Estimating the local economic impacts of HOPE VI

Abstract

This study identifies and measures the demonstrable changes to local political economies that can be reasonably attributed to HOPE VI redevelopments. It examines the extent to which the developments have contributed to increases in surrounding property values, decreases in serious crimes, additional regional economic activity, and changes in local tax revenues. It weighs these benefits against the public costs associated with the program.

Despite the expenses associated with HOPE VI, the redevelopments generate significant net social welfare benefits. In most cases, the collective tenant and neighborhood benefits exceed the net public costs of redevelopment. In addition, the redevelopments spark additional regional economic activity and contribute to an increase in the local tax base. HOPE VI's effects are far from uniform, however, and depend on the location of the redeveloped property, the characteristics of project funding, the strength of the local real estate market, and the presence of other development pressures.

Keywords:

• HOPE VI
• development/revitalization
• neighborhood

Notes

¹Such costs and benefits are difficult to calculate, particularly since they may increase or decrease over time. (The social costs of relocation, for example, presumably diminish as tenants build friendships and social networks in their new communities.) Although the research on HOPE VI resident experiences is somewhat mixed (Clampet-Lundquist 2004; Popkin et al. 2004), our suspicion is that the benefits of leaving severely deteriorated public housing outweigh the costs. If nothing else, the tenants' new location – be it within the redeveloped site or not – tends to be safer than their previous residence (Popkin and Cove 2007).

²Turner et al. (2007) estimated the costs and benefits associated with three prototypical properties, using a variety of national averages and assumptions from previous research. The study did not attempt to document actual benefits resulting from specific properties, however.

³The Section 8 costs subtract out the landlords' assumed 15% profit margin.

⁴The value of implied subsidies to public housing tenants are computed over a 30-year horizon using a 6% discount rate. Because of uncertainty regarding the expected future value of crime benefits, these have been set to zero for periods beyond 2006. Thus, the estimates of crime benefits are likely to be biased downward.

⁵For example, we determined that the southern impact area surrounding the Townhomes on Capitol Hill (Washington) extended only to Interstate 295. The various impact area maps are available from the authors upon request.

⁶The specifications for our model are found in Appendix A. Because Kansas City regulations prevent commercial data vendors from using the city's property sale information in their databases, we have relied on 1990 and 2000 median home value data at the census block group level to estimate changes in the areas surrounding Guinotte Manor and Villa Del Sol. We associate the difference between the change in the HOPE VI block groups and that of the city overall to the HOPE VI redevelopment. Because we cannot tie changes in property value trends to specific events in Kansas City, our ascribed real estate impacts in these two areas must be viewed as considerably more uncertain than those in the other cities. Appendix B provides the results of our adjusted interrupted time series analysis for New Holly in Seattle. The findings for other sites are available from the authors upon request.

⁷We explored but generally were unable to distinguish separate effects for the developments' announcement, demolition, move-in, and completion.

⁸HOPE VI projects are the antitheses of randomized experiments; the selection of sites by the housing authorities and HUD involved a variety of factors, including those likely to enhance the likelihood of community impact. It is not possible to separate out the impacts of these factors, especially given the limited number of HOPE VI projects.

⁹The publicly owned portion of a HOPE VI property also has value, though the amount of that value depends on the purpose for which it is used. From an economic perspective, we treat the public portion of a HOPE VI property as a 30-year asset. Absent regular infusions of replacement capital (data for which are unavailable, and thus have not been counted in the total HOPE VI costs), the buildings' economic value will depreciate to 0 over the 30 years. Assuming no restrictions on land use, the residual value of the property at that point would consist primarily of the value of the land. (Given the unknown dynamics of the local market at that point, estimating the residual value is impossible.) We are essentially capturing the publicly owned property's value over the next 30 years through the value of the implied rental subsidy received by the public housing tenants. The next section addresses this benefit. This discussion is not meant to suggest that the HOPE VI property would not have other value after 30 years. In particular, it may still have value as public or affordable housing, which is more of a political value than an economic one. As this study focuses on the economic costs and benefits associated with HOPE VI, we do not attempt to quantify non-economic values.

¹⁰Because we do not measure gains in the neighborhood rental market, our measure of real estate benefits is likely to be biased downward.

¹¹The concern, of course, is that the reduction of crime in one area simply shifts it elsewhere, therefore generating no net public welfare benefit. Short of tracking crime patterns in each of the neighborhoods of our selected cities and then conducting in-depth qualitative analyses of areas with increases in crime to determine if those increases could be related to the decreases surrounding the public housing site(s), there is no good way of answering the question. We leave such analysis for future research.

¹²Detailed findings from the New Holly analysis are found in Appendix C. The results of the analysis for the other sites are available from the authors upon request.

¹³A more detailed description of the RIMS II model is found in Appendix D.

¹⁴In estimating the impacts on aggregate income, we took into account the number of mortgage applications relative to the number of households. We assumed that the greater the number of loans, the greater the likelihood that the HMDA incomes reflect the broader neighborhood socio-economics. Detailed findings from the New Holly analysis are found in Appendix E. The results of the analysis for the other sites are available from the authors upon request.

¹⁵We do not have data on actual revenues from these areas pre- and post-redevelopment. Our calculations focus on the increases that would have occurred, assuming that both the tax rates and the number of non-HOPE VI related taxable properties remained constant.

¹⁶The ratios are based on data from the 2006 American Community Survey. The basic assumption is that the state allocates revenue among its various counties. Other fiscal impacts that could result from the HOPE VI developments but are not directly addressed in this study would include reductions in the costs associated with general city administration, criminal justice and incarceration, and public service provision.