By john a. powell & Stephen Menendian
January-March 2016 Issue of Poverty & Race
Segregation as a problem is gripping our cultural consciousness for the first time in a long while, beyond the domain of social scientists and city planners. David Simon (the creator of the acclaimed HBO series, “The Wire”) wrote and produced a mini-series for HBO this past year, “Show Me a Hero,” illustrating the dynamics of segregation in the post-civil rights period. Set in the late 1980s and early 1990s, the series visibly demonstrated the realities of metropolitan segregation, and the resistance to integration in both economic and racial terms through the lens of a fight over scattered-site public housing in Yonkers, New Jersey.
That fight had echoes in the Supreme Court’s decision last summer in Texas Department of Housing and Community Affairs v. The Inclusive Communities Project, Inc., illustrating how such issues are still a battleground. In its landmark decision, the Court affirmed the right of Dallas public housing residents to sue the state under a “disparate impact” theory that the government segregated low-income housing in the Dallas metropolitan area. In the process, the Court acknowledged the continuing realities of residential segregation across the
United States. (Full disclosure: The authors’ team contributed an amicus brief cited by the Court on this point, and john powell is on the board of the Inclusive Communities Project, Respondent).
Segregation from the 19th Century to the 21st
But how are we to understand segregation? Why does it matter? What does it look like, and how do we measure it?
Segregation in the nineteenth and twentieth centuries was relatively easy to understand. Segregation in the nineteenth century was not primarily a housing phenomenon, but was instead institutional and ingrained in everyday life. There was clarity on the divide between Black and white space and behavioral norms, even as people lived and worked in close physical proximity. Courthouses, theaters, restaurants, hotels, factories, buses, and schools were segregated by law and culture, but Blacks and whites lived in the same cities and towns; municipalities, jurisdictions, and neighborhoods were not needed to do the work of separating. Blacks and whites lived physically close and socially and culturally far apart. Segregation was embedded not only into law, but culture and commonsense understanding. Although this was most visible in the South, it was also true in many respects in the North.
In the twentieth century, segregation evolved and took a different form that centered on housing. By the 1920s, the federal government began playing a much larger role in the housing market, and, for the first time, with federal direction and funding, housing segregation took root and proliferated, especially in the post-war era, even as the other institutionalized forms of segregation were being dismantled in the courts and elsewhere. In a sense, segregation morphed from occurring in schools and public transportation to city blocks and suburban lots.
In some ways, the effort to integrate public spaces and dismantle formal segregation under law is what triggered a countervailing effort to re-inscribe segregation in our institutions, structures, and living patterns. The government played a key role in both movements. The challenge to Jim Crow laws in the courts and in other parts of the federal government was paralleled by federal investments in residential infrastructure that became the locus for new expressions of segregation. Federal support was tied to the promotion and acquiescence of jurisdictional structures that created a racially segregated housing market.
At the same time, the attack on segregation was narrowed to explicit and de jure expressions. This permitted rampant housing segregation that ultimately translated into de facto educational and opportunity segregation. When the Supreme Court ruled in 1974 that courts could not de-segregate schools across district lines without a demonstration of discriminatory intent, it was a signal and invitation to white flight and an escape hatch from the Brown mandate. Whites could obtain the benefits of opportunity hoarding without asserting a desire for racial separation.
Although housing remains the locus of segregation in the twenty-first century, we have witnessed notable shifts and new complexities emerging. By the early 2000s, there was increasing evidence of urban change. Poverty was growing in the suburbs, but especially inner-ring suburbs. A growing number of non-white families moved into these suburbs, and led to additional white flight. Inner-ring suburbs were replaced by second ring or outer-ring suburbs, as the wealthy continued to move further toward the periphery. As a result, poverty began to migrate to the suburbs, and by the 2000s, there was more poverty than ever in the suburbs. Metropolitan demographics may be shifting yet again.
The United States in the twenty-first century is a more diverse, multi-racial and multi-ethnic society than in the twentieth century, and while our metropolitan regions remain deeply segregated, the patterns of segregation are taking new, and deeply troubling, turns. Segregation today is more complex than a simplistic portrait in Black and white terms or in terms of residence. It turns out that understanding segregation, its harmful effects, and how to measure it, is more complicated than ever.
Segregation is both an intuitive concept and a slippery one. As noted, we understand segregation as living apart. Accordingly, Massey and Denton define segregation as “the degree to which two or more groups live separately from one another, in different parts of an urban environment” (1988). Yet it’s clear that people may be segregated in varying ways, as we have already seen. Even in terms of residence, people may be segregated across a municipal border, a geographic barrier or they may be clustered or concentrated within a larger region. Similarly, groups may be clustered and concentrated or scattered in clusters.
Segregation is not only a complex phenomenon, but measuring it has become more complex and challenging as well. The measures of segregation that have helped us understand Black-white segregation by residence may no longer be the best measures to understand segregation in an increasingly diverse, multi-racial society. Understanding the strengths and limits of such measures helps illuminate the full extent to which we continue to “live apart.”
The most well-known and widely used measure of segregation is the dissimilarity index. The dissimilarity index measures how evenly various racial groups are spread across neighborhoods within metropolitan areas. A dissimilarity index score indicates the percentage of a subgroup that would have to move to achieve integration. A score of 100 indicates that every neighborhood has residents of only one particular group (“complete segregation”), whereas a score of zero indicates proportional representation of each group throughout the metropolitan region (“complete integration”).
According to the dissimilarity index, Black and white segregation peaked in 1970 at 81.4, and has been in a gradual decline every decade since. Since then, it has fallen to 73.0 (in 1980), 67.2 (in 1990), 63.1 (in 2000), to 59.0 in 2010. If one looked only at dissimilarity values since 1970, one might conclude that the United States was well on its way towards residential integration. However, this simple portrait of segregation largely misses the mark.
First and foremost, the rate of decline between each decennial census has fallen significantly. Each decennial census indicates less progress. Between 1970 and 1980, dissimilarity scores fell by nearly nine points. However, between 2000 and 2010, it fell by three. This might not be so troubling if it weren’t for a second fact, that the measured rate of segregation in 1970 is considered very high, not far from that measured in South Africa under apartheid. Our current measure of segregation, 59.0, is still considered a very high level of segregation. In real world terms, it means that more than half of African Americans (or whites) would have to move residence to achieve a fully integrated society. Modestly improving dissimilarity index scores should not mask the persistence of segregation in our society.
Critically, however, segregation scores as measured by dissimilarity index scores were dramatically better in the first half of the twentieth century, with values of 30.9 in 1910, 43.2 in 1920, and 58.1 in 1930. In other words, the United States today is more segregated than it was in the 1920s, a period associated with lynching and Jim Crow terrorism.
Although the dissimilarity index remains the most widespread measure of segregation, by the mid-1970s, social scientists were beginning to question whether it was the best measure for a complex phenomenon. Although the dissimilarity index indicates the distribution of populations within a given area, it cannot tell us much about the particular patterns within that geographic area or region, such as whether populations are clustered, concentrated, or scattered.
Perhaps most importantly, improvements as measured by the dissimilarity index could be attributed, for example, to the small numbers of affluent and middle-class Black families that have moved into traditionally segregated and predominantly white neighborhoods, but tell us little about the overall integration of most African-Americans.
Another measure of segregation seeks to measure the “exposure” of groups rather than their evenness by looking at whether groups share a common residential area—not just how well distributed they are throughout an area. The two primary measures of exposure, the interaction and isolation indices, measure flips sides of the same coin: the degree of exposure of the average minority group member to the average majority group, and the degree of isolation as measured by the average exposure of minority members to the same group. Using this approach, as of 2010, the average white resident of a metropolitan area resides in a neighborhood that is 75.4 percent white, 7.9 percent Black, 10.5 percent Hispanic, and 5.1 percent Asian. In contrast, a typical African-American resident lives in a neighborhood that is 34.8 percent white, 45.2 percent Black, 14.8 percent Hispanic, and 4.3 percent Asian. Although understated by the dissimilarity index, these figures reflect demographically different worlds.
Since the average African American lives in a neighborhood that is 35 percent white, the exposure score of African Americans to whites is 35. In contrast to the dissimilarity index, exposure index scores over the last 70 years suggest an inverse pattern. Segregation, as measured by exposure, declined from 35 in 1950 to 31 in 1980, but has slowly risen to pre-1950 levels, suggesting that although our metropolitan areas may appear more integrated, actual neighborhoods are as segregated as they were more than a half century ago, before the civil rights era.
As serious as these concerns are, both the exposure indices and the dissimilarity index suffer from another significant deficiency, in that they can only be used to measure the distribution of two groups relative to each other. When trying to measure levels of segregation in a multi-racial society, different measurements are needed.
Perhaps the best measure of segregation for multiple groups is the entropy index, which measures the average difference between a group’s proportion within a geographic area compared to that of a larger region as a whole. For example, you would be able to use this index to measure the degree of segregation of all racial subgroups within a neighborhood, community or municipality compared to the larger metropolitan area or region. Another advantage of this approach is that it does not depend on the overall size of the subgroup, and therefore indirectly measures concentration as well.
Recent data shows that seventy-five percent of African Americans in the country live in only 16 percent of the Census Block Groups in the United States. As a corollary, 30 percent of African Americans live in Census Block Groups that are seventy percent African American or more. Since racial groups tend to be heavily concentrated, the entropy index may be a superior measure of segregation.
A Segregated Society
Some measures reflect segregation that is more pronounced than in the past, and others suggest gradual improvements. Regardless of the particular measure used, the data reflects a society that is highly and persistently segregated nearly 50 years since the passage of the Fair Housing Act. The growth of our metropolitan regions and the complexity of our racial dynamics masks considerable levels of persistent segregation.
In particular, it appears that people are segregated most visibly across municipal borders rather than within them. In an amicus brief we filed in the most recent Fisher case before the Supreme Court (supporting the continuing use of affirmative action as a consideration in collegiate admissions), we mapped inter-district school segregation in Texas, and the results were stunning. Despite the fact that Texas has no racial or ethnic majority (non-Hispanic Whites constitute 44 percent of the population), only 13 percent of the 1,024 school districts in the state have no racial majority. Nearly 40 percent of those districts have a racial/ethnic supermajority, meaning that more than 75 percent of the students are members of the majority race of that district. A regional analysis might understate or overlook entirely this lived reality of segregation.
Although inter-district segregation is persistent and severe, perhaps the most disturbing expression of segregation is visible only at the regional level. The “middle-out” patterns of urban change that defined the twentieth century—the movement from urban centers to the suburban and exurban periphery—is reversing and contracting, in what some call “the great Inversion.” The demand for urban residential housing by millennials and young professionals is leading to rapid gentrification in many of our major urban metropolises.
The effort of many lower- and middle-income residents to escape urban blight between 1990-2010 led the wealthier to peel away into second ring and outer ring suburbs. But what was once an escape route has now become a trap.
Despite our cultural diversity, we inhabit a segregated society. Even as our metropolitan regions become more diverse, the lived experience of segregation persists as we balkanize into different neighborhoods and communities. Why does this matter?
Segregation matters most because it shapes and reflects the distribution of opportunity. Some have suggested it is not necessary for blacks and other people of color to live close to whites. But segregation has never been simply about the sorting of people, but about sorting of opportunity and the resources that shape life chances. When this is considered it not just interracial contact we are concerned about when we talk about segregation and integration, but the full breadth of opportunity properly understood.
Segregation occurs by race and income in a dynamic, interconnected pattern. Because we live in a segregated society, families and groups not only do not live together, but they are not invested in each other. Segregation is a mechanism for what Charles Tilly called “opportunity hoarding.” Segregation ensures that group membership correlates to artificial geographic and institutional boundaries. If white families live in one school district, and black families live in another, even if they are adjacent, then they are not invested in the success of the other.
So what should we do? First, we must better understand the dynamics and damage of segregation both on excluded and marginalized groups and the overall health of society in order to develop more effective interventions.
Second, we must develop and refine research methods that can support effective interventions and responsive policies. We have worked on several such approaches. Since segregation is in part about opportunity, we have focused on the structuring and access to opportunity. This can be demonstrated through opportunity mapping. This was a methodology used in Thompson v. HUD, one of the largest public housing cases of the last generation, and the underlying methodology is being adopted by HUD as part of its new AFFH rules. It is too early to know how this will in fact work. But we know that hoarding by neighborhood or jurisdiction is a problem.
Since segregation is about exclusion, we must not only understand how exclusion happens, but how we might achieve inclusion. Instead of just focusing on structural and cultural discrimination, we should be also be focusing on structural and cultural inclusion. This recognition informs the Haas Institute Inclusiveness Index, which we have been developing as a diagnostic instrument to better understand why certain regions or countries fare better in terms of inclusion across group boundaries.
We should affirmatively fashion our structures to work to support the outcomes we seek. For example, there is a great deal of research that supports the importance of racial, economically- and ability-integrated schools. We should look at how our housing programs and the siting of public housing or use of federal funds gives families and students access to high-performing, well-integrated schools, or why it fails to do so. We should also look at jurisdictional arrangements through a similar lens. Our courts too often accept and fail to disrupt practices that perpetuate segregation or that would permit remedial action. Our goal must be to move beyond legal schemes limited to intent and make our structures and culture affirmatively address the “othering” performed by segregation and exclusion from opportunity.
john powell , a PRRAC Board member, directs the Haas Institute for a Fair and Inclusive Society at the Univ. of California-Berkeley. firstname.lastname@example.org
Stephen Menendian is Assistant Director and Director of Research, of the Haas Institute for a Fair and Inclusive Society at the University of California, Berkeley.