The Wages of Sin Are Death, Prison, and Taxes

Mark Kleiman (here, here, and here) and Juan Non-Volokh (here and here) have been having a polite debate on the subject of alcohol taxation. Click the links to read Iain Murray's and Stuart Buck's comments on the same topic.

Mark’s main point is that alcohol consumption creates a variety of negative externalities, such as drunk driving accidents, increased public health costs, criminality, and so on, and a tax would internalize the externality. His back-of-the-envelope calculations indicate that each beer consumed creates external costs of about $1. Juan replies by emphasizing that a tax would fall on all consumers of alcohol, not just the minority of abusers who create the vast majority of the social ills associated with alcohol (a harm Mark concedes but considers justified on net). Juan also observes that alcohol consumption can also create health benefits in terms of preventing heart disease, and some of that beneficial consumption would be deterred by a tax.

All of this is fine, but I think there’s a point that both Mark and Juan have given inadequate attention. Taxes are by no means the only way to internalize externalities. There are both civil and criminal punishments for the ill effects of drunk driving and other criminal behavior. If you actually cause harm to another person or property, you can be sued in civil court. If you get caught driving drunk, you can lose your license and go to jail. If you’re still allowed to keep your license, your insurance premiums will assuredly rise. Note that all of these types of punishment are focused on the real abusers, not the responsible drinkers. All of these facts indicate that if the external cost of a drink is $1 (Mark’s estimate), the optimal tax would have to be less than $1. To figure out how much less, we would need to calculate the expected dollar value to the consumer of the potential punishments that would result from the marginal drink. And don’t forget that taxes on alcohol already equal about 25% of the net-of-tax value per gallon (the average price per gallon of malt beverage is about $1). So we’re already about 25% of the way to Mark’s estimated external cost, without even having taken into account non-tax punishments.

Mark could still be right about the optimal tax needing to be raised (though not by the whole $1 per drink). The answer depends on two main things: first, the expected dollar-value of jail time, loss of license, and other non-tax punishments; and second, whether you’re willing to assume all policies other than the tax rate remain fixed. If the size of non-tax punishments is increased, the optimal tax will naturally decrease. And keep in mind that some of the “external” costs that Mark cites are not inherently external, such as the public health costs in socialized healthcare systems (like Medicare, Medicaid, and public ERs). As frequent readers of this blog will know, this is one of my hobby horses: when you socialize health costs, you create a political incentive for the regulation and taxation of personal lifestyle choices. Alcohol is but one example; others include eating fatty foods, engaging in high-impact sports, and being sexually promiscuous. Would we also support placing excise taxes on these other activities, which unlike alcohol currently are either untaxed or taxed at the same rate as everything else?

Oh, one other thing: the strategy of setting a tax equal to the marginal external cost only makes sense if the market is actually a competitive one in other respects, so that the price of the product is relatively close to its marginal cost of production. But in the case of alcohol, a slew of state and federal rules protect wholesalers from competition (and the laws also prevent suppliers and retailers from dealing directly with each other -- alcohol products must go through the wholesale sector). As a result, there is good reason to believe that alcohol prices are already a good bit higher than marginal cost of production.

[Note: the statistics I’ve cited here all came from my head, as a result of some research I did about a year ago. They do have original sources, of course, which I could track down if I had to, but I’m too lazy to do it right now.]