Issue #138, November/December 2004


The Reality of Deconcentration


A “moral panic” over crime in central cities, combined with a demand for reform of the most troubled public housing developments, led to a profound shift in the late 1980s in how this country housed poor people. After decades of encouraging cities to locate public housing in the poorest areas and setting aside most units for very low-income people, the federal government changed direction and began a policy to deconcentrate the poor from their neighborhoods.

The government did this in response to three factors. First, recent social science research showed that poverty had become highly concentrated in America’s urban areas. The 1990 census revealed that eight million people lived in neighborhoods where more than 40 percent of the population lived below the poverty level. Between 1970 and 1990, the number of these high-poverty neighborhoods more than doubled. Research also showed that the level of social pathologies such as violent crime in these neighborhoods was extremely high.

A second factor was that the nation was in the middle of an uproar over drugs and violent crime in the cities, directed primarily at the urban poor. Each session of Congress from the mid-1980s through the 1990s tried to outdo the previous one in getting tough on crime, increasing sentences and expanding criminal justice activities in poor neighborhoods. City governments and police departments initiated various forms of “community-based policing” and multi-agency task forces to combat the problem of violent crime in the poorest sections of our cities. National community development policy reflected the preoccupation with criminal justice in the federal “Weed and Seed” program, where the “weeding” referred to the removal of criminal elements. Congress’ 1994 “Three Strikes and You’re Out” sentencing law became “One Strike and You’re Out” for poor residents of public housing.

Third, the National Commission on Severely Distressed Public Housing was created in 1989 to plot a strategy for improving conditions in the nation’s worst public housing developments. These large developments concentrated very poor households in small geographic areas, typically in already disadvantaged central city neighborhoods, and the deteriorating physical and social conditions in those developments produced ready havens for gang and drug activity and their related violent crime.

In the early 1990s, two innovations in federal housing policy were introduced, both aimed at spatially deconcentrating (or dispersing) public housing families. The Moving To Opportunity (MTO) program, authorized in 1992, was a pilot program that provided Section 8 vouchers to residents of public housing so that they could move out of public housing and into lower-poverty neighborhoods. The program was modeled on the Gautreaux program in Chicago, which provided vouchers to black public housing residents so that they could move into more integrated neighborhoods. MTO, Gautreaux and smaller efforts like them were called “mobility programs,” because they enabled poor families in high-poverty neighborhoods to move into better, more integrated or more middle-class communities.

The second policy innovation, also introduced in 1992, was the HOPE VI program. HOPE VI was designed to rehabilitate or redevelop the nation’s worst public housing developments. Rather than rehabilitation, the program’s focus became demolition and redevelopment, and it expanded beyond the 6 percent of public housing (or 86,000 out of more than a million units) that had been initially identified as the most severely distressed. The program typically removes old buildings and redevelops sites into lower-density, mixed-income developments with only a fraction of the original number reserved as public housing units. Because of the reduced number of units, many of the original residents are unable to move back on site. Thus, like MTO, HOPE VI results in the dispersal of public housing residents into other neighborhoods.

For the purpose of deconcentrating the poor, there are two important differences between MTO and HOPE VI. The first is that, under MTO and other mobility programs, participation is voluntary. Families sign up to participate, just as they did for the Gautreaux program before it. With HOPE VI, families have no choice but to move, because their housing is torn down. The second important difference between these two approaches is that participants in mobility programs typically are limited to moving to low-poverty or low-minority neighborhoods, while those who are involuntarily displaced through HOPE VI have no limits on where they can relocate. These two program characteristics – voluntary vs. involuntary dispersal, and limited vs. unlimited relocation choice – are very important for considering the likely success of these approaches in deconcentrating poverty.

Neither program is likely to have the desired effect in high-poverty neighborhoods. Voluntary programs, though justifiable on other grounds, will not make a dent in concentrated poverty for two reasons: a) these programs cream, that is, they take only those families most likely to succeed in their new environments; and b) they will never reach sufficient scale to noticeably affect overall settlement patterns. MTO, Gautreaux and other voluntary mobility programs apply to only a subset of the poor.

In addition, as Sudhir Venkatesh and Isil Celimli point out, dispersal programs incorrectly assume the poor can relocate as easily as the middle class does. In fact, very real resource constraints limit the ability of public housing families to abandon existing support networks, and these constraints limit the attractiveness of dispersal strategies to poor families.

Due to the self-selection inherent in voluntary dispersal programs, and to the screening that these programs apply to applicants, participants are likely to be more motivated and possess more human capital than the families that do not participate. Program operatives choose the families they think will succeed, based on these families’ being organized enough to pass home inspections and other steps in the application process. Still, mobility programs typically have a low success rate, because fewer than half the applicants who are accepted are able to lease an apartment.

Even if mobility programs expand and serve more families, there is an upper limit to their growth. This is defined by the political realities of the destination communities. Low-poverty areas are not anxious to receive large numbers of poor, public housing families, and there will typically be political backlash if current residents feel that these families are being forced into their neighborhoods. In fact, this very type of resistance ended the expansion of the MTO program in 1995. Mobility programs face a paradox – they must remain small to remain politically viable, but smallness ensures they will never address concentrated poverty adequately.

Another problem with mobility programs is that they are not really about improving the conditions in high-poverty neighborhoods. MTO is not designed to improve the areas where public housing residents currently live. In fact, when these programs cream the most motivated families, those neighborhoods are left with even greater concentrations of the disadvantaged.

Displacement and dispersal approaches like HOPE VI have different limitations. Large-scale redevelopment will reduce concentrations of poverty at the redevelopment site. The HOPE VI track record across the country has shown this to be the case. But the public housing families who are displaced and relocated typically reconcentrate in other poor neighborhoods nearby. Very rarely do these families relocate to low-poverty suburbs. Well over half of HOPE VI relocatees either move into other public housing or use vouchers to rent units on the private market. Public housing units are, of course, likely to be in low-income neighborhoods. Families using vouchers are also likely to move into low-income areas, because it is there that they will find units with rents that are eligible for the program and landlords who are willing to rent to them.

Other aspects of HOPE VI show that, as it has been carried out in many cities, the program is not so much about improving the conditions for previous residents as it is about reclaiming urban neighborhoods for middle- income families. The Chicago example described in this issue illustrates the higher criteria used to screen tenants, effectively keeping a large number of original residents from moving back into HOPE VI sites. The Earle Village HOPE VI project in Charlotte, NC, was cited as “a significant blighting influence, holding back an otherwise promising market environment in the surrounding area,” according to a 2003 report by The Urban Institute. Similar outcomes of rapid neighborhood turnover from low-income to middle-income populations have been documented near Cabrini-Green in Chicago, Techwood Homes in Atlanta, St. Thomas in New Orleans and other HOPE VI sites.

The greatly reduced number of units and the much stricter tenant screening standards employed after redevelopment make it difficult, if not impossible, for most families to move back and enjoy the benefits. Families that do not return to a public housing site do the best they can in other neighborhoods of the city.

Beyond being ineffective in dealing with the problems of concentrated poverty, strategies to deconcentrate may, in fact, be counterproductive. The greater the acceptance of deconcentration in any given area, the less support there will be for affordable subsidized housing. This is because the way deconcentration is being carried out through federal housing policy equates concentrated poverty with subsidized housing. People perceive that subsidized housing “anchors” poor people in neighborhoods, and the concentration of subsidized units then directly contributes to a concentration of poor people. Central city neighborhood groups and city officials convinced of the need to deconcentrate poverty conclude that they must at least not add subsidized housing in core neighborhoods, and, in fact, might benefit from reducing the number of units (through demolition or conversion). This has been a central dynamic in the Minneapolis – St. Paul region, where public policy shifted in the 1990s away from supporting rental housing improvements and toward subsidizing more homeownership opportunities. Deconcentration justifies the stance of some low-income neighborhoods that they have produced their “fair share” of affordable housing – in fact, more than their share. At the same time, the deconcentration argument provides the basis for low-poverty neighborhoods to continue to oppose subsidized housing, by linking social pathologies to concentrated poverty, and concentrated poverty to subsidized housing. In the end, there is something perversely uniting about the deconcentration argument – it leads to almost universal resistance to subsidized housing.

Adapting federal housing policy to serve the cause of deconcentrating poverty was in many respects an easy decision. Our preoccupation with the criminal threat of low-income inner-city residents reinforced the impulse to disperse the poor. The physical decay of public housing suggested an urban renewal approach that would remake central city neighborhoods, simultaneously forcing the removal of the existing population. Because this population was politically weak, it was unlikely to mount an effective defense against such measures. So it was that deconcentration manifested itself as the demolition of poor people’s homes and their forced removal from the communities they lived in.

Efforts to address the problems of concentrated poverty did not have to turn out this way. Any number of different approaches might have been tried. For example, research in Columbus, OH, conducted in the 1980s showed that concentrated poverty has more to do with the exodus of the middle class out of central cities than with the movement of the poor into those areas. We might have addressed those factors that were pushing or pulling working- and middle-class families out of core neighborhoods. Instead, community development approaches were neglected (at best) and even blamed for the problem (at worst). If we were bent on forcing a portion of the population to alter their living environments or mobility choices, we might have done something to curtail the ability of the affluent to sequester themselves behind walls and gates and to use exclusionary zoning to deny housing and neighborhood alternatives to low-income families. Such an approach would have meant, of course, manipulating the mobility decisions of a much more powerful constituency.

There are examples we can look to for inspiration. In some cases, communities outside the central cities have developed inclusionary zoning, requiring developers to keep a percentage of new units affordable so that low-income families can rent apartments. In Montgomery County, MD, one of the wealthiest suburban counties in the nation, more than 12,000 units of low- and moderate-income housing have been built through inclusionary zoning since 1974. Cities such as Boston are offering financial assistance to lure middle-class families into buying homes in struggling neighborhoods and renting part of them out. These approaches give poor people options for staying close to their old homes, but also open up possible routes for those who want to leave.

As the Chicago and Pittsburgh examples in this issue illustrate, the poor often recognize community where others do not. Demolition and dispersal approaches assume that public housing families want to move out. In the Pittsburgh case, as Mindy Fullilove points out, urban redevelopment can bring about emotional and social losses; and Venkatesh and Celimli show that many Chicago residents have returned to their old neighborhoods to patronize local institutions well before “neighborhood improvements” are put in place.

Even if we insisted on tearing down our old public housing and building anew, we might have done it differently than what has occurred in many HOPE VI projects. We might have done it as they are doing it in the Hill District of Pittsburgh, making sure that the new public housing is built before the dilapidated units are removed. Or by making sure that all families who want to can stay on site and continue to live in the redeveloped community, taking advantage of the improvements made through the program. If we had done even that, then Pat Murphy’s article would have been the story of a commonplace outcome, and not the remarkable exception that we know it to be.

Edward G. Goetz is director of the Urban and Regional Planning Program at the University of Minnesota’s Humphrey Institute of Public Affairs. This article is based on excerpts from his book Clearing the Way: Deconcentrating the Poor in Urban America, published by The Urban Institute Press in 2003.