Do Tuition-Free College Programs Increase Voter Turnout and Affect Election Outcomes?

This study asks whether large-scale tuition subsidies increase political participation. Using 16.4 million FAFSA records linked to California voter files and a regression discontinuity design, the authors examine the Cal Grant, the largest tuition-free four-year college program in the U.S. They find that each of the 2.6 million awards from 2010–2019 increased recipients’ voter turnout by 4–12 percentage points in 2020, raising statewide turnout by about 1 percentage point and Biden’s margin by 0.5 points. One in every 66 voters participated because of the subsidy, with effects concentrated among Democrats and independents, suggesting significant civic externalities from education spending .

Full Citation and Link to Article

Firoozi, Daniel, and Igor Geyn. 2025. “Do Tuition Subsidies Raise Political Participation?” February 19, 2025. Available at https://www.aeaweb.org/articles?id=10.1257/pol.20240422&from=f

Extended Summary

Central Research Question

The paper investigates whether large-scale tuition subsidies, such as the California Cal Grant program, significantly increase political participation, particularly voter turnout, and whether these civic effects are large enough to influence electoral outcomes. The study also examines the mechanisms behind such changes and whether the effects generalize to other major financial aid programs, including the federal Pell Grant. The authors are especially interested in the potential partisan implications of these programs, given evidence that college-educated youth lean toward the political left in the United States.

Previous Literature

The work builds on a longstanding body of research linking education to civic engagement. Classical economists like Adam Smith and Milton Friedman, as well as historical policymakers, justified public education on the grounds of producing informed, politically active citizens. Prior empirical studies generally find positive associations between educational attainment and political participation, including voting, political knowledge, and civic involvement. However, most causal evidence comes from compulsory schooling laws or small-scale programs, often at the primary or secondary level. Results for secondary education are mixed, and there is little credible evidence on the civic externalities of higher education subsidies. Existing studies of tuition subsidies have largely focused on labor market outcomes, such as graduation rates and earnings, rather than political behavior. While some work using instruments like college proximity or military conscription has explored the enrollment–civic participation link, results have varied. This paper addresses a gap by using large-scale administrative data and a strong identification strategy to estimate the causal impact of a major tuition subsidy on voting.

Data

The authors link 16.4 million Free Application for Federal Student Aid (FAFSA) records from 2010–2011 through 2020–2021 to the L2 California voter file, which contains detailed information on registration status, party affiliation, and election participation for approximately 21 million registered California voters. The FAFSA data, provided by the California Student Aid Commission (CSAC), include information on income, assets, Cal Grant receipt, GPA, housing plans, and ZIP code of origin. The sample focuses on in-state residents without bachelor’s degrees who meet asset thresholds and whose family incomes lie near the program’s income eligibility ceilings for Cal Grant A awards between 2017–2018 and 2019–2020. Additional analyses extend to earlier cohorts and to a second program, the Pell Grant, using 2.5 million student records.

Methods

The study employs a fuzzy regression discontinuity design (RDD) exploiting the income eligibility thresholds for Cal Grant A awards. The running variable is family income, standardized to the relevant threshold for each applicant’s cohort and family type. The RDD compares students just below the income cutoff (eligible) to those just above (ineligible), under the assumption that these groups are otherwise similar in both observable and unobservable characteristics. Robustness checks include McCrary density tests to detect manipulation, covariate balance tests, and placebo threshold tests. The preferred specification focuses on 2017–2019 FAFSA cohorts to minimize biases from pre-2017 threshold visibility and to reduce attrition from out-of-state moves. The authors estimate reduced-form effects of eligibility, then instrument Cal Grant receipt with eligibility to obtain causal effects on political participation. They also examine heterogeneity by demographics, GPA, and political environment, and explore mechanisms such as peer socialization, voter reciprocity, civic trust, and income effects.

Findings/Size Effects

The primary finding is that receiving a Cal Grant increases voter turnout substantially. For the 2017–2019 cohorts, turnout in the 2020 general election rose by about 10 percentage points from a baseline of 56 percent, implying that the 2.6 million grants awarded over the 2010s generated roughly 259,000 additional votes statewide—a 1 percentage point increase in total turnout. The entire increase is among Democrats and independents; there is no detectable effect for Republicans. The program also increased cumulative turnout across all post-treatment elections by similar margins. Using conservative assumptions, the authors estimate that the Cal Grant raised Biden’s 2020 margin in California by about 168,000 votes (0.5 percentage points), meaning one in every 66 voters and one in every 55 Biden voters participated because of the program.

Results for earlier cohorts (2010–2019) show smaller but still significant increases in turnout, around 4–7 percentage points. Effects appear to peak in recent years, coinciding with increased program generosity and improved administration. Null effects for the 2020–2021 cohort, whose instruction was largely remote during the COVID-19 pandemic, suggest that in-person socialization on campus is a key mechanism.

The Pell Grant analysis supports external validity. A notch in Pell Grant generosity yields turnout increases similar in per-dollar terms to the Cal Grant, with over 80 percent of the effect among Democrats and independents. The Pell Grant effects are notable given that the additional aid at the notch is roughly one-tenth the value of a Cal Grant.

Mechanism analysis indicates that peer socialization—especially from living on campus—plays a central role. Cal Grant receipt increases on-campus living rates by about 17 percentage points and reduces living with family by about 9 points. This aligns with evidence that turnout effects appear soon after award receipt, vanish during remote learning, and are not fully explained by changes in enrollment patterns. Reciprocity, civic trust, and income effects may contribute but are less consistent with the data. Reciprocity is undermined by the timing of largest effects (2020, when state offices were not on the ballot), and income effects are inconsistent with the lack of rightward partisan shifts and the similar per-dollar impacts across income levels.

Conclusion

The study concludes that large-scale tuition subsidies can generate substantial civic externalities, significantly increasing political participation among young, college-educated voters, particularly those with higher GPAs. These effects are concentrated among left-leaning voters and can be large enough to influence election outcomes, even in a large state like California. The Cal Grant’s estimated impact in 2020—259,000 additional votes and a 0.5 percentage point boost to the Democratic margin—illustrates the magnitude of these effects under conservative assumptions.

The findings raise important questions for the political economy of education finance. If subsidies disproportionately mobilize voters favorable to one party, partisan policymakers may have incentives to adjust funding levels for electoral advantage, regardless of labor market returns or overall social welfare. The results suggest that civic externalities should be included in cost–benefit analyses of higher education spending, alongside earnings impacts, and that policy designs enabling in-person, on-campus engagement may maximize these civic benefits. The authors emphasize that the Cal Grant’s effects occurred even without measurable labor market gains for recipients, underscoring the distinct and substantial civic value of such programs.

The paper’s broader contribution is to extend the literature on education’s civic returns beyond enrollment and attainment, using robust causal methods and large administrative datasets. By demonstrating that tuition subsidies can mobilize substantial numbers of new voters, the study provides both empirical evidence for the longstanding civic rationale for public education funding and a cautionary note about its partisan implications. The external validity analysis with the Pell Grant strengthens the case that these effects are not unique to California’s context or the Cal Grant’s design, but may generalize to other large-scale tuition subsidy programs.