The 24 male and 32 female subjects included an equal number of black and white students at a small university in the United States. An equal number of black and white students was randomly assigned to receive one of four crime descriptions that varied in terms of defendant race (black or white) and the type of crime (burglary or embezzlement) committed. Burglary was chosen as a blue-collar crime and embezzlement as a white-collar crime. The authors hypothesized that a black defendant who committed a blue-collar crime would be judged more harshly than a white defendant accused of the same crime. It also was predicted that a white defendant accused of committing a white-collar crime would be judged more harshly than a black defendant accused of the same crime. Students were asked to rank the perceived severity of the crime on a 9-point scale ranging from moderately severe to extremely severe. Subjects also were asked to indicate the likelihood that the defendant would repeat the crime, and subsequently to recommend jail sentences and bail amounts. As hypothesized, the white embezzler received longer jail sentences than the black embezzler, and the black burglar received longer jail sentences than the white one. 15 references. (Author abstract modified)