When We Put Money Where People’s Mouths Are, What Should They Eat?

The principal aim of the SNAP program, formerly “food stamps,” is important and simple: to ensure that poor people struggling with food insecurity have enough to eat. The program has always been of great public health importance, but in the aftermath of the great recession, more than one in seven American families found themselves on the SNAP rolls. Enrollment has recently declined, due likely to improving economic conditions. This shows the system works as intended, helping those who truly need it.

But even simple ideas can run into complications. SNAP funds cannot be used for tobacco or alcohol. This makes intuitive sense, since these are not necessary to alleviate the food insecurity SNAP is intended to combat. But it raises a question on a slippery slope: should SNAP funds be used for anything that does more to promote ill health than good? When that question is asked, sugar-sweetened beverages, a.k.a. sodas, come immediately to mind. They are of no nutritional value, and decisively implicated in obesity, diabetes, and other maladies.

Some argue that SNAP funds, which come from taxpayers, should not be used for items like soda that are unnecessary (water is a perfectly adequate, preferable, environmentally friendlier, and less expensive answer to thirst); harmful; and associated with additional costs.

What additional costs?

Well, I have frequently lamented that despite its good intentions, SNAP uses some $80 billion a year in taxpayer revenue to help poor people choose poor food that results in poor health. Then, that poor health requires treatment, much of it covered again by taxpayers through Medicaid, at a cost enormously higher than SNAP itself. Everybody in this scenario loses: the SNAP beneficiaries have something to eat, but also have diabetes, and need coronary bypass; the taxpayers’ good money is ill-spent on bad health; and the government has an empty war chest.

There is, in fact, evidence of just this. Colleagues and I found, in a study published in 2014, that diet quality was lower in SNAP participants than in matched, SNAP-eligible non-participants.  In other words, the financial support provided through SNAP helped people make their diets, and by extension their health, worse.

There is nothing very surprising about this. We know that in general poverty and food insecurity are associated with poorer quality diets and health, and poorer quality food choices- in part because people facing socioeconomic disadvantage live in communities that conspire against health both by what they do offer (e.g., a particular density of fast food restaurants), and what they don’t (e.g., appealing resources for physical activity and recreation; affordable fresh produce, etc.). We know, as well, that the modern food supply has been booby-trapped to induce maximal eating. This effect is greatest in the most highly-processed foods, the very foods that prevail in the fast-food outlets and bodegas that are apt to populate a community where SNAP is in wide use.

That’s important, because it indicates clearly that bad choices associated with SNAP support are not the “fault” of the SNAP beneficiaries. There is a perfect storm driving poor people to poor food, and from there, to poor health.

The solution proposed by some is to fix the problem at the source, and limit the use of SNAP foods only to foods deemed suitable. But I trust you see the many problems with that. For starters it is quintessentially paternalistic. In effect, Big Brother gets to decide what you can have for breakfast. For another, it requires universal agreement on what foods are suitable, which in turn means some agreed-open standard for the nutritional quality of all foods. I know from personal experience that challenge can be met, but there is as yet insufficient consensus on the matter to inform food policy.

So we wind up between the proverbial rock and hard place, with an apparent choice between giving out carrots (i.e., no-strings-attached financial aid for food purchases), and wielding a stick (i.e., telling poor people what food they can’t buy).

A new paper suggests there is a middle path, and as someone who likes to walk just such roads- I am delighted to see it.

A study in JAMA Internal Medicine compared the effects on diet quality of providing just incentives for uncontroversially ‘healthy’ foods; just restrictions on uncontroversially ‘unhealthy’ foods; both; or neither, in a population of SNAP participants. The best results were seen when incentives and restrictions were combined.

The alternative to outright food restrictions would be financial disincentives, so that foods comfortably catalogued as ‘unhealthy’ came with a penalty. A SNAP-specific application would levy such a penalty within the SNAP system only. A population-wide approach to the same disincentive would be a tax, like the soda tax approved recently in Berkeley, California and Philadelphia, Pennsylvania.

Whatever approach to financial disincentives is applied, it’s easy to view it as the proverbial stick, and contrast it with the carrot of financial rewards for salutary selections. I have mused before in just that direction, as did my friend and colleague, Dr. Marlene Schwartz, editorializing on the new study.

My argument has been that I favor carrots to sticks, both when working to advance public health, and the forward motion of my actual horse. Experience in the saddle, however, makes the case for both- however seldom and gently the stick is wielded. The new study findings say much the same about SNAP.

Why? In the absence of restrictions or disincentives, the addition of incentives to the SNAP program seem prone to increase overall eating. Given the associations between poverty and obesity, and between obesity and chronic disease, an increase in total calories consumed is decidedly counterproductive. The real goal is to reduce intake of ‘bad’ foods, and increase intake of good in their place. Disincentivizing the former, while incentivizing the latter, aptly conforms to just such swaps.

With validated metrics for the overall nutritional quality of foods that enable us to go beyond the narrow categories of ‘uncontestedly good’ or ‘uncontestedly bad’ foods, we might think in terms of more nuanced gradations. We could, if so inclined, disincentivize the least nutritious entries in any food category, while incentivizing every movement up an objective scale. Future studies are required to determine how much greater the impact of this soup-to-nuts approach might be, but evidence to date suggests such comprehensive nutrition guidance is constructive all on its own.

Once we have the best methods worked out, there is no need for this approach to remain relegated to the SNAP program. The private sector would have all the same reasons to replicate the approach, promoting human health while reigning in costs.

For now, the evidence we have for getting those reigns to work argues for a generous accompaniment of carrots, with strategic, judicious applications of some kind of stick.

One thing’s for sure. When we put literal money where people’s mouths are, the intent should be to succor, not sicken. Data, like those newly provided, should help direct us there from here.



David L. Katz

Director, Yale University Prevention Research Center; Griffin Hospital

President, American College of Lifestyle Medicine

Senior Medical Advisor, Verywell.com

Founder, The True Health Initiative

Follow at: LinkedIN; Twitter; Facebook

Read at: INfluencer Blog; Huffington Post; US News & World Report; Verywell; Forbes


When We Put Money Where People's Mouths Are, What Should They Eat?

Show Buttons
Hide Buttons