Racial bias is small or non-existent in death penalty sentencing disparities

The disparity in death penalty sentencing could also be consistent with racial gaps in very violent crimes too. And it is possible that police encounters are not racially biased, while death penalty sentence is, and vice versa. Blacks have been over-represented relative to their proportion of the general population by 3:1 in terms of executions. This is almost always taken by some academics and the media as evidence that the death penalty is racially biased. The disproportionality argument is repeated without giving any consideration to the logic behind it. This argument fits with the ideological views of death penalty opponents, including those of mine. I’m against it as well. This disproportionality in the percentages of each race executed must be compared not with their proportion of the general population, but with the percentage of each race eligible to be included in those sub-populations (the correct target population is each race’s proportion of murderers with its proportion executed). Black homicide rate averages 12.7 times higher than the white rate. Blacks are over-represented in every homicide category, ranging from 66.7% of drug-related homicides to 27.2% of workplace homicides [1]. The Radford University’s Serial Killer Information Center finds that blacks have been 57.9% of serial killers in the U.S. from 2000 to 2014; whites 34%, Hispanics 7.9%, and Asians 0.06% [2]. And: “Although they are over-represented among death row populations and executions relative to their share of the U.S. population, blacks are underrepresented based on their arrests and convictions for murder [3]”. In California, Nevada, and Utah, blacks were overrepresented on death row relative to their proportion of murders in those states, but that in all other 28 states they were underrepresented. In Nevada the percentage of black murder offenders during that 22-year period was 30.2%, and the percentage of blacks on death row was 33.1%. In Utah the respective percentages were 8.6%and 10.5, and in California they were 33.8%and 35.3%. Blacks are overrepresented in proportion to the proportion of murders they commit in these three states. These differences are minuscule compared to the remaining 28 states, where they were underrepresented to a large degree [4]. It seems from these data that white murderers are proportionately more likely to both receive a death sentence and to be executed for death-eligible homicide. A 2001 U.S. Justice Department (2001) study of federal death-eligible cases reached a similar conclusion in federal murder cases: “United States Attorneys recommended the death penalty in smaller proportions of cases involving Black or Hispanic defendants than in those involving White defendants; the Attorney General’s capital case review committee likewise recommended the death penalty in smaller proportions of involving Black or Hispanic defendants…. In the cases considered by the Attorney General, the Attorney General decided to seek the death penalty for 38% of the White defendants, 25% of the Black defendants, and 20% of the Hispanic defendants. Why we should see a disproportional number of white murderers receiving the death penalty relative to blacks is a question yet to be addressed in any strong systematic fashion. The 2006 Rand study didn’t find racial bias (https://www.rand.org/news/press/2006/07/17.html). And the 2000 U.S. Department of Justice didn’t find it too (https://www.justice.gov/archive/dag/pubdoc/deathpenaltystudy.htm).

With all the evidence now showing that black murder defendants are less likely to be sentenced to death than white murder defendants, the race issue has become victim- centered rather than defendant-centered. The literature consistently shows that killers of whites (regardless of the killer’s race) are more likely to receive the death penalty than killers of other racial groups. I think that is your goal with these stats. However, the issue of victim-centered racial disparities in death penalty is still riddled with moral panic and divided along ideological lines. Because homicide is overwhelmingly intraracial, this is the primary reason that whites are more likely to receive the death penalty. Prosecutorial reluctance to seek the death penalty for blacks might be revealing a devaluation of black victims vis-à-vis their white counterparts, and a perception that black-on-black crime is not a threat to the white power structure.

The famous Baldus et al. study did find that “The odds of a defendant who killed a white victim are 4.125 times greater than the odds of receiving a death for a defendant who killed a black victim.” The real difference is the difference between a considerably less shocking 99% and 96%, if you one doesn’t confuse probabilities with odds, or the ratio between the respective odds represented by their probabilities.

In the Mauro and Gross (1989) datum, 28.8 percentage of blacks who killed whites under felony circumstances received a death sentence versus 6 percentage of blacks who killed other blacks under similar circumstances. Whites who killed blacks under felony circumstances received a death sentence 18.2% of the time [6]. Thus whites who kill blacks are more likely to receive the death penalty than blacks who kill blacks, although this must be viewed in light of the fact that whites only commit about 5% of interracial murders [3]. However, concern about race-of-victim bias motivated the National Institute of Justice to commission three independent teams to examine the role of race in the application of the death penalty in federal cases [7]. And they did find that death penalty-eligible violent crimes committed by black are specially heinous and violent: “When we look at the raw data and make no adjustment for case characteristics, we find the large race effects noted previously — namely, a decision to seek the death penalty is more likely to occur when the defendants are White and when the victims are White. However, these disparities disappear when the data coded from the AG’s [Attorney General’s] case files are used to adjust for the heinousness of the crime. For instance, [one of the studies] concluded, “On balance, there seems to be no evidence in these data of systematic racial effects that apply on the average to the full set of cases we studied.” The other two teams reached the same conclusion.[One team] found that, with their models, after controlling for the tally of aggravating and mitigating factors, and district, there was no evidence of a race effect. This was true whether we examined race of victim alone . . . or race of defendant and the interaction between victim and defendant race.”[the third study’s author] reported that his “analysis found no evidence of racial bias in either USAO [U.S. Attorney’s Office] recommendations or the AG decisions to seek the death penalty.”

A recent study by Sharma et al. (2013) looked at all first-degree murder convictions in Tennessee from 1976 to 2007 [8]. They noted that prosecutors sought the death penalty for 76% of white defendants and 62.6% of black defendants, and that 37.3% of white defendants for whom the death penalty was sought received it versus 23.6% of black defendants. Prosecutors sought the death penalty in 64% of the cases where the victim was white, and in 33% of the cases where the victim was black. When controlling for a variety of aggravating and mitigating factors, as well as demographic, and evidentiary variables, they found that the killing of a law enforcement officer or child, a history of prior violent offenses, and all evidentiary (scientific, co-perpetrator testimony, confession, and strong eye witness testimony) variables were by far the strongest predictors of receiving a death sentence for defendants of any race. Their most important finding was that victim’s race did not play any significant independent part, but victim’s sex (female) did. The racial make-up of the crime (black offender/white victim; white offender/white victim, etc.) had no significant independent effect, but race of the defendant did, with whites being significantly more likely to receive the death penalty than blacks over the 30-year period. Specifically about female victims of black attackers, there are interesting data on homicide and rape and racial differentials. During a 45-year period, blacks were arrested for rape an average of 6.52 times more often than whites. More interesting than the black/white ratio of rape arrests is the fact that although the great majority of homicides are intraracial, because of the frequency with which blacks choose white victims (up to 55 percent according to some accounts), there is evidence that rape is an interracial crime [9][11]. The main common explanation for this fact is that the racial composition of a city, representing the pool of rape victims or offenders of a particular race, and the degree of black-white residential segregation are significant predictors of the racial patterning of rape [10]. Felson and South victimization survey in 26 U.S. cities found that blacks chose white victims 41.3 percent of the time. The number of white offender/black victim rapes was approximately 1%. They also found that black rapists in cities with populations that are about 50% percent black, a white women is 2.4 times more likely to be raped by a black man than a white man. Felson and South do some tests and they think a “macrostrutuctural opportunity model”, the available pool of offender and level of segregation, and the opportunities of contact, explain these differentials and the intraracial aspect of this crime against whites. I find this explanation plausible, but elusive. Their own datum shows that in a 50/50 split in racial composition (or 50/50 in the “available pool” of rape victims), white women were still 2.4 times more likely to be raped by a black man than by a white man.

Baldus looked at a number of mitigating and aggravating contexts and argued that with zero or one aggravating factor in a murder case there was no racial discrimination regardless of the racial makeup of the victim/offender dyad. Similarly, with multiple aggravating factors (such as a prior homicide conviction, multiple victims, child victims, torture, and so forth), there was no discrimination and the risk of a death sentence was high regardless of the racial makeup of the victim/offender dyad [5]. It was at the middle range of aggravating circumstances where the correct sentence is less clear that racial disparities appear and here is where the famous 4.3 odds ratio came from.

In an analysis of the same Georgia data used by Baldus, Katz (2005) examined all aggravating and mitigating circumstances [12]. In the 1,082 homicide defendant sample, 141 cases involved a white victim and a black perpetrator. In 67.1% of white victim-black perpetrator (W/B) cases the victim was killed in the course of a robbery compared to 7.4% in black victim cases, and in 70.6% of the W/B cases the victim was a stranger compared with 9.6% of black victim cases. Katz (2005) also indicated that “White victim homicides show a greater percentage of mutilations, execution style murders, tortures, and beaten victims, features which generally aggravate homicide and increase the likelihood of a death sentence” (p. 405). Katz (2005) cites a number of other studies finding similar results in 10 different states; that is, once the full array of aggravating and mitigating factors are considered there is little or no discrimination evident in white victim-black perpetrator cases that is not accounted for by aggravating circumstances and other legally relevant variables. Cassell (2008) explains the nature of the victim-offender relationship: “Black-defendant-kills-white-victim cases more often involve the murder of a law enforcement officer, kidnapping and rape, mutilation, execution-style killing, and torture — all quintessential aggravating factors — than do other combinations [13]”.

Paternoster and Brame (2003) looked at 1,130 death-eligible homicide cases in Maryland and found no race-of-defendant bias but significant race-of-victim effects. That is, cases in which the offender was black and the victim was white were about 3.5 times more likely than any other offender/victim dyad to receive a death sentence [14]. On the other hand, statisticians Berk, Li, and Hickman (2005)looked at the same data and arrived at the opposite conclusion: “When race surfaces, cases with a Black defendant and White victim or ‘other’ racial combinations are less likely to have death sentences imposed”. Berk, Li and Hickman (2005) conclude that: “For both capital charges and death sentences, race either played no role or a small one that is difficult to specify [15]”.

Jennings et al. (2014) estimated the white victim effect using traditional statistical (logit regression) models controlling for 50 legal and non-legal factors and found that “the “’White victim effect’ on capital punishment decision-making is better considered a ‘case effect’ rather than a ‘race effect’” [16]. In other words, each case is unique in that it contains a multitude of case characteristics (aggravators and mitigators) and evidentiary qualities that have to be considered. Given the ability to case-match (albeit, imperfectly), this quasi-experimental approach is currently the best that we have to tease out any discriminatory effects that may be present in sentencing.

Another study of 1,163 capital cases in North Carolina found an additional complication (Bjerregaard et al., 2015).This study found that blacks who killed whites in “high severity” circumstances (multiple aggravating “heinous, atrocious, and cruel” factors) had an increased probability of a death sentence, but blacks who killed whites at lower levels of severity (a shooting death in the process of a botched robbery, for example) had decreased probabilities of receiving the death penalty [17]. Under low severity circumstances, a black defendant who kills a white victim had almost half the odds of receiving a death sentence than a black defendant who kills a black victim or a white defendant who kills either a white or black victim. This is an extremely odd finding that is difficult to explain, and the authors did not try to.What is revealing is that the average number of aggravating factors in black defendant/white victim cases(23.4% of the cases) was 2.20 while the average in white defendant/black victim cases (3.6% of the cases) was 1.47 factors (p. < .01). The average number of aggravating factors in black/black cases (30.9% of cases)was 1.96, and for white/white (42.1% of cases) it was 1.84.This supports the earlier Katz (2005) study that black defendant/white victim cases are more likely to have multiple aggravating factors. It is true that white victim cases are more likely to draw the death penalty for black and white defendants alike, but there may be valid legal reasons why this is so. The latest studies using more sophisticated quasi-experimental methods (propensity score matching) find that white victim cases tend to have many more aggravating and fewer mitigating circumstances involved regardless of the race of the defendant. It is these “case characteristics” rather than “race characteristics” that may account for the white victim effect in racial death sentencing gap, and not simply racial bias.

[1] Fox, J. A., Levin, J. A., & Quinet, K. (2011). The will to kill: Making sense of senseless murder. Pearson Higher Ed.

[2] Aamodt, M. G. (2015). Serial killer statistics. Retrieved February, 24, 2016.

[3] Robinson, M. B. (2008). Death nation: The experts explain American capital punishment. Prentice Hall.

[4] Blume, J., Eisenberg, T., & Wells, M. T. (2004). Explaining death row’s population and racial composition. Journal of Empirical Legal Studies.

[5] Baldus, D. C., Pulaski, C., & Woodworth, G. (1983). Comparative review of death sentences: An empirical study of the Georgia experience. The Journal of Criminal Law and Criminology.

[6] Hauhart, R. C. (1991). Death and discrimination: Racial disparities in capital sentencing by Samuel R. Gross and Robert Mauro: Northeastern University Press.

[7] Klein, S. P., Berk, R. A., & Hickman, L. J. (2006). Race and the decision to seek the death penalty in federal cases. Rand Corporation.

[8] Sharma, H., & Scheb, J. M. (2013). Race and the Death Penalty: An Empirical Assessment of First Degree Murder Convictions in Tennessee after Gregg v. Georgia. Tenn. J. Race Gender & Social Justice.

[9] LaFree, G. D. (1982). Male power and female victimization: Toward a theory of interracial rape. American Journal of Sociology.

[10] South, S. J., & Felson, R. B. (1990). The racial patterning of rape. Social Forces.

[11] O’Brien, R. M. (1987). The interracial nature of violent crimes: A reexamination. American Journal of Sociology.

[12] Scheidegger, K. (2012). Rebutting the myths about race and the death penalty.

[13] Cassell, P. (2008). In defense of the death penalty. Journal of the Institute for the Advancement of Criminal Justice.

[14] Paternoster, R., Brame, R., Bacon, S., Ditchfield, A., Beckman, K., & Frederique, N. (2003). An empirical analysis of Maryland’s death sentencing system with respect to the influence of race and legal jurisdiction. University of Maryland.

[15] Berk, R., Li, A., and Hickman, L. (2005). Statistical difficulties in determining the role of race in capital cases: a re-analysis of data from the state of Maryland. Journal of Quantitative Criminology.

[16] Jennings, W. G., Richards, T. N., Smith, M. D., Bjerregaard, B., & Fogel, S. J. (2014). A critical examination of the “White victim effect” and death penalty decision-making from a propensity score matching approach: The North Carolina experience. Journal of Criminal Justice.

[17] Bjerregaard, B., M. Smith, J. Cochran, J. & S. Fogel, S. J. (2015). A further examination of the Liberation Hypothesis in capital murder trials.Crime & Delinquency.