From a small pilot program created by John F. Kennedy’s first executive order in 1961, the food stamp program (now called the Supplemental Nutrition Assistance Program, or SNAP) has grown to be the largest federal food assistance program. However, there have always been major questions about whether the hunger and nutrition goals of the program justify its design.
A good way to see the logic of SNAP supporters is to ask, “Why give poor people food stamps rather than money?” Some fear that low income recipients may “waste” the money for some other goods than food, and want to preclude that possibility. Agricultural and grocery interests piggy-back on the belief that food stamps will increase food demand more than giving recipients money, which promises them higher sales and profits. Politicians do the same, because they can assert greater effectiveness at reducing hunger with the funds.
Unfortunately, a major problem with that shared understanding is that money is fungible. A person can redirect money put to ones use to any other use as he or she chooses.
Consider a numerical example. The average family SNAP benefit was recently reported as $256. But the vast majority of eligible families already spend more than that on food. Say that a family already spending $400 per month on food received $256 in SNAP benefits. Would it all go to food purchases? That would be very unlikely. In fact, it is possible that no SNAP dollars would be used to expand food purchases. A recipient who wanted to spend more on anything other than food, from alcohol to child care, could simply replace $256 in cash food purchases they are already making with SNAP resources, freeing up $256 in cash for whatever purpose the recipient wishes, just as if they were given cash directly.
Consequently, for the vast number of recipients, SNAP accomplishes the same thing as a cash program, except for requiring substantially higher administration costs. But that means the primary arguments for food stamps rest on a major analytical error.
That error is masked by the fact that SNAP benefits increase the effective incomes of recipients (by $256 a month in the example above). Since estimates of consumer behavior suggest that an increase of $1 of income would tend to increase food purchases by about 20 cents, the likely effect of an effective $256 increase in income would be just over $51 in added food purchases, with the rest going to other uses. However, SNAP supporters can then claim success in increasing low-income families’ food purchases, when the exact same result would have occurred with cash assistance.
There are some low-income recipients, however, who are forced to buy more food than they would choose by the food stamp program (their SNAP allotment exceeds what they would have spent on food, even if they received the benefits as cash). These are disproportionately low-low-income families, because, with more limited incomes, they would buy less food, and they would get larger SNAP allotments. But while they can be pointed to as a greater success in inducing more food purchases, we must realize that to the extent food stamps increase food consumption compared to giving recipients money, they do so at the expense of other goods recipients judge to be even more important. And given the tight constraints on their consumption of almost everything, it is unsurprising that there is little evidence that poor families waste large amounts of income.
Despite SNAP’s $70-some billion annual price tag, backed by logically confused support and misleading evidence of increased food purchases, what does it do about hunger and nutrition, its main targets? After all, as Senator Hubert Humphrey claimed in a hearing over four decades ago, “The food stamp program plays a very critical role in enabling millions of low-income families to have a better diet.” Unfortunately, a 2010 GAO report concluded that, “the literature is inconclusive regarding whether SNAP alleviates hunger and nutrition in low-income households.” In other words, we get very little nutrition bang for a great many bucks.
Nutrition and hunger advocates have noticed SNAP’s limited effectiveness in those dimensions and advocated changes in the system. However, despite their understanding that SNAP achieves far less than proponents claim, their suggested “reforms” also fail to incorporate a recognition of the fungibility of money.
This was made clear in a January 14 New York Times article by Anahad O’Connor, titled, “In the Shopping Cart of a Food Stamp Household: Lots of Soda.” It was triggered by a November 2016 USDA report comparing the types of food bought by SNAP participants with those of non-participants. Attention was focused on three findings: for the SNAP recipients, soft drinks accounted for 5% of food purchases, “sweetened beverages” accounted for 9.3%, and a broad category of “junk foods,” including sweetened beverages, desserts, salty snacks, candy and sugar accounted for 22.6%.
Nutrition and hunger advocates jumped on those numbers as apocalyptic, even though the USDA study itself cited national survey data showing “food purchases, consumption patterns, and dietary outcomes among SNAP participants and higher income households are more similar than different,” echoed by their study results that non-SNAP consumers spent 4% rather than 5% on soft drinks, 7.1% rather than 9.3% on sweetened beverages, and 20% rather than 22.6% on junk foods. NYU professor Marion Nestle concluded they meant SNAP was “a multi-billion-dollar taxpayer subsidy of the soda industry.” David Ludwig, of Boston Children’s Hospital, concluded that “let’s not use government benefits to pay for foods that are demonstrably going to undermine public health.”
A reform such advocates, along with several city and state governments, commonly endorse is to ban purchases of sweetened sodas from SNAP eligibility. Others want to expand the ban to all sweetened drinks or all junk food. To date, such proposals have not gotten USDA approval, but many are trying to leverage the latest study findings into permission.
Reconsider our family already spending $400 per month on food, and then getting $256 in SNAP benefits. The percentages for those “unwelcome” categories would increase soda spending from $20 to $32.80, sweetened beverage spending from $37.20 to $61.01, and junk food spending from $90.40 to $148.26. The consequence is that even exempting the broadest category, SNAP’s effect would still be the same as giving cash. $250 in allowed food (now paid for with SNAP benefits rather than cash) plus $148.26 in disallowed food (still paid for in cash) is less than the $400 cash previously spent on food. Incorporating the fungibility of money, the $256 in SNAP benefits still gives the family $256 more to spend any way it wants, despite the restriction.
For years, food stamps have been promoted with logically invalid arguments. Ignoring the fungibility of money has been an important component. SNAP supporters ignore the fungibility of money to overstate the increases in food consumption produced by SNAP, which also overstates the stimulative effect on agriculture and grocery interests (not to mention ignoring that the taxes imposed on others to fund SNAP will decrease their similar purchases). Further, any food consumption increases come at the expense of other consumption that consumers deem even more valuable. When we look beyond food purchases to effects on hunger and nutrition, however, clear benefits cannot be conclusively demonstrated. That justifies criticism of SNAP. Unfortunately, many critics make a similar mistake when they want “reforms” that would exclude unwelcome food from the program that will similarly leave the results essentially the same as if recipients received cash (except for much higher administrative costs, which would reduce the actual benefits received for a given program budget, harming recipients).
Money’s fungibility undermines rationales for current SNAP policy, as well as proposed SNAP reforms. But it also undermines intended or promised effects of policies in several other areas, including housing subsidies, heating subsidies, bond financing of construction, state lotteries with proceeds dedicated to education, and foreign aid. Further, in a broader sense, such common failures to understand a very simple characteristic of money across many areas provide one more gaping set of holes in any trust that our “public servants” are expert enough to make us better off than we would make ourselves with our own resources.