«A Journal of Policy Development and Research HoPe VI Volume 12, Number 1 • 2010 U.S. Department of Housing and Urban Development Office of Policy ...»
Crime Effect Estimation Methods As with the property value analysis, we are ultimately interested in the difference between the observed crime rates around the HOPE VI redevelopments and the rates that would have occurred in the absence of HOPE VI investment. In both cities, we estimated the counterfactual by assuming that violent crime rates in the HOPE VI areas would have tracked those in other traditional public housing neighborhoods had the redevelopments not taken place. We therefore calculated crime rates for traditional public housing areas before and after demolition of the HOPE VI sites and applied the post-demolition rate in those areas to the pre-development rate at the HOPE VI property to establish the counterfactual.
Estimated Crime Effects Exhibit 6 displays the actual numbers of Part I crimes committed near traditional public housing sites and at HOPE VI sites in Boston and Washington in the years before the HOPE VI interventions and also shows the actual numbers at the sites in the years following re-occupancy. It compares the actual figures with the estimated numbers of violent crimes that would have taken place at the HOPE VI sites in the absence of redevelopment.
Crime rates in three of the four HOPE VI sites declined. In Boston, violent crime in the Mission Main and Orchard Gardens neighborhoods declined by 51.5 and 66.2 percent, respectively. In Washington, violent crime in the Townhomes area fell by 75 percent. In each of these cases, the decline in the HOPE VI neighborhoods far exceeded the drop in violent crime in the neighborhoods with traditional public housing sites. Our interrupted time series model indicates that for these three areas a decline of at least 48 percent occurred in violent crime relative to the estimated crimes that would have occurred in the absence of the HOPE VI redevelopments. Only in the Wheeler Creek area did we observe higher crime rates after redevelopment than we would have expected in its absence. Although the number of crimes in the Wheeler area after redevelopment was lower than the number before, the decline was not as great as that for other public housing sites in the District of Columbia during the study period.
It is quite possible that larger declines in crimes occurred than we have estimated in the areas surrounding both Washington developments. Whereas the average pre-HOPE VI crime figures for Boston included figures from 1993 to the year of demolition, crime data for the Washington sites Unfortunately, these data do not contain a breakdown of the types of crimes committed except at the aggregate level (all public housing sites); we therefore estimated the types of crime committed at the individual site level based on the proportion of those committed at the aggregate level.
were available only back to 1997. We therefore averaged the number of crimes in the Townhomes and Wheeler Creek areas from 1997 to 1999 and from 1997 to 2000, respectively. These times coincided with the HOPE VI construction periods when few, if any, residents were on site. We would therefore expect the averages for these periods to be lower at the HOPE VI sites than at traditional public housing sites with large numbers of residents.
To what extent did the observed declines in Part I crimes around the selected HOPE VI sites simply reflect a displacement of crime from these areas to other parts of the cities? BHA Chief of Police Steve Melia recalled that some of the raids designed to clear what he called the “bad elements” out of the old Mission Main and Orchard Park projects led to complaints of vandalism and property damage from residents along different transit lines spreading out from the public housing sites. Yet the number of such complaints decreased with subsequent raids, and the police’s efforts received widespread public plaudits. DCHA officials could not remember any similarly localized complaints associated with the Ellen Wilson Dwellings or Skytower-Valley Green projects.
Short of comprehensive, indepth analyses of police reports, arrest records, and crime patterns over time in communities throughout both the Boston and Washington areas, no good way has been found either to prove or to disprove the crime displacement hypothesis. Such analyses were beyond the scope of our study, but we welcome any followup examinations of these issues. The lack of ongoing complaints from particular Boston or Washington communities charging that such HOPE VI-related crime displacement had occurred, coupled with the general decline in crime in the HOPE VI areas and the two cities overall during our study period, leads us to infer that any such displacement was relatively minimal.
HOPE VI Effects on Neighborhood Income Neighborhood Income Data Data on resident incomes—particularly at the neighborhood level—are somewhat limited. As a proxy, we have used information on the incomes of individual mortgage applicants in the two cities. The federal Home Mortgage Disclosure Act (HMDA) requires such data to be reported, along with information on the date of the application, the race and ethnicity of the applicant, and the census tract of the subject property. Our analysis focused on applicants hoping to finance properties in the census tracts corresponding to the ZIP Codes for which we obtained property sales data.
From 1992 through 2006, there were 397,222 mortgage applications in the sampled Boston tracts.
Of these applications, 84,280 were for properties in census tracts containing traditional public housing, and 686 were for homes in the census tract containing the Orchard Garden HOPE VI site.
The limited number of owner-occupied homes in the census tract containing the Mission Main site resulted in very few mortgage applications from that area; we therefore have not attempted to evaluate income changes near that site. During the same period, homebuyers in the selected Washington tracts received 353,309 mortgages—including 55,922 for properties in neighborhoods with traditional public housing and 5,851 for homes in the tracts containing the Townhomes on Capitol Hill and Wheeler Creek Estates. In each of the sampled HOPE VI tracts, the aggregate number of mortgage applications in the study period equaled nearly three-fifths of the number of households
living in the area. Thus the changes in applicants’ average incomes are reasonably good proxies for changes in the neighborhoods’ average incomes overall.12 Neighborhood Income Effects Estimation Methods Our statistical analysis of HOPE VI impacts on neighborhood income closely mirrors that of our property value analysis, in that we have used interrupted time series methods to estimate trends in resident incomes in neighborhoods with and without HOPE VI redevelopments. The differences in income levels and trends before and after the HOPE VI investment represent the conceivable influence of HOPE VI.
Estimated Neighborhood Income Effects Exhibit 7 displays the key parameters of the neighborhood income models.13 The estimated effects on income levels and trends are similar for the Boston and Washington sites, with both cities reporting significantly more favorable trends in income levels after redevelopment. The level coefficient was significantly higher for the period before redevelopment, but the trend coefficient was significantly lower. Thus the average mortgage applicant’s income was declining before the HOPE VI intervention but rising afterward. As the trend increases over time, the positive effects on incomes could increase significantly.
Exhibits 8, 9, and 10 display the implied path of income change for the neighborhoods encompassing Orchard Gardens, the Townhomes on Capitol Hill, and Wheeler Creek Estates. We compared the implied path with the estimated income path in the absence of the HOPE VI redevelopments.
Because the trend in the relative mortgage applicant’s income was very negative before the HOPE VI
In estimating the effects 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 reflected the broader socioeconomic character of the neighborhood.
The dependent variable is the log of applicant income/1,000. Appendix B shows full regression estimates.
Dollar Amount (in thousands) investments, the net effect on the applicant’s income was very large following redevelopment. As of 2006, average annual incomes were estimated to be about $40,000 higher in the Orchard Gardens area, $48,000 higher in the Townhomes area, and about $25,000 higher in the Wheeler Creek area than they would have been in the absence of the HOPE VI redevelopments.
The analyses of crime and neighborhood incomes provide evidence that is generally consistent with the finding that the HOPE VI redevelopments resulted in increased residential property values in three of the four sampled neighborhoods. These findings suggest that gentrification has been a factor in these communities, regardless of whether it was an outcome of the HOPE VI redevelopment. One criticism that has been leveled against HOPE VI is that its emphasis on reducing the concentration of poverty to create more mixed-income communities ultimately promotes gentrification and the displacement of poor people. Our study explicitly did not address those issues. It could easily be argued, at least in the selected Washington communities, that the HOPE VI investments actually resulted in a net increase of housing available to very low-income people, given that the HOPE VI properties were largely vacant before redevelopment. Even if HOPE VI did contribute to gentrification and displacement in the selected Boston communities, our qualitative analysis identified multiple factors independent of HOPE VI that contributed to the communities’ economic improvement.
Cityscape 119Zielenbach and Voith
Interpreting the Findings It is important to keep in mind that the HOPE VI redevelopments did not take place in a vacuum;
numerous other factors affected the surrounding neighborhoods before, during, and after the redevelopment process. This section adds a qualitative analysis to the quantitative findings in order to understand the relative importance of the HOPE VI interventions in catalyzing neighborhood change.
Mission Main The neighborhood encompassing Mission Main experienced dramatic increases in home sale prices, reductions in crime, and increases in average resident incomes almost immediately after the demolition of the distressed public housing property. The improvement is reflected by the fact that a waiting list exists for the market-rate apartments at Mission Main.14 To what extent would this improvement have taken place without the overhaul of Mission Main?
Pam McKinney, a local real estate appraiser, contends that Mission Main’s redevelopment has helped unlock the potential of the area. She thinks that Northeastern and Wentworth Universities would not have expanded across Huntington Avenue had Mission Main not been redeveloped, and that the HOPE VI project made it possible to think of the site as an extension of the Fenway and Longwood neighborhoods to the north and west. BHA officials disagree somewhat with her assessment, believing that the hospitals, universities, and medical centers in the area would have ultimately expanded anyway. Esther Schlorholtz, who worked for the city government before becoming senior vice president for Community Reinvestment at Boston Private Bank, emphasizes that, although the redevelopment of Mission Main was important to the community, it was one of multiple local revitalization efforts at the time. She notes that other subsidized housing properties in the area had expiring affordability restrictions and that developers were rehabilitating a number of them.
Orchard Gardens The Dudley Square area experienced significant increases in property values and reductions in crime after the demolition of Orchard Park and the development of Orchard Gardens. These positive changes were less substantial than those in the Mission Main area, however. Dudley Square also saw improvements in its commercial conditions. By 2007, local retail vacancies had fallen to less than 3 percent. New construction and major renovation projects were under way throughout the community. Commercial rents had increased from about $15 per square foot to about $24.50.
Nightlife had started to return to the area, as evidenced by increased use of the ballroom at Hibernian Hall, a longstanding local landmark completely renovated by Madison Park Community Development Corporation (CDC), and by the creation of a 10,000-square-foot establishment that included a restaurant, sports bar, and jazz club. Although Dudley Square’s economic and social conditions are far better than they were a decade ago, the community’s real estate market remains weak. Many of the area’s commercial tenants are social service agencies. Dudley Square has not Although conditions are much better, Mission Main still has problems. A gang is based at the property, and the BHA is concerned about onsite security.
yet attracted a critical mass of middle-income tenants, in part because Section 8 voucher holders occupy most of the “market-rate rental” apartments in the area.
The importance of the Orchard Gardens redevelopment to Dudley Square’s improvement is somewhat unclear. Although the neighborhood does not have the large institutional presence of the Mission Main area, it has been a major focus of city revitalization efforts. As part of his initial mayoral campaign in 1993, Thomas Menino promised to restore Dudley Square as a major commercial area, and the city of Boston injected considerable public resources into the area in the mid-1990s. The area was designated as both a Boston Main Street district and a federal Enhanced Enterprise Community, and it was a centerpiece of the city’s Empowerment Zone application.