Attracting Diversity: Magnet Schools and Integration in a Large Urban District

• In Event: Roundtable Session 38
  In Roundtable Session: Understanding School Choice as a Geographically Relational Phenomenon Across North America: Using Geographic Information Systems Approaches

Abstract

Objectives: We examine the impact of magnet schools on patterns of attendance by race/ethnicity and socioeconomic status in a large urban district with an extensive magnet program. The focal district, one of the first in the nation to adopt magnet schools as part of its desegregation efforts in the 1970s, has continued to steadily expand its portfolio of magnet programs. At present, nearly 1 in 4 district students attends some type of magnet school. We examine how the district’s current magnet system is helping the district meet its intent of improving the diversity of its schools, addressing two key questions: 1) To what extent is student access to magnets equitably distributed across the district?, and 2) To what extent do magnets integrate or segregate students by race/ethnicity and socioeconomic status?

Methodology & Data: We use data on over 200,000 students from a large urban district for the 2012-13 school year, spatially merged to attendance zone boundaries and neighborhood block-level Census ACS data. First, we estimate a series of spatial interaction models (Fotheringham & O’Kelly, 1989) to model and visualize the extent to which magnet schools are accessible to students of different racial/ethnic and socioeconomic groups, attending to differential access by magnet type (e.g., International Baccalaureate, language immersion). Second, we compare current patterns of attendance to “counterfactual” patterns of attendance that would be expected in the absence of magnet programs. Finally, we will employ spatial conditional logit models to predict student enrollment patterns as a function of student, neighborhood, and school characteristics.

Preliminary Findings: We find that a handful of magnets are achieving modest increases in integration, in that they have student enrollments that are significantly more diverse than their surrounding neighborhoods and the zoned public schools which students would otherwise attend. However, the overall impact of magnets is segregative; as such, the overall level of segregation in the district is higher than it would be if students attended their zoned schools. Moreover, the highest-achieving magnet programs are particularly segregative; in particular, drawing white students in from more diverse attendance zones. While results suggest that the geographic location of magnet programs plays an important role in worsening segregation, parental decisions and student eligibility for academically selective magnets are perhaps more salient factors.

Significance: Consistent with concerns of critics (Saporito, 2003), findings suggest that rather than achieving the imperative of integration, the district’s magnet schools may be worsening inequities in an already segregated district. Moreover, our findings suggest that locating prestigious magnet schools in more disadvantaged neighborhoods is unlikely to yield dramatic increases in integration. In particular, findings call into question the continued practice of using academically selective admissions criteria for magnet schools. It is important to note that results fail to capture the role that magnet programs may play in stemming the tide of “white flight” from the district (Bifulco, Ladd, & Ross, 2009; Smrekar & Honey, 2015); students that elect to participate in magnet programs may be those whose parents would elect to move to neighboring suburban districts if magnet options were not available to them.

Authors

• Meredith Paige Richards, Southern Methodist University

• Kori James Stroub, The University of Texas - Austin