Should We Close Schools on the Basis of Performance?
Note: This week, Deven Carlson, an Assistant Professor of Political Science at Oklahoma University, is guest-blogging. His work explores education policy and politics.
Policies at all levels of government promote the closure of low-performing schools as a strategy for improving student outcomes. At the federal level, No Child Left Behind listed closure as a possible consequence for schools that failed to make Adequate Yearly Progress (AYP) for five consecutive years. Similarly, the federal School Improvement Grant (SIG) program identifies closure as one of the approved "turnaround" strategies that districts receiving funds through the program can implement. Among states, legislators in 15 statehouses across the country have passed automatic charter school closure laws, which require charters to shut down if they fail to meet specified performance requirements. And countless districts—Chicago, Washington D.C., Cleveland, New York City, Philadelphia, Detroit, and Pittsburgh, to name just a few—have implemented large-scale closure initiatives in recent years.
The logic of closing low-performing schools is clear: Shutting down bad schools will remove students from these contexts and facilitate their transition to a better school. Improved academic outcomes will follow. In addition, the resources that had been devoted to the closed schools can be reallocated to those that remain open, which may contribute to their improvement.
This seemingly tight logic contributes to support for closure policy in the abstract—it is easy for policymakers to think, "We should definitely shut down bad schools." Of course, closure policy is not implemented in the abstract. It is implemented in specific districts and schools, affecting actual teachers, students, families, and neighborhoods. Perhaps not surprisingly, then, when it comes time to identify specific schools to be shut down, closure policy routinely proves to be immensely controversial.
The disconnect between support for closure policy in the abstract and closure policy in reality is illustrated by the fact that very few schools were actually shut down under NCLB or SIG. Further, just about every large-scale district closure effort has sparked strong resistance from parents and the broader community. This resistance often resulted in the initial closure plans being scaled back, sometimes substantially.
So should we shut down low-performing schools on the basis of performance? The logic underlying these policies seems sound, and we are accumulating a body of evidence indicating that shutting down low-performing schools can lead to student achievement gains for students whose school was closed. Indeed, we have findings to this effect from a well-done study in Michigan, as well as from analyses I have done with my colleague Stéphane Lavertu using data from Ohio. In addition, very recent work from New York City provides evidence that closure of low-performing high schools increased the likelihood of graduation for students who would have otherwise attended those schools. Together, these factors provide clear support for policies that lead to the closure of low-performing schools.
However, let me also suggest three reasons why we might think twice about shutting down low-performing schools. First, identifying bad schools may not be as straightforward as it initially seems, particularly if we think of school quality as extending beyond contributions to student reading and math scores (as I think most of us do). Even if we only talk in terms of test scores, many states' approaches to measuring school performance leave something to be desired and basing closure decisions on these metrics could lead to shutting down some schools that are actually effective. And if we go beyond test scores to consider schools' effectiveness in subjects other than reading and math, as well as in extracurricular activities, then identifying low-quality schools becomes even more complex.
Second, school closure undeniably leads to disruptions in the lives of numerous individuals. Closure clearly disrupts the schooling experiences of students who are displaced, and there is lots of evidence that mobility in and of itself is harmful to students. The effects of closure may also spill over and affect kids attending schools that receive displaced students—the study from Michigan provides evidence of negative spillover effects. Finally, shutting down a school is likely to disrupt the lives of faculty and staff employed by the school. When considering potential closures, it is important for policymakers to do their best to offset or minimize these disruptions. For displaced students, this means trying to ensure the availability of appreciably better schooling options. These options may exist in some districts, but in others they may not.
Third, schools often contribute to the identity of a community and serve as the pillar of a neighborhood. Closing schools may impose very real costs on the broader community. Will closing a school reduce the value of nearby properties? Will it lead to greater levels of crime in the area? What about the pride that local residents and alums had in the football team, marching band, or debate team? Some of these issues are easier to quantify and convey than others, but all of them—and others—should probably be considered when deciding whether to shut down a low-performing school.
All of this is to say that policies promoting the closure of low-performing schools are neither unambiguously desirable nor uniformly disastrous. As with most things in education policy, the wisdom of shutting down schools on the basis of performance is almost certainly context-specific. Each case has its own set of factors that should be carefully considered—with the input of a wide variety of stakeholders—before a decision is made.