The first part of the article sets out the provisional results of the analysis of the social balance sheets for the year 2003. Since not all the social balance sheets are available for that year as yet, the study is based on a reduced population constructed according to the constant sample principle, comprising 40,630 enterprises employing around 1,372,000 workers in 2003. The main results of the analysis of that reduced population are as follows. Employment declined at an annual average rate of 0.8 p.c. in 2003, but the rate of job losses slowed down slightly at the end of the year. While the number of full-time workers declined, the number of part-timers increased once again. Net staff departures slowed down between 2002 and 2003, particularly in the case of workers with primary or secondary school level skills. On the other hand, net recruitment of highly skilled staff remained below the levels recorded the previous year. Results confirm that employers have tried to use the most socially acceptable ways of implementing job cuts, as there was only a small rise in redundancies between 2002 and 2003, in contrast to the numbers taking normal or early retirement and the numbers whose fixed-term contracts were not renewed. At the same time, there was little change in the structure of employment by type of contract, where permanent contracts predominate by a long way, as the numbers of both fixed-term and permanent contracts declined. On the other hand, there was an increase in the use of agency workers, providing evidence of the steady improvement in the situation of enterprises in the reduced population. Staff costs were up by 1.6 p.c. between 2002 and 2003. Since employment in terms of full-time equivalents (FTEs) was down by 1 p.c. on average and the working hours per FTE remained unchanged, costs per hour worked increased by 2.7 p.c. Regarding the training policy, participation rates stagnated in 2003, as did the percentage of working hours devoted to acquiring skills or keeping them up to standard. The indicator of the amount spent on training declined : it represented just 1.2 p.c. of the wage bill. This renewed fall negates the efforts made here in the late 1990s, so that altogether, between 1998 and 2003, there has been no progress towards the objective of 1.9 p.c. fixed in the 1998 interprofessional agreement. The second part of the article analyses the characteristics of the enterprises according to their geographical location ; firms operating in only one region were distinguished from those with establishments in more than one region. In the latest case, the social balance sheets were allocated into the three regions of the country according to the one in which the largest number of jobs was recorded by the National Social Security Office. The analysis presented is an initial appraisal of the regional characteristics of firms which have to submit a social balance sheet. The average working hours recorded in the six groups of enterprises, namely single-region and multi-regional enterprises operating in Brussels, Flanders and Wallonia, reflects the differences in the structure of activity, since working hours within the actual branches of activity are relatively uniform in the different regions. That is less true of staff costs, in any case for single-regional enterprises, as hourly costs are systematically lower in Wallonian enterprises, for each branch of activity, and higher in Brussels enterprises, even though the firms are of the same average size. As regards the training policy, single-regional Wallonian enterprises are still lagging behind their counterparts in the other two regions, despite the progress made in terms of participation rates. For multi-regional enterprises, the weaker results in Wallonia are probably due in part to the smaller average size of the firms, and the preponderance of industrial enterprises. The latest devote fewer resources to their training policy than the financial, real estate and business services branches, for example, which account for 55 p.c. of the staff employed by Brussels multi-regional enterprises, and than the trade, transport and communications branches which represent over half of the workforce of multi-regional enterprises in Flanders.
