The Labor Market Effects of Welfare Reform

The recent reform of the federal welfare system was meant to encourage recipients to leave welfare and enter the workforce. If the reform is successful there are likely to be effects felt throughout the low–skilled end of the labor market. As former welfare recipients enter the labor market, they may exert downward pressure on wages or displace employment of others already in the labor market. Since there has been limited changes in eligibility for federal welfare programs from which to draw inferences, the magnitude of these labor market effects are uncertain. This study analyses an earlier welfare reform, the elimination of the General Assistance program in Michigan in October 1991, that may provide useful evidence on the effect of the 1996 Federal reform. General Assistance was a large–scale, state– administered program that provided benefits to people who fell through the cracks in federal anti–poverty programs. In all, about 82,000 able–bodied adults lost benefits. Comparisons with changes in labor market outcomes in other states suggest that the labor–market entry of former GA recipients in Michigan led to a 0.9 to 2.6 percentage point increase in employment among high school drop–outs and a 1.2 to 2.7 percent decline in weekly hours. There is little evidence of a systematic effect on hourly earnings among men; however, earnings among women may have fallen by as much as 5.8 percent. JEL Classification: H53, I3, J2. ∗Ph.D. candidate. I owe a special thanks to David Card for insightful advice and encouragement. I also thank Alan Auerbach, Hilary Hoynes, Warren Hrung, Lawrence Katz, Jonathan Leonard, Daniel McFadden, Enrico Moretti, Jonathan Siegel, Miguel Urquiola, and participants at the UC Berkeley Public Finance Seminar and Labor Lunch, and the NBER Universities Research Conference on “The Impact of Recent Federal and State Reforms in Public Assistance and Social Insurance Programs” for their comments. Jared Bernstein and David Webster of the Economic Policy Institute provided me with the CPS monthly interview data; and Julie Cullen provided the GA caseload data. I am of course responsible for any errors or omissions.