Introduction 

1Along with situations of insalubrity, overcrowding and excessive rental charges, housing evictions are an important dimension of inadequate housing in contemporary cities. They have a major impact, above all at an individual level. In a recent survey on the consequences of eviction in France, the Abbé Pierre Foundation showed that one to three years after an eviction, one household in three still had no roof over their head and was still living in a precarious or temporary form of housing (with a third party, in a hotel, in a mobile home, at a campsite, in a squat) or on the street. The survey also showed that, following an eviction, 29 % of people were unable to continue working, 43 % of households with children faced difficulties in school, and 71 % of households suffered health or psychological problems [Portefaix and Rothhahn, 2022]. Evictions are also very costly for the community, in terms of mobilising the judicial and social welfare systems as well as in economic terms. In this regard, eviction prevention pilot projects in Austria and Germany have shown that € 1 spent on housing maintenance saves € 7 in emergency shelter and social integration costs [Kenna et al., 2016].

2However, very little is known about the reality of housing evictions, in terms of numbers and spatial distribution as well as in terms of the profiles and paths of the people concerned. Already at the beginning of the 2000s, two researchers referred to eviction as a “hidden housing problem” [Hartman and Robinson, 2003]. Although the subprime mortgage crisis of 2008 drew a lot of media, political and academic attention to this issue on an international scale, there have been few publications regarding the situation in Belgium. For example, in 2015, two researchers noted the very limited knowledge available on the issue of housing evictions in Flanders, and pointed out that this lack of knowledge was even greater in the two other regions of the country [Verstraete and De Decker, 2015].

4This article is a response to this lack of knowledge on evictions in Brussels, and aims to contribute to a greater visibility of this issue among researchers as well as institutional and civil society stakeholders. In order to do this, we propose a well-thought-out, quantified and spatialised inventory of housing evictions in the Brussels-Capital Region. In the first section, we present the data collection strategy implemented for this purpose as well as our methodological choices. The second section presents and analyses the results obtained, in an attempt to answer several questions: How many evictions? Are some parts of the city more affected than others? What is known about the profile of evicted tenants and landlords who have called for an eviction?

1. Methodology

1.1. Defining evictions

6In this sense, there are two main types of eviction, distinguished by the authority issuing the eviction order: an administrative eviction, which involves the decision of a mayor, and a judicial eviction, which involves the decision of a justice of the peace. Any eviction which takes place outside these procedures is illegal under Belgian law. These are also referred to as informal evictions, and are in essence very difficult to perceive and quantify.

7On the one hand, administrative eviction concerns housing whose condition threatens the safety of the occupants or public health. It follows an order of uninhabitability issued by a municipality or a decision of the Direction de l'Inspection régionale du Logement (DIRL). In both cases, the mayor is responsible for carrying out the eviction.

8On the other hand, judicial eviction concerns cases where a landlord has his property vacated by a tenant household which has failed to meet its contractual obligations (unpaid rent, deterioration of the property, failure to leave at the end of the lease, etc.) or by an occupant household which occupies the property outside a regulatory lease. It follows a judgment of eviction issued by the justice of the peace of the judicial district in which the dwelling in question is located.

9The remainder of this article focuses on the analysis of judicial evictions, whose numbers are much higher than those of administrative evictions (see below).

1.2. Counting the number of evictions

12No official public statistics on housing evictions exist in Brussels or in the two other regions of the country, either for the administrative or the judicial procedure. The multiple stakeholders involved at each stage of these procedures makes the production and centralisation of uniform data difficult and thus contributes to the invisibility of this issue.

13The Observatoire de la Santé et du Social de Bruxelles-Capitale [Amerijckx et al., 2019] estimates that in all likelihood, administrative evictions are very rare in Brussels. After a survey of the Brussels municipalities (only 6 responded to the survey), the Observatoire only recorded one administrative eviction during the period from 2013 to 2017. It is likely that this number is higher, yet the absence of a register of administrative evictions makes it impossible to confirm this.

15Our count focuses on an intermediate but crucial stage of the judicial eviction process, namely that of the judgment authorising the eviction (stage 2). This is the decisive stage in the procedure, as the eviction is no longer a virtual threat from this moment, but becomes a legally enforceable reality. For the tenants concerned, the countdown has now begun: if they do not find a solution for rehousing within the time limit set by the judgment, they will be evicted, willingly or not, along with their belongings. Thus, in all cases, the delivery of an eviction order implies a forced departure even if, in the end, the judgment is not carried out by a bailiff with the help of the forces of law and order. Moreover, most of the negative consequences mentioned in the introduction are already in effect: receiving an eviction order is a source of intense stress for the tenants concerned and, in the short term, there is a strong risk that they will find themselves in a temporary housing situation (emergency accommodation, staying with relatives or friends, etc) or in inadequate housing (squalid, too small, too expensive, etc), which exposes them to new risks of eviction and has an impact on the other aspects of their lives (work, school, health, etc). We therefore consider that an eviction order must in fact already be considered as an eviction, even if the circumstances of the latter are more dramatic in cases where the procedure goes as far as forced execution.

16In Brussels, eviction orders have never been counted systematically. It must be said that this work is particularly time-consuming, as these orders are only available at the registry of the justice of the peace in each judicial district (of which there are 19), in paper format and without a specific directory. For this study, we carried out a comprehensive inventory of eviction orders issued during an entire year (2018, with collection beginning in 2019), after obtaining permission from each justice of the peace to access archived records. The information contained in these orders (address of the dwelling, age and place of birth of the evictees, address of the landlord and terms of the eviction) was systematically collected, anonymised, cleaned (deletion of eviction orders concerning a garage, for example) and encoded in a database. These data collection and encoding tasks required nearly 1 000 hours of work.

2. How many, where, who and at whose request?

17This section presents the results obtained by processing our comprehensive list of eviction orders issued in Brussels in 2018. We explore four questions, with regard to the scale of the eviction issue in Brussels (how many?), the spatial distribution of evictions (where?), the profile of evicted tenants (who?) and the profile of landlords who have obtained an eviction order (at whose request?).

2.1. Every day in Brussels, 11 households receive an eviction order

20In more than 8 out of 10 cases, the eviction orders issued by the justice of the peace are due to rent arrears, whose average amount is € 2 900 (Figure 2). A quarter of eviction orders are due to unpaid rent of less than € 2 000. Among the other reasons for eviction, the most recurrent is the failure to leave at the end (early or full term) of a lease or temporary occupation agreement (7 % of orders). The other reasons for eviction are marginal (neighbourhood disturbance, disturbance affecting use and enjoyment, damage, subletting, etc).

21Nevertheless, we must point out the fact that these results are limited due to the data source. The transcripts of eviction orders issued by a justice of the peace are usually not very detailed, merely providing a legal version of reality. However, as rent arrears are enough to justify an eviction, any other reasons which lead to an eviction order do not need to be taken into account. In addition, rent arrears may be the result of issues other than strictly financial ones. For example, an eviction for non-payment of rent may conceal the fact that the tenant voluntarily withheld a portion of the rent in order to get the landlord to carry out certain works on the property. But this strategy provides legal grounds for eviction. There will therefore be an eviction order as in any other case of rent arrears.

2.2. Housing evictions are not only a problem in working-class neighbourhoods

23Figure 3 shows the rate of housing evictions for the neighbourhoods in the Brussels-Capital Region. The size of the circles is proportional to the number of eviction orders issued for dwellings in each neighbourhood.

24The map highlights a significant spatial disparity of evictions in Brussels, marked by a contrast between the eastern and south-eastern neighbourhoods of the city, with eviction rates below the regional average (1,3 %), and neighbourhoods located mainly in the west or north of the city where eviction rates exceed this average very regularly. Moreover, in some municipalities, the eviction rate varies significantly from one neighbourhood to another, notably in Uccle, Anderlecht and Schaerbeek.

25However, this geography does not strictly follow the socio-economic geography of the city. While the wealthy neighbourhoods in the south and east have the lowest eviction rates, the central working-class neighbourhoods have eviction rates which are sometimes higher (in Cureghem, Forest and Laeken, in particular) and sometimes lower (in Molenbeek, in particular) than the regional average.

26This result is confirmed by the existence of statistically significant (at a 95% confidence level) but nonetheless limited correlations between the eviction rate according to neighbourhood and several socio-economic indicators, including the proportion of job seekers, average taxable income per capita and the proportion of single mothers with children (Table 1).

Table 1. Correlation coefficients between the eviction rate and various socio-economic variables at neighbourhood level in the Brussels-Capital Region

Indicators according to neighbourhood (source: IBSA)	Correlation coefficient with eviction rate
Proportion of job seekers in the population aged 18-64 (%) 2018	0,52
Proportion of North African population (%) 2018	0,48
Proportion of single mothers aged 20-49 with children (%) 2018	0,41
Proportion of people aged 0-17 in total population (%) 2018	0,37
Proportion of Sub-Saharan African population (%) 2018	0,30
Average monthly rent per dwelling (€) 2018	-0,37
Average living area per capita (m²) 2001	-0,43
Proportion of dwellings with basic comfort (%) 2001	-0,44
Average taxable income per capita (€) 2015	-0,49

28In addition, several neighbourhoods in the outer ring, in the south, west and north of the region, have higher than average eviction rates. It is interesting to note that these neighbourhoods are also destinations for households leaving central working-class neighbourhoods [Van Hamme et al., 2016]. This spatial concordance leads us to formulate another hypothesis: part of the evictions affect (relatively) precarious households which – by moving out of the central working-class neighbourhoods to less dense neighbourhoods in the outer ring, in the west or north of the Region – have put themselves in difficulty by renting better quality housing with high rental costs.

29It can also be noted that the eviction rate is correlated significantly with the average monthly rent per dwelling, but negatively. This statistical relationship at neighbourhood level suggests that households outside the most expensive segments of the rental market, which are inaccessible to households with limited incomes, are more at risk of eviction. Correlations with indicators of average surface area of housing per capita and of the proportion of housing with basic amenities confirm that evictions are more common in neighbourhoods with a high amount of low-quality housing. The review of eviction orders also shows that a rental dispute caused or aggravated by the poor condition of housing can lead to eviction.

30These analyses would benefit from further refinement, but the limited availability of data at neighbourhood level on housing conditions, costs and transactions in Brussels represents a heavy constraint.

2.3. Children and foreign-born persons appear to be numerous among those evicted

31The information available on households affected by an eviction order is fragmentary and the coverage of the data varies from one judicial district to another, reflecting diverse encoding (or non-encoding) practices throughout the Brussels-Capital Region. The main limitation is that this information relates to the person who signed the lease agreement, rather than to the individuals who actually occupy the dwelling. Therefore, in particular, the possible presence of children is omitted in eviction orders. Similarly, the slight overrepresentation of men (56 %) among those listed on eviction orders does not mean that there are proportionately more men than women who are evicted, as it is possible that this small difference is due to an overrepresentation of men among leaseholders. Note that this observation does not coincide with research in the United States, which indicates an overrepresentation of women among people who are evicted [Desmond, 2016; Hepburn et al., 2020].

32The eviction orders provide the year of birth of the tenants in almost three quarters (74 %) of the cases. These data indicate that nearly 60 % of those affected by an eviction order in 2018 were born between 1970 and 1989 (Figure 4). This means that they were between the ages of 28 and 49 at the time of the court hearing, i.e. in age groups where the likelihood of having dependent children is highest. Furthermore, the table of statistical correlations calculated at neighbourhood level (Tableau 1) shows a positive variation between the eviction rate and the proportion of 0-17 year-olds in the total population. The presence of many children among the evicted households is therefore highly likely. This observation is consistent with findings from the American studies showing that households with children are more affected by evictions [Desmond et al., 2013; Desmond, 2016].

34Finally, the civil status and occupation of the evicted tenants are only provided in an extremely small number of eviction orders (9 % and 7 %, respectively), thus preventing an analysis.

2.4. Evictions not only at the request of private landlords

35Finally, the eviction orders make it possible to determine the profile of the landlords who make claims to the justice of the peace. In 2018, more than 8 out of 10 housing evictions were at the request of private landlords, 67% of whom were individuals, and 14%, private companies (Table 2). The issue of housing evictions in Brussels is therefore above all a question of exclusion through market mechanisms.

Table 2. Number and rate of evictions according to type of landlord in the Brussels-Capital Region, figures for 2018

Landlords requesting eviction	Eviction orders		Eviction rate (α/β)	Data used in the denominator for the calculation of the eviction rate (sources: Bruxelles-Logement, Perspective, Census 2011) (β)
	Number (α)	%
natural person	2626	67%	1,3%	TOTAL - (A+B+C+D)
company	541	14%
AIS	123	3%	2,6%	AIS housing 2018 = A
SISP	433	11%	1,1%	social housing on 1/4/19 = B
property management agency (Region and municipalities)	85	2%	1,3%	regional housing on 1/1/17 + municipal housing on 31/12/18 = C
CPAS	93	2%	6,7%	CPAS housing 2018 = D
Total	3908	100%	1,3%	housing available for rent 2011 = TOTAL

36Nevertheless, sometimes (in 18 % of the cases), eviction orders are issued at the request of the Agence Immobilière Sociale (AIS) or public organisations, Société Immobilière de Service Public (SISP), Centre Public d’Action Sociale (CPAS) or the property management agency of the region or municipality. Among these, the SISP has a lower than average eviction rate, even though it houses only low-income populations, but the rent charged is significantly lower than in the private sector. On the other hand, both the AIS and the CPAS have higher eviction rates than the regional average, by a factor of 2 and 5 respectively.

38Finally, with regard to evictions requested by private landlords (individuals or companies), it is interesting to note that 13 % of them live at the same address as the evicted tenant and 19 %, in the same municipality. Nearly one third of evictions are therefore obtained by landlords who live (very) close to the home of the evicted tenant. In contrast, more than a third (36 %) of evictions are obtained by landlords who do not live in Brussels. In geographical terms, the proximity between the evicting landlord and the evicted tenant is greater in the central neighbourhoods of the city, whereas evictions in neighbourhoods on the outskirts are more often requested by landlords who are more distant, i.e. who live in another municipality of Brussels or outside Brussels (Figure 5).

39By comparing these maps with the map showing the geography of eviction rates (Figure 3), we can see relatively low rates in central neighbourhoods where there is a high proportion of landlords who live near the place of eviction. This result leads us to hypothesise that the proximity between landlords and tenants might encourage forms of “amicable” settlement, thus avoiding evictions, or the use of alternative strategies by the landlord to force the tenant to leave without going to court. Being a landlord who lives in the rental property implies a less affluent profile and therefore perhaps a greater willingness to find solutions which avoid the costs of legal proceedings. Unfortunately, the lack of available official data on property in Brussels prevents us from bringing these results into perspective.

Conclusions

41The vast majority of these judicial evictions concern privately rented housing (81 %) and are justified by rent arrears (86 %) of a median amount of € 2 900. Although the profiles of the evicted households are not well known, it seems very likely that many children and foreign-born persons are among them. Moreover, the spatial distribution of these evicted households indicates an increased prevalence of the problem in the working-class neighbourhoods of the city centre – although an internal disparity exists, which remains difficult to interpret – as well as in many of the neighbourhoods of the outer ring which are neither working-class nor affluent. Affluent neighbourhoods in the south and east of the region have the lowest eviction rates.

42These observations lead us to underline the structural dimension of housing evictions in Brussels: for a large part of the population, the root cause of these evictions is not so much unpaid rent as unaffordable rent. On the one hand, the burden of rents which are too high in relation to available income suffocates many households financially, to the point of being evicted for unpaid rent. On the other hand, this same burden of high rents makes it difficult for households to leave their homes for other reasons, such as family or work, to the point where some have nowhere else to go given their situation. Excessive rents can thus be the source of evictions for non-departure at the end or termination of a lease.

43In this perspective, it seems essential to us to consider housing evictions not as an individual issue related to “bad tenants” or “unscrupulous landlords”, but as a deeply systemic and therefore political issue. According to American sociologist Matthew Desmond, author of a highly acclaimed study in Milwaukee [Desmond, 2016], housing evictions are intimately linked to the commodification of housing and the insecurity of the working classes. In other words, the system of housing production and allocation is at the root of inadequate housing, of which evictions are one of the most extreme outcomes on a personal level.

44In Brussels, this system is dominated by market logic with very little public regulation (except for the duration of a lease), thus leaving landlords with a great deal of leeway in setting prices. Landlords are thus encouraged to maximise their rental profits while many tenants are forced to pay high rents until they can no longer afford to pay. In this system, except in a minority of cases in which a landlord requests an eviction in order to occupy the property him/herself, eviction is the legal tool which guarantees the lucrative function of a property, allowing a landlord to get rid of a tenant legally when he/she is unable to pay the fixed rent.

45Moreover, households in difficulty on the Brussels housing market have very few possibilities to find housing elsewhere, due to the lack of social housing supply which meets their needs. In addition, temporary housing for evicted households is in very short supply in the Brussels-Capital Region (i.e. only 517 units; source: Bruxelles Logement) and must also meet the demands of other situations of inadequate housing (protection of abused people, reintegration of prisoners, etc). Without the possibility of non-market housing, these households are forced to occupy housing which is unsuitable for their family and financial situation, thus exposing themselves to new risks of eviction.

46The lack of regulation restricting the market logic which dominates the housing production and allocation system in Brussels is still reflected in the scarcity of data on housing evictions. The study which this article is based on is the first to have conducted a comprehensive count of the number of housing eviction orders in Brussels for a single year. A project for the development of a monitoring system on evictions is currently in the hands of the Brussels government. We feel that its implementation is essential in order to give more visibility to the issue of housing evictions in Brussels, have a better understanding of the different dimensions and allow it to be monitored over time. However welcome it may be, the quantitative monitoring of housing evictions does not erase the social destructiveness inherent in the eviction process. In order to reduce the number of evictions, it is necessary to intervene in the structures of housing production and allocation, and reduce the domination of a profit-making approach.

We would like to express our warm thanks to all of the people who collaborated on our study: Isaline Wertz, Julien Descamps, Chaïma Bakkali Tahiri, Larissa Pâques, Lina Achour, Pierre Marissal, all of the judges and clerks of the Brussels justices of the peace, as well as all of the people who shared their experiences regarding housing evictions with us.