ameliorative.

Indeed, territorial boundaries might be just the sort of structural limits that could discipline American criminal justice. In other fields, scholars and reformers have turned to structural concepts to advance egalitarian aims. In immigration law, for instance, advocates have made preemption and commandeering claims to protect immigrants’ rights when equal protection law has proven inadequate to the task.309See Kaufman, supra note 106, at 1387 (describing the effort to use structural concepts to protect immigrants’ rights in debates over, among other things, sanctuary cities and the Deferred Action for Childhood Arrivals (DACA) program). See generally Hiroshi Motomura, The Curious Evolution of Immigration Law: Procedural Surrogates for Substantive Constitutional Rights, 92 Colum. L. Rev. 1625 (1992). Critics of the American criminal justice system might make a similar move, looking to the territoriality principle to constrain criminal law imperialism and law enforcement excess where arguments based on equality, liberty, and criminal procedure rights have failed.

My own inclination is toward this latter path—toward embracing territorialism as a means to regulate domestic criminal law. But I may be on the wrong side of history. The modern criminal justice system may simply be too functional, too contractual, and too interdependent to revive old ideas about criminal jurisdiction. That is certainly one conclusion to draw from Part II. For those who seek to tame the harshness of American criminal justice, there may be more mileage in abandoning or redrawing the boundaries of criminal law than in enforcing them. The real point is that all of this is possible. The territoriality principle is a norm, not a rule. And the uncritical assumption that criminal law has to be territorial has obscured some interesting and open questions about the future of the field.

Conclusion

Territorialism is an organizing principle in domestic criminal law. The presumption that a government’s power to criminalize and punish conduct stems from its authority over some piece of land dates to the founding. But over time, that presumption has weakened and warped in ways that raise foundational questions about the scope and legitimacy of American criminal law. The basic terrain of criminal law has shifted. In response, we have to ask why borders matter and whether we can sustain a criminal legal system without a clear sense of the source and limits of state power.


* Associate Professor, New York University School of Law. For helpful comments and suggestions, thanks to Greg Ablavsky, Rachel Barkow, Will Baude, Maggie Blackhawk, Mary Bosworth, Rick Brooks, Adam Cox, Kevin Davis, Justin Driver, Antony Duff, Barry Friedman, David Garland, Andrew Hammond, Don Herzog, Carolyn Hoyle, Marcel Kahan, Lucy Kaufman, Daryl Levinson, Ian Loader, Matt Lockwood, Troy McKenzie, Liam Murphy, John Rappaport, Julian Roberts, Sam Scheffler, Steve Schulhofer, Tom Schmidt, Linda Silberman, Lucia Zedner, and participants in the All Souls Seminar Series at the University of Oxford, the N.Y.U. Colloquium in Legal, Political, and Social Philosophy, the N.Y.U. Furman Scholars Seminar, and law faculty workshops at the University of Chicago, the University of Connecticut, Seton Hall, and Wayne State. For excellent research assistance, thanks to Gaelin Bernstein, Jessica Coffrin-St. Julien, Rachel Ellman, Rebecca Hutchinson, Elaina Marx, Evan Ringel, Matt Rosenthal, and Eli Watkins. For incisive and generative edits, thanks to the editors of the Michigan Law Review.