Is Belgium ready for EMU? A look at national, sectoral and regional developments

Summary for Economic Policy: The starting date of the European Monetary Union (EMU) is coming close. After several years of intense efforts, eleven European Union (EU) countries are prepared to go ahead. Belgium is one of those countries. This is a step of great importance. Gradually, the attention is shifting from the preparation of EMU to the question whether EMU will work. In Belgium much of this discussion focuses on the performance of the labour market. Belgium is a country with relatively high structural unemployment. The country is also confronted with a substantial regional gap in economic growth and unemployment. This diverging economic performance gives rise to politically sensitive interregional transfers between the North and the South. There is a widespread concern that the Belgian labour market will fail to cope efficiently with the need to adjust to changing economic conditions. While a macroeconomic wage restraint with respect to the major trading partners is ingrained in the so-called ‘law to safeguard the Belgian competitive position’, the country has been repeatedly urged by institutions such as the IMF and the OECD to promote wage and labour market flexibility, as well as to increase labour mobility. The prime minister, Jean-Luc Dehaene, pleaded recently for less government intervention in the wage formation process and for a greater wage differentiation that takes into account the diverging conditions that exist in various sectors of the economy. This position is shared by Karel Vinck, the head of the Flemish Employer Organisation (VEV), who views an increased role for sectoral negotiations between unions and employers as a means to achieving wage differentiation in the Belgian economy. And also Karel Boone, the president of the Belgian Employer Federation (VBO), emphasises the responsability of the sectoral level in the Belgian wage negotiations. Is this concern about insufficient labour market adjustment warranted ? The answer to this question requires a detailed understanding of the structural features in the Belgian economy. Subsequently, one must assess whether wages adjust to those economic shocks. Both issues are the focus of this study. After a theoretical discussion of the statistical framework (which can be skipped by the technically less inclined reader), the second part of this paper offers a detailed picture of changes in production that occurred in the Belgian economy during the period 1985-1995. Those production shocks were strong and, to an important degree, can be statistically decomposed in developments at the national, regional and sectoral level. Sectoral shocks are changes that are observed within the same industry in Flanders, Brussels and Wallonia but that are not shared by other sectors. Those sectoral changes are the dominant driving force in the Belgian economy. In our statistical exercise, the sectoral component explains an average of 47% of total output variation and 76% of the output changes explained by our statistical model. Even higher numbers are obtained when changes in employment and labour productivity are considered. The message is clear : the main source of differentiation in the Belgian economy is found at the sectoral level. These average figures mask a lot of interesting variation in individual industrial and service sectors. We distinguish between a small group of sectors that experience a structurally better growth performance than a group of poor performers. In between those two groups, a large number of primarily manufacturing industries are situated with a comparable growth performance. Looking at the cyclical pattern of individual sectors, there is not much evidence for very different nor for very similar sector-specific business cycles. One interesting exception is the countercyclical role of the government. Apparently, the government increases spending on public services in times of an economic slowdown. In this paper, regional shocks are defined as output developments that occur in all sectors of a specific region. We find evidence of such shocks but they are of secondary importance when compared to sectoral changes. Apparently, many of the observed regional trends are caused by developments in specific sectors that are primarily located in one region. Having said this, the well-known regional division of the Belgian economy emerges clearly from the statistical model. Flanders is characterised by a better medium-term growth performance, a different cyclical pattern and a greater similarity in regional output changes than either Brussels or Wallonia. By definition, national shocks are found in all sectors and in all regions. From 1985 to 1995 they accounted for 9% of the total output variation and for 14% of the production changes explained by our model. National factors matter mostly for a limited group of sectors including Metal Products, Construction, Other Market Services and Other Manufacturing. In an efficient labour market, we expect wage growth to reflect the dominating role of developments at the sectoral level. In the third part of the study, we found this not to be the case in the Belgian economy. While wage levels to some degree reflect long term sectoral productivity differentials, wage growth displays little differentiation across sectors. This lack of sectoral wage flexibility amplifies the impact of the production shocks on employment. Weaker sectors are faced with wage increases which they can ill afford. As a result, the employment performance gap between stronger and weaker sectors widens. In the end, the Belgian economy is faced with a substantial and widespread need for employment reallocation, for which the rigid labour market with limited labour mobility is ill prepared. A greater sectoral wage flexibility would avoid many of those problems. What about regional differentiation in wage setting ? There is no sign that Belgian wage increases are taking into account regional differences. This comes mostly at the expense of the weakest region, Wallonia, which suffers from higher unemployment than would be obtained if regional wage moderation were feasible. Quite likely, greater sectoral wage differentiation would contribute a lot to regional wage flexibility. Wherever necessary, this sectoral approach could be supplemented by greater attention to the regional dimension in wage setting. Such complementary approach would reduce the need for politically sensitive long-term interregional transfers and in this way enhance political cohesion in the country.