Implementation experiences will be the focus of several stages of the evaluation. During the first half of 1999, extensive contacts will be made with local WtW grantees to document the local context of their programs, the specific interventions they are attempting, their success in recruiting, and the factors that have shaped both their plans and their success in implementing them. The earliest of these contacts (some already made) are part of the process of recruiting sites for the indepth study component of the evaluation. By summer 1999, intensive site visits will be under way to selected sites. A process analysis report based on this field investigation will be prepared by fall 1999. Implementation experiences, services provided, and job placement outcomes for all WtW grantees will be examined again in a grantee survey in fall 1999 and reported on in early 2000. Comparison of that second survey and the first survey reported on here will provide a systematic measure of implementation progress.
The question of ultimate concern is which program approaches work best. Efforts are now under way to recruit grantees who will participate in the evaluation impact analysis. In an estimated 10 sites, random assignment of program applicants or referrals will be used to create a program group and a control group. This randomassignment process will begin in some sites as early as spring 1999 and in others as late as fall 1999; it will continue in each site for at least 12 months. Comparisons of employment outcomes for the program and control groups in each site (based on TANF and wage records and followup surveys) will be used to determine whether the special WtW program services helped participants achieve a level of employment success beyond that achieved by individuals who had access only to the standard services that would have been vailable in the absence of the local WtW programs that are the focus of the impact analysis. Such impact findings will be made available based on followup of the program group and control group over a minimum of 18 months and should be available in fall 2000.
1. Other factors are also contributing to slow enrollment. Some grantees note that, when they receive a list of longterm TANF recipients who are supposed to be potential WtW participants and attempt to contact them, many have already found jobs and left the TANF rolls. Others turn out to be exempt from TANF work requirements and thus may not face immediate pressure to participate, although they will still be affected by time limits and are eligible to receive WtW services.
2. The President's reauthorization proposal, if passed, would address some of these issues. The proposal retains the program focus on those most in need, while simplifying the eligibility criteria to avoid excluding individuals who are truly among the hardest to employ.
3. The survey question instructed respondents to answer "low" if "services are not yet being delivered or have just begun," to discourage grantees from answering based on their plans rather than on actual experience.