In December 2000, a U.S. District Court judge ruled that the University of Michigan could provide preference in admission policies to minority students. He relied partly on expert social science testimony, which concluded that such policies advance racial and ethnic diversity and improve the education of all students, not just the minority students admitted under the policy. Shortly thereafter, however, another District Court ruled that preferences by the University of Michigan law school, designed to achieve the same goal, did not pass Constitutional muster. In his decision, the judge accepted some "facts" about the possible beneficial consequences of diversity, but he expressed doubt as to whether such facts "trumped" constitutional notions of equal protection. He was later overruled. It is unclear how much of a role social science evidence will play in the forthcoming Supreme Court decision that is expected to address these inconsistencies. However, key philanthropic organizations such as the Ford Foundation, believing that such evidence will play some role, are continuing to finance major survey research projects exploring the possible benefits of diversity for all students in colleges and universities. In this article, the authors outline the historical and legal context of this debate, showing how social science data gradually became a major factor in it. Then they describe the procedures and findings of their survey, and discuss their implications for the current debate.