I didn't comment much on the Duke Lacrosse team "rape" fiasco, and in retrospect I'm glad I didn't. But this data is more salient than a cable news marathon:
If you go to the website reporting the annual National Crime Victimization Survey, as many people do, you can look up the rape statistics in "Table 42: Personal Crimes of Violence 2005: Percent distribution of single offender victimizations, based on race of victims, by type of crime and perceived race of offender."
Under "Rape/Sexual assaults" the survey reports 111,490 rape/assaults in 2005 in which a white was the victim. The "perceived race" of the offender was reported as white in 44.5 percent of cases, black in 33.6 percent of cases, "other" in 19.6 percent of cases.
Where the victim of rape was black, in 36,620 cases, things were rather different. The "perceived race" of the offender was reported as black in 100.0 percent of cases. White offenders? "0.0*" percent.
The asterisk means that the sample included ten or fewer reports. The federal crime statistics show that white-on-black rape was almost non-existent in the United States of America in 2005.