Perfecting Issue Preservation

Response - Online Edition - Volume 98


John Langbein famously observed that Americans “live under a criminal procedure for which we have no adequate theory.”[1] If any more evidence for that point is needed, Darryl Brown provides it in his article, Does It Matter Who Objects? Rethinking the Burden to Prevent Errors in Criminal Process.[2] In particular, Brown’s target is the venerable rule that a defendant must preserve objections to erroneous rulings at trial in order to perfect them for later appeal. Trial lawyers ignore this rule at their—or, more accurately, their clients’—peril. And it is the bane of appellate criminal lawyers, who all too often discover errors that escaped trial counsel’s notice when combing the record for grounds for appeal. Brown convincingly argues that the traditional forfeiture rule, despite its long lineage, deserves closer examination and reconsideration.

As someone who has made the mistake of repeatedly writing about error in criminal justice,[3] I read Brown’s argument with great interest. And Brown ably convinced me that conventional wisdom about who should bear the burden of bringing errors to a court’s attention is woefully undertheorized. In particular, Brown’s move to analyze adjudicative error from the perspective of accident prevention in other legal contexts is both clever and generative of insights. Moreover, Brown made a persuasive case that normative judgments about fairness, rather than a careful cost–benefit analysis, may better explain the status quo.

What I am less certain of, though, is whether Brown has met his burden of persuading us that we should adopt his proposed rule: that the law should “plac[e] the duty to prevent errors on the party who commits the error (or who benefits from the judge’s error), through the use of a de facto reverse-forfeiture penalty.”[4] In this short Response, I explain why I am not fully persuaded. In Part I, I discuss why Brown’s rule may not necessarily prevent errors as much as he hopes, and may instead significantly increase reversals and retrials. In Part II, I question whether that cost is worth bearing, by interrogating the concept of “error” and its multiple possible meanings. That inquiry leads me to a limited defense of our system’s current approach to forfeited legal errors. I conclude by suggesting more modest reforms that, in my view, follow from Brown’s significant insights.

I. The Dynamics of Reform

Brown’s argument for changing which side bears the risk of error proceeds in a consequentialist framework, using the language of cost–benefit analysis. That is, Brown starts from the premise that “the justification for forfeiture rules is fundamentally instrumental” and that “[t]heir overriding purpose is to optimize parties’ error-preventing behavior by encouraging care in preventing errors and deterring delays in raising claims of error.”[5] I find this consequentialist framework sensible in this context and will evaluate Brown’s proposal by considering its possible costs and benefits.

That said, we should be clear at the outset what the relevant universe of costs and benefits includes. In particular, I would stress that the goal is not merely to minimize or eliminate errors at the trial court level. If that were the goal, the best rule would be one that required mandatory reversal on appeal for any error, regardless of whether it was objected to, and no matter how small its possible effect on the judgment. That, of course, is not the way things work[6] nor is it the rule for which Brown is arguing.[7] Why don’t we structure the rules that way? For one, it is because reversals are costly.[8] Trials take time and impose significant burdens on victims and other witnesses; starting over from scratch may increase the risk of incorrect outcomes given that evidence may dissipate over time.

At the same time, not all trial court errors are created equal. The errors Brown is concerned with include trial court rulings or actions by a party opponent that conflict with governing law in some way. Some such errors—say, the trial court instructing the jury incorrectly on the reasonable-doubt standard[9] or perhaps the court denying the defendant the right to a jury trial altogether[10]—are quite troubling and seriously call the fairness and reliability of a trial into question. But other errors—such as a prosecutor inadvertently introducing a small piece of largely immaterial hearsay testimony during direct examination of a government witness—may be significantly less consequential. Indeed, it is important to stress that only some trial court errors (in the sense Brown is talking about, and which I will call “legal errors”) implicate the more important kind of error we are concerned with in criminal justice: what I will call “outcome errors,” such as when an innocent person (or, at least, a person who should have been acquitted had governing law been followed) is convicted. Although some legal errors have a strong tendency to produce outcome errors, many do not.

Given this calculus—reversals are always costly but errors are not necessarily so—we should not necessarily design the the adjudicative system to minimize legal errors vel non. Instead—and with apologies to Adrian Vermeule[11]—the goal should be to optimize, not minimize, legal errors. I do not think Brown would disagree with me on this point, but I emphasize it because at various points Brown uses language that suggests that error minimization should indeed be the goal of adjudicative rules.[12] In doing so, I wonder whether Brown is subtly substituting intuitions about outcome errors in a situation when we are really talking about legal errors. But more on this later.

The point of all this, for now, is simply to establish something that I think should not be terribly controversial: Brown’s rule has the potential to increase, perhaps significantly, the number of appellate reversals in criminal cases. If so, we would need to think hard about whether that is a price worth paying for a possible reduction in a number of legal errors experienced in the system. In the next Part, I will suggest how we might feel about that cost–benefit inquiry. Here, though, I would like to consider the possibility that Brown’s proposal represents a kind of free lunch. If he is right that the party benefiting from an error is the “least cost avoider,” then perhaps putting the burden to guard against error on that party will significantly deter legal errors without actually requiring any more reversals; the threat alone might be enough to change behavior and thus to produce benefits.

Perhaps, but I’m skeptical. Let’s consider how Brown thinks his rule will work. He argues that under the current regime, “forfeiture rules encourage parties to attempt rule violations in their own tactical actions.”[13] He contends that changing the burden would reduce this incentive by “encourag[ing] parties to use greater care to prevent their own errors (or judicial errors in their favor).”[14] At some level this has to be right; any rule that increases the perceived consequences for failing to prevent errors should have some effect on error-prevention behavior. But what would the likely magnitude of this change be?

There is some reason to suspect that it will not be significant. I think Brown is surely right that both sides have some incentive to introduce, or at least not studiously to prevent, errors at trial. Focusing just on the prosecution for the moment, a zealous prosecutor may try to push the envelope as much as she can in the hopes of getting a conviction. Certainly the pages of appellate reporters are full of many examples where a prosecutor benefited by convincing a trial court to accept a dubious legal argument.

But is Brown right that forfeiture rules play a significant role in this dynamic? I am not so sure. In most instances, a prosecutor must attempt to introduce a legal error before knowing whether the defense will object. That means that the prosecutor is taking the risk that (1) the other side will object and (2) that the trial court will overrule the objection, thus introducing possible grounds for appeal that could result in a victory being overturned on appeal. Yet prosecutors choose to take that gamble and introduce the error nonetheless. Why? Brown might argue that the forfeiture rule introduces some possibility that the error will be immunized from appeal, thus increasing the expected value of error-producing conduct.[15] And such a conscious cost–benefit calculation may well explain this dynamic in some number of situations.

But I think that in many cases, the explanation for the prosecutor’s behavior is simpler: the prosecutor does not believe she is introducing an error. Motivated reasoning and other cognitive biases are powerful forces,[16] and they can play a major role in shaping lawyers’ perceptions of what is the correct answer to a disputed legal question. As Christopher Schroeder puts it:

[A]ny lawyer who has litigated a case at some time has had an experience that speaks to the explanatory power of motivated reasoning. Lawyers are trained to advocate vigorously for their clients, a task that involves presenting the best legal arguments to advance their clients’ interests. Finding the best argument for a client does not mean, of course, that one has found the best argument in the case—the best argument of the other side might be better. This is not, however, how lawyers experience the process of analyzing facts, researching arguments, testing theories, and developing their eventual legal position. Because lawyers are motivated to find persuasive arguments to support their clients’ positions, the theory of motivated reasoning suggests that they will find arguments, evaluative techniques, and evidence that vindicate their clients’ positions to be the most persuasive.[17]

There is no reason to think prosecutors are immune from these forces. Indeed, there is a great deal of evidence that prosecutors suffer from significant cognitive biases that cause them to believe in the righteousness of their prosecutions even in the face of overwhelming evidence of a defendant’s innocence;[18] it is no large leap to think that a similar dynamic would apply to prosecutors’ evaluations of legal arguments.

This observation leads to my biggest objection to Brown’s view of prosecutors as the “least cost avoider” for errors at trial. In our adversarial system, the party harmed by a possible error will, in my view, always be better positioned to take note of the error precisely because that party is much more likely to perceive it as an error. Brown may be right that “the acting party likely would be more knowledgeable about the specifics of their tactics,”[19] but motivated reasoning may significantly undercut any knowledge advantage.

A further problem that I don’t think Brown sufficiently grapples with is the asymmetry of incentives created by the Double Jeopardy Clause. Because of that clause, an acquittal at trial will be final and unappealable—even if it was produced by a manifest legal error for which the defendant’s attorney strenuously argued, and even if the prosecution pointedly and persuasively objected to that error at trial. Thus, while both prosecutors and defense attorneys have some incentive to seek legal errors at trial in order to obtain trial victories, the defense’s incentives are unmitigated by any cross-cutting incentives created by the risk of reversal on appeal (other than the rare situations where the government can obtain an interlocutory appeal).

Defense lawyers thus must keep an eye out for the other side’s errors, but they need not care at all about their own. Prosecutors, by contrast, must pay careful attention to potential defense errors (since the defense need not care about them at all) but must also police their own errors to some degree, so as to make sure that any victory will not be subject to appellate reversal. Prosecutors, then, already have a greater responsibility for monitoring errors at trial as compared to the defense. Brown’s proposal would enhance that asymmetry further, giving prosecutors more to worry about and defense attorneys less reason to care about the other side’s errors (since any errors the defense misses will still be potential grounds for appeal later on).

It is possible that prosecutors would be better positioned to take account of trial errors relative to defense attorneys, notwithstanding the significantly greater burden that Brown’s rule would place on them. But that could be true only under certain assumptions. First, prosecutors on average could be more effective lawyers than defense attorneys, perhaps as a result of higher pay or other selection effects.[20] Brown, however, explicitly rejects this assumption.[21] Perhaps prosecutors, even if equally skilled, are better positioned to take note of errors because they have fewer demands on their time. And indeed, defense attorneys—at least those who provide defense for the indigent—often face crushing caseloads.[22] But whether a general workload disparity is sufficient to make prosecutors better positioned to keep track of errors during a criminal trial, when attorneys’ attention may be totally consumed by the trial, is at least highly uncertain. That is especially true once we factor in the asymmetrical incentives created by double-jeopardy protections, which mean we are starting from a baseline in which prosecutors already have to be more attentive to errors during trial than are defense attorneys.

For these reasons, it is hard for me to be confident that the prosecution really is the least cost avoider in the error-avoidance context. At the least, it is hard to confidently assert that prosecutors are going to be so much better at avoiding errors that Brown’s proposal will have no costs in terms of additional reversals. Brown’s proposal may be worth adopting even if it requires paying those costs, of course, and I do not take Brown as contending that his proposal will be perfectly reversal-neutral. But if the proposal will have costs, we must then squarely determine whether we think those costs are worth paying in light of the benefits of the proposal. It is to that inquiry that I now turn.

II. The Price of Reform

Let us assume that Brown’s proposal will entail some costs in terms of reversals of additional convictions—although we cannot know exactly how many. The reform will thus come at a price. Is it a price we should be comfortable paying? To answer this question, we need a better sense of what our goals are, or should be, when designing the rules of the adversarial criminal process.

Here is what Brown says: “[E]rrors in the adjudication process are no different from accidents in tort law. . . . The law of procedure, like tort law, is focused on minimizing the cost of errors (or, in tort terminology, accidents).”[23] This assumption is, in my view, critical to Brown’s argument. If we agree that the overriding purpose of procedure is to reduce error, then we may be comfortable purchasing a reduction in errors at the price of some number of additional reversals.

But is he right? Here, I think much turns on what we mean by “error.” Earlier, I distinguished between legal errors and outcome errors. I think we might well conclude that the goal of procedure is to minimize outcome errors: factually incorrect outcomes (an innocent person being convicted or a guilty person going free), or at least legally unjust results (such as someone being convicted when under governing law they should have been acquitted). But I do not think the goal is necessarily to minimize legal errors, at least across the board.

Let me use one of Brown’s own formulations to explain why. He says that “one of the core ambitions for any public litigation system” is “minimizing inaccurate or unfair judgments that are the products of rule-violating procedures.”[24] I don’t disagree with this formulation, but it requires us to determine which rule violations actually create “inaccurate or unfair” results. Brown’s analysis treats all errors similarly, but I am not sure all are created equal. When we consider some kinds of criminal-procedure rights that are enforced at trial, we will find that some of them might present a lower risk of leading to “inaccurate or unfair” results than others, at least when applied to situations where a defendant made no objection at trial. Much depends, I think, on how exactly we conceptualize the right in question.

Consider the Sixth Amendment Confrontation Clause, which provides that “[i]n all criminal prosecutions, the accused shall enjoy the right . . . to be confronted with the witnesses against him.”[25] In some instances—such as in a trial that motivated the rule in which Sir Walter Raleigh was condemned on the basis of questionable ex parte accusation that Raleigh was not able to cross-examine[26]—the confrontation right may guard against false convictions. But in other cases, the rule operates to keep highly probative evidence away from the jury simply because the declarant is unavailable. In Crawford v. Washington, for example, the rule barred the government from introducing statements by the defendant’s wife that cast doubt on the defendant’s claim of self-defense—not because the government refused to call her as a witness but because the defendant barred her from testifying under the state’s marital privilege evidentiary rule.[27] Had Crawford’s attorney not thought to invoke the Confrontation Clause, would we think there was any inaccuracy or unfairness in the resulting conviction, simply because there was a potential procedural argument that the defendant did not think to invoke?

How much unfairness is involved turns, I think, on how we think of the confrontation rule. It is partly about giving the defendant a certain satisfaction in getting to face his accusers before being condemned, and avoiding the outrage one might feel at being denied that privilege (as Raleigh was[28]). If one accepts that conception, however, one might not feel there is any great unfairness involved when a defendant does not even ask to confront his accusers. Put another way, whether the defendant invokes the right at trial might tell us a great deal about how much unfairness is actually resulting from the error. Indeed, some might even conclude that it is a good thing when a defendant does not seek to confront her accusers in this situation, since that failure might lead the finder of fact to make a more accurate judgment.

Consider another example. The Sixth Amendment also guarantees that “the accused shall enjoy the right to a speedy . . . trial.”[29] In Barker v. Wingo,[30] the Supreme Court concluded that whether the defendant asserted her right to a speedy trial would be “entitled to strong evidentiary weight in determining whether the defendant is being deprived of the right.”[31] Although the Court rejected a rule that would always require the defendant’s assertion of the right, it concluded that “failure to assert the right will make it difficult for a defendant to prove that he was denied a speedy trial.”[32] The Court so concluded because it recognized that in some instances, defendants might not want speedy trials for strategic reasons,[33] and also that the defendant’s contemporaneous assertion of her rights would provide significant information about whether the defendant was actually being harmed by the deprivation of the right.[34] Again, the defendant’s assertion of her rights is key to helping us figure out whether the result below was indeed unfair, and thus whether reversal seems justified.[35]

One need not totally agree with my conclusions about these examples to accept the larger point that they are meant to illustrate: not all legal errors are created equal. Some legal errors do create a significant risk of an erroneous, or at least unfair, outcome. But others do not. Some number of putative “errors” may be thought of simply as situations where the defendant was given a certain legal entitlement that she could have, but chose not to, exercise. Even if a trial court’s admission of unconfronted testimonial hearsay without objection is technically an “error,” the mere fact that a defendant could have, but not did not seek to, confront an accuser at trial does not mean that we should conclude the result is erroneous or unjust. The goal of the design of adjudicative rules may be to prevent errors defined unfair or erroneous outcomes, but that is not the same thing as concluding that adjudicative rules should be designed to minimize or prevent “errors” defined as all instances where a defendant failed to exercise some legal right she conceivably could have exercised at trial.

What I am left with, then, is the conclusion that the procedural rules should not necessarily care about providing remedies for every error that passes without objection in the trial court. Instead, perhaps, the right approach is for appellate courts to provide relief only for those forfeited errors that create a meaningful risk of erroneous or unjust outcomes. What this regime sounds very much like, of course, is the status quo. Under existing law, appellate courts may notice a “plain error” that was not objected to at trial—but only if it “affects substantial rights”[36] and also “seriously affect[s] the fairness, integrity or public reputation of judicial proceedings,”[37] a category that includes, but is explicitly not limited to, convictions of actually innocent defendants.[38] That is, the law allows courts to ignore a defendant’s failure to object to a legal error in precisely those circumstances where we think that error contributed to an outcome error.

For these reasons, assuming that Brown—as the party seeking to change the status quo—bears the burden of proof, I am not yet ready to conclude that he has discharged his burden by convincing me that we should be willing to accept some significant number of additional reversals in order to prevent some number of legal errors in criminal trials. Nonetheless, his analysis contains a great number of insights, and they provide grounds for at least some modest reforms of existing procedural rules, a point to which I now turn.

Conclusion: Toward Modest Reforms

Even if I am not ready to adopt Brown’s reforms wholesale, his insights have persuaded me that certain more modest reforms are advisable. First, I found particularly important and valuable Brown’s observation that “because they apply to bilateral activity, forfeiture rules create incentives for both parties, and those incentives conflict.”[39] As Brown ably explains, prior evaluation of forfeiture rules have inappropriately focused only on one side of the adversarial process, and any accounting of the desirability of the rule needs to consider the possible incentives for both sides of any potential rule. Moreover, any reforms that reduce the incentive of one side to introduce errors are worth considering.

Here is one potential reform that could do that: rules could somewhat relax the demanding plain-error standard for situations where the error represented a prosecutor’s potentially deliberate attempt to introduce an error. Rules could provide that if an appellate court determines that (1) a forfeited error was plain at the time of trial (and thus should have been plain to the prosecutor) and (2) the error was largely caused by the prosecutor (rather than by the trial court erring on its own initiative[40]), a less onerous plain error test would govern. By providing a greater likelihood of appellate review, such a rule change might discourage prosecutors from seeking to introduce obvious errors in the hopes that the other side would fail to object. Like Brown’s proposal, it would mean that the government would “lose[] immunity from appellate review for errors it caused.”[41]

In addition, Brown’s insights about the incentives to introduce error extend beyond the context of forfeited errors. In some instances, defendants will object to errors that prosecutors introduce, but they will not be able to persuade the trial court. In this domain of preserved error, Brown’s analysis helps us realize how the law could do more to penalize prosecutors for leading the trial court astray.

Consider one simple change: what if every time the defendant objected to a prosecutor’s tactics, the rules required the prosecutor to explain on the record, while responding to the objection, the effect his tactics would likely have on the jury? Such a change might put the prosecutor in a difficult position. If she says that the tactic should have no effect on the jury whatsoever, the trial court might conclude that there is no harm to the prosecutor in sustaining the defendant’s objection. If, though, the prosecutor acknowledges how the tactic is designed to persuade the jury, it may be harder for the government to argue later that the error was harmless beyond a reasonable doubt, as the government would normally be expected to do if the defendant is convicted and then appeals.

I do not know for certain whether these reforms are worth adopting, though I suspect they are. What I do know is that they, like Brown’s more ambitious proposal, merit consideration. Whether or not Brown has met his burden of persuasion in convincing us to go ahead and adopt his reforms, he has certainly convinced me—and, I hope, others—that we need to think much more deeply about the incentives and effects of our rules about forfeiture and issue preservation. Here, as with far too many of our other procedural practices in criminal justice, adherence to longstanding practices and pithy, oft-repeated slogans appear to substitute for meaningful justification for the status quo. By showing us the need to rethink our settled practices in this area, Brown’s article is an important step in the right direction.

  1. .John H. Langbein, The Origins of Adversary Criminal Trial 9 (2003).
  2. .See generally Darryl K. Brown, Does It Matter Who Objects? Rethinking the Burden to Prevent Errors in Criminal Process, 98 Texas L. Rev. 625 (2020).
  3. .See generally Daniel Epps, Harmless Errors and Substantial Rights, 131 Harv. L. Rev. 2117 (2018); Daniel Epps, One Last Word on the Blackstone Principle, 102 Va. L. Rev. Online 34 (2016); Daniel Epps, The Consequences of Error in Criminal Justice, 128 Harv. L. Rev. 1065 (2015); Daniel Epps, The Right Approach to Harmless Error, 120 Colum. L. Rev. F. 1 (2020).
  4. .Brown, supra note 2, at 632.
  5. .Id. at 631 (footnote omitted).
  6. .See, e.g., Fed. R. Crim. P. 52(a); Chapman v. California, 386 U.S. 18, 22 (1967); see generally Epps, Harmless Errors and Substantial Rights, supra note 3.
  7. .See Brown, supra note 2, at 658–59 (arguing that “the harmless error standard . . . reinforces the right incentives for litigants for error prevention in the trial stage” (footnote omitted)).
  8. .The Supreme Court explained the “substantial social costs” of reversal of a criminal conviction in United States v. Mechanik, 475 U.S. 66 (1986):

    [I]t forces jurors, witnesses, courts, the prosecution, and the defendants to expend further time, energy, and other resources to repeat a trial that has already once taken place; victims may be asked to relive their disturbing experiences . . . . Thus, while reversal “may, in theory, entitle the defendant only to retrial, in practice it may reward the accused with complete freedom from prosecution,” and thereby “cost society the right to punish admitted offenders.” Even if a defendant is convicted in a second trial, the intervening delay may compromise society’s “interest in the prompt administration of justice,” and impede accomplishment of the objectives of deterrence and rehabilitation.

    Id. at 72 (citations omitted).

  9. .See, e.g., Sullivan v. Louisiana, 508 U.S. 275, 277 (1993).
  10. .See, e.g., Duncan v. Louisiana, 391 U.S. 145, 146 (1968).
  11. .See, e.g., Adrian Vermeule, Optimal Abuse of Power, 109 N.W. U. L. Rev. 673, 676 (2015) (arguing that abuse of power should not be “minimized, but rather optimized”).
  12. .E.g., Brown, supra note 2, at 645 (“[P]rocedural law seeks to minimize these errors . . . .”).
  13. .Id. at 644 (emphasis omitted).
  14. .Id. at 650.
  15. .Brown argues that a prosecutor who has chosen to introduce an error “faces one of two outcomes: either the defense will object, in which case the prosecutor will have to play by the rules, or the defense won’t object, in which case the prosecutor gains whatever advantage the rule-violating tactic provides because the error is immune to appellate review (or at worst faces review under the highly favorable plain error standard).” Id. at 645. Yet this analysis ignores the possibility, discussed above, that the district court will agree with the prosecutor even in the face of an objection.
  16. .For a fascinating discussion of the problem of motivated reasoning in the legal context, see Dan M. Kahan, Foreword: Neutral Principles, Motivated Cognition, and Some Problems for Constitutional Law, 125 Harv. L. Rev. 1, 26 (2011).
  17. .Christopher H. Schroeder, Causes of the Recent Turn in Constitutional Interpretation, 51 Duke L.J. 307, 356–57 (2001).
  18. .See, e.g., Barry C. Scheck, Conviction Integrity Units Revisited, 14 Ohio St. J. Crim. L. 705, 715–17 & nn.37–44 (2017).
  19. .Brown, supra note 2, at 648.
  20. .For an account of this disparity, see Bryan Furst, Brennan Ctr. for Justice, A Fair Fight: Achieving Indigent Defense Resource Parity 8–9 (2019), https://‌‌/sites‌/default‌/files‌/publications‌/2019‌_09‌_Defender‌%20Parity‌%20‌Analysis‌V7.pdf []. Senator Kamala Harris recently introduced legislation aimed at eliminating this compensation gap. See Jacqueline Alemany, Power Up: Kamala Harris, Seeking to Change Her Narrative, Proposes Relief for Public Defenders, Wash. Post (May 8, 2019), https://‌‌/news‌/powerpost‌/paloma‌/powerup‌/2019‌/05‌/08‌/powerup-kamala-harris-‌seeking-to-change-her-narrative-proposes-relief-for-public-defenders‌/5cd1bf0‌d1ad‌2e506550b2f97‌/ [].
  21. .See Brown, supra note 2, at 648 (“On average, the parties should have lawyers who are equally skilled in recognizing rule violations.”).
  22. .See, e.g., Furst, supra note 20, at 7–8.
  23. .Brown, supra note 2, at 633.
  24. .Id. at 643.
  25. .U.S. Const. amend. VI.
  26. .See Crawford v. Washington, 541 U.S. 36, 44 (2004) (discussing Raleigh’s trial).
  27. .Id. at 40.
  28. .See id. at 44 (“Suspecting that Cobham would recant, Raleigh demanded that the judges call him to appear . . . .”).
  29. .U.S. Const. amend. VI.
  30. .407 U.S. 514 (1972).
  31. .Id. at 531–32.
  32. .Id. at 532.
  33. .See id. at 521.
  34. .See id. at 531–32.
  35. .In the speedy trial context, reversal means that the charges against the defendant must be dismissed with prejudice. See Strunk v. United States, 412 U.S. 434, 439–40 (1973).
  36. .Fed. R. Crim. P. 52(b).
  37. .United States v. Olano, 507 U.S. 725, 736 (1993) (quoting United States v. Atkinson, 297 U.S. 157, 160 (1936)) (alteration in original).
  38. .See id. at 736–37 (“An error may ‘seriously affect the fairness, integrity or public reputation of judicial proceedings’ independent of the defendant’s innocence.”).
  39. .Brown, supra note 2, at 643 (emphasis omitted).
  40. .While Brown’s rule would apply to errors introduced by a judge, I am less confident that this is desirable, given that courts—unlike prosecutors—lack the incentive to introduce errors that Brown identifies.
  41. .Brown, supra note 2, at 635.