We study the determinants of the level and the evolution of per capita hours worked in a panel of OECD countries since the 1970s. Following Pesaran (2006), our empirical strategy allows for the possibility of cross-sectionally correlated error terms due to unobserved common factors, which are potentially nonstationary. We find that much of the variation in per capita hours worked across countries and over time can be explained by differences in the level and structure of taxes and government expenditures. Differences in (the evolution of) labor and product market institutions have much less of a role to play. Our results show that a careful treatment of the time-series properties of the data is crucial.