In an NYT online commentary, Tina Rosenberg argues that low-income workers should not be allowed to use SNAP benefits (food stamps) to purchase foods that are not “healthful.” The main argument she makes involves WIC, a federal food program that provides vouchers to purchase certain specified foods that supplement the diets of low-income children under age 5 and certain women who are pregnant, breastfeeding, or postpartum. According to Rosenberg, changes in late 2009 to the WIC food package may explain a small but significant drop in obesity rates the CDC recently observed among low-income preschoolers in 18 states between 2008 and 2011.

I personally support the 2009 WIC food package changes, but I’m quite skeptical that they contributed much to the small declines in obesity in some states seen in the particular CDC study cited by Rosenberg, and don’t think they provide a policy template for SNAP.

On the first point, it’s unclear from the study Rosenberg cites whether there was a significant nationwide decline in obesity among low-income preschoolers between 2008 and 2011. However, there is recent published research that looks at 2003 and 2010, mostly before the implementation of the WIC rules, and finds a slight nationwide decline in obesity among low-income preschool children, after an increase between 1998 and 2003. So, the best evidence we have shows obesity among low-income preschoolers leveling off and slightly declining before the WIC changes would have had any impact. Another limitation with the study Rosenberg cites, which is based on low-income children who participated in federal nutrition programs during the economic downturn, is that the downturn brought a lot ofnew children into these programs, potentially confounding their results, something the authors acknowledge.

On the second point, I don’t think I have any big philosophical objection to very modestly adjusting the list of foods that SNAP benefits cannot be used to buy—beer, wine, liquor, cigarettes, tobacco, vitamins, food that will be eaten in the store, and hot foods—to include carbonated soft drinks with high-fructose corn syrup. But SNAP is a very different program than WIC. When WIC was authorized as a permanent program in 1975, it was specifically designed as a supplement to SNAP. And WIC benefits have always been limited to a short list of specified “supplemental foods,” ones defined as containing nutrients known to be lacking in the diets of young children and pregnant/new mothers. The WIC rules USDA adopted in 2009 are consistent with this—they simply further refine a long-existing supplemental food package, rather than creating one that never existed in the first place, as would be the case in SNAP. Although I worry about the extent of USDA’s capture by Big Food, the carefully documented case the agency has made against adopting a WIC-style restricted food package in SNAP is quite compelling.

Finally, the single biggest problem with the Rosenberg piece (aside from it’s offensive title) is her hasty dismissal of providing financial incentives to encourage the purchase of fruits and vegetables. For example, Rosenberg acknowledges that the Healthy Incentives Program, a SNAP pilot project operating in the lowest income country in Massachusetts being evaluated with a rigorous random assignment methodology, has increased fruit and vegetable consumption among participants by 25 percent compared to the control group. But Rosenberg dismisses it, and similar programs, as not reaching enough people and too expensive.

Given that the average incentive in the HIPS pilot was only $3.64 a month, making the same incentive available to all SNAP households nationwide would cost under $1 billion in additional benefits a year, an amount equal to about 1/80th of total SNAP expenditures, and a blip in the overall federal budget. By comparison, Rosenberg asserts that what she calls her “harsher” strategy would be “virtually free.” Meanwhile, here is what USDA has said about the potential costs and burdens of Rosenberg’s preferred policy: “There are more than 300,000 food products on the market, and an average of 12,000 new products were introduced each year between 1990 and 2000. The task of identifying, evaluating, and tracking the nutritional profile of every food available for purchase would be substantial. The burden of identifying which products met Federal standards would most likely fall on an expanded bureaucracy or on manufacturers and producers asked to certify that their products meet Federal standards.”

Finally, Rosenberg might have noted that there is a federal food program that already uses incentives on a nationwide basis. That program? WIC, which provides mothers who breastfeed with an enhanced food package and allows them to participate in WIC longer than non-breastfeeding mothers.



Post-script on “Obesity” and Well-being

In reading my draft of this post, CEPR Research Assistant Stephan Lefebvre notes that media accounts of  weight trends often come bundled with classist and “fat-shaming” assumptions, and that the body-mass index (BMI) used to define “obesity” does not provide a direct, objective measure of individual health and well-being.

This is an important point that is quite relevant to Rosenberg’s piece, which focuses on solely weight-height trends among low-income children, even though most children and adults above the BMI cutoff used to define obesity are not low-income or income-eligible for SNAP, and the relationship between socio-economic status and obesity among adults is not particularly clear-cut. For example, among non-Hispanic white adult males, there is no significant association between income and obesity presence, while among non-Hispanic blacks and Mexican Americans, low-income adult men are less likely to be obese than higher-income ones.