Individual Submission Summary
Share...

Direct link:

How Much Do Voters Care About Student Learning?

Thursday, November 5, 3:30 to 5:00pm, Property: Boston Marriott Copley Place, Floor: 4th Floor, Room: Nantucket

Abstract

One of the most consistent findings in the empirical research on school board elections is that voters appear to place little (if any) weight on student academic outcomes when casting their ballots. In the most comprehensive study to date, tracking 50,000 school board elections across 16 states, Kogan, Lavertu, and Peskowitz (2025) find that student learning rates "have only a modest association with incumbents' reelection rates." Their estimates suggest that moving a school district from the bottom 10 percent nationally to the top 10 percent in terms of achievement growth increases the odds that incumbents are reelected by only 2 percentage points.

One explanation for these patterns is that voters simply don't care about standardized test scores -- perhaps because they view such tests as too narrow measures of achievement or because they embrace a broader, multidimensional vision of what makes schools effective. Another explanation is that voters do care about achievement but that aspects of the informational environment, including ballot design, reduce the salience of academic performance at the point of decision-making. Our study seeks to examine which of these explanations is most likely to be responsible using data from two conjoint experiments embedded in nationally representative samples of the Cooperative Election study.

Our first study shows that voters do indeed care about academic achievement. In particular, we find:

1. Improving district achievement growth from the 10th to 90th percentile increases the probability that voters choose to reelect an incumbent by 20 percentage points, roughly ten times larger than the relationship observed in actual, real-world elections.

2. Academic considerations trump other documented determinants of vote choice, rivaling shared partisanship in terms of the absolute effect size.

3. Strikingly, even voters who claim to personally oppose standardized testing reward performance measured using such tests -- to the same degree as voters who explicitly support testing.

4. We find no meaningful differences in how much voters value academic achievement when we examine subgroups living in high vs. low achievement (or achievement growth) school districts, suggesting that differential prioritization of learning among the school board electorates does not explain differences in real world school quality.

In our second study, we randomly vary whether voters also receive candidate partisan information, to examine whether the growing shift toward partisan school board elections is likely to further crowd out academic quality considerations. This experiment is complete, but we are waiting for the data from YouGov to complete the analysis (expected in February 2026).

Authors