An empirical examination of the factors associated with the commutation of state death row prisoners' sentences between 1986 and 2005.

• Author: Kraemer, John

Commutation is usually a death row prisoner's last hope of evading his or her capital sentence. However, unlike many other stages of the death penalty process, little research focuses on the factors that affect decisions to commute or allow a death sentence to go forward, and that which has been conducted utilizes data which is now nearly a decade old. This paper seeks to examine personal and demographic factors associated with commutation decisions and to resolve inconsistent findings in the prior research. Using the statistical method of multiple logistic regression, this paper finds statistically significant disparities in the odds of commutation by sex (women have an eleven-fold increase in odds of commutation), race (nonwhite prisoners have twice the odds of commutation), geography (southern prisoners have less than one-fifth the odds of commutation), and education (college educated offenders have one-fifth the odds of commutation). After adjusting for other factors, this research does not find evidence that, across the run of cases, criminal history or severity significantly influence commutation decisions. This research, while unable to generate conclusions about any individual case, provides evidence that executives' commutation decisionmaking is driven more by personal characteristics--some of which are troubling--than by criminal culpability.

1. BACKGROUND

   The King may pardon any felon: but it may be objected that if he pardon one felon, he may pardon all, to the damage of the commonwealth; and yet none will doubt but that is left to his wisdom ... And the wisdom and providence of the King is not to be disputed by the subject; for by intendment they cannot be severed from his person, and to argue ... that by his power he may do ill, is no argument for a subject. (1) A death row inmate's petition for clemency is ... a "unilateral hope." The defendant ... appeals for clemency as a matter of grace.... Under any analysis, the Governor's executive discretion need not be fettered by the types of procedural protections sought by respondent. (2) Separated by nearly four centuries, these, decisions bear striking similarity: the power to commute belongs to the executive and is generally beyond substantive review. In the United States, constraints on the power to commute are controlled by a fractured decision of the Supreme Court in Ohio Adult Parole Authority v. Woodard. (3) Four justices found essentially no judicial review of clemency decisions, deeming it instead an act of grace. (4) Three others joined Justice O'Connor and would review for minimal procedural safeguards, finding a due process violation if clemency was granted or refused, for example, by flip of a coin or if the prisoner was arbitrarily barred from whatever procedures the state had created. (5) Justice Stevens alone argued that the executive could not "ignore the commands of the Equal Protection Clause and use race, religion, or political affiliation as a standard for granting or denying clemency" (6) and that clemency procedures, if created, must meet "the most basic elements of fair procedure." (7)

   Lower courts generally have followed the O'Connor position and found commutation procedures to be acceptable so long as they are not so whimsical as a decision made by the flip of a coin (8) or with disregard for procedural requirements guaranteed by state law. (9) However, one court found that there is no liberty interest in a commutation, so changes to the clemency procedures cannot trigger a due process claim. (10) Several courts have found no right to present any particular information during clemency hearings. (11) Additionally, several courts have explicitly found that the substantive grounds for commutation are left to the sole discretion of the executive. (12) At least one court has suggested equal protection claims cannot be made in clemency proceedings, (13) which is troubling in light of this paper's finding that race and sex appear to be implicated in executive decision making about commutation.

   Though commutation is largely unregulated by the courts--or perhaps, because of this--there is a compelling need to consider the factors that influence commutation decisions. The Supreme Court has spoken of executive clemency as "the 'fail safe' in our criminal justice system." (14) At the same time, critics have denounced commutation as too arbitrarily and infrequently used. (15) This Note quantitatively examines those factors associated with contemporary commutations of state prisoners, updating the empirical literature on this subject by nearly a decade. The results show meaningful disparities in commutation on the basis of sex, race, and geographic region. Controlling for other factors, women are much more likely to be commuted than men, (16) as are people of color when compared to whites. (17) Furthermore, prisoners sentenced in a southern state are far less likely to receive a commutation than their counterparts from other regions; (18) the same is true for prisoners with college educations. (19) Perhaps most disturbing, this Note finds no evidence that, across the run of cases, past criminal history or severity of the capital offense are associated with the likelihood of commutation. (20)

   This paper's primary strength over prior research is that it uses a data set nearly a decade newer than any examined previously. This is particularly important because this paper finds evidence of significant changes in commutation behavior since the last papers' data were collected. Additionally, it includes more nuanced measures for severity of the underlying offense and the offender's criminal history. However, unlike some of the other research, it does not include an investigation of government structures relating to decision-making, often referred to in the literature as "structural factors." (21) This exclusion should not, though, impact the associations discerned in this study between the measured variables and commutation.

   Part II of this Note examines prior empirical research on commutation decisions, looking first at demographic factors, (22) and then at crime-specific and political or structural factors. (23) Part III lays out the methodologies used in this Note to identify key factors associated with commutation decisions. The findings of this original research are described in Part IV, along with suggested legal and policy implications.

2. PRIOR EMPIRICAL RESEARCH ON FACTORS ASSOCIATED WITH COMMUTATION

   1. Demographic Factors

      Only a handful of studies have used empirical evidence to examine the factors associated with commutation of death sentences in the modern judicial system, (24) and many of the reported results are inconsistent. The four primary studies were conducted by William Alex Pridemore, using data from 1974 to 1995; (25) Laura Argys and H. Naci Mocan, using data from 1977 to 1997; (26) Jeffrey Kubik and John Moran, using data from 1972 to 2000; (27) and Michael Heise, using data from 1973 to 1999. (28) These studies take two broad forms: investigations of how political or structural factors impact commutation decisions (29) and investigations of the effects of prisoner characteristics on commutation. (30) Because the newest research is based on data now nearly a decade old, there is a need for examination of more recent data to better understand factors presently associated with decisions to commute or allow executions to go forward. This study both examines more recent data and controls for historical trends in the frequency of commutations. The results of this research are reported in Part IV.

      It is well established that female capital defendants are much less likely to be sentenced to death than male capital defendants. (31) There is also strong evidence that a high proportion of women sentenced to death are not executed. (32) The Pridemore study estimates that the odds of execution instead of commutation are more than sixty times greater for men than women. (33) The research focusing on the most contemporaneous data has generally found a statistically significant ten to twenty-fold increase in the odds of commutation for women, when compared to men (and controlling for other factors). (34) However, precise estimates of the effect of sex on commutation are difficult to obtain because of the relatively small number of women whose sentences of death go to final disposition.

      Fairly strong statistical evidence supports the claim that black defendants are more likely to be sentenced to death than similarly situated white defendants. (35) Additionally, there is very strong evidence that the race of the murder victim has a substantial effect on the likelihood that an offender will be sentenced to death. (36) However, the effect of race on commutation is more uncertain. Heise and Pridemore did not find race to be a significant predictor of commutation. (37) On the other hand, Argys and Mocan found that prisoners of black, Hispanic, or other racial or ethnic heritage are more likely to be commuted than white prisoners. (38) They also found that state governors who are white are more likely to commute white prisoners' sentences than minority prisoners' sentences (39) and that Democratic governors are more likely to commute the sentences of minority inmates than whites. (40)

      It is possible that the relatively small number of observations in all of these studies could mask a subtle race effect in those studies that did not find one. Additionally, if a race effect exists, its interpretation is difficult based on these data. One could plausibly argue that such an effect mitigates racial disparities in sentencing, (41) but one might also argue that such an effect reflects impermissible racial considerations in carrying out the sentence. Unfortunately, no studies have investigated the effect of the victim's race on commutation; if executives weigh the lives of black victims less heavily--and murders are disproportionately...