Using data on marijuana and cocaine use from the 1984 National Longitudinal Survey of Youth, the authors examine the hypothesis that drug use reduces labor market productivity, as measured by wages. From an analysis that controls for the probability of employment and the endogeneity of drug use, they find that although long-term and on-the-job use of marijuana negatively affected wages, the net productivity effect for all marijuana users (both those who engaged in long-term or on-the-job use and those who did not) was positive. No statistically significant association was found between cocaine use and productivity.
Labor Market Effects of Urban Riots: An Experimental Assessment