Sohoni, Deenesh, & Saporito, Salvatore
Mapping School Segregation: Using GIS to Explore Racial Segregation Between Schools and their Corresponding Attendance Areas
College of William and Mary
Examines whether student enrollment in non-neighborhood schools changes levels of racial segregation in public schools across urban school districts by comparing the racial composition of schools and their corresponding attendance area.
Journal Name or Institutional Affiliation:
American Journal of Education
Vol. 115, No. 4, pp. 569-600
- The difference between the percentage of children who are White in schools and their corresponding attendance boundaries is 5 percentage points at the elementary school level and 6 percentage points at the middle and high school levels. This difference is due to White children attending schools of choice (i.e., private, magnet, and charter schools) at higher rates than non-White children.
- Where we would expect schools to contain nearly equal proportions of White and non-White students is precisely where white children are the most underrepresented in schools relative to their neighborhoods.
- As students move into higher school levels, and as school boundaries become more racially balanced, the discrepancy between the percentage of white children in catchment areas and the schools that serve them will grow.
- At the elementary school level, the results indicate that 18 of the 22 school districts have schools that are more segregated than their corresponding catchment areas.
- As with elementary schools the majority of districts (18 of 21) have dissimilarity scores that are higher in schools than their corresponding boundaries, and only one of these districts (San Diego City) has considerably lower levels of segregation in schools than their catchment areas.
- Results indicate clearly that the availability of non-neighborhood schools exacerbates segregation within school districts because the distribution of White and non-White students within them is even less than in school catchment areas.
- Increased school segregation results from two processes: first, White children exit integrated neighborhood-based public schools at a greater rate than non-White children; second, White and non-White children are redistributed in private, charter, and magnet schools more unevenly than they are distributed across residential areas (as defined by school catchment areas), further contributing to racial segregation in schools.
Journal Article Historical Analysis
Choice, Public Schools, Racial Composition, Segregation, Urban Schools
Method of Analysis:
Unit of Analysis:
- Sample includes 2,853 private elementary schools, 2,195 private middle schools, and 913 private high schools within the 22 largest school districts in the country.
- School districts provided maps of attendance boundaries
- Combined district maps with Census Bureau's TIGER files which include streets and other features
- Additional census data provided demographic data for each district boundary area
- Racial composition of schools derived from Common Core of Data (CCD), available from the U.S. Department of Education’s National Center for Educational Statistics. The CCD describe the number of children by grade level and race for virtually every public school (including charter and magnet schools) in the country.
- The CCD also provided enrollment data by grade level and race.
- For private schools, data from the 1999 Private School Survey was used that contained information on virtually every private school in the country, including their addresses, racial enrollments, and the number of students in each grade.
- DV: Racial segregation in public schools, attendance of school of choice
- IV: Enrollment in private schools, race