Most contemporary sociological theories of crime predict that blacks will be overrepresented among arrestees in common law personal crimes. These theories differ, however, in the extent to which this overrepresentation is attributed to disproportionate involvement in criminal offenses vs. criminal justice system selection biases. Studies that have relied upon official data have generally supported the differential involvement hypothesis, whereas studies relying on self- report techniques generally have supported the differential selection hypothesis. National victimization survey data on victims' reports of racialcharacteristics of offenders are introduced as a third measurement technique in order to shed additional light on this controversy. These data for rape, robbery, and assault, are generally consistent with official data on arrestees and support the differential involvement hypothesis. Some evidence of differential selection for criminal justice processing is found; however, most ofthe racial disprportionality in arrest data is shown by victimization survey data to be attributable to the substantially greater involvement of blacks in the common law personal crimes of rape, robbery, and assault. These results suggest that traditional admonitions against using arrest data as an index of involvement in these crimes may be overly cautious. In fact, the results imply that more caution should attend the use of self-report data in this vein and that more attention should be given to sampling and instrument concerns in self-report techniques. As currently used, the method may not be adequate for assessing the correlates of serious illegal conduct. The results also suggest that research emphasis be placed on those theories, such as the subcultural and differential opportunity perspectives, which attempt to explain differential racial involvement in these common law personal crimes. (RACE AND INVOLVEMENT IN COMMON LAW PERSONAL CRIMES, MICHAEL J. HINDELANG, American Sociological Review 1978, Vol. 43 February: 93-109.)
For those who would like a translation from sociologese to English, Hindelang found that data from victims gathered by research interviewers matched arrest data; in other words, there was no targeting of blacks. The victim data did not match self-report data very well, however, indicating that blacks were underreporting their crimes. In a later study reported in Measuring Delinquency, Hindeling asked teens if they had ever been taken in by police, then checked law enforcement records to see if they were lying. All demographic groups were honest except black males.
By the way, did you notice that this guy is (actually was--he died young) a liberal? He suggests that we look at differential opportunity theories to explain why blacks are so violent. Since the system denies them equal opportunities, they get really angry and shoot at their neighbors. This guy was no conservative who went into this research with the goal of getting the system off the hook. The same can be said for most researchers who find no bias.