Proponents of socioeconomically and racially diverse schools, myself included, frequently point to data showing that charter schools, on average, have greater concentrations of poverty and racial isolation than traditional public schools. For supporters of school integration, these increased levels of de facto segregation are reason to be skeptical about the current direction of the charter school sector.

But it is important not to oversimplify this concern to mean that school choice in general is bad for diversity. In theory and in practice, school choice has the ability to help or hinder school integration.

A pair of recent studies illustrates this point well. Myron Orfield and Thomas Luce, researchers from the Institute on Metropolitan Opportunity at the University of Minnesota Law School, looked at how the racial/ethnic diversity of schools in the Twin Cities metropolitan area has been affected by Minnesota’s Open Enrollment law. Under Open Enrollment, all districts are required to allow students from anywhere in the state to attend, subject to space constraints. According to the study, Open Enrollment transfers had an overall negative effect on school diversity. In 2009-10, about a quarter of all transfers were “integrative,” moving students to districts where their race/ethnicity was underrepresented as compared to their sending district. However, more than a third of all transfers were “segregative,” moving students to districts where their race/ethnicity already represented a greater share of the student population than in their sending district.

In another recent study, released by the National Center for the Study of Privatization in Education at Columbia University’s Teachers College, researchers from the University of Arkansas examined how charter school enrollment has affected the racial composition of schools in Little Rock, Arkansas. Researchers Gary W. Ritter, Nathan C. Jensen, Brian Kisida, and Daniel H. Bown found that both charter schools and traditional public schools in the Little Rock metropolitan area had similar levels of racial integration. However, roughly three-quarters of charter school transfers increased the racial integration of traditional public schools in the Little Rock school district by making the racial/ethnic demographics of the school a student was leaving closer to the average demographics for the metropolitan area.

These two studies do not give many clues about why Open Enrollment increased school segregation in the Twin Cities but charter enrollment decreased school segregation in Little Rock. The differences may have to do with the number or type of choice-based schools available, or with the patterns of residential segregation in each area. But we know from other examples that the effects of school choice on school segregation can also be shaped by school and district policies.

Charter schools can intentionally foster diversity by locating in areas with untapped potential for diverse public school enrollment and targeting recruitment to create a socioeconomically and racially diverse lottery pool. Districts can follow models like the one used in Cambridge, Massachusetts, in which families rank their top three choices and students are assigned to schools based on an algorithm that balances parent preferences with a control to make sure that the percentage of low-income students at each school is reflective of the district has a whole.

It should not be surprising that the effects of school choice on school diversity vary—but it is a reminder that we should think about how school choice can be structured to help, not hurt, integration efforts.