Between 1990 and 2015, Seattle’s neighborhoods saw a notable decline in racial segregation.

It would make sense, then, to think that the city’s public schools had also become more integrated. Not so.

In fact, they were headed in the opposite direction. In 1990, only 3 percent of schools were intensely segregated — that is, at least 90 percent of students were nonwhite — but by 2015, that number had spiked to 17 percent.

That’s not entirely surprising. During that time, a high-profile Supreme Court case made it more difficult for Seattle to integrate its schools by race. But new research looking at America’s 100 largest cities shows that the diverging trends in Seattle — neighborhoods growing more diverse, as their schools grow more segregated — is not an anomaly.

The analysis finds that, between 1990 and 2015, 72 percent of U.S. cities saw their neighborhoods grow less racially segregated, by one measure. Sixty-two percent saw their schools grow more segregated over that same period.

“There is this incredibly striking trend,” said Ryan Coughlan, a professor at Guttman Community College, CUNY, who conducted the research. “It raises all kinds of alarm bells and questions as to what that’s about.”

Most cities did not see schools segregate as much as Seattle did. And 27 cities, like Durham, North Carolina and Memphis, Tennessee, saw both their schools and neighborhoods grow less segregated in concert over that 25-year period. But overall, the study finds that integrating neighborhoods didn’t predict integrating schools in those same cities.

The analysis has significant limits, particularly when looking at single cities. It uses one of several possible ways to measure segregation: the degree to which the racial breakdown of students in individual schools or neighborhoods mirror the demographics of the rest of the district or city. In most cases, charter schools are not included. That makes the data less useful in places like Detroit, where charters now enroll half the city’s students.

The big-picture trend matters, though. More integrated schools have long been shown to improve academic outcomes for low-income students and students of color. Living in a more integrated neighborhood has also been linked to long-run benefits for younger kids.

“Because of the connections between integration and educational opportunities, the dramatic increase in school segregation alongside the decrease in neighborhood segregation requires the immediate attention of school leaders, policymakers, and the public as a whole,” Coughlan wrote.

What’s behind those trends?

The analysis, published last month in the peer-reviewed Peabody Journal of Education, can’t say why that’s happening. The end of many legal desegregation orders during that time likely played a role. Coughlan also hypothesizes that the rapid increase in school choice, through charter schools and other means, had something to with it.

“These are 100 different urban areas with very different circumstances,” Coughlan said.

The paper’s starting point is 1990, before the first charter school law passed in the U.S. Since then, school choice has rapidly grown, through charters and other means.

In Seattle, though, charter schools are almost certainly not the cause of its increase in school segregation, since the city has very few.

Another city that saw a major spike in school segregation along with a modest decline in neighborhood segregation is Charlotte. Like Seattle, it’s seen a resegregation of schools in the wake of high-profile court cases.

Charlotte also has a number of charter schools; there and elsewhere in North Carolina, other research has found that charters have likely exacerbated segregation. (Coughlan’s Charlotte data does not include charter schools.) Nationally, research has shown that charters either exacerbate school segregation or have no effect on it.

“The broader literature at this point I think shows that charter schools do not integrate schools,” said Ann Owens, a sociologist at the University of Southern California who studies segregation.

Other research has shown that the existence of different school options can promote neighborhood integration (also described as gentrification). That could help explain Coughlan’s results, with a family’s ability to opt out of a neighborhood school encouraging their move to a neighborhood they wouldn’t otherwise have considered.

The disconnect between housing and schooling trends has important implications. For one, it means that divided neighborhoods shouldn’t be used as an excuse to do nothing about divided schools, said Tomas Monarrez, a researcher at the Urban Institute who has studied school boundaries.

“Neighborhood segregation is the result of a long, long history of discriminatory policies both on the part of private agents and the federal government,” he said. “School systems have gotten to ride that and say segregation’s not our fault.”

Instead, he argued, school leaders should be taking affirmative steps to integrate schools, and recognize that they may have to continually adjust their policies. “School attendance boundaries don’t have to replicate neighborhood segregation,” Monarrez said. “You can gerrymander school attendance boundaries to decrease it.”

City-by-city data

You can look up how residential and school segregation changed in your city from 1990 to 2015 below. Keep in mind that a city and its corresponding school district do not always overlap perfectly — the school data for Indianapolis, for example, includes just the Indianapolis Public Schools, the city’s central district but one of 11 districts in the city.

Segregation, here, means the degree to which the demographics of students in individual schools mirror the rest of the city’s public school students. This captures whether different groups of students are spread evenly across schools in a city, but it doesn’t say much about cities where virtually all students are students of color. Most segregation occurs between rather than within school districts.

Source: “Divergent Trends in Neighborhood and School Segregation in the Age of School Choice,” Peabody Journal of Education.