The case against automatic reversal of structural errors.

• Author: Shepard, Steven M.

NOTE CONTENTS INTRODUCTION I. THE DEFENDERS OF AUTOMATIC REVERSAL II. THE DOCTRINAL CONSEQUENCES OF AUTOMATIC REVERSAL A. A Three-Generational Theory of Appellate Review of Error B. The Diminished Meaning of "Bias" C. The Diminished Right to Conflict-Free Counsel D. The Confused Meaning of "Reasonable Doubt" E. Responses to Objections III. A NEW DOCTRINAL FRAMEWORK FOR CLASSIFYING ERROR A. Arizona v. Fulminante: The Current Framework for Applying Automatic Reversal B. A New Framework for Applying Automatic Reversal: Chapman v. California Revived CONCLUSION INTRODUCTION

The rule of automatic reversal requires appellate courts to reverse all criminal convictions tainted by certain errors. The Supreme Court calls these errors "structural errors" and distinguishes them from "trial errors" that do not require an automatic reversal. (1) When an appellate court confronts a new type of error, the court must decide whether to treat it as a structural error (which must be reversed automatically) or as a trial error (which does not always require reversal). This Note describes an important and previously undocumented doctrinal consequence of labeling a type of error as "structural": over time, the rule of automatic reversal narrows the definitions of structural errors and thereby weakens the procedural safeguards that protect defendants. To avoid that consequence, this Note proposes a new framework for deciding whether a type of error is "structural." The proposed test is straightforward: a type of error should only be labeled as "structural" (and therefore reversed automatically) if it never contributes to a verdict.

To illustrate, consider the Supreme Court's 2006 decision in United States v. Gonzalez-Lopez. (2) There, the defendant hired an out-of-state lawyer to represent him; that lawyer filed a motion for admission pro hac vice so that he could participate in the trial. (3) The district court erroneously denied that motion, and as a result the defendant went to trial without his preferred lawyer. He lost. On appeal, the government conceded that the district court was wrong to deny the pro hac vice motion, but argued that the error was harmless. To decide the case, the Court had to determine whether this type of error--depriving the defendant of his counsel of choice--is a "structural error" requiring automatic reversal, or a "trial error" that does not always require reversal. (4) To make this determination, the Court applied the test it set out fifteen years earlier in Arizona v. Fulminante. (5) Under that test, the error is "structural" because it "bears directly on the framework within which the trial proceeds"; as a structural error, it must be reversed automatically. (6)

Under this Note's proposed framework, by contrast, the Court would have asked instead whether this type of error--the erroneous deprivation of defendant's counsel of choice--never contributes to a verdict. The answer to that question is "no"--this type of error contributes to the verdict in almost every case. Therefore, under this Note's proposed test, the Court would have held that this type of error is a "trial error" and would then have asked whether the error was harmless in this particular case.

To determine whether the error was harmless, the Court would have applied the test for harmlessness articulated in Chapman v. California, according to which a constitutional error is harmless if on appeal the prosecution proves "beyond a reasonable doubt that the error complained of did not contribute to the verdict obtained." (7) Had the Court applied the Chapman test, it would have found abundant doubt regarding whether the particular error in Gonzalez-Lopez contributed to the verdict and likely would have concluded that the error therefore was not harmless. (8) Because the error was not harmless, the Court would have reversed and remanded for a new trial. In short, regardless of whether the Court labeled this type of error as a "structural error" or a "trial error," the outcome would have been the same: reversed and remanded for a new trial.

But the outcomes of future cases will be affected by the Court's decision in Gonzalez-Lopez to treat this type of error as structural. As Part II demonstrates, over the long run the rule of automatic reversal narrows the definitions of structural errors and thereby weakens constitutional safeguards that protect defendants. Part II examines three procedural rights: the right to an unbiased judge; the right to a defense counsel free from conflicts of interest; and the right to an adequate jury instruction on the meaning of reasonable doubt. The doctrinal fates of these rights suggest that courts have weakened all three to avoid applying the drastic remedy of automatic reversal. It is therefore likely that in the long run, Gonzalez-Lopez will generate case law that approves trial courts' denials of defense lawyers' pro hac vice motions; that case law, in turn, will make it more difficult for defendants to secure the counsel of their choice.

Because the rule of automatic reversal tends to narrow the definitions of the rights that the rule protects, this Note proposes a new, more limited test for determining whether a given type of error is structural. Under this test, only types of error that never contribute to the verdict would be labeled "structural" and automatically reversed. Part III describes and critiques the current Fulminante test for determining when an error is "structural" and describes the proposed alternative. First, however, Part I discusses the scholarship on automatic reversal.

1. THE DEFENDERS OF AUTOMATIC REVERSAL

   With few exceptions, legal scholarship on automatic reversal urges the courts to extend the rule of automatic reversal to more errors of criminal procedure. This Part describes that literature and shows that legal scholars have so far failed to appreciate the problem described below in Part II, namely, that the rule causes subsequent courts to narrow the definition of the right protected by the rule. To be sure, existing scholarship is correct that automatic reversal should apply to some types of error--namely, to those types of error that never contribute to a verdict. As Part III discusses in detail, automatic reversal is the only workable rule for those types of error. But in addition to those types of error, scholars have urged courts to extend automatic reversal to many other types of error that do contribute to a verdict. In so doing, scholars may have indirectly contributed to an erosion of the procedures that protect defendants.

   Some scholars argue that every constitutional error, simply because it violates the Constitution, must be reversed automatically; anything less than automatic reversal, these scholars suggest, is not enough to right a constitutional wrong. (9) This argument now appears somewhat dated; the Supreme Court has made dear that many constitutional errors can indeed be harmless. (10)

   Professors Tom Stacy and Kim Dayton argue that the rule of automatic reversal deters "law enforcement officials" from committing procedural errors. (11) But Stacy and Dayton offer no reason to think that a reversal impresses law enforcement officials more than does, say, a sternly worded opinion that both condemns the error and affirms the verdict. Nor do Stacy and Dayton consider whether the rule of automatic reversal may lead appellate courts to define the scope of error more narrowly, and thereby ratify conduct that would otherwise have been condemned and forbidden.

   Scholars also argue that automatic reversal is a cost-efficient way of maximizing the number of correct appellate decisions. If certain types of error are almost always prejudicial to the defendant, the argument goes, those types of error ought to be reversed automatically to save appellate courts the effort of deciding whether a particular error was harmless in a particular case. Professor Philip Manse made this argument two years after the Court decided Chapman; (12) Professor William Landes and Judge Richard Posner have since restated it with more theoretical rigor. (13) The problem with Landes and Posner's defense of automatic reversal is that they assume that each type of error is clearly defined. In other words, they assume that appellate courts can effortlessly determine whether, for example, a trial court's denial of a defense lawyer's pro hac vice motion was correct or erroneous. But that assumption is false: types of error are not so clearly defined--in many cases, it is debatable whether an error occurred at all. The initial decisions that establish an error rarely define the error in terms that are specific enough to decide easily whether later trials are erroneous or not. Because errors are vaguely defined, appellate courts have great latitude to find that, in a later trial, no error occurred at all. Part II, below, describes three areas of the law in which appellate courts have done just that.

2. THE DOCTRINAL CONSEQUENCES OF AUTOMATIC REVERSAL

   This Part examines in detail three errors that trigger automatic reversal: judicial bias; deprivation of conflict-free defense counsel; and failure to instruct the jury on the meaning of reasonable doubt. In each case, the rule of automatic reversal has created a conflict between two distinct purposes of appellate review. One purpose of appellate review is retrospective: the court must examine the fairness of the particular trial at issue in order to affirm a fairly obtained verdict and reverse an unfairly obtained verdict. A second purpose of appellate review is prospective: the court must articulate the best practices and procedures for future trials. The following Section describes a theory of how automatic reversal forces an appellate court to choose between those two purposes; the later Sections use that theory to explain the doctrinal history of the three errors described above. The final Section offers responses to expected objections.

   1. ...