Evaluation of Los Angeles County GAIN Program, 1997-2000
The Evaluation of Los Angeles Jobs-First GAIN project evaluated the results of changing Los Angeles' Greater Avenues to Independence (GAIN) program, a human resource focused welfare-to-work program into Los Angeles Jobs-First GAIN, a labor force attachment strategy. Some of the most important changes included: (1) strengthening the job club and instructing case managers to make immediate job activities referrals to almost all clients;(2) locating job developers in all GAIN offices; (3) establishing goals and performance-based contracts to reinforce the Work First message within the GAIN program structure; (4) changing program guidelines, instructions and staff training to communicate clearly to clients the central importance of finding employment; (5) informing enrollees that with liberal earnings disregards, even minimum wage jobs are financially advantageous.
The major research hypotheses for the impact evaluation were that: (1) it is possible to convert a large, diverse, mature education-focused welfare-to-work program to a Work First program; (2) the new program would produce earnings and recipiency changes of greater magnitude the original program; (3) the new program would produce earnings increases and welfare reductions with much smaller expenditures per enrollee than its predecessor; (4) having profited by experience, Jobs-First GAIN would produce equal or better results than previously evaluated Work First programs. Several additional hypotheses were provided in the context of TANF implementation: (5) Jobs-First GAIN could effectively operate on a large scale, serving a diverse welfare population; (6) the program would speedily move many recipients into jobs thereby minimizing the number who reach the TANF time limit unemployed; (7) the programs economies of scale would achieve large savings to government budgets.
The evaluation consisted of an implementation/process study, an impact study and a cost-benefit study. The methodology included a rigorous random assignment design. Data collection was primarily from administrative records, with a two-year survey of approximately 250 experimental and control group cases.