Michel Mignolet

Regional Employment Disparities in Belgium: Some Empirical Results

Employment and unemployment are unequally distributed over space and Belgium is no exception to the rule, its regions recording strongly contrasting performances. The Flemish region displays high levels of participation rate in the labour market and low rates of unemployment. By contrast, Wallonia and Brussels are characterized by poor results for both employment and unemployment. Moreover, the interregional divergence in the performance of the labour market shows no significant tendency to decrease. See Appendix 1 for a breakdown of Belgian spatial administrative divisions. Over the period extending from 1973 and 2001, Flanders recorded an increase of 33 percent in the number of employees, while Brussels experienced a decline of 1 percent and Wallonia had a slow increase of 6 percent. This employment pattern feeds the political debate in Belgium, putting a strain on the national consensus between the Flemish-speaking Northern region and the French-speaking Walloon region. Within the regions, a similar performance disparity is observable between districts. For example, Bastogne and Charleroi are districts of the same region, Wallonia. Between 1997 and 2001, the number of jobs in Bastogne increased by 98 percent, while Charleroi lost 19 percent of its labour force. Figure 6.1 illustrates, for the different districts, the regional and inter-regional employment disparities in employment rates in 2001. To explain the interregional discrepancy, it has long been argued that poor employment results recorded in Wallonia relative to Flanders are primarily ascribable to an initially adverse industrial mix. However, recent work has questioned this hypothesis, stressing instead the mediocre sectoral growth performance in Wallonia (see Estevao 2003). This chapter sets new light on this issue, notably by using a highly disagregated employment series. More precisely, we intend to examine regional employment disparities in Belgium and their evolution in the recent past by developing a shift-share analysis and an econometric approach and applying this to employment statistics.