The Effect of Magnetization: Magnet schools and school demographic change

• In Event: Student assignment, school choice, and diversity in contemporary American public school districts

Abstract

Introduction: This paper connects two key directions for research on school assignment policy. First, what are important goals for school assignment policies to pursue? Historical work points to at least three potentially conflicting goals: increasing racial and socio-economic diversity, increasing enrollment and maintaining market share against charter and private school alternatives, and progressively distributing the burden of reassignment and access to special academic programs.  Second, how do specific school assignment mechanisms distribute these goals? All students are assigned to schools via one of two broad mechanisms: via geographic attendance boundaries or via application (choice). These mechanisms can be grouped into four broad “systems” that can be characterized based on whether students have access to geographic assignment and whether application options are limited only to certain schools, or encompass all district schools. For example, substantial attention has been paid to fully choice systems in large urban centers where all students participate in choice lotteries that are present at all schools (e.g. Boston) (Levinson, 2015). However, the majority of urban school districts in the U.S. likely operate “mixed” school assignment systems that combine geographic attendance boundaries with limited choice options placed strategically at certain schools. Based on historical work in Durham, North Carolina, I hypothesize that in such “mixed” systems, it is the location and structure of choice programs that determines their impact on school demographics, market share, and reassignment burden (Rossell, 2003). Research Questions: Magnet schools provide an important opportunity to study how school assignment mechanisms affect three school assignment goals. This paper measures the effects of magnetization on school demographics, and how these effects vary with four mediating characteristics: 1) the stated goals of the school assignment system, 2) the structure of school assignment “systems” 3) the structure of magnet programs, and 4) the location of magnet programs.  Methods: To answer these research questions, I employ an event-study difference-in-difference (DiD) approach that compares schools magnetized at different times between 1980 and 2020 for eight urban North Carolina school districts. I use archival sources to create a novel magnetization database to supplement existing public data on magnet schools. These districts contain the vast majority of North Carolina’s magnet programs, have been early and sustained adopters of magnet schools, and have experimented with a wide range of magnet school structures, located across a range of neighborhood contexts, and implemented over a long timespan. They thus allow a relatively comprehensive comparison of how a wide range of magnetization strategies work to shape school demographics as part of “mixed” student assignment systems.   Results: Preliminary results from Durham County indicate that magnet structures and location produce significant combined effects. Certain magnet structures have larger impacts on school demographics, and this impact varies by the placement of magnet programs in mixed-race or predominantly black neighborhoods.  Implications: Better understanding the effects of different school assignment mechanisms on a wider range of important outcomes is important for helping school districts design more effective school assignment policies that target important goals.

Author

• Vitaly Radsky, University of North Carolina at Chapel Hill
  Presenting Author