Concept
Prosecutions Versus Punishments
Same offense and lesser-included offenses
Prosecutions Versus Punishments
The sections so far have asked one question, whether two charges define the same offense, and have answered it with the elements comparison, the rule for lesser-included offenses, and the legislature's unit of offense. The question carries so much weight because the Double Jeopardy Clause is usually described as three separate protections. It guards against a second prosecution for the same offense after an acquittal, against a second prosecution for the same offense after a conviction, and against multiple punishments for the same offense.1 The first two operate when the government tries to bring a second case, and the third operates when a single court imposes more than one punishment. Each of them comes into play only when the offenses are the same, so the same-offense inquiry is the gate every double jeopardy claim must pass, and the setting of the claim determines what a finding of sameness accomplishes once the gate is open.
The two settings share one yardstick because of how the test developed. The comparison of statutory elements was announced in Blockburger v. United States, which raised a multiple-punishment claim, and the Court later carried it into the successive-prosecution setting, which is why the version applied to second trials is called the Blockburger-Brown test after Brown v. Ohio.2 Under that test, two statutes are the same offense unless each requires proof of a fact that the other does not.3 The test is therefore a single comparison of two charges, used both to decide whether a defendant can be punished separately on each and to decide whether a defendant can be prosecuted later after a conviction or acquittal involving the same incident.4 Because one comparison produces the answer, offenses that count as the same for barring consecutive sentences at one trial also count as the same for barring a later prosecution.
The Justices have not always agreed that one definition serves both settings, but the disagreement should not be overstated. In United States v. Dixon, five Justices apparently agreed that the definition of same offense should be identical in the two contexts, although no opinion commanded a majority of the Court.5 A dissenting opinion argued for a different definition in each context, on the ground that the protection against successive prosecutions serves different purposes from the protection against multiple punishments.6 The decisions still show differences of detail between the settings, and the next two sections take each in turn. Blockburger remains the common starting point, and the differences lie in what the Clause protects in each setting.
The purposes differ because the harm differs. After an acquittal, a second prosecution gives the government a rehearsal: it learns from the first trial where its case was weak and tries again, so the first verdict never becomes final. After a conviction, a second prosecution lets the government seek a heavier punishment for the same wrong than the first judgment imposed. In both situations the protection works by closing the courthouse door on the second case, which is why a finding of sameness operates as a bar. At a single trial no second case exists, and the prosecutor may charge overlapping offenses together. The only risk is that the court will add up sentences beyond what the legislature authorized, so the Clause polices the total punishment rather than the number of charges.
Students go wrong when they let the setting change the comparison. The elements comparison is the same whether the claim arises from a second trial or from consecutive sentences, and the setting changes only what the finding of sameness does: it bars the second prosecution in the one case and raises the question of legislative authorization for stacked punishments in the other. The next section begins with successive prosecutions, where the Clause serves a policy of finality.
Successive Prosecutions
The same-offense comparison does its most familiar work when the State tries to bring a second case, and in that setting the Clause is best understood as a guarantee of finality. In Brown v. Ohio the Court explained that where successive prosecutions are at stake, the guarantee serves a constitutional policy of finality for the defendant's benefit, protecting the accused from relitigation of facts underlying a prior acquittal and from attempts to secure additional punishment after a prior conviction and sentence.8 Those are two different dangers, and each follows from the structure of a second trial. After an acquittal, a second prosecution gives the government a chance to study the first trial, repair the weak parts of its case, and try again, so no acquittal would ever settle anything.9 After a conviction and sentence, a second prosecution lets the State ask a new factfinder for more than the first judgment imposed. In both situations the State has the larger resources, and a rule that left it free to keep trying would let its persistence, rather than the legislature's penalty scheme, set the defendant's total exposure.
Brown takes the Blockburger comparison, which was devised for the problem of cumulative sentences, and applies it to later trials without change. The Court reasoned that if two offenses are the same under that test for purposes of barring consecutive sentences at a single trial, they necessarily are the same for purposes of barring successive prosecutions.10 In the Court's words, "Where the judge is forbidden to impose cumulative punishment for two crimes at the end of a single proceeding, the prosecutor is forbidden to strive for the same result in successive proceedings."11 The logic runs through the division of roles. The legislature decides how much punishment a wrong may carry, and a judge who cannot add a second sentence to the first at one sitting has been told the legislature authorized only one. If the prosecutor could split the same matter into two trials, the total punishment would reach the same forbidden figure by a back door.
The facts of Brown show the carryover at work. Nathaniel Brown stole a car in East Cleveland on November 29, 1973, and nine days later was caught driving it in Wickliffe, where he pleaded guilty to joyriding and received thirty days in jail and a $100 fine.12 After his release he was indicted in Cuyahoga County for auto theft and for joyriding on November 29, the date of the original taking.13 Under Ohio law joyriding was the lesser included offense of auto theft, so the Court held the two to be the same offense: a prosecutor who proves auto theft has necessarily proved joyriding, and the lesser offense requires no proof beyond what the greater requires.14 Brown's first case had run through to conviction and sentence, so the second prosecution was exactly what the finality policy forbids, an effort to obtain additional punishment for the same offense after the first judgment was fixed.
The bar also has a prerequisite that the elements comparison cannot supply. The Clause forbids a person to be "twice put in jeopardy," so a second prosecution is barred only if the defendant was in jeopardy the first time.15 The rule for when jeopardy attaches belongs to a separate reading, and a short statement is enough here. In a jury trial it attaches when the jury is empaneled and sworn, and in a bench trial when the first witness is sworn, because those are the moments at which the defendant first faces the risk of a guilty verdict.16 A proceeding that ends before that point does not stop the government from prosecuting the same case again.17
Because a successive-prosecution claim has a threshold and then a comparison, it is best run in an order that puts the cheap question first.
Attachment comes first because a failure there ends the analysis no matter how closely the two offenses overlap. Students lose this issue when they find that two charges are the same offense and announce a bar without asking whether jeopardy attached in the first case. The next section turns to the other setting, a single trial in which no second case exists and the question is how much punishment the legislature authorized.
Cumulative Punishment at One Trial
When the State tries both charges together, no second case exists for the Clause to close off, so the same-offense inquiry has a narrower job. In Brown v. Ohio the Court described it: where consecutive sentences are imposed at a single trial, the guarantee is limited to assuring that courts do not exceed legislative authorization by imposing multiple punishments for the same offense.19 That limit follows from the division of roles between the branches. The legislature defines crimes and fixes their penalties, and a sentencing court carries out that schedule. A sentence therefore offends the Clause only when it carries the defendant past what the legislature authorized, which makes the question at a single trial one of legislative intent. That is why the elements comparison works differently here than it does before a second trial.
In this setting the Blockburger comparison operates as a presumption about what the legislature intended. Two violations that are the same offense under the elements test are presumed not to carry separate punishments, on the assumption that a legislature that writes overlapping statutes and says nothing more does not mean to punish one wrong twice.20 In Albernaz v. United States the Court treated the test as a method of ascertaining legislative intent when nothing more concrete is available, and it regarded the question of what punishments are constitutionally permissible as the same question as what punishment the legislature intended.21 A presumption suits the task because overlapping statutes are usually silent about how they combine, and a court needs some neutral way to fill that silence. The elements comparison supplies one, and it costs the legislature nothing, since a clear statement of a different intent displaces it.
Blockburger itself arose at a single trial, and it shows the other side of the presumption. The defendant there was convicted on three counts of selling morphine to the same purchaser and was sentenced to five years in prison and a $2,000 fine on each count, with the prison terms running consecutively.22 Two of those counts charged the same sale, one as made outside the original stamped package and the other as made without a written order from the purchaser.23 Because each provision required proof of a fact the other did not, the Court held that the one sale had produced two offenses, and the consecutive sentences stood.24 The Court added that if the penalty seemed too harsh, the remedy had to come from Congress rather than from judicial construction.25 Different offenses leave a sentencing judge free to stack penalties, because the legislature has defined two crimes and attached a penalty to each.26
The presumption gives way when the legislature speaks clearly. In Missouri v. Hunter the Court held that where a legislature specifically authorizes cumulative punishment under two statutes, the court's task of statutory construction is at an end, and the prosecutor may seek and the trial court or jury may impose cumulative punishment in a single trial, whether or not the two statutes proscribe the same conduct under Blockburger.27 The defendant there received separate sentences for armed robbery and for the related enhancement offense of armed criminal action, even though all the elements of one offense were included in the other.28 The legislature, not the court, decides how much punishment a wrong carries, so a clear statement that a robber who uses a weapon is to be punished under both provisions answers the only question the Clause leaves open at a single trial. Intent may appear in the legislative history or in the language and organization of the statutes, and courts generally infer authorization when the offenses sit in different statutes or distinct sections and each unambiguously sets out its own punishment.29 The same power lets a legislature divide a continuous course of conduct into separate offenses, which is the unit-of-offense point taught earlier.30
The outcome for felony murder and its underlying felony shows how much depends on the legislature. Harris bars a separate later trial for the underlying felony except in extraordinary circumstances.31 When both are charged in one prosecution, however, the answer turns on legislative authorization, and the courts of some states have found that cumulative punishment is authorized while the courts of others have found the authority lacking.32
When no authorization appears, the court must decide what to do with the extra conviction, and the answer is to vacate it. A defendant convicted at one trial of two counts that make up the same offense, such as a greater and a lesser included offense, cannot keep both convictions.33 The Court applied that rule to a defendant convicted of a federal conspiracy offense and of the continuing criminal enterprise offense. The enterprise offense required proof of a conspiracy that would also violate the conspiracy provision, so the conspiracy offense contained no element missing from the enterprise offense. Under the elements comparison it was therefore a lesser included offense of the enterprise charge.34
The sentence on the lesser conviction ran concurrently, and the government argued that a second conviction with a concurrent sentence is no punishment at all. The Court disagreed for two reasons. A federal statute required a $50 special assessment for every conviction and the assessment had been imposed on both, so a second conviction amounted to a second punishment.35 A second conviction on the record could also delay parole eligibility, increase a sentence under a recidivist statute for a later offense, be used to impeach the defendant's credibility, and carry the stigma of any conviction.36 Concurrency fixes how much time is served, but the conviction goes on producing consequences after the sentence is over. The government's last argument, that Congress meant to allow both punishments, also failed, because the Court adhered to the presumption that Congress authorized only one, and one conviction with its concurrent sentence had to be vacated.37
The focus on total punishment also explains a resentencing decision. In Jones v. Thomas a defendant who had fully served the shorter of two consecutive terms learned on appeal that separate sentences for both offenses were impermissible, and the Court found no double jeopardy bar when the trial judge vacated the shorter sentence and credited the time served against the longer one, because the credit kept the total punishment within what the legislature had authorized.38 Set Hunter beside the conspiracy and enterprise case and one variable stands out. In each, one offense was included in the other, so each pair was the same offense under the elements comparison. The sentences survived in Hunter because the legislature had specifically authorized them, and a conviction fell in the enterprise case because nothing showed that Congress meant to authorize a second punishment.
Students lose this issue when they stop at the elements comparison and announce that cumulative sentences are barred, because express legislative authorization overrides the presumption that same-offense counts carry a single punishment.39 The reverse error is to treat a concurrent sentence as a cure, which overlooks that the second conviction still counts as punishment.40