MacCoun (1997) notes that the total harm (of a given type, for a given drug) is the product of average harm (per incident of use)×total use (number of incidents of use). Thus, if reductions in average harm bring about increases in total use, there is no guarantee that a drop in average harm will produce a decline in total harm. Critics of harm reduction strategies (such as needle exchange) often contend that such approaches send the wrong message—potentially undermining attempts to discourage drug use. Although that claim is rarely articulated in any detail, MacCoun (1998) has examined several plausible interpretations. One is relevant to the present discussion: specifically, the notion that a reduction in the average harmfulness of an incident of drug use may make drugs more desirable. MacCoun (1998) cites evidence from other domains suggesting that interventions to reduce risk can have offsetting behavioral effects; for example, the presence of air bags tends to make people drive slightly faster, everything else being equal. Yet in the domains that have been studied, the safety interventions have been beneficial on the whole, despite these offsetting effects. MacCoun argues that the available evidence suggests that needle exchange programs reduce net harm in this same sense. But he argues that as a general principle, all drug interventions should be evaluated for their effects on several key dimensions: the likelihood of using a drug, the quantity and frequency of drug use, and the average harmfulness (by various criteria) of each incidence of use.
No responsible analysis of the harmful consequences of drug use can ignore the possibility that many of the harms of drug use are either caused or augmented by the legal prohibition against these drugs and its enforcement. Drug prohibition is inevitably a source of government intrusion into citizens’ lives. Many (but not all) overdoses occur due to the unknown purity and potency of illegally purchased drugs. The sharing of contaminated syringes is largely a consequence of the artificial scarcity created by their illegality. And much of the criminality and violence associated with drug use (but by no means all) is due to the high price of illegal drugs and the conditions of their sale in illegal markets.
Consider the drug-crime link. Goldstein (1985) has provided a useful distinction between three types of links between drugs and violence. First, psychopharmacological violence is attributable to the fact that the ingestion of some drugs (e.g., alcohol, stimulants, barbiturates, PCP) makes some individuals more excitable, irrational, or violent than they would be otherwise. Second, economically compulsive violence occurs when ad-