Jeff Strnad has written an excellent article on this subject, "Conceptualizing the 'Fat Tax': The Role of Food Taxes in Developed Economies."

Here is the link:

http://papers.ssrn.com/sol3/papers.cfm?abstract_id=561321

And here is the abstract:

Public health scholars and practitioners in several countries have called for a broad system of taxes on unhealthy foods, possibly combined with subsidies for certain healthy foods. The typical motivation for these fat taxes is the public health perspective, the idea that it is socially valuable to make health outcomes better regardless of how individuals might trade off risky dietary behaviors with those outcomes. This paper assesses possible alternative rationales for such a tax, examining arguments based on behavioral and cognitive failure as well as traditional economic arguments.

The strongest role for a broad, health-based tax on foods would be as a health-insurance system component. Ideally, the tax amount would approximate the expected medical cost from consuming each unit of food, and the tax revenues would cover claims for the associated disease conditions as they arose. This tax/claim system would serve four major insurance-based purposes. First, it would reduce moral hazard by causing individuals to make ex ante payments covering the expected covered costs of risky eating behavior. Second, it would tend to alleviate adverse selection problems by providing implicit risk adjustment for the associated maladies. Third, it would address the incomplete markets problem that consumers are restricted to short-term coverage by introducing a long-term component into the system. Finally, it would aid in separating inherent risk over which individuals have no control from risk that is a function of behavior. Serving these purposes would be especially valuable in developed country systems where coverage is universal or a certain minimum treatment package is available regardless of ability to pay.

Whether a fat tax served to further insurance system objectives, general public health goals or other purposes, there are substantial implementation and scientific causality issues. The health impact of particular foods is subject to scientific uncertainty and often depends heavily on the other components in each individual's diet. Disease responses to food products are non-linear in quantity consumed and vary substantially across individuals. Individuals also are heterogeneous with respect to their preferences and self-control. A food tax scheme will result in an industry response and may spur both constructive and detrimental innovation.

The paper examines many non-insurance-based rationales for a fat tax, including correcting externalities or internalities (problems of self-control), alleviating bounded rationality (inability to be adequately informed at low cost), and addressing various cognitive or behavioral errors. The optimal form of the tax might be affected by such errors. For instance, a tax based on energy density rather than components such as fat might best address the obesity epidemic. Preliminary estimates show that such a tax, if focused on foods that are nutritionally poor as well as energy dense, might be quite large compared to existing prices for many such foods. The paper also considers the possibility of basing the tax partially or wholly on certain biomarkers (weight, blood chemistry, etc.) and briefly considers litigation as an alternative to taxation.

Will, it seems like you want to weigh the pros and cons of a proposed fat tax against a world with no taxes and no subsidies. Is that realistic, or even useful? We are indeed living in a welfare state (is that really SO scary?), and Twinkie-makers get their share of subsidies from Congress, which may well lead to lower cost products and overconsumption of fatty foods.

(Better perhaps to concentrate on ending subsidies than adding new taxes to offset the externalities fueled by those subsidies, but that's just a political question, I think.)

Yes, everything in the world is taxed or subsidized ... indeed it is hard to design a tax system without touching on pretty much every human activity in some way. Once you swallow that bitter pill, though, you gotta start somewhere, and the fact that we have other products with negative externalities (cigs, motorcycles, cell phones) doesn't make much of a case for avoiding talking about a fat tax.

Readers also may want to check out Katie Pratt's paper in addition to Strnad's. My thoughts are here: http://vic.typepad.com/taxingblog/2005/04/pratt_on_the_fa.html

I certainly agree with you, though, that dedicating the money to the already over-subsidized health care system is not a good idea.

Nice post.

I don't mean to be conceptualizing a world with no taxes and subsidies, but rather doing the opposite. Given that we subsidize almost all unhealthy choices, why on earth should we specifically pick on eating Fritos? I mean to be employing the sort of basic argument from the second-best. It would be nice if we didn't subsidize bad choices with federal tax dollars, but we do. So does it make any sense to start selectively picking on particular bad choices? I think not.

If promiscuous sex, unprotected sex, driving cars, reading drowsily by candlelight, shaving with straight razors, working office jobs rather than jobs that give one exercise, and so on and so on and so on are all subsidized by the health care system, picking on a particular practice doesn't really restore us to economic rationality.

You say that pointing out other pathologies of the system doesn't let us avoid talking about the fat tax, but why not? Maybe "you gotta start somewhere" but that's not an excuse for starting here rather than with, say, dangerous construction jobs.

And I'm really not sure that its true that "you gotta start somewhere". We could just debate whether we ought to subsidize unhealthy choices as a general matter, rather than have federal regulators start giving the thumbs-up or thumbs-down to ever unhealthy human activity.

Oh. But to the extent the justification for singling out, say, saturated fat is that we for some reason subsidize the production of saturated fat, it would be truly silly to keep the federal subsidy and try to fight it back with a federal tax. We could just repeal the subsidy.

Wait, aren't Twinkies and the like eaten disproportionately more frequently by those of lower incomes? It would make quite a bit more sense to end subsidies for corn-syrup manufacturers.

I fail to see how subsidizing all unhealthy activities is better than subsidizing some unhealthy activities.

Starting with the fat tax makes some sense because there's a fair amount of research done and it seems relatively easier to get a handle on. And, someone brought it up. I'm equally happy to talk about dangerous construction jobs. Or we can just start a big list.

Anyway, Will, at least we agree that in a sensible world, we should fight subsidies for agribusiness, repeal the employer deduction for health care, and perhaps then reevaluate to see what externalities remain. I readily acknowledge that the real story behind the fat tax is that it appeals politically to a demographic that, say, a tax on dangerous construction jobs does not.

The "fear" in going after federal-unhealthy-activities-subsidies retail rather than wholesale is that it will mostly just be a license for moralizers to pick on activities that they find morally dislikable. I'm not sure the gains of not-subsidizing some activities are worth the losses in equality, where some people's vices are federally subsidized and others are taxed.

I should add that this fear only comes to fruition if there are swing votes and voters who would be unwilling to, say, regulate consumption of sweets on strictly moral grounds, but who will vote for the selective non-subsidy of the same things, on the grounds that one must start someplace. C.f. Eugene Volokh on the mechanisms of the slippery slope.

The healthcare system would be so much more efficient if there was free market competition between healthcare providers. High deductible insurance would still be there if you got really sick.