Has Hope Vi Transformed Residents' Lives? New Evidence From The Hope Vi Panel Study

Abstract

Since the early 1990s, the HOPE VI program has been the United States' largest, most ambitious community revitalization program. HOPE VI sought to transform not only distressed public housing communities, but also to transform residents' lives and help them to become self-sufficient. This paper uses new evidence from the HOPE VI Panel Study on how HOPE VI families have fared. The long-term findings from this research paint a more positive picture than many critics had predicted, showing that the program has had profound benefits for many public housing families—particularly those who have relocated to less poor communities. However, the long-term results highlight the limitations of the HOPE VI approach, particularly the lack of impact on economic outcomes. These findings point to the need for new and creative strategies for addressing some of the worst consequences of concentrated poverty.

Keywords:

• Housing policy
• urban regeneration
• social housing
• HOPE VI

Acknowledgements

The authors wish to thank their collaborators on the HOPE VI Panel Study research: Megan Gallagher, Jennifer Comey, Mark Woolley, Elizabeth Cove and Beata Bajaj.

Notes

¹ The studies were conducted by The Urban Institute and its partner, Abt Associates Inc. The HOPE VI Resident Tracking Study was supported entirely by a grant from HUD. The HOPE VI Panel Study was supported by a consortium of funders, including HUD; the John D. and Catherine T. MacArthur Foundation; the Annie E. Casey Foundation; the Rockefeller Foundation; the Robert Wood Johnson Foundation; the Fannie Mae Foundation; and the Chicago Community Trust.

² For critiques of HOPE VI and the potential impact on residents, see Bennett et al. (2006); Goetz (2003); Keating (2001); Venkatesh et al. (2004).

³ See Popkin et al. (2004b) for a discussion of the rate of return issue.

⁴ Because of the retrospective design, the sample under-represents unassisted tenants and others who were more difficult to locate. In general, those who are difficult to find are those who move frequently, double-up with another family, are homeless, or have moved out of the area. These former residents are likely to have experienced more problems than those we were able to survey.

⁵ Other research has examined the issue of neighborhood impact. For a review, see Turner et al. (2009).

⁶ For a full discussion of this issue, see Buron et al. (2007), p. 2.

⁷ The level of problems reported was substantially higher than that for poor renters nationally in the American Housing Survey. See Popkin et al. (2002) for a discussion of these issues.

⁸ All respondent names are pseudonyms.

⁹ For other studies that have examined rates of return, see Buron et al. (2002); Holin et al. (2003); and National Housing Law Project (2002).

¹⁰ Perceptions of disorder are highly correlated with crime rates, and are often a better predictor of levels of fear (Perkins & Taylor, 1996).

¹¹ See Sampson et al. (1997). The concept of ‘collective efficacy’ comes from the Project on Human Development in Chicago Neighborhoods and is intended as a measure of neighborhood health. It is highly correlated with crime rates and other indicators such as low birth weight.

¹² With such small numbers of respondents living in mixed-income communities, it is not possible to see accurate statistical trends, but given that they experienced the same improvements in housing quality and neighborhood safety, it is likely that they have experienced the same benefits in terms of quality of life as those who received vouchers.

¹³ Trends for the small numbers of homeowners, unassisted renters, and mixed-income movers were similar, but the sample sizes are too small to permit meaningful analysis.

¹⁴ Behavior Problems Measure: Respondents were asked to indicate how often the child exhibited any one of the seven specific negative behaviors, taken from the Behavior Problems Index: trouble getting along with teachers; being disobedient at school; being disobedient at home; spending time with kids who get in trouble; bullying or being cruel or mean; feeling restless or overly active; and being unhappy, sad, or depressed. The answers ranged from ‘often’ and ‘sometimes true’ to ‘not true’. The study tracked the proportion of children whose parents reported that they had demonstrated two or more of these behaviors often or sometimes over the previous three months.

¹⁵ Positive Behavior Measure: This scale requires respondents to rate how closely each of the following six positive behaviors describes their child: usually in a good mood; admired and well liked by other children; shows concern for other people's feelings; shows pride when doing something well or learning something new; easily calms down after being angry or upset; and is helpful and cooperative. The list of behaviors was derived from the 10-item Positive Behavior Scale from the Child Development Supplement in the Panel Study of Income Dynamics. Each behavior was rated on a scale ranging from 1 (‘not at all like this child’) to 5 (‘completely like this child’). The study tracked the proportion of children with at least five out of six behaviors rated relatively high (‘a lot’ or ‘completely like this child’).

¹⁶ Delinquent Behavior Measure: Respondents were asked if over the previous year their child had been involved in any of the following five activities: being suspended or expelled from school; going to a juvenile court; having a problem with alcohol or drugs; getting into trouble with the police; and doing something illegal for money. The study tracked the proportion of children involved in two or more of these behaviors.

¹⁷ See, for example, Buron et al. (2002) and Orr et al. (2003).

¹⁸ Many health problems vary significantly by gender and race, and because over 88 per cent of the adults in the HOPE VI Panel Study are women and 90 per cent are black, a sample of black women nationally is used as the comparison group. The national data cited in this testimony are published by the US Department of Health and Human Services, calculated from the National Health Interview Survey in 2005. National Health Interview Survey data are broken down by sex and race, but not further by poverty status. Nationally, approximately one-third of all black women live in households with incomes below the poverty level. Therefore, the comparison data are biased slightly upward in terms of better health because of the relatively better economic well-being of the national population of black women compared with the HOPE VI sample. However, even limiting the comparisons to similar gender, race and age groups, adults in the HOPE VI study experience health problems more often than other demographically similar groups.

¹⁹ Indication of mental health was based on a scale derived from the CIDI-12, or Composite International Diagnostic Interview Instrument. The series includes two types of screener questions that assess the degree of depression and the length of time it has lasted. The index is then created by summing how many of the seven items respondents reported feeling for a large share of the past two weeks. If a respondent scores three or higher on the index, their score indicates a major depressive episode.

²⁰ The study tested the difference in the probability of employment with and without a specific employment barrier for an unmarried, high-school-educated, African-American female respondent using a housing voucher and facing no additional employment barrier. Unless otherwise noted, statistical significance is reported for probability values of 5 per cent or less.

²¹ A GED is the equivalent of a high school diploma.