Concept
Ashe v. Swenson in Action
Collateral estoppel
Ashe v. Swenson in Action
The reasoning in Ashe v. Swenson rests on what the record of the first trial contained, so the facts are best taken in the order the two trials occurred. In the early hours of January 10, 1960, six men were playing poker in a basement in Lee's Summit, Missouri, when three or four masked men armed with a shotgun and pistols broke in, robbed each player of money and personal property, and fled in a car belonging to one of the victims.1 Four men, Ashe among them, were arrested that morning and charged with seven separate offenses: the armed robbery of each of the six players and the theft of the car.2 Because each victim supplied a separate charge, the State could try the same four men more than once, each time under a different player's name, and the case asked whether the Constitution places any limit on that sequence.
Ashe went to trial in May 1960 on the charge of robbing Donald Knight.3 The State called Knight and three of the other players, and each described the holdup and itemized his own losses, so the proof that a robbery had occurred and that Knight had lost property was unassailable.4 The proof that Ashe was one of the robbers was weak. Two witnesses thought there had been only three robbers and could not identify him, a third, who was Ashe's uncle by marriage, could say only that Ashe's voice "sounded very much like" that of one of the robbers, and the fourth identified him only by "size and height, and his actions."5 The defense cross-examined briefly and aimed its questions at the weakness of the identifications. It made no attempt to question the testimony about the holdup or the losses, offered no testimony of its own, and waived final argument.6
The trial judge told the jury that if it found Ashe to be one of the participants in the armed robbery, theft of any money from Knight would sustain a conviction, and that a participant was guilty even if he had not personally robbed Knight.7 The jury, which was not instructed to explain its verdict, found Ashe "not guilty due to insufficient evidence."8 That phrase reports that the State's proof fell short without saying where, and the gap matters because the Court had to decide what the jury resolved without any statement from the jury itself. The record supplied the answer. With the robbery uncontested and the instructions making participation enough for guilt, the only question left for the jury to decide against the State was whether Ashe was one of the robbers.9
Six weeks later Ashe was tried for robbing Roberts, another of the players.10 He moved to dismiss on the strength of his acquittal, and the motion was overruled.11 The witnesses were mostly the same and the instructions were virtually identical, but the testimony on identity was now substantially stronger, because witnesses who had been hesitant at the first trial testified with greater assurance.12 Ashe was convicted and sentenced to 35 years in the penitentiary, and he then sought habeas corpus relief on the ground that the second prosecution had put him twice in jeopardy.13
In deciding whether the second trial was permitted, the Court rejected a hypertechnical reading of the first verdict. Collateral estoppel in criminal cases is applied with "realism and rationality," and where the earlier acquittal rested on a general verdict, as is usually so, the court examines the record of the prior proceeding, including the pleadings, evidence, and charge, and asks whether a rational jury could have grounded its verdict on an issue other than the one the defendant seeks to foreclose.14 A stricter test would amount to rejecting the rule in every case that ends in a general verdict of acquittal, since such a verdict never states its own grounds.15 Because the inquiry is a sequence of questions put to the record, it can be run in order.
Run on the first trial, the steps produce one answer. The second prosecution required proof that Ashe was one of the robbers. The record of the first trial contained no indication that the jury could rationally have found that an armed robbery had not occurred or that Knight had not been a victim of it, so the single rationally conceivable issue was whether Ashe had been one of the robbers, and the jury's verdict found that he had not.16 The second trial was therefore an attempt to litigate again an issue that the first jury had resolved in Ashe's favor.17
The change of victim did not alter that conclusion. The Court framed the question as whether, after a jury had determined by its verdict that Ashe was not one of the robbers, the State could constitutionally hale him before a new jury to litigate that issue again, and it set aside as beside the point whether Missouri could validly charge six offenses or impose six punishments after a single trial.18 The reasoning runs through the charge the State could no longer bring. Once a jury had found, on conflicting testimony, a reasonable doubt that Ashe was one of the robbers, the State could not have tried him again for robbing Knight, and it could not have offered the same or different identification evidence in the hope that another jury would find it more convincing. The Roberts trial was constitutionally no different, because the victim's name had no bearing on whether Ashe was one of the robbers.19 The trap in applying the case is treating a change of victim as a change of issue, since the two charges were different offenses that turned on the same contested fact.
The Court closed by rejecting the State's own account of what it had done. In its brief the State frankly conceded that after the acquittal it had treated the first trial as "no more than a dry run" for the second prosecution, reasoning that a good attorney who lost would refine the presentation in light of the first trial. The Court answered that this is precisely what the constitutional guarantee forbids.20 The record shows the mechanism at work: the witnesses whose identifications had been hesitant gave firmer ones the second time, so the weakness that produced the acquittal had been repaired at the defendant's expense.21 If a lost trial could serve as a rehearsal, the State could keep retrying a case, correcting whatever each defeat exposed, until some jury agreed with it, and the acquittal would carry no finality at all.
The single changed fact is the range of issues the first jury could rationally have decided. In Ashe the robbery was uncontested and the instructions made participation sufficient for guilt, so one issue remained and the verdict could be read as deciding it. In the altered version two issues are live, and a general verdict does not reveal which one the jury resolved. Chief Justice Burger's dissent pressed that very objection, that a general verdict leaves open whether the jury acquitted on a ground other than identity, and the next section takes up how courts read a prior verdict when more than one ground is open.22 Justice Brennan, concurring, would have gone further and required the prosecution to bring all charges arising from one transaction in a single trial. The majority did not adopt that rule, and Justice Harlan's concurrence said so expressly, so the joinder question is left for the discussion of charging decisions later in this chapter.23