Evaluating the Impact of Federal Human Service Programs with Locally Differentiated Constituencies

Publisher Website
Google Scholar
Abstract
Evaluations of large-scale federal programs require innovative methodological approaches for several reasons: (1) funds are often allocated by jurisdiction as opposed to needs; (2) local interpretation of program goals may vary widely; (3) target populations are often locally differentiated; (4) administrative organization and strategies for service delivery vary at the local level. A study of federally subsidized small-farm programs revealed a broad diversification of strategies and a range of client attributes both within and between local target populations, suggesting the need for an analytical model which would address program variation. Typical-client profiles were developed based on expectations that a given strategy would have differential effects dependent on client characteristics. Regression analysis was then used to assess the effect of program attributes on outcomes for each client group. Our results indicate that the strength of the relationship between program strategies and outcomes varies by group. The methodological approach used here should have broad application for the evaluation of large-scale human service programs.

This publication has 2 references indexed in Scilit: