Manhattan Institute report heralding the “end” of segregation uses a measure that masks important demographic and economic trends

In a study released this week, two Manhattan Institute researchers heralded the “end of the segregated century.” Harvard professor Edward Glaeser and Duke professor Jacob Vigdor showed that African American segregation levels have now declined to their lowest point since 1920, just after the beginning of the “Great Migration” of rural sharecroppers from the South to Northern industrial metropolitan regions.

From 2010 Census data, professors Glaeser and Vigdor calculate changes in what sociologists term “dissimilarity indices.” They find a national dissimilarity (or segregation) rate of about 55 percent for African Americans—in other words, “only” 55 percent of African Americans would now have to move to neighborhoods with more non-blacks in order to evenly distribute the black population throughout all neighborhoods in their metropolitan areas. This is a substantial decline from the segregation level of about 80 percent in 1970.

The report attributes this success primarily to legal prohibitions on housing discrimination. The Manhattan Institute aggressively promoted its report, which was then heavily covered in the media with headlines like “Segregation Curtailed in U.S. Cities, Study Finds” and “Study finds black segregation lowest in century,” and with celebratory op-eds by the report’s authors (“Desegregation Is an Unsung U.S. Success Story”;  “Why Your Block is More Integrated”).

But the celebrations are premature. Although the Manhattan Institute’s Census data were accurate, a 55 percent dissimilarity rate can hardly be called the “end” of segregation. And segregation can only get worse, not better. Because the epidemic of foreclosures has disproportionately affected African Americans, many blacks who were able to move to predominantly white neighborhoods in the last decade will undoubtedly relocate back to poorer and more racially isolated black neighborhoods. This likely reversal of dissimilarity decline will only show up in future census analyses.

More important, however, dissimilarity indices do not describe what most people mean when they think of a segregated society, and they don’t point policymakers to the most critical problems facing the country which, by more relevant measures, is becoming more segregated, not less.

Recent declines in dissimilarity have had complex causes: One is that low-income Hispanic (and in some regions, Asian) immigrants have moved into neighborhoods that previously were mostly black. This reduces the proportion of blacks in those neighborhoods (and thus causes a metropolitan area’s dissimilarity index to fall) but does little to integrate African Americans into white neighborhoods.

For policy purposes, a more appropriate index of segregation than dissimilarity is an index that describes the “exposure” of African Americans to the majority white population. By this measure, segregation is today greater than it was in 1940, and has remained mostly unchanged since 1950.  As John Logan and Brian Stults of Brown University’s US2010 Project have shown, in 1940, the average black lived in a neighborhood that was 40 percent white. In 1950 it fell to 35 percent—where it remains today. This average, of course, aggregates data from many neighborhoods where blacks have virtually no exposure to whites, and others where integration is advanced. Nonetheless, by this measure there has been no progress in reducing segregation for the last 60 years.

Another cause of reduced dissimilarity is indeed cause for celebration, but also has an unintended side-effect with dire consequences: The growth of the black middle class and the decline of housing discrimination has permitted more middle-class African Americans to flee the ghetto, frequently to inner-ring suburbs that are less homogenously black than the ghetto, although sometimes in transition from a predominantly white to a larger black population.

But as a result, inner-city ghettos are left without a middle class and are more homogenously poverty-stricken and hopeless. Some 20 years ago, William Julius Wilson observed the beginnings of this phenomenon and termed the remaining socially isolated ghetto residents the “truly disadvantaged.” Wilson noted that as urban industry and the black middle class departed, the lack of access to jobs in such neighborhoods led to greater single parenthood (because fewer men could support families) and left children without conventional (middle-class) role models. A few years later, in another now-classic work of sociology, Douglas Massey and Nancy Denton observed that, despite urban deindustrialization, white workers had little difficulty following jobs to the suburbs, but blacks were barred from doing so by the discriminatory policies of federal, state, and local governments. As a result, only blacks were left to deal with “falling retail demand, increasing residential abandonment, rising crime, spreading disorder, increasing welfare dependency, growing family disruption, and rising educational failure [that were] all concentrated simultaneously” in such segregated neighborhoods.

Indeed, a recent study of school reform in Chicago concluded that although much could be accomplished in schools serving disadvantaged students outside neighborhoods of such concentrated poverty, there is little hope of improving children’s education in “truly disadvantaged” neighborhoods.

The Chicago study, by Anthony Bryk and colleagues, found that many schools serving disadvantaged children that had well-developed and aligned curriculum, featured collaboration between teachers and principals, and made a concerted effort to involve parents and the community made substantially greater progress than schools without these characteristics. But Bryk and his investigators, to their great surprise, discovered that well-designed reform programs made little or no difference in schools serving neighborhoods of concentrated poverty where nearly all students moved frequently, were African American, and had low-income parents with relatively little formal education and a high likelihood of unemployment. These communities had high crime rates and inadequate community supports (such as health care and social service providers), and few adults exemplified the benefits of educational attainment. In communities of concentrated disadvantage, addressing these contextual factors was essential before school reform could take root. The investigators concluded, “Our findings about schooling in truly disadvantaged communities offer a sobering antidote to a heady political rhetoric arguing that all schools can be improved.”

And such neighborhoods have been increasing, not decreasing, in number. Another recent study of census data published by the Joint Center for Political and Economic Studies finds that over 20 percent of all African Americans now live in “high poverty” neighborhoods, unchanged from 2000. More than 40 percent of poor African Americans now live in high-poverty neighborhoods, compared to 15 percent of poor whites who live in such neighborhoods. Poor blacks are therefore nearly three times as likely to be “truly disadvantaged” as poor whites. (The Joint Center defines a “high poverty” neighborhood as one where 30 percent or more of the residents have incomes below the poverty line, but this definition can be misleading: The poverty line is very low, and neighborhoods with poverty rates of greater than 30 percent also inevitably house large numbers of residents whose incomes are barely above the poverty line, and whom most would also consider to be severely economically disadvantaged.)

Further reinforcing the point that a stark socioeconomic divide exists between African Americans and whites, Stanford University sociologists Sean Reardon and Kendra Bischoff, also using census data, have shown that income segregation among black families is now 60 percent greater than among white families (i.e., low-income black families are significantly more likely to live in neighborhoods dominated by other low-income black families) and has risen at a rapid rate in the last decade. Because poor African Americans once shared the ghetto with middle-class African Americans, income segregation among blacks was actually less than income segregation among whites in 1970. But it surpassed the white rate shortly afterwards, and has climbed ever since. Low-income black families are now much more isolated from middle-class black families than low-income white families are isolated from middle- and even high-income white families.

A special caution to the hyperbole surrounding the Manhattan Institute report, is the observation of Reardon and Bischoff that middle-class black families are still “much more likely to live in neighborhoods with low-income white neighbors than are comparable middle-class white families.”

These discouraging trends partly reflect growing economic inequality in the nation as a whole—now compounded by the disproportionate harm suffered by African Americans during the post-2007 economic collapse. Partly, persistent black segregation (defined as lack of exposure to whites and ghettoized concentration of black poverty) stems from exclusionary zoning laws adopted by most suburbs in the early to mid-20th century, with the not-so-disguised purpose of keeping those suburbs lily-white. As the Brookings Institution’s Jonathan Rothwell recentlyobserved, unless a concerted effort is made to force these suburbs to abandon such zoning and permit the construction of low- and moderate-income housing, there is little hope of reversing these trends.

But the reality that low-income whites are much more integrated into the middle-class population than are low-income blacks suggests that, declines in dissimilarity notwithstanding, an attack on exclusionary zoning—while necessary—will not alone be sufficient to desegregate the nation. Race-conscious policy remains necessary to undo the effects of racial residential rules established over the course of a century. Promoting the “end” of segregation, as the Manhattan Institute has done, can only undermine the political will that must be mobilized to embark on this course.