Monday, November 5, 2018
Henrik J. Kleven (Princeton) presents The EITC and the Extensive Margin: A Reappraisal at UC-Berkeley today as part of its Robert D. Burch Center for Tax Policy and Public Finance Seminar Series:
This paper reconsiders the impact of the EITC on the extensive margin of labor supply, combining evidence from all EITC reforms at the federal and state level in the United States. Starting from long-run evidence, I show that the labor supply of women who are eligible and ineligible for the EITC have evolved in a strikingly similar fashion over half a century, except for a unique period in the late 1990s. During this period, single women with children dramatically increased labor supply at the extensive margin, closing the entire gap between women with and without children. These patterns suggest that the only place for finding potential EITC effects is the 1993-expansion of the program. However, studying the 1993-reform is complicated due to the confounding effects of the strong economy, welfare reform and — potentially — changing social norms in the 1990s. Based on an event study analysis of the 1993-reform, it is argued that the data is consistent with no effect of the EITC.
This conclusion is based on two simple tests: (i) controlling for state-level welfare waivers and business cycles can explain the entire extensive margin increase for single mother until 1996, i.e. until the national welfare reform act. (ii) The DiD estimates for low-educated single mothers (eligibles) are the same as placebo DiD estimates for high-educated single mothers (ineligibles), i.e. triple-differences estimates are not significantly different from zero. Stacking all the EITC expansions at the federal and state level to increase precision, I show that one can rule out extensive margin elasticities greater than 0.1. While the changes in the 1990s could be explained in part by welfare reform and the strong economy, the similar patterns for low- and high-educated women suggest that the underlying mechanism may not be exclusively, or even primarily, related to economic incentives. I provide suggestive evidence of changing social norms during this period, and show that state-level variation in attitudes towards welfare recipients and working mothers can explain a sizeable part of the extensive margin increase by single mothers. Taken together, the evidence presented in this paper point to a new narrative regarding the historical changes in female labor supply in America in the 1990s.