Soda Taxes are Coming. What's the Problem?

The idea of taxing soda to pay for health care came up on the blog a couple of weeks ago, but now the Center for Science in the Public Interest is actually planning on proposing such a tax. That proposal is coming in for a fine round of snark (see Ann Althouse, Michele Bachmann, etc) so I thought I would self-plagiarize for a bit, and repeat why I can't really get excited about the whole controversy.

Soda taxes are the least popular item on the chart above. But why? Like cigarettes and alcohol, you aren't doing yourself any favors by consuming an abundance of soda. And like cigarettes and alcohol, unhealthy foods can have negative externalities -- namely, there seems to be some causal connection between obesity and the obesity of one's friends.

(It's also true that a tax on soda would, like a tax on alcohol or a tax on cigarettes, be fairly regressive. I think that's a fair objection, but it's one that should be applied consistently: Why not reject cigarette and alcohol taxes for the same reason?)

The best objection to the tax might be a good old-fashioned problem of line drawing. It's easy to draw a bright line between products that have alcohol and products that do not. It's easy to draw a bright line between products that have tobacco and products that do not. It seems quite a bit more challenging to draw the relevant distinctions between a sugary beverage with 97 calories per serving and a sugary beverage with 10 calories per serving.

I sympathize. But I'm not sure a problematic line-drawing issue is a good reason to get huffy over the tax.

Update: I think Kevin Drum has exactly the right solution to the line-drawing problem: Just tax the high-fructose corn syrup, not the products containing high fructose corn syrup! That seems like a smart idea.

The one point I would add to Drum's analysis is that you'd then have self-canceling public policy, since we currently give massive subsidies to the American sugar industry. It would be strange to subsidize sugar at one end, and tax it away on the other. At the very least it would create a gratuitous number of transaction costs.

Update # 2: I see my Atlantic colleague Corby Kummer has more on this.