This PDF is a selection from a published volume from the
National Bureau of Economic Research
Volume Title: Means-Tested Transfer Programs in the United
States
Volume Author/Editor: Robert A. Moffitt, editor
Volume Publisher: University of Chicago Press
Volume ISBN: 0-226-53356-5
Volume URL: http://www.nber.org/books/moff03-1
Conference Date: May 11-12, 2000
Publication Date: January 2003
Title: Housing Programs for Low-Income Households
Author: Edgar O. Olsen
URL: http://www.nber.org/chapters/c10259
365
6.1 Introduction
Unlike other major means-tested transfers, no low-income housing pro-
gram is an entitlement for any type of household. Despite the failure to
serve all eligible households who want to participate, federal, state, and lo-
cal governments in the United States spend substantially more on housing
subsidies to the poor than on other better-known parts of the welfare sys-
tem such as food stamps and Temporary Assistance for Needy Families
(TANF). The most widely cited figures for government expenditures on
housing subsidies refer to the direct expenditures of the U.S. Department
of Housing and Urban Development (HUD). They ignore the U.S. De-
partment of Agriculture (USDA) programs that account for about 20 per-
cent of all subsidized units, the tax expenditures on the Low-Income Hous-
ing Tax Credit ($3.5 billion per year in 2000 and growing rapidly), the
expenditures of state and local governments often funded by block grants
from the federal government, and the many indirect subsidies such as local
property tax exemptions and abatements received by all public housing
projects and many privately owned projects, the federal income tax ex-
emption of interest on the bonds issued by state and local governments to
finance housing projects, and the underpriced mortgage insurance re-
ceived by many privately owned projects.
Given the enormous amount of money that has been spent on means-
6
Housing Programs for
Low-Income Households
Edgar O. Olsen
“Those who do not remember the past are condemned to repeat it”
—George Santayana
Edgar O. Olsen is professor of economics at the University of Virginia.
The author is grateful to Neil Seftor for superb research assistance, Robert Olsen and
Michael Schill for detailed comments, and Fred Eggers, Dennis Fricke, Bob Gray, Steve
Malpezzi, Jim Mikesell, Kathy Nelson, Carla Pedone, Joe Riley, John Quigley, Mark Shroder,
Jim Wallace, and John Weicher for helpful conversations.
tested housing assistance over the years, the amount of research on the
most important effects of these programs is shockingly small. There is no
evidence on the effects of some major programs and little evidence on the
effects of other large programs. For example, there are only two studies of
the most important effects of Section 236, a program that still serves al-
most a million people. There are almost no studies of the important effects
of the Low-Income Housing Tax Credit (LIHTC), a program that has been
the fastest-growing housing program over the last decade accounting for
the majority of additional recipients of housing subsidies. The evidence on
the most heavily studied programs is old. For example, the studies of the
effects of the public housing program on housing consumption are based
on data from 1965 through 1977. At the midpoint of this period, the me-
dian age of public housing units was ten years. Today, the median age is
about thirty years. It stands to reason that the effect of public housing on
the housing consumption of tenants is very different today from what it
was at the time of these studies. Since direct HUD expenditure on public
housing in the form of operating and modernization subsidies is about $6
billion annually and the real resource cost of continuing to use these units
to house low-income households is much larger due to their opportunity
cost and the substantial local property tax abatement that they receive, this
is a sad state of affairs.
If a housing program consists of a set of eligible households and suppli-
ers operating under one set of rules, the United States has had an enor-
mous number of programs intended to improve the housing of low-income
households since the federal government became seriously involved in this
activity in 1937. There have been many programs as this term is commonly
used, each of these programs has typically had a number of variants, and
each change in the regulations produces a new program. For example, the
public housing program has at least twenty-nine variants. Each has its own
rules, and these rules have changed from time to time, although they have
many rules in common.
To keep the length of this paper within reasonable limits, it will focus on
four broad programs that account for the bulk of all subsidized rental units.
These programs are public housing, project-based assistance under the
Section 236 and Section 8 New Construction/Substantial Rehabilitation
Programs, and tenant-based assistance under the Section 8 Existing Hous-
ing Program. They illustrate the three basic approaches that have been
used to provide housing assistance: government ownership and operation
of housing newly built for occupancy by low-income households; govern-
ment contracts with private parties to build (or substantially rehabilitate)
and operate housing for these households; and subsidies to eligible house-
holds who select housing in the private market meeting certain minimum
quality standards and, under some variants, other restrictions. These pro-
grams account for about 70 percent of all subsidized rental units and about
366 Edgar O. Olsen
50 percent of all units for low-income households that have received fed-
eral housing subsidies.
Even though there have been almost no studies of their major effects, a
more comprehensive paper would have devoted considerable attention to
the LIHTC program (that serves more than a million households and has
been growing rapidly since 1988), Section 515 (a rental housing program
operated by the USDA in rural areas and small towns that produces few
new units each year but still serves more than half a million households),
and the HOME housing block grant program from the federal government
to state and local housing agencies on which we currently spend about
$1.85 billion annually, primarily for project-based assistance. In light of the
continuing interest in increasing the homeownership rate of the near poor,
it would also be desirable to devote considerable attention to several fairly
large homeownership programs—HUD’s Section 235 program, which
provided subsidies to about a half a million households since 1969, and
USDA’s Section 502 program, which has subsidized almost two million
households since 1949 and currently provides subsidies to about half a mil-
lion households.
The primary purposes of this paper are to (a) consider the arguments
that have been offered for housing subsidies to low-income households and
the implications of valid arguments for the evaluation and design of hous-
ing programs, (b) describe the most important features of the largest rental
housing programs for low-income households in the United States, (c)
summarize the empirical evidence on the major effects of these programs,
and (d) analyze the most important options for reform of the system of
housing subsidies to low-income households. The effects of these programs
that will be considered include effects on the housing occupied by recipi-
ents of the subsidy and their consumption of other goods, effects on labor
supply of recipients, the participation rates of different types of house-
holds, the distribution of benefits among recipients and all eligible house-
holds, effects on the types of neighborhoods in which subsidized house-
holds live and the effect of subsidized housing and households on their
neighbors, effects on the rents of unsubsidized units, and the cost-
effectiveness of alternative methods for delivering housing assistance.
Since we continue to seriously consider or embark upon new programs that
have the same basic features of older programs and the major effects of the
newer programs have not been estimated, an understanding of the older
programs is highly relevant for current discussions of housing policy.
Section 6.2 discusses the justifications for housing subsidies for low-
income households and goals consistent with these justifications, gives a
brief overview of the development of the current system of housing subsi-
dies, and describes in more detail the development and most important
rules of the major rental housing programs. Section 6.3 provides informa-
tion about the number of households served by major programs, direct
Housing Programs for Low-Income Households 367
federal expenditures on these programs, and the characteristics of the
households assisted. Section 6.4 discusses what can be said on theoretical
grounds about the effects of the programs and reviews the evidence on
these effects. Section 6.5 analyzes options for reform of the system of hous-
ing subsidies to low-income households. Section 6.6 summarizes the major
results.
6.2 Program Justifications, Goals, History, and Rules
6.2.1 Justifications and Goals
Without a clear understanding of the justifications for government ac-
tivity in a particular area, it is difficult (but not impossible) to conduct an
incisive evaluation of current programs or design better programs. For ex-
ample, a person who has not thought seriously about the justifications for
housing subsidies to low-income households might imagine that a housing
program for these households is successful if it induces them to occupy bet-
ter housing. However, a program of cash grants with no strings attached
would have this effect for all recipients whose income elasticity of demand
for housing is positive—that is, almost everyone. Indeed, it would be pos-
sible to devise a subsidy that led to a smaller improvement in housing for
all recipients than would result from an equally costly lump-sum grant. A
subsidy equal to a fraction of expenditure on all goods except housing
would have this effect for recipients whose price elasticity of demand for
these other goods is less than one. Should this be considered a successful
housing program?
In general, a justification for a government program is an explanation of
why we should have a program of that type. Obviously, this involves value
judgments. The value judgment underlying this paper is that we should
have a program of a particular type if and only if an appropriately designed
program of that type will lead to an efficient allocation of resources that is
preferred by everyone to the allocation in the absence of government ac-
tion. Although this simple view leaves much to be desired because it ig-
nores the impact of the multiplicity of external effects, market imperfec-
tions, and informational problems that justify other programs and hence
the design of a set of programs to deal simultaneously with all of these
problems, it is at least a step in the direction of clear thinking about policy
evaluation and design.
The major justification for housing subsidies to low-income households
is that some taxpayers care about these households but feel that at least
some low-income households undervalue housing. If some taxpayers feel
this way while others are either completely selfish or nonpaternalistic al-
truists, it is possible for the government to achieve an efficient allocation
and to make everyone better off as they judge their own well-being by pro-
368 Edgar O. Olsen
