The Returns to Formal Schooling in Turkey using Pseudo-Panel Data

The returns to education is of great interest in public policies and labor economics. It has been widely studied by many experts for decades. It has been acknowledged that in a Mincerian wage equation, ordinary least squares estimates are biased due to the endogeneity of education. One way to deal with this endogeneity could be removing individual fixed effects using panel data. However, education of an individual is fixed once their wage is observed. Thus, panel data would wipe out the information on education. On the other hand, using pseudo-panel approach the returns to education can be estimated making use of repeated cross section data. There are limited studies on returns to education on Turkey despite the fact that its crucial importance in the public policy. In this paper we estimate returns to education for Turkey using a pseudo-panel data approach. We make use of Turkish Household Labor Force data for the years 2009-2014 in order to construct pseudo-panel data. We find that one additional year of education increases individual wages by around 8.5 percent using ordinary least squares. However, using pseudo-panel fixed effect estimation leads a 9.3-percent rate of returns to education. Thus, we show that there is a downward bias in ordinary least squares estimates of returns to education in Turkey