The existence and persistence of large spatial disparities in unemployment within national economies is a central issue in regional economics. Explanations that exist within the literature for why regional unemployment disparities have been so much more persistent over time in the EU than in the US can be broadly separated to the hysteresis and equilibrium arguments. Before illustrating the analysis and strength of these arguments it is important to ascertain the scale of persistent unemployment disparities across EU regions and
U.S states, and why this may be a cause for concern.

Regional disparities in unemployment rates are dramatically significant in Europe. In 2001, the average unemployment rate in the
EU was 7.8%, but it ranged from 1.7% in Berkshire (UK) to 24.9% in
Calabria (Italy).[1] These disparities are present even within single countries, where the range of unemployment rates is much higher than in the US. Disparities in European regional unemployment rates are also extremely persistent. In the US the correlation between regional unemployment rates between 1997-2004 is only 0.3, indicating that over two-thirds of any regional unemployment problem is solved over that period. In Europe, however, the correlation averages 0.83, suggesting that very little adjustment takes place. (Figure 1 in Appendix)[2]
This is not a recent phenomena but persistence appears to have been the main feature of behaviour of regional unemployment disparities in
European regions since the early 1980s. Baddeley, Martin and Tyler provide this evidence in a report, using rank-order correlations between regional unemployment rates using five ‘core’ member states
[3] (data limitations prevent the comprehensive investigation of regional unemployment for all sub regions of the EU-in particular the southern member states.) The results illustrate a high degree of stability or persistence in the UK, Germany, Italy and France in contrast to the idiosyncratic regional unemployment levels in the US.

Significant persistent disparities in regional unemployment rates can have numerous problems. It entails considerable differences in the economic and social exclusion associated with joblessness. Research conducted within this field in England and Wales demonstrates how high unemployment is correlated with poor performance on a range of socio-economic variables. Localities with the highest unemployment rates also have the ‘highest proportion of lone parent families, the highest truancy rates and the highest proportion of pupils eligible for free school meals, which is an indication of poverty.’[4]
Furthermore Oswald (1997) reports survey evidence that shows unemployed people are much less happy people then the employed and
‘support the conclusion that losing one's job is amongst the worst things that can happen to a person.’[5] Regional unemployment disparities specifically break down informal support networks and asset prices drop sharply with economic distress becoming an acute social problem. Areas with persistent levels of high unemployment appear strongly associated with a large shadow economy. For instance, in Italy the employment in the informal sectors is estimated to be about 30% in the south and less than 10% in the North.[6] This relationship indicates high regional unemployment could reflect, and even to some degree cause, wider governance problems. All these problems outline the importance regional unemployment and also how there could be inefficiency in labour and capital markets since resources that are in excess supply in some regions are not productively used in others.

A theoretical framework (that encompasses different explanations) for analysing why regional unemployment disparities tend to be more persistent within the EU than within the US The framework is a demand-supply framework with the feature of a labour market that does not clear due to the ‘imperfect’ nature of the wage bargaining process. Within this framework, a region has an equilibrium rate of unemployment that is decreasing with the amount of labour demand, which is itself positively related to the level of demand for a region’s output and the efficiency with which a region is able to combine capital and labour to produce output. This framework allows us to analyse the impact of labour market shocks, namely labour demand shocks as they are much more frequent. Ceteris paribus, a negative labour demand shock causes a region’s rate of unemployment to increase, which is involuntary. Mechanisms that deal with such negative shocks so unemployment returns to its equilibrium level include the real wage flexibility mechanism and the out-migration mechanism. There are more mechanisms but these two seem to be empirically easier to analyse and are the most significant. Should none of these mechanisms operate then a region is said to exhibit pure hysteresis. Unemployment hysteresis, proposes that cyclical fluctuations will have permanent effects on the level of unemployment due to labour market rigidities. Thus, under this hypothesis – pioneered by Blanchard and
Summers (1986) – the level of unemployment is characterised as a non-stationary process.. This means that for a region to exhibit pure hysteresis a negative labour shock will have a permanent negative effect upon its unemployment rate.

The flexibility variant of the hysteresis explanation for persistent unemployment disparities may be clearly portrayed by contrasting the experience of local labour markets in Europe and those in the USA.
Thus for a negative labour demand shock the real wage should be able to adjust accordingly for unemployment to revert back to the original equilibrium point. The mechanism, in theory, is further enhanced as lower labour costs would attract firms to the region.

Mechanism 1: real wage flexibility

European labour markets have historically been associated with institutional inflexibilities that have hit unskilled workers particularly in manufacturing trade with the developing world, creating ‘entrenched localized problems of high unemployment and rising welfare bills.’[7] Yet in the USA wages and workers seem to be more flexible with workers more mobile between local labour markets.
Studies in this issue show how certain factors have influenced wage rigidity in European countries namely Insider-Outsider effects-where the ‘degree of real wage adjustment that would be required to make employment of job seekers profitable for potential employers,’ is prevented.[8] A further structural factor of youth wages also highlights the greater degree of flexibility in US labour markets compared to Europe. Greater flexibility of youth wages in US seem to contribute to the fact that youth unemployment rates (18-25 yr olds) are ‘consistently higher across Europe than they are in the US.’[9]
However further analysis on this argument of wage inflexibility in EU regions being a significant factor for persistent regional unemployment rates reveals conclusions that are at odds with conventional wisdom. Reports by Dercressin and Fatas (1995) and
Baddeley, Martin and Tyler (2000) provide evidence on the working of the real wage flexibility mechanism. Using a sample period of
1976-1994 the latter report concluded that real wage flexibility across regions of five EU countries compares favourably with that across US states. This undermines one version of the hysteresis explanation for why regional unemployment disparities are so much more persistent in the EU. It weakens the idea that EU regions exhibit pure hysteresis whilst US states do not because the real wage flexibility mechanism works better in the latter. Furthermore, the widening wage relativities associated with labour deregulation as seen in the US have contributed to ‘increased labour market insecurity, inequality and social polarization,’[10] which has sadly been emulated by the UK, the European country that has gone furthest towards labour flexibility. Thus those who promote the acceptance of an ‘American
Model,’ (labour market deregulation and flexible wages) to resolve
Europe’s uneven unemployment problem may be mistaken in assuming that wage flexibility is itself sufficient to increase regional employment.

The relative weakness of the real wage argument has meant other adjustment mechanisms must prevent adverse labour market shocks from resulting in prolonged disparities in unemployment across regions in the US. The out-migration variant of the hysteresis explanation provides greater substance to the issue. This mechanism assumes that when a region faces a negative labour demand shock a pool of involuntary unemployed migrate out of the region.

Mechanism 2:outmigration

From the work of Blanchard and Katz (1992), we know that regional unemployment disparities are not persistent across US states because of the working of the out-migration mechanism, thus negative shocks to regional labour demand is quickly corrected. Yet this mechanism does not operate for EU regions implying pure hysteresis and explaining the persistence of EU regional unemployment disparities. The cultural and language barriers obviously hinder labour migration across nations, yet even within individual European countries migration rates are relatively less then US states. Professor Oswald has provided key insight in this issue, illustrating how the out-migration mechanism is able to work better in the US than in Europe because the US has a greater private rental sector and a lower rate of home ownership. The evidence shows that nations with the fastest growth in home ownership had the fastest growth in unemployment (Figure 2 in Appendix)[11], suggesting that home owners are relatively immobile presumably because they find it much more costly then private renters to move around in search of new jobs. A shock to the housing market usually follows a negative shock to the labour market, thus greater home ownership leads to a higher level of negative equity making it more difficult to move.
The report identifies how most industrialised countries have experienced growth in home ownership with the two exceptions of the US and Switzerland. However, this only presents indirect evidence in favour of the hysteresis argument. So although the explanation works in favour of the US condition, it is limited at explaining the persistence of EU regional unemployment disparities.

The ‘equilibrium,’ phenomenon suggests that regional unemployment differences have different underlying long-run means across regions.
Blanchard and Katz 1992 provides information to support both explanations for US states, thus the equilibrium argument also holds as there are similar equilibrium unemployment rates across states.
Yet, for the EU regions it can be argued that the mechanisms for response to a negative labour market shock do work but the equilibrium unemployment rates are very different across regions. This evidence can be provided by Baddeley et al (1998), and with the exception of
Belgium regional unemployment disparities in the ‘EU core countries reflect some form of inter-area equilibrium in that they have different means in different regions because of varying demand-side, supply side or institutional characteristics.’[12] We can distinguish these factors from the Keynesian style view (of labour demand as the cause of large disparities across EU regions), and neoclassical style view (highlights variations in supply side and institutional factors such as strength of union organisation.) In favour of the Keynesian view regional variations in industry help to explain differences in equilibrium unemployment rates, which vary more across EU regions than
US states. Other areas face a persistent deficiency in demand because their economies are based on ‘industries that are no longer competitive in today’s global economy, or whose markets are in secular decline-typified by Europe’s industrial areas of northern member states.’[13] The neoclassical view takes the alternative interpretation focusing on the quality of the workforce and on structural restraint on the movement of workers and firms and even amenity levels within regions. The immobility of labour means different types of jobs in different regions are not substitutable opportunities for any given type of worker. The amenity issue demonstrates how areas that are more attractive with better facilities and climate are able to cope with a higher level of local unemployment. However these theoretical reasons are harder to prove with empirical evidence, thus there is no consensus over which argument is superior. A limitation in this respect is the fact that regional economics has only become a recent fashionable study and so research is not yet advanced enough to answer this question. It is hoped that research in the future can help to offer more clarity on this issue.

It is apparent in this complex issue that there is no clear reason for why unemployment disparities are so much more persistent across EU regions rather than across U.S. states. Although both hysteresis and equilibrium explanations tell the same story for the US, they differ for EU regions. It is clear that the out-migration variant is more important then the real wage flexibility model. Yet further research in the matter will aid future studies to assess the strength of the equilibrium argument and to ascertain whether the Keynesian or
Neoclassical view is superior.