Sleep, health, and human capital: Evidence from daylight saving time

Chronic sleep deprivation is a significant and understudied public health issue. Using BRFSS survey data from the United States and an administrative census of 160 million hospital admissions from Germany, we study the causal relationship between sleep and health. Our empirical approach exploits the end of Daylight Saving Time in a quasi-experimental setting on a daily basis. First, we show that setting clocks back by one hour in the middle of the night significantly extends people's sleep duration. In addition, we find significant health benefits via sharp reductions in hospital admissions. For example, hospitalizations due to cardiovascular diseases decrease by ten per day, per one million population. Using an event study approach, we find that the effect persists for four days after the time shift. Admissions due to heart attacks and injuries also exhibit the same characteristic four-day decrease. We also provide a series of checks to rule out alternative, non-sleep related, mechanisms. Finally, we discuss the benefits of additional sleep for the sleep-deprived as well as policy implications for nudging people to sleep more. Our findings illustrate the importance of public policies that target sleep deprivation.