This paper, via The Monkey Cage, is striking. Two political scientists, Mark Peffley and Jon Hurwitz, polled a bunch of white people with this question:

“Do you favor or oppose the death penalty for persons convicted of murder?”

Somewhat favor: 29%
Strongly favor: 36%

Okay, that's about what you'd expect. But then they polled another random set of white people with a slight variant:

“Some people say that the death penalty is unfair because most of the people who are executed are African-Americans. Do you favor or oppose the death penalty for persons convicted of murder?”

Somewhat favor: 25%
Strongly favor: 52%

No, that's not a typo. White respondents were more likely to strongly favor the death penalty after they were told that it's mostly African Americans being executed. (Not surprisingly, black respondents were far less eager to support the death penalty after learning about the racial disparity.)

Anyway, rather than chalk these results up to, say, genocidal tendencies among white folks, the researchers dove into a bunch of survey data and decided on this explanation: White survey subjects, they found, were more likely to emphasize personal failings as a primary cause of crime, rather than focusing on situational factors—poverty, inequality in the justice system, etc. So they're more likely to believe that, if black people are being disproportionately punished, it's because they "deserve" it, and hence, they reject any suggestion of unfairness so strongly—because it's inconsistent with their prior beliefs—that they run in the other direction.

I can't tell if that's the best explanation here, but either way, that's the result. Also, note that the death penalty debate has shifted terrain in the past decade. It used to be that opponents would argue over the morality of the death penalty, or appeal to the Constitution, but nowadays—especially after the rash of DNA exonerations that began in the early '90s—they tend to focus more on the risk of executing innocent people, as well as the disparate racial impact.

Those arguments actually seem to have worked somewhat—support for the death penalty nudged downward from 84 percent in 1994 to 66 percent in 2000—but Peffley and Hurwitz's work suggests that white enthusiasm for the death penalty isn't likely to change much in response to these new tactics. As noted, the racial disparity argument may, in some cases, even increase white support. What a world, eh?