Soda Tax, Obesity, and Incidence

Today’s Washington Post Eat, Drink, and Be Healthy column examines the possible role soda taxes could play in curtailing obesity. The column cites the research of Kelly Brownell saying:

The idea is simple: Slap an 18 percent tax on soda, and people will drink less of it. Since increased soda consumption is, Brownell says, one of the main contributors to our rising obesity rate, cutting back should lead to nationwide weight loss.

The Post is skeptical (and rightly so) of such arguments based on worries of government paternalism:

To some critics—and, I’ll confess, to me—Brownell’s approach smacks of paternalism and over-reliance on government intervention. Shouldn’t diet and weight be a matter of personal responsibility, not the government’s concern?

The anti-paternalism argument is just one of a number of arguments that can be made against such selective excise taxes.

Can we expect the tax on soda to truly reduce obesity as much as claimed by a number of studies? The decrease in consumption will likely not affect the people you are attempting to help. As an economist, let me be clear. I am not arguing that soda taxes won’t reduce consumption. As price goes up, of course, the quantity demanded in the market will fall. But that demand will not fall equally for every individual.

Let me illustrate this with an example. My name is Kail Padgitt and I am an addict. I “need” two cups of coffee every morning and a third in the afternoon. I share an office at the Tax Foundation with another economist, Mark Robyn, who is not an addict. He will, however, enjoy a cup occasionally in the morning. Let’s say the government decides that over consumption of coffee (3 or more cups a day) poses a national health crisis. It places a $2/lb excise tax on coffee. Mr. Robyn will decrease his consumption as he responds to the price increase and drinks Earl Grey instead. I will continue drinking coffee at only a slightly decreased rate because I am much less responsive to the price change. Thus the government programs will do little to curb my addiction (except make me poorer) and will take a healthy pleasure away from Mr. Robyn. When you look in aggregate there has been a decrease in consumption, but that decrease has not affected the group you wished to target.

This is not just economic story telling. A new paper at the NBER reports the same effect for alcohol.