Concept
The Four-Part Test
Plain error
The Four-Part Test
Rule 52(b) states its standard in a single sentence, and in United States v. Olano the Supreme Court turned that sentence into an ordered series of questions that courts of appeals now put to every forfeited claim.1 The first point Olano settled is what the Rule covers. The version the Court quoted referred to “plain errors or defects affecting substantial rights,” wording that could be read as naming two categories, plain errors on one side and defects affecting substantial rights on the other.2 The Court rejected that reading and explained that “error or defect” is more simply read as “error,” so that Rule 52(b) defines a single category of forfeited-but-reversible error.3 The current text, which speaks only of “a plain error that affects substantial rights,” fits that reading without strain.4 The single-category reading matters because, if there were two categories, an appellant could proceed on a defect that affected substantial rights even though it was not plain, and the plainness requirement would screen out nothing.
Once the category is fixed, the Rule's conditions operate as limits on what the court of appeals may do. A forfeited error “may be noticed” only if it is plain and affects substantial rights, and the court may correct it, either by vacating for a new trial or by reversing outright, only if it meets those criteria.5 The court must still consider the claimed error in deciding whether the judgment should be overturned, but it cannot grant that remedy unless Rule 52(b) applies.6 Olano identified three limitations on that authority and then a fourth, discretionary step, and together they form the Olano four-part test.7 A concurring opinion in United States v. Gaudin restated the same sequence compactly and in the same order: error, plainness, substantial rights, and then the court's discretion.8
The test governs only claims that were forfeited. An error raised by a timely objection goes to harmless-error review under Rule 52(a), a contrast taken up in the third part of this chapter.9 Within Rule 52(b), each step screens out a different kind of claim, which is why none can be skipped. The error step confirms that a legal rule was actually broken and that the defendant did not knowingly give up its protection. A defendant who validly pleads guilty, for example, cannot have the conviction vacated on the ground that there should have been a trial, because the right to trial is waivable and the plea waived it.10 The plainness step confines correction to defects that were obvious enough for the trial judge to catch without being told, so the objection requirement is not hollowed out by subtle claims first discovered on appeal. The substantial-rights step ensures that correction would change something, since a reversal that would produce the same result on retrial costs a second trial and buys nothing.
The fourth step differs from the others because it is permissive. Even a plain, prejudicial error does not oblige the court of appeals to act, because Rule 52(b) says the error “may” be considered and Olano reads that word as leaving the decision to the court's sound discretion.11 That step also adds a requirement beyond prejudice. In theory, an error that would require reversal if it had been preserved may not require reversal under plain-error analysis, although how much more serious the error must be remains unclear in practice.12 The trap for the unwary is that satisfying the first three steps does not entitle the appellant to relief, because the court still decides whether the error warrants correction.
The appellant carries the load throughout. Because Rule 52(b) authorizes a remedy only when its criteria are met, the party invoking the Rule has to bring the claim within each of them.13 The clearest statement concerns prejudice: under Rule 52(b) the defendant rather than the Government bears the burden of persuasion, which reverses the allocation on preserved claims.14 In the ordinary case the defendant must show a reasonable probability that the error affected the outcome of the trial, and an abstract possibility is not enough.15 Courts therefore describe plain-error review as a much higher standard than harmless-error review, and the Eleventh Circuit has called it an extremely stringent form of review that is harder to satisfy than clear-error review.16
The sequence can succeed, and the decision in United States v. Gray shows it doing so. The defendant had been sentenced before United States v. Booker, and the district court had treated the Sentencing Guidelines as mandatory and raised his sentence on a judge-found fact that the jury had not found.17 Gray raised no objection in the district court, so the court of appeals reviewed for plain error, stated the three showings the defendant had to make and the discretionary condition that followed, and found each satisfied. It then vacated the sentence and remanded for resentencing.18 The failure case turns on a single changed fact.
Both defendants stayed silent below, and both cleared the first two steps. What separates them is whether the record shows that the error made a difference to the result.
The four-step sequence is the federal formulation, and state courts phrase their plain-error tests in their own words. The Delaware Supreme Court in Clark v. State, where the defendant had not objected to jury instructions said to comment on the evidence, described plain error as a material defect apparent on the face of the record, basic, serious, and fundamental in character, and so prejudicial to substantial rights that it jeopardizes the fairness and integrity of the trial.19 That is Delaware law, and it reads as a single description of the kind of error that qualifies and not as a series of separate limitations. A question governed by Federal Rule of Criminal Procedure 52(b) is answered with Olano's steps. The next four subsections take those steps in order, beginning with the requirement that there be an error at all.
The Error Requirement
The first limit that United States v. Olano places on a court of appeals' power under Rule 52(b) is that there must be an “error.”22 The definition is short: a deviation from a legal rule is error unless the rule has been waived.23 Two consequences follow from it, and they point in opposite directions. A forfeited claim can still rest on an error, because failing to raise a matter on time does not undo the legal violation. A waived claim rests on none, because the defendant has given up the very rule the claim depends on. Most disputes at this step are disputes over which of those two descriptions fits the defendant's conduct at trial.
Forfeiture is the failure to make a timely assertion of a right, and the error requirement is built so that it survives forfeiture.24 Rule 52(b) speaks of a plain error that may be considered “even though it was not brought to the court's attention.”25 If a missed objection meant that no error occurred, every claim the Rule exists to reach would fail at the threshold, and the Rule would have nothing to operate on. Olano considered the argument that a question never presented to the trial court involves no error and leaves nothing to review, and it rejected that argument as not the theory Rule 52(b) adopts.26 The reasoning separates two questions. Whether the trial court departed from a legal rule is a fact about what happened in the courtroom, while whether the defendant spoke up decides how, and on what terms, the departure may be raised later. So if a legal rule was violated and the defendant did not waive it, there is an error despite the absence of a timely objection.27 Students lose this step when they treat silence at trial as proof that no error occurred, since silence only moves the claim into the narrow channel that Rule 52(b) provides.
In United States v. Griffith the defense preserved its objection to a prosecutor's remark but not its complaint about the judge's curative instruction, and the Fifth Circuit reviewed that second complaint under Rule 52(b).28 During rebuttal argument the prosecutor, annoyed by the defendant's running commentary at the defense table, broke off and asked whether he would like to take the stand and testify. Defense counsel objected and moved for a mistrial, and the district court denied the motion but told the jury to disregard the remark and reminded it of the presumption of innocence and the government's burden of proof. That instruction said nothing about the defendant's constitutional right not to testify, and the defense neither objected to it nor asked for a fuller one.29 The court of appeals found error “without difficulty,” because the prosecutor had violated a constitutional rule by commenting on the defendant's failure to testify and the district court had left that violation uncorrected by never instructing on the privilege.30 It defined error as “a deviation from a legal rule in the absence of a valid waiver,” and the missing objection to the instruction did not change that answer. The silence meant only that the court had to continue through the later steps.31
Waiver is the second half of the definition, and it differs from forfeiture in what the defendant did. Forfeiture is the failure to make a timely assertion of a right, whereas waiver is the “intentional relinquishment or abandonment of a known right.”32 The consequences differ in the same way: a forfeited right can still be considered on appeal if the remaining requirements are met, but a waived right cannot.33 The reason lies in what each state of affairs reveals about the defendant's choice. A forfeiting defendant has failed to act, usually through oversight, and nothing about the lapse shows that the defendant wanted the violation to occur. A waiving defendant chose to give up the right, so honoring that choice is fair, while correcting it would let a party reverse a deliberate decision once it turned out badly. The insistence on intentional relinquishment protects defendants in turn, because it keeps ordinary silence from being converted into a waiver that would foreclose all review.
Olano's own illustration is the guilty plea. A defendant who knowingly and voluntarily pleads guilty in conformity with Rule 11 cannot have the conviction vacated on the ground that a trial should have been held, because the right to trial is waivable and a valid plea waives it, so a conviction without a trial is not error.34 The word “valid” carries weight. In Boykin v. Alabama the trial judge accepted guilty pleas to five robbery indictments without asking the defendant anything, and the Court refused to presume a waiver of the privilege against self-incrimination, the right to a jury trial, and the right to confront accusers from that silent record.35 It treated acceptance of the plea without an affirmative showing that it was intelligent and voluntary as error plain on the face of the record.36 Boykin came from Alabama's automatic review of capital cases more than two decades before Olano, so it is not a Rule 52(b) decision. It does explain why a waiver must be shown and not assumed: a silent record is equally consistent with ignorance or coercion, so it cannot stand in for the knowing choice that waiver requires.
What counts as a valid waiver is not uniform across rights. Whether a particular right can be waived at all, whether the defendant must take part in the waiver personally, whether certain procedures are required, and whether the defendant's choice must be particularly informed or voluntary all depend on the right at stake.37 That variation is why Olano's example specifies a plea that is knowing, voluntary, and in conformity with Rule 11, and it is why a court cannot treat every affirmative word from defense counsel as a waiver. Invited error, an error a party cannot complain of on appeal because its own conduct encouraged or prompted the ruling, is the nearest label for conduct that goes beyond silence.38 In State v. Smith, a New Jersey decision, the prosecutor pointed out that the judge had left the “ordinary hearer” standard out of a terroristic-threat charge, the judge denied omitting it, and defense counsel agreed. The court held that by affirmatively representing that the charge did not contain the error, counsel induced or acquiesced in it, which precluded the defendant from asserting it as a ground for reversal.39 Smith is state law, and it shows what affirmative agreement looks like, but it does not establish that every affirmative statement by counsel is a waiver under Rule 52(b), because that question turns on the right involved.
Courts have long distinguished a deliberate decision not to object from simple oversight, though older decisions drew that line under other tests. Wainwright v. Sykes arose on federal habeas review of a state conviction, and one concurring opinion thought that competent trial counsel could well have chosen not to object to the defendant's in-custody statement, which was consistent with his trial testimony and portrayed him as acting under provocation.40 Justice White's concurrence added that counsel who knows the facts and the law and forgoes an objection as a matter of professional judgment has made a deliberate bypass, while ignorance of the applicable rules does not amount to one.41 Those opinions applied habeas-specific tests of cause, prejudice, and deliberate bypass, so they illustrate the contrast between choice and oversight without supplying Olano's standard.42 Estelle v. Williams, a 1976 decision about a defendant tried in jail clothing, shows how differently the older cases framed the question. The Court found it unnecessary to decide whether the defendant's silence was a defense tactic or simple indifference, because either way the silence precluded any suggestion of compulsion.43 That inquiry asked whether the State had compelled the defendant's appearance, so it did not ask whether a known right had been intentionally abandoned, which is the question Olano puts at this step.
The error requirement therefore screens out two kinds of claims: those in which no rule was broken and those in which the defendant gave up the rule that was broken. What remains is a violation of a rule the defendant kept, even if silence at trial let it pass unremarked, and that claim moves to the next question, whether the error is plain, meaning clear or obvious under current law.44
The Plainness Requirement
Once a court of appeals has found an error that the defendant did not waive, Olano's second limitation asks whether the error is “plain.” The Court treated that word as synonymous with “clear” or, equivalently, “obvious.”46 The minimum content of the requirement is that a court of appeals cannot correct an error under Rule 52(b) unless the error is clear under current law.47 That formulation directs the court to the state of the law when the question arose, which is a narrower inquiry than whether the trial judge was wrong. An error can therefore satisfy the first step, because a legal rule was violated, and still fail here because the rule's application to those facts was not obvious.
The demand for clarity follows from what Rule 52(b) is correcting. In the ordinary case the law is clear at the time of trial, so the defendant is on notice of a duty to object, and silence both forfeits the claim and deprives the judge of the chance to repair the problem while repair is still possible.48 Holding the defendant to that silence also deters a lawyer from strategically withholding an objection in order to seek reversal on appeal if the verdict goes badly.49 If an arguable error could be raised for the first time on appeal and still succeed, that deterrent would disappear, because a lawyer could take the verdict and press any debatable ruling afterward, and trial judges would answer for rulings that were reasonable when made. Limiting relief to obvious errors reserves the Rule for oversights that a competent judge should have caught unprompted, a point courts often put by asking whether the judge should have known better even if the lawyer did not.50
The Fifth Circuit in United States v. Griffith, drawing on earlier Supreme Court authority, treated an error as plain when it is so conspicuous that the trial judge and prosecutor were derelict in allowing it to stand even without the defendant's help in spotting it.51 The evidence rules express the same idea in Federal Rule of Evidence 103(e), which reaches an error so obvious that no formal objection should have been needed to alert the trial court.52 In Griffith the prosecutor, irritated by the defendant's running commentary at the defense table, asked whether he would like to take the stand and testify, and she told the court of appeals that she knew at once the remark was improper.53 The district court recognized the violation as well, rebuking her and telling the jury to disregard the comment, so the proper remedy should have been as obvious as the violation: a Fifth Amendment violation calls for an instruction on the privilege, and general instructions on the burden of proof and the presumption of innocence do not take its place.54 Griffith still lost, because he could not show an effect on substantial rights, which is why a plain error does not by itself earn relief.55
Other circuit decisions reach the same conclusion on different facts. In United States v. Bryan, the Fourth Circuit rejected a misappropriation theory of securities fraud as irreconcilable with Supreme Court precedent and held that the district court plainly erred by instructing the jury that it could convict on that theory.56 The excerpt states the plainness conclusion in a single sentence that cites Olano, so Bryan is better read as an example of the result than as an explanation of it. The common thread in these decisions is that the law supplied an answer the trial judge could have applied without any prompting.
The concurrence in United States v. Gaudin shows the contested side of the line in a decided case. The trial court there had not submitted the question of materiality to the jury, and the Court affirmed a court of appeals decision that had found that failure to be plain error.57 The concurrence observed that the Court did not review the plain-error determination and that the Government had not argued the error was not plain, but it added that every court of appeals to consider the question, except the Ninth Circuit, had held materiality to be a question of law, so it was “certainly subject to dispute” whether the error was “clear under current law.”58 The majority's own acknowledgment that it could not rule for the defendant while still adhering to the reasoning and holding of an earlier decision, Sinclair v. United States, points the same way, because a rule that must be set aside before the error appears is hard to call obvious.59 Set beside the hypothetical above, the difference between the two poles is how settled the law was, and students lose this step when they treat an error the reviewing panel finds persuasive as one that was obvious at trial.
Olano itself did not decide one variation: it expressly set aside the special case in which the error was unclear at the time of trial but becomes clear on appeal because the applicable law has been clarified.60 That reservation matters because the phrase clear under current law does not say which moment counts as current. Two circuit decisions from 1994 supply competing answers. The Second Circuit in United States v. Viola read current law to mean the law at the time of appeal, reasoning that a defendant has no duty to object to an instruction resting on firmly established circuit authority, since no legal right then existed to forfeit, and that demanding such objections would require defendants to foresee changes in the law and would invite frivolous objections.61 The D.C. Circuit in United States v. Washington took the other view, that the error must be plain under the law at the time of trial, with a separate inquiry for supervening decisions.62 Viola added that its treatment of supervening decisions, which puts on the Government the burden of showing no effect on substantial rights (a point that belongs to the third requirement), applies only when the decision alters a settled rule in the circuit, and that where the point was unaddressed or inconsistently decided the defendant still had a duty to object.63 These are circuit decisions responding to a question Olano left open, so an answer should present them as competing positions and not as a rule Olano announced.
Pre-Booker sentencing cases present that fact pattern. In United States v. Gray the defendant was sentenced before Booker, so the district court had treated the Guidelines as mandatory, and the court of appeals, reviewing after Booker, found plain error without separately discussing how the timing of the change affected plainness.64 The decision therefore illustrates a finding of plainness but does not answer the reserved question. On an exam, the plainness step is answered by asking what the governing law said and how settled it was, and when a stem describes a change in the law between trial and appeal, the answer should identify that Olano left the point open and lay out the competing positions.
Effect on Substantial Rights
A forfeited error that is an error and is plain has cleared two limitations, yet the court of appeals still has no authority to correct it unless the third is met: the plain error must “affec[t] substantial rights.”66 Those are the words Rule 52(a) uses for preserved error, and in most cases they carry the same meaning here, so the error must have been prejudicial, which means it must have affected the outcome of the district court proceedings.67 The requirement ties relief to actual harm for a practical reason: correcting an error that changed nothing sends the parties and the court system through a new trial or resentencing and buys no gain in accuracy or fairness. The Supreme Court later described the showing, in United States v. Marcus, as a reasonable probability that the error affected the outcome of the trial, which is more than a bare possibility that it did.68 Of the first three limitations, this is the one defendants find hardest to satisfy, because the claim was never raised below and the defendant must now prove that it mattered.69
The defendant, not the government, normally carries the burden of persuasion on this question. That allocation is the chief difference from preserved-error review, and the comparison with harmless error below develops it.70 The burden follows the defendant's silence. In the ordinary case the law was clear at trial, so the defendant was on notice of a duty to object, and by staying quiet both forfeited the claim and kept the judge from fixing the problem while a fix was still easy, which makes it fair to ask that defendant to show the mistake mattered.71 The Second Circuit in United States v. Viola recognized one exception, for a supervening decision that alters settled circuit law: a defendant cannot be expected to object to an instruction resting on firmly established authority, so the government must show that the error did not affect substantial rights.72 The exception is narrow, since it applies only when the new decision changes a settled rule in the circuit.73
The clearest picture of what a prejudice inquiry involves comes from Kotteakos v. United States, which Olano cites for the meaning of this phrase.74 Kotteakos was decided under the harmless-error statute that governs preserved error, so it supplies the shape of the inquiry and not plain-error doctrine itself. The Court directed reviewing judges to measure what the error meant to the jury in light of everything else that happened at trial.75 If the court is sure the error did not influence the jury, or had only a very slight effect, the judgment stands.76 If the court cannot say with fair assurance, after pondering all that happened without stripping the erroneous action from the whole, that the judgment was not substantially swayed, then substantial rights were affected, and a court left in grave doubt cannot let the conviction stand.77
Kotteakos also identified the wrong question to ask, which is whether enough other evidence existed to support the result apart from the phase the error touched. The right question is whether the error itself had substantial influence, because a court that asked only whether the remaining evidence would suffice would be deciding guilt for itself, and the verdict the error may have distorted belongs to the jury.78 The Court added that this difference is easy to overlook when the sense of guilt comes strongly from the record.79
The two trials the Court compared show how heavily the setting controls the answer. In Kotteakos the indictment charged one conspiracy, but the proof showed at least eight separate ones, run by participants who, with one exception, did not know one another, and thirty-two defendants stood accused at the outset, nineteen at the start of trial, and thirteen when the cases reached the jury.80 The Court found it highly probable that the error had a substantial and injurious influence on the verdict, and it described the defendants' substantial right as the right not to be tried en masse for a conglomeration of separate offenses committed by others.81 That a conviction would or might probably have followed a properly conducted trial was not the test.82 As Kotteakos described the earlier Berger decision, only two conspiracies involving four persons were proved there, and nothing showed prejudice or surprise from the variance, so no substantial right was affected.83 The errors in variance and instructions were identical in character, and what separated the two cases was scale: with so many defendants and schemes, a defendant tied to only one or a few of them had a far harder task preparing a defense and guarding against evidence about others being carried over to him, whether as harmless error or by psychological effect, despite the instructions.84 The same kind of error can therefore affect substantial rights in one trial and leave them untouched in another, because the inquiry measures what the error did within the record in front of the court.
Applied to a forfeited error, the same inquiry runs against the defendant, and United States v. Griffith shows how it looks when the defendant loses. The Fifth Circuit found plain error in the failure to give a specific Fifth Amendment instruction after the prosecutor asked the defendant, during rebuttal argument, whether he would like to take the stand, yet it refused relief at this step because the burden of persuasion lay with him.85 Reading the remark in context, the court noted that it was an isolated comment, that it did not strike at the heart of the defense, that the jury was told at once to disregard it, and that nothing suggested a calculated appeal to the jury.86 The court also relied on the strength of the government's evidence, which included tape-recorded conversations and roughly forty pounds of marijuana found in the defendant's home.87 That reliance does not conflict with Kotteakos, because the court used the evidence to conclude that the jury convicted on the evidence and not because of the defendant's refusal to testify, which is a judgment about the error's effect on the jury and goes beyond whether enough proof existed.88 Having found no effect on substantial rights, the court concluded that it had no authority to correct the error.89
The two halves of that pair differ in one fact, and it is the fact this step turns on: whether the error touched the evidence on which the verdict depended. Students lose this step when they stop once the error is plain, because clarity establishes only that a rule was clearly broken and says nothing about what the breach did to the verdict. An exam answer should therefore say what the error did to the outcome, point to the record facts that show it, and assign the burden of that showing to the defendant.
Olano drew this limitation with care and stopped short of saying that the phrase always means prejudice. The Court did not decide whether “affecting substantial rights” is always synonymous with “prejudicial.”90 It acknowledged that there may be a special category of forfeited errors that can be corrected regardless of their effect on the outcome, and it set aside the separate question of errors that should be presumed prejudicial when the defendant cannot make a specific showing.91 It said only that normally, though perhaps not in every case, the defendant must make a specific showing of prejudice.92 Both reservations rest on one idea: some defects damage the framework of a trial so basically that no measurement of effect on a particular verdict captures the harm, as Olano's reference to errors that deprive a defendant of basic protections, without which a trial cannot reliably serve its function, suggests.93
In Chapman v. California the Court noted that some constitutional rights are so basic to a fair trial that their infraction can never be treated as harmless error, and its footnote lists a coerced confession, the denial of counsel, and trial before a biased judge, citing Payne v. Arkansas, Gideon v. Wainwright, and Tumey v. Ohio.94 In preserved-error review, later opinions call defects of that sort structural errors, treat them as a very limited class, and apply a strong presumption that other constitutional errors are subject to harmless-error analysis.95 Those authorities describe preserved errors, so they show what the reserved category might contain without saying how Rule 52(b) treats a forfeited error of that kind, which is the point Olano left undecided. In Johnson v. United States the judge, not the jury, decided the materiality element and the defendant did not object, and the Court again reserved whether omitting an element of the offense automatically affects substantial rights, resolving the case instead at the fourth step, which the next section takes up.96
Discretion and Fairness
A forfeited error that is an error, is plain, and affects substantial rights has cleared the three limitations on appellate authority, and clearing them gives the court of appeals the power to correct the error without imposing a duty to do so. Rule 52(b) provides that plain errors affecting substantial rights “may be noticed” even though they were not brought to the trial court's attention.100 Olano took the word “may” at face value: the Rule is permissive, not mandatory, so a court of appeals that finds a plain, rights-affecting error has authority to order correction but is not required to use it.101 Three things pointed the Court to that reading, namely the Rule's wording, the nature of forfeiture, and the established appellate practice that Congress meant to continue.102 Because the choice belongs to the court, a fourth question remains after the first three: whether this particular error is one the court should use its power to correct.
The practice Congress meant to continue was old and tied to the stakes of criminal cases. Olano pointed to a 1913 Eighth Circuit statement that in criminal cases, where the defendant's life or liberty is at stake, federal courts may in the exercise of sound discretion notice forfeited error, and to a former Supreme Court rule under which the Court, at its option, could notice a plain error not assigned.103 The Supreme Court discussed that older rule in Weems v. United States, describing it as conferring “a discretion that may be exercised at any time” and as “not a rigid one.”104 Weems added that the Court felt less reluctance to act under the rule in criminal cases than in civil ones, and less reluctance still when the rights asserted are of such high character as to find expression in the Constitution.105 Two features of that history carry into the modern Rule: the power is discretionary, and the willingness to use it grows with the importance of what the error touched.
Because the Rule grants discretion without saying how to use it, Olano supplied a standard: the court of appeals should not exercise its discretion to correct a forfeited error unless the error “seriously affect[s] the fairness, integrity or public reputation of judicial proceedings.”106 The formula reaches Olano through United States v. Young from United States v. Atkinson, so the Court was restating a standard already in use.107 The concurrence in United States v. Gaudin lays the steps out in order and shows where this one falls: once the three prerequisites are met, the decision to correct the forfeited error “remains within the sound discretion of the court of appeals,” which should not use that discretion unless the error seriously affects fairness, integrity, or public reputation.108
Other formulations restate the same gate in different words, and they should be recognized as glosses and not as the federal standard. Harris v. United States, a District of Columbia decision that predates Olano, required an error in an unobjected-to instruction to be “so clearly prejudicial to substantial rights as to jeopardize the very fairness and integrity of the trial,” and it held that minor discrepancies between an instruction first read and the same instruction given again during deliberations fell short of that level.109 A dissent in the civil case City of Newport v. Fact Concerts, Inc. likewise said that any plain-error exception should be confined to the exceptional case where the error has seriously affected fairness, integrity, or public reputation.110 State courts supply their own wording. In State v. Smith, a New Jersey decision, the test asked whether a legal impropriety in a jury charge prejudicially affected substantial rights and had a “clear capacity to bring about an unjust result.”111 In City of Maryland Heights v. Heitz, a Missouri decision, review of unpreserved claims was limited to exceptional cases of manifest injustice or miscarriage of justice, and the court declined to exercise it.112 On a state-law question the state's own formulation controls, while a federal question calls for the Olano language.
This discretionary fourth step is a separate exercise of judgment, and it should not be treated as a second pass over the prejudice question. The third limitation asks what the error did to the outcome, measured against the record. The fourth asks about the proceeding as a whole: given that the error mattered, would leaving it uncorrected damage the fairness, integrity, or public reputation of the process that produced the judgment? A dissent in Neder v. United States drew the same line when it explained that, for an unobjected-to error, the deprivation of substantial rights does not by itself warrant reversal, and the appellant must also show the serious effect on fairness, integrity, or public reputation.113 In theory, then, an error prejudicial enough to require reversal if properly preserved will not always require reversal under plain-error review, although courts have found it hard to say how much more serious the error must be, and some opinions state little beyond their conclusion.114
Two reasons support placing a discretionary gate after prejudice. The first is economy: costly appeals and retrials can be avoided when claims of error are raised while the trial judge can still correct them.115 A rule that guaranteed correction whenever the first three limitations were met would let a lawyer stay silent, accept a favorable verdict if it came, and hold the error in reserve if it did not, so the discretion keeps silence from becoming a strategy that always pays. The second reason lies in the Rule's own purpose. Justice Brennan, dissenting in United States v. Frady, a collateral-review case, described the Rule as softening the harsh effect of an adversary system in which a defendant is generally bound by counsel's conduct, since it permits relief in exceptional cases despite counsel's failure to object.116 The same dissent stressed that the Rule permits rather than directs notice of error, that the power is to be exercised cautiously, and that resort to it belongs to exceptional circumstances.117 Plain-error review also reflects a judgment that appellate courts share some responsibility for the correctness of outcomes, and the fourth step is where the court decides whether a particular error calls on that responsibility.118
Since relief is discretionary, the defendant who sat silent carries one more burden. In United States v. Vonn the Court explained that such a defendant must show that substantial rights were affected and, because relief on plain-error review lies in the reviewing court's discretion, must further persuade the court that the error seriously affected the fairness, integrity, or public reputation of judicial proceedings.119 The allocation follows from the defendant's silence: having skipped the ordinary route of objecting, the defendant must justify the extraordinary one. In an appellate brief that means arguing the fourth step with record facts, and a court left unpersuaded simply declines to notice the error.
Cases in which relief followed show the step in operation. In United States v. Gray the district court had treated the Guidelines as mandatory and enhanced the defendant's sentence on a fact the jury had not found, and the court of appeals, reviewing for plain error after Booker, found plain error affecting substantial rights, separately concluded that the error seriously affected the fairness, integrity, or public reputation of judicial proceedings, and vacated the sentence.120 In Rosales-Mireles v. United States the Supreme Court held that a plain Guidelines miscalculation affecting a defendant's substantial rights will in the ordinary case seriously affect fairness, integrity, or public reputation and so warrant relief.121 The Court's description of the Guidelines as a meaningful benchmark for the initial sentence and for appellate review helps explain why: a wrong range distorts the starting point for the sentence the defendant must serve.122 Even there the holding speaks of the ordinary case, which keeps the step a judgment applied to a category of error, with no automatic relief.
Relief can also be withheld at this step even when the earlier steps are met or assumed. In Johnson v. United States the judge, rather than the jury, decided the materiality element in a perjury prosecution, and the defendant did not object. The Court reserved whether omitting an element automatically affects substantial rights, and it resolved the case at the fourth step, concluding that the error did not warrant correction in light of overwhelming and uncontroverted evidence on materiality.123 In United States v. Cotton the Court overruled Ex parte Bain, which had treated an indictment omitting an element as a jurisdictional defect, then reviewed the unraised omission for plain error and declined to reverse because, on the facts, the omission did not seriously affect the fairness, integrity, or public reputation of judicial proceedings.124 Cotton shows that even a defect once treated as automatically reversible can fail this step on its facts. Johnson shows what such facts can look like: uncontroverted proof on the very point the error touched.
Students lose this step when they treat correction as automatic once prejudice is shown, because a showing of prejudice satisfies the third limitation and says nothing yet about whether leaving the error in place would damage the proceeding's fairness, integrity, or public reputation. A complete answer states that Rule 52(b) is permissive, gives the fairness, integrity, or public reputation standard, places the burden of persuasion on the defendant who sat silent, and then applies the standard to the record by asking how serious the error was, how contested the point it touched, and whether a corrected proceeding would differ from the one already held.