Retrial After Other Mistrials

Retrial After Other Mistrials

The hung jury is the cleanest example of a mistrial, but trials end before verdict for many other reasons, and each of them raises the same retrial question. A mistrial is a termination of the trial by the judge before the judge or jury has reached a verdict, and it may be declared at either side's request or on the court's own motion.¹ The causes vary widely: jurors may be exposed to inadmissible evidence that makes a fair trial impossible, a key witness may become unavailable, or the court may discover after the trial has begun that the indictment is fatally defective.² A judge who calls the order a mistrial rather than a dismissal ordinarily means to leave the government free to try the case again, yet the label does not control the result. A ruling that resolves some or all of the factual questions bearing on guilt in the defendant's favor is an acquittal whatever the judge calls it, and an acquittal bars further prosecution.³

A second trial after a mistrial touches the interests the Double Jeopardy Clause protects. The defendant must again bear the time, money, and stress of a trial, and depending on when the mistrial came, the government may already have tested its case, repaired its weak points, and learned what the defense intends to do.⁴ Those costs explain why courts are reluctant to grant mistrials. The practical result nonetheless favors retrial, because a large share of mistrials rest on a jury's inability to agree or on trial error that makes a fair verdict impossible, so mistrials that bar a second trial are the exception rather than the rule.⁵ The reasoning is that a mistrial after jeopardy attaches overrides the defendant's interest in a verdict from the jury first chosen, and that interest yields in some cases to the public's interest in fair trials that end in just judgments.⁶ The law therefore asks whether the override was truly needed. If the trial could not produce an impartial verdict that would survive appeal, a new trial serves public justice. If the judge or prosecutor ended the trial for convenience, or to improve the government's position, a second trial would hand the government the repeated attempt the Clause forbids.

The analysis forks on a single question: did the defendant acquiesce in the mistrial? If the defendant opposed it, retrial is permitted only if the judge's decision met the manifest necessity standard. If the defendant requested the mistrial or consented to it, retrial is generally permitted unless the prosecution goaded or provoked the defendant into seeking it.⁷ The manifest necessity standard comes from United States v. Perez, decided in 1824, which held that a new trial after a mistrial declared without the defendant's request or consent is allowed when there is a manifest necessity for the mistrial or the ends of public justice would otherwise be defeated.⁸ The fork follows from who took the case away from the first jury. When the government or the court ends the trial over the defendant's protest, someone other than the defendant has given up the verdict, and that party must justify the choice. When the defendant makes the request, the defendant has chosen to forgo a verdict from that tribunal, and the manifest necessity test does not apply at all.⁹

The Supreme Court's decisions give manifest necessity concrete content, and Illinois v. Somerville shows the standard being met. Early in the trial the court discovered that the indictment was fatally defective. State law did not allow the defect to be cured by amendment, and any conviction on that indictment would have been automatically reversed. The trial judge declared a mistrial over the defendant's objection, and the Supreme Court held that retrial was permissible.¹⁰ The reason lies in the shape of the doomed trial. Had it continued, an acquittal would have barred any new prosecution despite the defect, while a conviction would have been thrown out on appeal, so the proceeding could end in a result that bound the government but never in one that would stand against the defendant. Somerville described two situations that justify a mistrial: an impartial verdict cannot be reached, or a conviction could be reached but would have to be reversed because of an obvious procedural error. This case fell in the second.¹¹

This hypothetical and Somerville both involve a mistrial over objection, so the objection alone does not decide anything. What differs is whether the first trial could still have produced an impartial verdict that would survive appeal. In Somerville it could not, and in the burglary trial it could, so the delay is a reason of convenience and carries none of the weight that a doomed trial does. On an exam, students go wrong by treating the word mistrial as the answer, because the label does not decide whether retrial is allowed. The first move is to ask who requested or consented to the mistrial, and only if the defendant opposed it does the inquiry turn to manifest necessity. The next three sections take up the standard itself, the procedure a court must follow before declaring a mistrial, and the special rule for mistrials the defense requests.

The Manifest Necessity Test

When a defendant opposes a mistrial, the government may try the case again only if the judge's decision to end the first trial was justified by manifest necessity.¹⁴ The standard comes from United States v. Perez, an 1824 opinion by Justice Story, which held that a court may discharge a jury without a verdict when there is a manifest necessity for doing so or when the ends of public justice would otherwise be defeated.¹⁵ The Supreme Court has acknowledged that the question turns on the facts of each case and that its decisions resist tidy categorization, but it has still developed guidelines for judging whether a mistrial was proper.¹⁶ Those guidelines are the subject here, beginning with the two core situations and then adding two further factors that sharpen the inquiry.

The test is a standard rather than a bright line because two legitimate interests pull against each other. Once a jury is sworn, the defendant has a constitutional interest in seeing that jury finish the case, because a new tribunal adds financial and emotional burden, prolongs the stigma of an unresolved accusation, and may raise the risk that an innocent person is convicted.¹⁷ The government and the public have their own interest, since a trial depends on the health and availability of witnesses, jurors, lawyers, and parties, on a crowded court calendar, and on the speedy-trial guarantee. In United States v. Jorn the Court reasoned that a mechanical rule forbidding retrial whenever a jury is discharged without the defendant's consent would be too high a price for the added security it would give, and that the defendant's right to a particular tribunal must in some circumstances yield to the public's interest in fair trials that end in just judgments.¹⁸ Illinois v. Somerville added a caution in the other direction: the test is not to be applied mechanically, and due regard must be given to the defendant's interest in having the original jury decide the case.¹⁹ The standard therefore comes into play when a proceeding is cut off before the merits are finally resolved, and the two core situations below are the working framework for deciding whether that cut-off was justified.²⁰

The first core situation is one in which an impartial verdict cannot be reached.²¹ A jury that is genuinely deadlocked is the clearest example, and Somerville also pointed to an earlier decision upholding retrial after a mistrial granted because one of the jurors was acquainted with the defendant and therefore probably prejudiced against the government.²² The reason a mistrial is necessary here lies in what a trial is for. A verdict is worth having only if it comes from a factfinder able to decide on the evidence, so when the factfinder is compromised, continuing the trial produces an outcome that neither side can trust. The consequences are also lopsided: a conviction by a jury biased against the defendant would be vulnerable on appeal, while an acquittal by a jury biased toward the defendant would be final and would deny the government the fair trial it is entitled to seek. Ending the trial protects both sides from a worthless verdict.

The second core situation is one in which a verdict of conviction could be reached but would have to be reversed on appeal because of an obvious procedural error.²³ The Court's examples include a trial in which one of the jurors was not qualified to serve and a trial in which it was discovered after jeopardy attached that the defendant had never entered a plea.²⁴ Somerville itself belongs here. The prosecution's indictment was found early in the trial to be fatally defective, state law forbade curing the defect by amendment, and any conviction would have been automatically reversed, while an acquittal would still have barred a new prosecution despite the defect.²⁵ A trial in that posture can end in only one result that binds anyone, because an acquittal would bind the government and a conviction would bind no one. Continuing would therefore leave the government with all of the risk and none of the reward, and a mistrial restores the case to a point where it can be tried on a sound footing. The qualifier matters, because the category covers errors that are obvious and that would require reversal, which is narrower than any complaint a defendant might later raise on appeal. A mistrial that fits neither core situation is presumably improper, and retrial after it would not be permitted.²⁶

The two categories are a starting point, because Arizona v. Washington made clear that further considerations bear on whether necessity exists.²⁷ One is whether the government or the defendant is responsible for the problem that prompted the mistrial. When the prosecution caused the difficulty, its claim to a second trial is weaker, since otherwise carelessness by the government would earn it a chance to rehearse its case. Even so, government fault is relevant but not determinative: the defect in Somerville lay in the prosecution's own charging document, and retrial was still allowed, particularly because the government gained no special advantage from a second trial.²⁸

The other factor is whether the trial judge considered less drastic alternatives to a mistrial, such as a curative jury instruction or a continuance.²⁹ In Jorn the trial judge ended the trial so that prosecution witnesses could consult attorneys about their privilege against self-incrimination, and the Court found no manifest necessity because the judge had not considered granting a continuance for that purpose.³⁰ The word necessity carries the logic. A mistrial costs the defendant the first jury, which makes it the most expensive remedy available, so a judge who has not tested a cheaper cure has not shown that nothing less would do. The next section takes up the procedure a court must follow before declaring a mistrial, which is how that inquiry is meant to take place.

The hypothetical and Jorn turn on the same variable, which is whether the mistrial was the only remedy the circumstances left. Both involve a mistrial over objection, so the objection alone settles nothing. In the first version and in Jorn a milder step was available and went untried, while in the altered version the judge found that no milder step could restore a fair trial.

Alternatives Before Mistrial

The last factor in the manifest necessity inquiry asks whether the judge considered less drastic alternatives, and it has a procedural counterpart that operates before the judge acts. Federal Rule of Criminal Procedure 26.3 requires the court, before ordering a mistrial, to give each defendant and the government three opportunities: to comment on the propriety of the order, to state whether that party consents or objects, and to suggest alternatives.³³ The rule speaks to each defendant and to the government separately. In a trial with several defendants, each is heard on its own position, and the prosecution has the same chance to be heard as the defense. The three opportunities serve different purposes, and seeing them separately shows how the rule feeds the substantive test from the last section.

The opportunity to comment puts the facts in front of the judge from the people who know the case best, before a step that cannot be taken back. The opportunity to state consent or objection fixes each party's position on the record, and that position matters later because it decides which branch of the retrial analysis applies. A mistrial the defendant opposed permits retrial only if manifest necessity existed, while one the defendant requested or consented to generally permits retrial unless the prosecution goaded the request.³⁴ Without a stated position, a reviewing court would have to reconstruct from a transcript whether the defendant acquiesced. The opportunity to suggest alternatives brings cheaper cures, such as a curative jury instruction or a continuance, before the judge while the trial is still alive and those cures can still be used.³⁵ This last function carries the most weight here, because a mistrial costs the defendant the first jury and is therefore the most expensive remedy on the table.

The link between the rule and the substantive test is clearest in United States v. Jorn. The trial judge there ended the trial so that prosecution witnesses could consult attorneys about their privilege against self-incrimination. The Supreme Court found no manifest necessity and stressed that the judge had never considered a continuance that would have let the witnesses consult counsel while the trial went forward.³⁶ A hearing of the kind the rule describes is where that question gets asked. Someone, whether the defense, the government, or the judge, raises the obvious possibility that a recess would solve the problem. When nobody asks, the record contains no showing that the mistrial was the only remedy left. The party who must justify a second trial is then left to defend necessity without that showing.

Rule 26.3 is a rule of procedure, and it does not itself decide whether retrial is barred. Its text attaches no retrial consequence to compliance or noncompliance.³⁷ The double jeopardy question remains whether manifest necessity existed. The absence of considered alternatives is one factor in that judgment, weighed alongside the two core situations: an impartial verdict that cannot be reached, and a conviction that would have to be reversed for obvious procedural error.³⁸ What the rule contributes is the record. A judge who heard the parties, tested the alternatives, and explained why none would work leaves a transcript that supports necessity. A judge who never considered a curative instruction or a continuance has a hard time showing necessity, because necessity means that nothing less would do.

Both versions involve a mistrial over objection and the same juror exposure, so neither the objection nor the exposure explains the different outcomes. The variable is whether the record shows that the judge looked for a cheaper cure before choosing the expensive one. In the first version the questioning of jurors and the consideration of an instruction and a recess answered that question, and in the second version, as in Jorn, nobody asked it.

Defense-Requested Mistrials

The last two sections dealt with a mistrial the defendant opposed, where retrial depends on the government showing manifest necessity. The position each party states under Rule 26.3 matters because a different rule governs when the defendant is the one asking to end the trial.⁴² When the defendant requests the mistrial, the manifest necessity test does not apply, and retrial is generally permitted.⁴³ A single question explains both that rule and its exception, and the question is whose choice ended the first trial.

The reason lies in what the manifest necessity requirement is for. The standard traces to United States v. Perez and governs a new trial after a mistrial declared without the defendant's request or consent, asking whether there was a manifest necessity for the mistrial or whether the ends of public justice would otherwise be defeated.⁴⁴ The requirement is a check on whoever ended the trial. A defendant has a valued right to have the first jury decide the case, and when the court or the prosecution takes the case from that jury over objection, someone must justify the loss. A defense request removes the premise of that justification, because the defendant has made the very decision the right protects. The Jorn plurality described the defendant's significant interest as an interest in deciding whether to take the case from the jury when events occur that might warrant a mistrial.⁴⁵ A defendant who answers that question by asking for a mistrial has chosen to give up a verdict from the first tribunal, so the State did not impose the second trial.

In United States v. Dinitz the defendant was the one who sought the mistrial, and the Court later cited the case for the point that the manifest necessity test does not apply when a mistrial is declared at the defendant's behest.⁴⁶ Dinitz therefore shows that the identity of the party who asks for the mistrial selects the governing standard. United States v. Scott supplies supporting context from a neighboring setting. The defendant there moved to dismiss two counts because of prejudice and the trial judge granted the motion, and the Court held that when a defendant himself seeks to have the trial terminated without any submission to judge or jury on guilt or innocence, the government's appeal from his success is not barred.⁴⁷ Scott concerns the government's right to appeal rather than retrial after a mistrial, but its logic carries over. A termination the defendant chose, before anyone has resolved the factual elements of the offense, leaves no acquittal for the Clause to protect, and it leaves no complaint that the State forced the defendant to abandon the first tribunal.

The rule has a limit, because a request counts as the defendant's own choice only if the government did not manufacture it. If the prosecutor or the judge goaded the defendant into seeking the mistrial, meaning deliberately pushed the defendant toward making the motion, retrial may be barred, and the same is true when the government acted in bad faith.⁴⁸ The Clause rests on the belief that the State, with all its resources, should not get repeated attempts to convict a person, each one adding expense, ordeal, and anxiety and raising the chance that an innocent person is convicted.⁴⁹ A prosecutor who could provoke a mistrial motion whenever a case was going badly would turn that belief upside down. The first trial would become a rehearsal in which the government learns the defense case and finds the flaws in its own, and the defendant's request would be a response to a trap the government set.⁵⁰ Bad faith or goading therefore marks the line between an ordinary defense decision and one the government created. The trigger is also narrow: the question is whether the prosecutor was intentionally trying to provoke the motion, so an error that merely prompts the defense to ask for a mistrial does not reach it.⁵¹

Both versions share the same prosecution error, the same inadmissible remark, and a defense motion that the judge granted, so none of those facts decides the result. The variable is whether the government intended to provoke the motion. That also separates this section from the opposed-mistrial analysis, where government fault was a relevant factor in judging necessity but was not determinative. After a defense request no necessity inquiry occurs at all, and the government's purpose becomes the question.⁵³ The exam trap is treating government error as goading, since a careless prosecutor and a prosecutor who engineered the motion lead to opposite outcomes.