The diurnal pattern of on-the-job injuries

Kenneth N. Fortson is a Ph.D. candidate, Department of Economics, Princeton University, Princeton, NJ. Shortly after 4:00 A.M. on March 28, 1979, mechanical equipment at the nuclear power plant at Three Mile Island, near Harrisburg, Pennsylvania, malfunctioned. In the course of responding to the emergency, operators working the late-night shift made errors that exacerbated the situation, resulting in the worst accident in the short history of U.S. commercial nuclear power.1 Seven years later and halfway around the globe, at 1:23 A.M. on April 26, 1986, negligence by night shift workers at the nuclear reactor in Chernobyl, U.S.S.R., led to an even more catastrophic nuclear disaster.2 In the popular press, it is often asserted, without explanation, that workplace injuries are more common at night.3 In the academic literature, economists have largely ignored the diurnal pattern of on-the-job injuries and, by extension, the economic ramifications. This article uses data on workers’ compensation claims from Texas to estimate the empirical distribution of injuries. The results show that the injury rate is high during off-hours late at night and low during the regular nine-to-five shift. The article also decomposes the factors causing the observed injury pattern and explores the possibility that the empirical injury cycle is merely an artifact of compositional changes in the age or industry and occupation of workers throughout the day. Late-night workers have longer shifts as well, so fatigue is examined as a possible explanation of the injury pattern. Both of these possibilities, however, are rejected as the lone explanation of the injury pattern. Instead, the article argues that there are inherent physiological implications of late-night work that make off-hours jobs more hazardous than daytime jobs. This is an important distinction because it suggests that, in scheduling work hours, firms should consider shift time in addition to factors such as shift length, which is merely correlated with late-night work and contributes to a higher injury rate, but is not unique to night work.