Derrick v. State

¶1dissenting.

¶2Prior to Strickland v. Washington, 466 U.S. 668, 104 S.Ct. 2052, 80 L.Ed.2d 674 (1984), the State of Texas had gone some distance in articulating the law of attorney effectiveness in the context of criminal defense. See Clinton & Wice, Assistance of Counsel in Texas, 12 ST. MARY’S L.J. 1 (1980). It is regretable that we should abandon this work simply because, as a matter of federal constitutional law, the United States Supreme Court has now laid down a method for evaluating the thresh-hold level of performance demanded by the United States Constitution.1 Indeed, it is probable that the Supreme Court’s relative generality in recent years is intended to spur the several states toward a more detailed and comprehensive assessment of their own constitutions and statutes, unhampered by precise rules from the federal courts when more general guidelines will do. See e.g. Abrahamson, Criminal Law and State Constitutions: The Emergence of State Constitutional Law, 63 TEX.L. REV. 1141 (1985). Ironically, the behavior of this Court has been slavish. It seems we do not want our freedom, and would rather the Supreme Court of the United States simply tell us what to do. I urgently dissent to this disturbing trend.

¶3In years past, it was natural that this Court’s interpretation of fundamental rights coincide rather exactly with Su*277preme Court case law. At a time when our nation’s highest court was actively expanding the scope of constitutional protection, the threshold generally exceeded that promulgated by state governments. Accordingly, the Supreme Court devised elaborate procedural schemes to promote implementation of its new agenda. But the applicability of federal constitutional rights to the states is now rather well-settled, and the Supreme Court has begun to express the scope of those rights in more general language. Compare Aguilar v. Texas, 378 U.S. 108, 84 S.Ct. 1509, 12 L.Ed.2d 723 (1964) with Illinois v. Gates, 462 U.S. 213, 103 S.Ct. 2317, 76 L.Ed.2d 527 (1983). The first part of the Stricklandholding reflects this trend. Thus, the level of performance which is to be expected from criminal defense attorneys is largely made to depend upon a case-by-case application of general principles by state appellate courts. But the second part of Strickland, supra 466 U.S. at 693,104 S.Ct. at 2067, 80 L.Ed.2d at 697, holds that the United States Constitution will not be offended should a state place the burden of showing harm upon the criminal appellant rather than upon the prosecution.

¶4I think it manifestly improper and unjust to assign the burden of persuasion in this way. Indeed, I believe that the Supreme Court will ultimately retreat from such a position itself, because it is essentially unworkable in practice. The harmless error rule is, at least conceptually, a response to the technicalities which have plagued the law of pleading, evidence, and procedure for hundreds of years.2 It flows naturally from the indubitable truth that goals of the criminal justice system are not well-advanced when convictions must be reversed for “errors” which did not deprive the accused of a fair and impartial trial. Unfortunately, determining whether a trial was fair has come to mean predicting whether the jury, or other factfinder, would have reached a different result absent the error. And no one has yet articulated a satisfactory method for making such predictions in a principled way.3

¶5What is so often apparent in the opinions of appellate court judges is that the harmfulness of an “error” can only be gauged by the personal reaction of an individual to it. The appellate court judge must imagine that he stands in place of a juror and, taking the record of the trial as a whole, must faithfully ask himself whether he would convict had the “error” not been committed. If he pretends, as a way of speaking, that the question concerns not himself but the foreman of an actual jury, or even twelve jurors collectively, no one is *278fooled into thinking that he really knows how a particular juror or group of jurors were affected by the “error”. We know that he speaks for himself and that, in saying the juror or jurors would not have reached a different decision, he means that they would have reached the same decision he reached.4 He cannot quantify the probable truth of his conclusion, nor can he even elaborate his reasons for it. He can say why the “error” would not have affected his own decision, or why it would have made no contribution to his own verdict, Tex.R.App.Proc., rule 81(b)(2), but he cannot say why he believes it would not have affected the decision of others. Anyone who cares to undertake an inquiry will likely find that no written opinion of any appellate tribunal has ever given a satisfactory basis for concluding that jurors would have returned the same verdict regardless of legal error in the case.

¶6I am generally opposed to the so-called “harmless error rule” for this very reason. In our system of justice, fairness and impartiality are produced, if at all, by operation of legal rules and by the assignment of adjudicatory responsibilities. Those who fashion these rules, including the legislative and judicial branches of government, may be expected to consider the efficacy of what they produce and to decide what rules and standards are necessary to achieve fairness in the system as a whole. When a procedural or evidentiary rule seems not to work well, or when it seems to produce unjust results, the remedy is to amend or repeal it. The harmless error rule does neither. It leaves the law fully intact, but authorizes appellate court judges to pardon the violation of any legal precept, constrained only by their personal views of fairness and justice. This is not the sort of method Texans have a right to expect in a government of laws.

¶7I do not mean to suggest that appellate court judges are less qualified than jurors to decide ultimate issues of guilt and innocence. But in our system of criminal justice, such decisions are not a part of their adjudicatory responsibilities, and it is an outrageous travesty to confer such authority upon them by subterfuge. The harmless error rule, particularly when created by judicial review, represents a seizure by the bench of veto power inconsistent with its proper role in government.5 Bound to apply the law, appellate court judges may not conscionably refuse to do so even when the consequence seems manifestly unjust. They may, of course, fairly interpret statutes and binding precedent so as to harmonize inconsistencies and effectuate legislative intent. They must necessarily strike down legislation which offends the supreme law of Texas and of the United States. They may even abandon judicial precedent outright whenever changes in the law and public policy indicate a need. But they may not simply except individual cases from the rule of law.

¶8The American jury trial has become an enormously complex thing, and the potential for malfunction commensurately great. Perhaps it is no longer reasonable to suppose that trial judges and attorneys can discharge their respective functions without violating the law themselves. If this is so, the problem runs much deeper than any *279formulation of the harmless error rule can hope to sound. But I am not convinced that the system itself is fraught with such traps that a capable practitioner cannot avoid failing his professional responsibility.

¶9For example, virtually all rules of evidence and many rules of procedure exist at the option of the parties. Unless irregularities in methods of proof are objected to, no error is likely to occur at all. Unless certain procedures are requested, such as the assessment of punishment by a jury or specific instructions from the trial judge, their omission is usually not improper. Our case law is literally bursting at the seams with examples of waiver and forfeiture which run the gamut of potential errors that never became actual. “Harmless error” exists only to excuse that which, after everything else is said and done, should be inexcusable — the contumacious refusal of a trial judge or a lawyer to do what the law requires, even when he knows or should know better.

¶10Perhaps there is a small minority of cases in which some rule of law, not susceptible of greater precision, was violated in a way such that the logical possibility or probability of a different result was not appreciably affected. I am inclined to think, however, that no such case exists in which a better formulation of the rule, or even its complete abandonment, would not produce a more sensible remedy to the perceived injustice than a dubbing of “harmless” by judicial fiat. Still, given the fallibility of human prediction, I am willing to concede that the prosecution should be permitted an opportunity to prove the harmlessness of any given error in its unique factual context. This, of course, means persuading an appellate court to a high degree of confidence that the error did not contribute to the result, even though such errors do affect the result in most cases. See e.g., Tex.R.App.Proc., rule 81(b)(2).

¶11Having found that an error was properly, specifically, and timely preserved, that it was neither waived, forfeited, nor cured, and that the violated precept either cannot or should not be changed, it represents the most extreme disrespect for our system of justice to further require that the injured party bear the burden of proving harm. There should be no rule of law in our jurisprudence the violation of which is harmless in more than a miniscule number of conceivable cases, e.g. LaPoint v. State, 750 S.W.2d 180 (Tex.Cr.App.1988), nor have we any need of such a rule. Obliging a party who has been wronged to further show that he does not fall within a class this insignificant is illogical and antisys-temic.

¶12Yet we know from Stricklandthat the United States Constitution condones such an obligation. That does not mean, however, that the State of Texas must condone it. As has long been clear from our published opinions in this area of law, the effective assistance of counsel is not to be found in a catalog of “dos and don’ts”. We have not yet undertaken to prescribe a list of exercises, all or most of which must be performed by attorneys in every criminal case. Nor even have we sought to compile fatal defects which, if they occur, will necessarily vitiate a criminal conviction. I believe that the effective assistance of counsel must be gauged by the overall performance of an attorney, bearing in mind that it is his responsibility to investigate the facts of the case, examine the evidence, research the law, and present his client’s case to judge and jury in a way which seems best to him. His judgments must be educated, informed, and reasonable, but they are his judgments, constrained only by the ethics of his profession, the instructions of his client, and his own preparedness to make them.6

¶13*280Thus, I have no difficulty joining the United States Supreme Court in its conclusion that “[t]he benchmark for judging any claim of ineffectiveness must be whether counsel’s conduct so undermined the proper functioning of the adversarial process that the trial cannot be relied upon as having produced a just result.” Strickland, 466 U.S. at 686,104 S.Ct. at 2064, 80 L.Ed.2d at 692-693. But, in my view a “just result” is one produced when the system designed to produce it functions properly, within the tolerances prescribed by law, and not one in which the verdict happens to comport with extrasystemic, individual, or majority notions of justice. Therefore, if “counsel made errors so serious that counsel was not functioning as the “counsel” guaranteed by the Sixth Amendment”, Strickland, 466 U.S. at 687, 104 S.Ct. at 2064, 80 L.Ed.2d at 693, the trial cannot be relied upon, in my view, to have produced a just result.

¶14I cannot, without violating my commitment to the system itself, subscribe to the doctrine that an accused whose representation was so grievously deficient as to satisfy the first prong of Stricklandmust also “show that the deficient performance prejudiced the defense.” Strickland, 466 U.S. at 687,104 S.Ct. at 2064, 80 L.Ed.2d at 693. And I am especially reinforced in my belief by the Supreme Court’s peculiar idea that prejudice is shown to exist when “counsel’s errors were so serious as to deprive the defendant of a fair trial, a trial whose result is reliable.” Strickland, 466 U.S. at 687, 104 S.Ct. at 2064, 80 L.Ed.2d at 693. It bears repeating that the result of a trial is reliable if, and only if, it has been arrived at in the manner required by law. It is never reliable when it has not been. All trial results in which the accused has not received the effective assistance of counsel under circumstances where he is entitled to such assistance are unfair by definition and, therefore, necessarily unreliable. Our system prescribes no method for determining reliability outside of itself.

¶15Consequently, to the extent that Strickland’s second prong requires an analysis different than the first, it is surely wrongheaded, as are all typical formulations of the harmless error doctrine which impose upon appellate court judges a responsibility to predict what the outcome of trial would have been absent the error. Because the system offers no objective method according to which such predictions can be made, reviewing courts must determine the question subjectively. It would nearly always be possible, of course, to summon back the jurors and require them to testify regarding the basis for their verdict.7 In this way, some reasonably pertinent evidence might be obtained concerning the probability of a different result. But again, our system seems to preclude any such methodology.8

¶16Accordingly, requiring the defendant to “show that there is a reasonable probability that, but for counsel’s unprofessional errors, the result of the proceeding would *281have been different”, Strickland, 466 U.S. at 694,104 S.Ct. at 2068, 80 L.Ed.2d at 698, produces a catch-22 in the law. It imposes a burden of proof on the one hand and withholds the only real evidence available to meet it on the other. Appellate court judges are left to speculate regarding the probable impact of deficient representation on the unknown cogitations of particular jurors, which is precisely what the Supreme Court proceeded to do in Part V of its Stricklandopinion.

¶17Even assuming that such jurors “reasonably, conscientiously, and impartially applied] the standards that govern the decision”, without “unusual propensities toward harshness or leniency”, Strickland, 466 U.S. at 695, 104 S.Ct. at 2068, 80 L.Ed.2d at 698, it remains the province of the factfinder to weigh evidence, determine the credibility of witnesses, and compare inferential strengths and weaknesses against the standards of proof prescribed by law. Each of these functions is highly subjective and is an adjudicatory responsibility of the jury alone under our system of criminal justice.

¶18To know the probability of a different result, therefore, requires some knowledge of actual juror beliefs concerning the evidence adduced at trial. Yet the Supreme Court has held that “evidence about the actual process of decision, if not part of the record of proceeding under review, ... should not be considered in the prejudice determination”, even though “[i]n making this determination, a court hearing an ineffectiveness claim must consider the totality of the evidence before the judge or jury.” Strickland, 466 U.S. at 695, 104 S.Ct. at 2068, 80 L.Ed.2d at 698. It follows from this that the reviewing court, although bound by the evidence, is obliged independently to determine its credibility, to weigh it, and to assess the strengths of any inferences from it, since the probability of a different result cannot meaningfully be decided without knowing these things. A powerful bit of evidence, utterly disbelieved, is no evidence at all, and can alter dramatically the result of a trial. A compelling inference from evidence given little weight is a weak inference indeed. These things are for the jury, not for appellate court judges, and under our system of justice fair and reliable verdicts can only be expected when defendants have an absolute right to insist that all questions of weight and credibility be decided by a jury.

¶19The harmless error rule, in its various guises, is now, perhaps more than ever, the most popular judicial expedient for avoiding reversal in the face of undisputed trial error.9 Since harmless error is an appellate doctrine, having no logical place at trial, the burden of proving it is one of persuasion only. Additional evidence, beyond that contained in the record of trial, need not be adduced by either party. Indeed, no mechanism even exists to do so, unless we propose to sanction new trial hearings for this general purpose at which the accused is permitted to examine jurors regarding the basis of their verdict. Otherwise, the probability of a different outcome, absent the error, must be gleaned somehow from a record which necessarily includes no evidence directly pertinent to the inquiry.

¶20How one assigns the ultimate burden, therefore, will usually be dispositive of the issue. In short, the party to whom it falls will usually lose. The fact that this is not *282invariably so only serves to demonstrate that application of the harmless error rule is essentially unprincipled. Accordingly, where convictions are reversed based upon an appellate finding of harmfulness, one senses that the driving force behind the decision is a vague feeling of injustice on the part of some higher tribunal. Unable to articulate its reasons in logical fashion, the court instead speculates about unquantifiable risks. Its language typically represents a policy rationale against the kind of error in question, not an explanation of why the error is harmful in some cases and not in others. If called upon for a more principled analysis, the court can reliably be expected to respond with an apologetic protest that each case must be decided upon its own facts. General principles, of course, have no place in a scheme where each case is utterly unlike every other.

¶21Appellate court judges are not supposed to engage in factfinding. In particular, their province is thought to preclude substitution of judgment for that of the jury. Thus it is much easier for them to conclude that even a rational juror might as likely have found the accused guilty had the error not been committed. The conclusion is difficult to gainsay. Indeed, in light of our exposure to the aggregate technicalities of trial practice, including the myriad rules of evidence and procedure, and particularly as compared with ordinary notions of right and wrong common to the general public, it is not at all improbable that a verdict of guilt would result just as often in any given case if the trial judge committed fifty or hundred errors more than he did in fact.10

¶22Deciding what should matter in litigation has always been the province of law. If it is now to be measured against the probability of a different jury verdict, then it would be far better to abandon the legal rules in straightforward fashion than to disguise their repealer with a candy coating of harmless error. The violation of a legal rule at trial is, or should be, a serious matter. If it is not, even in most cases, it should either be rewritten or abandoned. That which does not serve, in the best judgment of a community, to accomplish the legitimate ends for which it was intended desperately needs reforming.

¶23The harmless error component of the Stricklandtest has been criticized by commentators fairly uniformly. Some take the view that a rule of actual prejudice should not apply to Sixth Amendment violations at all, since the Constitution means to preserve the efficacy of adversarial process, a systemic prerequisite to fair trials in this country. Gabriel, The Strickland Standard for Claims of Ineffective Assistance of Counsel: Emasculating the Sixth Amendment in the Guise of Due Process, 134 U.PA.L.REV. 1259 (1986). Others argue that the rule, even if legitimate, improperly assigns the burden of proof. [O’Brien] Comment, Judicial Jabberwocky or Uniform Constitutional Protection? Strickland v. Washington and National Standards for Ineffective Assistance of Counsel Claims, 1985 UTAH L.REV. 723; Genego, The Future of Effective Assistance of Counsel: Performance Standards and Competent Representation, 22 AM. CRIM.L.REV. 181 (1984). Both positions have merit, in my view, although the latter is easiest to defend.

¶24Certainly if we have a need to embrace the harmless error rule as a safety valve for that minority of cases in which violation of a legal precept, although error, was unlikely to have actually affected the fairness of a trial or its outcome, there is no excuse for not burdening the prosecution with an obligation to persuade appellate courts of this extraordinary fact. Fault has nothing to do with it. It is unfortunate that ineffective counsel, brought about not by any shortcoming of the state but by deficiencies in the performance of a defendant’s own lawyer, should result in a benefit to him and a waste of judicial resources. *283But representation by an attorney whose performance is so deficient as to violate his own client’s constitutional rights is no windfall. It is exactly the kind of egregious failure that should undermine everyone’s confidence in the verdict, unless that confidence can persuasively be restored by the prosecution.

¶25For these reasons, I dissent to this Court’s wholesale adoption of Strickland. Its harmless error component is, at best, needless, and its assignment of the persuasive burden clearly wrong. We are not bound to apply it under the Texas Constitution, and should not do so. Therefore, expressing no opinion with respect to the constitutional adequacy of defense counsel’s performance in the instant cause, I am unable to join the majority’s opinion.11