It may seem self-evident that government-subsidized housing advances the interests of blacks. The argument is simple: Because lower-income blacks can only afford inferior housing, government subsidies are needed to put better housing within their reach. Yet housing subsidies may do more harm to minorities than good. In some cases, they have actually strengthened the incentive to discriminate on the basis of race.
This is the conclusion to be drawn from what was probably the most celebrated New York housing discrimination controversy of the Eighties: the Starrett City racial quota system. Built in Brooklyns East New York section in the mid-1970s, Starrett City is one of the largest housing complexes in New York. A 46-building, 5,470-unit state and federally subsidized complex, it houses 23,000 people. Between 1975 and 1988, when they were struck down by the courts, quotas kept 65 percent of Starrett Citys occupants white. The management kept separate waiting lists for whites and blacks and let apartments sit empty, keeping prospective black tenants waiting, while it looked for white families willing to move into the integrated project.
Because it was a quota system that appeared to discriminate against blacks, the Starrett policy made strange bedfellows. The Reagan Administration Justice Department made common cause with the NAACP in challenging the Starrett policy as discriminatory. Advocates of Starretts quotas, including housing design reformer Oscar Newman, defended them as necessary to forestall white flight. Opponents of the quotas, like the planner Paul Davidoff, viewed the Starrett policy as overt discrimination, asserting that it was unconscionable for a state-supported project to adopt such a policy in a metropolitan area where many communities were not yet integrated.
Both views are easy to agree with and impossible to reconcile. Entirely missing from the debate, however, has been an understanding of the role that government housing subsidies played in creating an incentive to discriminate. Starrett received a variety of government funds: Mortgage interest rates were subsidized by both the Federal Government and the state, and individual tenants received more than $21 million in federal rent subsidies. The unanticipated consequences of such subsidies are central to understanding the dynamics of race in the Starrett City case, and in the housing market in general.
How can housing subsidies abet racial discrimination? By preventing landlords from using other, more defensible criteria in selecting tenants. Owners of rental property generally select their tenants by estimating the effect of their presence on the long-term value of the property. Will they pay rent regularly? Will they treat the property with respect? Will their presence encourage other tenants and prospective tenants to regard the property as a desirable place to live? In answering these questions in any given case, property owners make judgments that reflect their own attitudes toward such characteristics as income and, unfortunately, race. The owners of highly subsidized properties like Starrett City find themselves impelled by government regulation to select subsidized tenants without regard to their earned income. Because landlords are unable to use income as a predictor of behavior, they fall back on a racial quota system to limit the number of what they think may be problem tenants. That, at least in my view, appears to have been true in the complex saga of Starrett City.
Starrett City was conceived in 1964 by the United Housing Foundation, an organization formed by a consortium of unions eager to build new, inexpensive housing for their members. Modeled after Co-op City in the Bronx and designed by the same architect, it was intended to be unadorned housing owned as a cooperative. The union group turned to bond-supported state financing available for middle-income housing under the terms of the 1955 New York Private Housing Finance Law, known as the Mitchell-Lama law. It is important to note that because Mitchell-Lama subsidies were less generous than the subsidies available to publicly owned housing, Mitchell-Lama projects tended to be placed either on land whose cost to the developer was reduced by urban renewal subsidies or in declining urban areas being deserted by the middle class. Starrett City fell into the latter category.
In 1971, after some $300 million in bonds had been floated and ground had been broken for the projects power plant, United Housing became concerned about the total cost of the massive project, its mortgage subsidy notwithstanding. The union feared that rising costs would force it to raise its prospective rents from $25 to $45 per room per month, putting them beyond the means of those for whom they were intended.
In other words, in taking its reading of the housing market, the union was skeptical of the demand for Starrett City. This skepticism did not deter the state, however. When the union group walked away, the State Division of Housing was determined to continue the project despite the warning signal that it might be too costly for its potential market. The Division of Housing turned to a private real estate firm, the Starrett Housing Corporation, to take over. Starrett insisted on operating the project as rental, not cooperative, housing, and immediately found itself in hot water with the white middle-class neighborhoods on Brooklyns south shore, particularly Spring Creek and Canarsie.
When Starrett went to the Board of Estimate to obtain a property tax abatement that was vital to its financing, it met with organized opposition from churches, synagogues, and parent-teacher associations. Now that the giant project was no longer reserved for union members, its future neighbors feared that racial change would lead to crime and lower property values. Among those testifying at the Boards public hearing was Daniel Shapiro, president of Congregation Bnai Israel in Spring Creek. We are hearing a lot of what this will do for families coming into the community [by] giving them houses, Shapiro said. Let us talk about what is going to happen and why we need to protect the families now living in Spring Creek. We dont have to tell anybody here what the word East New York means—ghetto, poverty.... I implore every one of you at this session [to] protect this area . . . as an integrated community, and not have it turned totally ghetto.
Opponents within the community proved willing to support Starrett on the condition that it promise to maintain a ratio among its tenants of 70 percent white to 30 percent black, a provision to which Starrett informally agreed. Thus, right from the start the issues of subsidies and race were intertwined. Because Starrett was perceived to be making it possible, through state subsidy, for newcomers to enter the community at prices lower than homebuyers would pay, the neighborhood wanted assurance that the newcomers would meet middle-class community standards. Lacking a means of screening new residents on the basis of their ability to pay (or, as implied by the original plan, on the basis of union membership), their prospective neighbors asked instead that the newcomers be screened on the basis of race.
Starrett was intended as a middle-income development. Why was that not enough to calm the fears of Spring Creek and Canarsie? Perhaps because they were racists who feared having even middle-class blacks in their part of Brooklyn. But at a time when the states Urban Development Corporation was announcing ambitious plans for subsidized housing outside ghetto areas, perhaps they were concerned that even if Starrett was intended as a middle-income project, there was no guarantee it would remain one. And indeed it did not, much to the professed surprise of Starrett management. When Starrett began, in late 1973, to rent its first units, its management had no plans to select tenants on racial grounds. Instead, management set out to fill the project with middle-income tenants for whom it was conceived, convinced, as general manager Robert Rosenberg has recalled, that in the process it would realize the racial mix it had promised. Toward that end, it established a minimum-income policy for tenants.
Once it began to take rental applications, however, Starrett City found that it could not come close to attracting a sufficient number of middle-income applicants to fill its units. Most of those who came to the office at the massive construction site were poor and black. Even after advertising the project extensively in middle-income areas and adding amenities such as built-in dishwashers and a community center with a swimming pool, Starrett could not attract the thousands of middle-income tenants it was built to serve. It was only at this point that Starrett asked the State Division of Housing to permit the establishment of a quota system to be known as the managed waiting list. Starrett would attract middle-income whites by assuring them there were limits on the numbers of blacks to be admitted. Starretts management wanted to accept only middle-income tenants, but because not enough of them wanted to live in the project, Starrett decided to leave units vacant rather than lower its prices and rent to poor blacks. The State Division of Housing agreed to increase the projects subsidy allocations to cover the expected cost of the forgone rents.
Thus, subsidies enabled Starrett to institute its quota policy. A privately financed project would not have been able to afford to leave apartments vacant until whites applied for them. But more fundamentally, subsidies created the need for the quota policy in the first place. It was only the availability of state housing finance bonds and a low-interest construction mortgage that permitted Starrett to be built on its grand scale, without regard to market demand for middle-income apartments in its section of Brooklyn. Starrett, in fact, was built to counter market forces that were drawing middle-income families to the suburbs, not to serve them. Its scale lay at the heart of its difficulties. A smaller project, designed with a realistic sense of how many middle-income families might choose to live there, would have had much less difficulty filling its units with middle-income tenants. Management would then have had no need to resort to racial quotas.
Despite additional aid from the state, Starretts huge scale continued to be a problem. The combination of its managed waiting list and cost increases—particularly oil costs—threatened to push the project into bankruptcy. Indeed, its management has acknowledged that the complex was saved by another dramatic bailout. In the waning days of the Carter administration, the Department of Housing and Urban Development decided that in addition to its state and federal mortgage interest subsidies, Starrett would be eligible for a massive infusion of federal Section 8 rent subsidies for poor tenants. Over time, subsidized tenants came to dominate the project originally planned for the middle class. As of July 1990, according to Starretts figures, fully 4,909 of the projects 5,470 units (89 percent) received some form of rent subsidy for the poor. By that same date, between 35 and 40 percent of Starrett families, including many elderly households, were earning less than $10,000 a year.
But even in this new context, Starretts management continued to insist on its managed waiting list—despite lawsuits, first (in 1979) by the Columbia University Law School Fair Housing Clinic and the Manhattan-based Open Housing Center, and later (in 1984) by the U.S. Justice Department. Throughout the protracted litigation, Starrett asserted that its goal was racial integration as a means of preventing the deterioration identified with all-black projects. Attorney Morris Abram, a Reagan appointee to the U.S. Civil Rights Commission who was generally viewed as a skeptic of affirmative action quotas, argued that the managed waiting list was an unfortunate necessity: In the circumstances that confront Starrett City today, race-conscious tenant selection methods are the only way in which we can even hope to put the problem of race finally behind us.
What Starrett did not say, but its data show, is that its choice of tenants was already narrowed by its dependence on federal subsidy. It did not have a broad income pool from which to draw; its choice, in essence, was between a project dominated by white poor and one dominated by black poor. Because virtually all tenants in the project received rent subsidies, Starrett had neither the incentive nor the ability to select tenants on the basis of their ability to pay. A housing project filled by two-earner families, regardless of race, might well have been acceptable to the surrounding community as well as to management. But because of its need to fill thousands of units, Starrett had to turn to rent subsidies and, despite mounting legal pressure, stick with a system that differentiated among prospective tenants on the noneconomic criterion of race.
In November 1988, the U.S. Supreme Court refused to review a Second Circuit Court of Appeals decision barring Starretts further use of its managed waiting list. It is fascinating to note that, in the years since, the now-renamed Starrett at Spring Creek has returned to its original method for maintaining the value of its property: economic, not racial, differentiation among tenants. Citing fears that the federal housing subsidy stream might not be reliable, Starrett has taken the position that it will now rent, whenever possible, to middle-income tenants—the original target population of the project—rather than to new tenants requiring deep subsidy. During the first two years of this new policy, Starrett rented 120 middle-income units, none to a family earning less than $25,000 and the largest percentage (30 percent) to those earning between $50,000 and $60,000 a year. Ninety-nine of the 120 new tenants—82 percent—were black. In a subsidized market, Starrett chose to discriminate on the basis of race. In renting to nonsubsidized tenants, Starrett has looked to the ability to pay, and blacks have benefited.
The Starrett City story suggests broader implications for race relations and black upward mobility. It may be generally harmful for blacks when the state supports the construction of large-scale, mixed-income subsidized housing developments, as New York has done to an unusual extent. Consider two other Mitchell-Lama projects not far from Starrett City in Brooklyn and Queens: Fairfield Towers (1,146 units) and Linden Plaza (1,500 units). Both suffer the kinds of maintenance and social problems that were feared by defenders of Starretts quotas. Like Starrett, these projects were built for the middle-class but without regard to the extent of middle-class demand, and subsequently have had to combine the poor and the better-off. When middle-class residents are combined with those requiring deep subsidy, solid, striving citizens are inevitably plagued by underclass behavior—a fact of life at Fairfield and Linden. Projects combining the classes, or low-income projects built in middle-class neighborhoods, make it more difficult for the emerging black middle class to consolidate its economic gains, to stake out buildings and neighborhoods of its own, in which price is a barrier to those who do not adhere to middle-class norms of behavior. Moreover, the critical mass of middle-class residents necessary to secure such neighborhoods is diminished because the state lures some of them to subsidized units.
Most ominously, when predominantly black, mixed-income projects decline because of the difficulties posed by the underclass poor, it reinforces the view that too many black people are bad for a neighborhood. Thus whites generalize that they must be wary of blacks, rather than of underclass newcomers of whatever ethnicity. Some of the racial tension in Canarsie, as Jonathan Rieder has written, can be traced to the presence of public housing projects preponderantly occupied by low-income black families on the fringe of a neighborhood of one- and two-family homes. By opening middle-class neighborhoods to poor black families who have not yet worked their way up the socioeconomic ladder, such subsidized housing invites middle-class whites to identify race as the problem and to become intolerant even of the middle-class black homebuyer who would otherwise easily fit into the neighborhood.
Thoughtful people may well support a government effort to moderate racially segregated residential neighborhoods in an effort to change school segregation. But when government subsidizes the market, it changes social dynamics in ways that can pose risks for those it seeks to help. In New York, where there is both more subsidized housing and seemingly more racial tension than in just about any other U.S. city, policymakers ignore the relationship between the two at the citys peril.