Abstract
An empirical analysis is conducted on two panels of 18 Organisation for Economic Cooperation and Development (OECD) countries to test whether the elasticity of hourly productivity to working time is negative and decreasing with working time itself. If so, the decreasing returns on working time could be indicative of a fatigue effect that increases with working time. We find that the elasticity of productivity per hour to working time is negative and decreases with working time, but its coefficient is not strongly significant. This study offers empirical support for the hypothesis of a fatigue effect that increases with working time, but with some reservations.
Acknowledgement
The views expressed in this analysis are those of the authors and do not reflect the position of the institutions in which they are employed.
Notes
1In this article, Δ before a variable means a difference of the first order for the short data set and an average annual change for the long data set and variables in lower case correspond to their logs.
2For the short data set, gdpcap–1 is the value of the previous year. For the long data set, it is the value from the beginning of the current sub-period.
5 We do not have a sufficient number of observations in the long data set to use a threshold below 1825 hours or above 2025 hours. For consistency, we use the same thresholds on both datasets.
6The first-stage estimation results of this IV estimate can be asked to the authors.