For over 30 years, many neighborhoods across the nation have been receiving financial assistance from the federal government to improve housing and other community facilities in the form of grants from the Community Development Block Grant (CDBG) program.
According to George Galster, Christopher Walker, Christopher Hayes, Patrick Boxall, and Jennifer Johnson, much research has been done on “where and how CDBG funds have been spent, which groups have been the prime beneficiaries, how efficient the plans and their implementation have been, and what political forces lie behind these allocations,” but little has been done to measure the program’s impact. Galster and his coauthors have conducted a study that endeavors to fill this void.1
The CDBG program was established by the Community Development Act of 1974. The program provides for federal funds to be allocated to local governments to assist them in accomplishing various goals, including “arresting the deterioration of property and neighborhood and community facilities, removing conditions detrimental to health and safety, conserving the housing stock, improving community services, promoting income integration and neighborhood diversity through spatial deconcentration of assisted housing and revitalization of deteriorating neighborhoods, and stimulating private investments in areas with population outmigration and stagnating tax bases.”2
Policymakers have come to expect that recipients of federal funds demonstrate the effectiveness of their program expenditures. This is true of the CDBG program as well.
The authors make several key formulations that shape their assessment of the CDBG program. First, they point out that neighborhood improvements resulting from CDBG investments can be both direct (upgrading the housing stock) and indirect (funding a project that would make the neighborhood attractive to private investors). Next, they note that many community development practitioners and scholars maintain that “a critical mass of improvements is needed to trigger changes in the perception of investment prospects in a distressed neighborhood, but that once this critical mass [or threshold] is achieved, the pace of neighborhood improvements accelerates.” Further, the authors expect that regardless of whether the CDBG investments are above or below the threshold, the neighborhood effects of the investments will depend on the general conditions of the city and the preexisting conditions and growth plans in the targeted neighborhoods.3
The authors devised a statistical model to measure the relationship between the annual CDBG expenditures per poor resident (averaged over the 1994–96 period) in neighborhoods (identified by census tracts) across 17 large cities and subsequent changes in several neighborhood indicators from 1994 to 1999.4 The sample cities were chosen to cover all U.S. regions and include the following: Birmingham, Boston, Charlotte, Cleveland, Columbus (Ohio), Denver, Fort Lauderdale, Houston, Indianapolis, Long Beach, Los Angeles, Milwaukee, Oakland, Portland (OR), Providence, Tulsa, and Washington.
The authors adjusted the CDBG spending for the “poor population in the neighborhood because theoretically the impact of a given amount of spending should differ depending on the depth of local needs and regulatory requirements that such spending primarily benefit low- and moderate-income people.” They also conducted some preliminary analysis to arrive at three indicators that would be reasonable proxies for the range of neighborhood outcomes of most interest to scholars and local policymakers. These indicators included “the home purchase mortgage approval rate, the median amount of the home purchase loans originated, and the number of businesses.” The authors used data from several governmental databases, which they supplement with information from CDBG grantees.5
Even after settling on CDBG expenditures per poor resident as the independent variable in their estimating equations, the authors did not find any “statistically significant, positive relationship between spending and changes in [their] neighborhood indicators when [they] analyzed either the full sample of census tracts or only those with non-zero values of CDBG spending.” It wasn’t until the authors focused their analysis on census tracts that had spending above the sample-average CDBG expenditures per census tract that they observed statistically significant results. They computed the annual sample-average CDBG spending to be $86,737 from 1994 to 1996. Thus, the sample-average CDBG expenditures approximated a threshold that had to be surpassed in order for CDBG investments to yield statistically significant impacts.
While the authors are able to establish that a threshold or critical mass of CDBG expenditures existed, they are not able to identify it with any precision. Therefore, they feel confident only in claiming that “below roughly $87,000 in annual average expenditure, significant neighborhood payoffs from CDBG are unlikely to be observed, rather than in stating the precise level above this amount at which sizable effects begin to ensue.”
In further analysis, the authors investigate the influence of neighborhood and city conditions on the capacity of CDBG investments to generate impacts. They chose trends in median sales prices of single-family homes as the best measure of the neighborhood’s trajectory before 1994–96. The authors found that “for all three outcome indicators, CDBG investments yielded the highest per-dollar payoffs (as evidenced by the size of the coefficients) in neighborhoods already experiencing a strong upward trajectory of housing prices.”6
The authors also addressed the issue of spatial targeting of CDBG spending. During the early years of the CDBG program, spending was widely dispersed by grantees. However, an amendment to the program in 1977 urged communities “to define areas for strategic investment … where concentration of public resources would produce a demonstrable difference over a ‘reasonable’ period of time.” However, the federal effort at targeting CDBG funds was abandoned in the early 1980s by pressure at the local level in favor of distributing investments more widely across urban neighborhoods. The authors indicated that their research strongly supports the spatial targeting of CDBG funds so as to reach a critical mass (or threshold) in order to demonstrably improve neighborhood conditions.
Galster and coauthors point out some caveats to their study and suggest areas for further research. They note that their analysis did not include possible neighborhood improvements that are not visible and thus did not spur additional investment. “For example, investments to the underground infrastructure (water and sewer lines, for example) may be critically important to sustaining urban services to a poor neighborhood, but private investors may not see them.” The authors recommend that future research include more comprehensive neighborhood indicators to capture the positive effects of investments that are not readily seen.
There were also deficiencies in the data. The authors found that information on CDBG expenditures was incomplete or missing for nearly all cities. They caution that the procedures they used to allocate some CDBG spending, while reasonable, might have biased the measured effects toward zero. They suggest that a replication of their analysis with a more complete and accurate database would be instructive.7
Finally, the authors urge that future researchers strive to verify the “notion of a threshold and, if possible, to identify more precisely its value and the degree to which it depends on neighborhood and city context.”