Duke 2008 | ASHE | Equity and Efficiency in Health and Healthcare

Nearly a third of a person's life is spent sleeping. In spite of the amount of time spent in slumber and the potential for sleep to affect a person's ability to function during the day, sleep has received little attention in economics. Research from other disciplines suggests that sleep affects productivity and the quality of life in general (Webb, 1993). Sleep patterns appear to have a genetic component, but individual choice also plays an important role (Webb, 1993; Benca, 1993; Partinen et al, 1983). Our research uses a 'rational-choice' framework to explore the interplay between individual choice and the body's innate need for sleep.

Existing models of sleep rely on a biological model of human behavior, ignoring that individuals have some control over the timing of sleep, and to a more limited degree, the 'typical' amount of sleep they consume. This choice aspect of sleep is often on display in stress situations such as students 'cramming' for exams, medical residents working extended shifts, or individuals rushing to meet a work deadline. Adjusting to 'jet-lag' is another common experience where choice comes into play. These intertemporal choices have fueled a large market for sleep-inducing and sleep-preventing pharmaceutical products.

Previous economic research on the topic of sleep has used very simple, static models of choice (Biddle and Hamermesh (1990)). Our research uses computer simulation to solve a dynamic model of intertemporal sleep choice. Previous mathematical models of sleep originate in biological and medical disciplines and do not account for the ability people have to choose, to some degree, the quantity and quality of sleep. We use our framework to explore issues of worker productivity, the development of human capital over time, and the use of sleep-aid and sleep-preventing products. Our computer simulation model follows the pattern used in Cardon and Showalter (2007) which examines a complex dynamic model of health expenditures calibrated to match observable behavior. We calibrate our model using the American Time Use Survey.

This research project serves two purposes. First, it provides a deeper theoretical understanding of an individual's decisions about how much time to devote to sleep. Second, the theoretical model will guide future empirical work about how sleep affects important outcomes such as wage growth, health, and educational performance. Taken together, these lines of inquiry will go a long ways towards filling the sizeable gap in our knowledge about something that we spend much time on, but about which we know little.