St. Louis, 1947. To construct its Gateway Arch, the city demolished a downtown African American neighborhood, displacing residents to new black areas like Ferguson.

FROM 1957 TO 1968, Congress adopted civil rights laws prohibiting second-class citizenship for African Americans in public accommodations and transportation, voting, and employment. Although not without challenges, these laws were effective. Ending segregation in housing, however, is much more complicated. Prohibiting discrimination in voting and restaurants mostly requires modifying future behavior. But ending de jure segregation of housing requires undoing past actions that may seem irreversible.

President Kennedy’s 1962 executive order attempted to end the financing of residential segregation by federal agencies. In 1966, President Lyndon Johnson pushed to have a housing discrimination bill passed, but in a rare legislative defeat, the Senate killed his proposal. Two years later, civil rights advocates tried again, and this time the Senate eked out by the narrowest of margins a Fair Housing Act that prohibited private discrimination in housing sales and rentals; shortly after, pressured by national emotion following the assassination of Martin Luther King, Jr., in April 1968, the House of Representatives enacted the law. For the first time since 1883, when the Supreme Court rejected the Civil Rights Act’s ban on housing discrimination, government endorsed the rights of African Americans to reside wherever they chose and could afford.

This law is now a half-century old. You might think that fifty years would be long enough to erase the effects of government promotion of and support for segregation. But the public policies of yesterday still shape the racial landscape of today.

Where other civil rights laws have fallen short, the failures have been in implementation and enforcement, not in concept. Their design was straightforward. If African Americans were permitted to vote freely, their political power would be no different from that of others. If discrimination were prohibited in hiring, it would take some years for African Americans to gain comparable seniority, but once they did so, their workplace status would no longer be inferior. Once we prohibited segregation in hotels and restaurants, patrons of any race could be served. Likewise, if segregation was abolished on buses and trains, passengers of both races could sit in any empty seat the next day. The past had no structural legacy—we could use the same buses and trains, and no gargantuan social engineering was needed to make the transition to integration.

Ending school segregation was much harder, but how to achieve it was clear: districts could revise local school attendance boundaries so that children of either race could attend their neighborhood schools, and districts could upgrade the lower-quality schools that African Americans had attended, so that all facilities would be equal. Certainly there was massive resistance after 1954, when the Supreme Court ordered the dismantling of separate black and white school systems, yet in principle school desegregation in most locales was easy. And if achieving it the next day was politically difficult, the time required should have been a matter of years, not decades. Unlike desegregating housing, desegregating schools required not undoing the discrimination that previous generations received but only practicing integration going forward.*

As it has turned out, schools are more segregated today than they were forty years ago, but this is mostly because the neighborhoods in which schools are located are so segregated. In 1970, the typical African American student attended a school in which 32 percent of the students were white. By 2010, this exposure had fallen to 29 percent. It is because of neighborhood segregation that African American students are more segregated in schools in states like New York and Illinois than they are anywhere else. Throughout the country, not just in the South, busing of schoolchildren was almost the only tool available to create integrated schools—because few children lived near enough to opposite-race peers for any other approach to be feasible. Were housing segregation not pervasive, school desegregation would have been more successful.

Yet unlike the progress we anticipated from other civil rights laws, we shouldn’t have expected much to happen from a Fair Housing Act that allowed African Americans now to resettle in a white suburb. Moving from an urban apartment to a suburban home is incomparably more difficult than registering to vote, applying for a job, changing seats on a bus, sitting down in a restaurant, or even attending a neighborhood school.

Residential segregation is hard to undo for several reasons:

•Parents’ economic status is commonly replicated in the next generation, so once government prevented African Americans from fully participating in the mid-twentieth-century free labor market, depressed incomes became, for many, a multigenerational trait.

•The value of white working- and middle-class families’ suburban housing appreciated substantially over the years, resulting in vast wealth differences between whites and blacks that helped to define permanently our racial living arrangements. Because parents can bequeath assets to their children, the racial wealth gap is even more persistent down through the generations than income differences.

•We waited too long to try to undo it. By the time labor market discrimination abated sufficiently for substantial numbers of African Americans to reach for the middle class, homes outside urban black neighborhoods had mostly become unaffordable for working- and lower-middle-class families.

•Once segregation was established, seemingly race-neutral policies reinforced it to make remedies even more difficult. Perhaps most pernicious has been the federal tax code’s mortgage interest deduction, which increased the subsidies to higher-income suburban homeowners while providing no corresponding tax benefit for renters. Because de jure policies of segregation ensured that whites would more likely be owners and African Americans more likely be renters, the tax code contributes to making African Americans and whites less equal, despite the code’s purportedly nonracial provisions.

•Contemporary federal, state, and local programs have reinforced residential segregation rather than diminished it. Federal subsidies for low-income families’ housing have been used mainly to support those families’ ability to rent apartments in minority areas where economic opportunity is scarce, not in integrated neighborhoods. Likewise developers of low-income housing have used federal tax credits mostly to construct apartments in already-segregated neighborhoods. Even half a century after government ceased to promote segregation explicitly, it continues to promote it implicitly, every year making remedial action more difficult.


FROM THE end of World War II until about 1973, the real wages and family incomes of all working- and middle-class Americans grew rapidly, nearly doubling. African Americans, however, experienced the biggest growth toward the end of that period. In the 1960s, the income gap between them and white workers narrowed somewhat. The incomes of African American janitors and white production workers grew at the same pace, and the gap between them didn’t much narrow, but more African Americans, who previously would have been employed only as janitors, were hired as production workers, and they made gradual progress into better jobs in the skilled trades, at least in unionized industry. African Americans remained mostly excluded, however, from highly paid blue-collar occupations—the construction trades, for example. In most government jobs (teaching, the federal civil service, state and municipal government) but not in all, African Americans made progress: they were hired in city sanitation departments, for example, but rarely as firefighters. Overall, African American incomes didn’t take off until the 1960s, when suburbanization was mostly complete.

From 1973 until the present, real wages of working- and middle-class Americans of all races and ethnicities have been mostly stagnant. For those with only high school educations, or perhaps some college, real earnings declined, as production workers with unionized factory jobs were laid off and found employment in service occupations where the absence of unions meant wages would be much lower.

Just as the incomes of all working-class Americans, white and black, began to stagnate, single-family home prices began to soar. From 1973 to 1980, the African American median wage fell by one percent, while the average American house price grew by 43 percent. In the next decade wages of African American workers fell by another percent, while the average house price increased yet another 8 percent.

By the time the federal government decided finally to allow African Americans into the suburbs, the window of opportunity for an integrated nation had mostly closed. In 1948, for example, Levittown homes sold for about $8,000, or about $75,000 in today’s dollars. Now, properties in Levittown without major remodeling (i.e., one-bath houses) sell for $350,000 and up. White working-class families who bought those homes in 1948 have gained, over three generations, more than $200,000 in wealth.

Most African American families—who were denied the opportunity to buy into Levittown or into the thousands of subdivisions like it across the country—remained renters, often in depressed neighborhoods, and gained no equity. Others bought into less desirable neighborhoods. Vince Mereday, who helped build Levittown but was prohibited from living there, bought a home in the nearby, almost all-black suburb of Lakeview. It remains 74 percent African American today. His relatives can’t say precisely what he paid for his Lakeview house in 1948, but with Levittown being the least expensive, best bargain of the time, it was probably no less than the $75,000 he would have paid in Levittown. Although white suburban borrowers could obtain VA mortgages with no down payments, Vince Mereday could not because he was African American. He would have had to make a down payment, probably about 20 percent, or $15,000.

One-bath homes in Lakeview currently sell for $90,000 to $120,000. At most, the Mereday family gained $45,000 in equity appreciation over three generations, perhaps 20 percent of the wealth gained by white veterans in Levittown. Making matters worse, it was lower-middle-class African American communities like Lakeview that mortgage brokers targeted for subprime lending during the pre-2008 housing bubble, leaving many more African American families subject to default and foreclosure than economically similar white families.

Seventy years ago, many working- and lower-middle-class African American families could have afforded suburban single-family homes that cost about $75,000 (in today’s currency) with no down payment. Millions of whites did so. But working- and lower-middle-class African American families cannot now buy homes for $350,000 and more with down payments of 20 percent, or $70,000.

The Fair Housing Act of 1968 prohibited future discrimination, but it was not primarily discrimination (although this still contributed) that kept African Americans out of most white suburbs after the law was passed. It was primarily unaffordability. The right that was unconstitutionally denied to African Americans in the late 1940s cannot be restored by passing a Fair Housing law that tells their descendants they can now buy homes in the suburbs, if only they can afford it. The advantage that FHA and VA loans gave the white lower-middle class in the 1940s and ’50s has become permanent.


THE REDUCTION of discriminatory barriers in the labor market that began in the mid-twentieth century did not translate easily into African Americans’ upward mobility. Movement from lower ranks to the middle class in the national income distribution has always been difficult for all Americans. This reality challenges a fantasy we share: that children born into low-income families can themselves escape that status through hard work, responsibility, education, ambition, and a little luck. That myth is becoming less prevalent today, as more Americans become aware of how sticky our social-class positions are.

Imagine that we lined up all American families in order of their incomes from highest to lowest and then divided that line into five equal groups. In discussions of mobility, it is usual to call the richest fifth the top (or fifth) quintile, the next richest fifth the fourth quintile, and so on. If we were fully an equal-opportunity society (and no society is), children whose parents have incomes in the bottom quintile of the income distribution would have equal chances of having incomes as adults anywhere in that distribution. In other words, of children in the bottom quintile, one fifth would remain, as adults, in that bottom quintile. Another fifth would have incomes in the fourth quintile; another fifth would have climbed to the middle, or third quintile (we can call this the “middle class”); another fifth would rise to the second quintile; and another fifth would land in the top quintile, having the highest incomes.

In fact, however, the United States has less mobility than many other industrialized societies. Of American children born to parents whose incomes were in the bottom income quintile, almost half (43 percent) remain trapped in the bottom quintile as adults. Only 30 percent of children born to parents in the lowest-earning quintile make it to the middle quintile or higher.

African Americans have even less mobility. For those born to parents in the bottom income quintile, over half (53 percent) remain there as adults, and only a quarter (26 percent) make it to the middle quintile or higher. Considering the disadvantages that low-income African Americans have had as a result of segregation—poor access to jobs and to schools where they can excel—it’s surprising that their mobility, compared to that of other Americans, isn’t even lower. Two explanations come to mind. One is that many African Americans heed the warning that they have to be twice as good to succeed and exhibit more than average hard work, responsibility, and ambition to supplement a little luck. The other is that our affirmative action programs have been moderately successful. Probably some of both are involved.


MEDIAN WHITE family income is now about $60,000, while median black family income is about $37,000—about 60 percent as much. You might expect that the ratio of black to white household wealth would be similar. But median white household wealth (assets minus liabilities) is about $134,000, while median black household wealth is about $11,000—less than 10 percent as much. Not all of this enormous difference is attributable to the government’s racial housing policy, but a good portion of it certainly is.

Equity that families have in their homes is the main source of wealth for middle-class Americans. African American families today, whose parents and grandparents were denied participation in the equity-accumulating boom of the 1950s and 1960s, have great difficulty catching up now. As with income, there is little mobility by wealth in America. In fact, intergenerational wealth mobility is even less than intergenerational income mobility.

An equal opportunity society with respect to wealth would operate similarly to an equal opportunity society with respect to income. No matter how wealthy your parents, you would have an equal chance, as an adult, of ending up anywhere in the national wealth distribution. But nearly half (41 percent) of children born to parents in the least wealthy fifth of American families remain in that lowest quintile as adults. Another 25 percent make it to the next-lowest wealth quintile, meaning that only one-third of children born to the least wealthy American families make it as high as the middle quintile in wealth.

As is true with income, African Americans are also less mobile in wealth than whites. Fewer than one-fourth of African American adults whose parents were in the bottom wealth quintile make it to the middle wealth quintile. Nearly twice as many (42 percent) white adults whose parents were in the lowest wealth quintile make it that far. Since African Americans were mostly prevented by government racial policy from owning single-family homes in the suburbs, it is not surprising that this would be so.

This difference becomes especially significant in that white families are more often able to borrow from their home equity, if necessary, to weather medical emergencies, send their children to college, retire without becoming dependent on those children, aid family members experiencing hard times, or endure brief periods of joblessness without fear of losing a home or going hungry. If none of these emergencies consume their savings or home equity, families can bequeath wealth to the next generation.

In 1989, the most recent year for which such data are available, 6 percent of black households inherited wealth from the previous generation. Of those who inherited wealth, the average inheritance was $42,000. Four times as many white households—24 percent—inherited wealth, and the average inheritance was $145,000. In that year 18 percent of black households received cash gifts from parents who were still living, in an average amount of $800. About the same share of white households received such gifts, but the average amount—$2,800—was much greater. This, too, is the consequence of government’s twentieth-century racial policy in housing and income.


ONE REASON low-income African Americans are less upwardly mobile than low-income whites is that low-income African Americans are more likely to be stuck for multiple generations in poor neighborhoods. Patrick Sharkey, a New York University sociologist, analyzed data on race and neighborhood conditions and reported his findings in a 2013 book, Stuck in Place. He defines a poor neighborhood as one where 20 percent of families have incomes below the poverty line. In 2016, the poverty line was about $21,000 for a family of three. In a neighborhood where 20 percent of families have incomes below poverty, many more families are likely to have incomes just above it. Notwithstanding the government’s official poverty line, most of us would consider families to be poor if they had incomes that were below twice that line, $42,000 for a family of three. The federal government itself considers schoolchildren whose family incomes are nearly twice (185 percent) the poverty line to be too poor to pay for their own lunches without a subsidy. Families like theirs are also unable to move to middle-class neighborhoods, either by saving for down payments or by renting apartments at market rates. So Sharkey is reasonable when he considers such neighborhoods to be “poor.”

He finds that young African Americans (from thirteen to twenty-eight years old) are now ten times as likely to live in poor neighborhoods as young whites—66 percent of African Americans, compared to 6 percent of whites. He finds that 67 percent of African American families hailing from the poorest quarter of neighborhoods a generation ago continue to live in such neighborhoods today. But only 40 percent of white families who lived in the poorest quarter of neighborhoods a generation ago still do so.

Forty-eight percent of African American families, at all income levels, have lived in poor neighborhoods over at least two generations, compared to 7 percent of white families. If a child grows up in a poor neighborhood, moving up and out to a middle-class area is typical for whites but an aberration for African Americans. Neighborhood poverty is thus more multigenerational for African Americans and more episodic for whites.

The consequences of being exposed to neighborhood poverty are greater than the consequences of being poor itself. Children who grow up in poor neighborhoods have few adult role models who have been educationally and occupationally successful. Their ability to do well in school is compromised from stress that can result from exposure to violence. They have few, if any, summer job opportunities. Libraries and bookstores are less accessible. There are fewer primary care physicians. Fresh food is harder to get. Airborne pollutants are more present, leading to greater school absence from respiratory illness. The concentration of many disadvantaged children in the same classroom deprives each child of the special attention needed to be successful. All these challenges are added to those from which poor children suffer in any neighborhood—instability and stress resulting from parental unemployment, fewer literacy experiences when parents are poorly educated, more overcrowded living arrangements that offer few quiet corners to study, and less adequate health care, all of which contribute to worse average school performance and, as a result, less occupational success as adults.

Certainly some children overcome these difficulties. But the average child living in a poor household is less likely to escape poverty as an adult, and the average child living in a poor household in a poor neighborhood is even less likely to do so. The cycle can be broken only by a policy as aggressive as that which created ghettos of concentrated poverty in the first place.


BECAUSE AMERICANS vary greatly in their economic and social circumstances, any government program will affect different Americans differently even if, on its face, the program treats all alike. A sales tax, for example, applies equally to all but will be more of a burden to lower-income consumers than to higher-income ones. The legal jargon for this is that it has a “disparate impact” on different groups. In a society where everyone’s situation is different, disparate impacts are unavoidable, but we can try to minimize them—in the case of a sales tax, by exempting grocery purchases.

Once de jure segregation was established, African Americans and whites were not affected similarly by subsequent race-neutral policies. The Fair Housing Act prohibits housing programs whose disparate impact on African Americans reinforces their segregation, unless the programs have a legitimate purpose that cannot be accomplished otherwise. But the Fair Housing Act does not prevent disparate impacts from other, nonhousing programs that build on preexisting residential patterns. Unlike the activities that comprise de jure segregation, these programs need not have the intent of harming African Americans (although sometimes they may) but they do harm nonetheless. Several seemingly “race-neutral” programs have reinforced the disadvantages of African Americans that were initially created by race-conscious housing policy.

Along with the mortgage interest deduction, another policy that on its face is race-neutral but has a discriminatory effect is our national transportation system. We have invested heavily in highways to connect commuters to their downtown offices but comparatively little in buses, subways, and light rail to put suburban jobs within reach of urban African Americans and to reduce their isolation from the broader community. Although in many cases urban spurs of the interstate highway system were unconstitutionally routed to clear African Americans away from white neighborhoods and businesses, that was not the system’s primary purpose, and the decision to invest limited transportation funds in highways rather than subways and buses has had a disparate impact on African Americans.

Transportation policies that affected the African American population in Baltimore illustrate those followed throughout the country. Over four decades, successive proposals for rail lines or even a highway to connect African American neighborhoods to better opportunities have been scuttled because finances were short and building expressways to serve suburbanites was a higher priority. Isolating African Americans was not the stated purpose of Maryland’s transportation decisions, though there also may have been some racial motivation. In 1975, when Maryland proposed a rail line to connect suburban Anne Arundel County and downtown Baltimore, white suburbanites pressed their political leaders to oppose the plan, which they did. A review by Johns Hopkins University researchers concluded that the residents believed that the rail line “would enable poor, inner-city blacks to travel to the suburbs, steal residents’ T.V.s and then return to their ghettos.” Maryland’s state transportation secretary stated that his office “would not force a transit line on an area that clearly does not want it,” failing to explain how he balanced the desires of a white suburban area “that clearly does not want it” with the desires of urban African Americans who needed it.

Forty years later little had changed. In 2015, Maryland’s governor canceled a proposed rail link to Baltimore’s west-side black neighborhood, saying the funds were needed for highway improvements. The NAACP Legal Defense Fund then filed a complaint with the U.S. Department of Transportation, claiming that Maryland’s priority for highways over mass transit had a disparate impact on African Americans. The case was still pending when the Obama administration left office.


ACTIONS OF government in housing cannot be neutral about segregation. They will either exacerbate or reverse it. Without taking care to do otherwise, exacerbation is more likely. The federal government now operates two large programs to address the housing crisis faced by the poor and near-poor, most of whom, in many metropolitan areas, are African American. Without an intent to do so, each program has been implemented in a manner that deepens racial segregation. One, the Low-Income Housing Tax Credit, subsidizes developers whose multiunit projects are available to low-income families. The other, Housing Choice Vouchers (popularly known as “Section 8”), subsidizes families’ rental payments so they can lease housing that they would not otherwise be able to afford.

In the tax credit program, communities can veto developers’ proposals, something that officials in middle-class areas don’t hesitate to do. Many policy makers urge developers to build in already segregated neighborhoods in the hope (usually a vain one) that their projects will revitalize deteriorating areas. Developers themselves also prefer to use tax credits in low-income neighborhoods because land is cheaper, it is easier to market new apartments to renters in the immediate vicinity, and there is less political opposition to additional housing for minorities and lower-income families. These conditions ensure that tax credit projects will have a disparate impact on African Americans, reinforcing neighborhood segregation. An analysis of all tax credit units nationwide, completed through 2005, found that about three-fourths were placed in neighborhoods where poverty rates were at least 20 percent.

In the Section 8 program, landlords in most states and cities can legally refuse to rent to tenants who use housing vouchers, although a few jurisdictions prohibit such discrimination. The voucher amount is usually too small to allow for rentals in middle-class areas. A family that receives a voucher may find that the only way to take advantage of it is to move to a neighborhood even more segregated than the one where they were already living. As a result, few families with children who used Section 8 vouchers rented apartments in low-poverty neighborhoods in 2010, while over half rented in neighborhoods where the poverty rate was 20 percent or more, including some who rented where poverty was extreme—40 percent or more. Where vouchers are used to rent suburban apartments, these apartments are frequently in segregated enclaves within otherwise middle-class suburbs.

In 2008, the Inclusive Communities Project (ICP), a Dallas civil rights group, sued the state of Texas, claiming that the operation of the tax credit program had a disparate impact on African Americans, violating the Fair Housing Act. In the city of Dallas, 85 percent of all tax credit units for families were in census tracts where at least 70 percent of residents were minority. The ICP had been attempting to promote racial integration in the Dallas area by helping African American families with Section 8 vouchers find affordable apartments in predominantly white neighborhoods, but it was impeded because so many of the tax-subsidized family housing developments approved by the State of Texas were in heavily minority and low-income areas.

In June 2015, the Supreme Court ruled in the ICP case that the disproportionate placement of subsidized housing in neighborhoods that had been segregated by past government policy could violate the Fair Housing Act, even if the placement was not intended to intensify segregation. But the opinion, written by Justice Anthony Kennedy, also allowed that placing subsidized units to support the revitalization of deteriorating neighborhoods could also be legitimate. So it is not evident how much of a nationwide push toward desegregation will result from the ICP case.

Gentrification of private housing in urban areas, redevelopment projects, and highway routing have forced low-income and minority families to search for new accommodations in a few inner-ring suburbs that are in transition from white to majority minority. When the tax credit and Section 8 programs subsidize the movement of low-income families into such suburbs, and not into predominantly middle-class ones, they contribute to segregation. Ferguson, Missouri, outside of St. Louis, is such a place. When the Section 8 and tax credit programs failed to offer opportunities to settle throughout the St. Louis metropolitan area, they contributed to Ferguson’s transformation from an integrated to a predominantly minority and increasingly low-income community.


As public housing towers like Pruitt-Igoe in St. Louis were taken down in the early 1970s and their sites redeveloped, residents were forced into other segregated neighborhoods.

Civil rights advocates and local housing officials face a difficult conundrum. The ongoing income stagnation of working-class families and the growing distance between job opportunities and affordable housing makes the need for subsidized housing more pronounced. Government officials can satisfy more of that need by using scarce Section 8 funds and supporting tax credit developments in segregated neighborhoods where rents and land are cheaper and where white middle-class voters place fewer hurdles in the way. In the long run, however, African Americans will be harmed more by the perpetuation of segregation than by continued overcrowding and inadequate living space. Neither is a good alternative, but short-term gains may not be worth the long-term costs.


* Whether African Americans were entitled to compensation for having received inferior and unconstitutional educations prior to 1954 is an important issue but not the subject of this book.

† In addition to de-unionization of the less-skilled workforce, the decline in the real (inflation-adjusted) minimum wage has also contributed. We tend to romanticize this economic history, saying that good factory jobs have been replaced by bad service jobs. But in truth, there is nothing better about banging hubcaps onto cars on a moving assembly line than about serving hamburgers in a fast food restaurant or changing bed linens in a hotel. The difference between these types of jobs is mainly that industrial jobs were frequently unionized and service jobs are not. Service jobs might even be a more secure source of stable income if they were unionized and protected by an adequate minimum wage standard. An automobile assembly plant can move overseas, but a hotel or fast food restaurant cannot.

If you find an error please notify us in the comments. Thank you!