This study uses administrative data and other public records to examine the impact of unionization on the test-based achievement of California charter schools between 2003 and 2013. Using a difference-in-differences approach, the authors estimate that unionization boosted math achievement by about 0.2 standard deviations but did not significantly affect reading achievement. These estimates differ from those of Hart and Sojourner (2015), who analyzed similar data but found that unionization had no impact on test scores. However, in the appendix of the more recent study, the authors provide a mostly convincing defense of their methodological choices.
A quarter of California charters were unionized in 2013, and between them these schools accounted for a third of charter enrollment. However, for obvious reasons, the authors exclude “conversion” charters that are automatically unionized because they are legally bound to the district contract. So their analytic sample includes just forty-four charter schools that switched from non-unionized to unionized between 2003 and 2013.
Overall, students in unionized charters score about 0.5 standard deviations higher on math and about 0.3 standard deviations higher in reading than students in non-unionized charter schools. And teachers in these schools are more experienced and far more likely to be on track for tenure than those in non-unionized charters. Yet relative to non-unionized charters, “switchers” have lower test scores, more minority and English Language Learner students, and are unusually concentrated in large urban school districts. (Nearly half are located in Los Angeles, Oakland, and San Diego.)
As the authors note, unions might improve school performance in a number of ways. For example, they might attract better teachers (via higher wages), reduce teacher turnover, and boost “teacher voice” (thereby improving the flow of information within a school). However, they might also do harm insofar as they prevent administrators from rewarding high-performing teachers or terminating underperforming teachers, and insofar as they engage in rent-seeking behavior.
But this is just speculation; one weakness of the study is its inability to account for the mechanism through which unions may boost achievement. For example, the authors estimate that unionization leads to a 0.8 year decline in teacher experience but has no effect on class size, the share of teachers with a master’s degree, or the fraction with a tenure-track position (none of which makes very much sense).
Another major question mark is the degree to which the study’s findings are generalizable. After all, unionization doesn’t occur in a vacuum. So it’s possible—perhaps even likely—that working conditions and administrators at the schools in question were below average. Thus, even if the authors have succeeded in isolating the impact of unionization for these schools, it doesn’t necessarily follow that every charter would be higher-performing if it unionized.
In short, the study suggests that voluntary unionization is potentially beneficial for low-performing charters. However, when it comes to high-performing charters—not to mention traditional public schools, where both theory and experience suggest that unionization may be more problematic—there are still plenty of reasons to be skeptical.
SOURCE: Jordan D.Matsudaira and RichardPatterson, “Teachers’ Unions and School Performance: Evidence from California Charter Schools,” Economics of Education Review (September 2017).
