Evaluating Education Governance: Does State Takeover of School Districts Affect Educational Inputs and Outcomes?

• In Event: What Makes School Districts Effective?

Abstract

Local school boards have historically had primary authority for running educational systems in the U.S. but little is known about the merits of this arrangement for academic performance. State takeovers of struggling districts represent a rare alternative form of educational governance and have become an increasingly common policy response in the event of persistent low performance. Takeovers also tend to target low- performing systems where low-income learners and students of color are concentrated (Wong & Shen, 2003; Morel, 2018). Therefore, the impact of takeover on student achievement outcomes could contribute to either narrowing or widening race- and/or class-based achievement gaps, depending on the direction of any effect.

We track all state takeovers nationwide from the late 1980s, when the first takeovers occurred, through 2015 and describe the nature and predictors of these takeovers (n=107 takeovers). Our tracking was based on previous studies that tracked takeovers for subsets of time, media reports, publicly available government documents, and correspondence with state education agency officials. We find that takeovers have become more common and shorter over time. We also observe that districts serving larger concentrations of Black students are more likely to be taken over than those serving fewer, even controlling for academic performance among other characteristics. Districts receiving a larger share of revenue from their state governments are also more likely to experience takeover.

After describing takeovers, we merge our tracking data with a new data source – the Stanford Education Data Archive (SEDA) – that allows for cross-state comparisons of student achievement outcomes. We analyze these data using difference-in-differences methods to estimate the effect of state takeovers that occurred between 2010 and 2015 (n=33 takeovers) on educational inputs (educational spending, teacher-student ratios, and the size of the local charter school sector) and student achievement (test scores in math and English Language Arts). We use matching methods to identify a set of comparison districts that are identical to the takeover districts on a number of dimensions that predict takeover (both in terms of district demographic characteristics and also political context). In other words, we compare whether any changes in outcomes for takeover districts after takeover were different, on average, from changes in those same outcomes for very similar districts that did not undergo takeover over the same period.

We find no evidence that takeovers benefitted either English Language Arts or math achievement, and some evidence of small negative effects in English Language Arts, though we observe notable heterogeneity of effects across districts. Takeovers also increased the share of a district’s students served by charter schools but this shift does not explain the test score declines. We find no significant effects on class size or greater educational spending. Our findings suggest that policymakers should be cautious about deploying state takeover as a strategy for improving academic outcomes in contexts where conditions are very different from those rare cases of takeover that have generated achievement gains for students.

Authors

• Beth Schueler, University of Virginia

• Joshua Bleiberg, Brown University