The cost of employment protection in the presence of match - specific learning 1

This paper studies the effects of the introduction of employment protection policies that raise the costs of dismissal. Macroeconomists studying these policies have thus far focused primarily on their effect on the flexibility of firms in adjusting their employment. This study highlights that in the presence of worker heterogeneity and match-specific learning, these policies have an important effect on average productivity by influencing productivity as a function of tenure and the tenure distribution of employed workers. Moreover, this productivity effect is different depending on the nature of the learning process that is the source of productivity gains with tenure, whether it entails learning-by-doing or learning about match quality. Using structural estimates from a previous study that imply that learning about match quality is the dominant learning process, I show that by hindering experimentation in the economy, the imposition of dismissal costs reduces average productivity by 3.2%. Moreover, I show that even if the first-best policy of removing these dismissal costs completely is not available, removing dismissal costs for low-tenure workers through the liberalization of the use of fixed-term contracts can undo most of the above productivity loss. I would like to thank the participants at the Stanford macro lunch, the Society of Economics Dynamics 2001 Conference, the NBER 2001 Summer Institute and the 2001 Summer Workshop of the Hungarian Institute for Economic Research and Richard Rogerson, in particular, for helpful comments. All errors are mine. Department of Economics, Northwestern University, 2003 Sheridan Road, Evanston, IL 60208-2600, USA and Institute for International Economic Studies, Stockholm University, 10691 Stockholm, Sweden. E-mail: [email protected]