Are public housing projects good for kids?

https://doi.org/10.1016/S0047-2727(99)00065-1Get rights and content

Abstract

One goal of federal housing policy is to improve the prospects of children in poor families. This paper examines the effect of public housing participation on housing quality and educational attainment. Using the SIPP, we show that living in projects is associated with more negative outcomes for children, although this appears to be due to unobserved heterogeneity. We control for the endogeneity of project participation using TSIV techniques which combine information on project participation from the CPS with information on outcomes from the Census. We find that project households are less likely to suffer from overcrowding or live in high-density complexes. Project children are less likely to have been held back. Thus, our results run counter to the stereotype that housing projects harm children.

Introduction

Since 1937, the federal government has subsidized the housing costs of some low-income families, with the stated goal of improving the quality of housing inhabited by the poor. Given that poor families with children make up 60 percent of the public housing caseload (most of the rest are households headed by the elderly and/or disabled), it is clear that a second important goal is to improve the life-chances of recipient children.

The real costs of this assistance (in 1996 dollars) have grown steadily over time, from $7.3 billion in 1977 to $26 billion in 1996. The number of households assisted has also risen from approximately 3.2 million in 1977 to 5.7 million in 1996, and annual outlays per unit have approximately doubled over the same period to $5480 (Committee on Ways and Means, 1996). However, public dissatisfaction with large public housing projects has remained high. More than 25 years ago, Henry Aaron wrote that “Over the years public housing has acquired a vile image — highrise concrete monoliths in great impersonal cities, cut off from surrounding neighborhoods by grass or cement deserts best avoided after dark…. This image suggests that any benefits inhabitants derive from physical housing amenities are offset by the squalid surroundings” (Aaron, 1972, p. 108). Many would argue that, if anything, the situation has worsened, as horrifying stories about large projects such as the Robert Taylor Homes or Cabrini Green in Chicago routinely appear in the national news.

As a result, the character of low-income housing aid has changed dramatically over time,1 as money has been diverted away from ‘project-based’ aid toward ‘household-based’ aid given in the form of certificates and vouchers that can be applied toward rents in the existing private housing market.2 Moreover, since 1982, appropriations for new construction of public housing projects have fallen sharply (Committee on Ways and Means, 1996).3 And in 1995, the Department of Housing and Urban Development (HUD) put forth a plan that would have eventually replaced all ‘project-based’ assistance with housing certificates provided directly to individual households (Government Accounting Office, 1995).4

The aim of voucher/certificate programs is to assist families without consigning them to the projects. But newspaper accounts not withstanding, there is little evidence that projects actually harm children. Basic economics suggests that families would not move into public housing projects unless they were better in at least some respects than the alternatives they faced. Aaron’s intriguing hypothesis is that families in projects tradeoff physical housing amenities and reductions in rental payments against neighborhood characteristics that are bad for their children. But many projects and project neighborhoods may actually be superior to the housing and neighborhoods that families would have occupied in the absence of assistance. And reductions in rental payments may or may not be spent on goods and services beneficial to children. Thus, it is important to look directly at the effects of housing assistance on housing quality and on child well-being.

We first use data from the 1992 and 1993 waves of the Survey of Income and Program Participation (SIPP) to show that living in a project is associated with poorer outcomes, a finding which provides a baseline for our subsequent analyses. However, we find that this sample is too small to yield reliable estimates using instrumental variables techniques. We turn instead to the two-sample instrumental variable (TSIV) technique developed by Angrist and Krueger, 1992, Angrist and Krueger, 1995 to combine information on the probability of living in a project obtained from the 1990 to 1995 waves of the March Current Population Survey, with information on outcomes obtained from the 1990 Census. The instrument common to both samples is an indicator equal to one if the household is entitled to a larger housing unit in a project because of the sex composition of the children in the household. Families entitled to a larger unit based on the sex composition are 24 percent more likely to live in projects. Using TSIV to control for unobserved characteristics of project residents, we find that project families are less likely to suffer from overcrowding and more likely to live in buildings with fewer than 50 units. And children in these families are 11 percentage points less likely to have been held back in school one or more grades. Thus, there is little evidence that the typical child living in a housing project is harmed by being there, and there is some evidence that living in projects may actually improve both living conditions and child outcomes.

The rest of the paper is laid out as follows: Section 2 gives additional background information about the public housing programs. Section 3 discusses methods, while Section 4 describes the data. Results appear in Section 5, and a discussion and conclusion follow.

Section snippets

Background

As noted above, public housing ‘projects’ tend to have very bad reputations. Yet, the publicity generated by the worst projects tends to obscure great heterogeneity between projects. Approximately 3300 public housing authorities own and operate about 13,200 developments with a total of about 1.4 million units. Seventy percent of these authorities operate fewer than 300 units, while the 40 largest agencies operate 1786 or more units and account for 36 percent of all public housing project units.

Methods

The 1992 and 1993 waves of the SIPP have information about both project participation and interesting child outcomes, which enable us to estimate baseline ordinary least squares (OLS) regressions of the effects of project participation on child outcomes. To do so, we estimate models of the formOUTCOME01PROJ2X+u,where the OUTCOME variables include measures of housing, neighborhood, school quality, and grade repetition which are discussed in greater detail below, and X is a vector of

Data

The data from the SIPP comes from topical modules that were added to the ninth wave of the 1992 panel, and the sixth wave of the 1993 panel. Thus, all of the questions were asked in 1994. In addition to questions about public housing participation, parents were asked to subjectively rate their home, neighborhood, and the quality of the school on a scale of 0 to 10. Parents were also asked whether their children had repeated any grades for any reason, about the number of times that their

OLS estimates of the effects of project participation

OLS estimates of Eq. (1) are shown in Table 2 for the SIPP outcome variables. We find that after controlling for observable characteristics, households in projects are still less satisfied with their housing and their neighborhood than other households. Moreover, project children are still more likely to have changed schools. However, there are no significant differences between project and other children in school ratings, extra-curricular activities, and grade retention once observables are

Discussion and conclusions

Although it is widely assumed that public housing projects are bad for children, there is little empirical research on this question. A likely reason is that there are few large data sets that combine information about project participation, housing quality, and child outcomes. In this paper, we combine information from several sources in order to take a first look at the effects of project participation on housing quality and on educational attainment, a very important child outcome.

In view of

Acknowledgements

We thank Joshua Angrist, Caroline Minter Hoxby, Lawrence Katz, Jeffrey Kling, Edgar Olsen, Steve Pishke, James Poterba, two anonymous referees, coeditor Thomas Piketty, and seminar participants at Harvard, MIT, the NBER Summer Institute, RAND/UCLA, Michigan, Michigan State, UC Berkeley, UC Davis, and UC San Diego for helpful comments. Janet Currie thanks the Canadian Institute for Advanced Research and the NICHD for support under grant #1R01 HD 31722.

References (36)

  • E. Olsen et al.

    The benefits and costs of public housing in New York City

    Journal of Public Economics

    (1983)
  • H. Aaron

    Shelter and subsidies: who benefits from federal housing policies

    (1972)
  • J. Angrist et al.

    Children and their parent’s labor supply: evidence from exogenous variation in family size

    American Economic Review

    (1998)
  • J. Angrist et al.

    The effect of age at school entry on educational attainment: an application of instrumental variables with moments from two samples

    Journal of the American Statistical Association

    (1992)
  • J. Angrist et al.

    Split-sample instrumental variables estimates of the return to schooling

    Journal of Business and Economic Statistics

    (1995)
  • W. Apgar

    Which housing policy is best?

    Housing Policy Debate

    (1990)
  • J. Atlas et al.

    From ‘projects’ to communities: redeeming public housing

    Journal of Housing

    (1993)
  • B. Barrington et al.

    Differentiating characteristics of high school graduates, dropouts, and nongraduates

    Journal of Educational Research

    (1989)
  • K. Butcher et al.

    The effect of sibling sex composition on women’s education and earnings

    The Quarterly Journal of Economics

    (1994)
  • R. Cairns et al.

    Early school dropout: configurations and determinants

    Child Development

    (1989)
  • Case, A., Katz, L., 1991. The company you keep: the effects of families and neighborhoods on disadvantaged youth....
  • D. Coggon et al.

    Housing in early life and later mortality

    Journal of Epidemiology and Community Health

    (1993)
  • Condon, M., 1991. Public housing, crime, and the labor market: a study of black youths in Chicago. Harvard University,...
  • Crews, A., 1996. Do housing programs for low-income households improve their housing? Maxwell School of Citizenship and...
  • M. Ensminger et al.

    Paths to high school graduation or dropout: a longitudinal study of a first-grade cohort

    Sociology of Education

    (1992)
  • C. Fischer et al.

    Crowding studies and urban life: a critical review

    Journal of the American Institute of Planners

    (1975)
  • O. Galpin et al.

    Helicobacter pylori infection and overcrowding in childhood

    Lancet

    (1992)
  • J. Grissom et al.

    Repeating and dropping out of school

  • Cited by (194)

    • Investing in early childhood development in preschool and at home

      2023, Handbook of the Economics of Education
    • Housing vouchers reduce residential crowding

      2022, Journal of Housing Economics
      Citation Excerpt :

      Researchers have long been interested in the direct effects of housing assistance on recipient households. Popular outcomes of interest include school outcomes (Currie and Yelowitz, 2000; Horn et al., 2014), economic self-sufficiency (Kling et al., 2007), and general well-being (Katz et al., 2001). In relation to overcrowding, Currie and Yelowitz (2000) use an IV strategy to conclude that public housing reduces overcrowding.

    View all citing articles on Scopus

    The STATA programs and data used in this study may be obtained from the authors.

    View full text