David Washington, facing severe personal and financial pressures, embarked on a short crime spree in Florida during which three people were killed. He voluntarily surrendered, confessed, and—on advice of counsel—pleaded guilty to multiple charges, including three counts of capital murder. Washington waived a jury for the penalty phase and was sentenced by the judge. At sentencing, Washington's attorney chose not to seek a continuance, did not request a psychiatric evaluation, did not conduct or present a substantial mitigation investigation, and called no character or background witnesses. Counsel's strategy was to accept responsibility, express remorse, and avoid opening the door to damaging information that could undermine the plea for mercy. The judge imposed the death penalty on each murder count. After the Florida courts denied relief, Washington sought federal habeas corpus, alleging ineffective assistance at the penalty phase based on counsel's limited investigation and failure to present mitigating evidence. The federal district court denied relief, but the court of appeals disagreed in part. The Supreme Court granted certiorari to clarify the standard for ineffective assistance of counsel and to decide whether Washington's showing entitled him to relief.
What is the proper standard for adjudicating Sixth Amendment ineffective-assistance-of-counsel claims, and, under that standard, did Washington demonstrate that his lawyer's performance at capital sentencing was constitutionally deficient and prejudicial?
A defendant claiming ineffective assistance of counsel must show both: (1) Deficient performance—counsel's representation fell below an objective standard of reasonableness under prevailing professional norms, with courts applying a highly deferential review and indulging a strong presumption that counsel's conduct was reasonable strategic choice; and (2) Prejudice—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. Courts may resolve an ineffectiveness claim on the prejudice ground without addressing performance. Strategic choices made after thorough investigation are virtually unchallengeable, and choices made after less than complete investigation are reasonable to the extent they are supported by reasonable professional judgments. In capital sentencing, prejudice asks whether absent counsel's errors there is a reasonable probability the sentencer would have concluded that death was not warranted.
The Supreme Court adopted the two-prong deficiency-and-prejudice standard for ineffective-assistance claims and held that Washington failed to meet it. His counsel's decisions were within the wide range of professionally reasonable judgment, and, in any event, Washington did not show a reasonable probability that additional investigation or mitigating evidence would have changed the death sentence. The Court reversed the court of appeals and denied habeas relief.
The Court began by articulating the constitutional baseline: the Sixth Amendment guarantees not error-free representation but reasonably effective assistance, sufficient to ensure a fair and reliable adversarial process. To implement that principle, the Court set forth a two-part inquiry. On performance, the Court emphasized deference to counsel's professional judgments and warned against the distorting effects of hindsight. The relevant measure is objective reasonableness under prevailing norms, not ideal or perfect lawyering. Strategic decisions made after adequate investigation (or after a reasonable decision that further investigation is unnecessary) are entitled to great deference. Applying those principles, the Court found that counsel's approach—advising guilty pleas, waiving jury sentencing, and making a restrained plea for mercy—reflected a reasonable strategy: to highlight Washington's remorse and acceptance of responsibility and to avoid opening the door to highly damaging evidence concerning the crimes and Washington's background that could have eclipsed any modest mitigation. On prejudice, the Court defined the standard as a reasonable probability of a different result, meaning a probability sufficient to undermine confidence in the outcome. This is a lower threshold than "more likely than not," but still requires a substantial, not speculative, showing. The Court authorized lower courts to address prejudice first when that is dispositive. In Washington's case, even assuming certain investigative omissions, the record contained overwhelming aggravating circumstances. The mitigation Washington faulted counsel for not presenting—such as generalized character evidence or a psychiatric evaluation not clearly indicated by counsel's interactions—did not create a reasonable probability the sentencer would have rejected death. Thus, confidence in the outcome was not undermined. The Court also situated the new standard within existing doctrine, noting limited categories where prejudice is presumed (for example, actual or constructive denial of counsel or certain conflicts of interest) but concluding those exceptions did not apply. Ultimately, the two-prong test was designed to respect the wide latitude afforded to defense counsel while still providing a meaningful check where deficient lawyering renders a result unreliable.
Strickland supplies the ubiquitous framework for ineffective-assistance claims in criminal cases. It governs trial, sentencing, plea bargaining, and appellate representation, and its concepts—objective reasonableness and reasonable probability—anchor both classroom analysis and litigation strategy. The decision powerfully influences defense practice: attorneys must conduct reasonable investigations into facts and mitigation, document strategic choices, and advise clients competently, particularly in capital cases. For habeas review, Strickland's deference to counsel often combines with statutory deference to state-court decisions, creating the so-called "double deference" hurdle for petitioners. Later cases (e.g., Hill v. Lockhart on guilty pleas; Wiggins v. Smith and Rompilla v. Beard on mitigation duties; Lafler v. Cooper and Missouri v. Frye on plea negotiations; Harrington v. Richter on habeas deference) elaborate but do not displace Strickland's core test.
Strickland v. Washington set the constitutional floor for defense representation by requiring a defendant to prove both objectively unreasonable lawyering and a resulting undermining of confidence in the outcome. The decision's deference to strategic judgments reflects respect for the adversarial system and recognition that competent advocacy varies with circumstances, while its prejudice requirement channels relief to cases where errors plausibly changed results.