Drug Legalization

A recent monograph published by the libertarian Cato Institute—Jeffrey A. Miron and Katherine Waldock, The Budgetary Impact of Ending Drug Prohibition (2010), available at www.cato.org/pubs/wtpapers/DrugProhibitionWP.pdf--offers an estimate of the budgetary cost to the U.S. government (federal, state, and local) of the federal and state legal prohibitions against the sale and use of marijuana, cocaine, heroin, and other mind-altering drugs. The lead author, Jeffrey Miron, has an economics Ph.D. from MIT and lectures in economics at Harvard; he has published extensively on the economics of the drug prohibition. His coauthor is a doctoral candidate at NYU's business school. I will summarize the monograph and then offer some thoughts of my own on the question of legalizing the illegal drugs.

The authors estimate that legalizing these drugs (which would require repealing both federal and state prohibitions) would reduce government expenditures by $41.3 billion per year, with about two-thirds of the savings accruing to state and local government. The savings would involve reductions in police expenditures, in prosecutorial and judicial expenditures, and in jail and prison expenditures. The authors estimate the reductions by multiplying the various expenditure categories by the percentage of arrests, prosecutions, and prison terms that are attributable to drug offenses. This is a crude method of estimation, because different types of criminal offense involve different amounts of police, prosecutorial and judicial, and prison resources; for example, the length of imprisonment for a particular type of offense is the best estimator of the prison costs for that offense, and the length varies across types of offense.

A further problem with the method of estimation is its disregard of fixed costs. Given fixed costs, a reduction in output will not reduce total costs by the same percentage as the reduction. At least not immediately; in the long run, a reduction in output should reduce total costs proportionately or nearly so, because in the long run all or at least most costs are variable.

So the $41.3 billion figure has to be taken with a grain of salt, but since it could be larger or smaller, it is a legitimate starting point for analysis.

In addition to reducing expenditures on law enforcement, legalizing the illegal drugs would, the authors argue, increase tax revenues (federal, state, and possibly local as well); they estimate the increase at $46.7 billion a year. They assume first, conservatively, that the demand for the drugs would not increase; that is, at a given price, the amount of drugs purchased would not increase. This is a counterintuitive assumption, since the illegality of the drugs discourages their purchase; but the authors point out that increased consumption of drugs might come largely at the expense of consumption of alcohol, tobacco, and other goods that are taxed as drugs would be.

However, the price of drugs would fall; although legalizing the drugs would result in the sellers having to pay taxes and incur regulatory expenses, and would result in advertising expenditures that might shift sales among sellers rather than increase demand, the authors plausibly assume that these effects would be offset by the elimination of the heavy costs that prohibition imposes on sellers, notably the threat of punishment and of gang violence.

The decline in price would have two effects, however, which might be largely offsetting from a tax-revenue standpoint: the amount of drugs sold would rise, but tax revenues per sale would fall because the tax rate would be based on the sale price. The net effect would depend on the elasticity of demand for the drugs. If demand is inelastic, this means that a fall in price will not generate a proportionate increase in amount purchased; since revenue is price times quantity, total revenue will fall when price falls. The authors note that the demand for drugs is believed to be inelastic, and if so then revenue will fall more than price if drugs are legalized, and this will reduce the amount of tax collected relative to a good the demand for which is elastic.

Additional tax revenues, however, would be generated by income tax paid by the sellers; drug gangsters do not pay income tax on their income from criminal activity.

From estimates of drug consumption and prices—estimates that must however be taken with a grain of salt, once again, because there are no reliable records of illegal transactions—the authors derive the $46.7 billion figure for increased tax revenues from legalization; this assumes tax rates similar to those on clost substitutes, such as alcoholic beverages.

The sum of public expenditures that would be avoided and additional tax revenues that would be generated is, according to the authors' estimate, $88 billion.

Their estimates are broken down by drugs. Of particular significance is their estimate that the total budgetary improvement from legalizing marijuana, the least controversial of the illegal drugs, would be only about $17.4 billion, of which only $8.7 billion would come from reduced expenditures on law enforcement (computed from tables 3 and 4 in the monograph), the rest representing increased tax revenues. And as we know from the recent decision by the Justice Department to continue enforcing the federal law against marijuana in California, despite that state's repeal of its state prohibition, both federal and state prohibitions must be repealed for the legalizing of tegalization of illegal drugs to be effective.

Although I think the authors' estimates are good enough to be a valid starting point for evaluation of the budgetary benefits of legalizing drugs, it is important to note that their monograph is not a cost-benefit analysis in the usual sense. True, the costs of police, judges, prisons, etc. are social costs; that is, they are resources that have opportunity costs. But tax revenues are a transfer payment rather than an increase in overall national wealth. The authors do not attempt to estimate whether taxing drugs is efficient relative to other taxes, though what is true is that taxing drugs is cheaper than prohibiting them, because collection costs are cheaper than the law enforcement costs that prohibition imposes.

Most important, the authors also do not consider the possible social benefits of prohibition. Prohibition reduces the consumption of mind-altering drugs. Of course there are mind-altering drugs that are not prohibited, and many of these are close substitutes. These include the numerous prescription drugs that have mind-altering effects very similar to those of the illegal drugs, and of course there is alcohol and cigarettes. Moreover, a tax on legalized drugs would raise the price to the consumer and thus moderate the effect of legalization on consumption. But if the tax is too high, it will result in reviving the illegal industry. And the authors probably underestimate the increased consumption that would result from a lower price or even the same price (brought about by a particularly stiff excise tax) because they don't mention concerns with impurities and with the stigma of being a "drug addict" that are created by the prohibition and would be substantially reduced by its repeal.

The question would then be whether the external costs of increased consumption of mind-altering drugs would exceed the savings in law enforcement costs from legalization. It seems doubtful that marijuana consumption generates significant social costs, but legalizing it would generate only modest cost savings--$8.7 billion a year, according to the authors' estimates. But cocaine, especially the crack form, along with heroin, ecstasy, LSD, methamphetamines, and perhaps others, may induce behavioral changes that cause social damage. Most leaders of black communities believe that rampant drug usage is highly destructive to their communities, and not only because of the gang activity that prohibition induces. Drug gangs would disappear with legalization and that would reduce the violence in those communities, but the effect might be more than offset by the effects of greater drug use.

Concern with the huge budget deficits of our federal, state, and local governments may gain the authors a more sympathetic reading than advocacy of repealing the drug laws usually does. From a budgetary standpoint, the authors are estimating an annual savings of almost $90 billion. But without an estimate of the social costs of increased drug usage, the path to repeal is blocked. It would a step in the right direction if the Justice Department would take the position that it will not enforce a federal drug law in any state that repeals its parallel prohibition of that drug; that way we might obtain experimental evidence of the social costs of illegal drugs.

Drug Legalization—Posner

Note to Drug Legalization Post