10/13/2010 05:10 pm ET | Updated Nov 17, 2011

Is Soda Food?

I have a sneaking suspicion that the "keep the government out of our business!" crowd will object to the plan by NYC's mayor, Michael Bloomberg, to preclude the use of food stamp assistance -- known as SNAP -- to purchase soda. But here's the irony in that: This is the government minding its own business, literally, more than anyone else's.

First, the Supplemental Nutrition Assistance Program is, of course, a government program overseen centrally by the USDA, with administration farmed out to state and municipal authorities. If the government is giving out the money, the notion that it has a say in what it can be used to buy may not be entirely outrageous. I trust we can all agree that there would be cause for protest if money or vouchers given to help sustain a hungry household were spent on, say, cocaine, or for that matter tobacco, right? Soda is not cocaine, or tobacco, but it is an at least nominally addictive substance, corrosive if not abruptly toxic to health, and devoid of any nutritional value food assistance is intended to help obtain.

But second, and at least as important, the ranks of SNAP participants -- much enlarged courtesy of our global economic woes-- are substantially the same ranks enrolled in Medicaid, government health insurance for the financially challenged. So here's the thing for all to ponder: should government food subsidies be usable for the very kinds of foods that are most strongly associated with bad health outcomes, for which additional government expenditures will be required? No private business would function so dysfunctionally, and this seems to bespeak the very kind of government inefficiency to which the shrink-the-government contingent objects.

Maybe, to the extent possible, the government should mind its own business. But almost any way the soda bubbles burst, this is their business.

So, I support the move. Food assistance dollars should pay for food. Only in a modern world habituated to glow in the dark comestibles could soda even be considered food. It is, in essence, a sugar -- or in its diet guise -- an artificially-sweetened chemistry experiment in a cup. Typically, a very large cup. At least as far from anything that truly deserves to be called food as it is from cocaine.

So it comes down to this: maybe food assistance dollars should be limited to food. Is there really cause for protest in that bold position?

Of course, it does raise a challenge the American Beverage Association will no doubt quickly voice: is soda demonstrably less food than, say, Fluffernutter; artificially colored hearts, moons, stars, and clovers; cheese-like spread (whatever that is, exactly); fruit roll-ups (that have, in fact, never met a fruit); and so on? Their anticipated protest is legitimate, and not in equal measure.

Legitimate, because soda is by no means the only detour from the realm of actual food into the realm of modern pseudo-food concoctions we eat routinely. But not legitimate because soda is a very good place to start the process of limiting dedicated food assistance dollars to food. It is about as devoid of redeeming nutritional properties as a food can be; as a beverage presumably intended to address thirst, it is readily replaceable with something that is available to almost everyone for free (or nearly free), namely- water; and, it is decisively and specifically associated with weight gain, obesity and compromised health.

But even as I defend and applaud the mayor's plan, I feel obligated to point out that we could do far better than this. We could make the SNAP program work better by applying a carrot, rather than a stick and by addressing food choice from soup to nuts, rather than just from Coca Cola to Pepsi.

Here's the idea: directly link the purchasing power of food assistance dollars to the nutritional quality of the food selected. The rationale for such an approach is both clear and compelling: If the government is providing money for food, then the more nutritious that food is, the better spent those dollars. And, just as bad food (or pseudo-food) can propel people toward ill health, true food can immunize them against it and confer vitality. I trust we can agree from all points on the political spectrum that when the government does spend its (and therefore, our) money, it should spend it wisely.

To make this work, we need a universally applicable, objective, reliable measure of overall nutritional quality. As regular visitors here know, we have one. The NuVal system assigns a score from 1 to 100 to any and all foods; the higher the number, the more nutritious the food. Some 75,000 foods have been scored this way to date.

Now, imagine taking these scores and for any given food category -- say, bread -- putting them into quartiles: the lowest scores; the next higher; the next; and the highest. Now, link the scores to SNAP vouchers so that a dollar is worth a dollar in the bottom quartile; it's worth $1.25 in the next; $1.50 in the next; and $2 in the top quartile. No punishment (stick) for choosing food of low nutritional quality, but a direct financial reward (carrot) for choosing food -- in any category -- of high nutritional quality. Such an approach could help make SNAP a means not only of procuring calories, but of actually procuring nutrition. Currently, the SNAP program prevents hunger, but does little or nothing to defend against obesity, diabetes and other chronic disease which, indeed, are disproportionately abundant among SNAP participants. Why not fix this?

And in so doing -- and here's a nod to my "less government" friends -- markedly reduce Medicaid costs? It is my hypothesis -- and that of a large group of esteemed colleagues -- that it would cost far less to subsidize and encourage nutritious foods this way, than it does to pay the costs of diabetes and heart disease care associated with poor food choices.

This group of colleagues -- from throughout the U.S. and indeed, the world -- has joined me in the development and submission of a grant application to the USDA to test this very hypothesis. We want to prove that if food assistance is directly linked to a reliable measure of food quality, health will improve, and costs will go down. We will learn the fate of the grant in November; more on that topic when we do.

For now, anticipating the objections to Mayor Bloomberg's proposal, I defend it. Even though I like using carrots better than sticks. Even though I prefer a soup-to-nuts over a soda-only approach.

I defend prohibiting the use of food assistance dollars for soda because, by any reasonable measure soda isn't really food.

Dr. David L. Katz