Broken Promises or Selective Memory Planning? A National Picture of HOPE VI Plans and Realities

Abstract

Government efforts to redevelop public housing often face a contentious gap between plans and realities. This paper compares 2014 U.S. Department of Housing and Urban Development (HUD) administrative data on housing unit counts and unit mixes for all 260 developments receiving Housing Opportunities for People Everywhere (HOPE VI) revitalization grants with data provided in the original HOPE VI grant award announcements. We find that HUD records undercount approximately 11,500 once-proposed units. The biggest changes were a 29% decline in the number of market-rate units and a 40% decline in homeownership units. The chief shortfall during implementation, therefore, was not with public housing units (although the HOPE VI program as a whole did trigger an overall decline of such units). To help elucidate the dynamics at play when the unit allocation shifts between initial grant award and implemented project, we include a series of five brief case studies that illustrate several types of unit change. Interviews with HUD staff confirm the baseline for record-keeping shifted during implementation once project economic feasibility became clearer; adherence to original unit mix proposals remained secondary. HUD prioritized its accountability to Congress and developers over its public law accountability to build the projects initially proposed to local community residents. Although these changes have sometimes been interpreted as broken promises, it is even clearer that HUD’s monitoring system exemplifies what we call Selective Memory Planning: when planners and policy makers, willfully or not, selectively ignore elements of previous plans in favor of new plans that are easier to achieve.

Keywords:

• Public housing
• income mixing
• HOPE VI
• implementation
• accountability

Acknowledgment

The authors are grateful for the assistance of Mark Joseph in obtaining HUD data, and for the assistance of Esther Chung and Diana Searl at MIT's Resilient Cities Housing Initiative.

Notes

1. HUD records for HOPE VI use the term affordable housing to connote housing that has some subsidy, but does not have the full deep subsidy associated with public housing. The definitional threshold for what counts as affordable housing tends to be highly variable.

2. The reports also list a large amount of detail on funding, as well as many pages on when units were produced and of what type, a phase compliance checkpoint report, and a phase narrative report. These sections are not the focus of this article and so are not discussed further here.

3. HUD continues to use the management system to track ongoing HOPE VI projects. We also have seen no evidence to suggest that the data reported in the management system is inaccurate for that point in time at which it is reported. We accept this to be the case to make data analysis feasible.

4. Although it would have been ideal to cross-reference these award announcements with the data contained in the original HOPE VI applications, HUD did not systematically retain copies of those applications.

5. Although it would be preferable to have the original grant announcements or grant proposals, HUD has been unable to locate them, despite repeated requests. As a result, we do not currently have information on some grants from the years 1994–1996.

6. Public housing residents and neighbors may have been exposed to project presentations separate from newspaper reports, since HUD required community consultation prior to submission of a HOPE VI proposal.

7. Such shortfalls may be perceived quite differently, however, since losses of public housing are subtracted from housing that was previously 100% public housing, whereas losses in market-rate units merely entail a smaller addition of such units to a place that formerly lacked any of them.

8. The mean, however, is skewed upward by the outlier Cabrini-Green in Chicago, Illinois, which experienced a large increase in proposed market-rate units after its initial HOPE VI proposal had been submitted.

9. As one anonymous reviewer usefully pointed out, the number of bedrooms proposed can be just as important as the number of units. To understand the social impact of a mid-course shift in unit provision in a way that would reveal which types of families will likely be accommodated, it is just as important to count the number of beds as it is to keep track of the number of front doors. Unfortunately, the dashboard data supplied by HUD do not provide sufficient detail about the number of bedrooms per unit to enable this article to provide a systematic analysis of this aspect.

10. Although all interviews with past and present HUD career officials and consultants were on the record, given current political volatility, we have left them anonymous.

11. HOPE VI projects also receive widely differing local funding matches. In the first round of HOPE VI funding, for example, McGuire Gardens in Camden, New Jersey, was redeveloped using 96.6% HOPE VI funds, whereas Mission Main in Boston, Massachusetts, relied on HOPE VI for only 31.3% of its funding (Holin et al., 2003, 26). Thereafter, as mixed-finance projects became increasingly ubiquitous, HOPE VI grantees frequently had to rely on the uncertainties of tax credit allocations, and often faced uncertainty about the availability of local matching funds (GAO, 2002). It is hardly surprising, then, that the complexity of funding can attenuate the implementation timetable or present insurmountable obstacles, making it more likely that realized projects will differ from initial proposals.

12. Of those fully completed projects where full income-mix data are available, there has been only a 2% loss of units between T2 and T3. But since the incomplete projects are larger on average, and have higher percentages of market-rate units and homeownership units, it is unclear what this will portend. Either these will get built out and start to mitigate the loss of these higher income units that occurred between T1 and T2—or the same trend established between T1 and T2 will continue, yielding an even greater shortfall in higher income units and a corresponding skew of income mixes toward lower income occupancy.

13. For example, SMP could be applied to climate change adaptation, transportation and infrastructure, business improvement districts, environmental protection, international development, or other arenas.

14. We thank an anonymous reviewer for suggesting these points.