Returns to Skills and the Distribution of Wages 543
but the available data suggest that, if anything, earnings and wage inequality continued
In contrast, there is evidence of a fall in wage inequality in some other European
countries during the same period. For instance, Verdugo (2013) shows that in France wage
inequality decreased continuously from 1969 to 2008; Naticchioni and Ricci (2009) ﬁnd
the same result in Italy from 1993 to 2006, and that for the highest percentiles this was
mainly related to a fall in the educational wage premium.
Within this context, the evolutionof wage inequality in Spain seems somehow puzzling.
As seen in Figure 1,1between 1995 and 2002, wages decreased throughout the entire
distribution, except for the highest deciles, and the wage distribution compressed around
a lower mean. Between 2002 and 2006, the wage distribution did not basically change,
and it is only between 2006 and 2010 when wage increases are observed for all deciles,
ranging from 10% to 20% as we move up along the wage distribution. This has also been
documented by Bonhomme and Hospido (2012) who found that wage inequality in Spain
decreased at a rate similar to that of the observed rise in other countries since the early
1990s up to the beginning of the current crisis, and by Casado and Sim´on (2013) who
offered further evidence of an increase in inequality from 2006 to 2010.
In this paper, we use a simple labour supply and demand framework to investigate
several possible explanations for the development of the Spanish wage structure during the
period 1995–2010. We ﬁrst document the importance of distinguishing between changes
in the composition of employment and changes in the returns to speciﬁc workers’and jobs’
characteristics to explain the observed wage evolution.Among the most relevantf actors that
might imply signiﬁcant composition effects on the supply side are the educational upgrade
of the labour force (see Lacuesta, Puente and Cuadrado, 2010) and huge immigration ﬂows
(see Carrasco, Jimeno and Ortega, 2008).2On the demand side, major factors include the
increasing importance of the construction sector, which leads to an increasing share of
low-paid workers (see Gonzalez and Ortega, 2013).3Using the quantile decomposition
technique proposed by Machado and Mata (2005), our results show that it is a decrease in
the returns to education which explains the lack of wage growth and the compression of
the wage distribution during the period 1995–2006.4In contrast, the widening of the wage
distribution after 2006 is largely explained by an increase of the returns to schooling.
1During the period 1995–2006, the Spanish economyexperienced a long and strong expansion with signiﬁcant real
GDP growth (an annual growth rate of approximately 4%), a decreasing unemployment rate (14 percentage points
over the period), and increasing employment creation. However, despite the large increase in labour demand, wage
pressures remained subdued with aggregate real wages decreasing at annual rates of −0.5% and −0.3% in 1995–
2000 and 2000–05, respectively (OECD, 2007).The recession started in 2008, with GDP growth rates of −3.7% and
−0.3% in 2009 and 2010, respectively, has seen a moderate increase in the aggregate real wage of 0.03% in 2009
and a moderate decrease of −0.008% in 2010.
2Another important change over this period is the variation in the gender composition of the labour force. In this
paper, we follow a long-standing tradition in the analysisof the wage structure and focus on male wage distributions
to abstract from fertility decisions which mainly affect female labour supply. Moreover, the increase in the female
labour force participation rate is likely to have changed the selection of women into work,which may have had an
independent impact on the female wage structure.
3Low real interest rates and lax credit conditions, together with some changes in the regulation of urban land,
contributed to a boom in the construction sector (see Arce, Campa and Gavil´an, 2012).
4Izquierdo and Lacuesta (2012) found similar conclusions for Spain using the non-parametric technique proposed
by DiNardo, Fortin and Lemieux (1996).
©2014 The Department of Economics, University of Oxford and JohnWiley & Sons Ltd