Bootleggers, Baptists, and (School) Building Closure

• In Event: Political Parties and American Democracy Mini-Conference III: The State of Polarization, Nationalization, and Identity in U.S. Schools

Abstract

Closing school buildings represents a third rail in local education politics, and there are growing bodies of research (1) examining the academic impacts of such closures on students and (2) the political feedbacks that result. In this study, I bring together original data sources and a credible causal (difference-in-differences) research design that speaks to both literatures.

First, I show that when districts close schools, they target lower-enrolling, lower-performing, and more heavily minority schools. Crucially, these schools are lower-performing because of the demographic composition of students they serve, not because these schools are lower in quality (as measured by year-to-year growth or “value-added”).

Second, I find that school closures have no measurable impact — positive or negative — on student achievement in the years after closure, debunking claims of both closure supporters (who argue that closing lower-achieving buildings will improve learning) and critics (who argue closures will hurt students). I conclude that this is largely because closures do not target low-growth buildings, the only context where closures have been found to produce positive effects on learning. This decision is strategic, designed to minimize the political costs of closure, resulting in the targeting of buildings in communities with less political power.

I do show, however, that closures produce significant cost savings by reducing staffing levels, providing support for one of the stated motivations. In addition to the large-N analysis, I also present original interviews with district leaders and community activists in school systems that have experienced significant closures, which confirm the finding that political, fiscal, and employment considerations largely drive both support and opposition to building closures, not sincere concerns about student learning. The findings are most consistent with the “bootlegger and Baptist” model of policy change.

These results have important real-world implications in light of the coming wave of closures due to declining public school district enrollment.

Author

• Vladimir Kogan, Ohio State University