Should We Tax Food—or Fat?

Published September 18, 2012

New York City Mayor Michael Bloomberg’s recent proposal to ban the sale of large sodas and other sugary drinks has received national attention. The ban would apply to virtually all sugary beverages—from energy drinks to sodas to pre-sweetened iced teas—at eating establishments such as fast food restaurants, delis, and sports arenas. Any sweetened drink larger than 16 fluid ounces would be prohibited.

Proposals such as this are increasingly common. Even cupcakes and bake sale school fundraisers are in the food police crosshairs. New York City public schools prohibit sales of unapproved home-baked goods. Massachusetts forbids students from handing out sweets such as cupcakes on their birthdays and strictly limits student bake sales. Montgomery County, Maryland prohibits all bake sales.

The federal government is expected to weigh in with its own rules soon. The Agriculture Department says the new rules will allow only infrequent bake sales during school hours.

A new report from the Institute of Medicine (IOM) goes even farther, recommending federal, state, and local governments increase prices on sugary foods, spearhead a marketing campaign to dissuade people from eating such foods, monitor the weight of pregnant women, implement new exercise requirements for children in school and daycare, and create a “food literacy” program for schools.

Food Police Overreach

These actions from the food police are misguided. There may be an argument for taxing fat, but there is no argument for taxing foods that fat people might eat.

If I become obese, does that create costs for you? In a rational world it wouldn’t. When I purchase health insurance, I could be charged an extra premium if my obesity creates above-average expected health care costs. Or if I become obese after acquiring insurance, my insurer could charge me more.

Unfortunately, misguided government regulations prevent premiums from reflecting real health care risks.

In a world in which my obesity creates external costs (higher premiums) for others, there is a straightforward way of “internalizing that externality,” to use an economic term. We can tax fat. The proceeds of the tax could be used to subsidize health care resources that obese patients tend to overuse. Or we could subsidize health insurance pools that attract a disproportionate number of obese enrollees.

Obamacare Makes Taxing Fat Easy

How would that work? Actually, President Obama’s health care law makes it rather easy. Under the law, every patient is supposed to have an electronic medical record, and one of the pieces of information that is supposed to be on that record is a body mass index (BMI) number.

Thus the IRS could match up everyone’s tax return with his BMI and impose a “fat tax” wherever appropriate. If we wanted to home in and make the tax even more optimal, we could combine the BMI measure with other health indicators. If someone has a health problem that is exacerbated by obesity—say, diabetes—a higher tax could be imposed.

There is also a less intrusive way to impose a fat tax. The IRS could simply tax everybody for it, and then non-obese taxpayers could escape the levy by attaching proof of a low BMI to their tax returns.

Regressive Food Taxes

I am not advocating any of this. I am merely pointing out that there are simple and direct “second best” ways of dealing with the external costs of obesity, if we are absolutely opposed to first best solutions. What is absolutely not justified is a tax on foods that obese people might eat.

Like other excise levies, the tax would be regressive—taking more of the income of low-income families than of high-income families. The costs of such a policy almost surely outweigh the benefits. Everyone would be made worse off so the social costs of obesity could be reduced.

Finally, such a tax makes the mistake of taxing inputs instead of outputs. It would be like rewarding students for the number of hours they study, or penalizing them for not studying enough, instead of basing rewards and penalties on what they did or did not learn.