Understanding School Segregation Data
If you’ve been confused about what the data and research say about school segregation — whether it is growing or shrinking – you’re not alone. Scholars argue over this, too.
Gary Orfield of the University of California, Los Angeles, and Sean Reardon of Stanford University have had a long-running, public dispute on the issue. Here’s what a panel of experts had to say about school segregation recently at the Education Writers Association’s annual seminar in Los Angeles.
‘Evenness’ v. ‘Isolation’
Views on whether segregation is rising or dropping in America depend on how it is measured. Academics sometimes refer to this as the “evenness” versus “isolation” debate.
Much of the academic community prefers the “evenness” measure, which looks at how racially balanced the school population is compared with the community’s residents.
For a simple example, if the residents of a school district are half white and half black, then an “even” distribution would be 50 percent black students and 50 percent white in each school. If one school’s population was 85 percent white while the other school’s was 99 percent black, then that would be “uneven.”
The issue gets trickier as the population grows and changes. Say, an influx of Latino immigrants moves into this “even” school community and the population changes to one-third black, one-third white, and one-third Latino. A school that was once 50 percent white might fall to 30 percent white and grow to almost 70 percent “minority” students — a combination of blacks and Latinos.
But just because a school is becoming less white doesn’t mean schools are becoming more segregated. They are still “evenly” reflecting the racial diversity of the residential population.
By this “evenness” measure, segregation decreased by a large margin from the 1960s through the 1980s. Segregation grew slightly in the 1990s, but then declined again.
Leaving aside small up-down fluctuations in the past two decades, you could say that schools are as segregated now as they were in the late 1980s, according to Meredith Richards, an assistant professor of educational policy at Southern Methodist University, who spoke on the EWA panel. According to the “evenness” measure, schools have not been resegregating along racial lines.
Within Districts or Across Them?
The problem with the evenness measure is that it only considers the demographics of each school district and the schools within it.
A suburban district with a 90 percent white population might abut an urban district with an 80 percent minority population. Each district is more or less evenly distributing its students among the schools within its borders. But there’s high residential segregation between this suburban and urban district.
Nationally, researchers have found increased racial segregation across district lines, as people of different races move into different catchment areas. In other words, segregation is going down somewhat within district boundaries, but going up across school boundaries.
Meanwhile, many journalists, school policy advocates and some academics have focused on two different measures: racial isolation and exposure. Racial isolation examines how much a student of one race attends school with peers of the same race (e.g., what percentage of black kids attend a school that is 90 percent or more black). Exposure measures how much a student of one race attends a school with peers of a different race (e.g., how many white kids are in the school of a typical black student).
By these isolation and exposure measures, segregation declined substantially between the 1960s and the 1980s. But since then, some analysts see steady increases in the number of schools where students are racially isolated and not exposed to white students. For some observers, this is evidence of re-segregation.
The data is complicated here, and different researchers and journalists arrive at different conclusions. When I look at federal data from the National Center for Education Statistics (tables 216.50 and 216.55 in the “Digest of Education Statistics”), I see that the typical black student is far less likely to attend a school that is more than 90 percent black. More than 22 percent of black students attended such all-black schools in 1995. Fewer than 15 percent of black youths attended an all-black school 20 years later.
‘Exposure’ at Issue
However, if you ask the question slightly differently — “What percentage of blacks attend a school that is 90 percent or more minority, including Latinos, Asian Americans and Native Americans?” — then the level of segregation has gone up. That is, back in 1995, 34 percent of black students attended a school that was 90 percent or more minority, meaning it had relatively few white students. By 2015, more than 40 percent of black students attended a school that was 90 percent or more minority. What this means is that we’re increasingly isolating poor black and Latino students together.
Alvin Chang of Vox produced an excellent explainer on the data confusion in March.
“I focus on (racial) isolation because that’s from the point of view of students,” he said during the EWA panel.
The biggest drawback of this racial exposure or isolation analysis is that it doesn’t take into consideration demographic changes. As the number of Latino students rises in our public schools, the percentage of minority students should increase. You don’t even need white flight or re-drawn school zones for it to happen.
In a 2013 study, Jeremy Fiel, now at the University of Arizona, reconciled the divergent findings between evenness and isolation. After controlling for these demographic changes, both measures show decreases in segregation since the 1990s.
School Choice and Segregation
Again, there are controversies here, too. University of Arkansas professor Patrick Wolf, who spoke on the EWA panel, raised questions about a December 2017 analysis by the Associated Press on school segregation. The AP concluded that 15 percent of the nation’s public charter schools were 99 percent minority. Wolf said the story misleadingly implied that charters were making school segregation worse.
Many charter schools are purposely situated in minority neighborhoods where students would have otherwise been going to traditional public schools that were equally dominated by minorities, Wolf explained. One 2014 study by Gary Ritter of the University of Arkansas, focused on the city of Little Rock, found that charters slightly improved integration in that city. Wolf said it’s important to do fine-grained city-by-city analyses.
The jury is also out on open-enrollment policies that allow families to choose any school they want in a district. In New York City, school choice made segregation worse because of the fierce academic competition to get into top middle and high schools. In Boston, the voluntary busing program, Metco, improved racial integration, but not students’ test scores, according to research published by the National Bureau of Economic Research.
A 2017 review by Else Swanson of the University of Arkansas saw evidence that school vouchers can enhance racial integration. Wolf’s study of school vouchers in Louisiana found that the public schools became less segregated after the voucher students left, but the private schools became slightly more racially homogenous.
Since 2007, when the U.S. Supreme Court banned the consideration of race in assigning students to schools, some communities have been trying to figure out other ways to desegregate.
University of Georgia associate professor Sheneka Williams, who spoke on the EWA panel, has studied post-2007 desegregation efforts in North Carolina and Kentucky. She said socioeconomic status, can be a good proxy for race. “It won’t work for every district. It depends on the demographics,” she said, explaining that a poor white population or a wealthy black population complicates the effort. Other communities have reconfigured school zones in novel ways to promote integration, using geography instead of income.
Successful desegregation programs tend to have broad community support and involvement, particularly support from the Chamber of Commerce. Even well-laid plans can be overwhelmed by population growth and change, Williams said.
Meanwhile, magnet schools are often poor engines of desegregation. While they may lure different races and incomes to the same school building, students are often tracked into high and low classes and have little contact with each other during the school day.
A Concluding Thought: Housing
Panel members suggested that one way to make further progress in school integration is to combat residential segregation. For example, Williams suggested that the federal government might consider financial incentives for people of different races and ethnicities to live in the same school district boundaries.