Study: Soft Drink Taxes Make People Slightly Thinner

Attempts to improve public health (and raise government revenue) by taxing soft drinks have been in vogue of late, most notably including New York Governor David Paterson’s abortive attempt to impose an 18% ad valorem tax on sugary drinks. But do soft drink taxes really affect public health? Via the National Center for Policy Analysis, I ran across an August 2008 working paper from Emory University that examines the relationship between taxes on soft drinks and body weight.

According to Jason Fletcher, David Frizvold and Nathan Tefft, taxes on soda do appear to affect obesity, but the effect is very small. They estimate that a 1% increase in the soda tax rate decreases mean adult body mass index by 0.003. So, even if sodas faced a 58% tax rate (similar to the average effective tax rate on cigarettes) the authors estimate that mean BMI would fall by just 0.16 points. For contrast, mean American BMI has risen by 2.3 points between 1990 and 2006.

One limit of this study is that looks just at general taxes on soda. Two-thirds of Americans face some sort of tax on soda purchases; for 41%, the tax rate is higher than for other food items. However, the average tax rate is low, at around 3%, and so the public health effects are infinitesimal.

Two alternative policy approaches might have greater health effects per dollar of tax levied:

• Tax only sugary drinks. Every state currently treats sugary drinks and diet drinks in substantially the same way for tax purposes. However, Paterson’s proposed tax would have applied only to sugary drinks. Consumers can only avoid a general tax on sodas by switching to other beverage categories, a fairly unpleasant move for people who enjoy soda. They could avoid a sugary drink tax simply by switching to diet soda, a closer substitute; therefore, more consumers might modify their behavior to avoid a sugary drink tax than an all-drink tax.
• Don’t allow food stamp purchases of sugary drinks. States are prohibited from levying sales tax on food stamp purchases, even when the food stamps are used to buy products that would ordinarily be taxable. So, a state can impose a tax on sugary drinks, but it can’t make food stamp recipients pay the tax. This is especially problematic because the Emory researchers found that soft drink taxes are most effective at discouraging soda consumption by low-income people. Barring food stamp purchases of sugary drinks would force stamp recipients to pay with taxable cash.

Of course, either of these actions would be paternalistic. A sugary drink tax might be a bad idea even if it would have good public health effects, because the tax would be complicated to administer, or would be regressive, or simply because paternalism is inherently objectionable. Such a tax would also be bad if the tax rate exceeded the marginal social cost of soda consumption, a likely outcome since political debates about sin tax rates rarely place much weight on the actual magnitude of the externalities that supposedly justify taxation.

But if we want a paternalistic policy to discourage soda consumption, there are likely more effective options than simply levying a general tax on drinks.