Magazine March 5, 2018, Issue

The Resegregation Myth

Civil rights leaders, including Martin Luther King, Jr., surrounded by crowds carrying signs, during the civil rights march on Washington D.C (Library of Congress/Handout via Reuters)
Racial groups are integrating of their own volition

The era of federally engineered school integration is coming to a close.

In the wake of Brown v. Board of Education, and in the face of a horrific racial caste system, federal courts took control of hundreds of school districts, primarily in the South. The courts not only took aim at legally enforced segregation, but redrew school-attendance boundary lines and often mandated that students be bused far from home to attend an integrated school. By the 1970s, schools in the South were more integrated than those in the North.

Since the 1990s and especially the 2000s, however, courts have been turning matters back over to local governments. More than half of the orders are gone; many more are no longer being actively monitored. It is still illegal for these governments to segregate their schools on purpose, of course, and some agreed to continue certain integration policies as a condition of regaining control. But in general, they are free once again to give each neighborhood its own school and leave it at that, and many have done so. Schools in these particular jurisdictions are, unsurprisingly, less integrated than they were when the federal government forced them to achieve a more even racial balance.

But contrary to a popular liberal narrative of nationwide resegregation, this has merely balanced out a fortunate (and mostly unengineered) trend of residential integration, leaving American schoolchildren writ large no more segregated than they were a couple of decades back — and roughly as segregated in schools as they are in their neighborhoods. This bodes well for the future. Assuming neighborhoods continue to integrate, schools will become increasingly integrated as well once desegregation orders are fully left in the past and their steady elimination no longer cancels out gains within neighborhoods.

In other words, we are — however slowly — integrating ourselves voluntarily and leaving a system of forced integration behind. The former development is one to be proud of, even as we still struggle to overcome our history and ensure that all children have access to a decent education. And the latter was inevitable, given our federalist system and the legally dubious nature of basing children’s school assignments on the color of their skin. In terms of policy, the future lies in empowering families to make the best decisions for their kids — and letting the benefits of integration flow from there — rather than in deliberately setting each school’s demographic profile through the brute force of government.

As of 1988, according to the University of California–Los Angeles Civil Rights Project, 44 percent of black children in the South attended majority-white schools. By 2011 that had fallen to 23 percent — and there’s a similar trend in the nationwide data. These and similar statistics are often deployed to support the notion that schools are resegregating. A recent piece in Vox, for instance, presented the Civil Rights Project’s numbers as a reason that school districts should start “gerrymandering” their attendance zones to integrate schools.

But such figures do not measure segregation; what they measure is black students’ exposure to white students, which is not the same thing. Most important, they do not account for the remarkable demographic transition that public schools have experienced over the past several decades. Hispanic enrollment has climbed from 14 percent in 1995 to about a quarter today, and non-Hispanic whites are now an outright minority among public-school enrollees nationwide. One should not expect black kids to attend majority-white public schools when public schools no longer have a white majority.

The rise in “segregation” disappears when one measures segregation properly. Here’s how the education researchers Sean F. Reardon and Ann Owens put it in a 2013 paper:

It seems fair to say that the last 25 years have been characterized by largely stable patterns of sorting of students among schools [the type of segregation that creates “unevenness” among schools in terms of racial demographics], while the racial/ethnic composition of the student population has changed substantially, a pair of trends that yields declining black-white exposure measures but no significant change in unevenness measures of segregation.

Even regarding the South itself, where desegregation orders are concentrated, Reardon and Owens note a finding that while segregation did increase in the 1990s, the change was “not large, and reversed following 1998.”

To be sure, white students in the U.S. might bring advantages to their peers that children of other racial and ethnic groups might not — richer and more influential parents, fluent standard English, the behavioral norms of the majority culture, etc. — and thus exposure measures do have their uses. But the country’s shifting demographics are not something that school boards have any power to change, and local governments should not be blamed for “resegregating” schools that are, in fact, becoming more diverse thanks to an influx of Hispanics and to a lesser extent Asians.

How did school segregation hold steady despite the end of so many desegregation orders? The answer is that residential segregation fell, progress that carried over to schools by virtue of the simple fact that most children go to school based on where they live. To perfectly integrate whites and blacks in 1980, you would have needed to move more than 60 percent of blacks to new neighborhoods, according to work by the Brookings Institution demographer William Frey. By 2000 that figure declined to about 50 percent, and it fell still further by 2010.

Frey writes:

One of the trends spurring this shift is the continued integration of southern communities that are magnets for both blacks and whites. . . . In the North, black population losses in cities, the destruction of large public housing projects, and increased suburbanization of blacks are contributing to declines in segregation.

Another impetus toward less segregation is the growth of the Hispanic and Asian populations. Although all minority groups still show a preference for members of their own group as neighbors, tolerance for other groups is strongest in settings that are already multiracial. . . . The 2010 census shows that some of the lowest black-white segregation scores are in areas with large or growing new minority populations.

Frey has also drawn attention to a growing pattern of intermarriage. Today about a seventh of newlywed couples are interracial.

It’s a long way to a society that mixes freely, but we are unambiguously headed in that direction, and our progress has staved off the school resegregation about which so many on the left fret.

One might still argue, of course, for the continuation of the desegregation orders, as schools would be even more integrated with them. These orders were immensely beneficial to black students decades ago, and there is a credible argument that integration still improves outcomes.

Desegregation orders went into effect at different times in different places, and affected children within the same families at different ages. This allowed the Berkeley public-policy professor Rucker C. Johnson to analyze whether black kids’ performance improved when the orders went into effect, and whether younger siblings — who attended the new integrated schools for a greater number of years — saw greater gains.

The results: “For blacks, school desegregation significantly increased both educational and occupational attainments, college quality and adult earnings, reduced the probability of incarceration, and improved adult health status; desegregation had no effects on whites across each of these outcomes.” Further, the effects seemed traceable to the higher school quality that black students experienced in integrated schools, measured through per-pupil spending and class sizes, rather than to integration with white peers per se.

Spending gaps are much less of an issue today, however. In fact, heavily poor and minority schools get more funding in most states. This doesn’t necessarily mean these schools are of equal quality to the ones middle-class white children attend — it takes more money to attract good teachers to poor neighborhoods, for example — but it raises the possibility that integration isn’t as needed as it once was.

According to a recent study by the National Center for Education Statistics, however, even once students’ socioeconomic backgrounds are taken into account, black students do worse in schools that are more than 60 percent black: a difference of seven points (out of 500) on the National Assessment of Educational Progress’s eighth-grade math test, relative to black students in schools less than 20 percent black. This is nearly one-quarter of the overall black–white gap, a finding consistent with other research using similar methods. Further, some studies have found that black students’ dropout rates rise when desegregation orders end, though one failed to find any effect on their test scores. On balance, the evidence suggests that, to this day, black students fare at least marginally better in integrated schools.

It’s not fully clear why this is the case — likely numerous mechanisms are at play — but a 2016 study by Reardon and two colleagues sheds some light on the question. They examined the black–white achievement gap in school districts and metro areas across the country, and included an astonishingly wide variety of statistical controls. What stands out in their full results is that racial segregation in itself is not correlated with a bigger black–white gap. Nor is the percentage of students who are black. Rather, what seems important is the extent to which black students are more exposed than white students to poor peers.

The much-discussed correlation between segregation and achievement gaps (when so many controls are not included) “may arise precisely because racial segregation generally leads to large racial disparities in exposure to poor classmates,” they write. This echoes a 2007 study finding that while metro areas with greater neighborhood segregation have greater SAT-score gaps, “much of the effect of neighborhood segregation operates through neighbors’ incomes, not through race per se.”

It’s also consonant with a growing body of research spearheaded by the Stanford economist Raj Chetty that documents, perhaps unnecessarily, that growing up amid concentrated disadvantage is bad for kids. Similar to the Johnson study of desegregation orders noted above, Chetty’s research demonstrates his point using siblings: When a family moves to a new area with better socioeconomic characteristics, the kids’ chances of upward mobility improve, and younger children are the most affected. And distressingly, while racial segregation seems to be declining of its own accord, economic segregation is rising.

Perhaps, then, we might get the benefits of integration — or at least a lot of those benefits — without using legally suspect race-conscious policies?

Like many victories of the civil-rights era, court-ordered school desegregation was an extraordinary measure justified by extraordinary circumstances. A system in which federal judges decide where children go to school, in place of local school boards, could not have lasted forever, and indeed was not intended to.

It is one thing to allow members of a community to challenge specific policies they believe to be discriminatory; it is quite another for the federal government to override local decision-making for decades at a time on the assumption that local officials can’t make decisions in a nondiscriminatory manner. The more time passes since the horrors of Jim Crow, the less plausible that assumption becomes, and the more problematic desegregation itself starts to look — even if school boards adopt it voluntarily — given that it involves sending children to one school or another based on the color of their skin.

Numerous court cases over the years limited the scope of the practice — especially in regard to segregation that had not resulted from government policy, as well as segregation between districts (instead of within them) — but the whole system began to collapse in 1991 with the Supreme Court’s decision in Board of Education v. Dowell. The Court emphasized that “federal supervision of local school systems [was always] intended as a temporary measure to remedy past discrimination.” A desegregation order should end, the Court found, if the school system is being “operated in compliance with the Equal Protection Clause” and it is “unlikely that the Board would return to its former ways” in the order’s absence.

In a dissent, Thurgood Marshall (joined by two others) protested that the Oklahoma school board at the heart of the case had maintained segregated schools for 65 years, 18 of which came after Brown v. Board — and that only 13 years had passed since a court had finally imposed a plan of its own. He further noted that, owing to the area’s residential segregation (in part the legacy of legally enforced housing covenants), ending the desegregation order could lead to the return of de facto one-race schools.

Perhaps these arguments should have carried the day in 1991. But another generation has passed since then, and at some point it becomes ludicrous to keep a local community under the federal government’s thumb as punishment for the sins of people who ran the community decades ago. This remains true even if the community’s neighborhoods might bear the mark of those sins well into the future, thanks to the tendency of residential patterns to persist over time. Sometimes it is simply not possible to fully erase the consequences of a past wrong.

The George W. Bush administration’s Justice Department helped to speed the end of desegregation orders by supporting localities’ efforts to be released from court oversight. And in 2007, the Supreme Court placed limits on voluntary integration plans as well. In Parents Involved v. Seattle School District No. 1, echoing decisions it had recently made regarding higher education, it ruled that schools may not collect information on individual students’ races and assign them to schools on that basis. However, schools were left free to consider race more generally, such as by ensuring that each school’s attendance zone includes a diverse group of neighborhoods.

The federal government also has some options for attacking residential segregation directly — options that the Obama administration availed itself of but the Trump administration has scaled back.

The 1968 Fair Housing Act has long been used (for example) to require suburbs to construct public housing in middle-class and wealthy neighborhoods or lose valuable federal funding, often through findings that facially neutral zoning laws were racially motivated or had a “disparate impact” on certain racial groups. In 2015, the Obama administration announced an even bolder approach to enforcing this law, debuting a plan to require municipalities and even whole metro areas to “take significant actions to overcome historic patterns of segregation, achieve truly balanced and integrated living patterns, promote fair housing choice, and foster inclusive communities that are free from discrimination.”

Writing for National Review Online, Stanley Kurtz called the plan “a de facto regional annexation of America’s suburbs.” “Once HUD gets its hooks into a municipality, no policy area is safe,” he explained. “Zoning, transportation, education, all of it risks slipping into the control of the federal government and the new, unelected regional bodies the feds will empower.” Ben Carson, President Trump’s secretary of housing and urban development, has effectively delayed the rule until at least 2020.

Not content just to coerce local governments into engineering integration, the Obama administration moved on to coercing the poor: The value of a housing voucher (issued through a program formerly known as Section 8) would now depend on the neighborhood the recipient chose to live in, rather than just the overall metro area. Those who lived in poor neighborhoods would actually see their vouchers cut, while more money would be given to those living in the expensive areas where regulators want more poor people to live. As the Wall Street Journal put it, the “announcement was greeted with outrage by an unlikely alliance of landlords, low-income tenant advocates and left-leaning local officials.” Carson suspended this plan, too, but the decision is being challenged in court.

Let’s be blunt: From a conservative perspective, and even from a simply practical one, there is no acceptable way for the government to fully integrate neighborhoods. There will always be the basic facts that some homes are more expensive than others, that for legitimate reasons (both aesthetic and social) neighborhoods tend to contain a limited range of housing types, that many people naturally gravitate toward others who are most like them, and that it will take decades for the marks of decades-ago redlining to fade.

Those racial marks are fading, if at an agonizingly slow pace, as we saw above. But economic segregation will always be with us — unless we double down on Obama’s project of forcing rich and middle-class neighborhoods to provide housing to people who can’t afford to live there. That project strikes a blow against local governments’ right to manage their own affairs, not to mention being a surefire route to a vicious backlash once the implications become clear to well-off voters of all races across the country. It is not a serious way forward.

But that’s not to say American housing policy is anything close to ideal — and many needed reforms could have the side benefit of spurring integration. Most pressingly, some of the nation’s most prosperous (and allegedly liberal) cities make it nearly impossible to build new housing, especially the tall apartment buildings that are appropriate to a thriving urban center, driving up prices and denying countless workers across the country the chance to move where the jobs are. Last year, the economists Chang-Tai Hsieh and Enrico Moretti estimated that these land-use restrictions “lowered aggregate US growth by more than 50% from 1964 to 2009.” This is rightfully a decision to be made by cities and states, not the federal government — but the tax benefits of recapturing this missing economic activity ought to be a large incentive to highly productive areas, especially now that their ability to hike taxes on existing residents has been limited by the recent cuts to the state-and-local-tax deduction. And as Frey notes, a city that is an economic magnet to new residents tends to be a racially integrated city.

In addition, federal policymakers should consider a reform once suggested by Eli Lehrer and Lori Sanders of the center-right R Street Institute: In areas with high unemployment and little work, let the jobless cash out their future unemployment benefits and use the proceeds to move somewhere better. Beyond helping people find jobs, this could help to break up concentrated poverty. In the same vein, cities should continue the trend of shifting away from housing projects and toward vouchers — but without placing the conditions on those vouchers that Obama tried to.

School integration, meanwhile, faces some of the same headwinds that housing integration does: There’s only so much that governments can do, and middle-class parents often fight back when their children are sent to school with less fortunate peers. But there are trends unfolding that highlight several routes to reform — ways to increase racial and economic integration by empowering families to make their own choices.

For libertarians and conservatives, vouchers have long been the holy grail of education-reform policy. And at least when they’re targeted toward lower-income students, vouchers do seem to increase integration. One can see this most easily, oddly enough, by looking at a report from Halley Potter of the liberal Century Foundation trying to throw cold water on it.

Potter freely admits that the voucher system of Washington, D.C., integrated schools, sending poor black children from public schools where they were overrepresented to private schools where they were underrepresented. Regarding a Louisiana program, Potter takes issue with a study claiming benefits for integration but then analyzes the data herself and finds the same thing, only less dramatically: Thirty-four percent of student transfers had unambiguously positive effects, meaning the kids went from schools where their race was overrepresented to schools where their race was underrepresented, while only 9 percent had unambiguously negative effects. (The rest had mixed effects — such as when a student’s race was overrepresented at both his original school and his new one.) Meanwhile, Potter cites a study finding that vouchers in Milwaukee had no effect in either direction on integration, though even that study noted that previous work had found positive effects.

Potter also raises concerns about a large voucher program in Indiana that began in 2011. This program has quite lenient criteria for participation — partial scholarships are now available to students whose families earn up to 370 percent of the poverty level — so, in theory, middle-class white parents might use it to pull their kids out of diverse public schools. The program is young enough that there’s been little scholarly work on its effects, but minority students are certainly taking advantage of it: They’re 30 percent of public-school students but 43 percent of those in the voucher program. In addition, the Huffington Post recently took a look at Fort Wayne, a city whose public schools were put under a desegregation order in the late 1980s — and today have arguably been “hit the hardest by the expanding use of vouchers.” The core finding: “Overall, private schools in Fort Wayne went from 84 percent white in 2011 to 74 percent white in 2017. Public schools went from 58 percent white to 55 percent white.” These are encouraging signs, to say the least.

Vouchers are slowly gaining traction, but as of 2015, only a few hundred thousand kids nationwide used them — and the research is mixed on whether these programs boost student achievement. So it’s also worth looking at less libertarian but more popular variations on the school-choice concept, such as charter schools, which now enroll about 3 million children and have repeatedly been shown to meaningfully improve kids’ outcomes, particularly in inner cities. Charter schools are independently operated public schools; they do not charge tuition, they admit children by lottery if they don’t have spaces for everyone who wants to attend, and they often face substantial regulation even though they are free from the normal public-school bureaucracy.

Like vouchers, charter schools often face dubious charges of exacerbating segregation. Nationwide, charter schools do have higher concentrations of poor and minority students than public schools have — but this is no surprise, because charters are located overwhelmingly in urban areas and are often intended specifically to serve the poorest students. If you compare charter schools with the public schools their students would otherwise have attended, you find there isn’t a difference in segregation.

And it would be entirely possible for charter schools to encourage integration rather than merely not hinder it. In another Century Foundation report, Potter and a co-author point out that state policies often incentivize charters to serve low-income populations. Relaxing these incentives — and locating charters in places where they’re accessible to a broader range of students — could turn them into forces for integration.

Failing that, cities can simply let students choose among public schools rather than requiring all students to attend a school based on where they live. Slots in oversubscribed schools can be handed out by lottery, possibly after admitting those who live nearby — so that all students are guaranteed access to their local school — or who have siblings already attending. (This was the context for the Seattle Supreme Court case mentioned above: Individual students’ classifications as “white” or “nonwhite” were used as one of these tiebreakers, and were given higher priority than how close to the school a student lived.)

To abandon the project of government-engineered racial balancing, then, is not to abandon the quest to ensure that all children have access to good schools. Indeed, it is not even to abandon the goal of integration. Americans of different races are, of their own volition, living side by side more and more as time goes on. And we can coax this trend along by helping people achieve their preferences, rather than by overriding those preferences.

Editor’s note: This piece has been emended since its original posting. The Dowell case originated in Oklahoma, not Kentucky, and Raj Chetty has moved from Harvard to Stanford.

In This Issue



Books, Arts & Manners


The Week

The Week

 In his new official portrait, as in his presidential rhetoric, Barack Obama just couldn’t quite get clear of the Bushes.

The Week

In his new official portrait, as in his presidential rhetoric, Barack Obama just couldn’t quite get clear of the Bushes.

Most Popular

Politics & Policy

Goodbye, Green New Deal

What will happen next with the coronavirus epidemic is unknown, but it seems certain to claim one very high-profile victim: the so-called Green New Deal. Good riddance. The current crisis in the U.S. economy is, in miniature but concentrated form, precisely what the Left has in mind in response to climate ... Read More
Politics & Policy

Goodbye, Green New Deal

What will happen next with the coronavirus epidemic is unknown, but it seems certain to claim one very high-profile victim: the so-called Green New Deal. Good riddance. The current crisis in the U.S. economy is, in miniature but concentrated form, precisely what the Left has in mind in response to climate ... Read More