One of the major contributions of the literature on human capital has been to demonstrate the role of schooling as a determinant of wages. However, the exact size of the effect on wages and how it should be estimated remain a source of both theoretical and empirical debate. This paper seeks to provide empirical evidence on the returns to education for the Spanish labor market.
The paper uses the instrumental variable approach proposed by Hausman and Taylor to assess the direction and size of the bias that affects standard OLS estimation, when some of the wage determinants are endogenous.
The results suggest that the returns to schooling are substantially higher once endogeneity is taken into account, rising from around 6 percent for each additional year of schooling to about 12 percent when the effect is estimated via instrumental variables. This is in line with research for other countries. Evidence is also found of endogeneity in the effect of job seniority, although in the opposite direction. That is, the Hausman and Taylor model finds little effect of seniority on wages, whereas OLS estimates suggest a larger effect.
Further research is needed to reconcile these results (downwards bias in the OLS estimate of the returns to schooling) with the theoretical notion of an upwards bias caused by the correlation between unobservable ability, wages and schooling.
The paper adds new evidence on wage determinants for the Spanish labor market and does so for men and women using alternative estimation procedures.
