A Dynamic Model of Weekly Sleep Timing: Endogenous Origins of Social Jetlag

Social jetlag is commonly attributed to externally imposed schedules that conflict with biological sleep timing. However, individuals also make daily decisions about how to allocate time between sleep and waking activities, suggesting that behavioral tradeoffs may contribute to within-week variation in sleep patterns.

To examine whether weekly patterns in sleep timing and duration, including social jetlag, can emerge endogenously from the interaction of biological sleep regulation and time allocation decisions.

We develop a dynamic model of sleep timing that integrates the two-process model of sleep regulation with decision-making over time use. Individuals choose sleep onset and offset each day while a stock of sleep evolves over time, determining sleep pressure. The model is calibrated to a steady-state schedule and simulated under alternative workweek structures, with and without preferences for leisure and direct, external constraints on the choice variables, such as fixed wake times and delayed sleep onset.

The model produces a stable weekly pattern in which individuals sleep less during the workweek and sleep longer on free days. Weekend sleep is also later, generating a workday-free-day shift in midsleep similar to the pattern commonly described as social jetlag. These patterns can arise even without imposing fixed wake times or required bedtimes, suggesting that the scarcity of discretionary time on workdays may contribute to delayed sleep and weekend recovery sleep. Adding realistic constraints, such as early weekday wake times and later weekend bedtimes, increases the size of the midsleep shift and brings the simulated pattern closer to empirical estimates of social jetlag.

The model suggests that behavioral tradeoffs over scarce workday leisure can account for part of the weekday-free-day variation in sleep timing and duration, producing shorter workday sleep, weekend recovery sleep, and later free-day midsleep. Social jetlag may therefore reflect both externally imposed schedules and endogenous responses to the changing value of waking time across the week.