Sleep Time: Media Hype vs. Diary Data

Abstract

Sleep duration has figured into claims of two trends promoted recently as dysfunctional in the mass media. One is the observation that the population at large is sleeping less than before. The second is that the annual change from Standard Time to Daylight Savings (or summer) Time causes adverse effects, largely through the loss of an hour’s sleep. This paper relies on recent Canadian and U.S. time-use data to empirically test both of these value-laden allegations. Analysis of Statistics Canada’s general social surveys containing time-use information in 1986, 1992, 1998, and 2005 shows that the mean duration of sleep was unchanged between 1986 and 1998 and actually declined by about 15 min a night in 2005, reflecting an earlier bedtime and unchanged arising time. Sleep duration is not constant in the population, though, and the media view might reflect the habits of population sectors such as the intelligence with great access to the media. The American Time Use Study sample of 20,720 respondents in 2003 enabled the analysis of time-use before, on, and after the dates of semi-annual time changes that year. These data showed that any sleep time “lost” in the spring-forward time change was insignificant and short-lived due to the fact that it occurred on the night between Saturday and Sunday, when people typically sleep much longer than on weekdays. While there are other time trade-offs observed after time changes, their explanation lies in other directions.