Association for Public Policy Analysis & Management

As a sizeable near-cash benefit, the Supplemental Nutrition Assistance Program (SNAP; formerly known as food stamps) can have important impacts on household well-being that extend beyond its intended aims to improve food security and nutrition. In particular, SNAP may promote health via impacts on nutritional well-being, and – because it can offset food costs – may free up resources that can spent on health-promoting activities or directly on health care. However, only a handful of studies have examined the effects of SNAP on health while accounting for concerns about selection that might bias estimates. Results of these studies have been inconsistent (Meyerhoefer & Yang, 2011), with some finding SNAP receipt to be related to better self-reported health, fewer sick days, and better birth outcomes (Almond, Hoynes, & Schanzenbach, 2011), but others findings poorer self-reported mental and general health (Bitler, 2015; Heflin & Ziliak, 2008).

This study examines the impact of SNAP receipt and benefit level on the physical and mental health of adults and children using a natural experiment – specifically, the temporary expansion of SNAP benefits from April 2009 to November 2013 provided through the American Recovery and Reinvestment Act (ARRA) – to address issues of selection. We use restricted data from the National Health Interview Survey (NHIS) from 2008 to 2014, restricting our sample to persons in SNAP-eligible and low-income SNAP-ineligible households. Following Nord and Prell (2011), we use differences-in-differences (D-in-D) models comparing changes in child and adult health outcomes between SNAP-eligible households and low-income but SNAP-ineligible households (termed “near-eligibles) across the ARRA benefit expansion and clawback periods. To compare periods immediately prior to and after the clawback, the ARRA SNAP benefit expansion period was defined as January 2013 through October 2013, and the SNAP benefit expiration period was defined as November 2013 through December 2014. SNAP eligibility was defined using household composition, income, employment, and elderly status, and state variation in broad-based categorical eligibility and other eligibility rules. Near-eligibles were defined as those with incomes just above their states’ thresholds (e.g., 130%– 200% FPL).

Controlling for a range of individual and household characteristics, year, month, and state fixed effects, results find that the health outcomes and health care use of SNAP-eligibles and SNAP near-eligibles did not significantly differ over the ARRA expansion or retraction. SNAP eligibles were 1.5 and 1.1 percentage points less likely to report that they delayed health care due to cost for their child, or that their child needed medical care but could not afford it, following the ARRA expansion, but both of these effects were marginally statistically significant. Future analyses will examine different specifications of the SNAP-eligible and near-eligible groups, and additional dependent variables related to behavioral health. Findings will shed light on the causal implications for food assistance participation on adults’ and children’s health in the United States and abroad.