Rethinking the timing of capital clemency.

• Author: Gershowitz, Adam M.
• Position: Introduction through I. What Do We Know at the End of Direct Appeals? B. Cases Where the Same Information Was Probably Available at the Conclusion of Direct Review, p. 1-31

This Article reviews every capital clemency over the last four decades. It demonstrates that in the majority of cases, the reason for commutation was known at the conclusion of direct appeals--years or even decades before the habeas process ended. Yet when governors or pardon boards actually commuted the death sentences, they typically waited until the eve of execution, with only days or hours to spare. Leaving clemency until the last minute sometimes leads to many years of unnecessary state and federal habeas corpus litigation, and this Article documents nearly 300 years of wasted habeas corpus review. Additionally, last-minute commutations harm the victims' families by delaying closure for years. And reserving clemency determinations for the very end of the process creates an information cascade that makes it harder for governors to grant clemency in meritorious cases. This Article therefore argues for a threshold clemency determination in capital cases at the conclusion of direct review, before any state or federal habeas litigation has begun.

TABLE OF CONTENTS INTRODUCTION I. What Do We Know at the End of Direct Appeals? A. The Clemency Basis Was Known at the End of Direct Appeals in Over Half of Capital Commutations 1. Doubts About Guilt 2. Defendant-Specific Characteristics: Mental Capacity, Mental Illness, History of Abuse, and Age 3. Questionable Evidence and Procedures 4. Different Treatment of Codefendants 5. Support for Clemency from Jurors and the Victim's Family 6. Ineffective Assistance of Counsel 7. Racial Discrimination 8. Failure to Preserve Evidence 9. Proportionality 10. Religious Conversion 11. Mixed Rationales B. Cases Where the Same Information Was Probably Available at the Conclusion of Direct Review C. Insufficient Information in Some Cases to Assess Whether Clemency Could Have Been Granted Earlier D. Cases in Which Crucial Information Came After Direct Review 1. Commutations Based on Recent Legal Developments 2. DNA and Other New Evidence 3. Ineffective Assistance of Counsel 4. Changes Related to the Defendant 5. Outside Influence II. The Case for Threshold Clemency Decisions Following Direct Appeals A. Efficient Clemency: Saving Hundreds of Years of Wasted Litigation 1. The Raw Numbers 2. The Story of Habeas Evolution B. An Earlier Clemency Decision Might Increase Commutations. 1. The Full and Fair Access Excuse 2. The Information Cascade Problem C. Moving Clemency to an Earlier Stage Benefits Victims' Families D. Responding to the Major Objection III. Implementing a Threshold Clemency Determination at the End of Direct Review Conclusion I am doing what a governor has to do.... I am taking a last, long look at it.

--Governor Buddy Roemer (1)

Introduction

In death-penalty cases, clemency is typically the last stage of the process. While governors have occasionally used their commutation powers to empty death row, blanket commutations are rare. (2) In the ordinary case, commutation comes only days or hours before execution. (3) Sometimes the inmate has already ordered his final meal when the governor steps forward to reduce the sentence from death to life imprisonment. (4)

There is a plausible argument for leaving clemency until the end of the process. In order to look holistically at a case, a governor or pardon board (5) must have all the information to make an informed decision. And important information may come from the years of appeals and postconviction litigation. For instance, it may only be after years of habeas corpus proceedings that governors would learn how terribly an inmate's lawyer performed or about new DNA evidence suggesting his innocence. (6) Therefore, the conventional wisdom is that governors and pardon boards should decide whether to commute a death sentence only at the very end of a case.

This Article challenges that conventional wisdom and advocates for a threshold clemency determination much earlier in the criminal justice process. Based on a review of every capital clemency decision in the last forty years, (7) I suggest that the governor or pardon board should make a clemency determination immediately after the conclusion of the direct appeals process but before any state or federal habeas corpus petitions are filed. (8) If governors or pardon boards decline to grant clemency at the end of the direct appeals process, they should retain the option to revisit that decision at the conclusion of the state and federal habeas corpus process. Put simply, clemency should still be last, but it should also be much earlier in the criminal justice process. To put the timing in perspective, the average time from conviction to execution is almost fifteen years. (9) Direct appeals typically take a few years, (10) with the habeas process covering the largest portion of the time. (11) The threshold clemency determination should therefore be made relatively early in the capital punishment process. No legal obstacle stands in the way of considering clemency at the end of direct review.

There are three reasons supporting the unorthodox proposal to consider clemency much earlier. First, when governors and pardon boards have commuted death sentences over the last few decades, they have usually based their decisions on information that was known before the habeas corpus process even began. (12) There have been sixty-six commutations for particularized reasons related to the inmate's case since the Supreme Court reinstated capital punishment in 1976. (13) To determine why governors or pardon boards commuted the death sentences, I reviewed news reports, direct appeals, and habeas corpus decisions, and I spoke with some of the attorneys involved in the cases. In more than half of the cases, the reason for the commutation was already known and fully developed at the conclusion of direct appeals. Governors and pardon boards could have avoided roughly 300 years of litigation and hundreds of millions of dollars in expenses in these cases if they had made their clemency decisions at the end of direct review rather than at the conclusion of the habeas process. (14)

The second reason for adding a threshold clemency determination earlier in the criminal justice process is that it may reinvigorate the use of executive clemency. Over the last half century, clemency has become a rarity. (15) While there have been more than 1,300 executions since the Supreme Court reinstated capital punishment in 1976, there have been only 66 individualized commutations. (16) By contrast, in the first half of the twentieth century, 1 out of every 4 or 5 death sentences was commuted to life imprisonment. (17)

In large part, the decline of clemency can be attributed to tough-on-crime politics. (18) There may also be another factor at play, however. Because habeas corpus avenues expanded in the 1950s and 1960s, the time between conviction and execution thereafter increased. (19) And as the Supreme Court and Congress imposed additional procedural obstacles in the 1980s and 1990s, (20) successful habeas claims declined. (21) Thus, by the time a clemency application lands on a governor's desk, it is now common for the inmate to have had more than a decade of direct appeals and postconviction habeas proceedings, with most, if not all, of these efforts having been unsuccessful.

Social science literature tells us that information cascades affect decisionmaking. (22) When an actor is asked to reverse a decision that many prior decisionmakers have upheld for many years, it is very hard to do so. By contrast, if we were to interject a threshold clemency decision early in the criminal justice process, before the years or decades of habeas litigation, it might be possible to limit the information cascade and spur governors or pardon boards to make a more independent decision. Earlier clemency review might therefore lead to more clemency.

The third reason for making initial clemency determinations before the habeas corpus process is to spare the family of the victims from emotional trauma. When a governor or pardon board commutes a death sentence, it is often very difficult for the victim's family to accept that decision. If the inmates' appeals and habeas petitions have been rejected for years or even decades, it is even harder for the victim's family to understand the last-minute commutation. It is far more preferable for the death sentence to be taken off the table earlier than for the governor or pardon board to quash the hopes of the family years or decades later.

The 2011 commutation of Shawn Hawkins in Ohio encapsulates the argument for a threshold clemency determination at the conclusion of direct appeals. Hawkins was sentenced to death in 1990 for two aggravated murders. (23) In 1993, the Ohio Supreme Court upheld his conviction and sentence, but Justice Pfeifer dissented, concluding that there was doubt about Hawkins's guilt. In particular, Pfeifer pointed to the possibility that a codefendant was the actual triggerman. (24) For nearly two decades thereafter, Hawkins litigated numerous state and federal petitions for postconviction relief. (25) With his habeas options exhausted, Hawkins petitioned for clemency in 2011 and relied heavily on Pfeifer's dissenting opinion from 1993. (26) In opposing clemency, the county prosecutor argued that "[m]ost of what Hawkins' counsel is now alleging and/or arguing has been litigated and examined during appellate review and is not new." (27) Yet after reviewing the old evidence, the pardon board was not certain that Hawkins was guilty and therefore recommended commutation to life imprisonment. (28) Governor Kasich concurred and, over twenty years after conviction, commuted Hawkins's sentence because of the very same concerns Hawkins had raised at the beginning of his case. (29) The victim's mother responded that "[i]f a man can spend 22 years in prison and still get clemency, then the system ain't working." (30) Hawkins's case is not unique. There are dozens of...