Inferring and Explaining

126 InferrIng and exPlaInIng the defendant’s record and the severity of the crime. When all the data were considered, the study concluded that murderers of white victims were 4.3 times as likely to receive the death penalty. Justice Brennan expressed this correlation in characteristically vivid language. At some point in this case, Warren McCleskey doubt- less asked his lawyer whether the jury was likely to sentence him to die. A candid reply to this question would have to tell McCleskey that few of the details of the crime or of McCleskey’s past criminal record were more important than the fact that his victim was white. Furthermore, counsel would feel bound to tell McCleskey that defendants charged with kill- ing white victims in Georgia are 4.3 times as likely to be sentenced to die as defendants charged with kill blacks. 5 I have discussed the McCleskey case with hun- dreds of students in the last several years. Many simply refuse to accept the following data. e 11 . When controlled for over two hundred non-racial variables such as the defendant’s record and the severity of the crime, the Baldus study concluded that murderers of white victims were 4.3 times as likely to receive the death penalty. It is, of course, true that life in the inner city is diferent from life in the suburbs and that black culture is in many ways diferent from white culture. Te shocking fgure that more than four times as manymurderers of whites receive the death penalty takes all that into account. I know some of you will continue to believe that “statistics always lie.” But the very same techniques that tell us that cigarette smoking causes cancer or that so-and-so will win next month’s election tell us that the connection between race and the death penalty in Georgia is for real. Tus the question before us is pro- ducing an explanation of why this correlation holds. Tere is no big mystery about the reason for this disparity. Te original study contained the crucial data. e 12 . District attorneys ask for a capital sen- tence in 70 percent of the cases involving a black defendant and a white victim. When the victim is black and the defendant is white, however, a mere 19 percent are even prosecuted as capital cases. Inone sense, theBaldus study’sdatabase isnot a sample at all but an analysis of the entire popu- lation of homicides in Georgia from the time the state rewrote its aggravated murder statute in response to Furman v. Georgia to the conclusion of the study in 1979. For the purposes of the Gregg trial, this was ideal, since it was the laws and behavior of legal ofcials in Georgia that were at issue. But Justice Blackmun, and certainly yours truly, believe that capital punishment, in general, is discriminatory. We can treat the Baldus study as telling us something about the death penalty in this country. t ′ 0 . Capital punishment in the United States is administered in a racially discriminatory manner. Such an explanation of the data in the Baldus study immediately invites two rival explanations that raise very diferent issues of bias.