U.S. DEPARTMENT OF LABOR
Employment and Training Administration
Washington, D. C. 20210
Sept. 2, 1997
|DIRECTIVE||:||TRAINING AND EMPLOYMENT INFORMATION NOTICE NO. 07-97, Change 1|
|TO||:||ALL STATE JTPA LIAISONS
ALL STATE WORKER ADJUSTMENT LIAISONS
ALL STATE EMPLOYMENT SECURITY ADMINISTRATORS
ALL ONE-STOP CAREER CENTER SYSTEM LEADS
Office of Regional Management
|SUBJECT||:||Revised Design for the Net Impact Evaluation of the Economic Dislocation and Worker Adjustment Assistance (EDWAA) Program|
Purpose. To inform States of changes made to the Employment and Training Administration's plans for a net impact evaluation of EDWAA.
Background. In September 1997, ETA distributed TEIN No. 7-97 which outlined a proposed design for a net impact evaluation of JTPA title III programs for dislocated workers and requested input from States and substate grantees. At the same time, the contractors for the evaluation--Mathematica Policy Research and its subcontractor Social Policy Research Associates--gathered information in a series of visits and discussions with a number of substate grantees. Input was also solicited from Congressional staff and public interest groups. This TEIN describes the changes made to the design as a result of the comments and discussions. Responses to issues raised by respondents to TEIN No. 7-97 are included as Attachment B.
Revisions to the Design. The revised design calls for the random assignment of eligible applicants to one of two treatment groups--one receiving the full array of EDWAA services, including retraining (full EDWAA group) and one receiving only EDWAA basic readjustment services (BRS-only group). Differences in average outcomes between these two groups can be characterized as the impact of offering EDWAA retraining to EDWAA eligibles.
The primary revision to the original design is a change in the definition of the treatment and control groups. The original design called for the random assignment of eligible applicants to the two treatment groups defined above and to a third group, the control group, which would be denied access to any EDWAA services. The new design eliminates the control group. This change means that under the new design every eligible client can be provided some EDWAA services.
The existence of a control group denied any EDWAA services in the original design was viewed as a major problem by respondents to TEIN 7-97. Program officials were averse to denying EDWAA assistance to eligible applicants. The new design offers substate grantees a way to participate in the study and still serve all eligible applicants. Concerns were also raised that denying EDWAA basic readjustment services to a control group would be difficult given the recent development of one-stop centers and the increased use of more integrated services. That is, in many substate areas if EDWAA applicants were denied basic readjustment services, they would still have access to similar services funded through other sources. This implies that in the original design the control group would look similar to the treatment group offered only EDWAA basic readjustment services. The new design removes this distinction by eliminating the original control group.
Although the new design changes the definitions of the treatment and control groups, it does not eliminate the plan to use random assignment. Random assignment is necessary so that the difference between the BRS-only group and the full EDWAA group validly measures the impact of offering retraining. Without random assignment, the credibility of the study would be seriously jeopardized.
ETA believes that the new evaluation design represents a reasonable compromise that balances the need for credible research with the realities of the changing EDWAA service environment and the concerns of State and local officials. The study will provide local programs with the information needed to gauge the impact of EDWAA retraining on participants and to evaluate potentially productive changes in service provision.
Inquiries. Questions and comments should be directed to Daniel Ryan, Office of Policy and Research, Employment and Training Administration, U. S. Department of Labor, 200 Constitution Avenue, N. W., Room N-5637, Washington, D. C. 20210 or by e-mail to firstname.lastname@example.org.