Association for Public Policy Analysis & Management

Recent popular and policy interest has focused on the idea of a Universal Basic Income that would provide an income floor to all members of a society. Such policies differ from traditional transfer programs by providing benefits to adults who are not “tagged” with characteristics indicating greater standards of need (such as caring for children or having a disability). However, as most traditional transfers do not benefit such adults, little is known about the effects of a basic income on the population of able-bodied adults without dependents. I exploit a novel source of variation in one of the few near-universal transfer programs in the United States, the Supplemental Nutrition Assistance Program (SNAP, also known as Food Stamps), to examine the effects of an income supplement on childless adults, thus helping to inform the debate about basic income and related policies.

The SNAP provides benefits of just under $200 per month to households with a single able-bodied adult without dependents (known as an ABAWD); such individuals make up about 10% of program recipients. This paper exploits the fact that SNAP benefits are available to ABAWDs differentially based on their local area unemployment rate. Normally, ABAWDs must work 20 hours per week to receive SNAP benefits for more than 3 months every 3 years; however, in areas with high unemployment, states can apply for waivers so that ABAWDs who do not work can receive continuous benefits from the program. The cutoff for an area with insufficient jobs is a complicated formula that compares state and county unemployment rates to a specified target based on national unemployment. Thus, in a year when the national cutoff is 6%, an ABAWD living in a county with a 6% unemployment rate will be eligible to receive benefits without working, but a comparable ABAWD living in a county with a 5.9% unemployment rate will not be eligible. This motivates my investigation of the effects of SNAP on ABAWDs through a regression discontinuity design, comparing counties just above and below the unemployment cutoff. This source of variation has not been previously used in the literature, and provides a rare source of local variation in the SNAP program (where most policies are set at the federal level), providing more convincing identification of the program’s effects.

My research has gathered historical data on unemployment and SNAP participation to track the counterfactual availability of ABAWD waivers over time. In preliminary results using the American Community Survey, I find that the work requirement in SNAP has a perverse effect; employment (of 20 or more hours per week) is actually higher for ABAWDs in areas just over the threshold that do not condition SNAP on work. Future research will examine the effects of the program on participants’ health, housing, food security, and other outcomes, providing new evidence on the value of transfer programs for childless adults.