Wage Returns to Experience and Tenure for Young Men in Italy

Published date01 November 2014
AuthorGiovanni Sulis
DOIhttp://doi.org/10.1111/sjpe.12058
Date01 November 2014
WAGE RETURNS TO EXPERIENCE AND
TENURE FOR YOUNG MEN IN ITALY
Giovanni Sulis*
ABSTRACT
This paper provides estimates of wage returns to experience-, firm-, sector- and
occupation-specific tenure for a sample of young Italian male workers. By com-
paring returns obtained using different estimators, I evaluate the importance of
endogeneity and selection problems generated by specific unobserved components
and individual fixed effects. After controlling for the role of collective bargaining
agreements and occupation categories, results indicate that general labour mar-
ket experience is the fundamental source of wage growth for blue and white col-
lars, while returns to firm tenure are insignificant. There is some evidence of
positive returns to sector and occupational tenure for white collars. Estimates
from different sectors suggest that union coverage can be relevant in offsetting
the role of search and matching in wage determination.
II
NTRODUCTION
The analysis of labour market transitions, employment duration and wage
growth has received renewed attention in recent years. In particular, this line
of research investigates the effect of labour market experience and specific ten-
ure on wages. This is not surprising, as correctly estimating true returns to
skills accumulated in the labour market is fundamental for the analysis of
wage dispersion and its evolution over time.
1
Furthermore, different theoreti-
cal models answer the question of why wages grow as experience and tenure
accumulate, resulting in the concave profile observed in the data. Human cap-
ital accumulation as individual investment in job-specific skills (Becker, 1964),
theories of deferred compensation as incentive mechanism (Lazear, 1981), and
search and matching models (Burdett, 1978; Jovanovic, 1979) are all able to
explain these empirical stylized facts.
Studying the returns to general and specific skills in the labour market
therefore turns out to be important for a number of reasons. First, it gives a
*Dipartimento di Scienze Economiche e Aziendali, University of Cagliari
1
The first wave of studies were focused on the United States, see Mincer and Jovanovic
(1981), Abraham and Farber (1987), Altonji and Shakotko (1987), and Topel (1991). Nowa-
days, recent contributions propose methodological advances (Buchinsky et al., 2010) or offer
evidence on European countries, or both: see Altonji and Williams (2005), Dustmann and
Meghir (2005) and Beffy et al. (2006).
Scottish Journal of Political Economy, DOI: 10.1111/sjpe.12058, Vol. 61, No. 5, November 2014
©2014 Scottish Economic Society.
559
picture of the overall individual wage growth over the life cycle, with implica-
tions for individual well-being and for the wage distribution (Farber, 1999). It
is also a good discriminating test of human capital and search models and
their implications in terms of accumulation (and loss) of human and search
capital. Finally, the slope of wage profiles varies across countries with differ-
ent institutional and wage bargaining environments, and this has various
implications for labour mobility and rent sharing (Teulings and Hartog,
1998).
However, correctly estimating wage returns to skills is not an easy task.
Standard OLS estimates are biased due to endogeneity and omitted variable
problems, and the direction of such bias is ex ante ambiguous. Hence, a large
literature has developed proposing different estimators to solve such endoge-
neity problems and/or including in the analysis previously omitted variables.
2
Altonji and Shakotko (1987) instrument firm tenure using deviations of
tenure from the job-match mean, thus controlling for both unobserved indi-
vidual- and match-specific effects, which are correlated with tenure and experi-
ence. In turn, the two-step estimation methodology proposed by Topel (1991)
uses an identification strategy partially based on job changes, as movers have
zero tenure and different levels of experience. This technique is not beyond
criticism, as job movers are selected in new jobs on the basis of their past
choices, thus not entirely solving the endogeneity concern. To overcome that
problem, Dustmann and Meghir (2005) identify returns to experience and ten-
ure using a sample of displaced workers. The intuition behind their identifica-
tion strategy is that displaced workers can be considered a random sample of
the population, as their status is not conditioned by their past choices, but it
is assumed to be substantially exogenous.
Neal (1995) also uses a sample of displaced workers as a source of identifi-
cation in the estimation of returns to industry-specific tenure, a variable that
was previously omitted in standard wage equations. In his paper, exogenous
displacement allows him to compare workers with the same level of firm ten-
ure, but different levels of sector-specific skills. Similarly, by using deviations
from the individual and job means as instruments for experience, firm- and
sector-tenure, Parent (2000) confirms that considering skills that are neither
firm-specific nor completely general strongly reduces wage returns to experi-
ence and tenure.
More recently, a new strand of literature includes occupational tenure in
wage regressions, thus explicitly considering the skills needed to perform tasks
in a particular occupation, which are transferable across jobs and sectors.
Kambourov and Manovskii (2009) and Zangelidis (2008a) convincingly show
that, after including occupational tenure, wage returns to firm and sector ten-
ure lose their capacity to explain wage growth. These results hold in decentral-
ized labour markets, characterized by high labour mobility and large returns
to specific skills, such as in the United States and the United Kingdom.
2
Needless to say, no estimator is completely unbiased, hence most of this literature focuses
on comparison of different estimators or comparison of the same estimator in different
contexts, or both.
560 GIOVANNI SULIS
Scottish Journal of Political Economy
©2014 Scottish Economic Society

To continue reading

Request your trial

VLEX uses login cookies to provide you with a better browsing experience. If you click on 'Accept' or continue browsing this site we consider that you accept our cookie policy. ACCEPT