Concept
Who Counts as a Sovereign
Dual sovereignty
Who Counts as a Sovereign
The source-of-power test asks where a government's authority to punish comes from. The federal government and each state pass it easily, because each exercises governmental power in its own right.1 The harder questions concern bodies that prosecute crimes but fit less neatly into the federal-state pattern: Indian tribes, municipalities, territories, and the federal government itself when it enforces a law that another sovereign wrote. In each of these the label does not answer the question. Calling a body a city, a tribe, or a territory says nothing about whether it defines and punishes crime on its own authority or on authority lent by someone else. If labels controlled, a government could change the double jeopardy consequences of its prosecutions simply by choosing how to describe its subunits. The doctrine avoids that result by treating a subunit as part of the sovereign it derives its power from, however the sovereign classifies it.2
Each of the four subsections below runs the same two steps: identify the source of the entity's power to define and punish crime, then state the result that follows. A tribe's power is its own, so a tribe may prosecute regardless of what the federal or state governments undertake.3 A municipality's power is the state's power exercised through a subunit, so a prosecution for a state crime bars a prosecution for a municipal violation, and the reverse is also true.4 The third subsection applies the same question to territories. The last situation inverts the usual setup, because one government, the United States, is the prosecutor both times, yet the two proceedings enforce the laws of different sovereigns, and the sovereign source of a law is enough to make its offense a different offense.5 The first subsection takes up tribes, because they show that the set of separate sovereigns reaches beyond the federal government and the states.
Indian Tribes
The federal-state and state-state cases might suggest that only the federal government and the states can be separate sovereigns, but the source-of-power test has no such limit. The test asks where a prosecuting entity's authority to punish comes from, so any government whose power is its own can qualify, whatever its place in the constitutional structure.7 Indian tribes are the clearest illustration, and the decision that settled the point is United States v. Wheeler.
In that case a member of the Navajo Tribe pleaded guilty in tribal court to contributing to the delinquency of a minor and was sentenced. A federal grand jury then indicted him for statutory rape arising out of the same incident.8 The district court dismissed the indictment on the theory that tribal courts and federal district courts are not arms of separate sovereigns, and the Supreme Court reversed.9 One incident had been prosecuted twice, which is the pattern the Double Jeopardy Clause seems built to prevent, so everything depended on whether the tribe and the federal government count as two sovereigns or as one.
The Court treated the source of a tribe's power to punish tribal offenders as the controlling question: is that power part of inherent tribal sovereignty, or is it an aspect of federal sovereignty that Congress delegated to the tribes?10 Tribes keep the attributes of sovereignty that have not been withdrawn by treaty, by statute, or by implication from their dependent status. The Court therefore looked at the Navajo Tribe's own history, reading its treaties with the United States and the federal statutes on jurisdiction over crimes by Indians. From those materials it concluded that the Tribe had never given up its power to punish its own members and that the power was not attributable to any delegation of federal authority.11 When a tribe punishes a member for violating tribal law, then, it acts as a separate sovereign and not as an arm of the federal government, and the tribal and federal prosecutions are not for the same offence within the meaning of the clause.12 Heath later described the tribe's prosecutorial power as derived from its "primeval sovereignty" rather than from a delegation of federal authority.13
The delegation question matters because it determines how many sovereigns are present. Had Congress handed the tribe its criminal power, the tribe would be exercising federal authority when it punished a member, both prosecutions would rest on one sovereign's power, and the second would be a successive prosecution by the same sovereign. The Court found no such delegation for the Navajo Tribe, so that route to a bar was closed. The contrast case is a territory such as Puerto Rico, whose authority to prosecute comes from the same source as the federal government's, and a later subsection takes it up.14
The Court also gave a practical reason that explains why the rule runs this way. Limiting dual sovereignty to the federal government and the states, as the defendant urged, would have produced "undesirable consequences": a tribal prosecution for a relatively minor offense would bar a federal prosecution for a much graver one.15 The mechanism is straightforward. A tribal prosecutor picks the charge for reasons rooted in tribal law and tribal circumstances, and the federal government has no voice in that choice. If the choice could shield the defendant, a tribe's decision to proceed on a lesser charge would fix the ceiling on federal enforcement of a separate federal law, much as Georgia's plea bargain would have capped Alabama's prosecution in Heath. Heath itself relied on Wheeler for the proposition that the sovereign set is not confined to the federal government and the states, and it restated the test as "the ultimate source of the power under which the respective prosecutions were undertaken."16
The same logic that permits a federal prosecution after a tribal one permits the reverse sequence. Because tribes exercise independent sovereign power, they can prosecute regardless of what the federal or state governments may also undertake.17 The reasoning of Abbate and Bartkus, which allowed a federal prosecution after a state one and a state prosecution after a federal one, carries over to the tribal setting.18 On an exam, a stem that stresses a tribe's reliance on federal treaties and statutes invites the argument that the tribe is a federal instrumentality. Students lose this issue when they treat a tribe's dependent status as proof that its prosecutorial power is federal. The analysis asks whether this tribe retained the power to punish its members or received it by delegation, and for a tribe like the Navajo the answer is retention.
Municipalities
The tribal cases showed that the set of separate sovereigns reaches beyond the federal government and the states, and municipalities mark the opposite edge of the same question. A city has its own ordinances, its own court, and its own prosecutor, yet it is not a separate sovereign from the state that created it. That result reverses the direction of the sections before this one. In the federal-state, state-state, and tribal situations, the dual sovereignty doctrine took the Double Jeopardy Clause out of the analysis and the second prosecution went forward. Here the doctrine does not apply, so the clause operates in the ordinary way and bars a second prosecution by the same sovereign when it is for the same offense.20
In Waller v. Florida a defendant was convicted of violating two city ordinances after he removed a canvas mural from city hall and then scuffled with police.21 The state then convicted him of grand larceny under state law for the same conduct, and the Supreme Court held that the second prosecution was barred.22 The Court reasoned that “the judicial power to try petitioner … in municipal court springs from the same organic law that created the state court of general jurisdiction,” and so the dual sovereignty doctrine did not apply.23 Organic law here means the foundational law from which governmental authority comes, which for a city and a state court alike is the state's own. The municipal court and the state court are therefore two channels for one source of power, and when the city punished the defendant, the state's power was the power being used.
The result follows from what an offense is. The clause bars a second prosecution for the same offense, and an offense is a breach of the law of a particular sovereign.24 A city ordinance is enforced under authority that the state has granted, so a violation of it breaches the state's law as exercised through a local body and is not an offense against some other sovereign. Municipalities and similar subdivisions therefore exercise the state's sovereign authority and hold none of their own, which is why a state prosecution for a state crime bars a later prosecution under a municipal law, and a municipal prosecution likewise bars a later state prosecution for the same offense.25 The same picture appears elsewhere in constitutional law, where municipalities are treated as instrumentalities through which states carry out their policies.26
The fact that moves the answer is where each body's authority to punish comes from. In Heath v. Alabama Georgia convicted the defendant of murder and Alabama then prosecuted him for the same killing, and the second prosecution was permitted because each state is sovereign with respect to the other.27 The Navajo Tribe in Wheeler likewise retained its own sovereign power to punish offenses against tribal law, so a federal prosecution could follow a tribal one.28 Each body in these comparisons has its own laws, its own courts, and its own prosecutors, so those features cannot be what separates the cases. The deciding fact is whether the body's authority to punish is its own, as it is for a neighboring state or a tribe, or is lent to it by another government, as it is for a city.
The rule holds however the state describes its subunits, because subunits take their power to exist from the sovereign.29 The reason runs through the state's own incentives. If the state's label controlled, a state could split one accusation into a minor municipal charge and a serious state charge, try them in sequence, and treat each as the act of a different sovereign. The defendant would face two trials by what is, in legal substance, a single government, which is the repeated prosecution the clause exists to prevent. A test built on the source of power closes that route, since a state cannot change the source of a city's authority by renaming the city.
On an exam, a stem that describes a city police department, a municipal court, and a city attorney, followed by a state charge for the same act, invites the argument that two governments have acted. Students lose this issue when they treat a separate prosecutor and a separate court as proof of a separate sovereign. The analysis asks where the city's authority to punish comes from, finds that it comes from the state, and then applies the ordinary same-offense question. If the two charges are the same offense, the second prosecution is barred, and if they are different offenses, the dual sovereignty doctrine was never the obstacle and the usual same-offense tests decide the matter.30
Territories
The tribal cases showed that a government outside the federal-state pair can be a separate sovereign, and territories show the test working in the other direction: a government with its own courts, its own criminal code, and its own prosecutors can still fail it. A U.S. territory such as Puerto Rico is not a separate sovereign from the federal government. Both draw their authority to prosecute from the same source, so they are not distinct sovereigns for Fifth Amendment purposes, and prosecutions by both for the same offence are not permitted.32
The decision is Puerto Rico v. Sanchez Valle, which holds that Puerto Rico cannot prosecute a person for an offense the federal government has already charged.33 Set it beside Bartkus v. Illinois, where a state prosecuted conduct that had already been prosecuted in federal court and was allowed to proceed.34 In both settings a government with its own laws and its own courts moves against a defendant whom the federal government has already pursued, so those features cannot explain the different results. The difference lies in where the second government's power comes from. A state's power to punish is its own, and each sovereign has an independent right to punish violations of its laws, which is why the doctrine admits the second prosecution.35 A territory's power to prosecute, by contrast, traces to the federal government itself.36
The reason follows from what the doctrine assumes. An offence is a breach of the law of a particular sovereign, so a defendant who has violated the laws of two sovereigns has committed two offences, and a prosecution by each is not a second trial for the same one.37 That reasoning needs two sovereigns on the stage. When a territory prosecutes, it exercises power that traces to the federal government, so the federal government's right to punish is the right being used on both occasions. The independent interest that would justify a second trial then has nothing to rest on, and the two prosecutions amount to one sovereign trying the defendant twice. The relationship parallels the one between a city and its state in the preceding subsection: a municipality's authority comes from the state and a territory's comes from the federal government, so each is a channel for another government's power and neither holds power of its own.
On an exam, a stem that gives a territory its own legislature, courts, and prosecutors invites the argument that it is a separate sovereign from the federal government. Students lose this issue when they treat a territory's own courts and criminal code as proof of separate sovereignty. The analysis asks where the territory's authority to prosecute comes from. Once the answer is the federal government, the dual sovereignty doctrine drops out, because it applies only when two sovereigns are involved, and the ordinary same-offence question decides whether the second prosecution is barred, a question that looks to the statutory elements of the two charges and not to the underlying conduct.38
Same Prosecutor, Different Laws
The preceding subsections asked whether the government bringing the second prosecution is a separate sovereign from the one that brought the first. Denezpi v. United States changes that variable. The United States was the prosecuting sovereign in both proceedings, but the first proceeding enforced the Code of the Mountain Ute Tribe, which is Indian law, and the second enforced the Major Crimes Act, which is United States law.40 Justice Barrett called this “a twist on the usual dual-sovereignty scenario,” because a single sovereign was enforcing the laws of two.41 The question was whether enforcing the federal law after the tribal code violated the Double Jeopardy Clause.
The Court held that the second prosecution was allowed.42 The Clause protects a person from being twice put in jeopardy for the same offence, and the Court read that language to look at the offence and not at the underlying conduct or action.43 The source of a law is a feature distinctive enough to make its offence different from another sovereign's offence, even when the elements of the two laws are equivalent.44 A tribal code and a federal statute trace to different sovereigns, so they define different offences. It follows that two prosecutions by one government are not forbidden merely because there are two of them, and the same person, the same prosecuting government, and closely matching elements do not by themselves produce a bar.
The result follows from the premise on which the whole doctrine rests. An offence is a breach of the law of a particular sovereign, so one act that violates the laws of two sovereigns is two offences, each owed to a different lawmaker.45 That premise concerns the law that was broken and says nothing about which officials carry the case into court. In Heath and Bartkus the two prosecutors happened to belong to two sovereigns, so the offences differed and the doctrine permitted both prosecutions. Denezpi removes that coincidence and the offences still differ, because the tribal code and the federal statute each define a wrong against their own source of authority. If the identity of the prosecutor controlled, the answer would turn on which government's officers happened to bring the case, even though the number and character of the wrongs the defendant committed had not changed.
Set beside the municipal cases, the decision shows the test working in both directions. In Waller the city and the state used different prosecutors and different courts, yet the ordinance and the state statute both drew their force from the state, so the successive prosecutions were barred.46 In Denezpi one government prosecuted twice, yet the two laws drew their force from different sovereigns, so the second prosecution went forward.47 The variable that decides both cases is the source of the law being enforced. The number of prosecutors and the number of courts do not decide either one.
On an exam, a stem that has the United States prosecuting twice invites the argument that one sovereign is trying the defendant twice, so the clause must bar the second case. Students lose this issue when they treat the identity of the prosecutor as the test. The sound analysis asks whose law each proceeding enforced. If the first proceeding enforced another sovereign's law, such as a tribal code, and the second enforced federal law, the offences differ and the second prosecution is allowed even if the elements match. If both proceedings enforced the same sovereign's law, the ordinary same-offence rules decide, and a repeat of the same offence is barred.48