Date of Award

12-1987

Degree Name

Doctor of Philosophy

Department

Public Affairs and Administration

First Advisor

Dr. Peter Kobrak

Second Advisor

Dr. Myron Ross

Third Advisor

Dr. Lyke Thompson

Abstract

The purpose of this study has been to make some contribution toward an understanding of the United States Department of Labor's (USDOL) JTPA Title IIA performance standards models and their policy implications. The major objective of the study has been to construct, evaluate, and document short-run econometric models which would be available for policy analysis and ex ante forecasting. The consequences of job training programs may have important effects on participants and the manpower service delivery system--hence, this study of the relationship between program activities and their consequences.

The State of Michigan has been using USDOL's JTPA Title IIA performance standards models since 1985. These models are utilized as the basis for both awarding annual incentive grants to Service Delivery Areas (SDAs) with good performance and also sanctioning SDAs with poor performance. An evaluation of USDOL's models was thus conducted to assess the validity and objectivity of these models.

The following evaluation criteria were used: (a) high explanatory power, (b) stability, (c) face validity, (d) lack of heteroscedasticity, and (e) lack of multicollinearity. It was found that USDOL's models had the following technical deficiencies: lower predictive power; definitional, measurement, heteroscedasticity and multicollinearity problems; and with biased regression parameters. These deficiencies may result in political and legal liabilities for the State.

Michigan performance standards models were therefore built upon three separate factors affecting SDA performance: (1) client characteristics, (2) program characteristics, and (3) economic characteristics. An econometric technique called ordinary least squares (OLS) was used to develop Michigan performance standards models. SDA was used as the unit of analysis in the development of Michigan performance standards models.

The results of the study indicated that USDOL's JTPA Title IIA performance standards models were invalid measures of Michigan's SDAs' performance. They also led to lower expenditures and higher placement rates of JTPA programs as reported by SDAs. Michigan JTPA Title IIA performance standards models, however, performed well in predicting Michigan SDAs' performance. It was also concluded that the Michigan performance standards models might have practical applicability to other states as well.

Access Setting

Dissertation-Open Access

Download

Share

COinS