Unit of Analysis: Household or Firm Versus Neighborhood or Community
In considering issues of outcome measurement and impact evaluation, we might consider measuring outputs or outcomes at the level of households, firms, or place. In conventional social program measurement or evaluation, data are generally desired at the individual or household level. However, for a variety of reasons, households may be either an infeasible or undesirable unit of analysis when it comes to the evaluation of CDFI activity.
For some CDFI product lines and outcomes, we may want to measure outcomes for households or, in the case of business-oriented activities, for firms. As Hollister (2004) argues, however, identifying appropriate comparison groups can be very challenging. Moreover, obtaining data from members of treatment groups – either firms or individuals – can be daunting. In the case of employment-focused business development, for example, gathering data on workers, and not just firms receiving assistance, can be difficult.
For other product lines and outcomes – particularly those with high degrees of relative spatial density – we may expect neighborhood or geographic effects, perhaps out of accident (the programs just happen to be clustered spatially) or out of design. Some CDFI activities are tied – to a lesser or greater extent – to some notion of place. Although she was not considering CDFIs per se, Ladd (1994) provided a typology for alleviating problems of lower-income people and places. First, people-based strategies focus on helping people but do not pay any attention to the places where they live. Examples might include CDFI minority businessfinancing programs that lend to firms regardless of their location throughout some large metropolitan area or state.
Second, place-based-people strategies focus on places – perhaps certain neighborhoods or types of neighborhoods – to increase the well-being of people living in such areas. But there is no primacy given to changing quality of the place. Place is a tool for helping local residents. Third, place-based strategies aim to improve physical and economic vitality with little attention to the impact on current residents. The condition of residents is not seen as the end so much as a means to improve the geographic community. Differentiating, for example, between a neighborhood that “improves” via incumbent upgrading and one that improves via inmigration and outmigration is not of concern.
Figure 6.2 provides a conceptual map of this strategic space. Particular product lines can be positioned within this triangular space. Some may be more likely employed as part of people-based strategies, while others (e.g., real estate-related programs) are pure place-based to place-based-people strategies. Yet others are quite flexible, strategically: a small business loan might be employed in a targeted, place-based strategy to revitalize a retail strip or might be used to encourage employment over the CDFI's entire service area.
Distinguishing among strategic approaches is critical to developing outcome
Figure 6.2 Strategic Space: Targeting Places, People, or Both
measures. Product lines employed as part of a place-based strategy are better candidates for geographic evaluation, while products used as part of a pure people-based strategy are less appropriate.
Related to spatial targeting are market size, mobility of capital and labor, and capitalization of benefits in residential property values. The size of labor markets, for example, can make targeting of employment effects in neighborhoods problematic. Firms can hire from broad distances, and there is some evidence that firms in minority neighborhoods tend not to draw their workforces from near their facilities as much as do firms in other locations (Ihlanfeldt 1999). The geographic scale of labor markets means benefits of employment-oriented CDFI interventions will be more spatially diffuse as compared to housing market investments.
Moreover, because business lending or investment frequently focuses on increasing a firm's stock of relatively mobile capital (equipment, inventory) or human resources, less goes into permanent improvements in spatially fixed real estate. Housing investment is more spatially fixed. Of course, some of the benefits of housing investment may diffuse spatially, especially over time, as beneficiaries move on to different residential locations, perhaps partly as a result of earlier CDFI interventions – for example, as new homeowners become established homeowners and move into a larger house. Overall, however, a larger portion of the benefits from housing investment is likely to accrue locally as compared to investments in job creation or retention.
When considering measuring effects spatially versus individually, it is important to keep in mind the general phenomenon of the capitalization of neighborhood amenities into residential property values. As various aspects of the quality of life of a neighborhood improve, residential values are expected to rise. In essence, the future value of those improvements are at least partially capitalized into property values. If many residents fix up their homes, for example, we expect to see an increase in property values (Ding, Simons, and Baku 2000).
Of course, property values should be used with some caution as a measure of neighborhood quality of life. If values appreciate very quickly, displacement of existing residents or speculative price bubbles are a concern. Moreover, many nonmarket neighborhood qualities, including various forms of social capital, may not be fully capitalized into property values.
Notwithstanding these limitations, moderate levels of appreciation – and avoidance of falling values – are seen as a positive neighborhood indicator. It is important to note that this signal of neighborhood quality of life is not relevant only to homeowners. Because property values reflect improvements in the quality of life for residents, it is expected that the quality of life for renters would also improve, assuming that increases in rents or taxes are not excessive.