The G.O.P.’s own leaders admit that the great Southern white shift was the result of a deliberate political strategy. “Some Republicans gave up on winning the African-American vote, looking the other way or trying to benefit politically from racial polarization.” So declared Ken Mehlman, the former chairman of the Republican National Committee, speaking in 2005.

And Ronald Reagan was among the “some” who tried to benefit from racial polarization.

True, he never used explicit racial rhetoric. Neither did Richard Nixon. As Thomas and Mary Edsall put it in their classic 1991 book, “Chain Reaction: The impact of race, rights and taxes on American politics,” “Reagan paralleled Nixon’s success in constructing a politics and a strategy of governing that attacked policies targeted toward blacks and other minorities without reference to race — a conservative politics that had the effect of polarizing the electorate along racial lines.”

Thus, Reagan repeatedly told the bogus story of the Cadillac-driving welfare queen — a gross exaggeration of a minor case of welfare fraud. He never mentioned the woman’s race, but he didn’t have to.

There are many other examples of Reagan’s tacit race-baiting in the historical record. My colleague Bob Herbert described some of these examples in a recent column. Here’s one he didn’t mention: During the 1976 campaign Reagan often talked about how upset workers must be to see an able-bodied man using food stamps at the grocery store. In the South — but not in the North — the food-stamp user became a “strapping young buck” buying T-bone steaks.

Now, about the Philadelphia story: in December 1979 the Republican national committeeman from Mississippi wrote a letter urging that the party’s nominee speak at the Neshoba Country Fair, just outside the town where three civil rights workers had been murdered in 1964. It would, he wrote, help win over “George Wallace inclined voters.”

Sure enough, Reagan appeared, and declared his support for states’ rights — which everyone took to be a coded declaration of support for segregationist sentiments.

Questions of whether or not the death penalty was applied fairly along racial lines surfaced in McCleskey v. Kemp. McCleskey argued that there was racial discrimination in the application of Georgia's death penalty. As evidence for this claim, McCleskey presented the results of an extensive statistical study by Professor David Baldus of the University of Iowa Law School and his colleagues. Baldus’ study collected information about all the capital defendants in Georgia—whether or not they were sentenced to death. This information allowed the researchers to control for hundreds of variables about the offender, victim and crime—thereby permitting a statistical comparison of cases in order to see what factors influenced whether a person was sentenced to death. Professor Baldus found, among other things, that:

• Fewer than 40% of Georgia homicide cases involve white victims, but in 87% of the cases in which a death sentence is imposed the victim is white. White-victim cases are roughly eleven times more likely than black-victim cases to result in a sentence of death.
• When the race of the defendant is added to the analysis, the following pattern appears: 22% of black defendants who kill white victims are sentenced to death; 8% of white defendants who kill white victims are sentenced to death; 1% of black defendants who kill black victims are sentenced to death; and 3% of white defendants who kill black victims are sentenced to death. (Only 64 of the approximately 2500 homicide cases studied involved killings of blacks by whites, so the 3% figure in this category represents a total of two death sentences over a six-year period. Thus, the reason why a bias against black defendants is not even more apparent is that most black defendants have killed black victims; almost no cases are found of white defendants who have killed black victims; and virtually no defendant convicted of killing a black victim gets the death penalty.)
• No factor other than race explains these racial patterns. The multiple-regression analysis with the greatest explanatory power shows that after controlling for non-racial factors, murderers of white victims receive a death sentence 4.3 times more frequently than murderers of black victims. The race of the victim proves to be as good a predictor of a capital sentence as the aggravating circumstances spelled out in the Georgia statute, such as whether the defendant has a prior murder conviction or was the primary actor in the present murder.
• Only 5% of Georgia killings result in a death sentence; yet, when more than 230 non-racial variables are controlled for, the death-sentencing rate is 6% higher in white-victim cases than in black-victim cases. A murderer therefore incurs less risk of death by committing the murder in the first place than by selecting a white victim instead of a black one.
• The effects of race are not uniform across the spectrum of homicide cases. In the least aggravated cases, almost no defendants are sentenced to death; in the most aggravated cases, a high percentage of defendants are sentenced to death regardless of their race or their victim's; it is in the mid-range of cases which, as it happens, includes cases like McCleskey's that race has its greatest influence. In these mid-range cases, death sentences are imposed on 34% of the killers of white victims and 14% of the killers of black victims. In other words, twenty out of every thirty-four defendants sentenced to die for killing a white victim would not have received a death sentence if their victims had been black.

The Supreme Court held, however, that a death-sentenced defendant cannot challenge his sentence as a violation of the constitutional requirement of "equal protection of the laws" by showing that it is consistent with a system-wide pattern of racial disparity in capital sentencing. To make out an equal-protection violation, a defendant is required to prove that some specific actor or actors in his or her individual case--the prosecutor or the judge or the jurors, for example, intentionally discriminated against the defendant on the ground of race in making a decision that resulted in the death sentence.

Ways in Which Race Can Impact Capital Sentencing

Race of the Victim

Nationally, nearly 80% of murder victims in cases resulting in an execution have been white, even though nationally only 50% of murder victims generally are white. A 1990 examination of death penalty sentencing conducted by the United States General Accounting Office noted that, "In 82% of the studies [reviewed], race of the victim was found to influence the likelihood of being charged with capital murder or receiving the death penalty, i.e., those who murdered whites were found more likely to be sentenced to death than those who murdered blacks." Individual state studies have found similar disparities. In fact, race of victim disparities have been found in most death penalty states.

Race of the Defendant

Nationally, the racial composition of those on death row is 45% white, 42% black, and 10% Latino/ Latina. Of states with more than 10 people on death row, Texas (70%) and Pennsylvania (69%) have the largest percentage of minorities on death row. Year 2000 census data revealed that the racial composition of the United States was 75.1% white, 12.3% black and 12.5% Latino/Latina. While these statistics might suggest that minorities are overrepresented on death row, the same statistical studies that have found evidence of race of victim effects in capital sentencing have not conclusively found evidence of similar race of defendant effects. In fact, while some studies show that the race of the defendant is correlated with death sentences, no researcher has made definitive findings that the death sentence is being imposed on defendants on account of their race, per se, independently of other variables (such as type of crime) which are correlated with defendants' race.

Race of the Jurors

In capital cases, one juror can represent the difference between life and death. A belief that members of one race, gender, or religion might generally be less inclined to impose a death sentence can lead the prosecutor to allow as few of such jurors as possible. For example, a Dallas Morning News review of trials in that jurisdiction found systematic exclusion of blacks from juries. In a two-year study of over 100 felony cases in Dallas County, the prosecutors dismissed blacks from jury service twice as often as whites. Even when the newspaper compared similar jurors who had expressed opinions about the criminal justice system (a reason that prosecutors had given for the elimination of jurors, claiming that race was not a factor), black jurors were excused at a much higher rate than whites. Of jurors who said that either they or someone close to them had had a bad experience with the police or the courts, prosecutors struck 100% of the blacks, but only 39% of the whites.