Food stamp cycle in the Washington Post

The Washington Post this weekend published a long and thoughtful feature by Eli Saslow titled: Food Stamps put Rhode Island town on monthly boom-and-bust cycle.

At precisely one second after midnight, on March 1, Woonsocket would experience its monthly financial windfall — nearly $2 million from the Supplemental Nutrition Assistance Program (SNAP), formerly known as food stamps. Federal money would be electronically transferred to the broke residents of a nearly bankrupt town, where it would flow first into grocery stores and then on to food companies, employees and banks, beginning the monthly cycle that has helped Woonsocket survive.

For research background, my dissertation in the late 1990s and the resulting article with Christine Ranney in the American Journal of Agricultural Economics provided the first nationally representative estimates of the monthly cycle in food spending and food use for food stamp program participants.  My 2002 article with Margaret Andrews in the Journal of Consumer Affairs used Electronic Benefit Transfer (EBT) data to get an even sharper view of the exact timing of food stamp and cash transactions over the course of the month.  In 2005, Jesse Shapiro provided more economic insight in the Journal of Public Economics.  In a 2012 article in the American Journal of Clinical Nutrition, Harvard researchers Cindy Leung, Walter Willett, and Eric Ding recommended further research on whether the monthly SNAP cycle could be related to risk of overweight.

One policy-relevant question is whether the federal government should ever consider providing benefits twice monthly.  Michigan considered such a policy in 2008, but I think it never came to pass.

One objection I have heard is that twice-monthly benefit delivery is paternalistic and might force low-income SNAP to conduct a potentially burdensome shopping trip.  This argument seems incorrect to me.  Let me lay out the case in an orderly fashion.

Whether benefits are delivered once or twice monthly, SNAP participants can freely choose whether to shop once or twice monthly.

• If benefits are delivered to the EBT card once monthly, then a SNAP participant who wanted to shop once monthly is perfectly satisfied, but a SNAP participant who wanted to shop twice monthly and smooth consumption over the course of the month must save half the benefits for use two weeks later.
• If benefits are delivered to the EBT card twice monthly, then a SNAP participant who wanted to shop once monthly must save half the benefits for use two weeks later, but a SNAP participant who wanted to shop twice monthly and smooth consumption over the course of the month is perfectly satisfied.

The key point is that the two policies are exactly equally paternalistic.  When middle-income speakers say that twice monthly benefit delivery is paternalistic, they implicitly assume that it is most natural for low-income people to shop once monthly and go hungry later -- a shopping pattern that the middle-income speakers would never use themselves.  I think it is the middle-income speaker's implicit assumption that really is paternalistic.

Any policy such as twice-monthly benefit delivery should be pilot tested.  The pilot should explicitly ask SNAP participants whether they had any shopping difficulties or budget difficulties, and whether they liked the new policy better than the old one.  It is possible that the new policy will reduce the occurrence of episodes of food insecurity at the end of the month.  The current policy should not be preferred based on implicit assumptions. SNAP participants deserve an influential voice in this decision.