Study Under Review: Kogan, V., Lavertu, S., & Peskowitz, Z. (2021). How does minority political representation affect school district administration and student outcomes? American Journal of Political Science, 65(3), 699-716.
For a full explanation of the PEARLS framework used in this review, see our [methodology page](/methodology.html).
The study draws on California school districts as its unit of analysis, using school board election results to construct treatment and comparison groups. The selection mechanism is the regression discontinuity design itself: districts where a minority candidate won by a narrow margin form the treatment group, while those where the same candidate barely lost serve as the counterfactual. This approach does not require a random sample from the broader population of districts, but it does restrict the sample to a specific subset, those with competitive close elections involving minority candidates. The abstract does not report the exact number of districts included, which limits transparency about sample size. The California focus raises immediate questions about generalizability: the state's district demographics, funding structures, and election laws differ meaningfully from other states. Readers should treat the findings as applicable to districts resembling those in the sample, primarily urban and suburban California districts with competitive minority candidacies, rather than to all U.S. school districts. The demographic composition of included districts is not fully detailed in available summaries, though the focus on minority representation implies variation in racial and ethnic composition.
The study uses a regression discontinuity design, which is a quasi-experimental method that exploits the arbitrary cutoff in close elections to estimate causal effects. When executed carefully, this design can isolate the impact of winning office from confounding factors that might also explain outcomes, because candidates who barely win and barely lose are assumed to be similar on all pre-election characteristics. The authors examine effects over up to six post-election years, which allows assessment of cumulative rather than immediate impacts. A meaningful design choice given that policy changes may take time to affect student outcomes. The study compares districts where a minority candidate barely won against those where the candidate barely lost, leveraging variation induced by the election threshold. This approach is stronger than cross-sectional correlation designs because it addresses selection bias: districts that elect minority representatives may differ systematically from those that do not, but the discontinuity approach controls for those differences by construction. The authors do not report randomization or blinding, which is standard in observational designs but worth noting. The design's validity depends on the assumption that candidates and voters cannot precisely manipulate the vote margin around the cutoff, a standard assumption in the literature that appears reasonable here.
The researchers analyze their data by comparing outcomes in treatment and control districts across multiple years, focusing on student achievement, district spending, administrative composition, and several intermediate outcomes. The reported finding of approximately 0.1 standard deviation gains in minority student achievement by the sixth post-election year is a cumulative estimate, which matters because single-year effects may be smaller or inconsistent. The achievement finding is notable for its symmetry: gains for minority students appear without corresponding losses for white students, which addresses a common concern in representation debates. The analysis also finds increases in capital funding and non-white principal representation, suggesting mechanisms through which representation might translate to outcomes. However, the study finds no effects on segregation, English language learner reclassification, or teacher staffing, which limits the scope of the mechanisms tested. The authors appear to assert causal effects based on their design, which the regression discontinuity approach supports when the assumptions hold. The interpretation is largely consistent with the results, though readers should note that the six-year window captures only medium-term effects and that longer-term outcomes remain unexamined.
The study was published in 2021, drawing on California data from school board elections. The policy context around minority representation in school governance has remained active since then, with ongoing debates about school board diversity, equity in resource allocation, and the role of elected boards in district outcomes. For practitioners in California or similar states with elected school boards, the findings offer directly applicable evidence. For practitioners in states where school boards are appointed or where district structures differ substantially, applicability is more limited. The study's focus on capital funding and principal composition speaks to concrete governance levers, though the lack of effects on staffing and segregation narrows the scope of what the evidence supports. The six-year follow-up window is both a strength and a limitation: it captures effects that persist beyond immediate post-election periods, but it does not address long-term trajectories that may diverge further.
The study appears in the American Journal of Political Science, a leading peer-reviewed journal in political science, which carries institutional credibility and suggests peer review was rigorous. The EdWorkingPapers series context indicates the working paper circulated publicly before journal publication, allowing community scrutiny. The funding sources for this specific study are not detailed in available summaries, though the authors' institutional affiliations, Ohio State University and the University of California, are research universities with standard academic funding structures. No conflicts of interest are apparent from the available information, and the study's design does not appear to favor a particular policy outcome. Readers should verify funding details in the full article if this is a concern for their context, but the publication venue and author affiliations suggest standard academic funding without obvious ideological or financial stake in the findings.
The study investigates multiple outcome domains (student achievement, capital funding, principal composition, segregation, ELL reclassification, and teacher staffing) which provides breadth. However, the focus on California only limits geographic generalizability, and the restriction to close-election districts raises questions about whether the findings apply to districts with less competitive politics or different demographic profiles. The study acknowledges that mechanisms are not fully tested: while capital funding and principal composition appear to move, the causal chain from representation to these intermediate outcomes to student achievement is not fully traced. The corroborating evidence from Brett's working paper on school boards and education finance points in the same direction, greater Hispanic representation associated with improved funding and achievement at high-Hispanic schools, which suggests the California finding is not isolated. However, Rouse et al. found no systematic impact of Latino representation on student outcomes despite increases in Latino administrators, which partially contradicts the downstream achievement finding. This discrepancy may reflect differences in methodology, time period, or district characteristics, but it indicates that the relationship is not yet settled in the literature. The study is comprehensive within its stated scope but does not resolve questions about mechanisms or generalizability beyond California.
Practitioners can reasonably conclude that electing minority representatives to school boards in competitive elections produces measurable gains in minority student achievement over several years without harming white students, at least in California-like contexts. The finding that capital funding and non-white principal representation increase under minority representation suggests that board composition influences resource allocation and staffing decisions, which are governance levers boards can actually control. The absence of effects on segregation, ELL reclassification, and teacher staffing indicates that not all equity-oriented outcomes respond to board representation, which calibrates expectations about what governance change can achieve alone.
Practitioners should not conclude that minority representation automatically brings the academic outcomes of minority students in line with those of other students or that the effect size found here will replicate in all districts. The 0.1 standard deviation gain over six years is meaningful but modest, and the single-state focus limits direct application to other contexts without adaptation. Practitioners should also not infer that board representation alone is sufficient. The study identifies correlates, not the full set of conditions required for gains.
Actionable takeaways: First, boards seeking equity-oriented outcomes should attend to how representation translates to capital budget decisions, where the study found clear effects. Second, the absence of staffing effects suggests that hiring and staffing policies may require different interventions or longer timelines to change. Third, districts outside California should treat these findings as suggestive rather than definitive, pending replication in their own state contexts. Given the regression discontinuity design's causal credibility but the single-state sample, practitioners should give this study substantial weight for its core finding while remaining cautious about generalizing the specific effect sizes or mechanisms to different settings.
Found a study on school board governance? Paste the DOI below and we'll add it to the PEARLS review queue.
We'll evaluate it using the PEARLS framework and post the review when complete.