Drug Testing in the Trucking Industry: the Effect on Highway Safety*

This paper uses a set of “natural experiments,” created by the passage of a U.S. Department of Transportation drug-testing mandate and 13 state testing laws, to examine the effects of testing truckers for illicit substances on highway safety. Since truckers do not bear the full costs of their driving and employers cannot contract on all aspects of their behavior, drug testing may be one means for companies to either screen or monitor employees and lower expected accident costs. Indeed, I find that testing led to a 9–10 percent reduction in truck accident fatalities. The social benefits of mandated testing appear to outweigh the costs of the program. However, the similarity between the effect of mandating testing and simply clarifying state law suggests that extending the right to perform drug tests may have been as effective at lower cost. In the 1980s, in an effort to step up its “War on Drugs,” the federal government began a widespread campaign to promote workplace drug testing. Trucking, and the transportation industry more generally, became the natural target of testing advocates. Although trucking companies implemented some safety and productivity measures on their own, such as the adoption of onboard computerized monitoring devices, drug testing was mandated by the federal government. Prior to state legislative and federal regulatory activity in the late 1980s, there was little testing in the industry. Thus, a priori, it is unclear if the benefits of testing outweigh the costs. I assess this issue by using a panel of states to look at the effects of drug testing on truck-involved highway fatalities. I exploit a set of “natural experiments” created by the passage of drug-testing laws by 13 states between 1987 and 1989 as well as the 1990 implementation deadline for a U.S. Department of Transportation (DOT) testing mandate to examine whether * I am extremely grateful to Claudia Goldin and Larry Katz for their suggestions throughout the research process. I also thank William N. Evans, Thomas N. Hubbard, Srikanth Kadiyala, Robert Kaestner, Jeffrey Miron, Sam Peltzman, Charles Rombro, Peter F. Swan, an anonymous referee for the Journal of Law and Economics, and the participants of Harvard University’s labor lunch series for their helpful comments. Financial assistance is gratefully acknowledged from the National Science Foundation’s Graduate Research Fellowship Program and the Lindesmith Center’s Fellowship in Drug Policy Studies. 132 the journal of law and economics drug screening is an effective policy instrument. I find that laws allowing drug testing reduced truck accident fatalities between 9 and 10 percent. These findings are robust to the inclusion of a broad set of traffic safety covariates and to several specification checks. On the basis of standard value-of-life estimates, the benefits of testing in the trucking industry, an industry in which one worker’s poor judgment can have multiplicative negative safety consequences, seem to outweigh the costs of the federally mandated testing program. Despite the fairly large estimated benefits, few firms were testing prior to the federal mandate, apparently because of uncertainty over the legal ramifications of drug testing. Indeed, the similarity between the effect of mandated testing on truck accident fatalities and the state laws clarifying the legal boundaries of testing suggests that extending the right to perform drug tests on employees in safety-sensitive positions may have been equally effective and would have come at a lower cost. I provide some suggestive evidence of this following the presentation of the cost-benefit analysis. While the present study establishes the net benefits of drug testing workers in safety-sensitive positions, it leaves open the question of the value of testing employees who, unlike truckers, do not pose significant safety risks. Why, for example, did nearly half of Fortune 500 companies have some type of drug-testing program by 1993? Do these firms view testing as a cost-effective measure of worker productivity or an easy way to signal compliance with federal drug-free workplace regulations and incentive programs? Although beyond the scope of this paper, drug testing in the broader workforce is an issue that continues to generate considerable debate among civil libertarians on the one hand and advocates of zero-tolerance drug policies on the other. This paper begins with an overview of drug testing. Section II provides a brief summary of related research, and Section III an informal framework for determining how testing may have reduced truck accident fatalities. Section IV uses the time-series variation in testing created by the mandate to look at the effect of testing on aggregate trends in truck accidents. Section V, the core of the paper, exploits the time-series and cross-sectional variation in the likelihood that truckers are screened for drugs, from the combination of state laws and the DOT mandate, to show that testing significantly reduced truck accident fatalities. Section VI presents concluding remarks. I. Overview of Drug Testing The mid-1980s mark a turning point in employee drug testing. In 1986, President Ronald Reagan issued Executive Order 12564, the Drug Free Workplace Program (DFWP), which directed all government agencies to make 1 See Tyler D. Hartwell, Prevalence of Drug Testing in the Workplace, 118 Monthly Lab. Rev. 35 (1996).