The future of Flemish social housing

Transcription

1 Workshop 16 - Institutional and Organisational Change in Social Housing Organisations in EU The future of Flemish social housing Sien Winters sien.winters@hiva.kuleuven.be Marja Elsinga m.elsinga@otb.tudelft.nl Paper presented at the ENHR conference "Housing in an expanding Europe: theory, policy, participation and implementation" Ljubljana, Slovenia 2-5 July

2 ENHR 2006 Ljubljana, Slovenia Workshop 16 Institutional and Organisational Change in Social Housing Organisations in EU draft The future of Flemish social housing Sien Winters & Marja Elsinga Higher Institute for Labour Studies (HIVA), K.U.Leuven OTB Research institute for Housing, Urban and Mobility studies, Delft University of Technology & 2

3 Contents Contents Placement of the research Flemish social housing The Flemish social rented sector in an international perspective Policy questions and future scenarios Bibliography Abstract Social housing in Flanders covers a share of 6% of the housing stock. At first glance the sector in Belgium could be considered as a residual housing sector. This paper aims to unravel Flemish social housing and focuses on the current problems. One of the dilemmas is that allocating social housing to the most vulnerable groups implies efficient use of social housing but also a risk of stigmatisation and of further loss of public support. Alternatively the target group of social housing could be enlarged, also to improve the liveability of neighbourhoods by way of mixed income neighbourhoods. The Flemish rent setting system implies that targeting to lower incomes results in lower rental revenues for the housing associations and can lead to financial problems if social housing finance is not adjusted. The paper first describes the history and position of the social housing sector in Flanders. Than the paper elaborates on the problems in social housing as they appear in the public and political arenas and as is shown by facts. Next the paper focuses on social housing in a number of other countries and more in particular on the way these countries cope with liveability problems and organise affordable housing expenses for those in need. The exploration is meant to inspire the search for solutions to the problems. The papers ends with scenario s for the future of social housing in Flanders and policy recommendations. 3

4 1. Placement of the research At the end of 2004, the Flemish government awarded the Knowledge Centre for Sustainable Housing Policy the task of conducting a study into the letting and financing of the social rented sector in Flanders.The reason behind the research was, firstly, the observation that the financing of the social rented sector suffers from many bottlenecks. Reference was made, for instance, to the declining incomes of the tenants and the increasing discrepancy between the incomes and expenditures of the housing associations while, on the other hand, affordability for the tenants is not always guaranteed. A second reason was the increasing signals from the field concerning liveability problems in certain neighbourhoods. The sector asked to be able to respond for example by expanding the target group for social housing, adjusting the allocation system and allowing a special treatment for problem tenants. Furthermore, it was regularly mentioned in the public arena that also modal Flemish families with earnings from employment should be able to enter social housing as it was earlier. Reference was made to the increasing influx of the unemployed, underprivileged, ethnic minorities, etc., and the associated problems. Finally, criticism was expressed by the social housing associations of the fairly strict regulation. In order to be able to provide a response to financial and social problems, the sector asked for more autonomy. The minister responsible for housing wanted to respond to all these problems with an appropriate policy. At that time, the policy options included raising the income limits, adjusting the allocation system, increasing the autonomy of the associations or the municipal authorities and a review of the financing system. The aim of the research was to provide direct policy advice. It was to provide an input both for the review of the Social Rent Decree which governs the main issues surrounding social letting and for the review of the financing system. The research studies the problems in the Flemish social rented sector, explores the possible policy options and formulates scenarios for future developments. In this paper, we present some of the results of this research, focusing on the target group, housing allocation and financing. The questions we deal with in this paper are as follows: how is the social rented sector organised in Flanders? Which problems are being flagged in the public and political arenas? What are the objective data? How is Flemish social housing characterised in an international perspective? What options are possible regarding the target group, housing allocation and financing? What are possible future scenarios for the Flemish social rented sector? 2. Flemish social housing 2.1 Definition of social housing The term social housing covers a wide range of initiatives within Europe, aimed at good quality, affordable housing. Much has been written in the literature about the need and definition of social housing. Maclennan & More (1997) claim that social housing is needed when the socially desired needs for housing exceed the effectively desired demand for housing. In general, it is agreed that this means affordable housing for vulnerable groups. A further search of the literature produces vari- 4

5 ous definitions of social housing (see, for example, Oxley, 1995, 2005). What is noticeable is that views are far from unanimous. Thus, for example, social housing can refer to both renting and purchasing. The owner of those homes can be a public or private owner and can be looking to make a profit in his business management, or not. Furthermore, the allocation of social housing can be regulated or non-regulated. In the latter case, the consumer s freedom of choice is key. The question which arises is then: how is affordable housing guaranteed? Is this via institutions, such as housing corporations, via object subsidies for the providers or via a form of income subsidy for the occupant? In all cases, such resources reduce the living expenses for the occupant. If we leave a housing subsidy in the form of an income supplement out of consideration, the characteristic difference between social and non-social housing seems to be definable at housing level: does a dwelling have a social objective (a primary goal) that is pursued through a nonmarket compliant, but lower rent? In accordance with this definition, in Flanders we find social dwellings in both the owned and in the rented sector. In the rented sector, the vast majority of the social dwellings are owned by social housing associations, which are private non-profit organisations. In addition, a limited part of the property is owned by the Vlaams Woningfonds (Flemish Housing Fund), by municipalities and Public Social Welfare Centres (OCMWs).. Most of these dwellings are subsidised and are let below the market rent. The dwellings let through the social letting offices can also be included in this definition of the social rented stock. This refers to a fairly limited number of dwellings which are not subsidised in terms of construction. When we talk about social housing in Flanders, we are usually referring only to the dwellings that have been built by the social housing associations. In describing the profile of social housing in Flanders, we confine ourselves to this market segment because this forms the largest portion of social housing and because reliable data exist for this segment, which is not the case for other segments in the social rented market Social housing policy in Belgium and Flanders Belgian housing policy has its roots in the late 19 th century and has been chiefly targeted, throughout its history, at promoting property acquisition by private individuals (see, e.g. Goossens, 1982, Descamps, 1997, De Meulder et al., 1999). Substantial premiums 2 and tax incentives were awarded to prospective owners. Significant tax advantages are still granted to owners today by the federal government. The social rented sector also originated at the end of the 19 th century, but the sector was mainly developed after the Second World War 3. Nonetheless, the social rented sector did not acquire such an important role in post-war reconstruction as it did in other European countries. During the period , 83% of new construction in Belgium was carried out by private individuals, approximately one third with government subsidies. The social housing associations represented 16% of new housing construction. Of this, two-thirds were intended for sale and one third for letting (Descamps, 1997). Until 1960, no income limit existed for social rented homes. After that date, limits were introduced. The target group was narrowly defined during some periods and during others very broadly (Goossens, 1997; De Decker, 2000). 1 No exact figures exist concerning the number of social dwellings let by local authorities. At the end of 2005, the number of homes let by Social Letting Offices (SVKs) was The main premiums were through the Moyersoen Act (1922) and the Detaye Act (1948). 3 In particular, the Brunfaut Act (1949) was a milestone. 5

6 In 1980, housing policy in Belgium became the sole responsibility of the three regions: the Flemish Region, the Walloon Region and the Brussels Region. The process of regionalisation was a gradual process and took a long time. It was not until 1997 that the Flemish Region had its own Housing Code (Vlaamse Wooncode - decree of 15 July 1997). Within this Housing Code, the social housing associations and their umbrella organisation, the Flemish Housing Corporation (Vlaamse Huisvestingsmaatschappij - VHM) 4, are declared the preferred partners for the implementation of social housing. The social housing associations are private associations whose largest shareholders are the municipal councils. Their boards of directors consist mainly of municipal representatives. The policy of the Flemish government differs from the previously pursued Belgian housing policy. Flanders clearly shifted the emphasis from promoting property acquisition to supporting tenants. This is clear from spending by the Flemish government: whereas, in 1994, expenditure on behalf of owners and tenants was approximately the same, from then on more resources were deployed year on year in favour of the rented sector. In 2003, the rented sector accounted for approximately three times the amount of Flemish government resources, compared to the ownership sector. In terms of expenditure to benefit tenants, in % went to investment support for social housing construction (Winters & Van Damme, 2004) and the remainder to direct support for private individuals. No housing allowance exists in Flanders. Currently, the possibility of introducing a housing allowance is being examined (Keulen, 2004). Nonetheless, the federal government also possesses important tools that have a significant impact on the housing market. For instance, the housing tax system benefits owners of homes many times larger than the budget which the Flemish government has available to pursue a housing policy. As a result, Flemish families are still strongly encouraged to purchase their own homes (Doms et al., 2001). Furthermore, rent legislation, which has also remained a federal power, is characterised as highly liberal legislation, partly because of its powerful protection of proprietary rights, the contractual freedom of the parties, the assumed equality of the parties and the largely suppletive nature of the legislation (Goossens, 1997). 2.3 Social housing regulation and finance in Flanders The rental of social dwellings is regulated in the Social Rent Decree ( Sociaal Huurbesluit, Flemish Government decree of 20 October 2000). This defines eligibility conditions and allocation rules, as well as rent setting. In order to be entitled to a social rented dwelling, maximum income limits apply, depending on the family situation. The net taxable income of single households should be below 16,715 per year. For families (at least two persons) the income limit is 25,073 (increased by 1,393 per dependent). The family may not own a home or fully enjoy the usufruct of property of the home. Prospective tenants should place their names on a register. The dwellings have to be allocated in chronological order of registration, taking into account a number of priority rules. Preference is given, for example, to people with a handicap if the available dwelling is adapted to their specific needs, to certain mobile home dwellers and to families leaving a home which has been declared uninhabitable, in cases of repossession. In order to be able to allocate correctly, the housing association must have internal allocation rules. These rules need to be compatible with the Social Rent Decree and have to be approved by the Flemish administration. The associations have to follow the rules strictly. They cannot take into account specific local circumstances (such as qual- 4 With effect from 1 July, the rights and obligations of the VHM are being assumed by the Vlaamse Maatschappij voor Sociaal Woonkrediet (Flemish Association for Social Housing Credit - VMSW). 6

7 ity of life in the housing complex) or special characteristics of the tenants. Exceptions are possible, but need to be justified by the association and accepted by the VHM. Rent calculation is also tightly regulated. The Social Rent Decree defines the formula which has to be used to calculate the rent. This formula does not result in an exact rent, but defines a margin within which the company can fix the rent. The main factors in the calculation formula are the updated cost price of the dwelling and the income and family situation of the tenant. Rents are recalculated on 1 January each year. The calculation is made as follows 5. Initially, the association decides the basic rent for the dwelling, which has to be between 3% and 9% of the updated cost price of the dwelling. The cost price includes all original building costs, as well as the costs of improvement and renovation work. Secondly, the basic rent is adapted to the income and family situation. For this, the income coefficient (ic) is calculated: I +1,672 euro ic = 17,551 euro Where I (income) is reduced by for each of the first two dependents. In the next step the adjusted rent is calculated as follows: Adjusted rent = basic rent * ic Larger families are then granted a deduction equal to 20% of the basic rent for 3 dependents, 30% for 4 dependents, 40% for 5 dependents and 50% for 6 dependents. Finally, maximum and minimum income limits apply. The monthly rent payable may not be higher than 1/60 of the income and no higher than the normal market rent. The lower limit is half of the basic rent. In addition to the rent, service costs and charges can be imposed. Although, at first glance, rent calculation is strictly regulated, associations can adjust rents according to financial needs, i.e. by raising the basic rent. Investment in new construction, renovation and infrastructure (roads, parks, etc.) is largely subsidised by the Flemish Region. Two subsidy systems are available. The first system provides the associations with soft loans to be paid back over 33 years. Interest rates on these loans depend on the average income coefficient (ic) of the association: the lower the average ic, the lower the interest rate. The loans to the associations are provided by the VHM and funded partly by subsidies, partly by private market loans. The level of subsidy is depending on market interest rates. In 2005, the subsidy was amounting to 28,74% of investment costs (VHM, 2006a). Although this financing system to some extent takes into account the expected revenues of the associations, it does not deal with the declining income of residents over the life cycle of the building. The other subsidy system is more generous and grants a direct subsidy of 60 or 65% of investment costs, but does not take into account expected revenues at all. Housing associations, especially those operating in the cities, complain that incomes from tenants are falling, which results in financial problems for the association. For three years now, some associations have already been receiving a subsidy aimed at bridging the gap between income (rents) and expenditure. 2.4 The problems as they appear in the public and political arenas De Decker (2005) provides an extensive discourse analysis of the development of language and vocabulary relating to the social rented sector and social tenants since Based on document analysis, De Decker demonstrates that the discourse evolved from a local topic into a broadly supported consensus about the social tenant as a problem and social neighbourhoods as uninhabitable. According to De Decker, this discourse was formulated and fuelled by people with impor- 5 We abstract some exceptional measures. 7

8 tant social and political positions and eagerly adopted by the media. It was based chiefly on incidents and was not supported by figures. In the early 1990s, according to De Decker, it was mainly the influx of ethnic minorities which was regarded as problematic. Gradually, new problem categories were added: criminals, people committing social fraud, tenants who did not pay their rent. Later, the problem group was even further expanded to include vulnerable families, people on a minimum income and, ultimately, to low wages in general. Various interested parties (directors of the associations, political parties, residents organisations, etc.) have pointed out that the target group for social housing has shifted from traditional family types to the long-term unemployed, ethnic minorities, single people and single-parent families (see e.g. Cools, 2004; VVH, 2004, hearing of Flemish Parliament, 16 December 2002). The influx of weak groups in this context is being directly linked to the existence of quality of life problems. Poverty is associated with inappropriate living behaviour noise nuisance, vandalism, harassment. The policy is being pressed into providing a solution. As with formulation of the problem, according to De Decker (2005), relatively broad consensus exists concerning the solutions. Virtually all political parties advocate raising the income limits in order to attract better target groups and to achieve a social mix. The housing associations advocate wide autonomy concerning the allocation of social rented dwellings, specifically so as to be able to turn away risk tenants or to be able to limit the influx of certain groups. The housing associations also propose the introduction of trial contracts or temporary contracts (VVH, 2004). Another problem to which reference is regularly made in the debate is the financial situation of the social housing associations. The sector specifies that the present subsidies are not high enough to be able to let the dwellings without making a loss. The VVH claims that, as a result of rents being linked to incomes, the incomes of the social housing associations have fallen, which is not being compensated by a proportionally increased level of subsidy (VVH, 2004). On the other hand, the voices of the groups highlighting the needs of the weak population groups are also being heard. For instance, the welfare sector points out that insufficient numbers of homeless people are able to transition into social housing (Van Menxel, 2005; Lescrauwaet, 2005). The Vlaams Overleg Bewonersbelangen (Flemish Occupants Interests Consultative Body VOB) has expressed criticism of the concept of social mix as a component of the allocation policy (Inslegers, 2005) 6. The Flemish Minister responsible for Housing is attempting to respond to some of the aforementioned problems. In the Policy Document , the Minister states that it must be possible, at the level of residential neighbourhoods, to achieve a balanced social mix, both within social housing and within the society as a whole, in order to provide a response to the local quality of life problems. The Minister states, following the Flemish coalition agreement, that the income limits will be reconsidered: An increase in the income limit would also give the lower middle class access to the social housing market. The aim of this is not only to promote the quality of life of social housing complexes, but also to improve the financial situation of social housing associations. In addition, when preparing the new Social Rent Decree, the Minister announced that attention would be devoted to dealing with social fraud and specific domicile fraud, harassment, rental property damage and rent arrears. In the mean time, draft amendments to the Flemish Housing 6 The VOB is the umbrella organisation of Flemish landlords unions, social letting offices and housing initiatives targeted at weak groups. 8

9 Code and adjustments to the Social Rent Decree are being prepared. No final decision has yet been made. 2.5 Profile of social housing in Flanders In Flanders, we have very little systematically collected data to be able to objectify several of the aforementioned problems. De Decker (2005) states that this is partly the cause of the negative discussion about social housing. Concrete information about marginalisation, ghetto formation and the opinions of users of social housing are virtually non-existent and could therefore not be included in the debate. We provide a brief summary of the main data that are available. These are primarily figures which the Flemish Housing Corporation collects from its associations, data from the 2001 Census and the results of a survey among social tenants (Pannecoucke et al., 2001). At the end of 2004, the number of dwellings owned by social housing associations was 135,910 (Annual Report 2005, VHM), which represents 5.7% of the Flemish housing stock. The share of social rented dwellings out of the total rented stock amounts to 22% 7. At the end of 2005, 58,215 people were on the waiting list for a social rented dwelling (VHM, 2006b) 8. In terms of housing typology, the social rented market is characterised, compared to the total Flemish housing stock, by an over-representation of apartments and an under-representation of single-family dwellings. Social dwellings are on average newer than other dwellings. Slightly more than half of the social dwellings was built between 1951 and Social dwellings have more conveniences and are better equipped than private rented dwellings (Pannecoucke et al., 2001). Social tenants have an income that is on average much lower than average Flemish families. At the end of 2004, 32.6% of social tenants had a taxable income of below 10,000 and 50.3% had an income below 12,000. There are no income distributions available for distinct family types in total population allowing us to compare these with the income distributions of social tenants. Distribution of fiscal incomes can be used as an approximation 9. The most recent available data are those from In that year 16,6% of fiscal households in the Flemish region had an income below , while 24,1% had an income below These lower incomes in social housing can partly be explained by a relatively high proportion of single people. On 1 January 2003, this proportion was 28.4% in the Flemish Region (NIS, based on National Register), compared to 41.2% for social housing (VHM). The results of a survey among social tenants (Pannecoucke et al., 2001) show that the social tenant has a weaker socioeconomic profile than average Flemish families. Among social tenants, relatively more are unskilled, unemployed, sick or disabled heads of household. In social housing, relatively more households also have a head of household who is of Moroccan, Turkish or non-eu nationality. Nonetheless, the maximum income limit is relatively high. Based on incomes for the year 2002 for single people and families, we roughly estimate that 52% of all single people would have 7 Figures for the total housing stock and the number of private rented dwellings based on socio-economic survey, Not included in this figure are those who are already renting a social dwelling and have requested a transfer. People who have already been waiting for longer than 10 years are also not included because these are highly specific requests. 9 One problem is that fiscal statistics do not include the lowest incomes. Another problem is that fiscal households do not correspond to households as defined in the Social Rental Decree. For taxation, couples living together but not married are seen as two households. Also family members living in but with an own income are seen as seperate fiscal households. 9

10 been entitled at that time to a social rented dwelling if only income were taken into consideration. For families on a single income, this proportion was 64%and, for families with 2 incomes, 24% 10. The fact that, despite this relatively high income level, the weak groups are still heavily represented, is probably related to the priority rules that apply to some groups. As indicated above, it is frequently said that the incomes of social tenants are becoming increasingly lower. Figures from the VHM show that this was indeed the case for the period The share of families with an income below 10,000 rose over that time from 28.0% to 34.3%. After 2001, this percentage stabilised and, between 2003 and 2005 it actually declined to the current 32.6% (VHM Annual Report, 2005) 11. However, this general picture can conceal major differences. For instance, it may be that the incomes of the newcomers are falling, but that this evolution is being compensated by rising incomes among sitting tenants. Figures from the VHM indeed show that incomes of new tenants are lower than incomes of all tenants. At the end of 2004, 52,0% of new tenants had an income below en 68,2% below , while these percentages were 32,6% and 50,3% for all tenants as specified above. The occupant profiles can also differ widely among social housing associations, so that it is not unlikely that average incomes in the urban areas are falling, while they are rising in other areas. At the same time, some data are available which reveal that a number of weak groups find it very difficult or impossible to obtain social housing. The welfare sector complains that the homeless have no legal priority when it comes to allocating social dwellings (Van Menxel, 2005). Lescrauwaet (2005, based on a survey in 1999) claims that 54% of the clients at hostels for the homeless could move directly into social housing because they can live independently, while only 4% of the 5,630 people received at the hostels in 1999 did actually move into social housing 12. According to Loopmans et al. (2005), asylum-seekers do in principle have access to a social dwelling, but actual allocation depends on the policy of the housing association in question. By contrast, people without documents cannot be allocated social housing because their existence cannot be confirmed by facts. The average rent for a social dwelling amounted to 207 in The percentage of tenants paying under 200 was 57.1%. The peak of the distribution is between 150 and 175. Compared to rents on the private market, it is cheap to live in social housing. No recent data are available about rents on the private market. In 2001, 27.6% of Flemish households paid less than (Census, 2001). In the same year in social housing, 80.3% paid less than that amount (VHM Annual Report, 2001). In contrast to the picture painted in the public arena, a survey among social tenants (Pannecoucke et al., 2001) reveals that there are no general problems with quality of life in social housing. Of the social tenants surveyed, 79% answered that they were sattisfied to very sattisfied with the dwelling and 75% satisfied to very sattisfied with the neighbourhood. Compared to a survey conducted in 1982, housing satisfaction had also increased. Satisfaction is high for most of the items surveyed, including neighbour relations, privacy and quality of life in the neighbourhood. Occupants are relatively the least satisfied with (lack of) community life and with the population composition, although only 10% expressed negative views of their own neighbours. Nonetheless, other differences are hidden behind this general picture, depending on circumstances. Differenti- 10 Rough estimations under several assumptions. Since no income distribution is available for different family types, we neglected the effect of dependents on income limits for social housing as well as on fiscal income disctribution. We also did not take into account the requirement of not being an owner. 11 Figures corrected for inflation. 12 We note that the research does not give a time period. Assuming that the average waiting time for a social dwelling is 2 to 3 years, it is possible that the percentage may be higher with the passage of time. 10

11 ated factors include the age of the dwellings, dwelling type and comfort level. A number of neighbourhoods certainly have problems with quality of life. For a description of the causes and possible solutions, we refer to Heylen and Tratsaert (2006). Objective data and studies are also difficult to find for objectifying the debate about financial problems. The relation between incomes of tenants and the financial position of housing associations is more complex than assumed, as the above data about the evolution of incomes indicate. Further research into the relationship between incomes, rents, cost prices and subsidies is needed in order to achieve an accurate picture 13. The VHM has conducted an interesting analysis of the consequences of the current system of rent calculation on affordability for social tenants (VHM, 2004). As previously indicated, the associations have a fair amount of autonomy in determining the basic rent. In order to keep the accounts balanced, the basic rent is often increased. Since the lower limit for the rent is half of the basic rent, and this limit especially applies to low income families, an increase in the basic rent usually has a detrimental effect on affordability for this group. Another problem is that the basic rent for new projects is usually high. For this reason, low-income families can elect only to apply for the older and cheaper dwellings. As a result, this can promote physical segregation. On the other hand, an affordability standard does apply for the higher incomes, specifically because the real rent is limited to 1/60 th of the income. In terms of affordability, this maximum limit could possibly be increased for these higher incomes. 3. The Flemish social rented sector in an international perspective Liveability of neighbourhoods and affordability of housing are central issues for social housing in many countries. The next sections deals with social housing in a number of other countries. First the organisation and features of social housing are described to get a clear picture of the autonomy of housing associations and the surroundings of the problems we focus on in this paper. Then we focus on housing allocation and ways to assure affordability in social housing and search for inspiration to find answers to Flemish policy questions. This overview is based on countries studies described in a forthcoming research report Organisation and target group of the social rented sector As described in section 2.1, a social rented sector can have both a private and a public nature. Table 3.1 gives a summary of the nature of the social rented sector in six countries. In Ireland, public landlords have exclusive rights; in the Netherlands private organisations have exclusive rights. The housing associations in Flanders are in a unique position: they are private organisations that provide for public participation through shareholding by the local authorities. In other countries, such as the UK, Germany and Spain, more parties are involved in the letting of social rented dwellings, not only public or private non-profit organisations, but also profit-making parties. The social rented sector is therefore not linked to an owner, but to the system within which dwellings are let. 13 With reference to preparation of a new financing system, thorough calculations have been performed within the VHM, but the results are not publicly available. 14 Winters S., Tratsaert K., Van Damme B., Elsinga M., Haffner M., Heremans F., Van Daalen G., (forthcoming), Op weg naar een nieuw Vlaams sociaal huurstelsel?, Kenniscentrum Duurzaam Woonbeleid, Leuven 11

12 Table 3.1 The social rented sector in several countries ownership participation Financial objective Flanders Private Public Non-profit Ireland Public Public Non-profit UK: council RSL Company/person Germany: company/person Municipality Public Private Private Public Private Private Non-profit Non-profit Profit Private Public Private public Profit Non-profit Netherlands Private Private Non-profit Spain: municipality Company/person Public Private Public private Non-profit Profit Source: Housing Statistics in Europe, 1998; Norris and Shiels, 2004; Scanlon and Whitehead, 2004 Table 3.2 Market shares and target group in social rented sector Market share Target group Flanders 6 Max income Ireland 9 Max income UK: council 14 Reasonable preference to weak groups RSL (registered social landlords Company/person 6 Germany: company/person Max income Municipality 5 Netherlands 35 Priority to low incomes Spain: municipality Company/person 3 Max income Source: Winters et al, forthcoming In the Netherlands and the UK, the social rented sector occupies a larger market share, i.e. 35 and 20% of the housing stock, as shown in Table 3.2. In both countries, we see a trend towards a decline in social rented dwellings owned by the government. In the Netherlands, he municipal housing companies have virtually disappeared and, in the UK, the number of RSLs (registered social landlords) is gradually increasing. These currently manage 6% of the UK housing stock. One difference between the UK and the Netherlands is that the social rented sector fulfils a safety net function in the former case and, in the latter, it is accessible to all households. In Spain, Germany and Ireland, as in Flanders, the social rented sector is modest in scope, varying between 3 and 9%. However, major differences exist between the social rented sectors in these countries. In Germany and Spain, a dynamic social rented sector exists, which is managed by private landlords. Dwellings are built with a subsidy and kept affordable for a certain period and they are also the subject of dwelling allocation rules. At the end of this period, these dwellings are free. In Ireland, the social rented sector fulfils the function of a social safety net and the sector is owned by local authorities. The social rented sector in the Netherlands has an unclear character; one the one hand a maximum income criterion applies, on the other hand priority rules apply to weak groups. The term target group requires two explanations. In many cases, a maximum income limit applies (Table 3.2). In some countries, this is then an absolute condition for being eligible for a social rented dwelling, such as in Germany and Italy. In the Netherlands, by contrast, the target group of public housing policy has priority; this group is defined by a maximum income but higher income groups can also enter the social rented sector. It should be remarked that the use of a maximum income criteria does not necessarily imply that social housing is only accessible for the most vulnerable groups. In Flanders for example around 50% of the Flemish households have an income below the maximum but priority is given to the most vulnerable households. 12

13 In addition to income, other target group descriptions apply. For instance, in the UK, reasonable preference is given to certain vulnerable groups that are described in the Homelessness Act. In Spain, those who have difficulty obtaining a privately owned property form the target group of the social rented sector. 3.2 Housing allocation Responsibility for housing allocation varies; Table 3.3 gives a summary. In Ireland, the central authority plays a prominent role in housing allocation. The Housing Act prescribes that local authorities must draw up a scheme of letting priorities and submit it to the minister for approval (Department of Environment, Heritage and Local Government, 2002a). This distribution of tasks bears close resemblance to the situation in Flanders, where the regional authority has established strict rules for the allocation of all social rented dwellings. Any deviations have to be centrally agreed. In the other countries studied, local actors are given a major role. In Germany, the Länder can further specify the target group for subsidised dwellings in their subsidy schemes. Municipalities can control landlords via the subsidies they allocate and the emergency rule. The arrangements for this differ locally. In Spain and the UK, the local authority is responsible for housing allocation. In the Netherlands, housing allocation is the shared responsibility of the local authority and the housing corporations. In housing allocation, a distinction can be made between the way in which those looking for housing are ranked and the way in which the dwellings are offered (see table). The simplest way to rank those looking for housing is registration time. In many countries, this was formerly the way in which those looking for housing were ranked for dwellings subsidised by the government. The perception that one candidate is more urgent than another and therefore needs a dwelling more quickly forms the basis for the systems that have been developed in all the countries studied. Based on personal circumstances and waiting time, candidates are awarded urgency points or are placed into a certain urgency class. The result is that one person seeking housing can take priority over another based on criteria which are, of course, often the subject of discussion. In all six of the countries studied, a points system is used or an allocation into urgency classes. A waiting list system can also imply that partial waiting lists exist those looking for housing can register for specific dwellings in specific areas, as is the case in Flanders, for example. In Spain, in many cases, dwellings are allocated by lottery. In this lottery system, a division into urgency classes is made and the candidates from the high-urgency class have more chance of obtaining a dwelling. Within the groups, chances are equal. On the one hand, this has an element of fairness because everyone has an equal chance; on the other hand, it brings uncertainty for those looking for housing. These people do not know where they are and they can also not assume that they can count on being given a dwelling after a certain waiting time. A second subject of the table is the way in which dwellings are offered to those seeking housing. A waiting list on which candidates are ranked according to the urgency points system forms the basis for the allocation of social rented dwellings in most cases. An alternative is what is known as a choice based letting model, in which dwellings are advertised and those looking for housing can respond to these advertisements. This system, developed in the Netherlands, has been followed up in the UK, where it is known as Choice Based Letting. Finally, in the UK and the Netherlands, exceptions have been created for housing complexes which have serious problems with quality of life. In the UK, this is known as estate profiling or a merit-based scheme. In the Netherlands, allocation may be related to lifestyle or include a 13

14 special treatment for problem households. In all cases, the issue is the selection of tenants who will not aggravate quality of life problems and will preferably alleviate them. In such experiments, the aim is roughly to objectify the criteria for allocation. However, clear differences exist when it comes to the way in which the criteria are applied. In cases of special treatment for problem households, the landlord has the final say and a candidate who does not meet the criteria is not eligible for a dwelling in the complex in question. In allocation based on lifestyle, more emphasis is placed on information about the lifestyle in the relevant neighbourhood which the person seeking housing must influence in his choice of a dwelling. In addition, in a number of cases, part of the lifestyle is established in a contract, which includes arrangements about use of the semi public spaces in buildings, for example, and the time when loud music is tolerated. The above shows that, both in the UK safety net model of the social rented sector and in the broader Dutch rented sector, local possibilities exist for steering the allocation of housing towards reducing quality of life problems. In both cases, implementation is a question of local made-to-measure solutions. In Flanders, it is possible to deviate from the maximum income limits. The condition for the acceptance of higher incomes up to a new certain maximum income is that the housing association have a liveability plan and an internal allocation scheme that has been approved by the Ministry of the Flemish Community, Housing Policy office. Summary of the priority criteria and the method of dwelling supply in six coun- Table 3.3 tries participation Priority criteria Method of supply Flanders central Points system Waiting list Ireland central Points system Waiting list UK local Points system Waiting list Small-scale application of: choice based letting - estate profiling, - merit based scheme Netherlands local Housing period Supply model Points system Waiting list Urgency classes Registration time Experiments with lifestyle and lottery Germany local Urgency classes Tenant s own initiative Municipality agrees to landlord recommendation 3 models: unless there are candidates - Landlord is free to choose; with higher urgency - Landlord chooses from three candidates; - Landlord cannot choose. Spain local Points system Waiting list Lottery Source: Winters et al, forthcoming 14

15 In Germany, an entirely different system of housing allocation applies. Specifically, freedom of choice plays a role here in terms of the landlord. The table above shows that the local authority determines to what extent the landlord has the possibility of choosing a tenant from a list of those seeking housing. The local possibilities for landlords vary from no choice to a free choice within urgency classes Subsidies and rent determination in the social rented sector Based on the objective of social housing, financing is always to be based on affordability of the dwelling for the social tenant. For tenants with a low income, this means that the net rent has to be squeezed in one way or another. We are talking in this context about the difference between the cost-effective rent and the social rent. The cost-effective rent is the rent that is needed to cover all costs which the landlord incurred for the dwelling. The social rent is the price paid by the social tenant. Based on the principle of affordability, it is logical that the lower the income, the lower the social rent should be 15. One first potential option is that the government bears full financial responsibility and therefore make up all of the difference between the social rent and the cost-effective rent. This can be through both a subject subsidy to the tenant and an object subsidy to the landlord, who then converts this into a reduced rent for the tenant. The effect for the tenant is the same in both cases, at least in so far as the rent is legally established. The alternative is that the landlord be made financially responsible for both affordability of rents and controlling the subsidy. He is then assumed to generate income himself in order to charge a social rent to lower income groups. The incomes can come from letting to higher income groups, from the sale of dwellings or from financial reserves accumulated in the past. In all the countries listed in Table 3.4, an object study for social rented dwellings is applied, a subsidy that can take various forms and makes a social, non-market compliant rent possible. The only exception to this is the Netherlands, where the object subsidies were abolished in 1995 and housing corporations are themselves deemed to subsidise the construction of social rented dwellings from their equity or through other activities that generate revenue. In addition to the object subsidies for the dwellings, for low income groups an income-dependent subsidy (housing allowance) is also available in the UK, the Netherlands, Germany and Spain. Ireland and Flanders are exceptions in this case. In these countries, the income-dependent rent in the social rented sector can be considered an alternative to housing allowance. The systems of determining rent in the countries under discussion are very diverse. In Flanders and Ireland, the rent depends mainly on the income of households; in Germany it now depends on the market rent (previously on the cost price of the investment and development), in the UK and Spain on the cost price and, in the Netherlands, on the quality assigned to a dwelling via the housing valuation system and the policy of the landlord. Flanders and the Ireland are remarkable in this context. In these countries, the rent and therefore the rental income depend on the income of the tenants. An increase in the influx of low-income groups results in a lower rental income. In both countries, financial problems exist with a number of social landlords (Department of the Environment, 1992). In Ireland, this is a public problem in Flanders this is a problem with the private housing associations. The present Flemish situation is very unclear. On the one hand, an investment subsidy is deemed to cover the difference between the incomes and costs of the associa- 15 When calculating the rent, it is best to take into account not only affordability, but also the quality of the dwelling and the risk of a poverty trap. We do not examine this further. 15

16 tions but, on the other hand, the rent calculation includes an internal solidarity between more and less prosperous tenants, i.e. tenants on a lower income receive more subsidies than tenants on a higher income. Complaints have been made by associations that the subsidies are insufficient to compensate all costs. For this reason, the Flemish government has now granted a subsidy three times to a number of housing associations to help make up the shortfall. Here too, the government is still clearly assuming final responsibility. 16

17 Table 3.4 Subsidising and determination of rent Object subsidy Rent subsidy Rent Dependence on tenants incomes Flanders Yes No Income-dependent Housing association/government Ireland Yes No Income-dependent Council UK: council RSL Company/person Germany Company/person Municipality Netherlands Spain: Municipality Company/person Yes Yes, for both social and private sectors Yes Yes, for both social and private sectors Cost-price dependent - Market price dependent No Yes, for both social and Depending on - private sectors regulation and market Yes Yes Cost-price dependent - Source: Norris and Shiels, 2004; Scanlon and Whitehead, 2004, Winters et al, forthcoming The position of the Flemish social rented sector The Flemish social rented sector is modest in scope, as it is in Ireland, Germany and Spain. What is special about Flanders is the combination of private property and control by the Flemish authority of dwelling allocation and rent determination. The private housing associations and the local authorities as shareholders have to deal with the results of a central policy regarding rent determination and dwelling allocation. The present system places the emphasis on centrally regulated fair allocation and leaves little room for local made-to-measure solutions when it comes to dwelling allocation. Furthermore, the combination of income-dependent rents and shared responsibility for financing is a peculiarity in the international context. The current subsidy and rent policy confronts housing associations with the financial consequences of a potential increase in the influx of low income groups. They are jointly responsible, but have no way of finding a solution other than setting rents as high as possible within the margins, drawing on their reserves or allowing financial problems to arise. 4 Policy questions and future scenarios 4.1 The policy questions The central question for the future of social housing in Flanders is which objective it is pursuing and which target groups it wishes to reach. One first possible option is that, given the limited scope of the social rented stock, a firm choice is made in favour of housing the weakest groups. Groups which should take priority in this case are families with a low or insecure income or those suffering from negative selection or discrimination. Possible arguments for such selective target group definition can be found in economic welfare theory, which states that intervention in dwelling allocation can best be limited, so as to interfere in the market mechanism as little as possible (Barr, 1998). However, in several European countries (the Netherlands for example), social housing is also open to average or higher incomes. This offers the advantage that weak groups live integrated among other income groups and that social housing acquires a broader public support. 17