Carmen Arace Middle School in Bloomfield, Conn., is 88 percent black.
Using one type of school segregation measurement, Arace could be considered racially segregated because black students there share the school with relatively few students of other races. This type of indicator is known as an “exposure” or “isolation” index because it indicates whether the school’s demographics promote isolation or exposure among students of different races.
However, this picture changes when segregation is examined against different yardsticks. “Evenness” indicators quantify the degree to which students of different races are evenly distributed among schools in a certain area. Such measures might suggest that Arace in not racially segregated because the school’s demographics are in line with the overall makeup of the 2,087-student Bloomfield school district, outside of Hartford, which is 72 percent African-American.
These two types of measurements—racial isolation/exposure versus evenness—are at the crux of an academic debate over whether the level of school segregation has changed since the early 1970s, when the flurry of post-Brown v. Board of Education desegregation activity slowed.
In a research synthesis scheduled to appear in the peer-refereed Annual Review of Sociology in July, sociologists Sean Reardon of Stanford University and Ann Owens of the University of Southern California find that the average levels of American school segregation by race have barely budged since the 1980s. They base these findings on a review of more than 100 studies, some of which examined evenness, some which examined isolation and exposure, and some of which employed both types of measures. In “60 Years after Brown: Trends and Consequences of School Segregation,” they conclude that “changes in segregation in the last few decades are not large, regardless of what measure is used, though there are important differences in the trends across regions, racial groups, and institutional levels.”
Definitions and Demographics
For example, segregation has increased in districts in which desegregation orders ended. But overall, Mr. Reardon and Ms. Owens find that school segregation has remained relatively stable, if disputed.
“Although there is disagreement about the direction of more recent trends in racial segregation, this disagreement is largely driven by different definitions of segregation and different ways of measuring it,” Mr. Reardon and Ms. Owens write. “There have been modest decreases in the exposure of minorities to whites, but these have been driven primarily by demographic changes in the school-age population.”
Between 1980 and 2010, the percentage of children who are white in the United States declined from 74 percent to 54 percent, according to Child Trends, a nonprofit, nonpartisan research organization in Bethesda, Md. During that time, the population of black children held steady at about 15 percent while the percentage of Hispanic children increased from 9 percent to 24 percent.
“The demographic changes are substantial,” Mr. Reardon said, “but they haven’t changed the amount of sorting among schools.”
In a response to Mr. Reardon and Ms.Owens’ review, Gary Orfield, Genevieve Siegel-Hawley, and John Kucsera of the The Civil Rights Project/ Proyecto Derechos Civiles at the University of California Los Angeles muster evidence from 17 years’ worth of data from their own organization and others to argue that black and Latino students are substantially more segregated today than they were a generation ago. Eight of the studies reviewed by Mr. Reardon and Ms. Owen were published by The Civil Rights Project or authored or co-authored by Mr. Orfield, the lead author of the response.
The conclusions in the response are largely based on measures of exposure and isolation. Although the Civil Rights Project authors have examined evenness in multiple studies, they suggest that such measures can produce misleading results, especially when used to simultaneously assess the segregation levels of multiple racial groups within a metropolitan area. Also misleading, they say, are studies that apply evenness measures to small, highly segregated jurisdictions, like central cities or virtually all-white communities.
‘Real World’ Preparation
Bloomfield Schools Superintendent James Thompson Jr. was not involved with the review or the response. But in an interview, he offered reasons why he believed it would be misleading to use a measurement that declared his district, which was never under a desegregation order, to be desegregated.
“It’s [racial segregation] an issue in terms of just resources, in terms of the resources that are valuable to these families,” he said. “Also in terms of diversity, it has another social benefit, in terms of students interacting within a diverse population, in terms of the paradigm for the real world that they’re going to eventually be a part of when they become ready for work or even college.”
Yet in their review, Mr. Reardon and Ms. Owens find that there is limited evidence on how or why racial segregation might worsen educational inequalities and, correspondingly, how desegregation might help. Because such evidence is limited, they contend that it is difficult to know which segregation measures to use under which conditions. For instance, if contact among people of different races is the active ingredient in school desegregation efforts, then measures of isolation and exposure are central. However, if what matters most is evenly distributing resources to schools with different demographic profiles, then evenness measures might be more important.
Evenness is also an important indicator for policymakers who might have limited control of the demographic profile of the population in their jurisdiction but could impact the degree to which students are proportionately distributed among schools, said University of Wisconsin doctoral candidate Jeremy Fiel, whose 2013 study in the peer-refereed American Sociological Review was cited by both the review and the response.
“If you think about how schools get the student body they have, first, they draw from the local population so the first thing is, what’s the composition of the area?” he said. “That depends on things like historical trends, fertility, migration. Given that population, schools get the student body they have based on the way students are distributed across schools. I see the unevenness as one of the causes of isolation.”
‘Evenness’ Critiqued
However, in their response to Mr. Reardon and Ms. Owens, the Civil Rights Project authors downplayed the importance of evenness. They offered extensive evidence suggesting that the benefits of desegregation result from interracial contact and experiences within schools. As a result, isolation and exposure are more important measures, especially since desegregation orders aimed to further their goals by increasing contact between historically more and less advantaged minority groups. The Civil Rights Project authors were especially critical of evenness measures that concluded that desegregation was decreasing because of the proportionate distribution of multiple historically disadvantaged groups, such as Latinos and African-Americans.
One report included in the review that used such an evenness measure (called a “dissimilarity index”) was a 2012 study of residential segregation by the Manhattan Institute, a right-leaning New York-based think tank. The study, entitled “The End of the Segregated Century,” concluded that, nationwide, neighborhood segregation had declined significantly since 1970.
One person who is critical of both the report and the dissimilarity index is Richard Rothstein of the University of California, Berkeley School of Law and the Economic Policy Institute, a left-leaning think tank based in Washington.
“I don’t think it’s reasonable to think of schools that are composed of low-income blacks and low-income Latinos as being integrated,” said Mr. Rothstein, suggesting that income status should also be considered when examining school segregation.
Mr. Reardon and Ms. Owens did find “limited evidence” that economic school segregation had increased by 1990. But Mr. Rothstein still believes that research suggests that race plays a role above and beyond poverty, especially for black students.
“You’re not going to improve the outcomes of black students by putting them into a school with primarily low-income Latino students,” he said. “The only integration program that will likely improve the outcomes of low-income black students is to put them in schools with middle-class students.”
