In June 1988 the Wisconsin Legislature authorized an independent evaluation of the Work Experience and Job Training (WEJT) Program and Community Work Experience Program (CWEP) initiatives. As required by the Research Design, the University of Wisconsin- Milwaukee Employment & Training Institute is submitting a January 1991 report on the WEJT/CWEP programs. This report should have been provided to the Legislature in January 1990. However, delays on the part of the Department of Health and Social Services (DHSS) and the Department of Industry, Labor and Human Relations (DILHR) in providing necessary data have resulted in a postponement of this report until January 1991.

This report consists of tabulations by county for the entire 1987 and 1988 population on AFDC and the entire population in WEJT and CWEP programs. The tables include a complete history of all reported earnings data for eight quarters beginning with the first quarter of 1988 for all participants. This study includes the entire adult population (caseheads and spouses) on AFDC in 1987 and 1988. It is not a sample. It does not, however, include Milwaukee County because the county's WEJT program was not fully operational until 1989.

The report also includes a detailed description of both the WEJT and CWEP experience in 1987 and 1988, and traces the experience of programs which have expanded from an estimated $7 million in 1987 to an estimated $40 to $50 million in 1989 and 1990. The financial analysis is limited to those expenses reported by counties currently available through DHSS and does not include any state administrative costs, statewide contracts, or any 1989 expenses.

While this report does not make conclusions about the success or failure of the WEJT or CWEP programs, the tabulations do reveal important information on the employment experience of AFDC recipients and of WEJT/CWEP participants. Fourth Quarter 1989 earnings are used as a benchmark of progress throughout this report. However, they are only one measure of program impact and the final evaluation will include other important measures including job retention, average wages, overall economic well being and welfare savings which result from participants finding long-term employment.

• Most 1987 recipients who left AFDC and had earnings in the Fourth Quarter of 1989 do not appear to have been in any welfare employment program in 1987, 1988 or 1989. This is similar to the experience in other states, where much of the AFDC population finds employment regardless of participation in welfare employment programs.
• There is little overall difference in AFDC reduction or earnings between WEJT/CWEP participants and those not in the program. This is consistent with evaluation findings in other states where programs show only modest program impact. Final analysis will need to control for differences between WEJT/CWEP program participants and non-participants.
• To avoid problems of "creaming", the AFDC welfare employment program was designed to focus on the most difficult to serve population, those unlikely to find employment without some intervention. However, most 1987 and 1988 WEJT counties indicated that they served all mandatory recipients. As a result, many recipients were served who likely would have found employment on their own. One indication of this is that the percent of AFDC clients without a high school diploma is the same for program participants and non-participants.
• The move to change the operation of welfare employment programs under WIN (Work Incentive Program) from a statewide Job Service contract to county-controlled programs has resulted in a wide diversity of program goals, administration, target groups and costs. While this diversity has resulted in much experimentation and has increased the level of county participation, some counties have considerable start-up problems as well as only minimal use of required work supplementation and CWEP components. As a result, costs per participant vary widely across counties.
• There is much overlap between welfare employment programs and JTPA (the Job Training Partnership Act), Wisconsin's other large employment program for the economically disadvantaged. In some CWEP counties cooperation with JTPA appears to have resulted in increased use of JTPA to complement welfare employment program efforts. In other counties there were declines which may to some extent have been the result of fallout from competitive bidding for WEJT/CWEP programs.
• The Rock County 1987 WEJT had formal control and experimental groups which allowed comparison between those clients in the traditional WEOP model of job search and limited services, and those clients in the county-based WEJT programs with enhanced services. Examination of the 1987 population shows little difference in outcomes between the two approaches with the control group outperforming the experimental group in AFDC reduction and percent with earnings and off AFDC but not necessarily for average quarterly wages. However, these differences are not statistically significant and require further analysis. The evaluators found a third available comparison group within the 1987 Rock County population consisting of those AFDC clients in neither the experimental or control group. The make-up of this third group is systematically different from the experimentals and controls and includes those exempt from work programs because they have young children (56 percent), those exempt because they are already working (8 percent), as well as those not served by either the WEJT or CWEP programs. However, the outcomes for this third group with no treatment in 1987 are not dramatically different from the control and experimental groups.
• As the WEJT/CWEP program expanded from an estimated $7 million in 1987 to an estimated $50 million JOBS program in 1990, the DHSS administrative staff has remained very small. This understaffing may explain the lack of financial record keeping, participant outcome data and timely progress reports. As a result there was no accounting of expenses by county and administrative entities for Calendar Year 1989 as of January 1991. Furthermore, DHSS delayed instituting a client tracking system until January 1990 which made it impossible to monitor program performance and participants' outcomes during the first three years of WEJT/CWEP. As a result, the Employment and Training Institute had to compile and edit over 60,000 client records, which were then verified by each county to establish how many participants there were in 1987 and 1988 and the types of program activities each participant received. This extremely time-consuming task on the part of the evaluators and the counties would not have been necessary if the proposed client tracking system had been put in place as first scheduled.