Results of the Policy Simulation
Conducted on Participation Rates
As noted in Section 2, changes to the benefit structure affect employment among current recipients and participation rates in two different ways. First, there is a behavioral effect: Because a recipient would take home more of her earnings, she works more. Second, there is a mechanical effect. A higher earned-income disregard and a lower BRR imply that a recipient remains eligible for cash assistance at a higher level of total earnings. Under the benefit structure in California, a woman with two children must work full time at about $8.75 an hour to be income-ineligible for CalWORKs. In many other states, full-time or even half-time work at the minimum wage makes a family income-ineligible for cash assistance. Thus, recipients who in another state would be income-ineligible remain eligible for, and often remain on, welfare in California.
Using the Q5 data, we can simulate the magnitude of the effect of benefit structure on the participation rate. Our analysis focuses on those working sufficient hours to be income-ineligible in other states (half-time to full-time at the federal minimum wage, $5.15 per hour). Because they are working more than 25 hours per week, they appear in both the numerator and the denominator of the federal participation rate in California, raising the rate over what it would be in a state with a lower level of maximum earnings. We tabulate the fraction of California welfare cases with earnings high enough to make them incomeineligible in other states. Assuming that every such person was participating according to the federal definition, we can compute the effect on the participation rate.
Table D.1 presents the results of such a simulation. It gives the percentage drop in the caseload that would occur if California adopted the benefit structure of each state (the benefit structures are described in Choesni et al., 2000), as well as the effect on the participation rate.1
This simple simulation ignores the behavioral response to the benefit structure. We can use conventional labor-supply models and estimates to make a rough____________________