Log in / Register
Home arrow Business & Finance arrow Reengineering community development for the 21st century
< Prev   CONTENTS   Next >

Conclusions and Implications for Policy

This chapter probes the possibilities for new research on impacts among CDFI subsectors and product lines. It focuses on research designs that might be used to measure impact, especially in the context of moving beyond evaluations of single CDFIs. I recognize the many formidable challenges involved in measuring impact, either through experimental or quasi-experimental methods. Moreover, I have argued that in many situations, well-implemented PM, which does not attempt to identify a convincing counterfactual, is preferable to impact evaluation, which is more ambitious and more vulnerable to error.

Recognizing that philanthropic and government funders and investors are under increasing pressure to allocate resources based on evidence of impact, this chapter attempts to move beyond the current state of affairs by proposing some potential research designs and methods that may provide us with more evidence than we currently have about CDFI impacts. Unfortunately, in some product-line areas, there appear to be very few prospects for feasible research in the near- to midterm. In some other areas, however, especially single-family housing lending in which spatial, neighborhood-level impacts are a substantial focus, there may be feasible methods for reasonably accurate impact evaluation.

Granted, these techniques are fairly sophisticated and have sizable data requirements. At the same time, data required are available in many places, sometimes at a sizable but usually not insurmountable cost. Moreover, in some places, county governments may provide the necessary data at a much lower cost, particularly to nonprofit or public-sector users.

Some may come away, understandably, with pessimism regarding the ability to demonstrate CDFI impacts. However, the intent is not to discourage but rather to provide a sober view of the challenges involved. Just as it takes resources to develop the CDFI field, it will take a significant investment to develop the data and methods needed to measure impacts. In my view, it makes sense to pursue the more feasible methods in the near term while exploring the possibility for improved data systems and availability for those areas and product lines where such data do not exist.

In considering the potential for research on CDFI impacts, one certainly needs a long-term view. There is a small but fairly steady stream of innovative research being conducted in the community development arena – some of it described in this chapter – that should continue to evolve. More applications of such methods could be applied to the CDFI sector. There is also a potential for more applications of true experiments, including geographically based experiments. Finally, improvements in data, especially in small-area data, will hopefully continue to develop and expand the tools available to conduct reasonable evaluative research.


I would like to thank Valerie Chang, Deborah Schwartz, and Michael Stegman, of the Mac Arthur Foundation, for commissioning this work and for providing feedback; Julia Sass Rubin for exchanges we had about topics in this chapter; Carla Dickstein, George Galster, and Robinson Hollister for commenting on earlier work; and practitioners and scholars who met at the Mac Arthur Foundation in March 2006 to discuss issues addressed here. I remain responsible for any errors, omissions, and opinions herein.

Found a mistake? Please highlight the word and press Shift + Enter  
< Prev   CONTENTS   Next >
Business & Finance
Computer Science
Language & Literature
Political science