Giving Compass' Take:

• Cody Tuttle finds that food stamp bans inflicted on former inmates after their release increases their chances of criminal recidivism. 

• What do formerly incarcerated individuals need in order to successful reenter society? How can funders work to increase access to the necessary support structures? 

• Learn how public health insurance affects recidivism. 

Since the late 1990s, state and federal prisons in America have released over half a million prisoners every year (Council of Economic Advisors (CEA) 2016). Upon release, these offenders face a myriad of obstacles that inhibit a successful transition into a new life as law-abiding citizens. To start, offenders have trouble finding work—survey evidence suggests over half are unemployed even a year after release (Schmitt and Warner 2010). Job searchers with a felony conviction are subject to extra scrutiny in the hiring process. Recent audit studies suggest that a felony conviction cuts probability of being called back by an interviewer in half (Pager, Western, and Sugie 2009). In addition, some occupational licensing rules bar felons from ever entering an occupation (Bushway and Sweeten 2007). Furthermore, offenders do not meet the requirements of the Unemployment Insurance program upon release and are frequently denied public housing by local Public Housing Authorities (CEA 2016). Finally, as a consequence of the 1996 welfare reform, many offenders are now banned from receiving Supplemental Nutrition Assistance Program (SNAP, formerly named food stamps) and Temporary Assistance for Needy Families (TANF) benefits. With this in mind, it may not come as a surprise that half of releasees are back in prison within five years of their release and three-quarters are rearrested within five years (CEA 2016). Recidivism in America may be at least partly the consequence of these barriers to reentry.

In this paper, I focus on one of those barriers, the SNAP ban, and ask how it affects recidivism outcomes, defining recidivism as a return to prison after release. It is particularly critical that we understand the effect of the SNAP ban because it is currently in effect in 27 states, and because survey evidence suggests SNAP is an important resource for offenders post-release (Wolkomir 2018). Approximately 70 percent of the former inmates in the Boston Reentry Study report receiving SNAP benefits even just two months after release (Western et al. 2015). Even more, SNAP benefits are an important component of income for recipients. Based on a representative sample of adult male recipients (not limited to offenders), SNAP benefits make up approximately 20 percent of their reported gross income. Finally, to the extent that SNAP availability has insurance value, it may also affect the decisions of non-recipients.

Specifically, I find that drug traffickers subject to the ban are about 9 percentage points more likely to return to prison after release than drug traffickers who have access to SNAP. An increase of this size is large for drug traffickers in Florida. Among those offenders who commit their trafficking offense in the 240 days before the cutoff date, about 16 percent return to prison at some point post-release. This implies that the SNAP ban increased recidivism among drug traffickers by about 60 percent. However, this estimate is based on the small sample of about 1,000 drug traffickers committing an offense sufficiently close to the cutoff date. Although I am able to reject a null effect of the ban, the estimate is noisy and the confidence interval is large. The 90 percent confidence interval on the main estimate is 1.7 percentage points to 17 percentage points, which implies the SNAP ban increased recidivism among drug traffickers by about 10 percent to 105 percent. Unfortunately, I do not have the statistical power to produce a more precise range of possible effect sizes.

Furthermore, the increase in recidivism is primarily driven by an increase in recidivism for financially motivated crimes (such as property crime and selling drugs). This result has important implications for state SNAP bans and for reentry policy in general. In fact, it is consistent with recent work by Munyo and Rossi (2015) showing that a disproportionate amount of recidivism happens on the first day of release and that first-day recidivism can be almost completely stifled by giving releasees a sufficient monetary stipend. Their work suggests that financial support can ease reentry. I provide further support for this idea by showing that recidivism increases after we decrease financial support to offenders by banning them from SNAP