| ----------------------------------------------------------------------- This is a prelininary rough draft of a paper presented at the Academy of Criminal Justice Sciences annual meeting in Washington, D.C. in April 2001. Please do not cite or quote. ----------------------------------------------------------------------- DISCRIMINATION AND THE DEATH PENALTY: BASIC PATTERNS Ernie Thomson INTRODUCTION In June 1972, a bitterly-divided United States Supreme Court ruled in a landmark decision that Georgia�s capital punishment system resulted in arbitrary, capricious, and discriminatory death sentences in violation of the Eighth Amendment ban on cruel and unusual punishment (Furman v. Georgia 1972). While the Justices in the Furman majority agreed that Georgia�s death penalty had been applied arbitrarily and capriciously, they were more ambiguous when it came to the issue of discrimination. This ambiguity, and its source, is hinted at in Justice Potter Stewart�s comment that These [Georgis] death sentences are cruel and unusual in the same way that being struck by lightening is cruel and unusual. For, of all the people convicted of rapes and murders in 1967 and 1968, many just as reprehensible as these, the petitioners are among a capriciously selected random handful upon whom the sentence of death has been imposed. My concurring Brothers have demonstrated that, if any basis can be discerned for the selection of those few to be sentenced to die, it is the constitutionally impermissible basis of race. But racial discrimination has not been proved, and I put it to one side. (p. 238) The Court�s ambiguity apparently derived from the fact that statistical evidence of systematic discrimination could not, by itself, prove that discrimination occurred in any particular case. Statistical evidence clearly showed that African American men had long been over-represented among those executed in the United States, especially in southern states like Georgia and especially when the alleged victims were white. For example, more than half of the men executed in the 1900�s in the United States were African Americans, while African Americans have never made up more than about twelve percent of the adult population. The evidence was even more striking for capital crimes other than murder. For example, about ninety percent of the men executed for rape in this century were African American and virtually all of the alleged victims were white women (see Gross and Mauro, 1989). Following the Furman decision, a number of states revised their capital punishment statutes and once again began issuing death sentences. Four years later in Gregg v. Georgia (1976), a now more conservative Supreme Court upheld several of these revised statutes that appeared to establish a more rational and predictable process for death sentencing. Among the new features of these statutes endorsed by the Court were bifurcated trials (separate guilt/innocence and penalty phases of capital trials), �guided discretion� for juries/judges in deciding sentences (consideration of aggravating and mitigating circumstances), and reviews of virtually all death sentences by state appeals courts, with some states including �proportionality reviews� by the highest state courts to ensure fair and proportionate sentences. Although the changes set in motion by the Court in Furman and Gregg appeared to lead to some improvement, early research indicated that the same patterns of arbitrary and capricious death sentencing continued, and specifically that race and ethnicity still played a disturbingly large part in death sentencing in the United States. Numerous studies of post-Furman death sentencing have indicated that minority men, especially African Americans, remain heavily over-represented among those receiving death sentences, especially when the victims are white (e.g., Baldus et al, 1990; Bowers 1984; Ekland-Olsen 1988; Gross and Mauro 1989; Keil and Vito 1990; Kleck 1981; Paternoster and Kazyaka 1988; Smith 1987). A 1990 evaluation of the extant research by the United States General Accounting Office found that the quality of the overall research supported the conclusion that race continued to exert a strong influence on death sentencing in the United States. Studies published since the GAO Report have found similar patterns (e.g., Karns and Weinberg 1991; Marquart et al 1994; Radelet and Pierce 1991; Thomson 1997). As clear evidence of the influence of race accumulated, some criminologists pointed out that the disparities might result from the fact that homicide rates are much higher among minorities than among whites (e.g., Kleck, 1981). But statistics clearly indicated that a large majority of the homicides committed by both white and minority defendants were intraracial, while most death sentences meted out to minority defendants involved interracial homicides (i.e., white victims). One study using Georgia data appeared to indicate that minority defendant/white victim homicides involved more "aggravating" factors than other homicides - they were more likely to be committed in the course of a felony, more likely to involve stranger rather than acquaintance homicides, or more likely to involve multiple victims (Heilbrun et al 1989). Because of severe design flaws in this study, though, these findings have been rejected by most researchers (see Thomson 1997). Virtually all of the other research on post-Furman death sentencing has found that the disparities result from discrimination and not from differences in case characteristics (see Baldus et al 1990, and Gross and Mauro 1989). As Thomson (1997) points out, research on discrimination in death sentencing is complicated by the fact that there are numerous decision points in the process. Police make decisions based on probable cause about whether or not to make an arrest. Prosecutors decide what charges to file, whether to offer a plea bargain or take a case to trial, and whether or not to seek a death sentence. Judges and/or juries decide the guilt or innocence of the accused and whether or not to impose death sentences on those convicted. Appeals courts evaluate the legal procedures involved and decide whether to uphold or overturn convictions and/or sentences. Finally, political authorities (usually state governors) eventually must decide whether or not an actual execution is to be carried out. Discrimination can occur at any or all of these decision points. The issue is further complicated by the fact that discrimination has usually been viewed simply as inequitable treatment directed at particular categories of defendants (i.e., minorities, men, the poor, etc.). More recent research, however, indicates that discrimination is based on characteristics of victims as well, and that the key to understanding patterns of discrimination is the race/ethnic combination of defendant and victim. Thus there are different patterns of disparities at the various decision points for white defendants accused of killing white victims, white defendants accused of killing minority victims, minority defendants accused of killing minority victims, and minority defendants accused of killing white victims (Thomson 1997). This paper represents an initial attempt to describe the basic patterns of post-Furman race/ethnic discrimination in death sentencing in the United States based on a review of empirical studies as well as some historical evidence that suggests that the patterns are deeply embedded in our legal system and culture. PATTERNS OF DISCRIMINATION Pattern 1. White defendant cases tend to be over-represented while minority defendant cases tend to be under-represented in death sentencing. Perhaps the most commonly cited evidence of death sentencing discrimination is the fact that while minorities make up less than one fourth of the adult population of the United States about half of all defendants sentenced to death in the United States are minorities. The disparity between the proportion of minorities in the general population and the proportion of minorities in the death-sentenced population is taken as evidence that minority defendants are systematically targeted for different and discriminatory treatment. The problem with this view is that the general population is the wrong basis for comparison. The proper basis for comparison is the proportion of (potentially capital) homicides committed by minority defendants. Minority homicide rates tend to be much higher than white homicide rates and, as a result, more than half of homicides are committed by minority defendants. If minority defendants commit more than half of all homicides and make up half of the death sentenced population, then there is no disparity that indicates that minority defendants in general are targets of systematic discrimination. As the Baldus study shows (Baldus et al 1990), in Georgia, for example, about one third of homicides involve white defendants while white defendants make up almost half of death sentenced defendants. More than two thirds of Georgia homicides involve minority defendants but only about half of death sentences involve minority defendants. Thus it is white defendants, not minority defendants, who are over-represented, and this is a typical pattern in death penalty jurisdictions. Pattern 2. White victim cases tend to be heavily over-represented while minority victim cases tend to be heavily under-represented in death sentencing. The main reason for over-representation of white defendants and under-representation of minority defendants is the fact that homicide tends to be an intra-racial crime (whites kill whites, blacks kill blacks, etc.) and white victim homicides are more likely to result in death sentences than minority victim cases (this is sometimes referred to as "victim-based" discrimination). In Georgia, for example, about forty percent of homicides involve white victims while more than eighty percent of death sentences result from white victim cases (Baldus et al 1990). Pattern 3. Cases in which minority defendants are accused of killing white victims are heavily over-represented while cases in which minority defendants are accused of killing minority victims are heavily under-represented in death sentencing. The classic form of racial discrimination indicated in the quote above from Raymond Fosdick's (1920) study of police departments relates to the combination of defendant and victim. The traditional targets of both the legal death penalty and lynchings were minorities accused of crimes against whites, while the much more frequent crimes by minority defendants against minority victims were routinely treated less seriously. The Georgia data (Baldus et al 1990) illustrate that this pattern still prevails as black defendant/white victim cases account for less than ten percent of homicides but almost forty percent of death sentences and almost seventy percent of executions. Minority defendant/ minority victim cases, on the other hand, account for more than sixty percent of homicides but only fourteen percent of death sentences and eight percent of executions. Pattern 4. Cases in which white defendants are accused of killing white victims are over-represented, while cases in which white defendants are accused of killing minority victims are under-represented in death sentencing. Cases in which white defendants are accused of killing white victims typically make up more than half of all death sentences in most (but not all) death penalty jurisdictions. This pattern is at least partly an artifact of the huge under-representation of minority defendant/minority victim cases. Cases in which white defendants are accused of killing minority victims rarely result in death sentences and almost never result in executions. Pattern 5. Most of the disparities originate at early points in the decision-making process (plea bargain offers, charging decisions, decisions to go to trial, decisions to seek death sentences, etc.) rather than later in the process (guilt/innocence, sentencing decisions). Several studies indicate that the proportions of cases for each defendant/victim combination emerge from the earliest decisions made by prosecutors regarding the cases. Prosecutors tend to take to trial (a prerequisite for a death sentence) a disproportionate number of minority/white cases while offering plea bargains or reducing charges in cases involving other defendant/victim combinations. Thus even non-discriminatory decisions later in the process (e.g., by juries) simply "pass on" the disparities to the next stage of the process. For example, Georgia data (Baldus et al 1990) indicates that Black/White cases are already heavily over-represented before the penalty phase of the process even starts and the proportions change very little as a result of sentencing decisions by judges or juries. Pattern 6. Most of the disparities result from discretionary decisions by prosecutors rather than decisions by sentencing agents (juries, judges) and the disparities tend to grow larger at each stage of the process where prosecutor decisions are involved. Traditionally, juries have been blamed for most death sentencing discrimination, and the literature abounds with anecdotal accounts of all-white juries and racism by jurors (and sometimes judges, prosecutors, and even defense attorneys). While this kind of discrimination does occasionally occur, it does not account for most of the disparities. Data from Georgia (Baldus et al 1990) and Robert Lilly's (date n/a) studies of executions by the US Military in the European Theater during World War II show an interesting pattern. In both sets of data, racial disparities increase at each stage in which prosecutor decisions are involved up to the sentencing phase. Sentencing agents (judges and/or juries) then add little or nothing to the disparities. Following sentencing, prosecutors typically pick up the cases again to handle appeals and processing the cases toward execution decisions, and the disparities tend to increase again. Thus most of the disparities emerge with prosecutor decisions early in the process and increase again only at the end of the process where prosecutor decisions are a key part of the outcomes. Pattern 7. The disparities do not result from differences in legally-relevant case characteristics (felony-related, multiple, child victim, etc.). Baldus et al (1990) in their study of Georgia, and Gross and Mauro (1989) in a study of eight death penalty states, used two different methods to examine what part case characteristics played in generating the disparities. Both approaches found that the disparities could not be explained by differences in case characteristics and concluded that most of the disparities resulted from discrimination rather than differences in aggravating or mitigating factors. Pattern 8. Most of the disparities result from prosecutor decisions in medium culpability cases rather than in low or high culpability cases. In death penalty jurisdictions, most high culpability homicides will be treated as death penalty cases, while very few low culpability homicides will be pursued as death penalty cases. Baldus et al (1990) found that in Georgia most of the disparities resulted from medium culpability cases and this is consistent with results from Gross and Mauro (1989) as well. This is also consistent with the view that most disparities result from prosecutor decisions, since it is medium culpability cases that allow the most prosecutor discretion in deciding whether to go to trial and seek a death sentence. Pattern 9. The basic patterns of disparities tend to be similar in most jurisdictions in the United States (e.g., state, federal, and military) although target groups differ in different jurisdictions (e.g., African Americans in the South, Hispanics in the Southwest). Gross and Mauro (1989) found similar patterns of disparities involving Black/White over-representation in eight different states, including midwestern states. Thomson (1997) found similar patterns in Arizona, where both Hispanics and Blacks accused of killing Whites were over-represented in death sentencing. Thomson (1997, unpublished paper) also found similar patterns in California where minorities accused of killing whites were heavily over-represented while minorities accused of killing other minorities were under-represented. There are numerous other studies in individual states and jurisdictions generally sgowing similar patterns. Pattern 10. The basic patterns of disparities tend to be similar for juvenile and mentally ill and mentally retarded defendants. Victor Streib, for example, (date n/a) found that the majority of juveniles sentenced to death in the 1980's were minorities, while more than eighty percent of the victims in these cases were white. Pattern 11. The basic contemporary patterns of disparities tend to be similar to historical patterns of disparities (including both legal executions and lynchings). Garfinkle's (1949) study in North Carolina found that Black/White homicides were over-represented while Black/Black homicides were under-represented in death sentencing. D.L. Abney's (date n/a) unpublished Master's Thesis on the history of the death penalty in Arizona suggests that most of the legal executions and lynchings in Arizona involved either White/White or Minority/White homicides. Tolnay & Beck (date n/a), in their key study of southern lynchings, indicate that most lynchings involved Black men accused of offending Whites. Copyright � 2009 Ernie Thomson. All rights reserved. email: [email protected] |