Regional wages: a solution to the lack of regional mobility edit

Jan. 30, 2006

Regional disparities among EU countries remain substantial. In the second half of the nineties, the coefficient of variation of income per capita across 50 regions in the EU was almost double than across the 49 states of the continental US. Policy makers will then need to make a clear choice, in favour of greater labour mobility or higher wage flexibility. Otherwise, the combination of wage rigidity and limited labour mobility will continue to lead to unacceptably high unemployment differentials among regions.

-->The aggregate figures do not fully convey the size of remaining territorial disparities within the EU, even when excluding the new members countries . In 2002, the latest year for which a fully comparable set of data is available, income per capita - in purchasing power standards - in Inner London was more than two and half times larger than the EU-15 average. Conversely, GDP per head in the Greek region of Dytikis was around 50 percent of the European value. Disparities are somewhat less pronounced at the national level, but far from negligible. Even if we neglect Luxembourg, income per capita ranges from 111 per cent of the EU average in Denmark to around 67 percent in Greece. These large, and sometimes persistent, gaps in GDP per person reflect both differences in productivity and in labour market performance. National unemployment rates cover a broad range from 11.4 percent in Spain to 2.2 percent in the Netherlands. At the regional level the gap is substantially wider, from 27 percent in Halle in Germany to 2.2 percent in Utrecht in the Netherlands. Perhaps more disturbingly, in Europe high unemployment regions tend to coincide with low income per capita regions. Poverty and unemployment become two faces of the same coin.

We would expect these large differentials in income and employment conditions to foster substantial migratory flow from low income and high unemployment regions toward high income and low unemployment regions. They do not. With the exception of the UK, internal migration flows represent a much smaller share of resident population than in the US, despite the much greater variability in the income and employment outcomes in Europe.

Would greater mobility among European regions foster convergence ? Should policy therefore act to encourage migration within Europe or, at the very least, within member countries ? Surprisingly enough, the role of migration in favouring convergence does not seem to be a major priority for European policy-makers. The European Commission in its Report on Economic and Social Cohesion lists a broad set of factors that hamper cohesion, but fails to mention the lack of mobility among them. Similarly, the sixth periodic Report on the Social and Economic Situation and Developments of the Regions in the EU notes - in a text box - that migration may be associated with faster convergence, but then does not pursue the issue any further. Only the demographic implications of migration are considered. By and large, when it comes to policy initiatives, European policy-makers hold a somewhat schizophrenic attitude toward labour mobility. Finance ministries may sometimes lament the lack of mobility as hampering the ability of regions to adjust to idiosyncratic shocks.

Their colleagues in other ministries however seem to be quite reluctant to engage on this issue.

With limited regional mobility, wage flexibility should be called on to play a key role in cushioning European regions from the impact of idiosyncratic shocks. Unfortunately, even there, the evidence is not reassuring. In the two countries where regional differentials are more pronounced, Germany and Italy, wages are largely determined by union bargaining at the national level and, as a result, little affected by regional shocks. There is furthermore substantive evidence that regional wages in Italy are largely unresponsive to local labour market conditions, particularly so in unionized sectors. The implication is twofold. First, wages will typically be set at a level which is too high for the low productivity region, thereby causing a significant rise in unemployment there. Second, and perhaps more crucially, the backward region will be ill equipped to respond to a negative localized shock, to the extent that regional wages are set largely according to economic conditions in the more prosperous region. Even more troubling, a favourable idiosyncratic shock in the rich region may push national wages up and lead to higher unemployment in the backward region!

Again, policy makers have not particularly active in advocating a more flexible system of regional wage determination. They would often argue that labour markets distortions are not the main issue in the backward regions. They also mention the fact that productivity levels are virtually aligned among regions. This however should not be taken as reason of comfort, as it simply reflects the fall in employment brought by high wages in the backward regions and the consequent rise in unemployment. The true fact is that unemployment in Southern Italy is still more than three times as high than in the North, It is hard to argue that the wage setting mechanism has little or nothing to do with this fact. Labour market distortions may not be the only issue but they are certainly part of the problem.