Economische groei zonder sociale vooruitgang: stand van zaken in Brussel

Dit artikel bestudeert de economische en sociale evoluties van Brussel in de laatste twee decennia. Ondanks een gunstige economische conjunctuur in die periode, blijft de sociale situatie verslechteren met werkloosheidscijfers die steeds hoger liggen en met sociale ongelijkheden die almaar toenemen. Hoewel zulks ook het geval is in de andere Belgische grote steden, is die situatie meer uitgesproken in Brussel. Dat wordt verklaard door een veelvoud van dynamische factoren, waaronder de almaar hogere opleidingseisen die aan de arbeidskrachten worden gesteld en die laagopgeleiden uitsluiten van de arbeidsmarkt. Die evolutie wordt nog versterkt door de ruimtelijke concentratie van kansarme bevolkingsgroepen, die zeer groot is in Brussel. Het artikel sluit af met de vraag of het Gewest wel een pertinent beleid voert, daar het de nadruk legt op de internationale positionering van Brussel.


Introduction
Over the past two decades, the major European cities have undergone paradoxical evolutions: whilst they have experienced economic growth superior to what is seen in other types of territory, social indicators continue to deteriorate, in particular as regards the most disadvantaged populations.According to many authors, the contemporary model of economic growth generates a growing social polarisation in itself, in major cities in particular (Sassen, 1991;Hamnett, 1996).However, theories regarding the complex relationship between the economic development of cities and social polarisation -and more generally regarding the processes of this polarisation -have not found empirical expression.
The existing data suggest that Brussels has experienced evolutions similar to the other major European cities, but empirical proof is lacking for a complete assessment of the social and economic evolutions which have taken place.A recent study attempted to fill this gap by carrying out a systematic study of the economic and social evolution of the five biggest Belgian cities, by distinguishing the centres of employment and the employment areas. 1   In this article, we therefore propose a comparative approach aimed at relating and explaining the economic and social evolutions in Brussels.In very concrete terms, we shall answer the following questions: -Has the economic situation in Brussels improved over the past two decades ?-Have social inequalities grown during this same period ?-How can the observed evolutions be explained?Following a presentation of the main theories describing the socioeconomic evolutions of cities (section 1), we shall analyse the social and economic evolutions of greater Brussels in comparison with the other major Belgian cities (section 2).In the third section, we shall attempt to understand why economic growth in Brussels no longer produces a noticeable improvement in the social situation -on the contrary.In conclusion, we shall propose a discussion of the policies pursued in the face of the declining social situation in Brussels.
Brussels Studies the e-journal for academic research on Brussels 1 1 The 'GROSE' study, conducted by IGEAT-ULB and COSMOPOLIS-VUB and financed by the Federal Science Policy in the framework of the 'Society and Future' programme.

Socially polarising economic growth in major cities -theoretical controversies
The observation of growing inequalities in western societies beginning in the 1980s is no longer the object of debate (OECD, 2006;Smeeding, 2002;Atkinson, 2003).It is also generally admitted that this process characterises metropolitan spaces in a particularly stark manner (Badcock, 1997;Hamnett, 1996).On the other hand, the causes of this evolution were the object of an intense debate during the 1990s, without really leading to any conclusions.
There are two simple yet conflicting theses.Either the new forms of economic growth are socially polarising because they produce a polarised labour demand in the major cities as regards qualifications, i.e. both highly skilled and low-skilled (Sassen, 1991), or the evolutions of metropolitan economies produce a demand for increasingly skilled labour, thus excluding those who lack the necessary qualifications (Hamnett, 1996).Let us take a closer look at the opposing arguments.
According to Sassen's thesis, globalisation favours global cities.More precisely, the evolutions since the 1980s produce a spatial process of concentration (of capital, strategic functions and economic command) to the benefit of global cities and decentralisation (of low-or semi-skilled productive functions) at urban, national and global level. 2he concentration of strategic activities in global cities is therefore accompanied by the decentralisation of industrial production and 'heavy services' (back office, call centres, etc.), except when they are related to local demand (food distribution, cement works, logistics, etc.).Global cities are therefore characterised by a dual labour market, with the growth -at the upper end -of very qualified functions andat the lower end -of basic captive services, in particular those intended for the upper classes (child care, leisure activities, hospitality business, etc.).The middle segments related to production therefore tend to decline.Hamnett (1996) differs with Sassen's thesis on a more empirical than theoretical basis.Although their theories coincide regarding the double observation of the regained health of metropolitan economies and of growing social inequalities, her research on London and the United Kingdom underlined a process of 'professionalisation'3 (i.e. a growing demand for skills) rather than a polarisation in the labour market.
In simple terms, manual labour and, more generally, low-skilled labour are declining, whilst the labour market offers more and more skilled employment.At the same time, major cities have experienced an increased demand for low-skilled labour, following a process combining industrial decline and an influx of low-skilled workers in the major cities.Consequently, instead of being due to the polarisation of the employment offer, social polarisation is explained by a growing requirement in terms of qualifications, which ends up excluding entire segments of the population from the Brussels Studies the e-journal for academic research on Brussels 2 labour market.This 'professionalisation' and a polarisation of income have also been identified in research conducted in the Netherlands (Hamnett), Paris (Préteceille), Helsinki (Vaattovaara, Kortteinen, 2003) and in American cities (Kloosterman, 1996).
Part of the difference between the two theses lies in the context in which they were produced: in Europe, the welfare state limits not only differences in income, but also the possibility for growth in precarious and underpaid employment, thus reducing the polarisation of the labour market (formal at least).That being the case, we agree with Hamnett's criticisms of the fact that it is not possible to understand social polarisation through a mere analysis of market forces, and that it is necessary for the interpretation to consider the importance of state regulations and the level of protection provided by the welfare state.This brings us to a third thesis which aims to go beyond the opposition between Sassen and Hamnett.It is based on the transition from Fordist capitalism to flexible capitalism4 at the beginning of the 1980s (Vandermotten et al., 2010;Boyer, 2006;Jessop, 1997).Like the theory of global cities, this reading has the advantage of proposing a theory which accounts for the regained health of metropolitan economies as well as growing social inequalities.In a macroeconomic context marked by a more unstable global demand, major structural evolutions underscore the transition towards more flexible capitalism: the expansion of markets ('globalisation') accompanied by a growing concentration of capital, the growing role of finance and the reorganisation of companies to return to their core business, together with a generalised use of subcontracting and a network organisation (Swyngedouw, 2003;Boyer, 2004).
At geographic level, these macroeconomic evolutions are expressed as a process of metropolitanization, i.e. of (re)concentration of activity (or of certain activities) in major cities.Why?
The city offers a large and diversified employment area, both skilled and low-skilled, flexible and available, and which corresponds best to the demands of this new context.The diversity of competences and stakeholders also favours innovation processes, often considered to be important in order to support growth.
The city also offers a critical mass of diversified stakeholders corresponding to the requirements of this type of technical/economic organisation: research centres, the SME fabric, decision-making authorities, multiple services, etc.It therefore provides the range of services needed by companies in the context of a return to core business.The city brings all of these stakeholders face to face within an economy which is more horizontal than before.
In a more global economy, major cities have a significant advantage as they constitute -at any rate through their physical transport infrastructures -the interface between regional and national levels and the global economy.
Brussels Studies the e-journal for academic research on Brussels 3 Finally, the strategic functions of command and management converge more and more in the major global cities of the developed world in a context of concentrated capital at global level.This results in stronger requirements in terms of qualifications.These characteristic evolutions of major cities -and not only global cities according to Sassen's definition -also explain the processes of growing social polarisation observed in them.Social polarisation is due as much to the calling into question of political evolutions which occurred during the neoliberal turning point in the 1980s, as it is to the 'technical' evolutions which lead to a growing demand for skilled labour in the labour market.Two elements are essential from this point of view: the flexibilisation of the labour market and the evolution of the state's role.On the one hand, the rise in structural unemployment in the 1970s and 1980s favoured an evolution of power relationships between unions and employers leading to a flexibilisation of labour, in particular through processes of the individualisation of work contracts and the permanent competition between different categories of worker.This flexibilisation was also favoured by the evolution of economic structures.From this point of view, contrary to Sassen's thesis, the transition of industrial and heavy sectors towards services was not translated into a polarisation of qualifications but rather into a polarisation of wages.In other words, we do not support the idea that semi-skilled industrial employment was replaced by low-skilled employment.A large part of this industrial employment was in fact deskilled, in particular in the Fordist sectors such as the car sector.On the other hand, it is certain that employment in sectors in which low-skilled industrial labour had acquired average salary levels and was capable of making itself heard was replaced by flexible employment, with more women employed and in non-unionised sectors.In other words, the dualisation in the labour market is probably related to the evolution of social power relationships rather than to the less skilled nature of employment.On the other hand, the state played a central role in these evolutions through the privatisation of state sectors which ensured a high level of protection for staff, through the relaxing of labour market regulations, and through the reduction (or reorganisation) of social redistribution policies.
However, the processes of qualification and polarisation in the labour market are not only the result of the transition towards more flexible capitalism.They are also inherent to capitalism itself, i.e. they are at work regardless of the regime of accumulation (Harvey, 2006).Indeed, the evolutions of capitalism are characterised by a permanent dialectic between skill requirements and deskilling.On the one hand, the process of increasing skill requirements is related to the growing importance of technology, research and management functions inside a company (executives, qualified technicians) as well as outside (education, administration, etc.).On the other hand, the search for productivity gains involves a segmentation of tasks (or 'Taylorisation'), which is usually characteristic of industrial sectors but tends to apply to entire segments of the service economy (Vakaloulis, 2001), from the retail trade to qualified sectors such as finance.These evolutions lead to deskilling and put an end to the destruction of deskilled employment, especially when the segmentation of tasks allows a high level of mechanisation.
The theoretical interpretation proposed here has the advantage of insisting on the unity of metropolitanization and social polarisation processes.This is also the case with the theory of global cities -at least in Sassen's version -but we feel it suffers Brussels Studies the e-journal for academic research on Brussels 4 from empirical as well as theoretical inadequacies, as we have mentioned above.From an empirical point of view, we shall explain that the reality of Belgian (and probably European) cities is better accounted for in Hamnett's analyses.On the other hand, they do not allow a unified reading of the sociospatial processes which we propose here.It is nonetheless necessary to bear in mind that the proposed theoretical reading constitutes a framework for interpretation and not a hypothesis which we shall prove.This would require in particular a focus on the impact of the evolution of social redistribution policies or a consideration of the subtle evolutions within the working world, which goes beyond the framework of this study.

The spatial framework of the study
As the study focuses on Belgian metropolitan areas, let us first define the spatial limits involved.One of the original aspects of this study is that it proposes a com-  parison of major Belgian cities with a functional and homogeneous definition of them.For the five major urban areas, we therefore distinguish between the centres of employment and their area of influence (the employment areas).
The centres of employment were defined by all of the municipalities which met the criteria of an employment rate above 1.255 and a high volume of employment.The resulting major centres correspond to the five major cities used by Luyten and Van Hecke (2007), with, however, some additional municipalities in Antwerp, Liege and Brussels. 6The employment area of each of the centres is defined by the municipalities which send more than 10% of their employed workers there (Figure 1).
These evolutions are in perfect keeping with the theoretical developments presented above, with a reconcentration of activity in the country's biggest city, in particular in strategic sectors such as finance and transport.However, although the economic performance of the Brussels centre of employment is higher than in the other major centres of employment, the suburban areas of the city experienced considerable economic growth.The added value figures are calculated based on regional accounts at the National Bank of Belgium.These sectoral data are therefore broken down into the data regarding paid employment and self-employment at municipality level to obtain municipal added value figures, thus allowing the employment areas to be put together.

Explanation of the indicator:
The ratio between the growth in employment and added value allows an evaluation of the extent to which economic growth generates employment.As the ratio gets higher, the growth generates more employment.

Growing social polarisation
Here we shall provide an overview of the interpersonal (or inter-household) inequalities in the major Belgian cities as regards their employment area, and evaluate the dynamics.We propose a comparative perspective of Brussels with respect to the four other cities, by using two types of indicator: inequalities of income and access to employment according to education level.Social inequalities have increased over the past twenty years.On the one hand, in terms of income per household (Figure 3), the evolution of the Gini coefficient (cf. the note under Figure 3) between 1985 and 2006 clearly demonstrates this.Let us also mention that whilst Brussels has experienced evolutions which are similar to the other major Belgian cities, it is also the city with the highest level of inequality throughout the period.On the other hand, in a context of the 'professionalisation' of the labour market accompanied by a rise in unemployment, access to employment differs more and more according to education level.Although a general rise in unemployment is observed between 1991 and 2006 (Table 2), this evolution is more pronounced for lower education levels.Figure 4 presents only the case of Brussels but the observation is also valid for the other cities, yet at different levels of unemployment.

The causes of social polarisation
In section 2, we explained the extent to which Brussels is in keeping with the evolutions of the other major European cities, with a paradoxical combination of economic growth and decline in social indicators.
The literature has suggested different explanations, in particular based on the evolutions of the welfare state and the structural evolutions of the labour market.Here we shall mainly elaborate upon the evolutions of the labour market which are at the centre of the debate between Sassen and Hamnett.The analysis of these evolutions allows an explanation of the mechanisms which intensify social polarisation.On the other hand, our study does not deal directly with the impact of tax and social redistribution policies.It is very difficult to measure their impact on the evolution of inequalities of income at regional or local level, even if it is obvious that they have played a significant role, in particular in the growth of inequalities at a higher level. 7he impact of the evolution of social legislations on situations at the bottom of the social ladder is almost impossible to evaluate.Finally, we shall discuss the relationship between sociospatial polarisation and the dynamics of social inequalities in Brussels.

A growing demand for skilled labour
In Brussels more than in the other cities in the country, an increase in the skill requirements of the labour demand is observed.In terms of education level, there has been a very clear increase in employment qualifications since the 1990s in all of the major Belgian cities -Brussels in particular.According to the labour force survey, between 1992 and 2005 the share of higher education graduates among workers who live in the Brussels-Capital Region rose from 37% to 50% (and then stayed around this level).However, these evolutions reflect those of the demand as well as the labour supply.The active population is increasingly qualified, with a rate of higher education graduates which rose from 33% to 46% in the BCR during the same period.This is due in particular to the arrival on the market of new generations of Brussels Studies the e-journal for academic research on Brussels 10 workers who are much more skilled than the former generations.Furthermore, in a context of structural unemployment as is the case in Brussels, employers favour the recruitment of people with a higher level of education even when it is not required.This reinforces the tendency towards the formal prerequisites for employment (Colard et al., 1997;Devillé H., 2008).
In order to overcome the limits of an analysis of the evolution of work centred on the education level of workers, we developed an analysis of the evolution of professions.We therefore had to adopt an a priori classification of professions according to their level of qualification.We used the classification into social classes devised by Eriksson-Goldthorpe-Portocarrero (EGP). 8As it is based on the type of contract and the autonomy of professions, we were able to use it as a good approximation of the qualification levels of professions.We shall refer only to the evolution of the most qualified (EGP I and II) and the least qualified categories (EGP VI and VII), and shall disregard the intermediate categories of employment.
The results leave little doubt regarding the intensity of the employment qualification process in the major Belgian cities and more so in Brussels (Table 3): the employment area of Brussels has the highest proportions of executive jobs and qualified technical professions, but it also experienced the highest growth in relative and absolute terms.The decline in low-skilled manual professions -in the service sector as well as in industry -was also more pronounced there in comparison with the Belgian average.In all of the cities, this qualification process was also more significant in the centre than in the outskirts of the employment areas.
These results must, however, be interpreted with caution.On the one hand, as we have already underlined, the classification of professions used is not based on a direct measurement of their level of qualification.On the other hand -and more crucially -the nature of professions evolves with time.Capitalism is marked by a permanent contradictory movement of qualification levels -with the increase in management functions for example -which conflicts with deskilling, resulting from the segmentation and automation of tasks.
Given the growing requirement for skills in the labour market, low-skilled people have more and more difficulties fitting in.This could already be deduced from Figure 4.More complex statistical analyses confirm that education level is the most significant variable to explain a person's situation in the labour market (cf.

Growth generating few jobs
The situation is worsened by the fact that economic growth in Brussels generates very few jobs (Table 1).Table 1 (last column) shows that 1% of economic growth in the Brussels employment area only leads to a 0.59% increase in employment, which is lower than the figures for the other cities.This is even truer at the level of the Brussels-Capital Region, where this figure falls to 0.5%: in other words, employment rises twice as slowly as added value.This specific feature of Brussels -which is also observed in the centre of Antwerp -is explained among others by the increasing employment qualifications in Brussels, which generate a low number of highly paid jobs.

The importance of the spatial dimension
The social polarisation of Belgian cities must also be understood in its complex relationship with sociospatial polarisation, i.e. the gaps between neighbourhoods from a social point of view.
Our study shows growing sociospatial polarisation, at least when it is measured using the unemployment rate indicator.The gaps between statistical sectors (neighbourhoods) as regards unemployment increased in the five major cities in the country between 1981 and 2006, but were at a much higher level in Brussels.If we focus only on the disadvantaged neighbourhoods, 9 the unemployment rates were structurally increased, even when the global situation of the labour market was improving throughout the employment area (Table 2).
Sociospatial polarisation must be understood in its dialectic relationship with social polarisation.On the one hand, in disadvantaged neighbourhoods, the characteristics of the population (young, often of foreign origin and with a low level of education) may explain the increase in unemployment, as these populations are confronted with growing difficulties in the labour market.But, on the other hand, the concentration of disadvantaged populations is likely to make integration into the labour market more difficult.This process is known as the location effect, which assumes that people -with the same characteristics -who live in a disadvantaged neighbourhood are more likely to encounter difficulties in the labour market than those who -with the same characteristics -live in more privileged neighbourhoods.This reality is highlighted by our 'disadvantaged neighbourhoods' Brussels Studies the e-journal for academic research on Brussels 12 9 Here we refer to disadvantaged neighbourhoods as defined by Kesteloot et al. (2006).
G. VAN HAMME, I. WERTZ, V. BIOT, « Economic growth devoid of social progress: the situation in Brussels », Brussels Studies, Issue 48, 28 March 2011, www.brusselsstudies.be4) which shows that the residents of these neighbourhoods in Brussels are twice as likely to be unemployed than the other residents in the employment area of the same age, sex, education level, household composition and nationality.However, we cannot conclude with certainty that a location effect comes into play with respect to the spatial concentration of disadvantaged populations, in particular because we were not able to verify the social and ethnic origin of individuals (for a more systematic analysis which overcomes these difficulties and confirms our results, see Dujardin et al., 2008).Furthermore, we were not able to consider population exchanges which may be due to the fact that unemployed people move to disadvantaged neighbourhoods and employed people move out of them.The GROSE study (2010) demonstrated the existence of this process (around 2001), although its extent does not seem to explain the significance of the location effect observed in the framework of our study.
Furthermore, the existence of a location effect is all the more plausible since it appears to be stronger in places where sociospatial polarisation is most pronounced, i.e. in Brussels.The difference in income between neighbourhoods (statistical sectors) is much higher in Brussels than in the other cities of the country: the coefficient of variation reached 0.27 in the Brussels employment area, whereas it was between 0.17 and 0.22 for the four other employment areas in 2001.
Finally, we must underline the importance of the opposition between the centre and the outskirts, in Brussels as well as in the other cities.This effect is described as a 'spatial mismatch', a term which is most often used to describe the situation in North American cities.It is the result of the contrast between the creation of employment in the outskirts and the concentration of disadvantaged populations in the centre.In Brussels, this effect is intensified by the linguistic border and the fact that it is difficult for the low-skilled populations in the centre to access employment created in the Flemish outskirts.However, the difference between centre and outskirts should not be overestimated.Our study shows that the outskirts also experienced a decline in low-skilled employment, especially around Brussels.In other words, it is misleading to believe that a solution to the unemployment of low-skilled people from the centre exists in the dynamism of the outskirts, as this mechanism could only have a marginal effect.
Brussels Studies the e-journal for academic research on Brussels 14

Discussion and conclusion
Since the 1990s, the nature of socioeconomic problems in Brussels has changed somewhat.Like many major European cities, Brussels -defined by its employment area -has therefore left most of the industry crisis behind.It has regained economic performance above the national average as well as that of the other Belgian cities, with the exception of Ghent.Even the Brussels-Capital Region is faring better than the other urban centres, although with markedly lower dynamics than in its outskirts.
However, social indicators continue to deteriorate: unemployment is rising -even more so in the disadvantaged neighbourhoods -social inequalities are growing and the fragmentation of space is becoming more pronounced.
The present article attempts to provide some theoretical and empirical elements to explain these paradoxical evolutions.Firstly, the contemporary growth model appears more and more to generate exclusion as well as social inequalities (Kesteloot & Loopmans, 2009).It involves a significant break with the post-war growth model which generated full employment and was accompanied by a considerable decrease in social inequalities.For Brussels, we have shown the very strong growth in skilled employment and the decrease in low-skilled manual employment.In accordance with Hamnett's thesis, during the 1991-2001 period, this resulted in the growing exclusion of low-skilled people from the labour market.But this mechanism alone probably does not account for the complexity of the evolutions observed; this is why we have proposed a more general theoretical framework which takes into consideration the huge changes in the socioeconomic regulation of capitalism since the beginning of the 1980s (the 'third thesis').Low-skilled employment created in the service sector stands out due to its precariousness and feminisation, as well as the low level of organisation of the employed workers (social sector, retail trade, hospitality business, construction, etc.) (Marissal et al., 2006).This results in a general decline in the situation of low-skilled workers.Let us also point out that this growth model generates few jobs, with the result that in the Brussels-Capital Region a 1% growth in added value only leads to a 0.5% growth in employment.Secondly, we have also shown that the growing sociospatial inequalities add to the difficulties of people who live in the most disadvantaged neighbourhoods.All things being equal in terms of nationality, age and education level, they have a much higher likelihood of being unemployed.Consequently, the very high spatial concentration of people in difficulty in Brussels -at a much higher level than in the other Belgian cities -probably increases the social difficulties.
Faced with these observations, it is necessary to question the political responses, in particular those of the Brussels-Capital Region.Above all else, it must be pointed out that the regional policies only have a limited hold on the social question in Brussels: the main political instruments are situated at other levels (social redistribution, orientation of the economic policy, etc.), the budgets are limited, and the political decision space does not correspond to the reality of the economic space in Brussels.
Nevertheless, the options considered in the narrow framework of the Region's margins for manoeuvre must be discussed in the light of the above observations.On the one hand, it must be acknowledged that discourses on economic development and international competitiveness are very present, as illustrated by the International Brussels Studies the e-journal for academic research on Brussels

15
Development Plan (IDP), among others.One may question the necessity of this type of priority for a city which in fact enjoys high economic performance and an important position in international urban networks.The policies aimed at reinforcing these strong points probably have only a very marginal effect.Moreover, our analysis suggests that this economic growth 'at a higher level' -based in particular on the highlevel services related to the international position of Brussels -reinforces the social inequalities within Brussels.The underlying idea of the trickle-down effect, i.e. that economic development at a higher level benefits everyone in the end, does not seem to withstand the examination of the facts in Brussels and, furthermore, has never been proven by those who uphold it (OECD, 2006).
On the other hand, unemployment is still too often interpreted in terms of a discrepancy between the labour offer and demand.According to Agoria Bruxelles (an employers' organisation), the situation is clear: 'whilst employers seek on average 33% higher education graduates, 30% upper secondary school graduates, and 37% primary or lower secondary graduates, the profile of unemployed people does not correspond to this, with the proportions being 11%, 19% and 70% respectively."As long as this gap continues to exist, the unemployment rate will not fall below 20% in Brussels" warns Dominique Michel' 10 (quoted on the website http://www.altereduc.be, in the news from 12 to 26/10/2004).Such reasoning is biased inasmuch as -in a context of structural unemployment -the recruitment of labour takes place to the detriment of the least skilled people (Marissal et al., 2006;Baudelot and Establet, 2000;Devillé, 2008).In other words, the most skilled workers are recruited first to fill the available positions, even when their qualifications are not required for the jobs in question.It is therefore more pertinent to compare the qualification profile of available jobs with that of workers on the whole, and not with that of unemployed people only.As employers hire the most skilled workers first, unemployed people are always much less skilled than the rest of the population, regardless of the actual needs of the labour market.These processes of cumulative deskilling therefore come in addition to the inadequate training of part of the population, thus emphasising the discrepancy between the qualification profile of unemployed people and the profile of available jobs.However, the public authorities do not hesitate to adopt the employers' discourse: 'One of the main difficulties in terms of the reduction of unemployment results from the discrepancy between the profiles of job seekers and the needs of companies' ('Contrat pour l'économie et l'emploi' (Contract for Economy and Employment) of the Brussels-Capital Region, 2005, p. 39).And now, training for workers has increased considerably in Brussels, economic dynamics have been relatively favourable, and unemployment has continued to follow an upward trend.It follows that unemployed people are also increasingly skilled: according to the Brussels employment survey, the proportion of unemployed people with at least an upper secondary diploma rose from 25% to 33% between  1992 and 2009. 11The idea that training policies are useless is far from us, but they are a very inadequate response to the social question.
Brussels Studies the e-journal for academic research on Brussels 16 10 President of Agoria Bruxelles.
11 According to the labour force survey, these figures are much higher, rising from 41% to 58% of unemployed people (ILO) during the same period, between 1992 and 2009.This difference is in particular related to the narrower definition of unemployment (ILO).
Faced with recent economic and social evolutions, a reorientation of urban policies appears to us to be necessary.We thus recommend a more integrated policy to confront these challenges, at individual level as well as in terms of development policies at regional level.On the one hand, at individual level, training policies have many difficulties to compensate for the initial deficiencies in training.We feel that it would therefore be judicious to elaborate social policies at individual level by including training policies, without limiting them to that.This would require more continuous and integrated support for individuals who have major difficulties with respect to the labour market.This type of policy exists in Scandinavian countries and obviously involves high costs.On the other hand, urban development policies would probably gain by not concentrating on the high-level sectors whose economic health is certain, but which are at the origin of an increasingly unequal social space.However, the accent could be placed on sectors which allow low-or semi-skilled employment to be provided, social ties to be recreated (retail trade, child care, social services, etc.) and genuine needs to be met.
journal for academic research on Brussels 5 G. VAN HAMME, I. WERTZ, V. BIOT, « Economic growth devoid of social progress: the situation in Brussels », Brussels Studies, Issue 48, 28 March 2011, www.brusselsstudies.be

Figure 1 :
Figure 1: Definition of metropolitan areas

Table 4 :
Individual determinants (exp b 1 ) of unemployment per employment area in 2001, according to binary logistic regression.Source: 2001 socioeconomic survey Note: The asterisks* indicate the statistical significance of each result.*** = 99% significant; ** = 95% significant; * = 90% significant; no asterisk = insignificant.1 Exp (b) may be interpreted as the likelihood of being unemployed with respect to a baseline equal to 1. Consequently, when exp b is higher than 1 for a variable, it increases the likelihood to be unemployed with respect to the baseline.This is all the more true when the value of exp b is high.

Employment area of Liege 1,68 1,09 0,65 Rest of the country 2,20 1,38 0,63
Table 4 below).Unfortunately, a rigorous comparison of the situation in 1991 and 2001 is not possible due to problems related to the measurement of education level in the two censuses used.

Table 3 :
The evolution of professions according to their qualifications in the major employment areas inBelgium, between 1991 and 2001.Source: 1991 population census; 2001 Socioeconomic Survey; INS (National  Statistical Institute). the e-journal for academic research on Brussels 13 G.VAN HAMME, I. WERTZ, V. BIOT, « Economic growth devoid of social progress: the situation in Brussels », Brussels Studies, Issue 48, 28 March 2011, www.brusselsstudies.be