Earnings effects of job and educational mismatch in the Czech graduate labour market

The paper provides the estimates of the impact of job and educational mismatches on earnings of graduates from the University of Economics, Prague. The dataset was obtained from the cross-sections REFLEX survey that gathered information about the labour market status of the graduates within the period 2008-2012. The earnings equations were estimated controlling the source of sample selection bias when using Heckman maximum likelihood procedure with the main Mincerian earnings equation, while the selection equation is a probit estimate of the probability to be employed rather than unemployed. We tested the hypothesis that educational and job mismatched workers earn less due to their lower competencies and skills in relative terms. We estimated earnings penalties associated to job and educational mismatches controlling for sample selection. The earnings penalty associated to educational / job mismatch is equal to 32.2 % and. 8.2 %, respectively. This finding supports the hypothesis there is positive selection into employment of the most skilled among workers whose individual characteristics are less on demand on the labour market.