In the first few days of August, the Ministry of Employment and Labour traditionally publishes the figures for unemployment as on 31 July of the current year. At the same time, the National Employment Service (Office National de l'Emploi or ONEm), which is run on a joint basis by employers' representatives and trade unions and manages unemployment insurance, publishes its own report which includes annual unemployment statistics for the preceding year.
How can unemployment figures be calculated objectively in Belgium? In late summer every year, conflicting press headlines reveal confusing news stories about unemployment.
In the first few days of August, the Ministry of Employment and Labour traditionally publishes the figures for unemployment as on 31 July of the current year. At the same time, the National Employment Service (Office National de l'Emploi or ONEm), which is run on a joint basis by employers' representatives and trade unions and manages unemployment insurance, publishes its own report which includes annual unemployment statistics for the preceding year.
The press immediately tries to enlighten the public regarding the employment situation in Belgium and risks some dangerous comparisons. Is unemployment going up or down this year?
According to ONEm, the national annual average number of fully unemployed persons in receipt of unemployment benefits (Chômeurs complets indemnisés demandeurs d'emploi- CCI-DE) fell to 477,740 in 1996 from 499,954 in 1995, that is, a drop of 22,214. The unemployment rate (calculated as a comparison with the total number of people insured against unemployment in 1994) fell from 17.12% to 16.36%, a drop of 0.76 percentage points).
According to the Ministry of Employment, the number of those out work decreased from 487,790 in July 1996 to 477,369 in July 1997 (a drop of 10,421 or 2.1% over the year), though the number of unemployed people rose by 4.5% between June and July 1997.
The simultaneous publication of both sets of figures means that any interpretation of employment data becomes a real headache unless the methods of calculation are explained. The mystery deepens as changes have been made from year to year both in the calculation methods and in the legal regulations surrounding definitions of unemployment.
For instance, unemployed people studying or in training (8,510 - up 15.4% between 1995 and 1996) are no longer counted as fully unemployed whilst those over the age of 50 (91,665 - up 26.5% between 1995 and 1996) have not been included in the definition since 1988.
Indeed, 32,044 long-term unemployed people - mainly married women now deprived of benefits (up 10,144 between 1995 and 1996) - have also disappeared from the calculations, as they are no longer in receipt of unemployment benefit. In fact, since 1996, the length of time for which one may draw unemployment benefits has been revised and reduced by a quarter.
The Ministry's communiqué now calculates the unemployment rate on the basis of a population representing at most 80.5% of all those seeking work, and in particular on the basis of those groups characterised by a decreasing or steady rate of unemployment.
Accordingly, professional expertise in statistics will soon be required to gain a precise idea of the unemployment situation in Belgium, not to mention some talent as a detective.
Eurofound recommends citing this publication in the following way.
Eurofound (1997), Pitfalls in reading Belgian unemployment figures, article.
