Project Description
Modeling the Demand for Cocaine
PI: Susan Everingham
Funded by: Office of National Drug Control Policy, U.S. Army, The Ford
Foundation
In a ground-breaking study, RAND characterized the market for a key illicit drug--cocaine--and drew implications for policy. To develop a more precise and quantitative understanding of cocaine trends, researchers created a model simulating changes in cocaine demand, where demand encompassed both the number of cocaine users and how much they consumed. Demand for cocaine was determined by constructing a two-state Markovian model that specified the flow of individuals into and out of two states: light cocaine use and heavy cocaine use. (People who used cocaine weekly or several times a month were defined as "heavy" users; all others were defined as "light" users.) The model was fitted to 20 years of historical data on cocaine use, which were derived from the National Household Survey of Drug Abuse (NHSDA) and other sources. The model predicts future prevalence (i.e., the total number of users) when given a scenario for incidence (i.e., the frequency with which non-users flow into cocaine use). By applying this model, RAND researchers were able to explain why cocaine consumption held constant during the 1980s as the number of new users dropped.
Related Publications:
Everingham, Susan S., and C. Peter Rydell, Modeling the Demand for Cocaine, RAND, MR-332-ONDCP/A/DPRC, 1994. (Summary available in Projecting Future Cocaine Use and Evaluating Control Strategies, RAND Drug Policy Research Center Research Brief, RB-6002, January 1995.)
Top