When Congress established the Temporary Assistance for Needy Families (TANF) program, it afforded states more flexibility in providing assistance to low-income families with children, but expected states to involve a high percentage of recipients in work or work-related activities. In fact, Congress specified minimum state work participation rates for families receiving TANF for each year from 1997 through 2002, and specified the types of activities in which families must participate in order to count toward the state's rate. Much of the current debate around reauthorization of the TANF legislation is focused on these rates and the establishment of broader participation requirements. State and local programs serve an important role as laboratories for identifying opportunities for innovation and improvement in the welfare arena. Programs that already strive to engage all or most TANF clients in activities can serve as models for others that are seeking to implement broader participation requirements or will have to respond to the new legislation. The proposed paper explores what we can learn about increasing engagement in work activities and work participation rates from seven such programs. This paper is based on findings from a study of work participation and full engagement strategies conducted by Mathematica Policy Research, Inc. (MPR) with funding from the Assistant Secretary for Planning and Evaluation (ASPE), U.S. Department of Health and Human Services.

The programs included in the study represent different service delivery models and have different federal participation rates, but all actively strive to engage all or most TANF clients in work or work-related activities. Researchers conducted in-person site visits to each program in the winter of 2003-2004 and collected administrative data from two programs to gather information on the extent to which programs are succeeding in engaging clients and the program policies and practices that contribute to or impede their success.

Analyses revealed that in most sites, substantially more TANF clients are engaged in program activities that the federal participation rates suggest. Many are engaged in activities that are acceptable at the state or local level but do not count toward the federal rate (such as mental health or substance abuse counseling), are in federally countable activities but for fewer hours than required in the federal rate calculation, or remain in activities (such as job club or job search) for longer periods than count in the federal rate calculation. Mechanisms for achieving high levels of engagement in activities include flexibility in defining the number of hours and types of activities required of clients, availability of a broad range of services to meet the individualized needs of clients, and consistent monitoring and tracking of client participation and performance. Study findings may be used to improve public welfare services provided to diverse and economically disadvantaged populations.