Some unpleasant arithmetics of regional unemployment in the EU. Are there any lessons for EMU?

Several studies have documented the weak response of regional wage differentials and labour mobility following region-specific (“idiosyncraticâ€) shocks in the average of the EU countries. This has been often taken as evidence of the rigidity of labour markets in European countries, as opposed to the flexibility of the USA. However, as such shocks by definition average to zero, one cannot make an explicit link between the (lack of) adjustment at regional level and aggregate unemployment. Moreover, the emphasis on the reaction to short-run idiosyncratic shocks is unlikely to explain the permanent differentials across regions, which characterise the regional distribution of unemployment in many EU countries. This paper tries to provide a better understanding of the regional distribution of unemployment and why region-specific shocks can matter for aggregate unemployment. It does so by explicitly considering the possibility of asymmetric reactions, so that unemployment rises more in poorer areas suffering an adverse shock than it declines in richer regions experiencing a favourable shock. The reason behind such asymmetries is the presence of a wage floor in the poorer regions resulting from policy centralisation, as for instance in the case of a national unemployment compensation system, which provides benefits that are uniform across regions. If such a mechanism is at work, aggregate unemployment tends to be “inflated†by region-specific shocks that are inequality- increasing. After presenting an illustrative model of the mechanism, the paper proposes a simple measure of the resulting “excess unemploymentâ€, based on the difference between the average (national) unemployment rate and the unemployment rate of the median region. It also examines the relationship between regional asymmetries in unemployment and the dispersion of productivity across regions, taken as proxy of the inequality-increasing shocks. The evidence, while not entirely conclusive, justifies two tentative policy conclusions, which are particularly relevant in the context of EMU: a) to avoid centralisation of labour market institutions at the EU level that may end up inflating aggregate unemployment; b) to effectively deploy regional policies to combat inequality- increasing shocks.