Insurmountable obstacles: structural errors, procedural default, and ineffective assistance.
INTRODUCTION I. THE REASONS FOR AUTOMATIC REVERSAL A. A Brief History of Harmlessness B. The Five Faces of Structural Error: Why Not Do Harmless Error Review? 1. Jury impossibility 2. Speculative impossibility 3. Lack of effect on verdict 4. Obviousness 5. Seriousness C. The Curious Case of Batson 1. Garden-variety jury impossibility 2. "Pay no attention to the man behind the curtain" 3. True impossibility 4. No valid verdict 5. Complications II. PROCEDURAL DEFAULT AND THE CAUSE AND PREJUDICE EXCEPTION A. Purpose and Mechanics of the Procedural Default Rule B. Ineffective Assistance as Cause III. STRUCTURAL ERRORS MEET PROCEDURAL DEFAULT: THE CONFLICT A. The Problem with Sykes Prejudice B. The Problem with Strickland Prejudice C. Counterargument: Is Habeas Different After Brecht? IV. THE MESS IN THE CASE LAW AND THE SOLUTION A. United States Supreme Court B. Courts of Appeals 1. The main split: to presume or not to presume? 2. Why collapsing Strickland and Sykes prejudice won't work 3. What do courts do when they refuse to presume? C. Possible Solutions CONCLUSION
Michael Edward Vansickel was accused of murder. (1) During jury selection, the judge allowed him only ten peremptory challenges--half the number to which he was entitled under California law. (2) His lawyer did not realize that he was entitled to the additional strikes, and when he'd used ten, a jury was seated, even though Vansickel's lawyer had more jurors in mind that he would've liked to strike. The judge's error came to light after the verdict, but the court denied Vansickel's motion for a new trial.
When Vansickel filed his federal habeas petition, the district court agreed that his due process rights had been violated. However, the court said, since he failed to object at the time the error happened, he couldn't raise a post-conviction objection on those grounds. Normally, a petitioner can raise this kind of procedurally defaulted claim if he can show that some external factor (here, his lawyer's poor performance) caused the default, and that the unraised error had prejudiced his defense. Vansickel pointed out that this very circuit had previously held that there was no conceivable way to show whether there had been prejudice in this type of jury selection error (how could he prove that different jurors would have decided differently?), and argued that prejudice must be presumed for such claims, but the court held fast despite the impossibility of its requirements. The Ninth Circuit affirmed the district court's denial of habeas despite the acknowledged due process violation because Vansickel had not met a procedural standard that the very same court, sitting en banc, had held was simply not possible for any defendant to meet. (3) As of this writing, Vansickel is still in prison at Salinas Valley. (4)
This sort of case presents a real puzzle in habeas law: courts have adopted two sets of rules in two different contexts that are incompatible when both contexts are implicated in a single case. The two rules--the structural error rule and the procedural default rule--have never been reconciled, and courts applying both without acknowledging their conflict end up with untenable results like those in the Vansickel case.
Courts and legislatures adopted harmless error rules to prevent a reversal for trivial errors where the trial's result was nonetheless accurate. But courts have always acknowledged that, logically, some errors must be exempt from this rule because it would be impossible to demonstrate whether or not they were harmless.
In a completely separate area of law, courts developed rules for federal habeas review of state convictions. One such rule is the adequate and independent state ground doctrine, which requires federal courts to reject claims that state courts rejected for procedural reasons on direct appeal or in state collateral proceedings (claims that are "procedurally defaulted"). Accepting the need to make accommodations in the interest of justice, federal courts also developed an exception to this rule of rejection, establishing in Wainwright v. Sykes (5) that a petitioner could still raise the defaulted claim if he could show cause and prejudice for failing to present the claim below. Further complicating the matter is the fact that the most common "cause" accepted in these cases is a claim of ineffective assistance of counsel--where the claim was not raised below because the petitioner's lawyer unreasonably failed to raise it. The claim of ineffective assistance of counsel itself requires a showing not only that the petitioner's lawyer performed deficiently at trial, but also that his failings influenced the outcome of the trial. Thus, some habeas petitioners must show prejudice twice before they are even allowed to raise their defaulted claims: once to satisfy the ineffective assistance inquiry invoked as cause, and a second time to satisfy the prejudice half of the "cause and prejudice" inquiry.
If prejudice is impossible to show on direct appeal in certain cases, it remains impossible to show on habeas. Thus, requiring a showing of prejudice to resurrect a procedurally defaulted claim is functionally equivalent to foreclosing the claim entirely. Courts do not appear to have intended to isolate the group of so-called structural errors and render them uniquely unavailable after procedural default, while leaving less serious claims available. Yet by developing these rules independently--the structural error rule on the one hand and the procedural default requirements on the other--courts have painted themselves into a corner. They cannot strictly follow both of their own rules without yielding the absurd result of the theoretically available claim guarded by an always-insurmountable barrier.
Courts have not clearly addressed this conflict; instead, they have made a muddle of it, either requiring the impossible showing without acknowledging its impossibility, presuming prejudice without analyzing why such a route is required, or confusing and conflating the various prejudice inquiries involved. Indeed, the courts of appeals are now split as to which rule must yield. Courts created this problem; they must sort it out. They could do so either by extending the structural error exception to any context where a showing of prejudice is required, or, at the very least, by incorporating the exception into the definition of prejudice for procedural default purposes.
No scholarly work has analyzed this conflict. Two treatises have recognized that there is some disagreement among courts as to how to treat prejudice inquiries on procedural default, (6) but neither one goes on to consider the reasons for or contours of the problem, or to present any solution.
This Note will proceed by independently exploring each of the two incompatible judicial doctrines--the harmless error/structural error complex and the procedural default rules (including the role of ineffective assistance). These explorations represent Parts I and II. I pause in each context long enough to take a good look at the factors motivating each rule; an understanding of the reasons for the rules will inform a discussion of how to reconcile them. Having laid this groundwork, I then demonstrate in Part III how the conflict of the two strands of reasoning is unavoidable if the rules are applied unyieldingly; we cannot have both a robust structural error rule and a robust cause and prejudice requirement. In Part IV, I examine the welter of case law attempting to deal with such issues and suggest the two possible solutions available to courts to clean up the mess.
I. THE REASONS FOR AUTOMATIC REVERSAL
To fully understand the conflict that courts have created for themselves, we first need to explore the two rules that are in conflict. The first of these is the structural error rule. Why is it that most claims require a showing of prejudice on direct review, but some do not? A survey of harmless error and the (somewhat confused) contours of the structural error rule will establish the foundation for an understanding of the true conflict between structural errors and procedural requirements imposed on habeas petitioners. This Part addresses the harmless error and structural error rules on direct appeal, the context in which they were created. After this discussion, we will examine the procedural default rule and its underpinnings. Finally, we will consider the conflict between the two doctrines.
A. A Brief History of Harmlessness
To understand exemptions from the harmless error rule, we must first understand why we have such a rule. The rule that a conviction should not be reversed for an error that was of no consequence to the proceedings is an old one, existing by federal statute and in all of the states since the early twentieth century. (7) The purpose of such rules is clear: they "serve a very useful purpose insofar as they block setting aside convictions for small errors or defects that have little, if any, likelihood of having changed the result of the trial." (8) Such a rule had proven necessary because "courts of appeal became so concerned with procedural perfection that a simple misspelling of a word on an indictment was sufficient grounds for a new trial." (9)
Harmless error analysis in its simplest form is basically a weighing test: in the basic case of evidentiary error, for example, a reviewing court will "simply review the remainder of the evidence against the defendant to determine whether the admission of the [improper evidence] was harmless." (10) Different standards of certainty apply in different contexts, (11) but the basic idea is the same: the analysis is an almost mathematical weighing of the strength of the case minus the challenged element.
Prior to 1967, courts generally agreed that constitutional errors could never be harmless. (12) That is, the rule existed to prevent trivial, technical errors from requiring retrials or release of criminals, and was not intended to allow serious violations to go unremedied. Indeed, the text of Rule 52 of the Federal Rules of Criminal Procedure, one of the statutory homes of the federal harmless error rule, states that "[a]ny error, defect, irregularity, or variance that does not affect substantial rights must be disregarded." (13) The harmless error provision in the United States Code contains similar language. (14)
The case of Chapman v. California changed this understanding, declaring that constitutional errors could indeed be harmless, and holding that "the court must be able to declare a belief that it was harmless beyond a reasonable doubt," (15) though there are still "some constitutional rights so basic to a fair trial that their infraction can never be treated as harmless error." (16) Exactly which rights these might be has remained unclear, though the Court has made a series of attempts to clarify, (17) leading to the creation of the category of "structural error" in the 1991 case Arizona v. Fulminante. (18)
Trial errors, the Fulminante Court explained, occur during the presentation of the case to the jury, while structural errors affect the entire trial from start to finish. (19) As we shall see, this seemingly simple test did little to clear things up. We still have some errors that can be harmless and some that cannot, and although there are certain types of errors that fall, by consensus, on one side of the line or the other, (20) there remains a large gray area. (21) Nonetheless, there is widespread agreement that some errors should not be subject to harmless error review. The next Subpart examines which errors these are and why this is so. Understanding why these errors require reversal on appeal will shed some light on why they create such problems when they crop up in contexts that independently require a showing of prejudice.
B. The Five Faces of Structural Error: Why Not Do Harmless Error Review?
The general idea of structural error as defined in Fulminante is that some errors "defy analysis by 'harmless-error' standards." (22) That is, the usual harmless error analysis of looking at all the evidence and assessing the strength of the case simply does not answer the question of whether reversal is necessary in these cases, since the problem is not with the quantity or quality of evidence, but rather with some other aspect of the process.
The "defiance" described in Fulminante might take several different forms: (1) an actual impossibility of demonstrating effect on the outcome, since the effect, if it happened, is protected by the privacy of the jury room; (2) a real difficulty in assessing the effect because it would require speculation about how the trial would have unfolded; (3) an impossibility of assessing the effect because the problem with which we are concerned may not have affected the outcome at all, but requires reversal for some other reason; (4) an unwillingness to spend the effort on harmless error review because the error is so likely to be prejudicial; or (5) a sort of vague assertion that the error is too "serious" to be harmless.
What follows is an attempt at categorizing the different reasons why courts have found certain claims to be exempt from harmless error review. Courts differ in how explicitly they identify their motivations; sometimes they are quite clear, spelling out the reasons why harmless error review is inappropriate, and in other cases, the language of the opinion is conclusory and the reasoning must be inferred. These categories should not be taken as exhaustive or exclusive; it is entirely possible that a claim could be categorized as exempt for multiple reasons. This is simply an attempt to identify what those reasons might be, and to illustrate them with actual cases.
The first three categories--jury impossibility, speculative impossibility, and lack of effect on the verdict--represent the primary reasons why an error might require automatic reversal. The last two--obviousness and seriousness--are explanations that are sometimes given by courts, but lack any real explanatory power. Understanding these reasons will help us to see, in Part III, why this rule cannot give way to procedural requirements in habeas claims.
In assessing the remedy for errors, courts use the term "structural error" in two slightly different ways. They sometimes use it to refer to the cases specifically designated by Fulminante as structural, generally relying on the language from Fulminante that a structural error is one "affecting the framework within which the trial proceeds, rather than simply an error in the trial process itself." (23) Other courts simply label any error they conclude cannot be reviewed under a harmless error test as structural error, apparently using a circular definition of the term amounting to "a structural error is an error that cannot be assessed under a harmless error standard." (24) This confusion is understandable; David McCord has identified no fewer than three distinct definitions of "trial error" and "structural error" just within the Fulminante opinion itself, (25) and he correctly observes that the post-Fulminante case law has done nothing to clear up the confusion. Nonetheless, most instances of structural error can be categorized by the reason the error is exempt.
1. Jury impossibility
We must presume prejudice in some cases because they deal with some feature of the jury beyond just what they saw or heard in the courtroom. Because jury deliberations are secret, we cannot assess the effect of the error where it concerns the composition or conduct of the jury during deliberations.
That is, it would be theoretically possible to assess the harmfulness of the error, but our rules do not allow us to make the necessary observations. We acknowledge that the presence or absence of certain jurors can have a profound outcome on a trial--this is why we allow challenges for cause and peremptory challenges (26)--but we cannot know what that effect might have been without observing the deliberations.
A clear example of this type of "jury impossibility" is the denial of peremptory challenges. For example, in United States v. Annigoni, the Ninth Circuit held that prejudice must be presumed for the erroneous denial of a peremptory challenge:
[There is a] dearth of information concerning what went on in the jury room. To subject the denial of a peremptory challenge to harmless-error analysis would require appellate courts to do the impossible: to reconstruct what went on in jury deliberations through nothing more than post-trial hearings and sheer speculation. In the context of an appeal based on denial of a peremptory challenge, there is inadequate evidence for an appellate court to determine the degree of harm resulting from the seating of a juror despite a defendant's attempted peremptory strike. (27)
Another class of cases that illustrates jury impossibility pertains to those involving improper selection of grand or petit juries. (28) These cases are complicated and present a number of challenges that will be addressed in more detail below, but among their troubling features is the impossibility of assessing what effect the discrimination might have had on actual jury deliberations. The courts do not tend to identify jury impossibility as the reason for their holding that prejudice is presumed, but the cases do clearly illustrate the problem: even without all of Batson's complicating factors, we would not be able to assess the impact of the seating of one jury instead of another without listening in on the deliberations.
2. Speculative impossibility
In some cases, courts refuse to evaluate the effect of the error because doing so would require unguided speculation about what might have happened had things been different. Of course, there is always a degree of speculation involved in assessing an error, but in these cases, there is no principle by which to gauge the error. In a classic trial error, such as the wrongful admission of a piece of evidence, the court still must speculate to a degree, but it does so only within an almost mathematical framework of the weight of evidence and strength of the case. By contrast, in these speculative impossibility cases, it would be hard to know where to begin, let alone where to end up.
A classic structural error--one often cited by courts, including the Chapman and Fulminante courts, in attempting to illustrate the distinction--is the presence of a biased judge. (29) The vast majority of cases discussing this error simply cite Tumey v. Ohio, (30) a leading case on judicial bias, and either Chapman or Fulminante for the proposition that the error cannot be harmless. (31) Though no cases I have found actually explain the reasoning, the reason for the rule is apparent enough. In attempting to assess the effect of the error, where would one begin? With the judge's first words to potential jurors during voir dire? With the first ruling she made on a challenge for cause? With various evidentiary rulings? With the charge to the jury? With the judge's facial expressions during closing arguments? Would a judge without an ulterior motive have acted differently at any or all of these stages? If she had, would the jury have noticed? Would the jurors have allowed it to influence their decision? Because a reviewing court cannot begin to assess the error with any degree of accuracy, but cannot rule out its effect on the verdict, prejudice must be presumed.
A second example of speculative impossibility, which the Supreme Court has explicitly recognized as such, is the denial of defendant's counsel of choice. In United States v. Gonzalez-Lopez, the Court noted:
It is impossible to know what different choices the rejected counsel would have made, and then to quantify the impact of those different choices on the outcome of the proceedings. Many counseled decisions, including those involving plea bargains and cooperation with the government, do not even concern the conduct of the trial at all. Harmless-error analysis in such a context would be a speculative inquiry into what might have occurred in an alternate universe. (32)
In addition, the Supreme Court has identified speculative impossibility in determining the effect of attorney conflict of interest, holding in one such case that "an inquiry into a claim of harmless error here would require, unlike most cases, unguided speculation." (33) Though the Court has not been completely consistent in its explanations for presuming prejudice in cases of conflict of interest, (34)
3. Lack of effect on verdict
Lack of effect on the verdict itself (meaning that some societal value other than accuracy of trials is implicated) is a species of impossibility. In such cases, we cannot demonstrate prejudice because we are conceding that the result at trial may well have been the same without the error. Reversal is required for some reason other than effect on the verdict. Whether reversal is an appropriate remedy in these cases is a question outside the scope of this Note. Either way, some such claims have been recognized by courts, and how we treat them will determine whether they can ever be raised as instances of ineffective assistance or can ever overcome a procedural default.
Perhaps the most obvious such claim is the right to self-representation. Since 1975, the Supreme Court has recognized a Sixth Amendment right to represent oneself. (35) Denial of this right is an error despite the fact that the vast majority of defendants would receive better representation, and a better chance at a favorable outcome, if they had had counsel. Nonetheless, the Supreme Court has noted that this right's "denial is not amenable to 'harmless error' analysis. The right is either respected or denied; its deprivation cannot be harmless." (36) In these cases, the harm is to the defendant's dignitary interest in representing himself; if it were judged by the usual harmless error standard, the defendant would lose in almost every case. Instead, courts presume prejudice to protect the right, despite the lack of what would usually be considered "harm."
A slightly less clear-cut example is the public trial right, another of the oft-cited "classic" structural errors. (37) Though public trial errors are thought to have some potential effect on the outcome of the trial, in that abuses are less likely when the trial is in the public eye, (38) the right also serves societal values of transparency and integrity in the judicial process. Indeed, in the classic public trial case Waller v. Georgia, the Supreme Court noted that "'[t]he harmless error rule is no way to gauge the ... societal loss that flows' from closing courthouse doors." (39) In these cases, then, we may be able to measure some effect on the trial using a harmless error-type analysis, but we cannot measure the full effect of the error.
It is also possible that some jury composition claims would fit in this category--especially Batson claims, where the problem is with the process of selection, not necessarily with the jury ultimately seated, and the injury is at least in part to the excluded jurors rather than to the defendant. (40) Batson claims present a unique set of challenges with regards to harmless error review; they will be discussed in more detail below.
In another class of cases, the Court does not engage in harmless error review because it would be a waste of time and resources. As the Court put it in United States v. Cronic, a case involving counsel's complete failure to function as a part of the adversary system, "There are ... circumstances that are so likely to prejudice the accused that the cost of litigating their effect in a particular case is unjustified." (41)
For example, in Cronic, a young real estate lawyer was assigned to defend the accused in a complex mail fraud case a mere twenty-five days before trial, while the government had spent over four years developing its case. (42) The Court noted that "the likelihood that any lawyer, even a fully competent one, could provide effective assistance is so small that a presumption of prejudice is appropriate without inquiry into the actual conduct of the trial." (43) In other words, when the circumstances prevent "meaningful adversarial testing" (44) of the case, the presence of prejudice is obvious. Similarly, in Estelle v. Williams, (45) the Court held that forcing a prisoner to stand trial in jail garb is "so likely to be a continuing influence throughout the trial that, not unlike placing a jury in the custody of deputy sheriffs who were also witnesses for the prosecution, an unacceptable risk is presented of impermissible factors coming into play." (46) There was no need to consider prejudice separately.
These cases, of course, do not represent a species of impossibility, but rather quite the opposite. Still, I include this category as an example of a type of case where the Court has not required a demonstration of actual prejudice.
At times, courts have simply noted that some rights are essentially too important to subject to harmless error review. This category is not analytically useful; it provides no principle by which we can separate trial errors from structural errors, and seems to encompass some of the other categories, but it appears often enough in opinions to warrant attention. Indeed, Chapman v. California, the case that originally subjected constitutional errors to harmless error review, appeared to carve out some exceptions based on the seriousness of the error rather than any particular characteristic thereof. (47) More recently, the Supreme Court has defined structural errors as those "so intrinsically harmful as to require automatic reversal." (48) This category is perhaps best seen as a shortcut--a designation of an error as structural without choosing an actual justification.
Having surveyed the range of justifications for the structural error exemptions, the reasons why the exemption must apply across the board are perhaps beginning to become clear. But before proceeding to that analysis, I will pause for one more brief detour to examine the troubles presented by one particularly important claim that, like structural errors, raises issues when a showing of prejudice is required.
C. The Curious Case of Batson
Structural errors are not the only claims that are usually exempt from harmless error review. In the landmark 1986 case of Batson v. Kentucky, the Supreme Court declared it impermissible for lawyers to dismiss potential jurors based on race. (49) Though it seemed, for a time, that Batson claims had fallen out of favor with federal courts, the Supreme Court has recently breathed new life into this area, applying a searching analysis to the behavior of a Texas trial judge and providing lower federal courts with new tools to hunt out racial bias in jury selection. (50) These claims are extremely common, especially in capital cases, and thus deserve a brief discussion here.
Federal courts of appeals and the Supreme Court have presumed prejudice for Batson errors, though generally without specifically deciding that doing so is required. (51) State courts have been less consistent, but still tend to presume prejudice. (52) Courts have not generally articulated clear reasons for making this presumption. And perhaps they have good reason not to.
Batson claims represent an uncomfortable fit with a harmless error regime because requiring proof of a different outcome would entail the court doing exactly what it has forbidden the lawyers to do: making an inference about how jurors would decide a case based on their race. Thus, while the claims are not generally subject to harmless error review, (53) they do not fit neatly into the categories outlined above for requiring exemption. I thus pause to consider them here as a kind of close cousin to structural errors, raising many of the same issues along with several vexing issues of their own.
There are a number of ways to approach this problem, and it is not at all clear that courts--or judges within a court--agree on the proper approach. I am not suggesting that one approach or another is necessary; rather, a survey of the possible approaches will suggest that no matter what route any given court takes, it should never require a showing of prejudice for a Batson claim in any context. In other words, all of these explanations should apply any time a Batson claim comes up, whether on direct appeal, in the simple habeas context, after procedural default, or as an instance of ineffective assistance of counsel.
Here are a few possible rationales:
1. Garden-variety jury impossibility
At least some courts have thought that demonstrating actual prejudice is not impossible at all. In Hollis v. Davis, for instance, the Eleventh Circuit actually found prejudice where the jury was all white and selected by a racially discriminatory process. (54) Likewise, the judges dissenting from denial of rehearing en banc in the Ninth Circuit case of Williams v. Woodford argued that the all-white jury likely produced a different result from a mixed-race jury. (55) These claims, while rare, are especially strong where a black defendant was convicted by an all-white jury. (56)
These courts express the belief that a mixed-race jury would have decided the issue differently. But then, they never actually prove the probability of a different outcome. Neither court totals up all the evidence and pinpoints what the damage would have been. They just assume there must have been an effect, given the racial composition of the jury, the race of the defendant, and the nature of the case. (57) Thus, these cases essentially fit within the category of "jury impossibility." Though claiming to assess the error, the courts are actually forced to presume prejudice because they cannot look into the jury room and see what transpired, nor can they speculate about what a different jury would have done.
2. "Pay no attention to the man behind the curtain"
Because Batson presents a conceptual paradox--requiring the simultaneous belief that race does not matter and that it does (58)--courts avoid further inquiry into the situation. They implicitly recognize that the rule will not bear logical weight, but that it still protects a value (nondiscrimination) that we as a society hold dear. Eric L. Muller has explicated the paradox and the conceptual inconsistencies on each side of the debate in great detail, and has made a convincing case that the Batson rule is, in its current form, incoherent. (59) Few if any of the Batson decisions actually explain the reasons for eschewing harmless error review, perhaps because there is no logically consistent reason for it. Any attempt at explaining would expose the logical flaws and perhaps force the court to abandon the idea.
Thus, rather than digging around and exposing this difficulty or turning away from a troubling discriminatory practice, courts look the other way and require reversal. (60) A showing of prejudice is not impossible in the usual sense of the word; courts simply do not want to look into it, because they might be forced to reconcile inconsistent viewpoints and sacrifice one treasured value or the other. If this is the reason, it should apply in any context; courts would no more want to decide whether counsel's error in failing to raise a Batson objection altered the outcome of the trial than they would want to dig into the question raised on direct appeal. Likewise, if a defaulted Batson claim is raised, and a good case for violation made, a court should not want to go through the prejudice analysis to ask if the Batson error might have changed the verdict, because it would require getting into the logical quagmire.
3. True impossibility
Perhaps equally troubling is the suggestion that it is impossible to show effect on the verdict because there is no effect: the defendant is not the one who was injured. (61) There are other reasons for requiring reversal in Batson cases; two in particular have been discussed widely.
First, automatic reversal adds to the deterrence value of these cases. Without automatic reversal, there would be no punishment for a prosecutor who used peremptory challenges impermissibly, since no defendant would be able to secure reversal based on the harm he suffered from the error. None of the other possible remedies--allowing excluded jurors to bring civil suits to vindicate their equal protection claims, for instance--would cause prosecutors to lose much sleep at night. Indeed, the Supreme Court has clearly rejected a range of alternative remedies for cases involving discrimination in the selection of grand juries. (62)
Harmless error review makes no sense when the reason for reversal is not harm to the defendant. Though there are "costs associated with this approach" of reversing a potentially accurate verdict, the costs "are outweighed by the strong policy the Court consistently has recognized of combating racial discrimination in the administration of justice." (63) Thus, even if it were possible to peer into the jury room and determine if race somehow played a role, it would not matter; reversal serves not to correct mistaken verdicts, but to make any strategy based on discriminatory jury selection entirely ineffective.
A second non-accuracy-based reason to reverse for Batson violations concerns the integrity of the judicial process. This harm is twofold: It is "not only to the accused.... It is to society as a whole." (64) In other words, there is indeed a harm to the defendant in being tried by a less-than-upright system, but we need not rely on that harm to demand reversal. Reversal serves instead to combat "cynicism respecting the jury's neutrality and its obligation to adhere to the law." (65)
This category of non-accuracy-related reasons is the most controversial, since it may allow defendants who had perfectly factually reliable trials to get another chance for reasons that have nothing to do with guilt or innocence. My purpose is not to question the wisdom of such a regime but rather to suggest that if we have decided that this is the rule for such cases, we ought to carry it through. If we need reversal for deterrence or to combat cynicism, then we need it whether or not defense counsel objected at precisely the right moment.
As a side note, this class of non-accuracy-related reasons might be the strongest argument against presuming prejudice for procedurally defaulted Batson claims. Habeas is not about vindicating societal rights; it exists to get people out of prison who, under the terms of the Constitution, really should not be there. A defendant whose trial included an equal protection injury to someone besides the defendant would arguably be getting a "free pass"--he is not one of those people that habeas is designed to protect.
This argument, however, leaves a void; the Supreme Court has been clear that even where the injury was primarily societal, the defendant's rights were also violated--he had a right to a trial with integrity, and a trial tainted by Batson violations is not such a trial. (66) Thus, even if we were required to leave aside the societal violation, we would be left with a loss analogous to the loss of the Faretta right to represent oneself (67)--the defendant was entitled to a certain type of process, whether its loss actually altered the verdict or not. Though it may seem costly to invalidate convictions to preserve these sorts of dignitary rights, the Supreme Court has been quite clear that reversal "does not render a defendant immune from prosecution, nor is a subsequent reindictment and reprosecution barred altogether." (68)
4. No valid verdict
In Sullivan v. Louisiana, the Supreme Court held that a faulty reasonable doubt instruction called for automatic reversal not because it infected the whole trial, but because no valid verdict of guilt beyond a reasonable doubt had been entered to begin with. (69) The defendant could not show that this verdict was different from the verdict he would have gotten absent the error because as a technical matter, there was no verdict.
A similar argument could be (but to my knowledge, has not been) made that Batson errors produce the same problem: the jury rendering the decision was not seated in accordance with the Constitution, so as a constitutional matter, it was not actually a proper jury, and its verdict was thus not a proper verdict. We cannot compare the old verdict to a hypothetical new verdict because there is no old verdict.
As a variation on this argument, we could compare a verdict delivered by a wrongly selected jury to a directed verdict of conviction in a criminal case--that is, the decision was not made by the right decisionmaker, and is thus invalid regardless of its factual accuracy. (70)
No matter which reason a court adopts for declining to engage in harmless error review--and I suspect that different courts, if pressed, would give different reasons--courts have been clear that we cannot require defendants to show that their trials would have turned out differently but for discrimination in jury selection as a condition of granting relief. (71)
It might be objected that no matter which explanation a court chooses, presumption of prejudice for Batson claims will create a mammoth practical problem on collateral review. If a Batson objection was not made at trial, there will likely not be a good record of the prosecutor's reasons for excusing the jurors he excused. (72) If the objection is contemporaneous, the judge can simply ask the prosecutor for race-neutral explanations at the time; but if nobody mentions the claim until years after the fact in a habeas petition, and the defendant then complains that the prosecutor's jury selection was improperly motivated, the prosecutor may well have forgotten the details of his rationale for excusing particular jurors from the long-ago trial.
Where a showing of actual prejudice is required to overcome the failure to raise the claim at the time, most, if not all, of the claims fail on that basis because, as just discussed, such a showing is not generally possible. (73) But if prejudice is not required to overcome procedural default, some of the claims will begin to proceed to the merits, as surely some defendants will demonstrate cause.
Defendants, assuming they can demonstrate cause, would only have to make a prima facie case that there had been discrimination. Prosecutors would then have the burden of rebutting that case, but will often lack the tools they would need to respond to the objection, even if there had been no hint of discrimination in their behavior. The cases arising after the fact in this manner would be unwinnable for prosecutors, and the Batson scheme for assessing motives would break down.
As distressing as this may seem, however, it is just another symptom of the disease ailing the Batson doctrine; the problem would not be caused by the lack of a prejudice requirement, but rather exposed by it. If Batson claims work very poorly when raised for the first time on collateral review, courts perhaps ought to address that question separately. The fact that Batson claims often involve no demonstrable effect on the actual verdict has nothing to do with the practical proof difficulties faced by prosecutors years after the fact. Thus, the prejudice requirement is useful to courts in this scenario, in that it prevents this vexing problem from ever reaching them. But it presents no solution.
We have now seen that harmless error analysis exists to prevent trivial errors from derailing convictions, that various forms of impossibility require a set of so-called structural errors to be exempt from the prejudice requirement, and that Batson claims are also generally exempt, though the reasons for exempting Batson claims are somewhat murkier. The next Part will consider an entirely separate line of cases, concerning procedural default, developed without reference to the structural error cases. But keep the structural error cases in mind; we will return to them in Part III, where we will consider their interaction with the procedural default cases. Then, we will see how lower courts have taken different stances on how to resolve the conflict between the two strands. But first, we will examine the background of procedural default.
II. PROCEDURAL DEFAULT AND THE CAUSE AND PREJUDICE EXCEPTION
The above Part has established that the chief reason for exempting structural errors from the prejudice requirement of the harmless error rule is that it would otherwise be impossible to satisfy the test. This Part will explore a rule that developed entirely independently, without reference to structural error. The subsequent Part will tie the two together.
A. Purpose and Mechanics of the Procedural Default Rule
Federal habeas corpus review of state criminal convictions exists to protect important constitutional rights that have not been protected in the state process.
It functions as a last resort for those whose trials were not conducted in accordance with the Constitution. (74) This review is provided by statute and functions not as an additional appeal but rather as a collateral attack on conviction, challenging some aspect of the trial process under federal law. (75)
If a petitioner has violated a state procedural rule at an earlier, pre-federal habeas point in the proceedings--for example, by missing a deadline, or by omitting a possible claim at one level of appeal and then attempting to raise it later--he may be barred from presenting the claim on federal habeas because there is an "adequate and independent state ground" for rejecting the claim, which is then said to be procedurally defaulted. (76) In other words, the state court would have rejected the claim, based not on the merits of the claim under federal law, but instead on its own state law rules; under these circumstances, the federal court cannot hear the claim. States can reject claims for substantive or procedural reasons, and both must be given effect by reviewing federal courts.
The Supreme Court, however, has treated state procedural grounds for rejecting claims in federal habeas review of state criminal convictions differently from substantive grounds. The Court initially exempted procedural errors from the adequate and independent state ground doctrine whenever it appeared that the error was unintentional; (77) later, it reinstated the requirements that federal courts reject claims that the state court would reject, even where the errors were accidental. (78) The unique situation of a federal court reviewing state criminal convictions supplies the rationale for this requirement:
In the habeas context, the application of the independent and adequate state ground doctrine is grounded in concerns of comity and federalism. Without the rule, a federal district court would be able to do in habeas what this Court could not do on direct review; habeas would offer state prisoners whose custody was supported by independent and adequate state grounds an end run around the limits of this Court's jurisdiction and a means to undermine the State's interest in enforcing its laws. (79)
The requirement that federal courts reject claims with procedural defects under state law is not, however, absolute; in bringing back the stricter requirements for adherence to state procedural rules, the Court included an exception, allowing claims to be heard despite procedural default if the petitioner could show cause for failing to raise the claim and prejudice from the error. (80) "Cause" means "the prisoner can show that some objective factor external to the defense impeded counsel's efforts to comply with the State's procedural rule," (81) such as ineffective assistance of counsel, unavailability of the factual or legal basis for the claim, or interference by the state. (82) "Prejudice" lacks even such a vague definition--it requires some kind of effect on the outcome, but beyond that, courts have not been specific or consistent as to the precise requirements of the test. (83)
These exceptions are clearly an attempt by the Court to create a safety valve; indeed, the Court has articulated a belief that the exceptions "will afford an adequate guarantee ... that the [procedural default] rule will not prevent a federal habeas court from adjudicating for the first time the federal constitutional claim of a defendant who in the absence of such an adjudication will be the victim of a miscarriage of justice." (84) This stated commitment to continuing to provide justice in the face of a procedural default is significant, considering the application of this rule in the case of structural errors.
B. Ineffective Assistance as Cause
One of the most common factors invoked and accepted by courts as "cause" is ineffective assistance of counsel, which, to add to the complication, incorporates its own prejudice requirement. Courts have been clear that to qualify as cause, ineffectiveness claims must meet the standard established by Strickland v. Washington--a test that requires both a showing that counsel's performance was deficient, and that there is a "reasonable probability" that but for the error, the outcome would have been different. (85) Strickland, of course, primarily exists as a stand-alone claim--a common one in habeas petitions. But the Court has been clear that to satisfy the cause requirement, the petitioner must satisfy Strickland as he would if he were raising it independently. (86)
A petitioner attempting to raise a defaulted claim and using ineffective assistance as cause essentially must make two showings of prejudice: first, he must demonstrate that his claim satisfies the Strickland prejudice prong, and then he must satisfy the prejudice test to overcome the procedural default. While these two requirements are technically separate, and need not use the same standard, most courts evaluating such claims have collapsed them, holding that a satisfaction of the Strickland standard satisfies the Sykes prejudice inquiry, or vice versa. (87)
In considering this role for Strickland, it is important to keep in mind that ineffectiveness is not the only way to demonstrate cause; any factor extrinsic to the case theoretically qualifies. Although ineffectiveness is the cause in many cases, a newly available factual or legal basis, or interference by government officials, for example, could qualify, (88) and there would be no ready-made prejudice analysis from the "cause" half of the test available to incorporate. The prejudice inquiry of Strickland is thus analytically distinct from the prejudice inquiry in Sykes.
While collapsing them is certainly convenient in a situation where both must be proven, we should bear in mind that they are not actually the same question. To satisfy the Strickland test, a "defendant must show that there is a reasonable probability that, but for counsel's unprofessional errors, the result of the proceeding would have been different. A reasonable probability is a probability sufficient to undermine confidence in the outcome." (89) The Sykes Court, on the other hand, left "open for resolution in future decisions the precise definition of the 'cause'-and-'prejudice' standard." (90)
Further, we must bear in mind that Strickland exists independently, as a substantive claim that can be raised on direct appeal or on habeas as its own grounds for relief under the Sixth Amendment. Though the Supreme Court has explicitly imported the Strickland test into the cause inquiry under Sykes, (91) such a move was not logically necessary. In theory, the Court could have defined the level of ineffectiveness required to overcome procedural default as different from that required for a freestanding Sixth Amendment claim. It is also worth noting that Sykes was decided in 1977, seven years before Strickland. It was only later that the Strickland standard was imported into this context, suggesting that the two need not be one and the same.
In sum, a petitioner raising a procedurally defaulted claim must satisfy the cause and prejudice test set forth in Sykes; petitioners using ineffective assistance as cause must also satisfy the Strickland test for prejudice. Sykes prejudice has never been explicitly defined, and while the two are analytically distinct, courts generally collapse them. Now that we have examined both the structural error rule and the procedural default rule in their separate contexts, we can see what happens when a structural error suffers from procedural default.
III. STRUCTURAL ERRORS MEET PROCEDURAL DEFAULT: THE CONFLICT
A. The Problem with Sykes Prejudice
If, as the Court has stated in its structural error cases, it is impossible to show prejudice--however defined--in some types of error, and if the rights implicated by structural errors can be sufficiently important to merit a grant of the writ of habeas corpus, (92) then setting up a system whereby it is impossible to make the required showing to overcome the default is as good as saying that structural errors, as opposed to trial errors, cannot be rescued from procedural default. A trial error might be salvageable via the cause and prejudice exception, but a structural error never is, because structural errors by definition cannot give rise to a showing of actual prejudice. They are thus doomed to fail this test in every instance.
Such a position would be absurd if the Court were to come right out and say that this is the rule, given the general agreement (no matter which definition you choose) that most structural errors are quite serious. (93) There is simply no justification for singling out this class of serious errors as ineligible to be considered on account of procedural default. Yet as the rules currently stand, a court rigidly applying them would reach this result. Indeed, Vansickel was denied the chance to raise his claim by a court that was not attentive to the clash.
Both of these rules are Court-created; the Court has defined which errors should be presumed prejudicial, and the Court has set the standards for overcoming a procedural default. It is thus up to the Court to reconcile its rules to yield outcomes that serve the goals they have articulated. One of the rules must give way.
Indeed, the Supreme Court seems to suggest that in a clash, the procedural default rule ought to yield. The Court has specifically noted that "[t]he terms 'cause' and 'actual prejudice' are not rigid concepts; they take their meaning from the principles of comity and finality.... In appropriate cases those principles must yield to the imperative of correcting a fundamentally unjust incarceration." (94) In other words, justice trumps comity. And yet, lower federal courts have not consistently applied this principle, and the Supreme Court has not stepped in. (95) This lack of clarity has left lower courts in a difficult position. If they attempt to comply with both commands, they will leave petitioners with no viable options; otherwise, they must choose to disobey one or the other rule handed down from above.
B. The Problem with Strickland Prejudice
Strickland claims are, in a way, more complicated because of their dual role. The considerations differ depending on whether the claim is a stand-alone claim of constitutional violation or whether it serves as cause. Specifically, as a stand-alone claim, its standard for what constitutes a violation of the Sixth Amendment is Court-defined. As cause, it serves as a procedural barrier to rights that have been separately guaranteed; while the Court may have set the standards for any given right, it has not done so with reference to Strickland prejudice, and the right as granted should not be abridged by a separate procedural barrier. This Subpart will tease out the differences between these two contexts.
In the substantive context, the Supreme Court has clearly and consistently defined ineffective assistance of counsel as a two-tiered inquiry (96) and has explicitly specified a set of cases where the prejudice prong of that inquiry is in- applicable. (97) Though we might quibble about which cases should be in that exempt category, and whether such a category is exhaustive or is better understood as evidence that presumption is sometimes acceptable, the Court has been forthright about its definition of what constitutes constitutionally ineffective assistance. Where there is no proof of prejudice, there is no violation. Theoretically, the Court could define the contours of this particular piece of the Sixth Amendment any way it chooses--including a formulation with an unyielding prejudice requirement.
But that formulation is not without its problems. If the rule is interpreted rigidly, without accommodating structural errors, petitioners will never be able to secure a new trial when counsel's error was failure to raise a structural error; and as we have established, structural errors are quite serious. Can the Court really mean that it is constitutionally ineffective for a lawyer to fail to raise objections to admission of the wrong sort of evidence, but that there is no constitutional violation if the same lawyer fails to object to an obviously biased judge? A thorough exploration of the contours of the basic Strickland rule is beyond the scope of this Note, but if and when the Court takes up the issue in the habeas context, it will have impact in the substantive Strickland context as well.
Delineating the contours of a Sixth Amendment right in nonsensical ways is one type of problem. Altering the availability of remedies for clearly defined rights is a different story altogether, and is more obviously problematic. (98) For instance, courts have been clear about the right to a trial without a biased judge; (99) this right is not defined by reference to prejudice. In contrast to ineffectiveness claims, in biased-judge claims there has been a violation whether or not a showing of prejudice has been made. Adding a prejudice requirement that cannot be satisfied to allow access to a remedy interferes with the substantive right that has been separately granted.
Thus, ineffective assistance as cause is conceptually different from ineffective assistance as a stand-alone Sixth Amendment violation. This is not to say that requiring prejudice for substantive claims is or is not fair where that showing is impossible; that question is beyond the scope of this Note. Instead, the point is that the conflict of standards is raised more squarely where ineffectiveness is used as cause. Some of the case law discussed below surrounding presumption of prejudice for Strickland claims deals only with the "substantive ineffectiveness claim" context; (100) because courts have addressed the true conflict so rarely, a consideration of these cases is necessary for an understanding of judicial thinking on the issue. Still, we must bear this difference in mind when reviewing the cases.
C. Counterargument: Is Habeas Different After Brecht?
It may be objected that there is no problem with making rights more difficult to vindicate on collateral review than they are on direct appeal. Indeed, the Supreme Court has been quite clear that standards on habeas may be more strict than standards on direct review. In particular, the 1993 case of Brecht v. Abrahamson explicitly approved a more demanding harmless error standard for claims being considered on habeas (without procedural default) than for claims being considered on direct review. (101) In Brecht, the Court explained that direct review is the primary vehicle for addressing errors in state court convictions, where habeas is meant as an extraordinary remedy for cases that would otherwise end in real injustice. (102) Why, then, should courts be forbidden to make claims less available on habeas than they are in other contexts?
The Brecht reasoning that habeas is different presents no answer to the problem at issue in this Note. The main concern is not that claims in general are more difficult to raise on habeas, but rather that these particular procedural requirements single out a group of serious violations even within the generally applicable habeas standards. In other words, Brecht allows claims to be treated differently when they arise in habeas petitions, as opposed to direct appeals, but does not provide any justification for treating certain claims differently within the group of habeas claims. The clash at issue here renders some serious claims unavailable while leaving others subject to usual habeas standards; if such a rule operated on direct review, it would be equally objectionable. Brecht thus provides no excuse.
We have seen that the Court has set up two standards that collide in certain cases; that the Court claims to envision these rules as safeguards, rather than rigid barriers; and that there is no good reason why some constitutional claims should be foreclosed while others are allowed. We will now turn to an examination of how courts have addressed--and failed to address--the conflict when it has arisen.
IV. THE MESS IN THE CASE LAW AND THE SOLUTION
Courts have not examined the question of presumed prejudice for structural errors in Strickland or Sykes contexts with any clarity. Four Supreme Court decisions--two from the 1970s and two more recently--have touched on the question without giving conclusive guidance, and the Courts of Appeals have grappled with it in various forms without coming to any clear conclusions. There is a circuit split on the question of presuming Strickland prejudice for structural errors, but the procedural default question has been mostly ignored: courts have either failed to recognize or acknowledge the problem, or they have decided that a presumption is necessary without examining the complexities or the possibilities. No decision I have found looks at only Sykes prejudice in this context; instead, where Sykes prejudice is discussed, it is always combined with Strickland prejudice. (103) Because, as established above, the two are not interchangeable, courts have obscured the issues in conflating them. A clearer, more reasoned analysis is necessary. What follows is an overview of the reasoning courts have supplied in resolving these cases, along with a suggestion for how to resolve them more clearly and fairly in the future.
A. United States Supreme Court
Two early cases--predating both Strickland and Fulminante--suggested that there need not be congruence between the treatment of a right in its substantive form and its treatment as grounds for overcoming waiver. In Davis v. United States, (104) the Court, in examining a grand jury discrimination claim waived under the Federal Rules of Criminal Procedure, held that "the presumption of prejudice which supports the existence of the right is not inconsistent with a holding that actual prejudice must be shown in order to obtain relief from a statutorily provided waiver for failure to assert it in a timely manner." (105) Three years later, the Court extended this holding to state waiver rules in Francis v. Henderson. (106)
Though a handful of lower court opinions have cited these two cases in holding that Strickland prejudice need not be presumed, (107) Francis and Davis do not seem to require this result, for several reasons. First, Fulminante and subsequent cases discussing structural error significantly changed the landscape surrounding the presumption of prejudice. The First Circuit put it well in Owens v. United States: "[T]he holding of Francis has been substantially weakened by the Supreme Court's subsequent pronouncement in Fulminante, Sullivan, and Neder that prejudice is impossible to quantify in cases of structural error." (108) In other words, Francis may apply to some rights requiring automatic reversal, but it cannot apply across the board to "structural errors," which did not yet exist as a category when the case was decided.
Second, these cases, even to the extent that they are still good law, do not compel any given result. They merely hold that the presumption of prejudice associated with a right does not require the court to presume prejudice to excuse a waiver. They certainly do not forbid courts from making such a presumption, should they find it necessary.
Additionally, by their terms, Davis and Francis apply only to a statutory waiver. The procedural default doctrine at issue in Sykes has its root in the judicially created adequate and independent state ground doctrine, and the prejudice requirement in Strickland was likewise announced not by a legislature, but by the Supreme Court interpreting the requirements of the Sixth Amendment. Though the state rules at issue, which the state courts relied on in refusing to hear claims, may be statutory, the default actually at issue in the federal habeas petitions is the adequate and independent state ground doctrine--not any given statutory rule.
Finally, Davis and Francis were both decided during the death penalty moratorium in the 1970s--Davis on April 17, 1973, and Francis on May 3, 1976. The death penalty had been put on hold by Furman v. Georgia in 1972; (109) death sentences were not allowed to resume until after Gregg v. Georgia, which was issued July 2, 1976. (110) Thus, when the Court held that the presumption of prejudice that supported the existence of a right was not inconsistent with requiring a showing of actual prejudice, no one's life was on the line. This observation, of course, changes nothing about the legal force of the opinions, but may suggest a reason why the issue is more pressing now than it was at the time.
More recently, the Supreme Court grappled with a related issue in a cluster of cases examining the reach of United States v. Cronic. (111) The chief case, Bell v. Cone, (112) dealt only with the question of whether the defense counsel's failings were sufficiently severe to warrant a presumption of prejudice, without touching on structural error. But a dizzying series of decisions in a second case, Bell v. Quintero, veered more into relevant territory. In Bell v. Quintero, the Sixth Circuit initially presumed prejudice for a Strickland claim for failure to object to a biased jury presented both as cause and as a substantive claim; (113) the Supreme Court granted, vacated, and remanded for consideration in light of Bell v. Cone's reiteration of the boundaries of Cronic; (114) the Sixth Circuit again presumed prejudice in reinstating its previous opinion; (115) and the Supreme Court denied certiorari, drawing a dissent from Justice Thomas. (116) The Sixth Circuit's holding on remand was framed not as a requirement that prejudice be presumed for cases of structural error, but rather as a rule that failure to object to a certain type of error constitutes complete denial of counsel--meaning that such cases are governed by Cronic (which entails a presumption of prejudice), not by Strickland (which requires that actual prejudice be shown). In other words, while the Sixth Circuit may have appeared to be saying that Strickland prejudice should be presumed for structural errors, the court was actually saying that failure to raise this particular error is more deficient than failure to raise most errors, pushing it into the category of failure to provide any kind of meaningful adversarial testing of the case.
While the Supreme Court disagreed with the Sixth Circuit's original application of Cronic to the facts of Bell v. Quintero, it was not dismissing an argument about impossibility, but was rather holding fast to the existing limits of Cronic. If this seems like a purely semantic distinction, it is important to realize that neither the Sixth Circuit nor the Supreme Court actually addressed the suggestion that prejudice should be presumed due to the impossibility of its demonstration. The Sixth Circuit rested purely on an interpretation of existing case law on one reason for presumption, and did not address the conflict of standards. The distinction is subtle, but the grounds for requiring reversal are slightly different from those I am suggesting. Further, the Sixth Circuit made no attempt to distinguish between the prejudice required for a substantive claim and the prejudice required for ineffectiveness as cause. (117) It thus failed to recognize the contours of the conflict.
Most recently, in Premo v. Moore, (118) the Supreme Court considered a substantive Strickland claim where counsel's alleged shortcoming involved failure to prevent a structural error. This case ultimately does not answer the question at issue here, for two reasons. First, there was no procedural default issue in the case; it concerned a substantive Strickland claim that was properly preserved throughout the process. Thus, its application to the cause inquiry is uncertain, and depends on the Court's continued commitment to absolute congruence between substantive Strickland claims and ineffective assistance as cause. Though the Court in Murray v. Carrier did of course adopt the Strickland standard where ineffective assistance is the Sykes cause, (119) it has not considered the intersection of this decision with the structural error cases and did not do so in Premo. Further, the case has absolutely nothing to say about Sykes prejudice and would have no relevance to a situation where procedural default was excused by some other variety of cause.
Second, Premo concerned a plea bargain, and the Court's rejection of a presumption of prejudice rested on this fact. Specifically, the Court noted that a "defendant who accepts a plea bargain on counsel's advice does not necessarily suffer prejudice when his counsel fails to seek suppression of evidence, even if it would be reversible error for the court to admit that evidence." (120) The prejudice question the Court considered was "whether Moore established the reasonable probability that he would not have entered his plea but for his counsel's deficiency," and "not whether Moore was sure beyond a reasonable doubt that he would still be convicted if the extra confession were suppressed." (121) Though the Supreme Court has indicated that it is possible to have a structural error in a case that ended in a plea bargain, (122) it is also entirely possible to have demonstrable prejudice in a plea bargain context (by satisfying the test from Hill v. Lockhart, which requires a showing that but for counsel's errors, the defendant would have gone to trial (123)) even though no showing would be possible in a trial context. The Premo Court's decision that the defendant could have suffered prejudice, but did not, thus does not decide the issue of whether a defendant who suffered ineffective assistance at trial in the form of failure to prevent a structural error must demonstrate actual prejudice for claims where such a showing is categorically impossible or otherwise not generally required.
In short, then, none of the four Supreme Court decisions that approach the territory of this conflict have said anything directly on point.
B. Courts of Appeals
At the level of the federal courts of appeals, there has been some discussion of the impossibility of demonstrating prejudice for structural error. Of the handful of cases discussing the problem, several deal with Strickland prejudice alone, never invoking the procedural default context. Though these cases are not directly concerned with the problem I have detected in the procedural default context, I discuss them here because they are the best indication we have of judicial thinking on the issue. Further, none of the courts looking at the issue made the important distinction between Strickland claims in the cause context and substantive Strickland claims, and there is no indication thus far that courts recognize the difference between the two contexts, nor that courts have recognized any actual or potential difference between Strickland prejudice and Sykes prejudice. As I will discuss below, combining the two is not problematic where the court presumes prejudice for both, but some of the cases requiring actual prejudice commit a logical error in combining them.
1. The main split: to presume or not to presume?
Three circuits--the First, the Sixth, and the Eighth--have held that Strickland prejudice must be presumed for structural errors, since a showing of actual prejudice would be impossible. (124) The argument that prejudice must be presumed in cases of impossibility has also surfaced in a handful of dissents. (125) Meanwhile, the Fifth and Eleventh Circuits have required a showing of actual prejudice, without addressing the fact that there is no way for defendants to make such a showing. (126) A dissenting judge in the Sixth Circuit even recognized the direct conflict between these two camps. (127) The Eighth Circuit, though grouped with the states presuming prejudice, seems to require the showing for Batson claims but not for structural errors generally. (128) The Ninth Circuit has reserved judgment, explicitly identifying the question as open. (129) (Though an earlier Ninth Circuit case required a showing of actual prejudice, (130) Withers, the most recent case in the Circuit, declares it an open question without citing the earlier case. (131))
Many of these cases do not deal directly with procedural default, either because there was no defaulted claim, or because the court disposed of the claim without reaching a Sykes prejudice inquiry. Of the cases that do consider and actually decide cause and prejudice, however, the more recent cases all presume prejudice, where older cases do not. Table 1 collects the cases implicating both the Sykes inquiry and the Strickland question.
These cases give rise to two observations. First, there is a neat dividing line as to when the cases were decided; all the cases requiring a showing of actual prejudice predate all of the cases presuming prejudice. Though the cases generally fail to cite one another, and thus it would be a stretch to consider this a genuine trend, it does suggest a general increased awareness of this problem. It may also be due to other factors, such as the major changes to habeas litigation engendered by the 1996 enactment of the Antiterrorism and Effective Death Penalty Act (AEDPA).
Second, absent in five of these six cases is any suggestion that Sykes prejudice and Strickland prejudice ought not be considered as coextensive. Thus even in reaching what I would consider to be the right result, the courts of appeals have obscured the nub of the issue. The only case to discuss the potential difference--Jackson v. Herring--identifies the fact that courts have defined the prejudice standard differently in different contexts, and requires that the Strickland standard be met to satisfy cause before proceeding to an analysis of prejudice. At no point, however, does the court address the reason why the standards must be kept separate. (132)
2. Why collapsing Strickland and Sykes prejudice won't work
This constellation of cases also reveals an additional important reason to differentiate between Strickland prejudice and Sykes prejudice: one of the chief arguments marshaled in requiring prejudice even for structural errors is applicable only to Strickland. A number of courts, apparently including the Supreme Court in the Quintero case, (133) rely on the following chain of reasoning: when the Supreme Court decided Strickland, it also discussed Cronic, a case recognizing a few very limited situations in which no prejudice is required to make out an ineffective assistance claim; the categories spelled out in Cronic, including complete failure of counsel to function as an advocate and actual attorney conflict of interest, compose the exclusive group of claims for which ineffective assistance does not require a showing of prejudice; that list does not include "structural errors"; therefore, we cannot presume prejudice for structural errors. Of the cases I found that required a showing of actual prejudice, three (counting the entire Quintero group as one) out of six explicitly rely on this argument. (134)
Not only is this reasoning flawed on its own terms--the existence of the Cronic cases could just as easily be evidence that presumption of prejudice is at times appropriate and should therefore be allowed for structural errors (135)--it is also inapplicable to Sykes prejudice. For courts treating Sykes prejudice and Strickland prejudice as a single inquiry when both are raised, (136) the error is clear. Because none of these cases raise Sykes with a different variety of cause, no court has confronted the Sykes/structural error issue without the Strickland issue. But if and when they do, this Cronic-based argument will not be available. It thus cannot function to make any sort of claim about Sykes prejudice. (I would also note that of the three cases rejecting presumption where the Sykes issue is presented, none explicitly rely on the Cronic reasoning, though the Eleventh Circuit cases, as read by the later case Purvis v. Crosby, actually do rely on this logic.) While cases exposing this shaky logic have not yet arisen, courts should not continue to decide cases based on a chain of reasoning that will, sooner or later, fall.
3. What do courts do when they refuse to presume?
The cases requiring an actual showing (those in the right-hand column in Table 1 above) generally do not confront the issue of impossibility even in passing, suggesting that courts are not insensitive to the possibility that defendants will be placed in an impossible bind, but rather that they have not realized that their decisions create this situation. One case in the Eighth Circuit--the above-mentioned Batson case Young v. Bowersox--confronts a petitioner's argument that a showing of prejudice would be impossible, but brushes it aside, claiming to be bound by a prior case and showing little concern for the appellant's predicament. (137) Otherwise, courts seem oblivious to the problem.
Some of the opinions simply say that a showing of prejudice has not or cannot be made, and that's that. (138) This makes sense for the jury and speculative varieties of impossibility; once committed to looking for prejudice, the court seems to have no choice but to point out that there is none.
More troubling, though, are the opinions that actually attempt to engage in some form of harmless error analysis. For example, in Vansickel, the court acknowledged that defense counsel likely would have stricken at least one additional juror had he been given the proper number of peremptory challenges, but then went on to explain:
The evidence against Vansickel was overwhelming. He was found at the murder scene with the murder weapon in his hand. He admitted to his former girlfriend that he had shot Howard. In the letters he wrote in jail, he stated that he fabricated the story about Dan. Further, at trial he conceded that he had, in fact, shot Howard. (139)
This has the character of a classic harmless error review, but it ignores the fundamental feature of this claim: the question is not whether the jury as seated would have seen things differently minus any given piece of evidence, in which case this weighing makes sense. Rather, it's whether a different jury would have reached the same result. The dynamics of juries are complex, and any number of factors could change the course of the discussion. The court gives no consideration to the identity of the potentially challenged juror or his replacement, and makes no distinction between the two types of analysis. Worse yet, the court acknowledges that harmless error analysis is not appropriate for these claims on direct review, but claims that it can be so on habeas, without explanation. (140)
This scenario of an actual attempt to perform harmless error analysis is more obviously problematic in the Batson context. Rather than presuming prejudice in the "pay no attention" scenario, (141) a few of the courts ventured into the unsettling territory of looking for prejudice in Batson cases. For example, in Jackson v. Herring, the Eleventh Circuit found no prejudice from racial discrimination in jury selection, partly on the grounds that "the crime in this case did not have any particular racial dimensions, which would cast doubt upon a verdict returned by a racially unbalanced, unconstitutionally composed jury." (142) The implication--that only a racially balanced jury could provide an accurate verdict in a case with racial undertones--is unsettling. The Jackson court's conclusion reads like a misunderstanding of the Supreme Court's holding in Ristaino v. Ross, a case which allowed the questioning of potential jurors about possible racial biases when the case has certain types of racial overtones. (143) But that case did not depend on the race of the jurors; it merely created a mechanism for attorneys to uncover racial prejudice not linked to the race of those who might possess that prejudice in cases where racial tensions run high. Suggesting that the race of the jurors themselves--rather than their attitudes--matters more in such a case is in tension with our dearly held concepts of racial equality.
In Hollis v. Davis, the Eleventh Circuit made the explicit assumption that jurors' race affects outcomes:
In Strickland terms, if we compared the result reached by an all white jury, selected by systematic exclusion of blacks, with the result which would have been reached by a racially mixed jury, we would have greater confidence in the latter outcome, finding much less probability that racial bias had affected it. (144)
Can these opinions really mean what they say--that the accuracy of a verdict depends upon the race of the jurors? And if they do mean this, why is this only the case when looking at Strickland prejudice, and not when a properly preserved Batson claim is presented on direct appeal?
These opinions are problematic for two reasons. First, they showcase courts' uncomfortable reasoning about race, pushing them straight into the paradox at the heart of Batson and disregarding the reasons, unclear though they may be, that courts generally do not apply harmless error review to Batson claims on direct appeal. Second, and more disturbingly, they may require petitioners to make these very arguments if they are to rescue their Batson claims from procedural default. If a petitioner is to attempt to demonstrate the prejudice these courts are requiring, he, too, will have to argue that the race of jurors matters.
C. Possible Solutions
Courts clearly need a new framework for analyzing the reviewability of procedurally defaulted structural errors, especially when ineffectiveness is used as cause. I see two possible solutions to this conflict. One would be to use a modified version of Strickland when it is used as cause (essentially overruling Murray v. Carrier, (145) the case that applied the newly minted Strickland standard to the cause inquiry). The other would be to change Strickland to accommodate structural errors, and continue to use Strickland uniformly, whether as a substantive claim or as cause.
First, however, we need a clear definition of Sykes prejudice. Indeed, the vague prejudice standard from Sykes is ripe for a clear and definitive definition anyway. Courts should clearly define Sykes prejudice as presumed in cases of structural error. In cases where cause is something other than ineffective assistance, such as government interference or a newly available claim, this would end the inquiry; no claim would be unavailable simply because it is structural.
Cases where ineffectiveness constitutes cause, however, require a bit more consideration. They could be resolved in one of two ways. First, we could use the Sykes prejudice inquiry as defined above to cover both the Sykes test itself and the prejudice arm of the ineffectiveness inquiry. As I have noted, many courts are already treating the two inquiries as coextensive, with both covered by the Strickland inquiry, (146) without acknowledging the problems created by (147) If, rather than shifting the Strickland rule into the procedural default context. importing Strickland wholesale into the cause inquiry, we use a modified version, it might be appropriate to continue collapsing the two prejudice inquiries. Another way to see this solution is that rather than allowing the Strickland prejudice test to stand in for the Sykes test, we would do the reverse, and let Sykes prejudice (with its structural error exception) stand in for the prejudice arm of the Strickland test when it arises as cause. If we do this, the dilemma disappears, and Strickland can remain on the books as we have always understood it, as a substantive claim.
The other solution would be to modify the Strickland rule--that is, to use the same rule for ineffectiveness in every context, but have the rule accommodate structural error by presuming prejudice in any case where a demonstration is theoretically impossible. This change would also allow an additional set of substantive Strickland claims to succeed, but might generate more resistance, as it would entail making a substantive change to a longstanding interpretation of the Sixth Amendment right to effective assistance.
Whether or not we collapse the two prejudice inquiries into one, we cannot use the current Strickland prejudice test for cases of structural error when that test is blocking access to a review that courts have agreed petitioners should have. Either of these two solutions would remedy the problem.
This Note has explored an understudied pocket of complexity surrounding the meaning of prejudice requirements and the reasons for dispensing with them. Structural errors merit reversal even without proof of actual prejudice. Procedurally defaulted claims cannot generally be raised on habeas without a showing of actual prejudice. These doctrines squarely conflict.
I have pointed out the logical error courts make in collapsing the Sykes and Strickland prejudice inquiries, and have drawn attention to the impossible situation in which courts have placed some petitioners. Courts can clean up this mess by clarifying the Sykes prejudice standard, and by avoiding the use of a prejudice standard in the cause inquiry that does not accommodate structural errors--either by using a modified version of Strickland when it comes up as cause or by making an across-the-board change to the way the Strickland prejudice test applies.
But as it stands, a defendant can spend his life in prison, or even be put to death, despite an acknowledged violation of his rights, without any articulated reason why this system is necessary. Indeed, Michael Edward Vansickel is in prison, despite the Ninth Circuit's realization that he had been given an impossible task. This Note has attempted to shed some light on the inconsistencies in the procedural rules courts use for habeas claims, showing how these inconsistencies have produced unjust outcomes. The analysis presented here has drawn attention to an internal inconsistency in the procedural rules for habeas and proposed a solution that will allow courts to reach a just outcome without sacrificing the important interests advanced by existing rules.
(1.) See Vansickel v. White, 166 F.3d 953, 954 (9th Cir. 1999).
(2.) California defendants facing the death penalty or life in prison get twenty peremptory challenges, which is double the usual number. See CAL. CIV. PROC. CODE [section] 231(a) (West 2011) ("In criminal cases, if the offense charged is punishable with death, or with imprisonment in the state prison for life, the defendant is entitled to 20 and the people to 20 peremptory challenges. Except as provided in subdivision (b), in a trial for any other offense, the defendant is entitled to 10 and the state to 10 peremptory challenges.").
(3.) United States v. Annigoni, 96 F.3d 1132, 1134 (9th Cir. 1996) (en banc).
(4.) Inmate Locator, CAL. DEPT CORRECTIONS & REHABILITATION, http:// inmatelocator.cdcr.ca.gov (search the "Inmate Number" field for "H42760") (last visited Mar. 9, 2012). Note that the California Department of Corrections and Rehabilitation's listing for Vansickel spells his name "Vansickle," while the court opinion spells it "Vansickel."
(5.) 433 U.S. 72 (1977).
(6.) See WAYNE R. LAFAVE ET AL., CRIMINAL PROCEDURE [section] 28.4(a) (4th ed. 2004) (recognizing the ambiguity, citing cases, and expressing a preference for a presumption of prejudice); BRIAN R. MEANS, FEDERAL HABEAS MANUAL [section] 9B:80 (2008) (noting that different courts have resolved the Sykes prejudice issue differently).
(7.) See Chapman v. California, 386 U.S. 18, 22 (1967); Tom Stacy & Kim Dayton, Rethinking Harmless Constitutional Error, 88 COLUM. L. REV. 79, 82-83 (1988).
(8.) Chapman, 386 U.S. at 22.
(9.) Jason S. Marks, Harmless? Constitutional Error: Fundamental Fairness and Constitutional Integrity, CRIM. JUST., Spring 1993, at 2, 2.
(10.) Arizona v. Fulminante, 499 U.S. 279, 310 (1991).
(11.) For example, constitutional errors on direct appeal must be harmless beyond a reasonable doubt, see Chapman, 386 U.S. at 24, and the same errors on habeas must be proven to have had "substantial and injurious effect," Brecht v. Abrahamson, 507 U.S. 619, 623 (1993) (quoting Kotteakos v. United States, 328 U.S. 750, 776 (1946)). The potential influence of these differences is discussed below in Part III.C.
(12.) See Stacy & Dayton, supra note 7, at 82-83 & n.16.
(13.) FED. R. CRIM. P. 52(a) (emphasis added).
(14.) "On the hearing of any appeal or writ of certiorari in any case, the court shall give judgment after an examination of the record without regard to errors or defects which do not affect the substantial rights of the parties." 28 U.S.C. [section] 2111 (2006).
(15.) 386 U.S. at 24.
(16.) Id. at 23.
(17.) See Rose v. Clark, 478 U.S. 570, 577-78 (1986); see also Stacy & Dayton, supra note 7, at 84.
(18.) 499 U.S. 279, 280 (1991).
(19.) See id. at 307-10.
(20.) A 1999 Supreme Court case lists the following as clearly structural errors: complete denial of counsel, a biased trial judge, racial discrimination in the selection of the grand jury, denial of self-representation at trial, denial of public trial, and a defective reasonable doubt instruction. Neder v. United States, 527 U.S. 1, 8 (1999).
(21.) See David McCord, The "Trial"/"Structural" Error Dichotomy. Erroneous, and Not Harmless, 45 U. KAN. L. REV. 1401 (1997).
(22.) 499 U.S. at 309.
(23.) Id. at 310; see, e.g., United States v. Soto, 519 F.3d 927, 930 (9th Cir. 2008) ("We hold that failure to give a Carter instruction is not a structural error, because it does not 'affect the framework within which the trial proceeds.'" (quoting United States v. Gonzalez-Lopez, 548 U.S. 140, 148 (2006))).
(24.) See, e.g., United States v. Navarro, 608 F.3d 529, 538 (9th Cir. 2010) ("'[S]tructural error' is a term of art for error requiring reversal regardless of whether it is prejudicial or harmless...."); United States v. Brandao, 539 F.3d 44, 58 (1st Cir. 2008) (defining structural errors as "constitutional errors that deprive the defendant of a fundamentally fair trial and thus may not be found harmless under Rule 52(a)'s harmless error standard").
(25.) See McCord, supra note 21, at 1412-17.
(26.) See, e.g., Hayes v. Missouri, 120 U.S. 68, 70 (1887) ("Experience has shown that one of the most effective means to free the jury-box from men unfit to be there is the exercise of the peremptory challenge. The public prosecutor may have the strongest reasons to distrust the character of a juror offered, from his habits and associations, and yet find it difficult to formulate and sustain a legal objection to him. In such cases, the peremptory challenge is a protection against his being accepted.").
(27.) 96 F.3d 1132, 1145 (9th Cir. 1996) (en banc); see also United States v. McFerron, 163 F.3d 952, 956 (6th Cir. 1998) ("Moreover, on a practical note, it would be virtually impossible to determine whether the erroneous denial of a peremptory challenge was harmless enough to warrant affirming a conviction.").
(28.) See Rose v. Mitchell, 443 U.S. 545, 551 (1979) (citing a long line of cases requiring automatic reversal for grand jury discrimination); Tankleff v. Senkowski, 135 F.3d 235, 248 (2d Cir. 1998) (holding that Batson claims--complaints that a lawyer exercised peremptory challenges on the basis of race--are not amenable to harmless error review, without explicating why); Ford v. Norris, 67 F.3d 162, 170 (8th Cir. 1995) (coming to the same conclusion under Batson's predecessor, Swain v. Alabama, 380 U.S. 202 (1965)).
(29.) See Arizona v. Fulminante, 499 U.S. 279, 310 (1991); Chapman v. California, 386 U.S. 18, 23 n.8 (1967).
(30.) 273 U.S. 510 (1927).
(31.) See, e.g., Cartalino v. Washington, 122 F.3d 8, 9 (7th Cir. 1997) (citing, inter alia, Tumey, 273 U.S. at 535).
(32.) United States v. Gonzalez-Lopez, 548 U.S. 140, 150 (2006) (emphasis added).
(33.) Holloway v. Arkansas, 435 U.S. 475, 491 (1978).
(34) it has made clear that an error requiring too much speculation must be presumed prejudicial.
(34.) Cf. infra Part I.B.4.
(35.) See Faretta v. California, 422 U.S. 806 (1975).
(36.) McKaskle v. Wiggins, 465 U.S. 168, 177 n.8 (1984).
(37.) See Arizona v. Fulminante, 499 U.S. 279, 310 (1991).
(38.) See Waller v. Georgia, 467 U.S. 39, 46 (1984) ("[T]he presence of interested spectators may keep ... triers keenly alive to a sense of their responsibility and to the importance of their functions...." (second omission in original) (quoting Gannett Co. v. DePasquale, 443 U.S. 368, 380 (1979))).
(39.) Id. at 49 n.9 (quoting People v. Jones, 391 N.E.2d 1335, 1340 (N.Y. 1979)).
(40.) See Powers v. Ohio, 499 U.S. 400, 415 (1991) (holding that defendants have third-party standing in the Batson context to bring excluded jurors' equal protection claims); see also Rose v. Mitchell, 443 U.S. 545, 556 (1979) ("The harm is not only to the accused, indicted as he is by a jury from which a segment of the community has been excluded. It is to society as a whole.").
(41.) 466 U.S. 648, 658 (1984); see also id. at 658 n.24 (citing cases where harmless error analysis was not necessary due to the high probability of prejudice); Estes v. Texas, 381 U.S. 532, 542-43 (1965) ("[A]t times a procedure employed by the State involves such a probability that prejudice will result that it is deemed inherently lacking in due process.").
(42.) Cronic, 466 U.S. at 649.
(43.) Id. at 659-60.
(44.) Id. at 656.
(45.) 425 U.S. 501 (1976).
(46.) Id. at 505.
(47.) Chapman v. California, 386 U.S. 18, 23 (1967) ("[T]here are some constitutional rights so basic to a fair trial that their infraction can never be treated as harmless error.").
(48.) Neder v. United States, 527 U.S. 1, 7 (1999).
(49.) 476 U.S. 79, 84 (1986).
(50.) See Miller-El v. Dretke, 545 U.S. 231 (2005).
(51.) See Eric L. Muller, Solving the Batson Paradox: Harmless Error, Jury Representation, and the Sixth Amendment, 106 YALE L.J. 93, 116 n.148 (1996) (collecting cases).
(52.) See id.
(53.) See id.
(54.) 941 F.2d 1471, 1483 (11th Cir. 1991).
(55.) 396 F.3d 1059, 1069-70 (9th Cir. 2005) (Rawlinson, J., dissenting from denial of rehearing en banc) ("[T]here is a reasonable probability that, but for the alleged Batson violation and the related ineffective assistance of counsel, there would have been a more racially diverse jury, [so] Williams was actually prejudiced by the seating of an all-white jury instead of a more racially diverse one.").
(56.) Cf EQUAL JUSTICE INITIATIVE, ILLEGAL RACIAL DISCRIMINATION IN JURY SELECTION: A CONTINUING LEGACY 14 (2010), available at http://eji.org/eji/files/EJI %20Race%20and%20Jury%20Report.pdf (noting that peremptory strikes are often used to exclude black prospective jurors, especially in capital cases).
(57.) In Hollis, the court was unable to perform harmless error analysis due to the absence of a transcript; they did express an intention, otherwise, to weigh the evidence and consider the probability of an acquittal by an unbiased jury. 941 F.2d at 1483. We do not know, however, whether the court, transcript in hand, would actually have been able to consider the effect of the jury composition, since the court would not have a transcript of what happened in the jury room but only of the evidence presented at trial; it thus could not assess the role of bias.
(58.) See Muller, supra note 51, at 96.
(59.) See id.
(60.) Occasionally, courts do attempt to conduct this analysis, and the results are unsettling. See infra Part IV.B.3.
(61.) See Powers v. Ohio, 499 U.S. 400, 415 (1991) (holding that a defendant has third-party standing to challenge the equal protection violations of the jurors, meaning that a defendant can also challenge the exclusion of jurors of his own race).
(62.) See Rose v. Mitchell, 443 U.S. 545, 558 (1979).
(63.) Id. at 557-58.
(64.) Id. at 556.
(65.) Powers, 499 U.S. at 412.
(66.) Id. at 411 ("The discriminatory use of peremptory challenges by the prosecution causes a criminal defendant cognizable injury, and the defendant has a concrete interest in challenging the practice.").
(67.) The Faretta right is named for Faretta v. California, 422 U.S. 806 (1975), which was the landmark case establishing the right to represent oneself in criminal proceedings.
(68.) Mitchell, 443 U.S. at 558 (internal quotation marks omitted).
(69.) 508 U.S. 275, 280 (1993).
(70.) See United States v. Martin Linen Supply Co., 430 U.S. 564, 572-73 (1977) ("[A] trial judge is prohibited from entering a judgment of conviction or directing the jury to come forward with such a verdict, regardless of how overwhelmingly the evidence may point in that direction." (citations omitted)).
(71.) At least one circuit--the Eighth--has treated Batson claims differently from traditional claims of structural error in the context of ineffective counsel claims under Strickland v. Washington, 466 U.S. 668 (1984). Compare McGurk v. Stenberg, 163 F.3d 470, 474-75 (8th Cir. 1998) (treating counsel's failure to inform defendant of his right to a jury trial as a structural error and presuming prejudice under Strickland), with Young v. Bowersox, 161 F.3d 1159, 1160-61 (8th Cir. 1998) (requiring a showing of prejudice for a Batson error as an incidence of ineffective assistance). The Bowersox court did not acknowledge the fact that prejudice is generally presumed for Batson errors, and parroted the language in Strickland, holding that "Young has not shown a reasonable probability that the results of the proceeding would have been different." Bowersox, 161 F.3d at 1161. We thus have little clue as to the Eighth Circuit's reasoning. For further discussion of this case, see Part IV.B.3, below.
(72.) The Batson analysis is a three-step process. First, a defendant must make a prima facie showing that there was discrimination in the jury selection process. Second, the prosecutor is permitted to offer his race-neutral reasons for striking the challenged jurors. The trial court then determines whether purposeful discrimination has been established. See Batson v. Kentucky, 476 U.S. 79, 96-98 (1986).
(73.) For a more detailed description of the concept of cause and prejudice, see Part II below.
(74.) Federal courts can, of course, also conduct collateral review of federal convictions. See 28 U.S.C. [section] 2255 (2006). These federal habeas claims present similar issues, and indeed, several of the cases discussed in Part IV arose in the [section] 2255 context. However, I generally focus on review of state convictions, as they present more challenges for the question of when review is appropriate, given concerns of comity and federalism that are absent in the [section] 2255 context. In other words, arguments in favor of review of state convictions will all apply in the federal context, but the reverse is not necessarily true.
(75.) See 28 U.S.C. [section] 2254.
(76.) See Coleman v. Thompson, 501 U.S. 722, 729-33 (1991) (citing Michigan v. Long, 463 U.S. 1032 (1983)).
(77.) See Fay v. Noia, 372 U.S. 391, 434, 439 (1963) (holding that federal habeas courts may hear claims that would otherwise be barred by state procedural rules unless the petitioner deliberately bypassed the state court).
(78.) See Wainwright v. Sykes, 433 U.S. 72, 85-87 (1977).
(79.) Coleman, 501 U.S. at 730-31.
(80.) See Sykes, 433 U.S. at 87.
(81.) Murray v. Carrier, 477 U.S. 478, 488 (1986).
(82.) See id.
(83.) See 2 RANDY HERTZ & JAMES S. LIEBMAN, FEDERAL HABEAS CORPUS PRACTICE AND PROCEDURE [section] 26.3[C] (6th ed. 2011) (citing a variety of open-ended formulations).
(84.) Sykes, 433 U.S. at 90-91.
(85.) 466 U.S. 668, 687, 694 (1984).
(86.) See Carrier, 477 U.S. at 488 ("So long as a defendant is represented by counsel whose performance is not constitutionally ineffective under the standard established in Strickland v. Washington, we discern no inequity in requiring him to bear the risk of attorney error that results in a procedural default." (citation omitted)).
(87.) See Johnson v. Sherry, 586 F.3d 439, 447 (6th Cir. 2009) ("There would be no prejudice--either with respect to Johnson's ineffective assistance of counsel claim or with respect to the showing of prejudice necessary to excuse procedural default."); Owens v. United States, 483 F.3d 48, 64 n.13 (1st Cir. 2007) ("We believe that these showings of prejudice overlap, and we resolve them simultaneously."); Quintero v. Bell, 256 F.3d 409, 413 n.3 (6th Cir. 2001) ("Since the two prejudice inquiries are so closely linked in this case, we discuss them together."); Hollis v. Davis, 941 F.2d 1471, 1478 n.7 (11th Cir. 1991) ("Of course, inadequate representation is only one of two components of constitutionally ineffective counsel; the inept representation also must have prejudiced the accused's defense. We defer discussing the prejudice component until we reach it in dealing with avoidance of the procedural bar." (citation omitted)). But see Maupin v. Smith, 785 F.2d 135, 139 (6th Cir. 1986) (holding that prejudice from the default and prejudice from the underlying violation are distinct). Other cases consider the two together without explicitly acknowledging that there are two standards. See, e.g., Vansickel v. White, 166 F.3d 953, 958-59 (9th Cir. 1999) (considering both under the heading of "prejudice" without specifying which prejudice standard is involved).
(88.) See Carrier, 477 U.S. at 488.
(89.) Strickland, 466 U.S. at 694.
(90.) Wainwright v. Sykes, 433 U.S. 72, 87 (1977).
(91.) See Carrier, 477 U.S. at 488.
(92.) Although the Supreme Court has recognized that a different--more state-friendly--harmless error standard applies on habeas rather than on direct appeal, see, it has also continued to acknowledge that some errors were exempt from this higher standard, and could demand reversal on habeas without a showing of prejudice. See id. at 629-30 ("At the other end of the spectrum of constitutional errors lie 'structural defects in the constitution of the trial mechanism, which defy analysis by harmless-error standards.' The existence of such defects--deprivation of the right to counsel, for example--requires automatic reversal of the conviction because they infect the entire trial process." (footnote and citations omitted) (internal quotation marks omitted) (quoting Arizona v. Fulminante, 499 U.S. 279, 309 (1991))).
(93.) Errors accepted as structural include complete denial of counsel, biased trial judge, racial discrimination in the selection of a grand jury, denial of self-representation at trial, denial of a public trial, and a defective reasonable doubt instruction. See Neder v. United States, 527 U.S. 1, 8 (1999).
(94.) Engle v. Isaac, 456 U.S. 107, 135 (1982).
(95.) The Court has denied certiorari in a number of cases raising this issue. See, e.g., Purvis v. McDonough, 549 U.S. 1035 (2006); Williams v. Brown, 546 U.S. 934 (2005); Ward v. Hinsley, 543 U.S. 1011 (2004); Bell v. Beck, 534 U.S. 830 (2001); White v. Vansickel, 528 U.S. 965 (1999); Young v. Bowersox, 528 U.S. 880 (1999); Davis v. Hollis, 503 U.S. 938 (1992).
(96.) See Strickland v. Washington, 466 U.S. 668, 687 (1984).
(97.) See United States v. Cronic, 466 U.S. 648, 659 (1984).
(98.) See Premo v. Moore, 131 S. Ct. 733, 743 (2011) ("The Fulminante prejudice inquiry presumes a constitutional violation, whereas Strickland seeks to define one."). The Premo Court rejected the Ninth Circuit's application of Fulminante's presumption of prejudice to a substantive Strickland claim where Sykes was not at issue. Thus, to the extent that its holding speaks to the prejudice requirement in the Strickland context, it is limited to the substantive Strickland inquiry, and says nothing about the cause and prejudice test. See infra text accompanying note 118.
(99.) See Tumey v. Ohio, 273 U.S. 510, 523 (1927).
(100.) See, e.g., United States v. Withers, 638 F.3d 1055, 1066-68 (9th Cir. 2011) (substantive ineffective assistance of counsel for failure to raise a public trial claim); Virgil v. Dretke, 446 F.3d 598, 608 (5th Cir. 2006) (substantive ineffective assistance of counsel claim for failure to use challenges for cause); Young v. Bowersox, 161 F.3d 1159, 1160 (8th Cir. 1998) (substantive ineffective assistance of counsel claim for failure to make a Batson objection).
(101.) 507 U.S. 619, 627 (1993).
(102.) Id. at 634-37.
(103.) A case raising Sykes prejudice without Strickland is currently before the Sixth Circuit. See Ambrose v. Booker, No. 11-1430 (6th Cir. argued Feb. 28, 2012). In Ambrose, a computer glitch allegedly gave rise to a fair cross-section violation; while the hidden nature of the glitch constitutes cause, the Sixth Circuit is considering the State's argument that the petitioner must show actual prejudice from the violation. See Brief for Appellee at 30-31, Ambrose, No. 11-1430.
(104.) 411 U.S. 233 (1973).
(105.) Id. at 245.
(106.) 425 U.S. 536, 542 (1976).
(107.) See Purvis v. Crosby, 451 F.3d 734, 742 (11th Cir. 2006); Hollis v. Davis, 941 F.2d 1471, 1479-80 (11th Cir. 1991).
(108.) Owens v. United States, 483 F.3d 48, 64 n.14 (1st Cir. 2007).
(109.) 408 U.S. 238 (1972).
(110.) 428 U.S. 153 (1976).
(111.) 466 U.S. 648 (1984).
(112.) 535 U.S. 685 (2002).
(113.) Quintero v. Bell, 256 F.3d 409, 412-15 (6th Cir. 2001).
(114.) Bell v. Quintero, 535 U.S. 1109 (2002).
(115.) Quintero v. Bell, 368 F.3d 892, 893 (6th Cir. 2004).
(116.) Bell v. Quintero, 544 U.S. 936 (2005).
(117.) See Quintero v. Bell, 256 F.3d at 413 n.3 (6th Cir. 200l) ("Since the two prejudice inquiries are so closely linked in this case, we discuss them together.").
(118.) 131 S. Ct. 733 (2011).
(119.) See supra note 86 and accompanying text.
(120.) Premo, 131 S. Ct. at 744.
(122.) See United States v. Gonzales-Lopez, 548 U.S. 140, 150 (2006).
(123.) 474 U.S. 52, 59 (1985).
(124.) See Johnson v. Sherry, 586 F.3d 439, 447 (6th Cir. 2009) ("Because the right to a public trial is a structural guarantee, if the closure were unjustified or broader than necessary, prejudice would be presumed."); Owens v. United States, 483 F.3d 48, 64 (1st Cir. 2007) ("If the failure to hold a public trial is structural error, and it is impossible to determine whether a structural error is prejudicial, we must then conclude that a defendant who is seeking to excuse a procedurally defaulted claim of structural error need not establish actual prejudice." (citations omitted)); Miller v. Dormire, 310 F.3d 600, 603 (8th Cir. 2002) (citing McGurk as requiting a presumption of Strickland prejudice where attorney's deficient performance causes a structural error); McGurk v. Stenberg, 163 F.3d 470, 475 (8th Cir. 1998) ("[W]e hold that when counsel's deficient performance causes a structural error, we will presume prejudice under Strickland."); see also Bell v. Jarvis, 236 F.3d 149, 165 (4th Cir. 2000) (noting in dicta that Strickland prejudice may be presumed for structural errors). The Eighth Circuit, however, has not been entirely uniform in its views and has failed to apply its holding from McGurk in the context of Batson claims. See Young v. Bowersox, 161 F.3d 1159, 1160-61 (8th Cir. 1998) (requiring a showing of Strickland prejudice for failure to raise a Batson claim).
(125.) See Vansickel v. White, 166 F.3d 953, 960 (9th Cir. 1999) (Reinhardt, J., dissenting) ("Without explaining how Vansickel or any other litigant could possibly make such a showing, the majority simply overrides our well established rule that prejudice as to the result need not, indeed cannot, be shown in jury composition cases. By doing so, it renders it virtually impossible for any defendant to vindicate his tight to due process if his attorney has committed a procedural default in such a case."); see also Williams v. Woodford, 396 F.3d 1059, 1069 (9th Cir. 2005) (Rawlinson, J., dissenting from denial of reheating en banc) (arguing that Sykes prejudice should be presumed for a Batson claim because it is structural, and also arguing that the petitioner can show actual prejudice from being tried by an all-white jury).
(126.) See Purvis v. Crosby, 451 F.3d 734, 742 (11th Cir. 2006) ("IT]he law of this circuit [is] that an ineffective assistance of counsel claim based on the failure to object to a structural error at trial requires proof of prejudice."); Virgil v. Dretke, 446 F.3d 598, 607 (5th Cir. 2006) ("[W]e do not hold that a structural error alone is sufficient to warrant a presumption of prejudice in the ineffective assistance of counsel context...."); Jackson v. Herring, 42 F.3d 1350, 1361 (11th Cir. 1995) (requiring a showing of actual prejudice where counsel failed to raise a Swain claim); Hollis v. Davis, 941 F.2d 1471, 1479 (11th Cir. 1991) (holding that a requirement of actual prejudice is compelled by Francis v. Henderson).
(127.) See Johnson, 586 F.3d at 449 (Kethledge, J., dissenting). This case is currently back before the Sixth Circuit. Johnson v. Sherry, No. 10-2699 (6th Cir. argued Oct. 27, 2011).
(128.) See Bowersox, 161 F.3d at 1160-61.
(129.) See United States v. Withers, 638 F.3d 1055, 1067 (9th Cir. 2011) ("The Ninth Circuit has not yet decided whether a trial counsel's failure to object to a structural error is presumptively prejudicial for purposes of the Strickland ineffective assistance of counsel inquiry."). Withers involved both a Strickland claim and a substantive public trial claim. The quote above is in reference to the Strickland claim. Because the case involved the review of a federal conviction and arose in a particular procedural posture, the required showing for the substantive public trial claim was only that Withers had a "non-frivolous" argument that he could show cause and prejudice. Thus, while the court did indicate that the structural nature of the error might establish prejudice, it did not issue a holding on the question. Id. at 1064.
(130.) See Vansickel, 166 F.3d at 958.
(131.) Withers, 638 F.3d at 1067.
(132.) See Jackson v. Herring, 42 F.3d 1350, 1361 (11th Cir. 1995) ("In making this inquiry, we bear in mind that the prejudice prong of Strickland is not co-terminous with the more general prejudice requirement of Wainwright v. Sykes.... Neither is it akin to the 'harmless error' standard of Brecht v. Abrahamson...."). The case never reaches the question of Sykes prejudice, finding that it need not address whether an ineffectiveness claim that is itself defaulted can satisfy the cause requirement--a debate beyond the scope of this Note. Id. at 1362.
(133.) See Bell v. Quintero, 535 U.S. 1109 (2002) (remanding for reconsideration in light of Bell v. Cone, 535 U.S. 685 (2002), which examined Cronic).
(134.) The three cases explicitly raising this argument are Bell v. Quintero, 544 U.S. 936, 937-38 (2005) (Thomas, J., dissenting) (explaining the Court's remand in its earlier Quintero decision); Purvis v. Crosby, 451 F.3d 734, 738 (11th Cir. 2006); and Virgil v. Dretke, 446 F.3d 598, 607 (5th Cir. 2006). The other three cases requiring a showing of actual prejudice are Bowersox, 161 F.3d at 1160-61; Jackson, 42 F.3d at 1361; and Hollis v. Davis, 941 F.2d 1471, 1479 (11th Cir. 1991).
(135.) See McGurk v. Stenberg, 163 F.3d 470, 473 (8th Cir. 1998).
(136.) See supra note 87.
(137.) 161 F.3d at 1160-61 ("Otherwise, Young argues, he is forced into the impossible position of showing how the outcome of the trial would have been different in the absence of a structural defect. We cannot accept this position."). The case the court cites for this proposition is Wright v. Nix, 928 F.2d 270, 273 (8th Cir. 1991), which never reached the question of prejudice and opined on the matter only in dicta.
(138.) See, e.g., Purvis, 451 F.3d at 738 (rejecting a Stricklandclaim for failure to object to a public trial violation because defendant could not "show that an objection from his counsel would have caused the factfinder to have a reasonable doubt about his guilt").
(139.) Vansickel v. White, 166 F.3d 953, 959 (9th Cir. 1999).
(141.) See discussion supra Part I.C.2.
(142.) 42 F.3d 1350, 1362 (11th Cir. 1995).
(143.) See Ristaino v. Ross, 424 U.S. 589, 595-98 (1976).
(144.) 941 F.2d 1471, 1482 (11th Cir. 1991).
(145.) 477 U.S. 478, 488 (1986) (holding that counsel must be constitutionally ineffective to qualify as cause).
(146.) See supra note 87 and accompanying text.
(147.) See supra Part III.B.
Amy Knight Burns, J.D. Candidate, Stanford Law School, 2012. Thanks to Bob Weisberg for his extensive guidance on this project. Thanks also to Betsy Candler, Bethany O'Neill, Brad Hall, Nancy King, Anya Goldstein, and Michael Miller for encouragement and discussion, and to the Honorable Marsha Berzon of the United States Court of Appeals for the Ninth Circuit for raising interesting questions that helped me to strengthen my analysis.
TABLE 1 Cases Discussing Presumed Prejudice for Procedural Default in Addition to Prejudice for Ineffective Assistance. Presuming Requiring Showing Quintero v. Bell Hollis v. Davis (11th Cir. 1991) (6th Cir. 2001, 2004) Owens v. United States Jackson v Herring (11th Cir. 1995) (1st Cir. 2007) Johnson v. Sherry Vansickel v. White (9th Cir. 1999) (6th Cir. 2009)