The Dream Revisited

Race Remains the American Dilemma

by Richard Rothstein | March 2014

Charles Clotfelter describes two definitions of “segregation.” I think it preferable to restrict the term to his first use, a “system of state enforced regulations as under Jim Crow in the U.S. or under apartheid in South Africa.” For statistics describing sub-population concentrations, we should use terms like “dissimilarity” to avoid distraction from the important phenomenon of segregation itself.

Preserving this distinction makes it possible to have a conversation with millions of Americans who, following Chief Justice John Roberts and his conservative colleagues, won’t call dissimilarity “segregation,” instead labeling it racial “imbalance,” or what is commonly termed “de facto segregation.” Roberts himself avoids the de facto phrase, apparently considering it an oxymoron – if it is de facto, then it is not segregation.

In denying the Louisville and Seattle school districts a right to pursue explicit racial integration (in Parents Involved in Community Schools, 2007), Roberts wrote: "The distinction between segregation by state action and racial imbalance caused by other factors has been central to our jurisprudence … 'Where [racial imbalance] is a product not of state action but of private choices, it does not have constitutional implications'." Roberts went on to say that schools in the two cities were imbalanced because they are located in imbalanced neighborhoods; racial housing patterns in these cities might result from "societal discrimination," but remedying discrimination "not traceable to [government's] own actions" can never justify racial classifications of students.

The public, including scholars and liberal policymakers, has largely accepted Roberts’ distinction between de jure segregation (by government action) and imbalance (statistical dissimilarity). But even if we accept the legitimacy of this distinction, the facts don’t fit it the way Roberts supposes. What was unsettling about the Parents Involved opinions was not Roberts’ view, entirely expected, but Justice Stephen Breyer’s dissent which, while otherwise stirring and compelling, nonetheless accepted that segregation in Louisville and Seattle (and elsewhere) is indeed mostly de facto. Breyer’s argument was that where de facto segregation exists, schools should be permitted to integrate even if they cannot constitutionally be compelled to do so.

But in fact, our schools in Louisville, Seattle and elsewhere are racially imbalanced not because their neighborhoods are de facto imbalanced, but because those neighborhoods were segregated by government policy whose effects endure, structuring the residential opportunities of African Americans.

In some small cities, and in some racial border areas, some racial school integration can be accomplished by adjusting attendance zones, establishing magnet schools, or offering more parent-student choice. This is especially true – but only temporarily – where neighborhoods are in transition, either from gradual urban gentrification, or in first-ring suburbs to which urban ghetto populations are being displaced. But generally, our most distressed ghettos are too far distant from truly middle-class communities for school integration to occur without racially explicit policies of residential desegregation.

Although Justice Breyer argued that racial integration should be permitted as a remedy even where schools are not legally segregated, his acceptance of the de facto notion lends unwitting support to policymakers hoping to avoid difficult (but necessary) discussions of race by advocating plans of socioeconomic rather than racial integration. They hope that this substitution of class for race can dissolve the perpetual “American dilemma” of a racial caste system, without arousing racist opposition.

But it won’t work. Although race and poverty are correlated, they remain quite different, both in degree and in geographic distribution. As Paul Jargowsky shows (see, for example, his recent Century Foundation report, Concentration of Poverty in the New Millenium), black poverty is more concentrated than white poverty: in 2011, seven percent of poor whites lived in high poverty neighborhoods, while a breathtaking 23 percent of poor blacks lived in such neighborhoods. As Patrick Sharkey shows in his 2013 book, Stuck in Place, if African American parents themselves grew up in high poverty neighborhoods, their children are likely to have the same debilitating experience, but if white parents grew up in high poverty neighborhoods, their children are likely to escape that environment and live in middle class neighborhoods.

These differences are frequently obscured in education policy, to unfortunate effect. In most cases, the only socioeconomic data available to policymakers is whether students are eligible for free or reduced-price lunch. This category encompasses both poor and near-poor families with incomes up to 135 percent of the poverty line (and eligible for free lunches), and low-income working class families with incomes up to 185 percent of the poverty line (and eligible for reduced-price lunches). In consequence, “lunch-eligible” vs. “non-eligible” is not a good proxy for black vs. white; it is not even a very good proxy for disadvantaged vs. advantaged. If “free-lunch eligible” could be separated from “reduced-price eligible,” it would be a closer proxy, but very few states or districts report separate data for free as distinct from reduced-price eligibility.

Analysts observe that lunch-eligible black children achieve at lower levels than lunch-eligible white children, and too often conclude that the black-white academic achievement gap cannot be driven by socioeconomic differences but must result from incompetent (or racist) teachers and school systems. The conclusion is unfounded. While some teachers and school systems are incompetent (in middle-class as well as in disadvantaged neighborhoods) poor black children are hobbled by much more serious poverty, and much more concentrated and multigenerational poverty, than poor whites. Even if all had highly qualified teachers, poor (free-lunch eligible) black children will have lower average performance than low-income (reduced-price eligible) white children because of the greater educationally-relevant disadvantages with which typical free-lunch eligible children come to school.

For instance, low-income black children are more likely to be mobile, with unstable housing, than low-income whites; they are more likely to come from single parent households, and their parents are likely to have lower educational attainment. As Paul Jargowsky and Patrick Sharkey show, they are more likely than low-income whites to live in concentrated poverty in urban neighborhoods (and for multiple generations), and so are more likely to suffer from lead poisoning and asthma, more likely to be stressed (and have stressed parents) from environments where crime, drug dealing, and violence is more prevalent, and will have families with less access to nutritious food, primary care physicians, and opportunities for safe physical activity. All these characteristics, and many more, make it more difficult to make educational progress, all other conditions (such as school quality) being equal.

We will not accomplish much reduction of the black-white educational achievement gap if we pretend it only reflects socioeconomic differences that are so grossly mismeasured by the free- and reduced-price lunch category.

There is a second unintended and unfortunate consequence of relying on lunch status for policy purposes. Pro-integration liberals have increasingly proposed balancing lunch-eligible and middle class students in schools. Many hope to accomplish racial integration by this device, while avoiding both Supreme Court rejection and white families’ racially motivated opposition.

But the device will, in many cases, accomplish less racial integration than is being sought. Because lunch-eligible African American children are more likely to be poor or close to it, while lunch-eligible white children are more likely to be working class with low incomes, some share of the effort expended to integrate by socioeconomic status will end up not integrating black with white students, but integrating white working class (including Hispanic working class) with white middle class students. This is certainly a desirable goal, but not the same as racial integration.

To accomplish more, we will have to be race-conscious. And before we can expect the public to consider this, we must first discredit the widely-accepted notion that residential racial imbalance is de facto, or as the late Justice Potter Stewart once reflected, the result of “unknown and perhaps unknowable factors such as in-migration, birth rates, economic changes, or cumulative acts of private racial fears.”

In truth, nationwide metropolitan residential racial segregation today results largely from the ongoing effects of racially explicit government policies, quite similar to Jim Crow in the U.S. South and apartheid in South Africa. These government policies, although no longer explicit, had and continue to have enduring effects.  Policies of de jure residential segregation (within the Chief Justice’s meaning--intentional and racially explicit) have been well-documented but largely forgotten, including by many contemporary advocates of racial equality. It is urgent that we refamiliarize ourselves with this history because if we become aware of how African American ghettos were government-created and sustained, we will conclude that residential integration is constitutionally mandated--opinions of the Chief Justice and Justice Breyer notwithstanding. With that conclusion, we may be more willing to challenge the conventional wisdom that racially explicit remediation is to be avoided, and less willing to seek socioeconomic proxies to avoid speaking about race.

Here are some ways that government created de jure segregation:

From its New Deal inception and especially during and after World War II, federally funded public housing was explicitly racially segregated, both by federal and local governments. Not only in the South, but in the Northeast, Midwest, and West, projects were officially and publicly designated either for whites or for blacks. Some projects were “integrated” with separate buildings designated for whites or for blacks. Later, as white families left the projects for the suburbs, public housing became overwhelmingly black and in most cities was placed only in black neighborhoods, explicitly so. This policy continued one originating in the New Deal, when Harold Ickes, President Roosevelt’s first public housing director, established the “neighborhood composition rule” that public housing should not disturb the pre-existing racial composition of neighborhoods where it was placed.
Once the housing shortage eased and material was freed for post-World War II civilian purposes, the federal government subsidized relocation of whites to suburbs and prohibited similar relocation of blacks. Again, this was not implicit, not mere “disparate impact,” but racially explicit policy. The Federal Housing and Veterans Administrations recruited a nationwide cadre of mass-production builders who constructed developments on the East Coast like the Levittowns in Long Island, Pennsylvania, New Jersey, and Delaware; on the West Coast like Lakeview and Panorama City in the Los Angeles area, Westlake (Daly City) in the San Francisco Bay Area, and several Seattle suburbs developed by William and Bertha Boeing; and in numerous other metropolises in between. These builders received federal loan guarantees on explicit condition that no sales be made to blacks and that each individual deed include a prohibition on re-sales to blacks.
The federal interstate highway system was constructed through urban areas with the explicit purpose of razing and then relocating black neighborhoods to be far from white neighborhoods, or of creating barriers between white and black neighborhoods.
Public police and prosecutorial power was used nationwide to enforce racial boundaries. Illustrations are legion. In the Chicago area, police forcibly evicted blacks who moved into an apartment in a white neighborhood; in Louisville, the locus of Parents Involved, the state prosecuted and jailed a white seller for sedition after he sold his home in his white neighborhood to a black family. Everywhere, North, South, East, and West, police stood by while thousands (not an exaggeration) of mobs set fire to and stoned homes purchased by blacks in white neighborhoods, and prosecutors almost never (if ever) charged well-known and easily identifiable mob leaders.

Other forms abound of racially explicit state action to segregate the urban landscape, in violation of the Fifth, Thirteenth, and Fourteenth Amendments. Yet the term “de facto segregation,” describing a never-existent reality, persists among otherwise well-informed advocates and scholars. The term, and its implied theory of private causation, hobbles our motivation to address de jure segregation as explicitly as Jim Crow was addressed in the South or apartheid was addressed in South Africa.

Many state policies to enforce residential segregation ended a half-century or more ago. Can we still consider racially separate neighborhoods de jure? I think yes. Equally explicit racial labor market policies ensured that in the mid-twentieth century when the white working and middle class grew in wealth and income, similarly situated African Americans were denied those opportunities. Government certification for exclusive collective bargaining of unions that openly excluded black workers is one example. Denial for explicitly racial reasons of minimum wage coverage to occupations in which black workers predominated is another. In consequence, most black families, who in the mid-twentieth century could have joined their white peers in the suburbs, can no longer afford to do so. Highways that were once constructed to separate white from black neighborhoods remain in place. A history of state-sponsored violence to keep African Americans in their ghettos cannot help but influence the present-day reluctance of many black families to integrate.

Today, when facially race-neutral housing or redevelopment policies have a disparate impact on African Americans, that impact is inextricably intertwined with the state-sponsored system of residential segregation that we established.   

Avoidance of our racial history is pervasive. We ensure the persistence of that avoidance for subsequent generations. In over 1,200 pages of McDougal Littell’s widely used high school textbook,The Americans, a single paragraph is devoted to 20th century “Discrimination in the North.” It devotes one passive-voice sentence to residential segregation, stating that “African Americans found themselves forced into segregated neighborhoods,” with no further explanation of how public policy was responsible. Another widely used textbook, Prentice Hall’s United States History, also attributes segregation to mysterious forces: “In the North, too, African Americans faced segregation and discrimination. Even where there were no explicit laws, de facto segregation, or segregation by unwritten custom or tradition, was a fact of life. African Americans in the North were denied housing in many neighborhoods.” History Alive!, a popular textbook published by the Teachers Curriculum Institute, teaches that segregation was only a Southern problem: “Even New Deal agencies practiced racial segregation, especially in the South,” failing to make any reference to what Ira Katznelson, in his 2013 Fear Itself, describes as FDR’s embrace of residential segregation in return for Southern support of his economic policies.

Socioeconomic integration should certainly be part of a broader assault against economic polarization. But even if this assault made modest progress, geographically separate black communities could be left behind. We are more likely to diminish racial segregation if we reacquaint ourselves with a history that imposes a constitutional obligation to dismantle this de jure legacy.

Richard Rothstein is a research associate of the Economic Policy Institute and senior fellow of the Chief Justice Earl Warren Institute on Law and Social Policy at the University of California (Berkeley) School of Law.

More in Discussion 2: Economic Segregation in Schools