BENEFIT-COST-ANALYSIS OF DRUG-ABUSE PREVENTION PROGRAMS - A MACROSCOPIC APPROACH

To date, benefit-cost analysis has rarely been used to justify the dru g abuse prevention field. However, there is an increasing demand for t his type of analysis as the field of substance abuse prevention enters a new phase-a phase characterized by a competitive marketplace, an in creased demand for accountability, and the desire to measure return on the money invested in prevention. In response, an effort is made to s timulate discussion and further research on the topic. This article fi rst determines the overall strategy for initiating benefit-cost analys is (BCA), followed by definitions of BCA and cost-effectiveness analys is (CEA). This is followed by the determination of some of the major v ariables used in BCA along with the algorithm for determining the bene fit-cost efficiency ratio (R) as it applies to the macro level analysi s. In estimating a value for the R, a decision has been made to incorp orate uncertainity into the BCA. In a macroscopic approach to BCA, fou r independent variables are identified for computing R. These independ ent and dependent variables are assumed to be random variables with no rmal distributions. The population means and standard deviations perta ining to these independent variables are estimated from the existing l iterature. In order to incorporate uncertainity into the computation o f R, ten measurements have been randomly selected for each of the four independent variables. Following this procedure, fifteen benefit-cost efficiency ratios are calculated by selecting one of the ten values a t random per variable used in the R equation. In this way, we were abl e to determine the most likely population benefit-cost efficiency rati o of 15:1, indicating that there is a $15 savings on every dollar spen t on drug abuse prevention. The 95 percent confidence level pertaining to the R has an interval from $13.7 to $16.1. This indicates that the population R resides within the range 95 percent of the time.