Do Legacy Preference Bans in College Admissions Increase the Racial and Socioeconomic Diversity of Enrolled Students?

Evans and Christensen (2025) ask whether banning legacy preferences in college admissions changes the racial and socioeconomic composition of enrolled students, and why effects differ across institutions. They analyze policy documents and enrollment data from seven case studies where legacy preferences were eliminated, including public systems and selective private colleges. Using before-and-after comparisons and difference-in-differences style estimates, they find mixed results. Some institutions saw modest declines in White enrollment (about 4–5 percentage points) and small increases in Hispanic or Asian enrollment, while most saw little change in Black enrollment. Gains in low-income enrollment were limited and inconsistent, suggesting context strongly conditions policy effects.

Why This Article Was Selected for The Policy Scientist

This article addresses a policy issue of broad and growing importance: how admissions rules shape access to selective higher education at a moment when traditional diversity tools are legally constrained. Legacy preferences sit at the intersection of equity, institutional finance, and political legitimacy, making their regulation consequential well beyond individual campuses. The timing is particularly salient following recent court decisions limiting race-conscious admissions, as policymakers increasingly look to formally race-neutral mechanisms to influence enrollment composition. Evans and Christensen (2025), who have contributed extensively to research on higher education policy and enrollment management, advance the literature by systematically documenting the heterogeneous policy landscape and pairing it with cross-institutional enrollment evidence.

Full Citation and Link to Article

Evans, B. J., & Christensen, C. L. (2025). The evolving landscape of legacy preference bans in postsecondary admissions: Evidence and policy implications from case studies. Education Finance and Policy. https://doi.org/10.1162/EDFP.a.433 

Central Research Question

This article asks whether eliminating legacy preferences in college admissions alters the racial and socioeconomic composition of enrolled students, and why the effects of such bans vary across institutional and policy contexts. Rather than assuming a uniform response to the removal of legacy preferences, the authors focus on understanding heterogeneity: under what conditions do bans lead to observable changes in enrollment, and when do they appear largely symbolic? The study also examines how differences in policy design—legislative versus voluntary bans, scope of coverage, and enforcement mechanisms—shape institutional responses. At a deeper level, the article interrogates whether legacy preference bans function as an effective substitute for race-conscious admissions policies in a post–affirmative action legal environment.

Previous Literature

The study builds on a well-established literature documenting the prevalence and magnitude of legacy preferences in selective college admissions. Prior research shows that legacy applicants are disproportionately White and high-income and that legacy status confers large admission advantages at elite institutions. Widely cited work by Espenshade et al., Hurwitz, and more recent studies using Harvard admissions data demonstrate that legacy preferences significantly raise admission probabilities even after accounting for academic credentials. Other strands of the literature examine the institutional rationale for legacy preferences, including yield management, alumni relations, and fundraising, though empirical evidence on donation effects remains mixed.

What has been largely missing from the literature is systematic evidence on what happens after legacy preferences are removed. Existing discussions are often normative or speculative, with limited empirical grounding. Some case-specific analyses, such as those focusing on individual universities, suggest potential gains in racial diversity, but these findings are not easily generalized. This article contributes by shifting the focus from whether legacy preferences matter—which is well established—to whether banning them meaningfully reshapes enrollment outcomes across different institutional settings.

Data

The authors assemble multiple data sources tailored to the descriptive goals of the study. First, they compile original data on federal and state legislative efforts to ban legacy preferences, drawing from congressional records, state legislative archives, and publicly available policy documents through early 2025. These data capture variation in policy scope, definitions of legacy status, inclusion of donor preferences, and enforcement mechanisms.

Second, the authors identify colleges that voluntarily eliminated legacy preferences using public announcements, press releases, and secondary reporting. Institutional characteristics are drawn from IPEDS and Barron’s Competitiveness Index, allowing comparisons between voluntary legacy enders, institutions affected by legislative bans, and other selective colleges.

Finally, the core outcome data consist of fall enrollment characteristics of first-time, full-time freshmen, including race, ethnicity, and receipt of federal grant aid as a proxy for socioeconomic status. These data are available for multiple years before and after policy changes in seven focal case studies. While the data do not include applicant-level admissions outcomes, they are appropriate for examining changes in the enrolled student body, which is the policy-relevant endpoint.

Methods

The analysis relies on structured descriptive comparisons rather than formal causal identification. For each case study, the authors compare average enrollment characteristics in the years before and after the legacy preference ban. To partially adjust for secular trends, they implement a difference-in-differences–style approach by subtracting contemporaneous changes observed in comparison institutions that did not alter their legacy policies.

The authors are explicit about the limitations of this approach. The estimates are not interpreted as causal due to potential confounding from concurrent policy changes, such as shifts in financial aid, recruitment strategies, or test-optional admissions. Additionally, the analysis cannot disentangle changes in admissions decisions from changes in enrollment yield across groups. Pre-trend comparisons are presented to assess plausibility, but the authors stop short of making strong identification claims. Methodologically, the approach prioritizes transparency and contextual interpretation over statistical sophistication.

Findings/Size Effects

The results reveal substantial heterogeneity across cases. In several settings, eliminating legacy preferences produced little to no change in the racial or socioeconomic composition of incoming freshmen. This pattern is most evident in contexts where admissions rates were high, legacy populations were small, or institutions faced binding constraints on the use of race in admissions.

In contrast, a small number of institutions experienced measurable changes. At Texas A&M University, the share of White freshmen declined by roughly 4–5 percentage points after the ban, accompanied by increases in Hispanic enrollment and modest gains in students from low-income households. At Johns Hopkins University and the University of Pittsburgh, White enrollment declined while Asian enrollment increased substantially, with smaller changes for Hispanic students and minimal changes for Black students. Across cases, increases in Black enrollment were limited or nonexistent.

Socioeconomic effects were similarly uneven. Only two of the seven case studies showed clear increases in the share of students receiving federal grant aid, and even these gains were modest in magnitude. In most settings, displaced legacy students appeared to be replaced by non-legacy students with similar socioeconomic profiles, limiting changes in income diversity.

The authors emphasize that these outcomes are consistent with structural constraints: the size and composition of the legacy pool, institutional selectivity, applicant volume, financial capacity, and state policy environments all condition the effect of legacy bans. The findings suggest that the maximum possible impact of banning legacy preferences is often mechanically small, particularly when legacy students comprise a limited share of the entering class.

Conclusion

This article provides a careful and policy-relevant assessment of legacy preference bans at a moment of heightened interest in admissions reform. Its central contribution lies in demonstrating that eliminating legacy preferences does not reliably produce large or uniform changes in student diversity. Instead, outcomes depend heavily on institutional context and complementary policies. The study advances the literature by pairing a systematic mapping of the policy landscape with comparative enrollment evidence across multiple settings.

While the descriptive methods are appropriate for the available data, the analysis would be strengthened in future work by the use of more explicit causal inference designs, particularly those leveraging applicant-level data or policy discontinuities. Nonetheless, the article clarifies the limits of legacy preference bans as a standalone tool and underscores the importance of institutional and policy context in shaping admissions outcomes.