Rompilla v. Beard

Under AEDPA, federal habeas relief required more than showing that the state court was wrong: Rompilla had to show that its application of clearly established federal law was objectively unreasonable. Strickland supplies that law. It requires deficient performance measured against prevailing professional norms and resulting prejudice, while giving substantial deference to informed decisions made from counsel’s perspective at the time.

The Court acknowledged that Rompilla’s lawyers did not wholly neglect mitigation. They interviewed Rompilla and some relatives and consulted three mental-health professionals. Rompilla was unhelpful, at times misleading, and his relatives offered little about his childhood. The Court therefore did not hold that every potentially useful lead—such as school, juvenile, medical, or alcoholism records—had to be pursued on the facts presented.

But one investigative omission was dispositive. Counsel knew that the Commonwealth would seek a death sentence partly by proving Rompilla’s history of violent felonies, would rely on his earlier rape conviction, and would use the rape victim’s testimony to emphasize the prior crime’s violent details. The relevant court file was public and readily available in the courthouse where the capital trial was being held, yet counsel did not review it before the penalty hearing.

Reasonable counsel had to examine a file that the prosecution had specifically identified as a source of aggravating evidence. Reviewing it would allow the defense to learn the details the Commonwealth would emphasize, assess whether the prosecutor was presenting the prior testimony selectively, discover mitigating or extenuating evidence, and prepare a meaningful response. Counsel could not reasonably substitute interviews with Rompilla and relatives for reviewing records known to be central to the State’s case.

The duty was especially important because the prior offense was similar to the charged murder and thus endangered counsel’s residual-doubt strategy. Without learning the full context of the earlier crime, counsel could neither effectively rebut its aggravating force nor determine whether residual doubt remained a viable mitigation argument.

Contemporary ABA standards reinforced the conclusion. They directed defense counsel to investigate facts relevant both to guilt and penalty and to seek information held by the prosecution and law-enforcement authorities. The Court treated those standards as guides to reasonable professional conduct, not as independently binding constitutional rules.

The Court rejected the claim that it was imposing a per se requirement to review every prior-conviction file in every case. Its conclusion rested on the specific circumstances: the prosecution’s announced plan to use detailed testimony from a similar violent prior offense, the prior conviction’s central role in establishing death eligibility, the file’s ready availability, and the resulting danger to the defense’s chosen sentencing strategy.

Because the Pennsylvania courts rejected Rompilla’s claim at Strickland’s performance prong, they did not decide prejudice. The Supreme Court therefore considered prejudice de novo. Pennsylvania did not meaningfully contest that, if counsel had reviewed the file, the omission prejudiced Rompilla.

The prior-conviction file contained prison records that would have opened substantial mitigation leads. Those records described a childhood in a poor environment, early involvement with juvenile authorities, repeated incarceration, alcohol-related problems, low cognitive test scores, and indications of schizophrenia or other mental disorders. This information contradicted the benign picture counsel had drawn from Rompilla and the relatives they interviewed.

The records would have prompted further investigation. Postconviction counsel found evidence that Rompilla’s parents were severe alcoholics; that his father violently abused his mother and children; that the children lived in terror, isolation, and extreme neglect; and that Rompilla was locked in a filthy dog pen, lacked adequate clothing, and attended school in rags. School and juvenile records also documented his mother’s frequent absences and intoxication and showed IQ results in the intellectually disabled range.

Armed with this background, competent mental-health experts identified organic brain damage, profound cognitive impairment, and an extreme mental disturbance, likely connected to fetal alcohol exposure and childhood conditions. They further concluded that these impairments substantially affected Rompilla’s capacity to appreciate the criminality of his conduct or to conform his conduct to law.

The sentencing jury heard none of this. Instead, it heard only sparse pleas for mercy and testimony supporting residual doubt. The omitted evidence created a fundamentally different mitigation case, and although a jury still could have imposed death, Strickland requires only a reasonable probability of a different result. The evidence might well have affected the jury’s assessment of Rompilla’s culpability, so the Court ordered Pennsylvania either to conduct a new penalty trial or to accept a life sentence.