There is no doubt that part of the group differences reported in chapter two are associated with differences in the distribution of individual char acteristics and preferences between non-Hispanic whites and minorities. To take just one example, it is well known that personal earnings tend to increase with age. It is also true that the propensity toward self-employment increases with age. Since most minority populations in the U.S. have a lower median age than the non-Hispanic white population, is it possible that the disparities documented in the last chapter are largelyor even entirely-due to differences in the age distribution of minorities compared to non-minorities? Or due to differences in other factors, say, marital status, that are probably unrelated to discrimination? In other words, can the large racial and ethnic disparities observed in self-employment rates and in the earnings of the self-employed be attributed to discrimination or can they be accounted for by differing individual traits among minority and non-minority business owners and by the ordinary processes of market resource allocation? The Croson and Adarand decisions have raised questions of this sort-questions thought by many to be long settled with respect to employment discrimination against wage and salary workers-to renewed levels of legal and policy impor tance. This chapter is an attempt to provide some renewed answers to these questions.