The Policy Lessons of Partial Desegregation

In April 1968, when Congress for the first time enacted a broad fair housing law in the United States, the degree of housing segregation experienced by African Americans was staggering. The simplest precise measure of segregation is the index of dissimilarity, which analyzes the degree to which two groups (for example, blacks and whites) in a large region (say, a city or metropolitan area) live in different small neighborhoods (say, a block or census tract). This index shows that in 1960, about 88 percent of urban blacks living in representative major cities would have to move to a different block to achieve the same residential distribution as urban whites.1 Moreover, most of the remaining 12 percent of blacks were moving into neighborhoods on the fringes of existing black districts that currently had white residents but that would soon resegregate.2 Nationally, only 2 percent or 3 percent of urban blacks lived in what could be fairly regarded as stably integrated neighborhoods.3

Matters had not changed perceptibly by April 1970, when the next decennial census was taken.4 This was not so surprising; the Fair Housing Act5 (FHA) had gone into full effect only at the end of 1969, and it was generally understood that under the best of circumstances, it would take some time for enforcement efforts to gear up and have an effect.6 But when the 1980 census rolled around, it, too, showed very high levels of segregation.7 Integration had indeed increased (in some places, as we shall see, rather dramatically), but in many major urban areas the drop in the index of dissimilarity amounted to only a few percentage points.8 At the time, it was perhaps understandable that most observers thought that, in this realm at least, civil rights laws had failed to bring about meaningful change.

In the wake of these results, two narratives arose in academia about fair housing. The more influential, liberal narrative was that fair housing laws were simply ineffective; political compromises had fatally weakened the law in 1968, and what little enforcement followed was no match for continuing white racism.9 The most powerful statement of these views came in Professors Douglas Massey and Nancy Denton’s book American Apartheid, which, though published in 1993, relied mainly on 1970 and 1980 census data; it became the standard college text for students of housing segregation and urban inequality and is still the most influential book-length work on the subject. A competing conservative narrative contended that fair housing laws had greatly expanded housing opportunities for African Americans, but that the desire for housing integration was not particularly strong among either blacks or whites, and that continuing segregation reflected these preferences.10

Both of these accounts were fundamentally pessimistic about the prospects for housing desegregation. Those in the liberal camp considered housing discrimination to be very deeply rooted. Even the passage of a stronger fair housing law, the Fair Housing Amendments Act of 1988,11 did not make Massey and Denton sanguine.12 The conservatives, as well as a number of economists who analyzed more extensive data in the 1990s and early 2000s, had a different brand of pessimism. They were persuaded that discrimination had in fact declined substantially; but what impact would fair housing laws have on deeply rooted patterns of segregation if whites were generally willing to pay a premium to live in predominantly white neighborhoods?13 Neither view was likely to prompt activist policy measures against segregation. Political leaders knew well how politically costly efforts at school integration had been in the 1970s; without any clear vision of a way forward, there were virtually no policy initiatives aimed directly at promoting housing integration.

Meanwhile, in a trend that was missed or ignored by most scholars, levels of black-white segregation continued to fall. The magnitude and contours of the decline were subject to interpretation (and are explored further below), but there could be no question that indices of dissimilarity were falling a little bit each decade, and by 2010 the cumulative decline in black-white segregation was substantial. The average index of dissimilarity across sixty major metropolitan areas, measured at the block level, fell from 0.92 in 1970 to 0.70 in 2010.14 Professors Jacob Vigdor and Edward Glaeser, two leading economists studying segregation and cities, published a widely noticed piece called The End of the Segregated Century in 2012, suggesting that housing segregation had, by 2010, changed fundamentally from the conditions that prevailed through most of the twentieth
century.15

The declines in housing segregation were substantially larger in some metropolitan areas than in others, and in the 1990s a number of scholars began to study what difference integration made for black outcomes. What they found was at first striking, and then overpowering: lower segregation had large, positive, and causal effects upon a wide array of black outcomes.16 Whether one considered health, mortality, teen pregnancy, high school dropout rates, unemployment, earnings, marriage, or poverty, black outcomes improved in both relative and absolute terms when the black-white dissimilarity level fell sharply.17 A large controlled experiment, started in the 1990s, similarly showed that improvements in economic integration for low-income families had powerful effects on long-term outcomes for preadolescent children in those families.18

These very strong findings about the harms of segregation have come just as segregation has, for the first time in decades, emerged as at least a second-tier political issue. In June 2015, the Supreme Court issued its first major interpretation of fair housing law in over two decades, blessing a muscular interpretation of a key ambiguity in the law.19 A few weeks later, the Obama administration issued regulations that for the first time obligate cities and counties that receive federal community-development funding to develop strategies for reducing housing segregation.20 The emergence of videos depicting police behavior toward black Americans, the “99 percent” movement, and a general increase in concern about rising inequality have all led to substantially more media attention to housing segregation in the past couple of years than had existed for quite a while.21 And recent scholarship on inner-city housing and segregation has captured a significant readership among the literate public.22

Into this fray steps Professor Nicholas Stephanopoulos, making two important points. First, Stephanopoulos shows that many legal academics have failed to notice or acknowledge the very significant declines in segregation that have occurred.23 Much of modern fair housing scholarship, he points out, is consequently out of touch with what is happening on the ground.24 Second, Stephanopoulos argues that the decline in housing segregation renders some important areas of civil rights doctrine partially obsolete.25 We must take stock of these changes and recalibrate our strategies, he contends, to keep the law relevant and usable.26 On both of these subjects, Stephanopoulos has some important and original things to say. But on both I would like to offer some friendly amendments to his arguments, and on the ultimate question—where should we go from here?—I suggest a different path.

I.  In Correcting the Conventional Wisdom about Segregation, We Should Avoid Replacing Old Generalizations with New Ones—Nuance Is Important

Professor Stephanopoulos performs a valuable service in pointing out that urban, black Americans are, as a whole, substantially more integrated than they were half a century ago.27 Legal academics—especially when writing on subjects with strong ideological valence, like race—often develop narratives about their subjects that are remarkably detached from on-the-ground facts. So it has been on the subject of housing segregation, in which, as Stephanopoulos documents, some academics in the field imply not only that segregation has been largely static, but also that it is worse than ever.28 At a recent fair housing conference I attended in Chicago, several speakers inveighed against the continuing pervasiveness of both housing segregation and racial discrimination, prompting sustained audience applause. When one speaker mildly observed that there had, in fact, been some progress, the room was silent.29

Activists seem to fear that acknowledging progress leads to complacency, but I disagree, and I think Stephanopoulos disagrees as well. Only by understanding where and how progress has occurred can we determine which strategies work and which don’t. Ignoring success in fair housing not only can lead to despair and political disengagement; it also puts ideological blinders on research. An astounding share of recent segregation scholarship in sociology, demography, and law is in thrall to outdated assumptions about segregation’s static nature or the pervasiveness of housing discrimination. The failure to consider new findings with an open mind slows progress, and impedes the kind of scholarly consensus that can spur policymakers into action.

Stephanopoulos’s detailed documentation of segregation’s decline, and his reasonably sophisticated summary of the literature on the factors behind the key trends, is thus a very welcome addition to the legal literature. However, on some important points, I think Stephanopoulos oversimplifies matters.

A.    Matters of Measurement

Stephanopoulos, following the argument of economist
Edward Glaeser, suggests that black-white segregation has followed a large, inverted “U” over the past 120 years: relatively low in the late nineteenth century, high in the middle of the twentieth, and now returning to moderate nineteenth-century levels.30 This picture is overly simplistic.

First, this picture is wrong about early urban conditions. Until very recently, indices of dissimilarity for periods before 1940 relied on city wards as their basic unit of geography, because census data did not generally report demographic information for any smaller subunit.31 Using these big, political units to measure segregation tends to seriously understate it, not least because wards in that era were often drawn to gerrymander blacks and dilute the black vote.32 But in the past few years, several independent scholars have been able to computerize original census manuscripts, allowing us to analyze demographic patterns at finer levels.33 When we examine black residential patterns in cities in the late-nineteenth and early-twentieth centuries at this higher level of resolution,34 it is clear that black-white dissimilarity levels were much higher than we had assumed—in major cities, the index was around 0.735 or higher (on a 0-to-1 scale, with 1 representing complete segregation), and not below 0.5 as Stephanopoulos suggests.36 This is important, because it means that urban areas now experiencing broad residential integration in the housing market are not returning to some “natural” state, but are venturing into fundamentally new ground.

Segregation intensified during the period from 1910 to 1940; by 1940, black-white dissimilarity indices measured at the block level—the finest level of detail available from the census—were above 0.85 in most major cities, and often above 0.90. These levels held generally steady from 1940 to 1970, and then began to decline, just as Stephanopoulos suggests.37 However, they did not fall nearly as far or as consistently as he claims. Measures of average “black-white dissimilarity” are lower if one uses “census tracts”—that is, entire neighborhoods—as the unit of measurement, rather than blocks.38 They are also artificially lowered by including relatively small metro areas that have comparatively few tracts and thus very large measurement errors in their dissimilarity indices.39 Table 1 shows metropolitan black-white dissimilarity indices calculated at the block level for sixty major metropolitan areas, and for the twelve metro areas with the largest number of black residents.40 (These twelve areas together are home to about half of the nation’s urban black population.)

In short, the inverted “U” Stephanopoulos describes is actually rather shallow. Black-white segregation was real even at the turn of the last century, and it remains high—not moderate—in most big cities today. Moreover, as Table 1 shows, the big declines in segregation occurred during the 1970s; the rate of decline has slowed markedly since then.  Celebrating the death of intense black-white segregation is, at this point, decidedly premature.

B.    Conceptual Clarity

Segregation and discrimination are often conflated. In too many articles—and even more in casual discussions and at conferences—legal academics and lawyers refer to housing segregation and housing discrimination as interchangeable concepts.41 They are, of course, very different things: housing segregation describes the degree to which groups live apart from one another, while housing discrimination describes the rate at which housing providers block access to housing to specific groups. It is possible to have low segregation alongside high discrimination—say, for instance, by imposing housing quotas in a public housing project. It is also possible to have high segregation alongside low discrimination: Orthodox Jews in New York are highly clustered, but this is for cultural and religious reasons, not because they are denied access in the broader housing market.42 Stephanopoulos clearly understands the difference between these concepts, but he is gingerly vague in discussing discrimination, observing, for example, that “discrimination and segregation have decreased in tandem over the last few decades.”43 This confuses more than it clarifies. A wide range of evidence suggests that housing discrimination fell dramatically nationwide from the late 1960s to the late 1970s, and that the ability of black Americans to move into previously all-white areas simultaneously increased, often dramatically.44 But, as I have noted, general declines in black-white segregation occurred much more slowly, and much more unevenly. This is an important puzzle to confront in explaining (and understanding) the course of housing segregation.

C.    Variability of Falling Segregation 

My biggest complaint about Stephanopoulos’s discussion of housing segregation is his suggestion that declines in housing segregation are steady and inexorable.

 In fact, the decline in segregation has been dramatically uneven: as noted above, the dissimilarity index remains above 0.75 in many of our largest northern urban areas, but has fallen to around 0.60 in a number of other areas, such as San Diego, Portland, and San Antonio, with intermediate declines in most other places.45 Stephanopoulos is familiar with the variability of segregation decline, but he does not note that it results from variations in demographic environment. We now know, for example, that essentially all of the metro areas that experienced significant desegregation in the 1970s also experienced large in-migrations of black Americans from other metro areas in the 1970s when drops in segregation made it easier for these movers to locate in white neighborhoods. Areas with large declines in segregation also tended to have a relatively modest-sized incumbent black population.46

In other words, Stephanopoulos exudes a certain complacency about housing segregation that I think is unwarranted. For most American metropolitan areas with very large black populations, there is no real prospect that desegregation on a large scale—where block-level dissimilarity falls to, say, 0.6—will happen spontaneously during our lifetimes. On the other hand, there is good reason to be excited about those cases where desegregation has occurred, because, if we learn from them, we may come to understand what strategies can help get desegregation going in the rest of the country. While many academics miss how much segregation has changed, Stephanopoulos misses the degree to which further progress in desegregation will depend on careful analysis and intelligent policy.

II.  How Does Black-White Housing Desegregation Affect Civil Rights Law and Policy?

Federal policies have almost never taken direct aim at housing segregation; even laws like the Fair Housing Act or the Community Reinvestment Act of 197747 never mention housing segregation specifically. But as Professor Stephanopoulos valuably points out, some key parts of civil rights law are premised, to one degree or another, on the existence of housing segregation.48 The second part of his article usefully catalogs many ways that housing segregation plays a role in these laws.49 Stephanopoulos goes further, however, by suggesting that three key pillars of civil rights law are losing their effectiveness because of housing desegregation.50 I disagree with many of these claims. I think Stephanopoulos gets too wrapped up in the narrow ways that desegregation trends undermine specific, traditional remedies, and fails to recognize how these same trends broadly further the ultimate goals of civil rights law. As a result, I think Stephanopoulos mistakenly focuses on patching up old remedies instead of considering the more important big picture: how to further the crucial goal of housing desegregation.

A.  Voting Rights

My general concern is well illustrated by Stephanopoulos’s discussion of voting rights. Stephanopoulos points out that a central goal of the Voting Rights Act of 196551 (VRA) is to increase minority representation, and that for several decades after the VRA’s passage, a key remedy for vote dilution was to create majority-minority districts.52 In principle, this should be more difficult to do as housing segregation decreases, simply because a desegregated minority group is, by definition, less geographically concentrated.

But this misses the larger point: the ultimate goal of the VRA is not to create as many majority-minority districts as possible, but to give nonwhites the same opportunities to participate in the political process as whites.53 If housing segregation decreases, one would hope that racial polarization in voting would decrease as well, and that more black Americans (for example) would become viable candidates in majority nonblack districts. President Barack Obama’s two terms are obvious evidence that this can happen; but there is other evidence as well. The number of black federal and state elected officials was 179 in 1970, 326 in 1980, and 436 in 1990.54 It rose to 659 by 2015,55 even though the urban black population was significantly more dispersed in 2015 than in 1990, and even though the black share of the voting-age population has been essentially unchanged during this period.

In Congress, the number of black representatives rose from seven in 1968, to seventeen in 1980, twenty-four in 1990, and thirty-six in 2000.56 The number in 2015 was forty-four—an all-time high,57 and very close as a percentage of the House of
Representatives (just over 10 percent) to the black presence in the national voting-age population (roughly 13 percent).58 Moreover, if we examine the demographic makeup of the districts that black congressmen represent, a clear trend is evident. In 1974, the sixteen districts represented by black Americans had a median black population of 65 percent; only three majority nonblack districts in 1974 were represented by black Americans.59 In 2015, the forty-four districts represented by black Americans had a median black population of only 51 percent; twenty-one of the districts had a nonblack majority.60 This would seem rather strong evidence that growing urban integration and greater difficulty in creating “majority black” districts are not intrinsic obstacles to black representation, but rather the reverse.

Another way of understanding this issue is by thinking about the political effect of racial integration. In ghettoized cities during the 1960s and 1970s, black communities generally bordered all-white, often tightly bound ethnic communities that feared black entry in part because so few examples of stably integrated neighborhoods existed; black entry usually meant eventual black resegregation and the demise of ethnically based institutions—churches, specialty shops, and social clubs—that served the earlier white community.61 Segregated conditions can, in this and other ways, lend themselves to racial hostility and polarization. When integration increases, these fears recede and race itself becomes a less defining issue. It seems intuitively plausible that progressive blacks and whites in integrated cities are likely to see one another as allies rather than rivals. Professor Jessica Trounstine has made a strong case that this is empirically demonstrable. Her analysis of over ninety large cities shows that less segregated cities are less racially polarized and tend to spend more on public goods.62

Of course, voting rights issues are still very much with us. But the key issues today have to do with voter ID laws, disenfranchisement of those with criminal records, and other matters that have the effect of reducing the number of minority voters.63 There are still strong remedies in the VRA for government policies that have an unjustified disparate impact on minorities. The quest for majority-minority districts should be a declining concern to the extent that segregation itself declines. Laboring to persuade civil rights activists that new remedies are needed, when there is no real evidence of an underlying problem, seems counterproductive.

B.  Fair Housing

Stephanopoulos contends that rising levels of housing integration also weaken fair housing law.64 Stephanopoulos argues that there are several problems, including that segregated conditions (a) are sometimes important for plaintiffs in establishing standing, and (b) can provide a basis for bringing suit under a “disparate-impact” theory.65

While both observations are technically true, the conclusions Stephanopoulos draws from them seem completely unwarranted. He suggests that the decline in segregation has a stark effect upon disparate-impact litigation, and that declines in segregation will necessitate a “reorientation [of fair housing law] from desegregation to antidiscrimination.”66 This is an odd claim, because one can make a good case that the trend in fair housing law has been more nearly in the opposite direction—or at least in a direction orthogonal to the trend Stephanopoulos posits.

To a rather overwhelming degree, litigation under the Fair Housing Act in the 1970s focused on discriminatory treatment. In the vast majority of cases, plaintiffs sought to stop private housing providers, such as apartment owners, home sellers, or real estate agents, from intentionally treating people differently on the basis of race.67 Disparate-impact cases were comparatively rare, even exotic; though the principle that a “disparate-impact” theory was possible under the FHA was recognized by the Eighth Circuit as early as 1974,68 there were only a handful of such cases in the 1970s and 1980s, and the Supreme Court did not recognize a disparate-impact cause of action under the FHA until 2015.69 Two-thirds of all the published court opinions addressing disparate-impact litigation under the FHA have appeared since 2000.70

There is not much doubt that the dramatic changes wrought by the FHA were due to these early disparate-treatment cases, especially the many such cases brought by the Justice
Department in the 1970s under the leadership of Frank Schwelb, who later became a judge on the Washington DC Superior Court. Schwelb’s team did a brilliant job of focusing on larger real estate operators and “patterns” of discrimination that affected large numbers of renters and home seekers.71 As noted earlier, it was in the 1970s that discrimination fell most dramatically, and the demographic effects of this change are plainly visible: the rate at which blacks moved into predominantly white neighborhoods that were far from an existing black enclave accelerated dramatically after 1970.72 This profound shift is what started some metropolitan areas down the desegregation path.

In contrast, I am not aware of any evidence that any disparate-impact case, or lawsuit based on desegregating housing, ever had a measureable effect on the level of segregation in a metropolitan area. Sometimes these private, high-concept suits had very good effects—such as Hills v Gautreaux,73 the suit in
Chicago that enabled thousands of low-income blacks to secure housing vouchers and to move into suburban neighborhoods.74 Sometimes these suits produced resegregation, such as the lengthy Yonkers litigation in New York.75 But none of them had a measureable effect on metropolitan levels of segregation.

Thus, I think that Stephanopoulos’s analysis has some mistaken premises. Fair housing law does not have to “reorient” toward a focus on intentional acts of discrimination against individual plaintiffs—that has always been its predominant focus. Nor is disparate-impact litigation being eroded into irrelevance by integration; such litigation is, in fact, becoming more common and established. The problem with disparate-impact litigation is that it has never been part of a coherent metropolitan-level strategy to significantly reduce segregation levels; it tends, instead, to surface as an ad hoc, opportunistic weapon against scattered targets.

C.  School Desegregation

Stephanopoulos’s third and final example of how housing desegregation may reshape civil rights law is in the realm of school desegregation. In introducing this topic, Stephanopoulos does an excellent job of describing the evolution of school segregation over the past half century.76 Although the Supreme Court decided Brown v Board of Education of Topeka77 in 1954, little desegregation of schools occurred until after 1964, when the Civil Rights Act of 196478 gave the Justice Department both a mandate and additional tools to do the job. Scores of southern school districts ended de jure segregation systems in the late 1960s and early 1970s. More problematically, the Justice Department also pursued desegregation remedies against de facto segregated school districts in the North and West; because these remedies often required busing and other aggressive remedies, they sometimes succeeded but often simply drove whites out of central cities and away from urban school districts.79

By the late 1970s, the strong political reaction against “forced busing” greatly slowed the rate at which school districts were brought under court supervision, and in the 1980s and 1990s the vast majority of such districts were released from supervision.80 As Stephanopoulos points out, this had two effects: school segregation in some (unsupervised) districts rose modestly, while school segregation in general was increasingly shaped by the prevailing level of housing integration.81 Across metropolitan areas, the partial correlation coefficient between housing and school segregation increased from 0.58 in 1970 to 0.94 in 1990.82 All the evidence I have seen suggests that the causation arrow between housing and school segregation predominantly points one way: schools follow housing.

In this discussion, Stephanopoulos’s analysis is thoughtful and on point. But he fails to draw what seems to me to be the obvious conclusion: if we want more integrated schools, we should focus on furthering housing desegregation. School administrators should make it their business to improve the learning outcomes of their students, and leave the broader social engineering to housing policy. This is not what Stephanopoulos suggests: he argues for more aggressive efforts to pursue school integration, both through race-conscious student assignment systems adopted by school districts themselves, and through more frequent and aggressive suits against school systems by civil rights attorneys.83 This seems to me plainly wrongheaded.

For one thing, it seems fairly clear now that when school integration happens as a result of neighborhood integration, it is also likely to be more meaningful.84 Students are less likely to be racially segregated within schools by track or magnet program; children are more likely to develop friendships with students of other races because they can play with those students after school or on weekends; parents are more likely to get to know one another across racial lines.85 One of the most interesting pieces of research in this field is the careful 2007 study by labor economists Jesse Rothstein and David Card.86 Using a unique data set that combined national registers on high school test scores with neighborhood demographic information, they made two important findings: first, that black-white housing integration had a powerful effect in narrowing the racial gap in test scores; and second, that school integration, by itself, had a much smaller effect.87 These results make perfect sense. We know from other research that about two-thirds of the black-white test score gap exists by the time children enter kindergarten; it arises from environmental factors that disproportionately affect young black children, such as lower birth weight, fewer children’s books in the home, younger mothers, and lower socioeconomic status.88 We know that at least some of these differences narrow when housing integration occurs.89 And, if neighborhood effects tend to mitigate test score differences, this is yet another reason why neighborhood-level school integration will reduce tracking and produce more cross-racial interaction and friendship.

There is a second compelling reason why housing integration is the proper path to school integration: experience shows plainly that while housing integration is self-sustaining, school integration is not. Over the long term, there is not a single instance of a major metropolitan area desegregating after 1970 and then returning to a condition of high black-white segregation.90 But this happens regularly in school systems when administratively or judicially designed interventions create a sharp, short-term drop in segregation, but segregation rises over the long term as whites leave the system or as courts become weary of running the schools.91 Housing integration is more resilient, and has much stronger and more positive effects on black outcomes, than school integration pursued in isolation.

Conclusion

Professor Stephanopoulos’s article is an important contribution. He has noticed and pointed out an enormously significant development that has been largely ignored in the legal literature: the substantial decline in housing segregation that has occurred—albeit at a slow rate that varies across different parts of the country—over the past half century.92 He is right that we should think carefully about what this trend implies about civil rights law and policy. And I agree with Stephanopoulos that desegregation, to the extent it has occurred, implies that some traditional remedies for black Americans are becoming outmoded.93

But I think the set of lessons we need to draw from the increase in integration is different than those Stephanopoulos suggests. Because integration has advanced at very different rates in different metropolitan areas, we need to understand—as precisely as we can—how integration happens (and how it doesn’t happen). We should recognize, as I think the evidence clearly shows, that core discrimination rates fell dramatically as a result of fair housing law enforcement. The key difference between regions of desegregation and regions of continuing segregation is not (primarily) the prevailing level of discrimination, but the demographic stability of neighborhoods that become integrated. Understanding how to foster and promote integration where it is currently scarce should be a crucial policy goal.94

The amount of political capital available to address racial inequality varies from one election cycle to the next, but it is always limited. We should spend our capital carefully, and for maximum effect. Stephanopoulos has identified a variety of strategies, seemingly almost as a mental exercise considering how the portfolio of late-twentieth-century civil rights strategies can be shoehorned into a new environment. His implicit assumption is that the political capital available for new initiatives is unlimited. But it is very limited, and we must shepherd it carefully. We now know that housing desegregation is possible, and we can probably figure out what sort of policies could make it spread more broadly and deeply. Housing integration is the one outcome that we know spurs progress in dozens of other outcomes. Our best strategy is to think comprehensively about how to solidify and strengthen an outcome we know is our surest path toward racial equality.

• 1Calculation by the author of the average index of dissimilarity for Atlanta, Baltimore, Boston, Buffalo, Chicago, Cleveland, Dallas, Detroit, Houston, Louisville, Miami, Milwaukee, New Orleans, Philadelphia, St. Louis, San Antonio, San Diego, San
  Francisco, and Seattle, as reported by Karl E. Taeuber and Alma F. Taeuber, Negroes in Cities: Residential Segregation and Neighborhood Change 32–34 (Aldine 1965). This number is higher than the comparable number reported by Professor Nicholas O.
  Stephanopoulos because it is based on block-level rather than tract-level data. See
  Nicholas O. Stephanopoulos, Civil Rights in a Desegregating America, 83 U Chi L Rev 1329, 1343–44 (2016). As I discuss below, tract-level analyses can provide an overly rosy view of the actual extent of desegregation.
• 2Norman M. Bradburn, Seymour Sudman, and Galen L. Gockel, Racial Integration in American Neighborhoods: A Comparative Survey 14–16 (National Opinion
  Research Center 1970).
• 3A national study of housing integration, conducted in 1967, found that fewer than 3 percent of urban black households lived in neighborhoods that were less than 10 percent black and not perceived as likely to resegregate shortly. Id at 84–86.
• 4Douglas S. Massey and Nancy A. Denton, American Apartheid: Segregation and the Making of the Underclass 46–47 (Harvard 1993) (demonstrating that block-level indices fell only 3.9 percent and 2.8 percent in northern and southern cities, respectively).
• 5Pub L No 90-284, 82 Stat 81 (1968), codified as amended at 42 USC § 3601 et seq.
• 6See Jean Eberhart Dubofsky, Fair Housing: A Legislative History and a Perspective, 8 Washburn L J 149, 166 (1969) (stating at the time of the act’s adoption that “[w]hether the fair housing law will be enforced remains to be seen”).
• 7Massey and Denton, American Apartheid at 46–47 (cited in note 4).
• 8Id.
• 9This narrative is examined and debunked in Jonathan Zasloff, The Secret History of the Fair Housing Act, 53 Harv J Legis 247 (2016).
• 10See, for example, Richard F. Muth, The Causes of Housing Segregation, in 1 Issues in Housing Discrimination: A Consultation/Hearing of the United States Commission on Civil Rights, Washington, D.C., November 12-13, 1985 3, 11 (1986).
• 11Pub L No 100-430, 102 Stat 1619, codified at 42 USC § 3601 et seq.
• 12Massey and Denton, American Apartheid at 211 (cited in note 4) (“It is not at all clear that the new amendments, as tough as they are, will succeed in overcoming the entrenched discriminatory processes that sustain the ghetto and perpetuate segregation.”).
• 13See, for example, David M. Cutler, Edward L. Glaeser, and Jacob L. Vigdor, The Rise and Decline of the American Ghetto, 107 J Polit Econ 455, 487 (1999); Muth, The Causes of Housing Segregation at 9 (cited in note 11) (discussing the premium that immigrants are willing to pay to live “among those of similar background”); Ingrid Gould Ellen, Sharing America’s Neighborhoods: The Prospects for Stable Racial Integration 3–4 (Harvard 2000).
• 14Calculation by the author, in collaboration with Professor Yana Kucheva of the City University of New York, using restricted-access block-level data. We calculated dissimilarity indices using constant (1980) definitions of metropolitan areas. The metropolitan areas included were the sixty analyzed in a major study of segregation from the 1980s. See generally Douglas S. Massey and Nancy A. Denton, Trends in the Residential Segregation of Blacks, Hispanics, and Asians: 1970–1980, 52 Amer Sociological Rev 802 (1987). Unlike Professors Douglas S. Massey and Nancy A. Denton, who relied on tract-level data, we were able to obtain uncensored block-level counts, which largely solves the problem of maintaining constant small-area geographies over time (because blocks, unlike census tracts, rarely change boundaries).
• 15See Edward Glaeser and Jacob Vigdor, The End of the Segregated Century: Racial Separation in America’s Neighborhoods, 1890–2010 *4, 10 (Manhattan Institute, Jan 2012), archived at http://perma.cc/5E85-NT99.
• 16See David M. Cutler and Edward L. Glaeser, Are Ghettos Good or Bad?, 112 Q J Econ 827, 828 (1997) (“A one standard deviation reduction in segregation eliminates approximately one-third of the difference between blacks and whites in most outcomes.”).
• 17See generally Elizabeth Oltmans Ananat, The Wrong Side(s) of the Tracks: The Causal Effects of Racial Segregation on Urban Poverty and Inequality, 3 Am Econ J:
  Applied Econ 34 (Apr 2011) (using an instrumental variable methodology to establish that causation runs from segregation to harmful outcomes).
• 18Raj Chetty, Nathaniel Hendren, and Lawrence F. Katz, The Effects of Exposure to Better Neighborhoods on Children: New Evidence from the Moving to Opportunity Experiment, 106 Am Econ Rev 855, 859–60 (2016).
• 19See Texas Department of Housing and Community Affairs v Inclusive Communities Project, Inc, 135 S Ct 2507, 2525–26 (2015) (holding that disparate-impact claims are cognizable under the Fair Housing Act).
• 20See generally Department of Housing and Urban Development, Affirmatively Furthering Fair Housing, 80 Fed Reg 42272 (2015), amending various sections of CFR Title 24.
• 21See, for example, Ray Sanchez, Race and Reality: The Scourge of Segregation (CNN, Dec 1, 2015), archived at http://perma.cc/BX9U-UBYX.
• 22In 2016, two books dealing with inner-city housing and segregation captured a significant general readership. See generally Matthew Desmond, Evicted: Poverty and Profit in the American City (Crown 2016); Mitchell Duneier, Ghetto: The Invention of a Place, The History of an Idea (Farrar, Straus and Giroux 2016). So far as I know, that’s never happened before.
• 23Stephanopoulos, 83 U Chi L Rev at 1331–33 (cited in note 1).
• 24Id.
• 25Id at 1333–38.
• 26Id at 1338.
• 27See Stephanopoulos, 83 U Chi L Rev at 1343–61 (cited in note 1).
• 28See id at 1343–61 (collecting sources). See also, for example, Stacy E.
  Seicshnaydre, The Fair Housing Choice Myth, 33 Cardozo L Rev 967, 970 (2012) (“[T]he failure to stem racial residential segregation has helped it to deepen, widen, and become seemingly intractable.”).
• 29I attended the John Marshall Law School Conference on Fair Housing Law, September 8–10, 2016, in Chicago, Illinois. Attorney Chris Brancart noted that progress has been made.
• 30Stephanopoulos, 83 U Chi L Rev at 1344 (cited in note 1).
• 31Claude S. Fischer, et al, Distinguishing the Geographic Levels and Social Dimensions of U.S. Metropolitan Segregation, 1960–2000, 41 Demography 37, 41 (2004) (using tract data, and explaining that “block or block-group level [census data] would be ideal, but much less data are available at that level, particularly for earlier years”); Cutler, Glaesar, and Vigdor, 107 J Polit Econ at 460 (cited in note 14).
• 32For a discussion of relative effects of using block versus tract data, see Taeuber and Taeuber, Negroes in Cities at 229–31 (cited in note 1). On the use of large political units like wards, see John R. Logan, Weiwei Zhang, and Miao David Chunyu, Emergent Ghettos: Black Neighborhoods in New York and Chicago, 1880–1940, 120 Am J Sociology 1055, 1061–63 (2015).
• 33For New York and Chicago, see Logan, Zhang, and Chunyu, 120 Am J Sociology at 1069 (cited in note 32).
• 34Angelina Grigoryeva and Martin Ruef, The Historical Demography of Racial Segregation, 80 Am Sociological Rev 814, 820–21 (2015) (describing the indexing of raw census data); Logan, Zhang, and Chunyu, 120 Am J Sociology at 1063–64 (cited in note 32) (describing the process of geocoding and mapping raw census data).
• 35See Allison Shertzer, Randall P. Walsh, and John R. Logan, Segregation and Neighborhood Change in Northern Cities: New Historical GIS Data from 1900–1930, 49 Historical Methods 187, 192 (2016).
• 36See Stephanopolous, 83 U Chi L Rev at 1343 (cited in note 1).
• 37See id at 1332.
• 38See William H. Frey and Dowell Myers, Neighborhood Segregation in Single-Race and Multirace America: A Census 2000 Study of Cities and Metropolitan Areas *1–3 (Fannie Mae Foundation Working Paper, 2002), archived at http://perma.cc/K6SN-GZU4 (“[A] more refined block group–based segregation measure permits the detection of segregation patterns for small racial groups or in small areas that are camouflaged when tract-based segregation measures are used.”).
• 39See Stephanopoulos, 83 U Chi L Rev at 1357 (cited in note 1).
• 40For a description of how this data was calculated, see note 14.
• 41See, for example, Florence Wagman Roisman, Living Together: Ending Racial Discrimination and Segregation in Housing, 41 Ind L Rev 507, 508 (2008). Professor Florence Wagman Roisman writes: “Although the 1968 Fair Housing Act has prohibited residential racial discrimination and segregation for forty years . . . the United States still is characterized by substantial racial discrimination with respect to the sale, rental, and occupancy of housing and by pervasive racial residential segregation.” Id. Of course, the Fair Housing Act did not outlaw segregation and, as noted below, discrimination rates fell dramatically in the 1970s. See text accompanying note 46.
• 42See John R. Logan, Richard D. Alba, and Wenquan Zhang, Immigrant Enclaves and Ethnic Communities in New York and Los Angeles, 67 Am Sociological Rev 299, 302–03 (2002) (discussing evidence that ethnic enclaves are unrelated to economic
  constraints).
• 43Stephanopoulos, 83 U Chi L Rev at 1350 (cited in note 1).
• 44In many urban areas, the number of blacks moving to predominantly white neighborhoods (not including those adjacent to existing black districts) increased three-, five-, or even tenfold in the 1970s. Richard H. Sander, Yana Kucheva, and Jonathan Zasloff, Moving toward Integration: The Past and Future of Fair Housing *99–171 (Harvard forthcoming 2018). See also generally Jonathan Zasloff, Between Resistance and Embrace: American Realtors, the Justice Department, and the Uncertain Triumph of the Fair Housing Act, 1968–1978, 61 Howard L J (forthcoming 2018) (on file with author).
• 45See Table 1 and note 14.
• 46This argument is developed in detail in Richard Sander and Yana Kucheva, Black Pioneers, Intermetropolitan Movers, and Housing Desegregation *16 (US Census Bureau, Center for Economic Studies Research Paper 16-23, Mar 2016), archived at http://perma.cc/5SU4-AHPN.
• 47Pub L No 95-128, 91 Stat 1147, codified at 12 USC § 2901 et seq.
• 48Stephanopoulos, 83 U Chi L Rev at 1333–38 (cited in note 1).
• 49See id at 1362–1415.
• 50Id at 1374 (arguing that desegregation “make[s] it harder for plaintiffs to show standing” under the Fair Housing Act); id at 1384 (describing “problems” and “obstacles” that are likely to arise in the context of voting rights law in the aftermath of integration); id at 1410 (suggesting that integration is at least a “contingent . . . asset” for school districts defending against desegregation suits).
• 51Pub L No 89-110, 79 Stat 437, codified as amended at 52 USC § 10101 et seq.
• 52Stephanopoulos, 83 U Chi L Rev at 1334–35, 1381 (cited in note 1).
• 53James F. Blumstein, Defining and Proving Race Discrimination: Perspectives on the Purpose vs. Results Approach from the Voting Rights Act, 69 Va L Rev 633, 689–701 (1983) (highlighting the emphasis on equal access to the political process, rather than proportional results, in the legislative history of the Voting Rights Act).
• 54Statistical Abstract of the United States: 1999 *298 (US Census Bureau), archived at http://perma.cc/V8PX-B6D2. See also Black-American Representatives and Senators by Congress, 1870—Present (US House of Representatives), archived at http://perma.cc/7C9-AXF6J.
• 55Harold W. Stanley and Richard G. Niemi, Vital Statistics on American Politics 2015–2016 38 (CQ 2015).
• 56Statistical Abstract of the United States: 1999 *293 (US Census Bureau 1999), archived at http://perma.cc/V8PX-B6D2.
• 57See Stanley and Niemi, Vital Statistics on American Politics 2015–2016 at 35–37 (cited in note 55).
• 58For general population percentages, see QuickFacts: United States (US Census Bureau), archived at http://perma.cc/ZPA3-PBG4.
• 59These figures come from an analysis by the author, using data from Michael Barone, Grant Ujifusa, and Douglas Matthews, The Almanac of American Politics: The Senators, the Representatives–Their Records, States, and Districts. 1974 (Gambit 1973).
• 60Analysis by the author, using data from The Almanac of American Politics: 2016 (Columbia 2015).
• 61See Michael T. Maly and Heather M. Dalmage, Vanishing Eden: White Construction of Memory, Meaning, and Identity in a Racially Changing City 6–13 (Temple 2016).
• 62Jessica Trounstine, Segregation and Inequality in Public Goods, 60 Am J Polit Sci 709, 713, 715 (2016). This has been my own observation as well; during the 2000s I lived in California’s thirty-third congressional district, which was represented by Diane Watson, a distinguished African American congresswoman. The district was only about one-quarter black, but there was little or no sign of racial polarization across its different sections.
• 63For recent discussions, see Jedediah Purdy, A Voting-Rights Victory in North Carolina (New Yorker, Aug 2, 2016), archived at http://perma.cc/FA3C-MVDT; Samuel Issacharoff, Voter Welfare: An Emerging Rule of Reason in Voting Rights Law, 92 Ind L J 299, 315–17 (2016).
• 64Stephanopoulos, 83 U Chi L Rev at 1362–77 (cited in note 1).
• 65For the discussion on standing, see id at 1363–64, 1369. For a discussion on disparate impact, see id at 1370–72.
• 66Id at 1374.
• 67See 2 The Federal Civil Rights Enforcement Effort–1974 30–35 (US Commission on Civil Rights 1974); Zasloff, 53 Harv J Legis at 277 (cited in note 9).
• 68See United States v City of Black Jack, Missouri, 508 F2d 1179, 1184–86 (8th Cir 1974).
• 69This occurred in Texas Department of Housing and Community Affairs v Inclusive Communities Project, Inc, 135 S Ct 2507, 2525–26 (2015).
• 70Calculation by the author in cooperation with Chris Anderson. For a history of disparate-impact litigation under the Fair Housing Act, see Chris Anderson, Establishing Disparate Impact in Housing *3–15 (unpublished manuscript, Jan 2017) (on file with author).
• 71See Sander, Kucheva, and Zasloff, Inside the Labyrinth at *101–04 (cited in note 44).
• 72Sander and Kucheva, Black Pioneers at *27–29 (cited in note 48).
• 73425 US 284 (1976).
• 74Id at 288, 305–06. See also Leonard S. Rubinowitz and James E. Rosenbaum, Crossing the Class and Color Lines: From Public Housing to White Suburbia 49–70 (Chicago 2002).
• 75For Professor Peter H. Schuck’s discussion of the Yonkers case, see Peter H. Schuck, Diversity in America: Keeping Government at a Safe Distance 231–57 (Belknap 2003).
• 76The material in this and the next paragraph is based on Stephanopoulos’s discussion of school desegregation. See Stephanopoulos, 83 U Chi L Rev at 1393–1415 (cited in note 1).
• 77347 US 483 (1954).
• 78Pub L No 88-352, 78 Stat 241, codified as amended at 42 USC § 2000a et seq.
• 79Stephanopoulos, 83 U Chi L Rev at 1408–10 (cited in note 1).
• 80Id at 1336, 1398, 1408–10.
• 81Id at 1398.
• 82Id at 1400.
• 83See Stephanopoulos, 83 U Chi L Rev at 1414–15 (cited in note 1).
• 84See Jeanne Brooks-Gunn, et al, Do Neighborhoods Influence Child and Adolescent Development?, 99 Am J Sociology 353, 353–57 (1993) (proposing to understand child development “in the context of a series of environments or ecological systems in which [children] reside,” including their neighborhoods and schools).
• 85See id.
• 86See generally David Card and Jesse Rothstein, Racial Segregation and the Black–White Test Score Gap, 91 J Pub Econ 2158 (2007).
• 87Id at 2159–60.
• 88See Roland G. Fryer Jr and Steven D. Levitt, Understanding the Black-White Test Score Gap in the First Two Years of School, 86 Rev Econ & Stat 447, 447–48 (2004).
• 89See generally Emily Walton, Residential Segregation and Birth Weight among Racial and Ethnic Minorities in the United States, 50 J Health & Soc Behav 427 (2009).
• 90For example, in every urban area where the black-white index of dissimilarity has fallen below 0.70 at any point since 1970, the index in that area remains below 0.70—usually substantially below. See note 14. It is important to note that this statement does not, arguably, apply to Hispanic Americans. In some urban areas, Hispanic-Anglo segregation has increased since the 1970s, though never to levels comparable to that experienced by African Americans in major urban areas. See Douglas S. Massey and Nancy A. Denton, Trends in the Residential Segregation of Blacks, Hispanics, and Asians: 1970–1980, 52 Am Sociological Rev 802, 815 (1987). However, Hispanic segregation trends over time can be misleading because the underlying population has evolved so rapidly (due to large-scale immigration) and Hispanics in a number of metropolitan areas make up a plurality of the population. See note 14.
• 91See Sarah J. Reber, Court-Ordered Desegregation: Successes and Failures Integrating American Schools since Brown versus Board of Education, 40 J Hum Res 559, 560 (2005).
• 92Stephanopoulos, 83 U Chi L Rev at 1344–45 (cited in note 1).
• 93See id at 1333–39.
• 94In our forthcoming book, Professor Kucheva, Professor Jonathan Zasloff, and I advance a detailed policy strategy for achieving substantial and sustained integration in metropolitan areas that remain highly segregated. See Sander, Kucheva, and Zasloff, Moving toward Integration at *263–96 (cited in note 44).