Presumed Prejudice

Presumed Prejudice

The sections above applied one allocation of burdens to every stage of a prosecution. The convicted defendant must prove deficient performance and must also prove prejudice, which ordinarily means a reasonable probability that, but for counsel's errors, the result of the proceeding would have been different.¹ Showing that a lawyer fell below professional standards therefore does not by itself earn a new trial, because the defendant must also tie the lapse to a worse outcome. A short list of settings departs from that allocation. In some of them a court presumes prejudice, treating the harm as established without any evidence that the result would have changed, and in others the court still asks about harm but lowers what the defendant must show.² This section explains why those departures exist and how narrow they are, and the two sections after it teach the two most important ones.

The two halves of this arrangement answer different concerns. The prejudice requirement keeps the Sixth Amendment attached to reliable outcomes. Claims of ineffective assistance are raised constantly and rarely succeed, and a system that sent convictions back for a new trial or sentence whenever counsel erred would have no way to bring cases to an end.³ That reasoning loses force in two situations. The first is a harm that cannot be reconstructed after the fact, because a rule demanding proof of the unprovable would deny relief to exactly the defendants who were hurt most. The second is a defendant who lost an entire proceeding, because the injury is then the lost proceeding itself and there is no worse result inside it to point to. Powell v. Alabama shows the first situation at its starkest. No lawyer was named or designated to represent the capital defendants until the very morning of trial, and the trial judge had merely appointed all the members of the local bar for the limited purpose of arraignment.⁴ The Court held that the defendants had not been accorded the right of counsel in any substantial sense, and part of its reasoning was that no investigation was ever attempted, so neither the lawyers nor the court could say what one would have uncovered.⁵

The same logic supports the three situations that United States v. Cronic identified as ones in which prejudice may be presumed: a complete denial of counsel at a critical stage, counsel's entire failure to subject the prosecution's case to meaningful adversarial testing, and circumstances that make it highly unlikely that any competent lawyer could render effective help. Outside those situations, the Strickland prejudice standard applies.⁶ A critical stage is a point in the prosecution at which the accused needs a lawyer to secure a meaningful defense. In United States v. Wade a post-indictment lineup qualified, because confrontations of that kind can settle the accused's fate and reduce the trial to a formality.⁷ Each Cronic situation describes a lawyer who is absent, or so completely emptied of function, that asking what a better lawyer would have changed has no meaningful answer.

The presumptions are narrow enough that even a serious structural error does not automatically produce one. A structural error is a trial defect so basic that, when it is preserved and raised on direct review, the conviction is reversed without any inquiry into harm.⁸ In Weaver v. Massachusetts the courtroom was closed to the public during jury selection, counsel never objected, and the defendant first raised the closure through an ineffective-assistance claim. The Court required the defendant to show either a reasonable probability of a different outcome or a violation so serious that it rendered the trial fundamentally unfair.⁹ The reason is that a claim first raised after the conviction became final puts the interest in finality at greater risk, and without a prejudice showing it would let a defendant sidestep the ordinary rules of waiver and forfeiture.¹⁰ The allocation also differs sharply from the harmless-error rule on direct review. There, the beneficiary of a preserved constitutional error must persuade the court beyond a reasonable doubt that the error did not contribute to the verdict, as Chapman v. California held.¹¹ Under Strickland the defendant carries the burden of proving harm, which is why a presumption, where one applies, changes the outcome of a claim.

Two of these settings receive full treatment next. The first is counsel's conflict of interest, where the lawyer's divided loyalty leads the Court to presume prejudice in some situations and to demand a lesser showing than Strickland's in others.¹² The second is counsel's failure to file an appeal the client wanted, where prejudice is presumed and the relief is the missed appeal rather than a new trial.¹³ On an exam, the trap is treating every serious lawyer error as presumptively prejudicial. The presumptions reach only narrow, identifiable settings, so a missed objection, an uncalled witness, or a weak closing argument still requires proof that the result probably would have changed.

Conflicts of Interest

A conflict of interest reaches the Sixth Amendment by a different route from an ordinary lawyering error. A lawyer serving two clients with opposing interests may do every task competently by ordinary measures and still fail the defendant, because the defendant is entitled to the lawyer's undivided loyalty, and a conflict removes it.¹⁵ The Court therefore treats a conflict as its own way of being denied effective assistance, separate from performance so poor that it undermines confidence in the verdict.¹⁶ These claims most often arise when one lawyer represents several co-defendants. A defendant has no right to joint representation, yet multiple representation can still be a basis for reversal when it goes wrong. Under the ordinary Strickland allocation the defendant must prove a reasonable probability of a different result. For conflicts the Court has lowered that showing, directing automatic reversal in some circumstances and requiring a lesser showing of injury in others.¹⁷

Which showing applies depends on what the trial court was told or should have known, and the reason traces to what divided loyalty does to a lawyer's work. It operates on strategic choices: whether to cross-examine a witness, whether to press a theory that shifts blame to the other client, whether to explore a deal that would require one client to testify against the other. Most of those choices leave no trace in the transcript, because the harm is the road not taken. Strickland asks a reviewing court to imagine what an unconflicted lawyer would have done and then to decide whether the result would have differed. For a conflict that two-step guess is close to impossible, and a rule demanding it would deny relief to the defendants whose lawyers were most compromised. The conflict rules therefore ask whether the conflict influenced or limited counsel's decisions, and they presume prejudice where the court was alerted and did nothing. That presumption also gives trial judges a reason to act. The court is the one actor who can cure a joint-representation problem before trial, by inquiring and, if necessary, requiring separate counsel, and a judge who knows that ignoring a timely objection risks automatic reversal has every reason to look into it at the outset.

The first scenario is the clearest. When counsel timely objects that adequate joint representation of co-defendants is impossible and the trial court fails to investigate, prejudice is presumed.¹⁸ In Holloway v. Arkansas counsel told the trial court before trial that he could not adequately represent the co-defendants, the court did not investigate, and prejudice was presumed.¹⁹ The presumption has a deliberately harsh edge. Where the trial court requires joint or dual representation despite a conflict and over a timely objection, reversal follows even if the evidence of guilt was overwhelming.²⁰ The rule targets the court's failure to respond, so the strength of the state's case plays no part in it.

A close variant arises when no one objects but the trial court knows or reasonably should know that a particular conflict exists. The judge then has a duty to inquire, and if it is found, on appeal or on remand after further inquiry, that an actual conflict infected the representation, prejudice is presumed.²¹ That presumption depends on what the court knew or should have known. Where no one raised an objection and the court had no reasonable basis to suspect a conflict, the court had nothing to investigate, and the defendant is left with the third rule.²²

Cuyler v. Sullivan supplies that third rule. A defendant whose lawyer simultaneously represented a co-defendant, and who made no timely objection, must show that an actual conflict of interest adversely affected the representation.²³ The phrase adverse effect has a specific meaning. It does not refer to Strickland prejudice, so the defendant need not show a substantial likelihood that the outcome of the trial would have been different.²⁴ It means that counsel was influenced or limited in the strategic decisions made because of the other client's interests, so the defendant must point to a concrete and negative effect on the lawyer's exercise of professional judgment.²⁵ If the defendant had to prove a different outcome as well, the reduced showing would collapse into Strickland's own and the conflict doctrine would add nothing. A defendant who raised no objection and cannot show that counsel was influenced or limited in any strategic choice by the other client's interests therefore loses the conflict claim, because no adverse effect is shown.²⁶

The reach of the Cuyler approach is itself uncertain. These conflicts are rare outside the simultaneous representation of co-defendants, and in Mickens v. Taylor the Court suggested, in dicta, that the presumption of prejudice associated with Cuyler may be limited to such cases.²⁷ Dicta does not settle the question, so the safe statement is that the relaxed showing is secure for concurrent representation of co-defendants. For other kinds of divided loyalty, such as duties to a former client or the lawyer's own interests, the governing showing is less settled, and an answer should say so.

The fourth step is where students lose points, because they import Strickland into it. Equating adverse effect with Strickland prejudice raises the defendant's burden above what Cuyler demands. The inquiry runs to the lawyer's choices rather than to the verdict, and a defendant who proves that counsel pulled punches because of the other client has met it even when the evidence of guilt was strong.²⁸

Forfeited Appeals

A lawyer who lets a client's appeal lapse causes a harm that the general Strickland prejudice test fits poorly, so the showing is modified in this setting.³² Ordinarily the defendant must prove a reasonable probability that the result of a proceeding would have been different. When the thing lost is the appeal itself, there is no worse result inside the proceeding to point to, and a court asked what the appeal would have shown must guess at arguments that were never briefed. The doctrine responds by asking what passed between client and lawyer. In one scenario the client wished to appeal and the lawyer never filed a notice; in the other the lawyer never discussed an appeal at all. The doctrine frames each scenario differently, and a careful answer keeps them apart.

The first scenario is the simpler one. In Roe v. Flores-Ortega the Court stated that when a defendant wishes to appeal and counsel fails to file a timely notice of appeal, prejudice is conclusively presumed.³³ A conclusive presumption means the defendant need not prove that the appeal would have succeeded or anything else about its likely outcome, because the failure to file is enough once the wish to appeal is shown. Other decisions have found a Sixth Amendment violation where counsel failed to file a timely notice of appeal, as in Evitts v. Lucey, and where counsel failed to file at the client's request after the client signed an appeal waiver, as in Garza v. Idaho. The presumption rests on the same reasoning that governs a defendant who lost an entire proceeding. The appeal was the only forum in which the conviction would have been reviewed, and once it is forfeited there is no record of what review would have accomplished. Requiring proof of a different outcome would ask the defendant to prove the result of a proceeding that never took place.

The remedy matches the harm. Because what the defendant lost was the chance to appeal, the relief is permission to file the missed appeal, and the conviction itself is not overturned.³⁴ The appellate court then decides whether anything in the trial was wrong. If that appeal succeeds and a new trial follows, a separate protection applies. In North Carolina v. Pearce the Court held that due process bars vindictiveness against a defendant who successfully attacked a first conviction, and that a harsher sentence after a new trial must rest on reasons that affirmatively appear, grounded in objective information about identifiable conduct after the original sentencing and placed in the record.³⁵ That rule keeps a restored appeal from becoming a trap, since a defendant who feared retaliation might otherwise decline to use it.

The second scenario arises when the client gave no instruction and counsel never raised the subject. Counsel's failure to discuss the advantages and disadvantages of an appeal, and to find out whether the defendant wants one, usually counts as unreasonable professional assistance.³⁶ The deficiency threshold is met if a rational defendant would have wanted to appeal, for example because nonfrivolous grounds existed for challenging a guilty plea, or if the defendant expressed an interest in appealing.³⁷ The logic is practical: a lay defendant often cannot tell which claims survive a conviction, so the lawyer's conversation is how the client's wish becomes known, and a lawyer who never asks has no means of learning it. Prejudice in this scenario is framed as a reasonable probability that an appeal would have been filed had counsel discussed the subject with the defendant.³⁸ The framing looks to whether the opportunity was lost, which matches the idea that the injury is the lost appeal, but the defendant still carries a burden here that the presumption in the first scenario removes.

A guilty plea often comes with a waiver of the right to appeal, which might seem to excuse counsel from raising the subject. In Garza v. Idaho the Court explained that an appeal waiver does not bar every claim that might be raised on appeal, and a defendant who signed one can still contest, for example, the voluntariness of the waiver itself.³⁹ The Court also noted that a defendant may explicitly ask counsel to file an appeal despite the waiver, a request counsel cannot ignore.⁴⁰ So the general rule that failing to discuss an appeal with a client who pleaded guilty is usually unreasonable applies even when the plea agreement contained a waiver.⁴¹ The reason follows from what the waiver leaves open. Because some claims survive it and the client may still want to pursue one, the lawyer needs the conversation to learn which situation exists.

The usual error is carrying the presumption into the failure-to-consult scenario. The conclusive presumption attaches to the defendant who wished to appeal and whose lawyer failed to file. The failure-to-consult scenario runs through the rational-defendant or expressed-interest threshold and then a reasonable-probability showing. An answer should state each framing beside the facts that trigger it and should avoid asserting that the presumption reaches a client who never asked.