The Effect of Means-Tested Transfers on Work: Evidence from Quasi-Randomly Assigned SNAP Caseworkers

We provide a comprehensive evaluation of the dynamic labor supply effects of the Supplemental Nutrition Assistance Program (SNAP) for a representative population using novel administrative data and an examiner design. We find no effects of SNAP receipt on the full sample of working-aged SNAP applicants. This is because the majority of working-aged applicants do not work before applying and experience no change in work if granted SNAP, consistent with this group facing barriers to work. The minority who work before applying appear to treat SNAP as insurance against negative shocks; they decrease work temporarily but work more in the longer-run.