What Do Welfare-to-Work Demonstrations Reveal to Welfare Refomers?

Under the new welfare system, states must design and institute programs that both provide assistance and encourage work, two objectives that have thus far appeared incompatible. Will states meet these new requirements? For many innovative programs, the randomized welfare-to-work experiments conducted over the last three decades may be the only source of observed data. While these experiments yield information on the outcomes of mandated treatments, the new regime permits states and localities much discretion. Using data from four experiments conducted in the mid-1980s, this study examines what welfare-to-work demonstrations reveal about outcomes when the treatments are heterogenous. In the absence of assumptions, these data allow us to draw only limited inferences about the labor market outcomes of welfare recipients. Combined with prior information, however, data from experimental demonstrations are informative, suggesting either that the long run federal requirements cannot be met or that these standards will only be met under special circumstances