1 Introduction
While the major unemployment differentials between EU counties have attracted a lot of
interest over the years, the issue of substantial disparities in regional unemployment rates
has only been addressed more recently (see, for example, Marston, 1985, Blanchard and
Katz, 1992, Decressin and Fatas, 1995, Jimeno and Bentolila, 1998, Baddeley, Martin and
Tyler, 1998 and Overman and Puga, 2002).
This paper aims to shed light in the dynamics of Spanish regional unemployment rates
and determine the driving forces of their disparities. The Spanish economy has one of the
highest unemployment rates in the EU and is characterised by severe regional disparities
(see Bande et al., 2005, 2006).
Elhorst (2003) argues that the issue of regional unemployment deserves special at-
tention for the following two main reasons. First, the magnitudes of regional disparities
are at least as large as the magnitudes of unemployment differentials among countries
(OECD, 2001). For instance, in 2005, the Southern Spanish region of Extremadura had
an unemployment rate of 16%; in contrast, the richer Northern Spanish region of Navarra
experienced a modest unemployment rate of 5%. Such big differentials have not been
witnessed by the EMU countries.
Second, regional unemployment differentials may be inefficient as they may reduce
GDP and put upward pressure on inflation. In addition, there is wide agreement that the
same nationwide unemployment rate may have different social repercussions depending
on the distribution of regional unemployment rates.1
The standard macro models explain unemployment differentials on the basis of the
differences in the institutions of the labour market like the wage bargaining mechanism,
the degree of social protection, the tax system, etc. However, although there are differ-
ences in the labour market institutions of different countries, there are no such differences
between the different regions of a European country.2
This led to the development of models that interpret unemployment disparities as
the result of scant inter-regional labour mobility or of regional differences in the labour
market - such as the sectorial composition of employment and the regional characteristics
of the unemployed workers. These explanations, although valid and relevant, only offer an
incomplete account of regional unemployment rates. The evolution of regional disparities
cannot be explained by labour mobility and idiosyncratic elements alone.
According to Marston (1985), the existence of regional disparities in unemployment
may reflect an equilibrium outcome. Each region tends to its own natural rate of unem-
ployment which is determined by demand side variables (such as the sectoral composition
of the regional demand), supply-side factors (such as differences in workers’ qualification
levels), and institutional variables (e.g., unemployment benefits and employment protec-
1 For example, consider the extreme case where a country has two regions of similar sizes. The social
impact of, say, a 10% national unemployment rate is not the same when both regions experience a 10%
unemployment rate, and when the unemployment rate in one region is 19% whilst in the other is 1%.
2 The legal systems of European countries ensure that regional differences in labour market institutions
are minimal.