The following HTML text is provided to enhance online
readability. Many aspects of typography translate only awkwardly to HTML.
Please use the page image
as the authoritative form to ensure accuracy.
Informing America’s Policy on Illegal Drugs: What We Don’t Know Keeps Hurting Us
context of detailed questions about drug use may distort recall and reporting.
Sources of data on drug use consequences are not well suited for supporting causal inferences. Indeed, the phrase “drug use consequences” is potentially misleading, as many apparent consequences may actually be spurious correlations. Relative to nonusers, heavy drug users are known to be disproportionately impulsive, less educated, less likely to be employed, more criminally involved, and less healthy. Drug use can cause or augment these characteristics, but it can also be a consequence of one or more of these characteristics. Isolating the causal role of drug use is a difficult challenge; there is a small body of laboratory experiments with adult volunteers, but this type of research is ethically and methodologically constrained.
These limitations for drawing causal inferences mean that existing data provide a fragile and incomplete foundation for recent efforts to estimate the aggregate consequences of U.S. drug use (Harwood et al., 1998; Rice, 1999). The inadequacy of these estimates has been documented by Cohen (1999), Kleiman (1999) and Reuter (1999). Moreover, these aggregated estimates are of limited value for policy analysis. They may facilitate budgetary planning and serve a rhetorical role in mobilizing public support for drug policy, but they provide little insight into the dynamics of the drug problem or its responsiveness to alternative strategies and tactics of policy intervention.
Some alternate methodologies for improving understanding of drug use consequences are available. One underutilized approach for understanding the relationship between drug use and its consequences is dose-response analysis—a standard methodology in pharmacology (Julien, 1998), epidemiology (Lilienfeld and Stolley, 1994), and technological risk analysis (Morgan, 1981). In a simple dose-response analysis, the strength or intensity of a given type of response (e.g., a physical symptom) is plotted (on the vertical axis) as a function of increasing dose (on the horizontal axis). Alternatively, the vertical axis depicts the percentage of subjects (e.g., laboratory animals, human participants) displaying the response in question. Dose-response curves show that organismic responses to many biological or technological stimuli have an S-shape, with relatively little response at very low doses, a steep rise in response probability or intensity, and an eventual plateau. These S-shaped curves graphically depict three key concepts (Julien, 1998:33). The potency of the stimulus is shown by the location of the curve on the horizontal (dose) axis; more potent stimuli are shifted toward the left end of the axis, so that smaller