No Rest for the Weary: Commuting, Hours Worked, and Sleep

This paper is the first to combine data from large nationwide surveys to investigate how commuting and work hours affect sleep. I estimate that 11-21\% of the marginal unit of time spent working and 22-30\% of the marginal unit of time spent commuting replace sleep. Controlling for these effects, commuting before 5 a.m. and after 9 a.m. each increase the likelihood of short sleep. I also find that time spent commuting and working and the prevalence of these strange commute times each contribute to unintentionally falling asleep at some time during the day, while early commuting in particular increases the likelihood of falling asleep while driving. Little of these effects are explained by reduced time spent sleeping, indicating that there are multiple biological channels through which commuting duration and timing impact road safety. None of these effects appear for non-workers as opposed to the employed, supporting the validity of the results. Overall, most of the effects are stronger for women than for men, though the prevalence of early commutes is particularly associated with less sleep among men.