THE STATE OF THE DEATH PENALTY.

• Author: Desai, Ankur

The death penalty is in decline in America and most death penalty states do not regularly impose death sentences. In 2016 and 2017, states reached modern lows in imposed death sentences, with just thirty-one defendants sentenced to death in 2016 and thirty-nine in 2017, as compared with over three hundred per year in the 1990s. In 2016, only thirteen states imposed death sentences, and in 2017, fourteen did so, although thirty-one states retain the death penalty. What explains this remarkable and quite unexpected trend? In this Article, we present new analysis of state-level legislative changes that might have been expected to impact death sentences. First, life without parole ("LWOP") statutes, now enacted in nearly every state, might have been expected to reduce death sentences because they give jurors a noncapital option at trial. Second, legislatures have moved, albeit at varying paces, to comply with the Supreme Court's holding in Ring v. Arizona, which requires that the final decision in capital sentencing be made not by a judge, but by a jury. Third, states at different times have created statewide public defender offices to represent capital defendants at trial. In addition, the decline in homicides and homicide rates could be expected to contribute to the decline in state-level death sentencing. We find that contrary to the expectations of many observers, changes in the law such as adoption of LWOP and jury sentencing, did not consistently or significantly impact death sentencing. The decline in homicides and homicide rates is correlated with changes in death sentencing at the state level. However, this Article finds that state provision of capital trial representation is far more strongly and robustly correlated with reduced death sentencing than these other factors. The findings bolster the argument that adequacy of counsel has greater implications for the administration of the death penalty than other legal factors. These findings also have implications beyond the death penalty and they underscore the importance of a structural understanding of the Sixth Amendment right to counsel in our system of criminal justice.

INTRODUCTION 1256 I. STATE DEATH SENTENCING PATTERNS 1262 A. Prior Research on State Death Sentencing 1262 B. Data Sources 1264 1. Death Sentencing Data 1264 2. Homicide Data 1265 3. LWOP 1266 4. Judge Versus Jury Sentencing 1266 5. Capital Defense Regimes 1266 C. Empirical Strategy 1268 II. FINDINGS: EXPLAINING THE DECLINE IN DEATH SENTENCING 1269 A. Overview of Findings 1270 B. Homicide and Capital Sentencing 1271 C. State Adoption of Life Without Parole 1274 D. State-Level Capital Defense 1278 E. State Ring v. Arizona Compliance 1282 III. PRACTICAL AND CONSTITUTIONAL IMPLICATIONS 1288 A. Structural Sixth Amendment Implications 1288 B. Policy Implications 1291 C. Eighth Amendment Implications 1292 D. Implications for Future Death Penalty Trends 1294 CONCLUSION 1295 APPENDIX A 1297 APPENDIX B 1308 APPENDIX C 1309 INTRODUCTION

Use of capital punishment is declining in America. Death sentencing has fallen to a modern low and executions are increasingly rare. (1) While nineteen states have abolished the punishment, that is not a good explanation for this steep decline, since none were states that had imposed death sentences in large numbers. (2) What explains this decline? While scholars and journalists have increasingly commented on this decline and speculated as to what might be causing it, (3) empirical research has just begun to examine the question. (4) A recent book comprehensively analyzes the great American death penalty decline, (5) and it relies on the research presented for the first time in this Article tackling a central question: why some states have experienced greater declines in death sentences than others.

To analyze the state of the death penalty in decline, we focus on three types of legal changes that many suspected might affect death sentencings and focus on a defense-lawyering effect--the strong impact that the creation of state-level capital defense offices has on reducing death sentencing. We conclude by discussing implications of these findings for Eighth Amendment arbitrariness claims, Sixth Amendment right to counsel arguments, and criminal justice reform conversations more broadly.

The American death penalty today produces the fewest death sentences seen in over three decades. In 2016, thirty-one defendants were sentenced to death and in 2017, thirty-nine defendants were sentenced to death. (6) In contrast, in the 1990s, several hundred people were sentenced to death each year. In 2016, only thirteen states imposed death sentences and in 2017, fourteen states did so, although thirty-one states retain the death penalty as a legal option. Virginia, which is third only to Texas and Oklahoma in the number of executions in the modern era, has imposed no death sentences since 2011. (7) In the past few years, Texas has imposed only a handful of death sentences annually. As Carol Steiker and Jordan Steiker have explained: there are abolitionist states; but also "de facto or virtually abolitionist states," which retain the death penalty but rarely impose it (like Colorado); "symbolic states," which impose many death sentences but do not conduct executions (like California); and "executing states" (like Texas), which irregularly impose death sentences and carry out executions. (8) The resulting uneven decline makes the trend in modern American death sentencing complex and a subject in need of empirical study.

One trend that might explain that striking decline in death sentences is the national decline in murder rates in the United States since the mid-1990s. We find a significant effect of the decline in homicides on death sentences, but an effect that is highly inconsistent across states. For example, Texas experienced a sharp drop in capital sentencing as the number of murders fell. (9) However, murders fell even faster in California, and death sentencing remained relatively high. (10) In a separate work, we find that declining murder rates more strongly correlate with death sentencing at the county, rather than the state level. (11) However, at that county level, other factors, such as the race of the victims of homicide and inertia within counties, more strongly predict death sentencing. (12) Although county-level practices, like the preferences of prosecutors, may strongly impact death sentencing, changes at the state level can also be expected to affect death sentencing. After all, states regulate the death penalty through adoption of legislation defining death-eligible offenses and the procedures for capital trials; states maintain the death row, states conduct executions, and states may subsidize--if not fund--the defense and prosecution of capital cases. (13)

In this Article, we examine three key state-law changes relating to the death penalty that might be expected to contribute to the uneven state of the American death penalty: (1) the enactment of life without parole ("LWOP") statutes for capital murder, (2) the requirement of a jury determination on the presence of an aggravating factor, and (3) the establishment of state systems of capital representation. Each of these types of state legal changes were ones that many observers expected to impact death sentences. These changes were made at different times in different states, so they could be examined using statistical models and regressions, controlling for "fixed effects" or other state-specific factors.

First, a possible explanation for the decline in death sentencing may be the rise of an alternative sentence: life without parole. Nearly every state has introduced a life without parole sentencing option, often making noncapital defendants and even nonmurderers eligible for true life in prison. (14) LWOP adoption has been supported by both tough-on-crime conservatives, who sought to end parole, and death penalty abolitionists, who hoped that jurors would prefer not to impose death if a satisfactory alternative is available. (15) Many assumed that if the jury is given the choice whether to sentence a person to death or LWOP, that more jurors would take advantage of a nondeath sentence certain not to result in release. (16) LWOP "provides assurance to juries and victims' family members that perpetrators will not be set free." (17)

Second, another important possible factor is state legislation requiring that the jury and not a judge make the decision whether to sentence a person to death. The Supreme Court held in its 2002 Ring v. Arizona decision that capital sentences imposed by judges violate the defendant's Sixth Amendment right to a trial by jury. (18) Some but not all states had already adopted such a procedure, and some states were slow to comply with this ruling, (19) recently prompting the Court to reaffirm its stance. (20) Some observers wondered whether this change might impact states where elected judges might have incentives to aggressively impose death sentences. (21)

Third, reduced death sentencing may be attributed to improved representation of capital defendants--a defense-lawyering effect. Capital defense in America is almost always indigent defense, and it demands far greater knowledge, skill, and effort on the part of an attorney. (22) Since 1932, the Supreme Court has required a court-appointed lawyer for any indigent defendant charged with a capital crime, (23) and the Court has affirmed that the Sixth Amendment requires some minimal level of investigation of mitigating evidence that might provide the jury with a reason not to sentence a person to death. (24) But achieving widespread competent defense is an "enormous social task," (25) as Anthony Lewis observed in the wake of Gideon v. Wain-wright. (26) Claims regarding ineffective assistance of counsel have long been the most common reason why death sentences are reversed during post-conviction proceedings. The...