College Students' Use of Cocaine

Abstract

After experiencing a period of rapid decline between 1986 and 1994, cocaine use is once again on the rise in the United States. The increased prevalence of use among college students is particularly troubling because of its potential impact on human capital acquisition and long-term labor market success. Merging information on the price of cocaine and marijuana from the U.S. Drug Enforcement Agency with data on cocaine use from the College Alcohol Study, we investigate the demand for cocaine in the college population. We find evidence that participation in cocaine use by college students is responsive to changes in the price of cocaine and marijuana and that cocaine and marijuana are economic complements for this population. Further investigation revealed significant differences in the demand for cocaine by those less than age 21 and those at least age 21, years, with the younger age group being more responsive to changes in the price of cocaine. No difference is found, however, in the demand for cocaine across gender.

Keywords:

• cocaine
• demand
• elasticity
• human capital
• marijuana

Notes

¹There is some evidence that alcohol use negatively impacts on college students' grade point average by reducing the hours they spend studying (Williams et al., 2003; Wolaver, 2002).

²Deficiencies of the STRIDE data are discussed by Horowitz (2001) and Manski et al. (2001) but include the fact that they are a convenience sample resulting from purposeful policing activities rather than a random sample of drug transactions. The timing and location of encounters are not only unrepresentative, they are erratic. The number of observations from a given location can vary dramatically from year to year.

³As he notes, however, that because of the nature of his sample (individual's who start using cocaine after the 1984 survey, when they are aged 19–27), his results may not generalize to younger cohorts, where the uptake of cocaine use is concentrated.

⁴DeSimone and Farrelly (2003) is the exception.

⁵Authors' own calculations. In 1985 there were fewer than 100 crack cocaine observations in the cleaned data file. By 1990, there were 2,094 and in 1992 there were 2,578. Crack observations have continued to exceed powder cocaine observations in number since 1992.

⁶In practice, this requires that we restrict our use of the STRIDE data to only those cities for which we had at least 40 observations for powder cocaine, which reduces the number of potential cocaine observations by 25% and the number of cities represented in the data by over 90%. Further restrictions on the number of observations did not dramatically influence the price series generated here.

⁷For each wave of the CAS, questionnaires were initially mailed to students beginning at the end of February. Mailings were timed to avoid the period immediately before and after spring break, so that students would be responding to behavior during a time when they were on campus. Student participation was voluntary and anonymous.

⁸The annual prevalence of crack cocaine use is 0.7% in the CAS, and monthly prevalence is less than 0.2%.

⁹The cities for which quarterly prices are published are Atlanta, Georgia; Boston, Massachusetts; Chicago, Illinois; Dallas, Texas; Denver, Colorado; Detroit, Michigan; Houston, Texas; Los Angeles, California; Miami, Florida; Newark, New Jersey; New Orleans, Louisiana; New York, New York; Philadelphia, Pennsylvania; Phoenix, Arizona; San Diego, California; San Francisco, California; Seattle, Washington; St. Louis, Missouri; and the District of Columbia. Sinsemilla marijuana is of a higher quality (more potent) than commercial-grade marijuana.

¹⁰The retail price is clearly the price that would be most relevant to young adults' consumption decisions. The choice to focus on commercial-grade marijuana rather than sinsemilla grade has to do with fact that there were fewer missing observations for these data. In addition, there is some evidence that commercial-grade marijuana dominated the U.S. marijuana market during this period (National Narcotics Intelligence Consumers Committee, 1998).

¹¹Tragler et al. (2001) show that there are national cycles to drug epidemics, particularly cocaine, suggesting that an independent time trend in use rates might exist.