In states as diverse as West Virginia, Florida, and Iowa, decisions regarding the creation, expansion, and possible extinction of charter schools rest with the elected school boards of traditional districts. Such arrangements arise from an entrenched doctrine of local control, which many take to mean that only elected boards can run schools, despite much evidence to the contrary. These boards can also curb enrollment in charters, so as to prevent students from having educational choices beyond their zoned district schools. While it is difficult to examine empirically the impact of school board composition on charter schools, an interesting new working paper makes a novel effort to do so.
School board composition is endogenously determined via the electoral process—meaning, for instance, that factors like district teacher salaries or student performance may influence who runs for office and whom voters ultimately elect. By leveraging a well-established empirical phenomenon called the ballot order effect—which has repeatedly shown that the candidate listed at the top of the ballot gains an electoral advantage—two economists from Syracuse University and the University of Rochester sought to overcome these methodological hurdles. Taking advantage of California’s randomized ballot ordering process, they estimated the causal effects of an additional educator elected to the board (conditional on the share of educators in the candidate pool) on various outcomes, including total enrollment in charter schools within the district, the number of district-authorized charter schools, teacher salaries, student achievement, and high school graduation.
They constructed school board candidate rosters for California spanning nearly two decades, which include each candidate’s vote shares, their ballot position, electoral outcome, and occupational background—which they use to categorize them as educators or not. They merge those records with school district characteristics and other dependent variables, including teacher salaries and student outcomes. Their sample includes over 14,000 candidates in California school board races between 1998 and 2015. Descriptively, 16 percent of candidates were educators and the average share of educators on each school board was 18 percent.
There are four key findings. First, an educator randomly assigned to the top of the ballot increases the number of educators elected by 0.14 individuals—an effect that equates to a 26 percent increase—which obviously shifts the composition of the overall board. Second, an additional educator elected to the board reduces charter school enrollment beginning two years post-election, with the effect peaking at four years post-election; whereas the number of charters decreases starting one year after the focal election, and continues linearly through six post-treatment years. Each additional educator elected to the board causes a 3-percentage-point decline in the charter enrollment share and 1.3 fewer charter schools in the district four years after the election. Third, the top ballot assignment leads to increases in certified teacher salaries (about 2 percent on average), which rises over time and shows a simultaneous shift away from capital outlay costs. Presumably reductions in the latter help to fund bumps in pay. Fourth, there are no impacts on high school graduation rates, but small decreases in reading scores are observed in the first three years after the focal election, before rebounding towards zero.
Importantly, survey data show that educators are more likely to be union-endorsed. In many ways, then, these findings reinforce what other studies have found. That is, union representation predicts higher teacher salaries and potentially negative effects on student performance—while also translating to less support for charters. The analysts write, in matter-of-fact fashion, that they interpret their findings “as consistent with educator board members shifting bargaining in favor of teachers’ unions.” Yep, so does this writer.
SOURCE: Ying Shi and John D. Singleton, “School Boards and Education Production: Evidence from Randomized Ballot Order,” Annenberg Institute at Brown University (April 2021).