assault (51 percent for NCVS and 57 percent for UCR), whereas a higher proportion of black youth are reported for robbery offenses (58 percent for NCVS and 60 percent for UCR). Third, one observes an important difference in the two sources of information in the proportion of youth in the “other” race category. In the NCVS, the proportion in the other race (e.g., American Indian, Asian) category ranges from 12 to 15 percent. In contrast, in the UCR, this percentage is only 2 percent. This may reflect the tendency of the police to categorize persons as black or white, because of their record-keeping system. Thus, it is possible that some of the respondents categorized as other in the NCVS data are classified as black in the UCR, inflating the rates for black juveniles.

Comparing the findings from these two important and different data sources does not answer the question about whether police arrest black youth inappropriately or excessively. This comparison does, however, reveal the complexity of trying to disentangle juvenile crime data by race and the urgent need for more focused examination of these issues.

In a further analysis, Greenfeld (1999) presented the results of calculations of juvenile court data based on rates in 17 states³ in 1994. From that analysis, Greenfeld concluded that black and white juveniles who commit robbery have nearly the same likelihood of being arrested, convicted, and punished with confinement. In contrast, Greenfeld found that for aggravated assault, black juveniles have a one-third higher likelihood of the offense being reported to the police (the rate was 52 per 100 offenders for black youth compared with 39 per 100 for white youth), a 50 percent higher rate of being referred to juvenile courts (rates for black youth were 22 per 100 compared with 15 per 100 for white youth), and a 60 percent higher rate of getting petitioned (rate of 16 per 100 for black youth compared with 10 per 100 for white youth), and a 50 percent greater likelihood of receiving institutional placement (3 per 100 for black youth compared with 2 per 100 for white youth).

This brief summary of crime rates indicates that black juveniles are overrepresented in some types of crimes. The question is why should black juveniles be more likely to engage in criminal behavior than whites? Such overrepresentation may be at least partially explained by considering how exposure to risk factors affects the probability of engaging in criminal behavior. The argument has been made that more minority children, and black children in particular, are subject to risk factors associated with crime and that these factors explain the disparity. For example,