Hume’s last criterion for causality concerns isolation. If marijuana actu- ally causes the use of hard drugs, other explanations should not account
for any association between them. At least two alternative explanations have some empirical support. One concerns the idea that the initiation of each drug arises from its own individual process. Miller (1994) calls this approach the statistical independence hypothesis. This hypothesis states that using one drug stems from its own availability, expectancies, and motivations that are separate from those related to another drug. Thus, the initiation of caffeine may stem from one process; the initiation of heroin may stem from another. The other idea that may account for stages of drug use focuses on the abuse of any and all substances as part of a cluster of larger problem behaviors. This problem behavior theory views substance abuse, unsafe sex, crime, and delinquency as all part of the same underlying trouble.
The rationale for individual processes follows statistical logic. People who participate in rare events likely engage in popular activities first. For example, more people view television than skydive. Thus, we would ex- pect that most skydivers watched TV before they leaped from a plane. This fact need not mean that television causes skydiving. The two acts probably arise from independent, individual processes. Yet the most com- mon one occurs first simply because it is more common. Comparable logic applies to drug consumption. People who use drugs will likely begin with those that are most common. Thus, individuals may use marijuana before cocaine because marijuana is more prevalent in our culture. Yet this fact need not mean that marijuana caused cocaine consumption. A massive study of four national samples including over 6,000 participants suggests that a large portion of the appearance of stages of drug use can be accounted for by statistically independent processes (Miller, 1994). This model does not account for all the data, but independent processes clearly contribute to the progression of drug use.
An alternative way to test the statistical independence hypothesis might examine neighborhoods where crack cocaine is more available than marijuana. If most drug users in such neighborhoods smoked crack before cannabis, the role of availability might receive some support. Under these circumstances, few could conclude that crack is a gateway leading to cannabis. Instead, people use the most available drug first and less avail- able drugs later or not at all.
Another alternative explanation of drug sequencing concerns problem behavior theory. According to this theory, a small group of adolescents engage in a cluster of actions that all may lead to negative consequences (Jessor & Jessor, 1977). These problem behaviors include drug consump-
tion, poor school performance, unsafe sex, and criminal activities. Ac- cording to this theory, the association between cannabis use and the con- sumption of other drugs does not arise because marijuana causes problems with other substances. Instead, both marijuana and the use of other intoxicants arise from the underlying problem orientation in a sub- set of individuals.
Many studies reveal strong correlations among the use of different drugs (Earleywine & Newcomb, 1997). Several potentially dangerous ac- tions also correlate with drug use (Jessor, 1998). Miller (1994) analyzed data from four national surveys including more than 6,000 participants and found that problem behavior theory may account for the appearance of stage-like progressions in substance use. He found a large subgroup who used many drugs, and another set of people who used no drugs at all. These results are consistent with the idea that an underlying “problem-proneness” may account for links between marijuana and other drugs. Essentially, cannabis does not cause cocaine consumption, but a subset of people who like marijuana also like cocaine.
In addition to statistical independence and problem behavior theory, a third set of findings also supports arguments against marijuana as an isolated cause of hard drug use. Studies that show personality traits cor- relate with the use of multiple substances may mean that a personality characteristic led to both marijuana use and hard drug use. These data suggest that the same personality traits that can lead to smoking cannabis can also lead to snorting cocaine. Thus, the marijuana may not cause the use of the other drugs; both stem from the same underlying character- istic. Although evidence for an addictive personality is clearly limited (Nathan, 1988), people who report strong desires for thrill, adventures, and sensations often use a greater variety and amount of drugs (Simon, Stacy, Sussman, & Dent, 1994).
These findings support the idea that marijuana cannot be isolated as the cause of the use of hard drugs. Simple exposure to cannabis is not strongly associated with the use of other intoxicants. Hard drug users do not always use marijuana first. Causes other than marijuana also lead to the consumption of heroin or cocaine. Nevertheless, some authors argue that marijuana may still contribute to the use of harder drugs, even if it is not a unitary cause. They assert that even if cannabis does not qualify as the cause of other drug problems, it facilitates the use of more sub- stances, increasing the likelihood of trouble.