designating or of inexpensive, publicly owned or financed housing units scattered throughout middle-class residential areas
For more than 25 years, "scattered-site" housing has been widely promoted as an alternative to the
large, dense public and assisted housing developments that too often concentrated poor families in socially
isolated and profoundly dehumanizing environments. Although many public housing authorities have invested
significant resources in scattered-site housing, there has been little systematic infonnation on its characteristics and perfonnance. Scattered-Site Housing fills an important vacuum in our understanding of this
crucial segment of the HUD-assisted housing stock.
Professor James Hogan of Seattle University has assembled survey, census, and other secondary data
to provide a rich analysis of several key locations where scattered-site housing has been employed. His
report indicates that scattered-site programs, though sometimes more costly to manage, are highly
popular with housing authority managers, residents, and communities at large. Tenants in scatteredsite housing feel welcome in their new homes and prefer their new neighborhoods, where they expect
their children to benefit from safer surroundings and better access to quality schools. Hogan's research
also offers important insight into factors that contribute to successful scattered-site housing, including
careful tenant screening, small-scale development patterns, good design, and attractive buildings that
"promote individual pride and care of property."
Nonetheless, scattered-site units, like other assisted housing, frequently encounter initial community
opposition, in part due to fears that their presence will reduce the value of nearby properties. HUD's
Office of Policy Development and Research has recently funded complementary research that will
gather and examine empirical evidence on scattered-site housing's effects (if any) on local property
values and communities. '
The persistence of concerns over poverty concentration and quality of life that originally led to the
wide spread use of scattered-site housing makes it certain that this development model will continue
to be of considerable policy interest. It is, for example, a key element of HUD's broader strategy for
transfornling America's public housing system. This excellent report offers policymakers and program
administrators alike a new, firmer basis for understanding the evolution, the variety, and the promise of
scattered-site housing.
I am grateful to the individuals and institutions that helped make the completion of this study possible. I wish to thank Seattle University for sabbatic support and summer faculty fellowships, my colleagues in the Political Science Department for their encouragement and occasional therapy, and Dean
Stephen Rowan of the College of Arts and Sciences for release from courses during a particularly hectic
quarter when a contract deadline had to be met.
This manuscript was enriched by the assistance of and numerous conversations with directors and
staff of dozens of housing authorities as well as the willing participation of scattered-site housing residents in Tucson and Seattle, who shared their views about the program and its impact on their lives and
on their children's futures. A great deal of the credit for the success of scattered-site housing belongs to
them. Many who provided assistance did so anonymously by releasing documents, making statistical
data available, and completing questionnaires. I am grateful to the Housing Authority of Savannah for
pennitting me to incorporate highlights from Nathan Belzer's 1994 survey of national scattered-site
programs. More than a dozen other housing authority staff and directors completed followup surveys
and were interviewed during site visits conducted between 1987 and 1991. Their names and affiliations
are provided in a directory appendix. Each has agreed to serve as a resource to readers seeking additional information about a specific community's scattered-site program.
I am particularly indebted to two housing professionals: Dorothy L. Lengyel, executive director of
HomeSight, a nonprofit community development corporation engaged in affordable housing in Seattle,
Washington, and Alex Polikoff, executive director of the Business and Professional People for the Public
Interest. Dotti has read sections of the draft, helped collect the data, coauthored "Experiences With
Scattered-Site Housing" published in Urban Resources (Winter 1985), made research presentations to
local audiences with me, and supported my struggles to bring this report to fruition. Alex Polikoff's
immersion in issues of housing discrimination and housing mobility programs is extensive and
longstanding. His encouragement of my research, along with his reading of early sections of the manuscript and his willingness to provide access to his personal files on the litigation history of Gautreaux,
are deeply appreciated. Alex's vision of housing options for the poor, incorporated in both the Gautreaux
Demonstration program in Chicago and HUD's Moving to Opportunity demonstration, have influenced
my thinking about the role of scattered sites in our national housing policies.
Various parts of the manuscript have been read in one of its several versions by David Wilson of The
University of Toledo; Peter Nelson, formerly of the University of Sydney; my beloved wife Susan K.
Hogan; and HUD technical representative John Goering, Office of Policy Development and Research.
Each of these individuals bears major responsibility for any readability the present version reflects. The
deficiencies and errors of omission are mine alone.
Finally, I am grateful for the continuous, daily support tendered me by my wife, Sue, our two children, and my extended family. The quarter century of my family's patient support, and their being there
for me as I fought the fatigue and worry of whether my research would ever bear fruit will never be forgotten.
It is easier to trace the origin and subsequent diffusion of scattered-site developments than to pin
point the exact moment when the idea was born and what it initially meant. The late 1950s and early
1960s are recognized as the incubation period for the scattered-site program. The May and June 1957
issues of Architectural Forum, discussed earlier, contained seeds of thought about the appropriateness and appearance of scattered-site housing.
Mario Cuomo (1974), in seeking revisions to the controversial Forest Hills project, searched the
Congressional Record hoping to locate the etiology of the scattered-site concept in the legislative history
of the 1949 Housing Act. Cuomo found nothing particularly helpful-rather intentions "painted in broad
strokes" of the "need for housing generally, the particular plight of the low-income, the need for Federal
subsidy-all now accepted general propositions." (Ibid., p. 45) What Cuomo sought was the application
of those propositions to the Forest Hills controversy, which called for "a study of the scattered-site
refinement of the principle." (Ibid.)
Cuomo never did find a practical guide to scattered-site programming. While exploring revisions to
the Forest Hills project, he did develop an understanding about what such a program should look like.
Cuomo even chipped away at the "vagueness of the terminology" as reflected in the following portion
from his diary:
Ifscatter site is taken to mean the building of projects that are exclusively or largely low-income
in middle-income areas, then I'm not sure that it is a workable concept. It has no real history and
there is no firm basis on which to conclude that it will work. On the other hand, if scatter site is
taken to mean the dispersal of public assistance so as to permit low-income people to live in
nonracially restricted areas, then it is no longer a question of workability but rather a matter of
law. (Ibid., p. 95)
. 1960s Housing Experimentation
Until the 1960s, "several elements of the national housing program tended to be destructive of
community," especially to low-income families concentrated in the ghettos of large cities. (George
Schermer Associates, 1968) A number of new programs surfaced in the early 1960s that made it possible
to "intersperse low-income families in neighborhoods in which there [was] a wider range of incomes,
thus reducing the scale and intensity of colonization of the poor." (Ibid., p. 52) These programs included
scattered-site projects, acquisition programs, leasing programs, and turnkey housing. (Ibid., p. 50)
The 1965 Housing Act authorized major new initiatives to provide good housing for the poor "in
other than traditional public housing." (Congress and the Nation, 1965, p. 187) The leased-housing
and rent-supplement programs looked promising to supporters of housing dispersal. Ellickson (1967)
called the Housing and Urban Development Act of 1965 "potentially the most important development
in Federal housing policy since the advent of urban renewal in 1949." (Ibid., p. 518) The lease program
(Section 23) could be an effective way to integrate public housing tenants into middle-class neighborhoods,
because "such tenants would be inconspicuous and therefore unobjectionable to the neighborhood residents." (Luttrell, 1966)
Section 101 of the Act was even more promising as a tool for the integration ofneighborhoods. That HUD "planned to use the rent supplement program to break up the ghettos" was
apparent in what Ellickson said could be accomplished:
The program opened up communities which, if left to their own devices, would never make
affirmative efforts to provide housing for low-income families. In particular, the rent supplement
projects could be located on the comparatively cheap land available at the urban fringe, thus
aiding in the deghettoization of the central cities and of their suburbs. (Op. cit., p. 530)
The perfonnance of the rent supplement program was disappointing. Major inhibiting factors were
meager funding from Congress and stubborn resistance from the Appropriations Committees. "The
underlying motive," wrote Keith (1973), "appeared to be primarily racial and, especially, a desire to
protect white suburban communities from incursions by housing projects to be occupied predominantly
by low-income Negro families." (Ibid., p. 178) The program had "little success in achieving a range of
incomes in its projects or in dispersing them throughout the community." (Taggart, 1970, p. 57) Since
the program was forced to serve "an extremely low-income clientele," very few middle-income families
wanted to live in rent-supplemented projects, which, in tum, made the apartments "unattractive to more
affluent neighborhoods." (Ibid., p. 61)
Housing policy of the early and mid-1970s provided "the first significant changes and additions to
the national housing program in 16 years." (Schermer, op. cit., p. 18) The policy stimulated discussions
of new programs to address the shelter needs of poor families, including the scattered-site strategy, a
not altogether new concept. (Ibid., p. 50)5 It was at this time that the housing literature began to report
endorsements of scattered sites, suggesting that "few persons argue [ d] anymore that subsidized housing ... should not be scattered." (Grigsby, et ai., 1977) Outside of the more densely populated areas,
"several smaller projects, scattered throughout the community, would seem more suitable than the
massive high rises required in the big cities." (Ledbetter, 1967) Supporters wrote, "It has often been
proposed that public housing should be built in relatively small, scattered projects which would
blend into their surroundings"; (McEntire, 1960, p. 324) and "the emphasis upon small, scattered-site
developments" (along with several other new approaches) "constitutes a versatile and practical kit of
tools for ... localities having the imagination and initiative to utilize them." (Schermer, op. cit., p. 57)
Other political institutions were acting to foster greater choice in the housing market. President
Kennedy's 1962 Executive Order 11063, Equal Opportunity on Housing, was intended to "secure open
occupancy in federally assisted housing." The impact of the order was marginal as only "20 percent of
the new housing rparket, and less than 1 percent of the existing housing stock were covered." (Sorenson,
1965, p. 482) Title VI of the 1964 Civil Rights Act forbade discrimination on the basis of race, color, or
national origin in programs that received Federal assistance.6 However, these two initiatives did not
dismantle the dual housing market, nor did they create "equal access to housing, or achiev[e] housing
integration." (Lief and Goering, 1987, p. 232)
HUD also possessed a mechanism, through its regulatory procedures, to affect the placement of
assisted housing. In 1967 HUD undertook a new policy of equal opportunity regulation pursuant to the
Civil Rights Act of 1964. Section 205.1 of HUD's Low-Rent Housing Manual cited "efforts of housing
authorities to persistently locate public housing in the ghetto." (Columbia Journal ofLaw and Social
Problems, 1970, p. 266) The regulation provided that "any proposal to locate housing only in areas of
racial concentration will be prima facie unacceptable," and returned to the Housing Authority for reconsideration. The significance of the section was that it created, "after 30 years, an explicit Federal policy
against discriminatory site selection." (Ibid., p. 267) Evidence of HUD's utilization of Section 205.1 is
sparse, however, and the Columbia lournal ofLaw and Social Problems inferred:
Even accepting the contention that the requirement is new and that competing goals exist, it is
possible-rhat as a 3-year-old rule living amongst 30-year-old rules, Section 205.1 could conveniently become the family's foster child. (Ibid., p. 269)
After the urban riots of the late 1960s and passage of the 1968 Housing Act, the Federal Courts also
played a key role in the evolutionary history of scattered-site housing. lones v. Mayer,7 Shannon et al. v.
HUD, lames v. Valtierra, Gautreaux v. Chicago Housing Authority, Hills v. Gautreaux, and U.S. v.
Yonkers Board ofEducation were important in legitimizing open housing, forcing HUD to develop fair
procedures for selecting sites, mandating scattered-site housing, and limiting the development of additional public housing in areas of racial concentration.
Spatial deconcentration or "opening up the suburbs" strategies were responses to post-World War II
population shifts and demographic changes. The 1960 and 1970 censuses confirmed what academics,
journalists, and the media had already noted: Massive population movement within and between metropolitan regions had created white suburbs and a bipolar city. (Harrigan, 1989, p. 156) That these shifts
were fostering urban apartheid was recognized, according to Robert Wood, as early as the mid-1950s.
(Wood, in Warren, 1969, p. 63)
That transformation and its consequences were highlighted in the Kerner Commission report, which
investigated the causes of the dozens of urban riots that erupted between 1964 and 1967. Trends from
1950 to 1966 indicated that nearly all African-American population growth had occurred within metropolitan areas, primarily within central cities. The vast majority of white population growth (77.8 percent)
was in the suburban portions of metropolitan areas. Central cities were becoming increasingly African
American, with urban fringe areas almost entirely white. (Report ofthe National Advisory Commission
on Civil Disorders, 1968, p. 243)
Suburbanization, "the peripheral expansion of the metropolis," not only separated city from suburb,
but it created "territorial fragmentation" where hundreds of small local jurisdictions controlled "access
to residential, educational, and even recreational opportunities within their borders." (Danielson, 1976,
p. 46) The net effect of the nlovement and rearrangement of population was "the separation of neighborhoods along socioeconomic lines" and the isolation of "private resources from those most in need of
public services." (Ibid.) The institutionalization of the new urban reality eventually would provide the
rationale for the dispersion of low-income households throughout the metropolitan region.
"Apart from being objectionable on moral grounds," enrichment programs are said to impose "very
large cost[s] in terms of economic inefficiency, while making the solution of many social problems
inordinately difficult." (Kain and Persky, 1969, p. 80) Economists Kain and Persky support the dispersal option for two reasons: (1) the concentration of minorities in the inner city exacerbated a number
of problems, including distance from suburban job markets, inferior education for children, and family
dysfunctions associated with life in the ghetto; and (2) efforts to make the ghetto livable were not likely
to work. (Ibid.)
Taylor (1971) also compares ghetto enrichment with ghetto dispersal and finds the former defective
when used alone, even though it "runs with the tide by aligning its advocates at least for a period of time
with those who seek to preserve the existing segregated structure of the city." (Ibid., p. 234) He identifies a number of defects of the dispersal strategy: (1) the "policy changes required to overcome current
demographic trends toward massive segregation" (Ibid., p. 247); (2) the "prospect that there will continue to be a good deal of separation in residential patterns" (Ibid., p. 249); and (3) the status of new
African-American entrants to suburban neighborhoods as a "relatively powerless minority." (Ibid.)
Free choice is the key to an effective strategy where "individuals and families have the economic,
political, and psychological resources to be self-sufficient" and to "exercise ... control over their own
lives, to select their own priorities and styles of life." (Ibid., p. 229) Taylor does not advocate an either/
or position. Rather, he sees as "the only practical and sensible strategy" one that embodies "ghetto
enrichment and development with a program designed to disperse the ghetto and provide access to jobs,
schools, housing, and services on an integrated basis." (Ibid., p. 252)
The foremost proponent of residential mobility options for poor households was Anthony Downs.
Opening Up the Suburbs (1973) reveals the author's strong normative opposition to residential exclusion. According to Downs, "one deeply entrenched division," the legal and political separation between
central city and suburb, has generated three destructive outcomes: (1) the perpetuation of a "host of
problems by concentrating the burdens of coping with poverty inside central cities"; (2) the prevention
of suburbs from "achieving ... improvements in their efficiency and quality of life"; and (3) the undermining of suburban efforts to achieve "true equality of opportunity." (Ibid., p. vii)
Downs' solutions for residential segregation and·central-city deterioration have not lacked for critics.
Piven and Cloward (1966), for example, offer two reasons to reject his strategy.12 First, they argue,
integration will not be achieved. White majorities in the "receiving community" will veto any project
that "threaten[s] their neighborhood." (Ibid., p. 20) Second, residential desegregation ignores the importance of the "ethnic community as a staging area for groups to build the comn1unal solidarity and power
necessary to compel ... access to the mainstream of urban life." (Ibid.) Spatial deconcentration, of
course, vitiates such mobilization efforts.
Gathering baseline data from comparable jurisdictions enriches case-study research. Profile information on the universe of scattered-site programs, or a significantly large sample from that universe, provides an important context for interpreting information extracted from the single case study. A portion of
that context was drawn from the Hogan/Lengyel 1982 survey and the descriptive information contributed
a decade later by the Belzer survey of larger Public Housing Authority (PHA) programs.
Both surveys inquired into the origins and characteristics of scattered-site housing, and each sought
the perceptions of housing directors of program impact. The Hogan/Lengyel instrument contained more
than twice as many questions as Belzer's survey (36 items ,to 16), and the 1982 inventory included
questions on population characteristics, neighborhood involvement, and identification of problems when
communities initiated scattered-site housing.
A sample of 13 jurisdictions was selected from the housing authorities who responded to the 1982
survey. Personal visits to Alexandria, Charleston, Chicago, East St. Louis, EI Paso, Lancaster, Philadelphia, Rochester, St. Paul, San Diego, Tucson, and Yonkers were undertaken between 1987 and 1992.
Although not a random sample, the programs were diverse in terms of region, size of authority, program
size, and time of program origin. In two respects, sample jurisdictions differed from the larger 1982
survey: (1) more than 70 percent of the residents of 11 of the 13 PHAs were nonwhite (a reflection of
the large scattered-site programs in Chicago and Philadelphia) and (2) half the families worked, with
only 39 percent dependent on Aid to Families with Dependent Children (AFDC).
Chapter 3 begins with a description of national profile characteristics of scattered-site programs in
1982 and 1994. This is followed by a description of the case-study communities, including their dispersal, demographic characteristics of scattered-site census tracts in 1990, and how the scattered-site
program is viewed when compared with project housing. The chapter concludes with summaries of four
common elements for nine of the programs: (1) background material on the city, Housing Authority, and
assisted housing programs; (2) description of the origins and characteristics of the scattered-site program; (3) program results; and (4) 1994-95 updates of the program subsequent to the first site visit.
In the context of the Baltimore "political firestorm" that ended financing for the second year of the
program (New York Times, March 28, 1995), one can legitimately question the realism of supporting
choice-in-residency strategies that aim to integrate neighborhoods beyond the ghetto. Middle- and upperincome households have for more than four decades been deserting the city, "seeking social separation
from the lower classes as well as better housing and more spacious surroundings." (Erber, 1977, p. 306)
Most Americans now live in the suburbs. What is troubling about that statistic is that it signals more
balkanization and widening of the gap between city and suburb, "between the haves and the have-nots,
as more affluent people physically distance themselves from the problems of poor urban residents."
(New York Times, December 12, 1995) Moreover, the country's increasingly conservative, antigovernment political climate is unlikely to support programs to unslum the slum through dispersal strategies,
even those as benign as the Clinton administration's MTO demonstration.
A major aim of this study has been to examine the scattered-site strategy when undertaken by PHAs.
Several housing specialists leveled criticisms against some of these new housing programs that fostered
"mobility, freedom of choice, and inclusiveness for the metropolitan area as a whole, while simultaneously encouraging and fostering the revitalization of the central slum areas." (George Schermer
Associates, 1968, p. 44) The following are typical remarks:
• It's too expensive. The cost per unit "rises appreciably with a diminution of units per project."
(Ledbetter, 1967, p. 502) There is "good evidence that large, multibuilding projects can be
substantially cheaper than those built on scattered sites." (Taggert, 1970, p. 34)
• "Neither the benefits of mixing nor the costs of separation have been adequately quantified."
The "presumed educational and employment benefits to be gained from dispersing low-income
families throughout metropolitan areas do not yet seem so certain as to justify the huge subsidies
which such an effort would require." (Grigsby, p. 108)
• Scattered-site housing in minute quantities is an "inefficacious sprinkle of a cure on a massive
disease. The few relocated assisted persons to scattered-site clusters are merely remarginalized
by virtue of their systemic isolation." (Tein, p. 1463) "[E]ven quite expensive dispersal programs
could do no more than modestly reduce existing low-income concentrations." (Grigsby, p. 108)
• "[L]and is simply unavailable, except at very high prices, in those neighborhoods into which
the proponents of [the scattered-site] approach wish to send low-income families. Where land is
available in large parcels sufficient to accommodate three- or four-unit projects, the expense, the
zoning and building code regulations, and the vigorous opposition from the residents of the
invaded neighborhood would pose grave problems." (Ledbetter, p. 502)
• "It is probably impractical to assume that very low-income, problem-ridden white and Negro
families can be moved into middle-class neighborhoods, or into racially mixed developments
and be expected to develop viable, mutually supportive communities." (George Schermer Associ
ates, p. 36) Moving "unstable, problem-ridden families" into stable social environments "runs
head-on into the aspirations, the values, and the self-interests of the nonpoor. One cannot be
optimistic about implementing this element beyond a very nominal limit." (Ibid., p. 34)
Many of these observations remain undocumented hypotheses; a few require additional research
to demonstrate their validity. However, this study and the works of Rosenbaum (1995), Peterson and
Williams (1994), and Burchell, et aI., (1994) on housing mobility programs contain data that counter
some of the earlier doubts about the efficacy of spatial deconcentration strategies, including the
scattered-site housing model.
This study did not examine the costs and benefits of the scattered-site housing model. The Belzer
survey and several directors from the case-study communities discussed in chapter 3 believed the program was more expensive, however. Bratt's review of studies completed between 1968 and 1988 on the
benefits and costs of public housing found it to be:
Less costly to subsidize a household through the Section 8 Existing Housing program than to
subsidize the construction of a new unit of public housing ... the cost of subsidizing households
in existing [author's italics] public housing units is no higher than the cost of subsidizing households in existing private units through the Section 8 program [and] although no definitive conclusions can be drawn about the present cost of building new public housing in comparison to the
other types of subsidized housing programs that have been utilized in the past, public housing
does emerge in a relatively positive position. (Bratt, 1989, p. 73)
The benefits of residential mobility have been carefully examined by James Rosenbaum and his
associates at Northwestern University, who studied the social, educational, and employment progress
of Gautreaux families for more than a decade. The study found the Gautreaux "[s]uburban movers were
significantly more likely to be employed after their move than Gautreaux families" who remained in the
city, and "children in suburban Gautreaux families generally fared well in school." (Housing andDevelopment Brief, 1994, p. 4) The "geography of opportunity" premise that "where someone lives has an
important impact on his or her social and economic prospects" (Rosenbaum, 1995, p. 266) was confinned, but needs to be replicated.20
Scattered-site housing, particularly the conventional public housing model examined in this study, remains an "inefficacious sprinkle" compared to the need for affordable housing. The Hogan/Lengyel and Belzer surveys reported that scattered-site units account for only 8 to 9.5 percent of the PHAs' inventory. Thus even though the program appears to produce highly successful developments, it will not make much of a dent in the broader problem where "[n]early three-fourths of very low-income renters pay over 30 percent of their income for housing, and more than 4 in 10 pay over half of their income for housing." (Bogdon, et al., 1992, p. 11) Downs (1994) and Rusk (1996) offer illustrations of how current mobility programs-Gautreaux in Chicago and inclusionary zoning in Montgomery County, Maryland-could be expanded and the "volume of out-movement" increased sharply. (Downs, op. cit., p. 108) Using figures from 44 of the Nation's largest cities, Downs calculates that "setting aside 5 percent of suburban housing units built in the 1980s would have pennitted 12.7 percent of the 1980 inner-city households (in the 44 areas combined) to move to the suburbs. In 13 of these areas the share would have exceeded 30 percent." (Ibid., p. 110) Rusk demonstrates how a significant reduction in Baltimore City's concentrated poverty population could be realized if a "Montgomery County-type, Moderately Priced Dwelling Unit policy were established" for the seven-county Baltimore Metropolitan Area. (Rusk, 1996, p. 111) The genius of Montgomery County's strategy of targeting communities in the "process of being created" is that there are "no neighbors already in place in the proposed subdivision, townhouse project, or apartment complex who are organized to object."
In that enlarged context, the major goal of housing policy is to equalize opportunities across space and social groups. This will require Federal leadership to n1itigate serious local constraints such as political fragmentation, fiscal crisis, exclusionary land-use regulations, unequal public expenditures for education, and the dual housing market. ... If housing policy is to equalize opportunities across space, the political agenda will require addressing these political realities.