Double Jeopardy Research Paper

A double jeopardy principle has been part of Western legal systems for thousands of years. The Code of Hammurabi, for example, in the nineteenth century B.C.E. sought to prohibit judges from changing judgments (law 15). The Greek philosopher Demosthenes said in 355 B.C.E. that the ‘‘laws forbid the same man to be tried twice on the same issue.’’ In the Roman Republic, an acquittal could not be appealed. St. Jerome in A.D. 391 interpreted a passage from the Old Testament to mean that not even God judges twice for the same act.

The English common law principle that there should be one punishment for one crime first manifested itself during the confrontation between King Henry II and St. Thomas Becket that occurred between 1164 and 1170. Henry, the great-grandson of William the Conqueror, enacted a law that required punishment in the king’s court of clergy who had already been punished in the church courts. In opposing this law, Becket relied on St. Jerome’s principle forbidding more than one judgment for the same act. After four of Henry’s knights killed Becket, the pope condemned Henry’s provisions permitting the double punishment of clergy. Henry relented and today, over eight hundred years later, courts still condemn double punishment.

The evolution of double jeopardy law from the twelfth century to today cannot be easily summarized, but the great English commentator Sir William Blackstone could state confidently in 1765 that there was a ‘‘universal maxim of the common law of England, that no man is to be brought into jeopardy of his life more than once for the same offence.’’ This ‘‘universal maxim’’ led directly to the Fifth Amendment double jeopardy clause, which is strikingly similar to Blackstone’s statement of the common law maxim. The Fifth Amendment provides: ‘‘nor shall any person be subject for the same offence to be twice put in jeopardy of life or limb.’’

There are two difficult concepts embedded in the arcane language of the double jeopardy clause—‘‘same offense’’ and ‘‘twice in jeopardy.’’ ‘‘Same offense’’ could be read literally to be the very same statutory offense—for example, the premeditated murder of V is the same offense as the premeditated murder of V but would not be the same offense as manslaughter of V even though manslaughter is a lesser form of homicide. From at least the seventeenth century, however, courts and commentators understood ‘‘same offense’’ in a broader way. Unfortunately, the precise outlines of this broader meaning have long been elusive. In part this is because ‘‘same offense’’ issues were not very troubling in Blackstone’s day. The common law recognized a relatively small number of criminal offenses and, aside from homicide, the offense definitions rarely overlapped.

The ‘‘twice in jeopardy’’ issue was also easy in the eighteenth century. If a defendant was acquitted or convicted of murder, he could not again be tried for that murder. Unlike the eighteenth-century English system, modern criminal procedure permits the judge sometimes to dismiss cases before a verdict is rendered—the jury might not reach the required vote (almost all jurisdictions require unanimous verdicts in criminal cases), or the case might be dismissed during the trial for some reason. Errors that could justify dismissing the case after trial begins include the failure of the prosecution’s key witness to appear, a remark made by the prosecutor or defense counsel that prejudices the jury, and a mistake made in the indictment that cannot be corrected. Most dismissals during a trial are called ‘‘mistrials.’’

Mistrials

In Blackstone’s day, a verdict was required for the double jeopardy principle to operate, but this was probably because outcomes short of a verdict were virtually unknown. The current standard for deciding when a mistrial is equivalent to a verdict, drawn from the 1824 case of United States v. Perez, is whether the first trial ended because of ‘‘manifest necessity.’’ If so, the first trial does not erect a double jeopardy bar to a second trial. If there was no ‘‘manifest necessity’’ to end the first trial, then a second one is a forbidden second jeopardy.

Three general principles can be drawn from the ‘‘manifest necessity’’ cases. First, if the defendant requests the mistrial and the judge grants it, this will almost always constitute manifest necessity for ending the first trial. This defendant can be retried. Second, if the judge decides that the jurors have been unfairly prejudiced—for example, by hearing something they should not have heard—the judge’s decision to terminate the trial will almost always constitute manifest necessity. This defendant, too, can be retried. Third, if the first trial ends because the jury might have acquitted—such as when the prosecution’s chief witness did not appear—there is no manifest necessity. If the state fails to produce enough evidence at trial, the defendant is entitled to an acquittal. This defendant cannot be retried.

Other kinds of reasons can lead to a mistrial—for example, one judge granted a mistrial because his mother-in-law died unexpectedly. In these miscellaneous cases, courts balance the reason for the mistrial, including how carefully the judge considered other alternatives, against the unfairness of asking a defendant once again to defend the criminal charge. In the case where the trial judge’s mother-in-law died suddenly, the appeals court held that there was no manifest necessity for the mistrial, in part because the judge did not consider asking another judge to take his place. The double jeopardy clause thus barred a second trial.

Multiple Punishment

There is another ‘‘twice in jeopardy’’ issue, one that may sound odd to the ear. Is a defendant placed twice in jeopardy if he is tried only once but convicted of two offenses that are the same offense? Courts have long assumed that it is double jeopardy to convict a defendant twice of the same offense whether the convictions occur in one trial or two. If the rule were otherwise, the prosecutor could often circumvent the double jeopardy clause by trying both offenses in a single trial (a procedure that would not have been available to prosecutors in the eighteenth century).

The linguistic oddness of finding that a single trial can be double jeopardy may explain why courts have developed the terminology ‘‘multiple punishment’’ to explain what the double jeopardy clause forbids in a single trial. As the U.S. Supreme Court has stated on numerous occasions, the clause offers three protections in addition to the ‘‘manifest necessity’’ principle—it ‘‘protects against a second prosecution for the same offense after acquittal. It protects against a second prosecution for the same offense after conviction. And it protects against multiple punishments for the same offense’’ (Brown v. Ohio).

Using this three-part description of double jeopardy protection, one way to understand the double jeopardy clause is that it constrains judges and prosecutors. If the prosecutor brings more than one charge for the same offense in a single trial, the judge can enter but one conviction. If the prosecutor follows a conviction or acquittal with another charge for the same offense, the judge is obligated to dismiss the second charge. The Supreme Court put the matter this way in Brown v. Ohio, one of its most important double jeopardy cases:

[T]he Fifth Amendment double jeopardy guarantee serves principally as a restraint on courts and prosecutors. The legislature remains free under the Double Jeopardy Clause to define crimes and fix punishments; but once the legislature has acted courts may not impose more than one punishment for the same offense and prosecutors ordinarily may not attempt to secure that punishment in more than one trial.

The multiple punishment issue sometimes arises when the legislature has ordered consecutive sentences for violations of more than one criminal statute. For example, a Missouri statute created an offense of ‘‘armed criminal action’’ to punish the use of a dangerous weapon to commit a felony. This statute stated that any sentence imposed under it ‘‘shall be in addition’’ to the punishment for the felony that was committed using the dangerous weapon. Is this explicit indication of legislative intent significant in deciding whether the consecutive sentences are multiple punishment? Yes, the Court held in Missouri v. Hunter. The presence of clear legislative intent to punish offenses consecutively means that consecutive sentences are not multiple punishments within the meaning of the double jeopardy clause regardless of how much the offense definitions overlap.

Second Prosecution After Conviction

The prosecutor can bring a second prosecution after a conviction unless the charges are for the ‘‘same’’ double jeopardy offense. Same offense issues arise when multiple criminal violations occur during a single criminal ‘‘transaction.’’ For example, R uses a knife to rob V. When another person, V2, attempts to prevent the robbery, R pulls a gun and threatens V2 with the gun, then also robs him. This defendant might have committed four criminal offenses— robbery of V1, robbery of V2, assault on V2 (based on the threat with the gun), and the offense of carrying a gun without a license. A prosecutor who wanted to charge all four offenses must know whether any of them are the same offense for purposes of double jeopardy.

This issue has caused the Supreme Court considerable trouble and is still at least partly unresolved. Since the time of Blackstone, it has been accepted wisdom that two different offenses are the ‘‘same’’ if one is necessarily included in the other—if proving the greater always proves the lesser. To use Blackstone’s example, a conviction of manslaughter bars a later trial for murder because manslaughter is a necessarily included offense of murder. Applying this principle, the Supreme Court held in Brown v. Ohio that auto theft is the same offense as joyriding because proving auto theft (taking a car without permission and with intent to steal) always proves joyriding (taking a car without permission). The theory here is that a lesser included offense is simply a species of the greater offense.

Some commentators have criticized this principle on the ground that the defendant who is first prosecuted for the lesser offense is never in jeopardy for the additional culpability manifested in the greater offense (the intent to steal, for example, required for auto theft but not for joyriding). The Court’s rationale seems to be that the prosecutor can choose to try the greater offense first. If, instead, the prosecutor chooses to try the lesser, the state is stuck with that choice.

Blackstone’s lesser-included offense understanding of ‘‘same offense’’ worked well for two hundred years. In 1889, the Supreme Court applied a version of Blackstone’s test in In re Nielsen and first clearly relied on the lesser-offense test in the 1932 case of Blockburger v. United States. The test is known today as the Blockburger test and is usually stated as follows: when the same criminal conduct violates more than one statute, offenses are different if each requires proof of an element that the other does not. If each requires proof of an element the other does not, then neither can be included within the other.

The Blockburger test answers the earlier robbery example. Robbery of V1 is not the same offense as robbery of V2 because the two robberies are based on different conduct. R could have stopped after robbing V1; when R does not stop, he has committed two robberies. On the facts of the hypothetical case, robbery is based on the same conduct as carrying a gun without a license, but these offenses are not the same offense because robbery does not require the use of a gun. But the threat of the gun that constituted assault on V2 is the same offense as robbery of V2 because robbery does require proof of force or threat of force.

Although the test is both relatively easy to apply and grounded in Blackstone’s wisdom, changes in U.S. criminal law have created difficulties for the Blockburger test. Today there are many overlapping, complex criminal offenses, and the same conduct will often violate two, three, four, or more criminal statutes. Modern statutes tend to be complex, and many require distinct elements. As early as 1958, well before the various ‘‘wars’’ on drugs, a single sale of narcotics violated nine different federal statues, each of which required an element that the others did not—for example, sale not in the original package, sale without a prescription, and sale knowing of unlawful importation. In Gore v. United States, the Court held that these three narcotics offenses could be punished consecutively. What remained unclear after Gore was whether separate trials could be based on a single sale of narcotics.

Reacting against the unfairness of multiple trials based on the same conduct, the Supreme Court in the 1970s began to suggest that there might be a greater protection against successive prosecutions than against multiple punishment in a single trial. The Blockburger test, the Court seemed to say, told us how many punishments were permitted but not how many trials. In 1990 in Grady v. Corbin, the Court held that successive prosecutions required a ‘‘same conduct’’ understanding of ‘‘same offense.’’ In addition to the Blockburger inquiry that focused on offense definitions, Grady read the double jeopardy clause to forbid a trial for any criminal charge that required proof of ‘‘conduct that constitutes an offense’’ of which the defendant had already been convicted. For example, manslaughter by auto would be the same offense as drunk driving if the defendant had already been convicted of drunk driving and the manslaughter required proof of the same drunk driving.

The rule proved difficult to apply and, perhaps more importantly, was difficult to justify. As Justice Antonin Scalia sarcastically asked in his dissent in Grady, how could the double jeopardy clause words ‘‘same offense’’ mean one thing when there was a single trial and something very different when successive prosecutions were involved? The Court abandoned the ‘‘same conduct’’ definition of same offense in United States v. Dixon, decided only three years after Grady. In Dixon, the Court held that there is only one definition of same offense—the Blockburger lesserincluded offense definition. If the criminal statutes themselves do not define the same offense when the elements are compared, it does not matter how often the same conduct is reprosecuted. Drunk driving would not be the same offense as manslaughter by auto if the latter offense could be proved by other kinds of reckless behavior even if drunk driving was the reckless conduct that killed the victim in the case being prosecuted.

Dixon did not solve all the same offense problems, however. For one thing, the five Justices who voted to overrule Grady disagreed among themselves about how to apply Blockburger to the complex statutes in Dixon. For another, there might still be a ‘‘same offense’’ difference between multiple punishments in a single trial and a second trial after conviction. Recall the Missouri v. Hunter rule that a specific legislative requirement of consecutive sentences made the punishments not multiple. But this does not necessarily solve the problem of multiple trials. One way to frame the question is whether the legislature can, by simply stating its intent to create separate offenses, make offenses not the ‘‘same’’ for purposes of successive prosecutions as well as for the multiple punishment doctrine. If, as the Court suggested in Brown, the legislature is free to ‘‘define crimes and fix punishments,’’ perhaps the legislature can create separate offenses under the double jeopardy clause by simply stating its intent to do so.

But the Court has never suggested that the multiple punishment principle from Missouri v. Hunter would extend into the successive prosecution context. Indeed, one member of the current Court, Justice Scalia, has argued just the opposite—that the multiple punishment doctrine is analytically separate from the successive prosecution doctrine. In the single trial context, the argument goes, the legislature can rebut the result of the Blockburger test because the rebuttal merely makes clear how many penalties the legislature intended to authorize, but the double jeopardy clause forbids the legislature to authorize more than one trial for the same offense as defined by the Blockburger test. This issue remains unresolved.

Second Prosecution After Acquittal

When the first trial ends in an acquittal, there can be no second prosecution for the same offense. In this way, acquittal and conviction provide the same double jeopardy bar. But the Court has expanded the role of the double jeopardy clause to protect acquittals even when the offenses are not the same offense. In Ashe v. Swenson, masked men robbed five poker players. When the prosecutor tried Ashe for robbing one of the players, the evidence that Ashe was one of the robbers was weak, and the jury found Ashe not guilty of that robbery. The prosecutor then tried Ashe for robbing another player. This time the eyewitnesses seemed more certain that one of the masked men was Ashe; the eyewitness who was least certain at the first trial was not called to testify. Ashe was convicted of this robbery.

The same offense rule is that different conduct gives rise to different offenses. Robbery of one victim is never the same offense as robbery of a second. Thus, Ashe could get no help from the same offense doctrine. If he had been convicted of robbing the first poker player, he could have been tried later for robbing the second one.

But the acquittal provided a broader ban against a second trial. The Court noted that the only issue in the first trial was whether Ashe was one of the masked men, which the first jury determined in Ashe’s favor. The Court held that the state could not force Ashe to defend that issue again. To permit the state to bring a prosecution for a different victim would, in effect, permit the second jury to overrule the first. It would also encourage prosecutors to structure later cases to hide evidence that turned out to be favorable to the defendant in the first trial. Forcing a defendant repeatedly to defend the same basic issue, while the state’s case gets better and better, can only increase the likelihood that innocent defendants will be convicted.

Appeals

Just as was true in the Roman Republic, a conviction today can be appealed and reversed, but an acquittal is final and cannot be appealed. As with the Ashe principle discussed in the last section, one justification is that appeal of an acquittal creates too much risk that an innocent defendant will be worn down by the superior resources of the state. A justification from outside the double jeopardy clause is that permitting an appellate court to reverse a jury’s acquittal would violate the Sixth Amendment right to trial by jury (this justification does not explain why acquittals by judges are also non-appealable).

While the jury should have the final say in deciding the facts that underlie an acquittal, the bar of prosecution appeal seems less persuasive when the trial judge has made an error that keeps some important fact from the jury. Suppose the trial judge suppressed a confession that was clearly admissible. The jury’s acquittal in this situation is based on incomplete information. In 1937 the Supreme Court in Palko v. Connecticut upheld the constitutionality of a state process that permitted the prosecution to obtain a new trial by appealing an acquittal infected by legal errors. The doctrinal framework of Palko was rejected in 1969 in Benton v. Maryland, however, and most commentators believe that the double jeopardy clause does not permit a prosecution appeal even on the ground of legal error.

Appeals are therefore tilted in favor of the defendant. A guilty verdict can be appealed and reversed, but an acquittal, even if clearly wrong, can never be reversed on factual grounds and probably not on any other ground. This is perhaps a fair price to pay to ensure that innocent defendants are not convicted after repeated trials and appeals.

Lower Courts

Thousands of lower court cases have applied the Blockburger lesser-included offense test to federal offenses and to offenses from all fifty states. Assault with intent to murder, for example, is a different offense from assault with a dangerous weapon (each requires proof of an element that the other does not). Burglary, which requires entry into a structure, is a different offense from larceny committed inside the structure (one can commit burglary without committing larceny and vice versa). But larceny is the same offense as grand larceny (larceny of property over a certain value), and assault is the same offense as assault with intent to rape or assault with intent to murder.

Lower courts generally recognize that Blockburger is just a presumption when applied to multiple punishment in a single trial. For example, Blockburger often pronounces different kinds of homicide offenses to be different double jeopardy offenses. The offense of homicide by auto is not the same Blockburger offense as intentional murder. The latter requires proof of intent to kill while the former requires proof that the killing was done by auto. Blockburger thus permits two homicide convictions for one killing (an intentional killing by means of an auto). Perhaps, however, the number of homicide offenses is better correlated with the number of victims than the number of superficially distinct statutes.

Dozens of lower courts have wrestled with this issue. Most have concluded, by one means or another, that the legislature did not intend to authorize two homicide convictions for killing a single victim. These courts thus use actual legislative intent to rebut the presumption about intent that is created by Blockburger.

Dual Sovereignty

Although ‘‘dual sovereignty’’ is really a variation of the same offense issue, it is usually treated separately. Suppose the federal Congress and a state legislature have identical criminal statutes. Can a defendant be charged and convicted (or acquitted) of an offense in federal court and then tried in state court? What if the defendant is first tried in state court? This issue is called ‘‘dual sovereignty’’ because the original thirteen states were separate political entities until they joined the federal union and gave up some of their sovereignty to the federal government. The states did not give up their right to define and punish crimes.

This issue, and its dual sovereignty implication, was recognized by the Supreme Court in the 1820 case of Houston v. Moore, but it has only been in the last few decades that the issue affected very many defendants. Congress initially did not create many criminal offenses and there was little overlap between federal and state criminal law. But there has been an explosion of federal criminal law in the last twenty years, and many defendants now potentially face successive state and federal prosecutions.

When the issue was first noted in Houston, the various opinions of the Supreme Court laid out the two basic approaches to the problem. Justice Joseph Story argued that it would violate double jeopardy for both sovereigns to prosecute the same offense, which he took to mean the same criminal conduct. Justice William Johnson saw the matter differently—it was not a question of prosecuting the same conduct but the same offense. Because each U.S. citizen owes allegiance to two sovereigns, the same conduct that violates state and federal criminal law was two offenses, in Johnson’s view, not one.

Johnson’s view ultimately prevailed. It is not double jeopardy for a defendant to be acquitted of federal bank robbery charges and then tried and convicted in state court for the same bank robbery. Nor is it double jeopardy for a defendant to be convicted in state court and then convicted in federal court. These cases drew a stinging dissent in Bartkus v. Illinois from Justice Hugo Black, who wrote: ‘‘If double punishment is what is feared, it hurts just as much for two ‘Sovereigns’ to inflict it than for one. If danger to the innocent is emphasized, that danger is surely no less’’ when the successive trials are brought by different sovereigns.

The dual sovereignty doctrine is controversial, but there are not very many instances of successive state and federal prosecutions. Both the federal and state governments have imposed limits on their ability to re-prosecute the same conduct. The federal limit is found in a Department of Justice policy that generally forbids prosecuting conduct that has already been prosecuted. There are exceptions for cases in which justice was not done in the prior prosecution—for example, the judge or prosecutor was corrupt or the jury entered an acquittal that was clearly against the evidence. More than half the states have enacted statutes that generally forbid a state prosecution to be based on the same conduct as an earlier federal prosecution. Although there is much to commend in Justice Black’s rejection of the dual sovereignty doctrine, the federal and state systems have adjusted to minimize the potential harm.

Bibliography:

1. AMAR, AKIL REED. ‘‘Double Jeopardy Law Made Simple.’’ Yale Law Journal 106 (April 1997): 1807–1848.
2. BLACKSTONE, WILLIAM. Commentaries on the Laws of England (1765), vol. 4. Chicago: University of Chicago Press, 1979.
3. CASSELL, PAUL ‘‘The Rodney King Trial and the Double Jeopardy Clause: Some Observations on Original Meaning and the ACLU’s Schizophrenic Views of the Dual Sovereignty Doctrine.’’ UCLA Law Review 41 (February 1994): 693–720.
4. FRIEDLAND, MARTIN Double Jeopardy. Oxford, U.K.: Clarendon Press, 1969.
5. GUERRA, SANDRA. ‘‘The Myth of Dual Sovereignty: Multijurisdictional Drug Law Enforcement and Double Jeopardy.’’ North Carolina Law Review 73 (March 1995): 1159–1210.
6. HERMAN, SUSAN ‘‘Double Jeopardy All Over Again: Dual Sovereignty, Rodney King, and the ACLU.’’ UCLA Law Review 41 (February 1994): 609–647.
7. KING, NANCY ‘‘Portioning Punishment: Constitutional Limits on Successive and Excessive Penalties.’’ University of Pennsylvania Law Review 144 (November 1995): 101–196.
8. KIRCHHEIMER, OTTO. ‘‘The Act, the Offense and Double Jeopardy.’’ Yale Law Journal 58 (March 1949): 513–544.
9. MOORE, MICHAEL Act and Crime. Oxford, U.K.: Clarendon Press, 1993.
10. POULIN, ANNE BOWEN. ‘‘Collateral Estoppel in Criminal Cases: Reuse of Evidence after Acquittal.’’ University of Cincinnati Law Review 58, no. 1 (1989): 1–57.
11. POULIN, ANNE BOWEN. ‘‘Double Jeopardy: Grady and Dowling Stir the Muddy Waters.’’ Rutgers Law Review 43 (summer 1991): 889–931.
12. RUDSTEIN, DAVID ‘‘Double Jeopardy and the Fraudulently-Obtained Acquittal.’’ Missouri Law Review 60 (summer 1995): 607–651.
13. SCHULHOFER, STEPHEN ‘‘Jeopardy and Mistrials.’’ University of Pennsylvania Law Review 125 ( January 1977): 449–539.
14. SIGLER, JAY Double Jeopardy: The Development of a Legal and Social Policy. Ithaca, N.Y.: Cornell University Press, 1969.
15. SHELLENBERGER, JAMES, and STRAZZELLA, JAMES A. ‘‘The Lesser-Included Offense Doctrine and the Constitution: The Development of Due Process and Double Jeopardy Remedies.’’ Marquette Law Review 79 (fall 1995): 1–193.
16. STRAZZELLA, JAMES ‘‘The Relationship of Double Jeopardy to Prosecution Appeals.’’ Notre Dame Law Review 73 (November 1997): 1–30.
17. THOMAS, GEORGE, III. ‘‘Sentencing Problems Under the Multiple Punishment Doctrine.’’ Villanova Law Review 31 (September 1986): 1351–1428.
18. THOMAS, GEORGE, III. Double Jeopardy: The History, the Law. New York: New York University Press, 1998.
19. ‘‘Twice in Jeopardy.’’ Yale Law Journal 75 (December 1965): 262–321.
20. WESTEN, PETER. ‘‘The Three Faces of Double Jeopardy: Reflections on Government Appeals of Criminal Sentences.’’ Michigan Law Review 78 (June 1980): 1001–1065.
21. WESTEN, PETER, and DRUBEL, RICHARD. ‘‘Toward a General Theory of Double Jeopardy.’’ Supreme Court Review (1978): 81–169.