Abstract
This paper examines the relationship between young men's hours worked and their use of marijuana, alcohol, cigarettes, cocaine, and other drugs using cross-section data from the 1991 National Household Survey on Drug Abuse (NHSDA), a nationally representative survey of the U.S. noninstitutionalized population age 12 and over. Our results indicate that substance use has little effect on the number of hours worked by young men in the past month, with the exception that young men who smoked 1 to 3 marijuana joints in the last month worked 42
more hours than nonusers. To assess the robustness of our 1991 results, we re-estimated identical models using data from the 1992 NHSDA, an independent cross-section that was collected using the same methodology as the 1991 survey. Comparing the 1991 and 1992 results, the 1992 data also show that substance use has little relationship overall to the number of hours worked. However, in contrast to the 1991 results, the 1992 results show that smoking 1 to 3 marijuana joints in the last month is associated with 41
fewer hours worked than nonusers. This paper is the first study to assess the robustness of drug use–labor supply results on adjacent cross sections. Our results demonstrate the value of re-estimating the drug use–labor supply relationship.