Jordan Steiker is the Judge Robert M. Parker Endowed Chair in Law and Director of the Capital Punishment Center at The University of Texas at Austin School of Law. Carol Steiker is the Henry J. Friendly Professor of Law and Faculty Co-Director of the Criminal Justice Policy Program at Harvard Law School.

Two terms ago in Glossip v. Gross, Justice Stephen Breyer, dissenting from the Supreme Court’s rejection of a lethal-injection challenge, set forth a comprehensive case against the American death penalty, calling for the court to revisit the question of its basic constitutionality. Over the past 40 years, several justices have questioned the constitutional viability of the death penalty, but Breyer’s dissent seemed more significant because it came at a time when the death penalty appeared newly vulnerable. Seven states had legislatively abandoned the death penalty within the past decade (although Nebraska has since reinstated it by referendum). Executions and especially death sentences were in free-fall, declining around 80 percent and 90 percent respectively from their yearly highs in the mid-to-late 1990s. The few death sentences issued in recent years were increasingly confined to a few counties within a few states, leading Breyer to the conclusion that the imposition of the death penalty was “unusual.” Breyer also argued that the death penalty had become “cruel” given evidence of wrongful convictions, continued arbitrariness in its administration and excessive (and unprecedented) delays between sentence and execution. The latter point has been of particular concern to Breyer, who has noted the suffering caused by prolonged death-row incarceration as well as the ways in which delay undercuts the deterrent and retributive rationales of the death penalty.

Though Breyer’s dissent was joined only by Justice Ruth Bader Ginsburg, many observers of the court believed that Breyer would not have called for the court to revisit the death penalty’s constitutionality unless he suspected that a majority might reach his conclusion. Justice Anthony Kennedy – who was then and likely remains the critical fifth vote – had issued his own remarkable opinion 11 days before Glossip condemning solitary confinement on death row, in a case in which that issue had neither been briefed nor argued.

This term’s capital cases suggest that the court is no closer to taking up Breyer’s challenge. The court continued to grant relief in cases involving appalling facts left unredressed by state or federal courts. Three of the four cases briefed and argued on the merits fell within this category. The court also ruled against one inmate, rejecting an extension of recent decisions forgiving state procedural defaults on federal habeas review. All four decisions seem unlikely to change the trajectory of the American death penalty, given the fact-specific nature of the cases and the reluctance of the court to speak broadly in its opinions. The court’s disinclination to interfere with state prerogatives absent egregious legal error was most dramatically illustrated by its willingness to allow the resumption of executions in Arkansas. State officials there had scheduled eight executions over an 11-day period because the scarce lethal injection drugs the state had on hand were set to expire. (Four executions were eventually consummated, including two on one day.) As the American death-penalty ship is sinking, the court seems content to continue its work at the margins, showing little inclination either to abandon its regulatory role or hasten the death penalty’s demise. Only Breyer has continued to press the court to hear broader challenges to the death penalty in several dissents from denial of review, reiterating his concerns about prolonged death-row incarceration.

The most attention-grabbing of the court’s interventions was its decision in Buck v. Davis reversing the U.S. Court of Appeals for the 5th Circuit’s unwillingness to consider a claim involving overt racial discrimination stemming from a 1997 Texas capital trial. Duane Buck was sentenced to death after his trial lawyer called an expert witness to testify about Buck’s dangerousness, and the expert suggested that Buck’s risk of future criminality was heightened because he was black. The case came to the court in a procedural quagmire. The fundamental injustice was the expert’s racist testimony and the failure of Buck’s lawyers at every stage to do anything about it. By the time the case arrived for review, the court was faced with the absurdly technical issue of whether the 5th Circuit should have granted a certificate of appealability on the question whether Buck’s prior habeas judgment could be reopened under the applicable federal rule. Ultimately, the court made clear that Buck’s lawyer had been ineffective in allowing the expert’s testimony; Buck also satisfied the requirement of prejudice because of the especially damaging nature of racist testimony. The state argued that the racist testimony was brief and minuscule in relation to the evidence received by the jury. In response, Chief Justice John Roberts issued the most memorable line of the opinion: “Some toxins can be deadly in small doses.”

The court’s willingness to cut through the procedural morass (and to pave the way for relief in Buck’s case) is emblematic of this term’s incrementalism. It is hard to imagine many other condemned inmates benefiting from this decision. Instead, the opinion has expressive value. In truth, when Buck is coupled with McCleskey v. Kemp, the court’s position seems to be that racism will command relief only when it appears in small, discrete doses. When racism appears in its broader, systemic form, the court retreats because of the impossibility of cabining relief.

Texas was an outlier in another of the court’s cases – Moore v. Texas – in which the court rejected the effort of the Texas courts to shrink the protection of the Supreme Court’s decisions prohibiting the execution of persons with intellectual disability. The highest criminal court in Texas, the Texas Court of Criminal Appeals, reacted to the Supreme Court’s ban by creating its own standard for reviewing findings of intellectual disability – a test that diverged markedly from the clinical approach endorsed by the Supreme Court. The CCA’s test emphasized, among other things, whether friends and family regarded the defendant as having intellectual disability. The CCA created its test because it doubted that the court meant to exempt all persons with intellectual disability and accordingly sought to limit relief to those whom average Texas citizens would regard as deserving of the exemption. The court reversed the CCA 5-3, but it was unanimous in rejecting the CCA’s invented test as unfaithful to the court’s decisions. Moore will require Texas to revisit a substantial number of cases decided under that approach, but the decision will have much less purchase outside of Texas where states have shown less overt hostility to the court’s categorical protection.

Even less dramatic was the court’s decision in McWilliams v. Dunn, reversing a death sentence because the defense was not afforded adequate psychiatric assistance pertaining to the defendant’s mental health. At James McWilliams’ trial, the prosecutor had cross-examined him about the neurological effects of his head injuries, eliciting the response, “I am not a psychiatrist.” McWilliams had not received expert assistance to review his mental health records, develop a defense strategy, prepare witnesses or testify. In these circumstances, the court found McWilliams entitled to relief and did not decide the question that had generated the cert grant – whether an indigent defense is entitled to a psychiatric expert who is part of the defense team or simply access to a neutral psychiatric expert available to both parties. Like Buck, McWilliams corrected an obvious miscarriage of justice without venturing much beyond the facts of the case.

Justice Neil Gorsuch’s presence was significant to the one defeat for a Texas capital defendant.  In Davila v. Davis, the court, divided 5-4, refused to extend its decisions allowing federal habeas review where a defendant receives ineffective representation both at trial and on state habeas (as in Buck’s case). Erick Davila alleged ineffective representation on appeal and on state habeas, and the court parsed the language and reasoning of its earlier decisions to cabin their reach. Least persuasively, the court cited a concern about opening the floodgates to federal habeas claims raising this issue, despite the fact that Texas cited only 10 cases (of which only one was meritorious) out of 7,500 habeas cases that had attempted to access federal habeas review in the one circuit that permitted the extension that the court disallowed.

The court’s hesitance to extend recent precedent is further illustrated by the court’s refusal to clarify the implications of a major decision of the 2015 term, Hurst v. Florida. Hurst required a jury rather than a judge to find facts rendering a defendant death-eligible, thus invalidating Florida’s statutory scheme permitting judicial override of a jury’s sentence. In the months following the decision, the court remanded several capital cases to Alabama, whose statutory scheme also permitted judicial override, for reconsideration in light of Hurst. Early in the 2016 term, Alabama’s highest court ruled that its judicial-override scheme was sufficiently different from Florida’s to survive constitutional invalidation. Despite its previous remands, the court denied certiorari, declining to determine whether Alabama was correct that no reconsideration of its judicial-override practices was necessary. The Alabama legislature has since repealed the judicial-override provisions, but dozens of the nearly 200 people on death row in Alabama have been sentenced by judicial override, so the question lingers unresolved.

As capital sentences diminish to new lows and executions become even rarer events, the court remains visible in the capital sphere, reluctant to give up the appearance of carefully monitoring the American death penalty. As it has in every decade since the resumption of capital punishment in 1976, the court grants review and decides capital cases at a rate wholly disproportionate to their numbers in the criminal justice system and out of line with their dwindling practical significance. The court’s docket of four capital cases in the 2016 term seems high in relation to the 30 death sentences and 20 executions nationwide in 2016. Though its capital-case docket remains robust, the court continues to avoid the sweeping questions about capital punishment raised by Breyer. And when it addresses the cases it chooses to hear, the court eschews broad rulings that would fundamentally alter state capital practices, focusing instead on correcting particularly egregious mistakes or injustices. In this respect, the court’s role relative to capital punishment has become similar to the role of capital punishment relative to criminal justice: an occasional reminder of some awesome power, often an expression of outrage, but little to show for the effort.