Essays in the Economics of Sleep and Genoeconomics

This dissertation examines how sleep-related behaviours and biological traits influence educational and labour market outcomes. Although essential for survival, only recently research has begun to explore sleep determinants and effects across multiple life domains, and knowledge about its causal impacts is still scarce. Even less is known about the implications of biological variation in sleep-related traits, as individual circadian differences, within the social context. This thesis contributes to fill these gaps. It does so by applying causal inference techniques to analyse time-diary and genetic data from representative samples of the US population, in three complementary chapters. The first chapter studies the effect of sleep duration on labour productivity. The binding 24-hours time budget constraint complicates the identification of its causal effects, since any increase in sleep necessarily implies a simultaneous, equivalent reduction in substitute activities. The net impact of sleep crucially depends on the activities that it replaces, and no meaningful interpretation can be given to unconditional effects. This, beyond the fact that sleep is under individual choice. The first chapter, co-authored with Marco Bertoni and Lorenzo Rocco, highlights this issue and addresses both challenges by estimating the impact of sleep relative to non-work activities on labour productivity among a representative sample of full-time employees. Using detailed time-diary data in a value-added framework, the analysis reveals that, once working hours are held constant, substituting sleep for other non-work activities does not significantly affect productivity. This finding underscores the importance of explicitly accounting for substitution patterns when evaluating the effects of sleep – but also of any other time use on any outcome. The second chapter moves the focus to timing. This solo-authored chapter examines chronotype, an expression of individual biological clock, the innate predisposition toward earlier or later sleep-wake cycle. In societies regulated by rigid schedules, this biological heterogeneity implies that some individuals are inevitably required to perform at hours misaligned with their natural rhythms, potentially achieving lower efficiency compared to similar peers whose biological clocks are better synchronised with social ones. I explore this phenomenon in the school context, comparing students from the same high school and cohort, as well as full siblings from the same family, and exploiting genetic variation in their “genetic predisposition" for chronotype - measured with a polygenic index (PGI) for this trait. Results indicate that, all else being equal, students with a more morning-oriented chronotype significantly achieve higher grades than their evening-oriented peers. This effect stems from closer alignment between biological and social rhythms, enhancing efficiency. The third chapter is motivated by the fact that, in today's round-the-clock society, non-standard arrangements such as irregular or night shifts are becoming increasingly widespread but are largely considered harmful due to a disruption of workers' circadian rhythms. However, individual heterogeneity in chronotype may induce workers to self-select into schedules better aligned with their natural biological rhythms, thereby mitigating risks. Using a DNA-based measure of chronotype, the third chapter explicitly tests this hypothesis but finds no evidence in its support. Therefore, at least at the beginning of their careers, individuals either lack awareness of their biological needs or opportunity to make biologically optimal occupational decisions, remaining exposed to the risk of circadian misalignment. Results also reveal a positive relationship between morningness and job satisfaction, plausibly arising from the synchrony between biological and social clocks.