Voters in Michigan in November approved a proposal to ban affirmative action in the admissions process at state universities.  Similar bans have previously won approval in California and Washington State. Various student groups opposed to the ban have filed suit; critics of affirmative action are looking to mount challenges in other states. Affirmative action has always been a politically sensitive issue but it is not the only issue in achieving equality in higher education. As the legal challenges to the Michigan ban work their way through the courts and spread to other states, now is a good time to address another sensitive, and maybe even more important issue: equality in college performance.
Most colleges provide the public with very little information about racial and ethnic differences in students’ grades and graduation rates. Nor do they provide much information about the effectiveness of their diversity programs. So what should prospective minority students and their parents expect after being accepted? Unfortunately, the answer is that race and ethnicity are important predictors of college performance. Recent research confirms that white and Asian students not only enjoy pre-college advantages in family income and school quality, but on average, they also benefit throughout their college experience in ways that black and Latino students do not.
For example, in the 2001 Duke University entering class, freshman grades were on average lower among blacks and Latinos than they were for whites and Asian Americans. Black-white differences narrowed, but remained significant,  even among students with similar family structures, social class backgrounds, middle and high school characteristics, and SAT scores.
Why do these racial and ethnic disparities continue? One explanation is stereotype threat; when race or ethnicity is emphasized in academic situations, minority academic performance declines. The core argument is that minority students underperform because they are trying so hard to avoid confirming pernicious stereotypes. However, when excellence is emphasized, the stereotype threat is deactivated and racial and ethnic performance differences fade or disappear. (An excellent "Frontline" interview with Claude Steele  explores this issue.)
In addition, there is also evidence that racial and ethnic disparities in college success are due to differences in students’ social and information networks. From parents, peers, staff, and faculty, students get a range of information, such as which courses to take, and the best path to a desired career. They also learn behaviors, such as how to balance social and academic demands on their time. Students who have families with a long history of college attendance are more likely to have access to information about college, and to relevant role models. Due to historical racial disparities, differences in access to these social and information resources tend to correlate with race and ethnicity.
Although these findings may be surprising to many people, they are not news to many in higher education. We have long known that we cannot simply admit diverse cohorts and expect that there will be no group differences in college performance. For decades, colleges have conducted a range of programs designed to increase comfort, skills, and connections among minority students, and to make campuses more receptive to traditionally underrepresented groups.
At Colgate University, Breaking Bread  requires members of disparate student groups to plan, prepare, and eat a meal together. By the end of the meal, the groups must have identified a collaborative campus event. Last year, the College Republicans and the Rainbow Alliance combined to bring Andrew Sullivan, a conservative gay-rights advocate, to campus. A strength of Breaking Bread is that it uses everyday activities -- preparing and eating a meal, as an opportunity to build bridges between groups that tend to have very little to do with one another.
Another noteworthy program is the Summer Institute for Diversity and Unity at Hamilton College, where faculty members spend three days off campus engaged in discussion groups about diversity. Participants use the experience to create new course syllabi, or to revise syllabi for existing courses. Over the past three summers, nearly 20 percent of the full-time faculty at Hamilton have participated in the program. This initiative promises a substantial impact on the campus climate because diversity discussions now appear throughout the curriculum, not just in a few courses.
However, these innovative and successful diversity programs are the exception rather than the rule. One reason that colleges don’t provide more information on their diversity programs may be that the programs are not properly evaluated. Far too many programs persist today because key administrators merely believe they work or are reluctant to ask hard questions about politically sensitive programs. This is troubling because the minority students of today are substantially more diverse than minority students in the 1980s. Without rigorous assessment, we cannot know if programs designed 20 years ago are effective for today’s students, or that the programs designed today will be effective for the students of the future.
It is imperative that colleges and universities scrutinize their diversity goals, programs, and outcomes. As with affirmative action, such examination is sure to produce a number of uncomfortable confrontations. Nevertheless, colleges and universities have a responsibility to take on this challenge. The parents who trust us with their children, and the students who trust us with their futures, deserve nothing less.
David Harris is vice provost for the social sciences at Cornell University. He, along with a team from Cornell and Colgate Universities and Hamilton, Hobart and William Smith, and Wells Colleges co-authored "Eliminating Racial and Ethnic Disparities in College Completion and Achievement,"  which was commissioned by the Teagle Foundation. A podcast of Teagle Foundation President W. Robert Connor, president of the foundation, interviewing Professor Harris about the report is available here.