Subsidized low-income housing in the United States is often assumed to be a welfare program of income redistribution in favor of the poor. In fact, in the United States low-income housing became primarily an instrument of urban renewal rather than Watson's program of income redistribution. Even to the extent that low-income housing has been a program of income redistribution, its benefits have been depreciated by segregation and discrimination. As a result of segregation, racial minorities have often been denied units altogether or assigned exclusively to housing projects located in low-income neighborhoods with inferior services and amenities. (Berry and Kassarda 1997)
Before the Civil Rights Act of 1964, segregation (at least racial segregation) was widely accepted, codified in some community laws, and the source of no noticeable opposition from federal housing agencies. The period from 1932, when the federal government first became involved in public housing construction and what was then called slum clearance), to the early 1960s, when discrimination and segregation in housing programs receiving direct or indirect federal financial aid were formally prohibited, will be referred to as the early period of subsidized housing.
The time since 1964, the high point of the civil rights movement with its important goal of ending segregation in the housing market, will be referred to as the modern period of subsidized housing. What distinguishes this period from the earlier one (at least formally) is the withdrawal of support for segregation by federal housing agencies. During the early period, subsidized low-income housing projects were, as a matter of public policy, concentrated in low-income areas and assigned to tenants on the basis of race, and that this process still characterized the housing programs, with no significant improvement, in the modern period. From its beginnings during the Great Depression, federally subsidized low income housing in the United States has rarely been a true welfare program. It has more characteristically been an adjunct to corporate city central business district redevelopment, serving as a receptacle for some of the residents displaced by the demolition of low-income or minority neighborhoods too close to CBDs; or a component of the industrial mobilization of World War II; or a way of subsidizing owners of private housing. (Berry and Kassarda 1997)
Federal programs have actually provided shelter to less than 10 percent of the eligible population at any point in time; the meager resources made available were completely inadequate to the scale of the problem. For example, in the 1930s about one-third of the population of the country was ill housed, but the fewer than 22,000 federally supported units that existed by 1936 (some still under construction) were not even enough to rehouse the more than 30,000 households displaced by federally supported clearance activities in 1932-1936. (Aaron 1992) Federally subsidized low-income housing in the United States has also exhibited pervasive patterns of racial segregation, both in acceptance of tenants and assignment to units and of income separation.
Before World War II, 43 of the 49 projects supported by the Public Works Administration and at least 236 of the U.S. Housing Authority's 261 were completely segregated racially, a few by means of racial sections within projects but most by admitting only tenants of a particular race. (Burchell et al 1995) This segregation surely kept some whites as well as blacks from benefiting from the federally subsidized housing of the era. Wartime public housing programs were segregated incidentally by being targeted to workers in defense industries which customarily discriminated against blacks and directly by deliberately excluding many of the blacks who did qualify even when units were available.
After World War II, antidiscrimination laws enacted by many localities outside the South led to reduced levels of segregation in the assignment of public housing units, at least for a time. However, the sustained removal of income-ineligible tenants, most of whom were white, greatly increased the proportion of blacks among public housing residents. Many of the remaining white tenants were elderly, and as separate public housing projects for the elderly became common after 1956 the older family projects became increasingly dominated by racial minorities, especially blacks. (Berry and Kassarda 1997) This trend became even more significant as new construction essentially ceased for family housing while elderly housing continued to be built and seems to extend to the sector of privately owned units subsidized by rent supplements.
In analyzing changes in the degree of income separation of subsidized housing projects over time, it should be noted that the distribution of housing projects among income areas in a community can change simply as a result of economic changes in the community. Deterioration or revitalization of the surrounding community can move a housing project from a moderate-to-high income area to a low-income area and vice versa. (Montgomery 1992) Income areas can also change as a result of natural disasters (e.g., earthquake, flood, fire) or such large-scale activities as freeway construction. Many of the changes in the distribution of subsidized housing projects among income areas cannot be attributed to actions or policies specific to public housing.
After 1956, public housing projects for elderly and handicapped households were built outside of central cities. The approximately 640,000 units of federally subsidized housing in 1960 included a substantial number outside of central city areas, the proxy for the low-income area. During World War II, radical, if temporary, change in the objectives of subsidized housing programs strongly influenced patterns of racial occupancy of housing projects and data collection methodology. (Aaron 1992) The subsidized housing stock underwent tremendous expansion, from about 100,000 units in 1941 to nearly 800,000 in 1945. (Aaron 1992) Although racial occupancy data are incomplete, it is clear the vast majority of war housing tenants were white. Blacks were largely relegated to permanent public housing, essentially that built or authorized before the war. Relatively more reliable racial occupancy data for these permanent projects were collected by the Race Relations Branch of the FPHA. (Burchell et al 1995) Since this constitutes most of the data from which it is calculated, the index of racial segregation for this period is likely to be biased toward housing projects with relatively more black tenants.
After World War II, the public housing picture again changed significantly as temporary war housing projects, mostly occupied by whites, were liquidated or demolished and as many tenants, more of them white than black, became ineligible on the basis of income for continued occupancy of permanent projects. In a relatively short period, blacks and other racial minorities came to predominate in conventional public housing. Between 1953 and 1964, racial occupancy data were collected somewhat systematically by the Race Relations Branch of the Public Housing Administration, but only for entire states, not individual housing projects or PHAs, and only for those states with open racial occupancy. (Montgomery 1992)
In the post- 1964 period, racial occupancy patterns in federally assisted housing were strongly influenced by the concentration of white tenants in elderly housing, first authorized by Congress in 1956, and of black tenants in more traditional family housing. These data tend to be relatively more detailed and conceptually consistent with one another than earlier period occupancy data. Despite their limitations, comparison of indices of racial segregation for the periods before and after the civil rights and fair housing legislation can yield some insights.
The analysis of trends in segregation indicates that the level of income separation in low-income housing declined significantly immediately after the civil rights and fair housing legislation of the mid to late 1960s. Patterns of racial occupancy of housing projects changed for a time from an almost complete separation of tenants by race to very modest levels of interracial housing. Structurally, the change in the level of racial segregation was not uniform across census regions.
During the pre-civil rights period, the level of racial segregation of tenants was lowest in the Northeast and Midwest and highest in the South. After the civil rights reforms of the 1960s, the level of segregation of tenants was highest in the Northeast and Midwest and lowest in the South and West. In some Northeast PHAs, including those of New York City and Philadelphia, there was a change from a relatively high degree of racial integration before the civil rights reforms to racial homogeneity afterwards. (Montgomery 1992) Segregation in the later period is generally by program, many family and elderly housing projects being occupied predominantly or exclusively by blacks and by whites respectively. Housing projects were less concentrated in low-income areas during the post-civil rights period. This was primarily a result of significant construction of elderly housing outside of central cities. A comparison of the indices of income separation for family and elderly housing projects indicated that elderly housing was significantly less separated by income than family housing.
Income separation, the disproportionate siting of public housing in low income areas instead of evenly or randomly across neighborhoods of different income levels, is hard to analyze because of the difficulties of geographically delineating the low- and the moderate-to-high-income areas and of locating projects in one or the other. Nevertheless, rough gauges of income separation may be obtained by making some assumptions. For the earlier period, we can assume that projects built in demolished slum areas and formerly industrial vacant sites were located in the low-income area. For the modern era, we assume that in the old core cities of metropolitan areas are low-income, other places moderate-to-high income. (Berry and Kassarda 1997)
Comparing the values of the housing indices before 1965 with those since suggests that the civil rights legislation of that time had no significant effect on the level of racial segregation in federally subsidized housing. The current picture of federally subsidized housing, then, is of a certain amount of rent subsidies to tenants of private landlords and of publicly owned housing characterized by a curious form of de facto segregation, by demographically targeted programs (whites in elderly housing, blacks in family housing) rather than by housing projects. Such structural patterns of segregation seem largely beyond the influence of the civil rights legislation of the 1960s