Evaluating the multidimensional nature of supported employment

The emergence of supported employment services in the United States has raised expectations concerning the viability of employment in promoting the work productivity, social integration and personal independence of persons with severe disabilities. Although several studies have offered preliminary insights on the efficacy of supported employment's contribution to worker productivity and earnings, little attention has been given to testing the other expected outcomes of such service activities empirically. This paper examines the multidimensional outcomes of supported employment in terms of its effectiveness and efficiency and offers several policy recommendations for improving these outcomes. Three sets of questions are addressed. The first examines factors that separately influence each of the multiple outcomes through ordinary least-squares regression techniques. A second set estimates recursive relationships among the multiple outcomes through the use of two-stage least-squares regression techniques. And a third set explores efficiency relationships between these outcomes and the costs of providing employment-related services. The data used in the study were originally collected in 1991 for a purposive sample of 40 programmes for estimating the economic costs and benefits of supported employment in Minnesota. Preliminary conclusions indicate positive degrees of both efficacy and efficiency in the programmes examined.