There is a well-established link between poverty and incarceration. For individuals who are incarcerated, their incomes prior to arrest are about 41% lower than those who are not incarcerated. There is also evidence that shoplifting at grocery stores increases in times of high unemployment—most recently during the COVID-19 pandemic. This begs the question: does providing greater food assistance reduce crime?
In my study, I use the Supplemental Nutrition Assistance Program (SNAP) to better understand how providing food assistance impacts a person’s criminal activity. Every month, participants on SNAP receive benefits on a debit card that they can spend at grocery stores and convenience stores. In order to receive SNAP benefits for an extended period of time, able-bodied adults without dependents (ABAWD) must work or participate in job training for at least 20 hours per week. However, states can increase SNAP access to ABAWD by getting a waiver to the work requirement policy—which they generally do in times of high unemployment.
In response to the Great Recession, Congress allowed all states to get a work requirement waiver, and in 2010, every police agency in my sample was in a region with a waiver for the majority of the year. Over the course of the next seven years, some states became ineligible for the waiver or decided to voluntarily reimpose work requirements, which allows me to compare arrests for police agencies that were in a region with a waiver (meaning SNAP access for able-bodied adults was greater) with police agencies that were in a region without a waiver.
I ran my analysis on different types of crimes reported in the FBI’s Uniform Crime Reports database. Specifically, I expected to find an impact of the waiver policy on property crime arrests, since they are generally more financially motivated than violent crimes. I analyzed the age group of people ages 18 to 49, since this is the group that must meet the ABAWD work requirement to receive SNAP benefits. About 45% of the ABAWD population on SNAP is female, so the work requirement affects males and females in about equal proportions, but I analyzed the data on men and women separately to see if there were differences in the effect based on gender.
For men, I did not find a significant change in arrests for any of the four crimes I studied. The estimate for property crime is that a police agency that was in a county with a waiver in place had about 4% fewer property crime arrests, but—as Figure 1 demonstrates—an effect of 0% is also within the margin of error. This is also true for robbery, burglary, and aggravated assault.
When it comes to the effect of SNAP on arrests for females, I did find a significant effect for two of the four crimes. For burglary and robbery, the results do not differ from those of the males. This is not very surprising since burglary and robbery are more serious crimes than theft and are much less frequent, especially among women. Then, when I look at theft, I do find that the waiver policy is estimated to significantly reduce arrests for women by about 7%, as seen in Figure 2. On the other hand, I find a somewhat surprising result for aggravated assault arrests—they are estimated to have increased by about 7% in response to the waiver policy.
It is difficult to explain why SNAP might increase assault. Another possibility is that my model is biased or prone to picking up false positives, when there really is no effect. I hope to continue working on this project to get a better understanding of the results.
In terms of absolute number of arrests, the effect on female theft is larger, since it is more common than assault. If result of theft reductions hold upon further analysis, then this would be an important finding for policymakers as they consider reimposing work requirements as the economy recovers from the COVID-19 pandemic. Since vulnerable populations face higher rates of unemployment, it is important that policymakers do not make any “snap” decisions to reimpose work requirements for safety net programs like SNAP.
That being said, the results also make it clear that simply providing food or income assistance will not single-handedly solve the overall problem of high incarceration in the U.S. That will take a deeper commitment to providing long-term economic opportunities for America’s most vulnerable populations.
Jonathan Ralstin is a senior at Indiana University.
Leave a Reply