Abstract
We study how eligibility expansions in the US Supplemental Nutrition Assistance Program (SNAP) affect take-up among previously eligible households and social welfare. Using an event study with administrative take-up data, we document a "woodwork effect": Expanding eligibility raises take-up among previously eligible households. An online experiment and administrative survey provide evidence that information frictions, rather than stigma, drive SNAP's woodwork effect. To interpret our findings, we develop a model of welfare programs with incomplete take-up and woodwork effects. Given our empirical results and certain modeling assumptions, expanding SNAP eligibility would raise social welfare.