Within-School Segregation in an Urban School District

  • The segregation of Blacks and of Whites across schools is higher than the segregation of Hispanics and Asians
  • In NYC, rates of across school segregation at the elementary level are primarily driven by residential segregation because students are typically assigned to their neighborhood schools.
  • Although within-school segregation is lower than across-school segregation, the relative severity of these two types of segregation varies considerably by race.
  • Hispanics are 40% less exposed to non-Hispanics in their classrooms than in the school district, and within-school sorting processes account for almost 20% of total segregation.
  • Total segregation between racial groups far exceeds total segregation between native-born and immigrant students.
  • The actual segregation levels are two to five times higher than the levels achieved under the random assignment simulation, suggesting that segregation is partially driven by non-random processes, such as systematic segregation or other sorting practices.
  • If test scores were the only criterion used to sort students, Blacks and Hispanics would not be very isolated.
  • A system based strictly on ability grouping would lead to higher Asian and White segregation, and conversely, lower Black and Hispanic segregation than current levels.