1The Brussels-Capital Region is completely surrounded by the province of Flemish Brabant. Consequently the interaction between both areas is intense and of long date. Suburbanization has been important for several decades: many young families move from the Brussels Capital Region to Flemish Brabant. As this migration often concerns higher income groups, it impoverishes the Brussels municipalities and increases the prosperity of Flemish Brabant. The number of moves in the opposite direction is much less pronounced, but there is a large commuter flow to the Brussels Capital Region. Many inhabitants of Flemish Brabant travel daily to the Brussels municipalities in order to work, but also for other social and economic activities.
2This contribution examines the extent of the structuring role the suburbanization from the Brussels municipalities is still playing in the relations between the Brussels Capital Region and Flemish Brabant, and if its content is changing. How large is the migrant group at present? Is it still mainly composed of young families? Does the nationality (at birth) of the migrants reflect the demographic evolution in the Brussels Capital Region? Are these really high income groups and what is their position on the labour market? We also explore differences between municipalities. If possible, we drill down to the level of the districts, which correspond to the old municipalities before the mergers of the 1960s and 1970s.
- 1 This article is based on “Dossier Wisselwerking Flemish Brabant en Brussels”, a study of the Steunp (...)
3The movement in reverse, the commute to the Brussels Capital Region, is also studied. Here we highlight the importance of the daily commute to work, but primarily examine the commute in other, much less documented areas: the commute to school, child care or several other social services.1
4This contribution focusses on the part of Flemish Brabant having the strongest interaction with the Brussels Capital Region. In a number of other municipalities in Flemish Brabant the influence of the Brussels Capital Region is indeed limited, mostly in the eastern part of the province, where the city of Louvain exerts its influence. We can read this on the map of the Brussels Capital Region drawn by Luyten & Van Hecke [2007] (figure 1). This urban region encompasses not only the agglomeration municipalities (contiguous development from the central city outwards) but also numerous other municipalities which were added to the urban region based on several indicators: the evolution of built surface and population, the migration from and towards the city, the work and school commute and the income ratios.
- 2 This also appears from other studies mapping the urban area surrounding Brussels, for example the s (...)
5The study area includes the Flemish Brabant part of this urban region, with the addition of the municipalities Liedekerke and Affligem in the extreme west of the province (figure 1). In recent years there has been a strong migration dynamic between these two municipalities and the Brussels Capital Region. We will call this area the “large periphery”. The term “Brussels” refers to the (administrative) Brussels Capital Region, unless indicated otherwise. The Brussels urban region also comprises a number of municipalities in Walloon Brabant.2 These municipalities are not included in our study, due to a lack of source material.
Figure 1. The Brussels urban region and the demarcation of the “large periphery”
Source: Luyten & Van Hecke, 2007
- 3 A move or migration is defined as a change of (official) domicile between two consecutive years, ea (...)
- 4 The migration balance between Brussels and the province of Flemish Brabant as a whole is not much l (...)
6The population evolution of the large periphery is strongly influenced by the presence of Brussels (table 1). In the years from 2008 to 2012, the migration balance with Brussels (the number of migrants3 from Brussels minus the number of migrants to Brussels) was more than 33 000 individuals4, or each year almost 12 individuals per 1 000 inhabitants. By way of comparison: in the same period, natural growth (the number of births minus the number of deaths) increased the number of inhabitants of the large periphery with 4 796 people. This averages to 1.7 individuals per 1 000 inhabitants per year. The migration balances with other Belgian regions are negative. More than 6 000 individuals move from the large periphery to the rest of Flemish Brabant rather than in the opposite direction. The migration balances with the rest of the Flemish Region and the Walloon Region are in the same order of magnitude. In the Walloon Region, the Walloon Brabant municipalities predominate.
Table 1. Migration movements to, from and within the large periphery (2008‑2012)
Source: Rijksregister, processing Steunpunt sociale planning
7In the large periphery, the growth from foreign countries is limited. In the period 2008‑2012 it amounted to 4 000 individuals, representing a yearly increase of 1.5 individuals per 1 000 inhabitants. By way of comparison: in Brussels, the migration balance with foreign countries was more than 100 000 individuals. Like other regional capitals, Brussels clearly has a reception function for foreign migration.
8Although population increase in the periphery is mainly determined by the inflow from Brussels, the number of moves between and within the municipalities of the large periphery is even larger. These movements do not impact the global population of the large periphery, but do change the composition of the population of a given municipality (in the case of moves between municipalities) or a district (in the case of moves within municipalities).
9The numbers used, reflect a five year period, in order to neutralize coincidental variations between the years. Non-recurrent events, such as new real estate developments, can also have a substantial impact, most certainly on the municipal or district level. The global migration balance in the large periphery however appears to be very constant during the period under review. In the longer term, the numbers are slightly higher than in the 1980s or 1990s, but much lower than in the period predating the oil crisis. This can be observed clearly in figure 2, showing the population evolution in the large periphery in the 1954‑2012 period. The total increase is mainly determined by the migration balance, which was much higher in the Fifties, Sixties and Seventies than in the present period. Even then the migration from Brussels was decisive [Van der Haegen, 1987].
10Figure 2 also shows the evolution of average income (as per income statement) in relation to the average income for the entire former province of Brabant. This indicator increases continuously in the large periphery, from far below the average (<1) to markedly above the average (>1). In Brussels (figure 3) the trend is reversed – despite a continuous outflow to the periphery, the population increased in the beginning and at the end of the period under review. This can be explained by migration from foreign countries.
Figure 2. Population and income evolution in the large periphery, 1954‑2012
Source: NIS/ADSEI, 1955‑2014
Figure 3. Population and income evolution in Brussels, 1954‑2012
Source: NIS/ADSEI, 1955‑2014
11The impoverishment of Brussels and the growing prosperity of the periphery suggest that mostly higher income groups are leaving the city, as observed in other studies [Loeckx et al., 2014]. The departure of mainly young families is also a classic suburbanization characteristic [Willaert, 2009]. Some recent studies have also shown that more and more individuals of non-Belgian origin are moving from Brussels [De Maesschalck, 2012]. In the following section, we will examine the migrant profiles in detail and look for changes in recent years.
- 5 Strikingly, female migrants are on the average younger than males and strongly concentrated in a nu (...)
12In the period 2008‑2012 mainly young families leave Brussels. The number of parents and underage children moving from Brussels far exceeds the average of those living in Brussels and the large periphery. Figure 4 (upper left) shows the population pyramid of the migrants leaving Brussels, being mostly people in their twenties or thirties and young children. On the other hand, large numbers of singles and children older than 18 (home leavers) move from the large periphery to Brussels. Figure 4 (upper right) illustrates that this mainly concerns young adults. The net result of these movements (figure 4, bottom) is an increase of mainly young people in their thirties and young children. This results in a decrease of the age average in the large periphery. Its average inhabitant is 41 years old, the average age of the last population pyramid is 25 years.5
Figure 4. Population pyramids of the migrants from Brussels to the large periphery, from the large periphery to Brussels and net balance in the large periphery
Source: Rijksregister, processing Steunpunt sociale planning
- 6 A subdivision has been made between Belgium, other countries of the EU (divided in three groups: No (...)
13Whereas this age and family profile has been mostly stable throughout the years, a different demographic variable has changed: the nationality at birth (hereafter “origin”). In figure 5 (top) we see the net growth in the large periphery from Brussels, subdivided according to origin.6 The number of individuals of Belgian origin decreases each year and is now less than half of the total population. Most striking is the increase in Eastern European origins. In figure 5 (bottom) children moving with their parents have been given the nationality at birth of the head of the family. This means that some children born as Belgians can be assigned a non-Belgian origin. Here also we notice a decreasing trend for individuals of Belgian origin, who now represent less than one third of the increase from Brussels. Although here also the increase of Eastern European origins is strongest, the much larger share of Northern African, Turkish and other non-EU origins is striking. These are also the nationality groups most often taking the Belgian nationality [Steunpunt sociale planning, 2010].
Figure 5. Nationality at birth of the net inflow from Brussels
Source: Rijksregister, processing Steunpunt sociale planning
14Therefore increasingly more people of non-Belgian origin are moving from Brussels, whereas the age and family positions do not change much in time. This means that the suburbanizers of non-Belgian origin fit the classic profile: mainly young families with children. Figure 6 shows the age distribution of the net inflow from Brussels to the large periphery for various origins. We can observe that the curve is indeed similar for all groups. The age differences are however slightly more pronounced for the Belgian origin, the Northern and Western European origin and the non-European origin from wealthy OECD countries.
Figure 6. Age distribution of the net inflow from Brussels according to origin, 2008‑2012
Here also, children moving with their parents were given the nationality at birth of the head of the family
Source: Rijksregister, processing Steunpunt sociale planning
- 7 Deboosere et al. [2009] calculated that 46% of the inhabitants of Brussels had a non-Belgian nation (...)
15Inhabitants of different origins do not have fundamentally different life ambitions, including the suburbanization aspiration. In the Belgian context this has not yet been specifically studied, but in other European countries it has [Clapson, 2000; Permentier & Bolt, 2006]. In Belgium, the suburbanization of migrants has long been blocked by the economic crisis of the 1970s and 1980s, which mainly hit the groups in the centre of the city [Kesteloot, 2001]. At present, suburbanization no longer appears to be a phenomenon restricted to the Belgian middle classes. The migrant’s origin more closely reflects the population composition of Brussels. We have no detailed figures for origins in Brussels7, but nationality statistics do exist. Figure 7 shows the population composition according to nationality in Brussels, for migrants from Brussels to the large periphery, for migrants from the large periphery to Brussels and within the large periphery. Here we can see that migrants from Brussels have slightly fewer non-Belgian nationalities than the Brussels average, but that the difference is not very large. It is also notable that migrants from the large periphery to Brussels much more often have a non-Belgian nationality than the average for the large periphery - although less often than the migrants from Brussels.
Figure 7. Nationality of migrants to and from Brussels and of inhabitants of Brussels and the large periphery (2008‑2012)
Source: ADSEI, 2014; Rijksregister, processing Steunpunt sociale planning
- 8 The reference individual gets a value of 1, each extra family member of 14 years of age or older a (...)
16Although suburbanization evidently results in an impoverished city and a more prosperous periphery, in reality not much is known about the (evolution) of the socio-economic position of the migrants themselves. The number of studies directly mapping the socio-economic profile of migrants is extremely limited. The exceptions are some publications using data from the 2001 socio-economic survey of [Van Criekingen, 2006; Van Criekingen et al., 2013]. We have used the more recent data of the Kruispuntbank van de Sociale Zekerheid. We have divided the Belgian population in four income groups or quartiles: the 25% of inhabitants with the highest incomes, the 25% with the lowest incomes and the two quartiles in between. This concerns incomes from work and benefits; we do not have information on moveables and immoveables. It also concerns the equivalent income: all incomes of a family are added and divided by a factor taking into account the composition of the family and possible economies of scale.8 The result is allocated to each family member.
17We see the result in figure 8, illustrating the distribution over Belgian income quartiles for the inhabitants of Brussels, the migrants from Brussels to the large periphery, the inhabitants of the large periphery and the migrants from the large periphery to Brussels. The lowest income quartile is strongly overrepresented in Brussels, the reverse is the case for the highest income quartile. The profile of the migrants to the periphery does not reflect this: the lowest income group is less represented, the highest more. In the large periphery we observe the reverse. It has relatively few low income and many high income inhabitants, but for the migrants the lowest income group is the largest and the highest is the smallest.
Figure 8. Equivalent income distribution (2007‑2010)
Source: KSZ DWH AM&SB, processing Steunpunt sociale planning
18In figure 9 on the upper part we see the net result of these movements, the balance. This shows that mainly the highest income group grows strongly in the large periphery, while the lowest income group grows least. On the bottom part, the total migration volume is shown. This is the addition of all movements, both from Brussels to the large periphery and the other way around. We observe that the number of movements is much larger for the low incomes (often renters). The balance however is relatively small, as they often move to Brussels.
Figure 9. Migration movements between Brussels and the large periphery according to equivalent income (2007‑2010)
Source: KSZ DWH AM&SB, processing Steunpunt sociale planning
19Here also there are few differences in time, while we know that the share of non-Belgian individuals does increase strongly. As with the age and family position, the group of non-Belgian origin reflect the classic suburbanization pattern: mostly high incomes leave the city. The average income is higher for individuals of Belgian origin than for non-Belgian origins, with non-EU origins scoring the lowest average income. However within these groups we observe that migrants from Brussels on the average always have a higher income than the people remaining in the city.
20For the position on the labour market, we observe the same trend as for income. The migrants from Brussels number more employed and children than the average for Brussels, and fewer unemployed and inactives (benefits, disability…). On the other hand, more unemployed and inactives move to Brussels than the average for the inhabitants of the large periphery. Very few pensioners move between Brussels and the large periphery. These movements result in a strong growth in the large periphery of the employed (45%) and children (37%). For the remaining 12% the position on the labour market is not yet known. This concerns individuals without any link with Belgian social security: househusbands and housewives, EU employees and other groups not covered by Belgian social security.
21The migration balance (and the growth through migration) is especially large in the municipalities bordering Brussels and foremostly in the districts nearest Brussels. In the west, the influence extends further, reaching the border of the province. The highest values are found in the southwestern valley of the Senne (Drogenbos, Ruisbroek, Lot, Beersel), in the northeast (Machelen, Diegem, Sint-Stevens-Woluwe), and in the northwest (Strombeek-Bever, Wemmel, Zellik). In the southeast, values are somewhat lower. Here the exchange with Brussels is less important and the number of immigrations is often more in balance with the number of emigrations (figure 10).
Figure 10. Average yearly migration balance with the Brussels Capital Region, 2008‑2012
Source: Rijksregister, processing Steunpunt sociale planning
- 9 The city of Brussels numbers 166 497 inhabitants, the north-eastern municipalities 191 890 inhabita (...)
22Looking at the Brussels municipality of origin, we notice very large spatial differences. Migrations indeed most often cover small distances. Figure 11 shows the migration balances for different groups of Brussels municipalities, composed according to their location. The groups in the northeast, southwest and northwest each number about 200 000 inhabitants. A fourth group is composed of the southeastern municipalities, totalling about 350 000 inhabitants. The city of Brussels itself is shown separately, because of its central location, and numbers the fewest inhabitants.9 Each time we observe that the large majority of migrants move to more or less neighbouring districts in Flemish Brabant. They are often also socio-economically and morphologically close. The housing market in the southeastern municipalities, for example, is often comparable to that on the other side of the regional border. The same goes for a number of municipalities in the valley of the Senne.
Figure 11. Average yearly migration balance with various municipalities of the Brussels Capital Region, 2008‑2012
Source: Rijksregister, processing Steunpunt sociale planning
- 10 A linear regression gives a R² value of 0.51.
- 11 A linear regression gives a R² value of 0.72.
23Although the group of southeastern Brussels municipalities numbers more inhabitants than the other groups, migration here is less pronounced. There are more home owners, who move less often than renters. An important factor however is population growth. This has been very limited in the southeastern municipalities, but very strong in a number of municipalities along the Canal. Watermaal-Bosvoorde for instance experienced a negative population growth in the last 10 years, while the population of Sint-Jans-Molenbeek increased with more than a quarter. More than half of the differences between Brussels municipalities, regarding the migration balance with the large periphery in the period 2008‑2012, are explained by the population growth in the last 10 years.10 In some municipalities the outflow was smaller than could be expected based on the population increase. This was the case in Sint-Jans-Molenbeek, where population growth was strongest, but also in Etterbeek, Elsene and Sint-Gillis, which have a relatively large inflow of young (high educated) adults from the periphery - linked to the presence of higher education. When we exclude these municipalities, the differences between Brussels municipalities are explained for more than 70%.11
- 12 Only for the group of 18‑29 years olds there are large spatial differences: in the southeastern per (...)
24The profile of the Brussels population varies significantly between municipalities. This also impacts the profile of the migrants and therefore the inflow in the neighbouring municipalities. It mainly concerns young families12, but there are large differences according to origin. Many individuals of Southern European origin, a significant group in the southwest of Brussels, move to the southwestern Senne valley. Many individuals of Turkish origin, strongly concentrated in the northeast of Brussels, move to the northeast Senne valley. In general few individuals of non-EU origin move to the southeast.
25We also observe large differences regarding income. Figure 12 illustrates the distribution over Belgian income quartiles of migrants per district. Whereas the southeast mainly receives higher incomes, we find many low incomes in several municipalities in the Senne valley, in the northeast and the west. This does represent an overwhelming majority of employed. Figure 13 shows the position on the labour market of migrants of active age (excluding children, pensioners and unknown). In all municipalities it concerns almost exclusively employed.
Figure 12. Average yearly migration balance with the Brussels Capital Region according to equivalent income, 2007‑2010
Source: KSH, Rijksregister, processing Steunpunt sociale planning
Figure 13. Average yearly migration balance with the Brussels Capital Region according to position on the labour market, 2007‑2010
Source: KSH, Rijksregister, processing Steunpunt sociale planning
26Brussels is by far the largest employment pool of the country, therefore attracting many commuters [Medina Lockhart & Vandermotten, 2009]. In 2011 more than 227 228 inhabitants of Brussels worked in Brussels, compared to 381 394 non-inhabitants. More than one fifth of the latter come from the large periphery (83 380 commuters). This only regards employees, as self-employed individuals are not taken into consideration. These core figures come from the Kruispuntbank van de Sociale Zekerheid, which means that foreign diplomats, EU officials and employees of NATO and other international institutions are not included. As most of these institutions are based in Brussels, the aforementioned data are a lower limit.
27The commute flow in the other direction, from Brussels to the large periphery, is limited. In 2011, 63 216 inhabitants of Brussels worked outside the city, of which 40 percent in the large periphery (25 183 commuters). The airport area in particular attracts many inhabitants of Brussels. Zaventem and Machelen are the only municipalities in the large periphery attracting more inhabitants of Brussels than the number of their inhabitants working in Brussels. Vilvoorde welcomes more employees from Brussels than Machelen in absolute figures, but has an even larger flow towards Brussels. Together with the municipalities with language facilities, Zaventem, Vilvoorde and Machelen have the largest commute volume - the sum of incoming and outgoing commuters.
- 13 Wemmel, Lennik and Drogenbos do have a relatively large inflow from Brussels in primary education, (...)
28Next to the work commute, there exists an important school commute to Brussels. Figure 14 (top) show the school commute in the education system of the Flemish Community. All flows between Brussels, the large periphery, the rest of the Flemish Region and the Walloon Region are included. The exchange with the Walloon Region is unidirectional, as the Region does not organise education in Dutch. We observe that more than 7 500 children from the large periphery visit Dutch-speaking schools in Brussels, while fewer than 3 500 children take the other direction. This means a net flow to Brussels of more than 4 000 pupils. This negative balance applies to almost all municipalities of the large periphery. The larger municipalities Dilbeek and Halle are notable exceptions, but in these cases the movement never concerns more than 5% of the pupil population. We find these flows primarily in secondary education; cross-border movements in primary education are relatively limited.13
Figure 14. Pupil flows in the school system of the Flemish Community (left) and of the Flemish Community and French Community in Brussels (right), 2009‑2010
*Only the total number of pupils from the Flemish Region enrolled in the school system of the French Community in Brussels is known. These pupils were distributed over the large periphery and the remainder of the Flemish Region based on the distribution in the school system of the Flemish Community in Brussels (87,3% from the large periphery and 12,7% from the remainder of the Flemish Region).
Source: Departement Onderwijs en Vorming ; IBSA ; Data processing: Steunpunt sociale planning
29On the other hand, the number of pupils living in Flemish Brabant who commute to Brussels is decreasing slightly from year to year. The growth of the number of pupils in Brussels Dutch-speaking education is therefore entirely due to the increase of the number of Brussels pupils. The growth of the number of pupils in the large periphery can largely be explained by the increase of the number of pupils in the large periphery itself: the increase of the number of Brussels inhabitants is rather limited.
30Of course this is only part of the story. 83% of the pupils in Brussels schools is indeed enrolled in the school system of the French Community (not counting the European and international schools). For pupils living outside Brussels but visiting schools in Brussels, this is 72%. In figure 14 (bottom), the Brussels French-speaking schools are added. The flow from Brussels to the large periphery stays the same – as the French Community does not organise education in the periphery – but the flow in reverse now counts almost 27 000 pupils. This results in a net flow of almost 24 000 in the direction of Brussels. In most municipalities with facilities, a majority of pupils is enrolled in French-speaking schools.
- 14 In Brussels this is 31%, in the large periphery 36% and in the Flemish Region 40%.
31Next to work and school commuting, many other movements cross the borders of the region, such as preschool child care. In order to measure this, the Steunpunt sociale planning, in partnership with Kind en Gezin, held a survey with child care initiatives recognized or supervised by Kind en Gezin. The survey assessed the domicile of the children, which is not (yet) being centrally registered. The results are shown in figure 15, constructed in the same manner as figure 14. This figure is corrected for the response rate, which was higher in the large periphery and the remainder of the Flemish Region (93% of capacity) than in Brussels (70% of capacity). It shows a net stream of more than 900 children from the large periphery to Brussels. This mainly concerns movements close to the regional borders. The number would be even larger if we added the French-speaking offer, as this only exists in Brussels, but data on this offer are lacking. Still, the offer in Brussels is relatively small compared to the number of children living in Brussels. It is below 33% of the Brussels 0‑2 year olds, the so-called Barcelona norm. In the periphery this norm is reached, but the offer is smaller than in the Flemish Region as a whole.14
Figure 15. Flows in child care, extrapolated to capacity (03‑09/02/2014)
Source: Survey on the domicile of children in care; Capacity data: Kind en Gezin
32In various others social services we also observe, in the Dutch-speaking offer, a flow towards Brussels, often combined with a limited offer. The Centra voor Algemeen Welzijnswerk (CAW), subsidized by Flanders, are an example of this phenomenon. Figure 16 shows the domicile of the clients of Brussels CAWs and the CAWs of Flemish Brabant. For the latter, a distinction was made between the establishments in the large periphery and in the rest of Flemish Brabant. We observe that many inhabitants of the large periphery are being assisted in Brussels, but that the number of inhabitants of Brussels with CAWs in Flemish Brabant is very small, both in the large periphery as elsewhere. This is partly due to the specialized offer in Brussels in the period reviewed, but even without this offer we find a net flow towards Brussels. This coincides with a very limited offer in the large periphery.
Figure 16. Domicile of individuals assisted by the Brussels and Flemish Brabant CAWs, 2008‑2011
Source: Steunpunt Algemeen Welzijnswerk, TELLUS client registration, processing Steunpunt sociale planning
33The population growth in the Brussels periphery is primarily determined by migration from Brussels. This often concers young families, as has been the case during the previous decades. This suburbanization impoverishes the city and makes the large periphery more prosperous. This observation is not new, but the socio-economic profile of the migrants has hardly been studied. This study has done so and confirms the image of the migrant having a higher income than the Brussels average, although this does not apply to all districts and municipalities of the large periphery. The observation that the large majority of migrants is active on the labour market however applies to all municipalities. It is also notable that this more often concerns individuals of non-Belgian origin. Whereas for many years suburbanization was almost exclusively limited to middle class people of Belgian origin, this no longer applies. More than previously the migrants from Brussels reflect the composition of the Brussels population, at least in this regard. In the group of non-Belgian origin, young families and higher income groups are also overrepresented.
34This continuous inflow into the large periphery is contrasted by a daily commute flow towards Brussels, for work, child care and numerous other social services. The evolution of these movements is not unequivocal and requires much additional research. The growing Brussels employment is increasingly taken up by inhabitants of Brussels itself, while commuting remains relatively constant. We have previously seen that the number of pupils from the large periphery following Brussels Dutch-speaking education is decreasing, while it strongly grows in the periphery. Does this mean that the inhabitants of the periphery and especially the migrants from Brussels are less focussed on Brussels than in the past? Or can it not be separated from the capacity problems in Brussels schools [Wayens et al., 2013]? The limited offer of social services in Flemish Brabant also influences the movements towards Brussels. It is however difficult to judge if an expansion of the social offer in the large periphery, which is presently far below the Flemish average, would indeed decrease these movements.
35As this publication maps the movements between Brussels and the larger periphery, it requires some qualifications. Brussels itself experiences a suburbanization movement from the central to the peripheral municipalities, as the city’s administrative borders do not coincide with its social borders. The differences between Brussels and the larger periphery are often more significant than those between neighbouring areas in Brussels and the larger periphery. This is shown for instance on figure 17, illustrating the number of pupils (in the schools of the Flemish Community) receiving scholarships because of low income. We notice higher values in the valley of the Senne, from Halle through Brussels up to and including Vilvoorde, whereas the east and southeast of Brussels approach the low values of the (south)eastern periphery. We have also seen that the migration movements between the southeast of Brussels and the neighbouring areas in Flemish Brabant mainly concern high incomes, which does not always hold true for other areas. This also we must keep in mind: an administrative border is not a real social border.
Figure 17. Pupils in primary and secondary education with scholarships (school year 2011‑2012)
Source: Departement Onderwijs en Vorming, BISA, processing Steunpunt sociale planning