In the 1996-97 Term, the Court decided six cases (two of them, summarily) in the death penalty and habeas area. Divided evenly between states' and prisoner's victories, all were either unanimous judgments or five-to-four splits among the justices.
Lindh v. Murphy, a split decision for the defense, constituted a significant triumph for supporters of habeas. In a non-capital setting, with Justice O'Connor in the majority, the Court held the general provisions of the Antiterrorism and Effective Death Penalty Act of 1996 ("AEDPA") inapplicable to habeas petitions pending at the time the Act was passed -- April 24, 1996. That represented good news for many petitioners, who feared that the Act's drastic curtailments of the writ would effectively slam the doors to federal court in their faces.
That news was tempered, however, by defense losses in two other habeas cases that divided the justices. Like Lindh, O'Dell v. Netherland and Lambrix v. Singletary involved questions of prospectivity. But these related to prior decisions favoring prisoners rather than a recent adverse statute. Hence, the holdings that the earlier rulings did not apply in collateral proceedings harmed the death-sentenced petitioners, who had sought to rely on those precedents. In O'Dell, Justice O'Connor once more provided the pivotal vote -- this time, for the state. Both opinions serve to remind habeas proponents that the "good old law" before the AEDPA only rarely helped defendants, especially in the last few years.
In the wake of Lindh, most thorny issues concerning this statute remain to be resolved by the Court; they are percolating in the lower courts. The key issue, which the justices deftly avoided in Lindh, is whether -- or to what extent -- habeas courts must now defer to state decisions on legal questions and "mixed" questions of law and fact. Traditional independent review has furnished a crucial safeguard for constitutional rights. Probably, the Court cannot skirt this land mine too much longer.
Additional looming issues involve the Act's novel limitations periods, standards for evidentiary hearings, and special (so-called opt-in) provisions for capital cases in states that create an acceptable mechanism for assigning and paying counsel in post- conviction proceedings. No jurisdiction has yet been found that meets the prerequisites of this chapter. When it does apply, it will give complying states significant benefits by further constricting federal scrutiny and speeding up the decisional process.
Thus far, while the Court has already granted review in the upcoming Term in five death penalty or habeas matters, only one (arguably) has anything to do with the amended statute. Spencer v. Kemna (96-7171) and Trest v. Cain (96-7901), non-capital cases, both pose issues relating to habeas. Spencer poses a technical question regarding mootness. Trest, more importantly, challenges the Court of Appeals' invocation of procedural default against a petitioner when the state never raised this affirmative defense. Buchanan v. Angelone (96-8400) invites the Court to vacate a death sentence in the context of a jury charge that made no mention of mitigation, despite a request for appropriate instructions.
Spencer, Trest and Buchanan were all cert grants at the defendant's behest. Buchanan, in particular, gives the petitioner grounds for optimism -- nothing justifies the judge's refusal to instruct the jurors on the vital concept of mitigation.
The two other pending death cases, by contrast, involve petitions filed by the state; they appear less promising for the defense. Ohio Adult Parole Authority v. Woodard (96-1769) presents the question whether clemency procedures should be subject to due process constraints. Calderon v. Thompson (97-215), from the often-reversed Ninth Circuit, will address (on rather tangled facts) the limits of the Court of Appeals' power to recall its mandate to prevent a potential miscarriage of justice.
Though narrow and technical on its face, Thompson actually raises concerns at the heart of death penalty jurisprudence. The majority below overturned a capital sentence because it believed that counsel's ineffective assistance "cast grave doubt on the reliability of the rape conviction and the rape special circumstance finding," in the absence of which the jury could not have imposed death. It justified doing so -- in the unusual posture of a vote to rehear en banc the panel's refusal to recall the mandate -- on the ground that "procedural misunderstandings" prevented a timely en banc call and the panel, which had upheld the sentence, "committed fundamental errors of law that would result in manifest injustice."
A plurality would also have granted the writ on the ground that the prosecution relied on two conflicting theories of the crime at separate trials of Thompson and his co-defendant. In the words of Judge Fletcher: "Here, little about the trials remained consistent other than the prosecutor's desire to win at any cost."
Judge Reinhardt, concurring, excoriated Judge Kozinski's dissent suggesting that whenever an en banc call is missed, even in the capital context, "the error can be corrected in a future case where the problem again manifests itself." He wrote: "This conclusion is bizarre and horrifying in its implications, and is unworthy of any jurist." A different dissent angrily dubbed Thompson's motion to recall the mandate an "end run" around the AEDPA's restrictions on filing second or successive petitions.
Thus, Thompson points up the extreme divisiveness of issues relating to capital punishment. One cannot imagine a majority opinion by the Court -- either way -- without a dissent as emphatic, if not as sharp, as Judge Reinhardt's. More broadly, the heated exchange in Thompson reflects the tension between the desire to expedite executions and the fear of killing innocent people (or, at least, those who may not be guilty of the crime qualifying them for death).
There is now, in fact, a considerable danger of executing innocents. A recent study by the Death Penalty Information Center documents 69 releases of death row inmates, from 1973 to the present, based on belated assessments of innocence. The number of such cases has been growing. On the one hand, the overall expansion of the death penalty makes mistakes increasingly likely. On the other, the evisceration of habeas and related trends, such as the withdrawal of federal funding from the death penalty resource centers, surely prevents some errors from surfacing.
The defendants cited in the study spent, on average, approximately seven years on death row prior to release. Currently, the lag between sentencing and execution runs about eight years. When "streamlined" procedures substantially reduce the latter delay, innocent defendants will predictably be killed before proof of innocence emerges -- the mistakes will literally be buried with them.