Testing labour supply and hours constraints

This paper provides empirical evidence that, at a given wage, individuals cannot freely choose the number of hours they work. The novelty relative to the existing literature (e.g. [Altonji, J., Paxson, C., 1986. Job characteristics and hours of work. In: Ehrenberg, R. (Ed.), Research in Labor Economics, vol. 8. Westview Press, Greenwich, 1-55]) is twofold. We use the US data on prime age males from the National Longitudinal Survey of Youth and we account for endogenous switching between jobs. Our results are: (i) the variance of the change in hours worked is more than six times higher for movers than for stayers; (ii) the intertemporal labour supply elasticity is positive and significant for movers and zero for stayers. This is further evidence for the presence of hours constraints. One important implication is that estimates of the labour supply elasticity that ignore these constraints are biased.