This paper analyzes the extent to which labor supply adjusts to incentives created by social programs. We find new evidence of highly elastic labor supply for single mothers in the United States, with sizable responses to the Earned Income Tax Credit (EITC) and welfare (AFDC/TANF) reforms during the 1990s. We reconcile some conflicting results in the literature by showing how the difference in differences design fails to identify a meaningful treatment parameter when a reform expands a pre-existing social program and when multiple programs change simultaneously. Finally, we use our quasi-experimental estimates to identify a structural model of labor supply with multiple tax and transfer programs. Model counterfactuals show that the effect of the EITC on labor supply depends on the regime of taxes and transfers in place. We conclude that evidence-based policymaking must explicitly model the tax and transfer system when using past reforms (ex post analysis) to draw inference about the effects of future reforms (ex ante analysis)

on the labor market.