Karen F. Parker is an associate professor of sociology in the Center for Studies in Criminology and Law and the Department of Sociology at the University of Florida. Her primary research interests include exploring the impact of racial stratification and urban economic restructuring on disaggregated violence. Mari A. DeWees is a doctoral candidate in the Department of Sociology at the University of Florida. Her current research interests focus on the social and legal aspects of violence, specifically the effect of structural conditions on gender-specific criminal involvement and the mobilization of law in cases of domestic violence. Michael L. Radelet is a professor in the Department of Sociology at the University of Colorado–Boulder. During the past two decades, he has testified in 60 capital cases and worked with 50 of the last 54 men and women executed in Florida.

Racial discrimination has long been recognized as a part of the American criminal justice system. Studies of racial bias have tended to focus on the flaws and failures within the legal process itself, rather than on extralegal factors that may also be involved. But just as the literature has definitively established that it is true that racial discrimination affects everything from arrest to voir dire to sentencing, so is it true that racial discrimination has origins and explanations outside of the legal system that contribute to the failures within that system. In this article, we investigate and expand upon our previous efforts to offer a structural explanation of racial discrimination’s impact upon the legal system. (See Karen F. Parker et al., Racial Bias and Conviction of the Innocent, in Wrongly Convicted: When Justice Fails 114 (Saundra Westervelt & John Humphrey eds. 2001).) To do so, we specifically examine the nexus among wrongful convictions, capital cases, and Hubert Blalock’s power-threat hypothesis. Further, we provide a snapshot of the changing perceptions of the American public on issues of wrongful conviction and capital punishment, suggesting that these changes may well affect the political and judicial actions that will influence racial discrimination within our justice system in years to come.

Such an investigation necessarily begins with reference to the literature on these subjects. Surprisingly, very few scholarly works have examined the role of race in the wrongful convictions despite ample evidence of racial disparities across multiple stages of the criminal justice system. (See Cassia Spohn et al., The Impact of the Ethnicity and Gender of Defendants on the Decision to Reject or Dismiss Felony Charges, 25 Criminology 175 (1987); John Kramer & Darrell Steffensmeier, Race and Imprisonment Decisions, 34 Soc. Q. 357 (1993); Michael Tonry, Malign Neglect (1995); Samuel Walker et al., The Color of Justice: Race, Ethnicity, and Crime in America (1996).) Although these studies are few in number, they have served to establish the existence of racial disparities in cases of wrongful convictions.

One landmark study is that of Bedau and Radelet in 1987, for which the most extensive database of cases of wrongful conviction in American history was assembled. (Hugo Adam Bedau & Michael L. Radelet, Miscarriages of Justice in Potentially Capital Cases, 40 Stan. L. Rev. 21 (1987).) They compiled information on 350 cases of wrongful convictions—326 homicides and 24 rapes—across American jurisdictions from 1900 to 1985. Bedau and Radelet found that among the 350 cases, approximately 43 percent involved black defendants. This finding led them to conclude that "the risk of a miscarriage of justice falls disproportionately on blacks when compared to their representation in the population, but not in comparison to their arrest rates." (Id. at 39.)

In their 1992 book, In Spite of Innocence, Radelet et al. updated the sample of defendants thought to be wrongfully convicted in capital or potentially capital cases to include 416 cases. They also supplied narrative-style depictions of several cases in which race was clearly an influential factor in the conviction of the innocent. (Michael L. Radelet et al., In Spite of Innocence 119–38 (1992).) They found that blacks were more likely than whites to be wrongfully convicted of capital crimes and that the probability of wrongful conviction of blacks occurs disproportionately to their representation in the U.S. population. This research presented significant evidence of miscarriages of justice; raised questions concerning the risk of executing the innocent, because 23 cases in the study involved defendants who were executed despite doubts about their guilt; and demonstrated the importance of race as an important correlate to wrongful convictions.

Two more recent studies broaden the connections between wrongful convictions and race. Both of these investigations focused on erroneous convictions for any type of felony, rather than rape and homicide cases alone. In the first, Huff et al. interviewed judges, prosecutors, public defenders, and exonerated inmates in order to construct a database of wrongful convictions. (C. Ronald Huff et al., Convicted But Innocent: Wrongful Conviction and Public Policy (1996).) They eventually established 205 erroneous convictions for various felonies. While precise racial classification of the defendants in those cases was not reported, the authors found evidence of racial disparities: "Many convicted innocents are white, some are even middle-class, but a disproportionate number are black or Hispanic." (Id. at 80.) In the second study, Scheck et al. examined 62 cases in which prisoners have been cleared by DNA evidence. (Barry Scheck et al., Actual Innocence (2000).) Of the 62 cases, 29 percent of the defendants were white and 57 percent were black. (Id. at 267.) Like the preceding studies, they found that miscarriages of justice and wrongful conviction fall disproportionately on blacks as compared to other racial and ethnic groups.

Another recent paper, by Holmes in 2001, offers an empirical analysis that examines race and other social characteristics among those wrongfully convicted of capital crimes. (William M. Holmes, Who Are the Wrongly Convicted on Death Row? in Wrongly Convicted: When Justice Fails 99 (Saundra Westervelt & John Humphrey eds., 2001).) Holmes’s study focused on cases of wrongful conviction where the capital cases were "overturned by appellate courts and sent back to the trial court, only to have the charges dropped by the prosecutor or the defendant found not guilty on retrial." (Id. at 102.) Specifically, he investigates several extralegal factors related to conviction, where the "wrongly convicted" include not only innocent defendants but also undoubtedly guilty defendants whose convictions were later reversed by appellate courts because of a wide array of due process errors. Based on 6,228 cases collected by the U.S. Bureau of Justice Statistics on prisoners on death row from 1970 to 1992—a figure that includes all capital prisoners in state and federal prisons during that time—Holmes examines characteristics such as race, ethnicity (that is, Hispanic background), sex, marital status, education, and the most serious offense of those wrongfully convicted. In our view, his most significant and striking findings involve race and ethnicity. In his initial analysis, he found distinct racial differences among those who had been wrongfully convicted. However, after further statistical testing that controlled for other study variables, he noted that strict racial differences did not account for wrongful convictions. Nonetheless, he found that ethnicity (specifically, Hispanic background), education level, and the seriousness of the offense were still all associated with rates of wrongful convictions, and he concluded that these factors ultimately predict a likelihood of who faces a greater risk of wrongful conviction in capital cases. Holmes’s study thus suggests that ethnicity may be an important consideration in addition to race and other social characteristics in cases of wrongful convictions.

Let us now focus on two studies of death row inmates who were released from prison because of evidence that they were innocent. The Death Penalty Information Center (DPIC) provides race and other characteristics of inmates released from death row on their website. (See www.deathpenaltyinfo.org.) Using data derived from the DPIC, Radelet et al. compiled extensive information on 68 cases involving men and women who were released from U.S. death rows since 1970 due to doubts about their guilt. (Michael L. Radelet et al., Prisoners Released from Death Rows Since 1970 Because of Doubts About Their Guilt, 13/3 T.M. Cooley L. Rev. 907 (1996).) They found that 45.6 percent of them were white, 41.2 percent were black, and 13.2 were other minorities. (Id. at 963.)

In 2001, Harmon analyzed 76 cases of inmates released from death row from 1970 to1998. (Talia Roitberg Harmon, Guilty Until Proven Innocent: An Analysis of Post-Furman Capital Errors, 12 Crim. Just. Pol’y Rev. 113 (2001).) Sixty-eight of Harmon’s cases replicate those used in the 1996 work of Radelet et al. In addition to court records, Harmon presented information gained through questionnaires sent to "key informants" for each case. These included defense attorneys (the most common source), journalists, and law professors who were familiar with the facts of the cases. Race was among the factors found to be associated with wrongful conviction. For example, according to court records, 58 percent of the defendants who had been erroneously convicted were minorities, while only 10 percent of the victims were minorities. Furthermore, Batson violations—i.e., those involving racially motivated peremptory challenges—were alleged by the defense in 29 percent of the sample cases, yet the appellate courts found such violations in only 1 percent of the cases. (Id.; see Batson v. Kentucky, 476 U.S. 79 (1986).) Key informants suggested that racial bias was an influential factor in 34 percent of the cases. (Harmon, supra, at 126.) Interestingly, these informants perceived more racial discrimination in the cases than was found by the courts. Harmon concluded that this disparity may be due to misjudgments on the part of the key informants regarding the impact of racial discrimination or, more likely, to the "reluctance of the courts to admit that serious race problems exist within the criminal justice system . . . [which] can cause public outrage and seriously undermine the legal system." (Id.)

These two studies establish both an increase in the number of documented cases of innocence among death row inmates and a racial gap among cases of the wrongfully convicted. In 2001 the present authors published a paper that depicted the disadvantages faced by minorities in the criminal justice system and offered explanations for minorities’ disproportionate high risk of being erroneously convicted. (Karen F. Parker et al., Racial Bias and Conviction of the Innocent, in Wrongly Convicted: When Justice Fails 114 (Saundra Westervelt & John Humphrey eds., 2001).) This study moved beyond existing research by proposing both individual and structural explanations for the racial disparities in the conviction of the innocent. As individual explanations, we proposed (1) overt and intentional racist behavior; (2) the increased probability of errors when making eyewitness identifications across races; (3) stereotyping of minorities by whites, i.e., that minorities are more likely to conform to a criminal stereotype and thus be convicted on weaker evidence; and (4) the lack of resources minorities are able to access, which may make them "easy targets." As structural explanations, we proferred (1) Blalock’s power-threat hypothesis and (2) arguments concerning the concentration of disadvantage among minority groups as compared to whites in urban areas. Overall, we claimed that the racial bias in wrongful convictions was present at both the individual and structural levels, indicating that race continues to be an influential component in the operations of the criminal justice system and our society.

Here we wish to build on our earlier claims concerning the racial disparities in the conviction of the innocent by providing empirical data and analysis. We also wish to expand our discussion of the importance of examining wrongful conviction in the larger context of race in our society and criminal justice system. In doing so, we first delineate the relationship between perceived minority threats to whites, the death penalty, and wrongful convictions. Subsequently, we provide an empirical analysis of the linkages between race, region, and wrongful convictions in death penalty cases. Before concluding, we discuss recent changes in public opinion and the political arena that may influence future studies on the relationship between race and wrongful conviction.

We believe that Hubert Blalock’s power-threat hypothesis, first formulated in 1967, provides an avenue for understanding the link between race and wrongful convictions. (H.M. Blalock, Jr., Status Inconsistency, Social Mobility, Status Integration and Structural Effects, 32 Am. Soc. Rev. 790 (1967).) The theory claims that as the relative size of the minority group increases, members of the majority group perceive a growing threat to their position and will take steps to reduce the threat of competition. Blalock argued that the relationship between minority percentages and the motivation of the majority group to discriminate assumes two forms: power threats and competition over economic resources. Regarding political threat, as the relative size of the black population increases, Blalock postulated that the state will increasingly perceive blacks as a threat to white political power and thus will intensify social control to maintain the dominant position of whites. Economic threat occurs as blacks compete for jobs, positions, and economic resources, and as they increasingly become a threat to the economic well-being of whites. As a result, social control efforts can be manipulated by whites to exclude blacks from participating in the economic sphere, which reduces the economic competition they pose for whites.

In previous tests of the power-threat hypothesis, researchers have found that the size of the minority population is significantly related to the increase of black lynchings between 1889 and 1931 in the South. (See Jay Corzine et al., Black Concentration and Lynching in the South: Testing Blalock’s Power-Threat Hypothesis, 61 Soc. Forces 774 (1983); Jay Corzine et al., The Tenant Labor Market and Lynching in the South: A Test of Split Labor Market Theory, 58 Soc. Inquiry 261 (1988); Stewart E. Tolnay et al., Black Lynchings: The Power Threat Hypothesis Revisited, 67 Soc. Forces 605 (1989).) Others report support for this hypothesis in terms of the treatment of blacks in the criminal justice system. (Martha A. Myers, Black Threat and Incarceration in Postbellum Georgia, 69 Soc. Forces 373 (1990); Stewart E. Tolnay et al., Black Competition and White Vengeance: Legal Execution of Blacks as Social Control in the Cotton South, 1890 to 1929, 73 Soc. Sci. Q. 627 (1992); J. Yates, Racial Incarceration Disparity Among States, 78 Soc. Sci. Q. 1001 (1997).) For example, when considering the availability of poor black labor for employers after the Civil War, Myers found that declines in labor posed a specific threat to cotton-producing landowners, which resulted in higher rates of incarceration imposed on both black and white males. Tolnay et al. found a link between economic competition and the frequency of legal executions of blacks in the South from 1890 to 1909 and 1910 to 1929. Overall, these studies provide support for the power-threat hypothesis and suggest that, as whites perceive blacks as a threat to their dominant position in areas with a high proportion of blacks, the likelihood of state control and exercise of its power increase.

Using data from the DPIC website, we have explored minority threat as a possible explanation for the persistent and growing racial bias in cases of wrongful conviction. Based on Blalock’s thesis, we predict that whites perceive blacks as a threat to their power, which fosters high rates of erroneous convictions among blacks more so than among other racial and ethnic groups. And, because of the larger presence of blacks in the South—where 19 percent of the population is black, as opposed to about 12 percent in the entire United States, and where the region has been marked historically by racial discrimination—wrongful convictions of blacks will be higher there than in other regions of the country.

Table 1 (see page 53) displays the racial characteristics of the 107 inmates exonerated from death row since 1973, as well as descriptive state and regional information where the miscarriage of justice occurred. While 41 states had the death penalty in 1973 (nine did not), the number of states rejecting the death penalty had grown to 14 by 1991. Today 12 states have rejected the death penalty. This fluctuation in the number of states implementing the death penalty over the time period we studied affects the data presented in Table 1. Thus we present information on 107 death row inmates who were wrongfully convicted across 25 states that hosted legalized executions.

Of these exonerated inmates, 45 percent were identified as black, 42 percent as white, and 13 percent as other racial or ethnic backgrounds. (See www.deathpenaltyinfo.org.) In comparison, blacks are estimated to make up approximately 12 percent of the U.S. population and about 19 percent of the population in the South. (U.S. Bureau of the Census, Census of Population and Housing, 2000.) In terms of regional differences, we find that more inmates are exonerated and freed from death row in the South than in other regions of the country, that is, 57 percent versus 43 percent.

Because we are also interested in whether blacks in the South are exonerated from death row more often than other race and ethnic groups in this region, Table 1 displays statistical information that allows us to compare across the racial groups by regional location (South versus non-South). For example, we estimate the number of blacks exonerated in the South at 63 percent. This statistic is calculated by dividing the number of blacks exonerated in the South (30) by the total number of blacks exonerated (48). By calculating the percentages in this way, we can make the proper statistical comparison across racial groups separately for each regional area. When taking into account the race of the death row inmates wrongfully convicted by region of the country, disparities become even more striking. As mentioned, blacks make up approximately 63 percent of those wrongfully convicted. They are followed by other race/ethnic groups (57 percent), and whites (51 percent). In all other regions combined, whites are more likely to be among those wrongfully convicted (49 percent), as compared to other race/ethnic groups (43 percent) and then blacks (37 percent).

We therefore find support for the claims that blacks and other minority groups are more likely to be wrongfully convicted disproportionate to their size in the population and in the South more so than in other regions. That is, the regional differences between the South and other areas of the country are larger for blacks than other racial and ethnic groups. Given the unique history of racial discrimination in the South, as well as the vast racial inequalities in political and economic opportunities found in the South relative to other regions (see Stewart E. Tolnay et al., Black Lynchings: The Power Threat Hypothesis Revisited, 67 Soc. Forces 605 (1989); Stewart E. Tolnay et al., Black Competition and White Vengeance: Legal Execution of Blacks as Social Control in the Cotton South, 1890 to 1929, 73 Soc. Sci. Q. 627 (1992); Jay Corzine et al., The Tenant Labor Market and Lynching in the South: A Test of Split Labor Market Theory, 58 Soc. Inquiry 261 (1988)), the observed racial differences in wrongful conviction of death row inmates are consistent with Blalock’s claims of political threat.

Our findings support and document the applicability of Blalock’s power-threat hypothesis. Specifically, this hypothesis contributes to our understanding of the racial disparities in wrongful convictions. The ability to execute a person is the most significant indicator of state power and control. As a result of discrimination, particularly in terms of economic and political displacement, minority groups (in this case, blacks) are often unable to protect themselves and are viewed as a threat to whites as their presence in the population increases. In this study, we find that a higher percentage of blacks face wrongful conviction, particularly in the South.

In recent years, evidence indicates that more Americans have come to believe that (1) racial disparities threaten the fair application of the death penalty, and (2) as long as America retains the death penalty, innocent people will be sent to death row (even if not executed). Although no public opinion polls probed these issues before the late 1990s, a 1985 Gallup Poll found that only 15 percent of those opposing the death penalty cited erroneous convictions as their main concern, and only 3 percent of its opponents rested their positions on concerns with fairness in its application. (Support for Death Penalty Highest in Half-Century, Gallup Reports, Nos. 232, 233 (Jan./Feb. 1985) at 3.) These changing tides of American opinion may well influence the rate of wrongful convictions in the future.

For example, a national opinion poll taken in May 2002 by the Gallup Organization found that 40 percent of respondents thought the death penalty was applied unfairly, while 53 percent thought it was applied fairly, and 7 percent had no opinion. (See www.pollingreport.com/crime.) An ABC News/The Washington Post poll from April 2001 found that 63 percent strongly or somewhat agreed that "the death penalty is unfair because it’s applied differently from county to county and state to state." Thirty-seven percent agreed that the death penalty is unfair "because it’s applied unequally to blacks compared to whites." The respondents, by a 51–43 margin, also supported halting all executions while a special commission studies whether or not it is applied fairly. Previously, an NBC News/Wall Street Journal poll in 2000 found that respondents were split evenly, 42–42, in their opinions of whether the death penalty is applied fairly. (Id.)

Innocence also is a concern for the American public today. A July 2001 Harris Poll found that 93 percent of Americans believed that innocent people are "sometimes" convicted of murder; only 3 percent said that this never happens. (Id.) In the April 2001 ABC News/The Washington Post poll mentioned above, 68 percent either strongly or somewhat agreed that "the death penalty is unfair because sometimes an innocent person is executed." A June 2000 Gallup Poll commissioned by CNN found that 80 percent of the respondents believe that within the five previous years "a person has been executed who was, in fact, innocent of the crime he or she was charged with." (Id.)

Expressions of public concern seem to have translated into political caution. Since January 2000, two states have declared moratoria on executions while questions about the fairness of their application could be studied. The first was declared in Illinois in January 2000 because of repeated instances where innocent defendants had been sentenced to death. A study commission was established and research conducted that found widespread racial disparities in the death penalty in Illinois. (See Glenn L. Pierce & Michael L. Radelet, Race, Region, and Death Sentencing in Illinois, 1988–1997, 81 Or. L. Rev. 39 (2002).) After reviewing the materials compiled by that commission, shortly before he left office in 2003 Governor George Ryan issued complete pardons to four death row inmates because of innocence and commuted the sentences of 167 others to terms of life imprisonment. The second moratorium was declared in Maryland in 2002 because of evidence of racial disparities. That state had also sent innocent defendants to death row. (Michael L. Radelet et al., Prisoners Released from Death Rows Since 1970 Because of Doubts About Their Guilt, 13/3 T.M. Cooley L. Rev. 907, 926–27 (1996).) However, when newly elected Governor Robert Ehrlich took office in early 2003, he canceled the moratorium and announced plans to once again proceed with executions.

Two recent U.S. Supreme Court decisions can be expected to have significant impact on the issue of racial disparities and the conviction of the innocent. On June 20, 2002, the Supreme Court reversed a 13-year-old decision and ruled that executing the mentally retarded offended evolving standards of decency, in violation of the Eighth Amendment. (Atkins v. Virginia, 122 S. Ct. 2242 (2002).) Four days later, the Court invalidated portions of the Arizona death penalty statute, ruling that juries, rather than judges, should determine whether a defendant receives a death sentence. (Ring v. Arizona, 122 S. Ct. 2428 (2002).) These two decisions suggest that the Supreme Court itself is narrowing the scope of the death penalty, another impact of changing American public opinion.

The Ring decision was based on the Sixth Amendment right to a trial by a jury of one’s peers. Arizona and eight other states allowed death sentences to be imposed by a single judge—sometimes with an advisory sentence by a jury—or by a panel of three judges. The same decision makers also acted as fact finders in determining the aggravating and mitigating circumstances. One criticism of these statutes has been that judges are not representative of the communities in which they serve; they tend to be older, richer, better educated, and more likely to be white and male. (Michael L. Radelet, Rejecting the Jury: The Imposition of the Death Penalty in Florida, 18 U.C. Davis L. Rev. 1409, 1424–26 (1985).) Of course, death penalty juries are not truly representative of their communities either, because all those who stand solidly opposed to the death penalty are not permitted to serve. However, they are more representative than is the judiciary. In 1985, for example, Radelet found that of the 341 Florida circuit judges, only 21 were women and eight were black. Arguably, the increased diversity of juries may make them more receptive to claims of innocence raised by minority defendants (especially if buttressed by testimony from other minorities), resulting in fewer convictions and/or death sentences for innocent defendants.

Atkins is likely to be the more important decision. This case not only reduced the range of cases in which capital punishment can be applied, but it did so by using, at least in part, public opinion data. At issue was a determination of evolving standards of decency and how these evolving standards might render some previously tolerated punishments to be cruel and unusual. The 6–3 decision came in the case of Daryl Renard Atkins, a Virginia prisoner with an I.Q. of 59, who was sentenced to death for a robbery-murder. The Justices in the majority noted that several groups of experts opposed the death penalty for the mentally retarded and noted the moral opposition to such executions by major religious groups. (122 S. Ct. 2242, 2249 n.21.) They also acknowledged the overwhelming opposition to such executions by other countries throughout the world. Finally, they took note of approximately 20 public opinion polls, which, though "no means dispositive," nonetheless showed widespread distaste for executing the mentally retarded (Id. at 2050 n.21).

In 1958 the Supreme Court had ruled that the cruel and unusual clause of the Eighth Amendment "must draw its meaning from the evolving standards of decency that mark the progress of a maturing society." (Trop v. Dulles, 356 U.S. 86, 101 (1958).) This ruling set the stage for a periodic reexamination of public opinion to determine if specific aspects of capital punishment—e.g., methods of execution, mandatory death sentences, or imposition of death sentences on specific groups—offend evolving standards of decency. One of the most significant aspects of Atkins is the receptiveness of the Court to such a wide range of data in evaluating these evolving standards of decency.

Public sentiments, such as those noted above, may one day be used by future courts to evaluate whether the public’s concerns about injustices in the application of the death penalty, based on racial disparities and on growing recognition of the inevitability of executing the innocent, rise to a level that suggests the punishment may violate the Eighth Amendment. Although many Americans voice theoretical support for the death penalty, that does not mean that they support it as applied.

Our study provides evidence of a racial bias in the operations of the criminal justice system, particularly in regard to the death penalty. As in the previous studies reviewed here, we find racial disparities in wrongful convictions among death row cases. Moreover, we also find that cultural, political, and economic differences among regions of the country may well factor into the relationship between race and wrongful convictions. Specifically, we posit that blacks are more likely to face wrongful conviction in the South than whites and other racial and ethnic groups. Additional research is undoubtedly needed to better understand the nexus between race and wrongful conviction within the larger structural context, and we urge scholars to consider regional differences that produce racial discrimination and unequal treatment under the law.

Return to Table of Contents - Spring 2003

Return to Criminal Justice magazine home page