The Supreme Court upheld the concept of affirmative action, but issued a challenge to administrators and scholars: in order for affirmative action to remain a viable admissions strategy, they must show “that no workable race-neutral alternatives would produce the educational benefits of diversity” (Fisher v. the University of Texas, 2013, p. 11).

The results of simulations in this paper suggest at least three important patterns. First, race-based affirmative action is consistently more effective than SES-based affirmative action at producing racial diversity in selective colleges. Socioeconomic-based affirmative action results in only modest gains in racial diversity. Our simulations suggest that unless SES-based affirmative action policies use a very high, probably untenable, weighting for lower-resource students, these policies are unlikely to result in the same racial composition in colleges as under current race-based affirmative action policies. Second, our models suggest that socioeconomic affirmative action results in considerable economic diversity in selective colleges. In contrast, race-based affirmative action alone yields relatively little socioeconomic diversity. One concern that may limit colleges’ use of socioeconomic-based affirmative action, however, is that it necessarily increases in the enrollment of students from the bottom of the socioeconomic distribution, and so it may carry a heavy cost in terms of financial aid (a factor that is not included in our stylized models).

Finally, it is clear that students’ knowledge (or lack thereof) regarding their probability of getting into selective colleges plays a large, and perhaps previously unrecognized, role in the sorting of minority students into colleges. All of our affirmative action policies were much more effective at increasing diversity when students knew which specific schools employed affirmative action practices and could use this information when estimating their likelihood of admission to these schools. If colleges wish to use affirmative action policies to encourage a diverse pool of applicants and enrollees, they will be more successful when those policies are made public because students respond to additional information that helps them more effectively allocate their applications.

The correlation between race and socioeconomic status is not strong enough for one to be used as an effective proxy for the other during the admissions process. Therefore, the mix of racial and socioeconomic affirmative action policies that a college employs will depend in part on how heavily it weighs racial and socioeconomic diversity. Schools that highly value racial diversity can obtain it most effectively with race-based affirmative action strategies; in order to obtain socioeconomic diversity, colleges must employ socioeconomic affirmative action. The pertinent question facing colleges is thus what kinds of diversity do we wish to see in our student body? Based on thoughtful answers to that question and a clear assessment of the costs associated with affirmative action policies, colleges can design admissions policies that are best suited to their goals.