logo SBA

ETD

Digital archive of theses discussed at the University of Pisa

 

Thesis etd-12092015-192549


Thesis type
Tesi di dottorato di ricerca
Author
GABBRIELLINI, CECILIA
URN
etd-12092015-192549
Thesis title
The Determinants of the Italian Wages
Academic discipline
SECS-P/01
Course of study
ECONOMIA POLITICA
Supervisors
tutor Prof. Fiaschi, Davide
Keywords
  • gender wage gap
  • generalized additive model
  • returns to education
  • wage function
Graduation session start date
06/01/2016
Availability
Full
Summary
The aim of this work is to estimate the determinants of the wage function in Italy, focusing on the crucial role of education, taking into consideration even the impact of years of experience (training on the job and learning by doing activities), controlling for individual characteristics, and sectorial and geographical variables.
The benchmark model for the empirical estimation of the returns to education is the relationship derived by Mincer (1974) between (log) hourly wages, schooling and experience. In all the chapters, the empirical analysis is carried out using a representative sample of Italian households, come from the Bank of Italy’s Survey of Household Income and Wealth (SHIW), for the period 1995-2012.

In the first chapter, in line with previous literature, we find that ordinary least squares (OLS) under-estimates the return to schooling. Considering the endogeneity of schooling, the return to an additional year in school increases. In addition, in the period considered, the findings show that the returns to schooling have changed from 5.4 percent to 7.9 percent. The highest level is recorded in 2006 and the lowest in 2012 thus the advantage to invest in education is decreasing in Italy. Moreover, a relative convenience to work in the public sector emerges as well as an evidence of a gender pay gap, in favor of men for all the period considered.

In the second chapter, we remove the hypothesis of homogeneity of the return to education and we estimate the wage-schooling and wage-experience profile to take into account all the shape of these variables. In addition, to analyze the effect of endogeneity on the non-monotonicity of the marginal rate of return to education, we use a control function approach for a semiparametric estimation. Results show that the wage−schooling relationship is non-linear. This implies that returns to education depend on the level of schooling. In particular, increasing returns are evident for workers with almost 8 years of schooling (junior high school). If we consider workers with almost 13 years of schooling (secondary school), the marginal effects across year continue to increase. On the other hand, decreasing returns are observed for workers with 18 years of schooling (tertiary education), from 2008 to 2012.

Considering the heterogeneity of education is relevant because allows to test if education can reduce or increase inequality. In order to take into account simultaneously endogeneity and heterogeneity of education, an instrumental variables quantile regression is applied in the third chapter. Our results show that, while returns to schooling are positive along wage distribution, a large degree of heterogeneity emerges for returns to education. In particular, gains are higher for individuals in the upper tail of the wages distribution than for those in the lower tail. This means that education have an inequality-increasing effect over time. Indeed, individuals with higher ability are in upper quantile of the wage distribution and seem to benefit more from formal education.
File