Labor market flexibility and investment in human capital ∗

We consider the effect of labor market rigidities on human capital accumulation and economic growth when some human capital is difficult to observe prior to employment. We distinguish between two types of human capital, those that formal schooling and test scores can measure (“knowledge”) and those that can be observed by employers only after a period of employment (“creativity”). We build a simple model to show when employers have limited discretion to set wages or terminate employment they favor the more reliable signals of “knowledge”, such as years of schooling and class rank, at the expense of “creativity”, which stands for non-testable skills. Individuals in rigid labor markets will therefore favor greater acquisition of knowledge at the expense of creativity, which results in distorted accumulation of human capital and lower growth. We explore the implications of our model for empirical issues in the literatures in education and growth, and for education policy. We show that accounting for labor market flexibility in cross country growth regressions provides an explanation for the observed low productivity of schooling in economic growth. A key policy implication of our model is that education reform in countries with rigid labor markets is unlikely to succeed unless it is combined with labor market reform that reduces distortions in the incentives for investments in human capital. ∗Please address correspondence to [email protected], Department of Economics, Virginia Tech, Blacksburg, VA 24061. For helpful comments we wish to thank Mine Cinar, Julio Elias, Hadi Esfahani, Lant Pritchett, Lyn Squire, and seminar participants at the University of Illinois, Urbana-Champaign, Virginia Tech, and the 10th ERF Annual Conference, Morocco, 2003. An earlier version of the paper was presented at the Northeastern Universities Development Consortium conference, Cornell University, October 2000. Thanks also to Joel Niamien and Rushad Faridi for able research assistance.