AEDPA’s Rorschach Test


Edward Lang is condemned to die. During his murder trial, the system malfunctioned twice: a member of the victim’s family was on the jury until the second day of trial, and defense counsel did not present or uncover the details of Lang’s terrible childhood. On Friday, the Sixth Circuit issued an opinion addressing whether the Ohio courts that adjudicated Mr. Lang’s appeal reasonably applied U.S. Supreme Court precedent. The panel split. Two judges said that the Ohio courts’ tolerance for these errors was reasonable. Judge Moore dissented.

In many ways, this case is more about the deferential standard of review required by the Anti-Terrorism and Effective Death Penalty Act, 28 U.S.C. § 2254, commonly referred to as “AEDPA” (ED-pah), than the constitutional questions it raises. Bill Clinton signed AEDPA into law after the Oklahoma City bombings. (The Intercept has an interesting story about AEDPA’s legislative history.) In Criminal Law 2.0, former Ninth Circuit Judge Alex Kozinski had this to say about the statute:

Hidden in its interstices was a provision that has pretty much shut out the federal courts from granting habeas relief in most cases, even when they believe that an egregious miscarriage of justice has occurred. We now regularly have to stand by in impotent silence, even though it may appear to us that an innocent person has been convicted.

AEDPA is a cruel, unjust and unnecessary law that effectively removes federal judges as safeguards against miscarriages of justice. It has resulted and continues to result in much human suffering. It should be repealed.”

Many others agree with that point of view. Because I am a federal public defender, you can probably guess what my view is.

I believe Lang’s case illustrates AEDPA in action and provides an opportunity for you to judge for yourself whether the AEDPA standard of review is good policy.

1.     The Remmer Claim

The judge, prosecutor, and defense counsel did not learn that one of the jurors was related to the victim until they noticed the juror nodding and smiling at the victim’s family members who were in the gallery. On the second day of trial, the victim’s father approached the prosecutor and revealed that the juror was the victim’s niece by marriage. When questioned by the judge, the juror confirmed the relationship, but said that she did not know much about the murder and had not shared with other jurors any information about her relationship to the deceased. The court dismissed the juror and then brought the remaining members of the jury back for questioning. The court told the remaining jurors that the juror “may have had a relative relationship with either a witness or a party or somebody that was involved in the case.” The judge asked the jurors if she had told them about this relationship, and the judge accepted the jurors’ silence as a “no.” The court asked no more questions, and the trial proceeded.

On appeal, Lang argued that the trial court did not adequately investigate or address the impact the biased juror had on the outcome. When faced with a claim of juror bias, Remmer v. United States, 347 U.S. 227, 230 (1954), requires trial judges to “determine the circumstances, the impact thereof upon the juror, and whether or not it was prejudicial, in a hearing with all interested parties permitted to participate.” Petitioners must prove actual bias.

The Ohio Supreme Court rejected Lang’s claim of juror bias, reasoning that the trial court had done enough by asking the biased juror and the other members of the jury whether she had discussed her relationship with the victim. Never mind that defense counsel did not ask to question the remaining jurors individually.

The Sixth Circuit majority said that the Ohio Supreme Court’s decision to sanction this one-question Remmer was not unreasonable or contrary to Supreme Court law. Why? The trial court removed the juror, and she did not participate in deliberations. Both the prosecutor and defense counsel were present when the judge questioned the jurors. And the biased juror’s assurances and remaining jurors’ silence was enough to dispel any concern that the deliberations were unaffected by her presence. Plus, Remmer articulates a generalized standard, and the Ohio courts get a lot of leeway when deciding whether they have complied with the Supreme Court’s generalized standards.

Judge Moore disagreed. She said the trial court’s inquiry “was less than minimal,” and the single question “was overly narrow because it focused only on whether Juror 386 had revealed her relationship to [the victim] to her fellow jurors, and not on whether Juror 386 had tainted the remaining jurors’ ability to be impartial through other biased comments.” She also believed that because “being forced to speak up in front of the rest of the jury panel [has] a depressing effect on [individual jurors’] ability or willingness to be forthcoming.” She argued that the trial judge had to engage in a more searching inquiry to determine whether the juror’s presence and sudden absence impacted juror impartiality. She compared the judge’s conduct to “a doctor trying to determine if a patient had caught an infectious disease from an afflicted acquaintance by asking only if the patient had shared a drink with that person” without “determining whether the two individuals had other interactions through which the disease could be communicated.” In Judge Moore’s view, the Ohio courts unreasonably concluded the Remmer inquiry was so limited.

The majority and Judge Moore disagreed about whether it is reasonable to believe the one-question Remmer hearing adequately addressed the potential for jury bias. What do you think?

2.     The Strickland Claim

Lang presented two reasons to believe his attorney rendered ineffective assistance: (1) the attorney failed to investigate and present mitigation evidence about his horrific childhood; and (2) during closing arguments, counsel described Lang’s childhood as “normal” when it was anything but. To prevail, Lang had to satisfy the Strickland standard: (1) counsel’s performance fell below an objective standard of reasonableness; and (2) that deficient performance prejudiced Lang. Strickland is already a deferential standard that gives defense counsel a lot of leeway to make strategic decisions. And the Supreme Court has said, federal habeas courts must give “double deference” to defense counsel’s decisions.

So, what did Lang’s attorney do during the mitigation phase? He called Lang’s half-sister and mother to testify. Both said that Lang lived with them until he was 10, when Lang’s father abducted him for two years. Lang’s father had been mostly absent from his life because he was incarcerated after setting fire to Lang’s apartment, molesting a girl, and raping Lang’s mother. They neglected to mention: that Lang’s father sexually and physically abused him, that Lang saw his father tie his mother up for days and force her to perform oral sex, that he shot out windows, that he stabbed Lang’s mother in the chest with scissors, and that Lang was in the house when his father set fire to it. These witnesses also did not describe the torture Lang endured when his father abducted him. His father locked Lang in a room for days and beat him with anything in sight. The jury did not learn that Lang started using drugs during that time to numb the pain. Also missing from the mitigation testimony was information about Lang’s older brother, who physically and sexually abused Lang and his sister.

These facts were missing from the record, in part, because Lang’s counsel did not obtain Lang’s foster care records corroborating these details until after the mitigation phase had ended and the jury had recommended a death sentence. The defense psychologist could not write a report with the benefit of all of the records. And Lang’s mitigation specialist met with Lang’s mother for only 25 minutes 10 days before the mitigation phase began. Lang’s attorney had met with her only once before, the day before the mitigation phase began. He chose either to overlook this evidence or never learned it in the first place. And then, during closing argument during the mitigation phase, counsel told the jury that Lang’s childhood before he was abducted was “normal.”

The Ohio courts rejected Lang’s claim of ineffective assistance of counsel, reasoning that counsel conducted sufficient investigation and made a strategic choice not to present evidence of how Lang’s father treated him before abducting him. The Ohio Court of Appeals suggested that strategy played a role because information about Lang’s early childhood would discredit his mother’s testimony because she was also abusive. Finally, the Ohio courts said Lang wasn’t prejudiced by his attorney’s under-preparation because the additional information about his horrific childhood was just cumulative.

The majority of the Sixth Circuit held that these determinations were reasonable, hypothesizing that Lang’s counsel chose not to call a psychologist or introduce records because they showed that Lang was violent, anti-social, and hyper-sexual at a young age. Also lacking, said the majority, was sufficient evidence proving that Lang’s counsel made decisions about the mitigation strategy without adequate information. The court faulted Lang for not offering direct evidence of counsel’s mitigation strategy. Finally, the court said that the Ohio courts reasonably concluded that the additional information about Lang’s early childhood was cumulative, and therefore was unlikely to impact the jury’s decision to impose a death sentence.

Judge Moore was not so impressed by the Ohio courts’ handling of Lang’s IAC claim. Citing the professional standards of the America Bar Association, she argued that trial counsel’s failure to find available documentary evidence and learn more about Lang’s childhood fell below professional standards in death penalty cases. The investigation was so lacking, she explained, that counsel could not have made a reasonable strategic decision not to present the information. She similarly rejected the Ohio courts’ view that presenting this evidence would have undermined the credibility of Lang’s mother and dehumanized Lang. Finally, Judge Moore could not accept the Ohio courts’ conclusion that Lang was not prejudiced by his counsel’s underwhelming performance. She faulted the Ohio courts for suggesting that this mitigation evidence would have to impact every juror, when the Supreme Court has said, when analyzing prejudice, the question is whether there is a reasonable possibility the omitted mitigation evidence would have influenced one juror. In her view, the Ohio court’s prejudice analysis was lacking.

The central disagreement was about whether the Ohio courts reasonably applied the Strickland standard to these facts. Was it reasonable to say that defense counsel made reasonable strategic choices? Was it reasonable to conclude that Lang suffered no prejudice? What do you think?

* * *

There is no doubt that AEDPA’s standard of review drove the outcome in this case. But Judge Moore posed a question that is also worth considering: When a man is condemned to die, does Due Process demand this little? She “do[es] not believe . . . that the protections guaranteed by our Constitution are so minimal.”

How about you?

No comments: