population consists of two distinct groups: a large group of individuals who experiment with illegal activities for a relatively short time during adolescence and then desist, and a small group of offenders who begin their criminal careers earlier, end them later, and offend at higher rates during their criminal careers. The latter group is responsible for the majority of the crime that occurs.
MacCoun (1998) also notes that there are several plausible causal interpretations of the basic findings that tobacco, alcohol, marijuana, and hard drug use are associated and tend to occur in a particular sequence. One interpretation suggests that the association is spurious or noncausal— specifically, the notion (discussed earlier) that the use of the early and late drugs in the sequence reflects some common risk factors, with timing determined by price and availability. Other interpretations are causal— for example, experiences with the early substances in the sequence might: (a) stimulate one’s interest or appetite for the later substances, (b) change one’s beliefs about the severity of health, legal, or social risks of drug use, (c) bring one into contact with a subculture of hard drug users, or (d) bring one into contact with hard drug sellers. Interestingly, the first causal interpretation is widely cited in the United States as a basis for stringent sanctions against marijuana; the latter three interpretations were influential in the development of Dutch drug policy, which seeks to separate “soft” and “hard” drug markets and cultures (see MacCoun and Reuter, 1997).
Few actual data are available to direct policy decisions about the targeting of prevention activities, but debates over appropriate targeting have appeared at the margins of the prevention literature. Although the field continues to be predominated by the gateway ideas, a few commentators have questioned this approach (Brown and Kreft, 1998; Gilham et al., 1997; Gilchrist, 1991). Brown and Kreft (1998) argue that the “no use” messages typically conveyed in universal prevention programs actually increase use among those most at risk for using. These youths are more knowledgeable about drugs and their effects than prevention curricula assume, and the naive messages conveyed in the programs serve to create cognitive dissonance in the minds of these youths.
Gilham et al. (1997) argue that the results of research on these universal prevention programs is misleading because the programs have no effect on the large proportion of the population that is not likely to use drugs even without benefit of prevention programming, but they may have large effects on the smaller population that is at risk. They argue that the more substantial effects on potential users are diluted in studies that report findings for the entire population.
For example, suppose that 98 percent of a population targeted for universal prevention programming will never use heroin, and that the