The Affordability of Housing in the
Netherlands: An Increasing Income
Gap Between Renting and Owning?
MARIETTA E. A. HAFFNER & HARRY J. F. M. BOUMEESTER
OTB Research Institute for the Built Environment, Delft University of Technology, Delft, The Netherlands
(Received May 2009; revised March 2010)
ABSTRACT Housing became more expensive in the Netherlands between 2002 and 2006, a trend
which has been demonstrated using various measures of affordability. The expenditure-to-income ratios calculated for households confirm that the average cost of housing rose for tenants and homeowners, as well as for most income groups generally. This contribution analyses the development of various components of household expenditure which contributed to these higher ratios. One of the most important considerations here is the fact that average household disposable incomes either fell (tenants) or remained stable (homeowners) during the four-year period under review. This leads to the question of whether these increasing income differences between renting and owning can be attributed to the business cycle alone, or whether they are part of a longer-term trend that will eventually result in a rental sector that provides housing for those on lower incomes. The findings suggest that a longer-term or structural widening of the income gap between renting and owning may indeed be taking place.
KEYWORDS: affordability, expenditure-to-income ratio, income inequality, the Netherlands
Introduction
The decreasing affordability of housing appears to be back on the agenda in many countries (Gabriel et al., 2005; Haffner & Heylen, 2008; Moore & Skaburskis, 2004; Quigley & Raphael, 2004; Whitehead & Yates, 2007). In the Netherlands, housing also became ‘more’ expensive in the period between 2002 and 2006. Since this period includes the economic downturn which followed the burst of the ‘Internet bubble’, the question then becomes whether this declining affordability was the result of an unfavourable economic climate or whether other trends can be discerned that may be causing structural changes in the affordability of housing. Structural changes could relate to the fact that the owner-occupied sector has attained ‘maturity’, since it has grown to be the largest sector in the housing market (Table 1). This happened much later in the Netherlands than in many
ISSN 0267-3037 Print/1466-1810 Online/10/060799 – 22 q 2010 Taylor & Francis DOI: 10.1080/02673037.2010.511472
Correspondence Address: Marietta E. A. Haffner, OTB Research Institute for the Built Environment, Delft University of Technology, PO Box 5030, 2600 GA Delft, the Netherlands. Email: m.e.a.haffner@tudelft.nl Vol. 25, No. 6, 799–820, November 2010
other countries, because for a long time a strong social rental sector that provided housing for large sections of the population prevented such a development (Van der Heijden & Haffner, 2000).
In order to analyse the causes of more expensive housing, the paper first presents a measure of affordability—the expenditure-to-income ratio—and the dataset. Next, since the period of analysis is the period including the aftermath of the ‘dotcom crash’, there is a discussion about the relationship between the economic downturn and the housing market and housing affordability. This includes a description of the most relevant policy instruments that affect affordability. The average affordability ratios for tenants and homeowners are then presented. From this it can be deduced which components of the ratios have contributed to declining affordability. As income developments appear to be a central influence, the paper will then focus on whether this is a structural development. The consequences of these income developments on the affordability of housing are examined in greater detail, before final conclusions are presented.
Method
Out-of-pocket Expenses
The term ‘affordability’, as it is used here, concerns the annual price or a rent paid for housing consumption. It could involve the user costs as price of housing consumption, or the cash flows, out-of-pocket expenses or outlays needed to finance access to housing (Hancock, 1993; Hulchanski, 1995; Stone, 2006a).
The definition of affordability of Maclennan & Williams (1990; see Whitehead, 1991 and Hancock, 1993, p. 129) which is often referred to, distinguishes two other elements alongside ‘a price or a rent of housing’. These are norms that are usually set by government, one concerning the reasonableness of price or rents in relation to household income and the other concerning the quality standard of the housing to be consumed.
In practice, however, when cash flows are measured, they do not usually involve a standard of housing quality, and are sometimes not about a standard of expenditures. It is difficult to design a norm, especially with the well-known expenditure-to-income ratio that measures housing expenditure as a part of the household budget. The proportion spent on housing gives
no indication of whether too much is being spent on housing in relation to other outgoings1
(Baer, 1976; Gabriel et al., 2003; Hulchanski, 1995; Stone, 2006a, 2006b). The amount of income in absolute units will play an important role when determining a norm.
Table 1. Size of housing market sectors (in %) in the Netherlands in 2002 and 2006
Tenure 2002 2006
Rented sector total 45.8 44.2
Social rented sector 36.8 35.2
Private/commercial rented sector 9.0 9.0
Owner-occupied sector total 54.2 55.8
With a mortgage 46.9 48.5
Without a mortgage 7.3 7.3
Total households 100.0 100.0
Source: Haffner & Heylen (2008); based on WoON (2006), OTB/TU Delft calculations.
Although there is no standard for measuring expenditures on housing, the method used in this study is the expenditure-to-income ratio, which is the standard that is usually applied by the Dutch ministry responsible for housing when dealing with issues of housing affordability (Haffner et al., 2008). For this purpose it is appropriate to show changes over time. Table 2 shows the various components of the ratio for renting and homeownership. These include gross expenditure (rent versus mortgage payments), government intervention in housing (housing allowance and income taxation) and necessary incidental expenditures such as utility costs.
Data and Analyses
In order to perform the analyses, the Dutch Housing Survey was used, which was called ‘WBO 2002’ in 2002 and ‘WoON 2006’ in 2006 (Haffner et al., 2008). Both surveys were conducted by the ministry responsible for housing for which the reference dates are 1 January of the years 2002 and 2006. More than 60 000 households were interviewed for WoON 2006, and a comparable number for the WBO 2002. Weighting factors are applied to the sample to make the results valid for all Dutch households, of which there are almost 7 million. In both surveys, income data from the Dutch tax administration were coupled with the survey data.
The databases refer to two cross-sectional databases and the analyses used descriptive methods. That means that average figures for groups are not necessarily applicable to individual households. A group from the 2002 survey may have a different composition to a group in 2006. This is valid when comparing tenants or owner-occupiers in 2002 with those in 2006 respectively. The outcomes described thus have to be interpreted as indicative.
Table 2. Components of housing expenses in relation to the incomeaof the tenant and the owner-occupier (nominal amounts)
Tenant Owner-occupier
Gross rentb Gross housing expenditureb
2 Housing allowances 2 Tax allowances
¼ Net rent ¼ Net housing expenditure
þ Necessary incidental expenditurec þ Necessary incidental expenditurec
¼ Total housing expenditure ¼ Total housing expenditure
Net ratio ¼ net rent/disposable income Net ratio ¼ net housing expenditure/ disposable income
Total ratio ¼ total housing expenditure/ disposable income
Total ratio ¼ total housing expenditure/ disposable income
Notes:aNet disposable household income, including income from business activities, social benefits and social insurance, child benefit, contractual savings, holiday allowance, sickness funds premiums (employer and employee) or sickness insurance premiums, bonuses. The net disposable household income for each month is the annual income registered with the tax administration divided by 12.
bGross rent and gross housing expenditure, including financing costs, property tax and other
property-owner expenditure, such as home insurance fees and ground lease charges. In property-owner-occupation it excludes owner’s maintenance.
cNecessary incidental expenditure, including expenditure on energy and local levies including local
property tax which is to be paid by the owner of the dwelling. Source: Haffner et al. (2008).
The Economic Downturn and the Affordability of Housing
Since the period under analysis, 2002– 2006, followed the ‘dotcom crash’ of 2001, the relationship between the economic downturn and housing affordability is a relevant one. The Dutch economy did not suffer a recession, but GDP growth remained modest at 0.3, 2.2 and 2.0 per cent for the years 2003, 2004 and 2005 respectively, although it showed a steady increase from its lowest point of 0.1 per cent in 2002 (CPB, 2008). This was accompanied by increasing unemployment, which reached 6.5 per cent in 2004 and 2005. Total disposable real household incomes, on the other hand, followed a different trajectory between 2002 and 2006, turning negative in 2003 (22.5 per cent) and 2005 (20.3 per cent), while staying positive in 2004 (þ 0.6 per cent). Since the Consumer Price Index (CPI) for all households remained relatively low in the period 2003– 2005, household disposable income in nominal terms, which is relevant for the analysis of this paper, was only negative in 2003.
The analysis of many recessions, credit crunches (contractions) and falls in asset prices has shown that such impacts on Dutch GDP and unemployment are not uncommon, but still relatively mild (Claessens et al., 2008; IMF, 2009; Reinhart & Rogoff, 2008, 2009). These studies also provide relevant economic statistics for housing, including the development of house prices and residential investment. These indicators will fall, more so in severe recessions than in less-severe recessions, and they will fall more in recessions that are accompanied by credit crunches or falls in house prices than in other types of recessions. The latter type of combined recessions will also take longer to recover from.
Precisely how the affordability of housing for households will be affected by a crisis will depend on the complex interaction of all the variables involved. For example, if only the supply prices of owner-occupied dwellings fall, access to that sector will become cheaper. If interest rates (and inflation) go up at the same time, whether access becomes more affordable will depend on the relative magnitude of both effects. When there is a decrease in the rate of new construction causing demand to exceed supply, house prices may increase over time.
Analysing the effects on the housing market may be complicated by the way in which the housing market is organised. If housing markets do not function as competitive markets, or only partially, as is the case in the Netherlands, the effect on affordability may be different.
The Dutch Regulated Housing Market and Affordability
The owner-occupied market may appear to be a true market at first sight, but on closer inspection it may turn out not to be a competitive market on the supply side. Supply, in terms of numbers of dwellings, appears to be very inelastic in the event of price rises. Figure 1
shows that in the second half of the 1990s house prices rose relatively steeply,2while (and
partly because) the supply of housing did not react to this increase in house prices (see Boelhouwer, 2005). Favourable tax treatment can also be regarded as a factor that has driven up prices when supply is inelastic (Van Ewijk et al., 2006).
An important question in this paper is how these price rises may have affected expenditures on housing. Large increases in gross expenditure on housing could have been expected, but for several reasons that did not happen. Since the mid-1980s, the favourable economic climate had led to a combination of rising household incomes and falling interest rates (Haffner & De Vries, 2009). Furthermore, some developments during the mid-1990s allowed house prices to rise further without causing cash flow problems for households. First, in 1993 mortgage requirements were eased by the inclusion of second household incomes when determining the
amount of the loan. Second, loans which took maximum cash advantage of the mortgage interest deduction in personal income tax became more popular. Endowment loans gained popularity in the 1990s and since 2000 the share of interest-only loans has also grown. With both types of loan, the full amount of interest paid is income-tax deductible against the marginal tax rate after the taxable imputed rent is taken into account for a maximum of 30 years. No repayment of the loan sum takes place. For the endowment loan, the borrower pays into a savings or stock account instead of repaying the loan.
Contrary to the more implicit blockage of the homeownership market on the supply side, in the rental market, market forces have little scope for working (Haffner et al., 2009). Implicit bricks-and-mortar subsidies, rent regulation and housing allowances ‘combine forces’ to keep gross and/or net rents in most of the rental sector more than likely below market rents, and even below cost-price rents in some cases. There will now be a more detailed description of how each of these instruments does that.
Since 1995, bricks-and-mortar subsidies to the owners of social rental dwellings (i.e. the housing associations) have been an ‘implicit’ or ‘non-government’ contribution. At that time, financial ties with the central government were cut. No more bricks-and-mortar subsidies were to be made available for the construction of new social rental dwellings (see also Ouwehand & Van Daalen, 2002). Nevertheless, in line with their non-profit status and due to the uneconomic investment that they can make, housing associations sometimes set a social rent, which is a gross rent below the cost price rent at the start of the lease. This enables social tenants to be implicitly subsidised by housing associations. There are no national regulations concerning this. The ‘good’ price-quality relationship of these dwellings makes them popular and local allocation systems determine which of the households that fall within the target group according to income are allocated social rental housing. This means that waiting lists are common.
Rent regulation in the Netherlands applies to the regulated rental stock which includes 95 per cent of rental stock, with gross rent levels at or below the level of around e615 per month in 2006. Rent regulation includes a maximum rent for each vacant dwelling and a
0 20 40 60 80 100 120 140 160 180 200 1980 1984 1988 1992 1996 2000 2004 2008 Index
1995 = 100 Real house price
Newly built owner-occupied dwellings
Figure 1. Real house price and new-build owner-occupied dwellings, 1978 – 2007 (Index: 1995 ¼ 100). Source: Haffner & de Vries (2009) based on Cadastre Netherlands and Statistics
Netherlands (OTB/TU Delft calculations).
maximum annual rent increase. The maximum rent for a dwelling is based on a dwelling assessment system which evaluates the quality of dwellings in the regulated rented sector using ‘quality points’. The score determines the maximum rent for a certain dwelling and is based on the characteristics of primarily the dwelling and its neighbourhood. As well as the maximum rent permitted, the yearly rent adjustment implemented on 1 July is also determined by parliament. In the period 2002 – 2006, this was set at the rate of inflation plus a specific percentage.
Within the limits of rent regulation, landlords have the freedom to determine rents and rent increases. Rents in the social rental sector tend to be well below the maximum rent, at around 70 per cent of the maximum rent allowed for a dwelling. This illustrates how responsibility for the affordability of rental housing is shared between government and the social rental sector. The gross rent of commercial landlords was 83 per cent of the maximum rent on average. As there are not many dwellings with a market rent, 83 per cent may be an indication of the market rent that is actually achievable, on average. However, it may also be the effect of the large number of dwellings (95 per cent) on the market that are offered for a ‘social’ rent and effectively subsidised via rent regulation. Since 83 per cent is a national average, which of these two explanations applies may well differ according to local circumstances.
Last, but not least, housing allowances influence the net rents of households in the rental sector. One Dutch tenant in three received a housing allowance in 2006, which was an average amount of e148 per month for each recipient (Ministry of VROM, 2007).
In order to set the level of the allowance, the household income and composition as well as the rental price (maximum e615.01 per month in 2006) are taken into account. Above that rent-level, housing allowances are not granted, as shown in Figure 2. Everyone pays a ‘basic rent’ which is income-dependent and is set at e196 for households on the lowest incomes. There is also a ‘quality discount’ limit up to which rents are fully subsidised. This implies that for dwellings whose quality is above the limit, the percentage of rent subsidised will be lower than 100 per cent. For young adults up to the age of 23, the quality discount limit is also the maximum rent (e339) that is subsidised. For other groups, the ‘top-off limits’ are different for small and large households, but up to this limit rent is
339.08 Quality discount 485.33/520.12 Top-off limits 615.01 Housing allowance limit
Youth Singles Two-person Multi-person Elderly
195.89 Basic rent Lower level Lowest income group
0% 0% 0% 0% 0% 100% 100% 100% 100% 100% 75% 75% 75% 50% 50% 75% 0% 0%
Figure 2. Housing allowance system, percentage of monthly rent covered by housing allowance within the set limits, 2006 – 2007. Source: Haffner et al. (2009; from Ministry of VROM,
2006/adapted by OTB).
subsidised at 75 per cent. Finally, there is a housing allowance limit of rent, the maximum rent, up to which the percentage of rent subsidised will be 50 per cent. This will only be applicable to elderly households and single persons.
In conclusion, the policy context (including the past context) influences the affordability of housing in the owner-occupied and rental sectors, although these effects will not always be explicitly visible in the components of housing expenditures as presented in Table 2. What can be seen clearly is government intervention by means of housing allowances and income tax policy.
More implicitly, the policy context affects gross rents for tenants and gross expenditure on housing by homeowners. In the case of renting, there are (implicit) bricks-and-mortar subsidies for social renting and the rent regulation for most (95 per cent) of the rental dwellings that affect the level of gross rent. In the case of owner-occupation, it is a combination of inelastic supply, rising prices and regulation on the mortgage market that influences the level of gross expenditure on housing. Mortgage interest tax relief on owner-occupied dwellings in income tax, is said to push up house prices in cases where supply fails to respond. It allows for higher mortgage loans to be taken out and for mortgage types that take full advantage of the mortgage interest deduction for the whole loan term.
Mediation of Economic Climate Effects on Affordability
A policy context that brings about a situation in which neither owner-occupied housing nor rented housing operate as competitive markets will impact on how house prices and rents will develop in response to a crisis, with regard to either the ‘dotcom bust’ that is relevant to the affordability analysis of this paper, or the current global credit crisis. First,
the lag in the supply of new dwellings has led to a shortage of housing on the market.3
This scarcity has brought about increases in house prices and reduced the risk of any price decreases, as shown in Figure 1. Even the increase in unemployment did not bring about any decrease in demand that may have pushed house prices downwards. As interest rates decreased in the period 2002 – 2006 (CBS, 2009), the tax system would have mitigated any changes in interest rates on out-of-pocket expenditures for existing homeowners because a relatively high proportion of mortgage interest is deductible from income tax. Marginal tax rates range from about one-third up to 52 per cent (DNB, 2008). Another mitigating effect for existing homeowners stems from the relatively low proportion of mortgage loans with variable interest rates, or with an interest rate fixed for a period of less than one year (15 per cent). This means that changes in interest rates have no immediate impact on affordability, and that this effect is delayed until interest rates are periodically reset. Owner-occupied housing thus seems to be cushioned from the effects of the crisis.
Because rents are regulated in the rental sector, good or bad economic times will have no effect on prices as long as the housing market remains tight. Landlords will not have to drop rents, and in economic downturns, the pressure on the rental market usually increases as waiting lists of potential candidates become longer.
In conclusion, changes in prices will mainly influence the affordability of existing homeowners in the longer term. In contrast, regulation in most of the rental sector and the subsidisation of social renting will usually prevent volatility in house prices from affecting the affordability of rental housing. If, however, unemployment increases as it did in the
aftermath of the ‘dotcom bubble’, the households that are affected will have a lower income than before and housing will become less affordable unless households move house.
Affordability
As shown above, the disposable income of households remained relatively stable or decreased in some of the years of the study because of the economic downturn after the ‘dotcom crisis’ of 2001. The questions then become how incomes developed in the period 2002 – 2006 in the databases and what the impact was of this on the affordability of housing for tenants and owner-occupiers.
Affordability in the Rental Sector
Table 3 shows the results of policy on affordability in 2002 and 2006. On average, gross monthly rents rose by 13 per cent, while housing allowances grew by 18 per cent, resulting in an average increase in net rents of 12 per cent. Of all the components that make up total expenditure on housing, expenditure on energy increased the most. Of energy costs, gas (þ 77 per cent) and electricity (þ 50 per cent) increased in particular, as shown in Figure 3. Local taxes decreased because of the abolition of local property tax for occupiers (rather than for owners) on 1 January 2006.
The increase in gross rents was only partly compensated for by an increase in the housing allowance (an average increase of e7.00 in housing allowance compared to an average increase in gross rents of e46.00). Furthermore, necessary incidental expenditure rapidly increased on average by e48.00. None of those increases in net rent or the related necessary incidental expenditures was compensated for by an increase in disposable income; in fact, disposable income fell on average by 2 per cent.
As a result of the increase in expenditure and the decrease in household income, the net expenditure-to-income ratio in the rental sector increased by an average of 3.6 percentage points to 24.2 per cent in the period 2002 – 2006, and the total ratio increased by almost six percentage points to 36.3 per cent.
Table 3. Housing expenditure and average disposable income in the rental sector in e per month, for 2002 and 2006 2002 2006 Change 2002 – 2006 Gross rent 365 411 13% 2 Housing allowances 40 47 18% ¼ Net rent 326 365 12% Net ratio (%) 21.6 24.2
þ Necessary incidental expenditure 129 177 37%
¼ Total housing expenditure 454 542 19%
Total ratio (%) 30.4 36.3
Disposable household income 1755 1714 2 2%
Sources: WBO (2002) (revised) and WoON (2006) (OTB/TU Delft calculations).
The Affordability of Homeownership
Table 4 shows that affordability in the owner-occupied sector developed similarly to that in the rented sector in the period 2002– 2006: gross expenditure rose (þ10 per cent), income tax effect rose more (þ 15 per cent), resulting in a net increase in expenditure of 8 per cent. As in the rental sector, total expenditure on utilities, especially gas and energy as shown in Figure 4, rose by more than 30 per cent.
The increase in monthly gross housing expenditure was compensated only partly by an increase in the fiscal effect of income taxes (on average e22.00 tax relief compared to an extra gross expenditure of e54.00). Furthermore, necessary incidental expenditures rapidly increased by an average of e58.00.
Disposable household income remained stable, causing the income-to-expenditure ratios to rise: the net ratio rose by one percentage point to 16.1 per cent; while the total ratio went up by 3.4 percentage points to 25 per cent. However, these changes in the owner-occupied sector were smaller in percentage terms than in the rental sector.
–20 0 20 40 60 80 100 % change
Gas Water Electricity Local levies Total
Figure 3. Change in the monthly necessary incidental expenditure* in the rental sector between 2002 and 2006, split up in the different components. Sources: WBO (2002) (revised) and WoON (2006) (OTB/TU Delft calculations). Note: *The expenditure on water has remained constant on
average.
Table 4. Housing expenditure and average disposable income in the owner-occupied sector in e per month, for 2002 and 2006
2002 2006 Change 2002 – 2006
Gross housing expenditure 548 602 10%
2 Tax allowances 151 173 15%
¼ Net housing expenditure 397 430 8%
Net ratio (%) 15.1 16.1
þ Necessary incidental expenditure 181 239 32%
¼ Total housing expenditure 578 669 16%
Total ratio (%) 21.6 25.0
Disposable household income 3056 3064 þ 0%
Sources: WBO (2002) (revised) and WoON (2006), (OTB/TU Delft calculations).
Rising Necessary Incidental Expenditure and Stagnating Incomes
As shown above, in the period 2002– 2006, average disposable household income remained stable in the owner-occupied sector and actually fell in the rental sector. At the same time, necessary incidental expenditures rose substantially. A comparison of the period 2002– 2006 with earlier four-year periods starting in 1990 gives greater insight into how this period may differ from previous ones.
Figure 5 shows that 2002– 2006 can be considered exceptional with net rent and total housing expenditure outpacing annual inflation and disposable household incomes declining. Expenditure on energy (i.e. the difference between total housing expenditure and net rent in the figure) increased exceptionally in this period compared to the net rent component of expenditure and also compared to the earlier periods.
For the owner-occupied sector, the same comparison is made between the period 2002 – 2006 with earlier four-year periods. Figure 6 shows that, as with the rental sector, the period 2002 – 2006 can be considered exceptional, with disposable household income being outpaced by annual inflation rates, and sometimes not rising at all, while necessary incidental expenditure in particular increased much faster than net expenditure on housing. The most recent period under study was the first in which the rise in incidental costs was much larger than the increase in net housing expenditure in both the rental sector and the owner-occupied sector. It was also the first time that average disposable household incomes remained stable and even decreased in the rental sector. The income distribution of households in both sectors requires further analysis in order to establish whether this was the result of the ‘dotcom crisis’ alone or whether it signalled a more structural development. Explaining the Gap in Income Between Owner-occupiers and Tenants
Owner-occupiers and tenants had to deal with different levels of expenditure and disposable household incomes in 2006 (and 2002) on average and in nominal terms, as
–20 0 20 40 60 80 100 % change
Gas Water Electricity Local levies Total
Figure 4. Change in monthly necessary incidental expenditure* in the rental sector between 2002 and 2006, split up into different components. Sources: WBO (2002) (revised) and WoON (2006), (OTB/TU Delft calculations). Note: *The expenditure for water has remained constant on
average.
shown previously. Even though monthly net expenditure in the rental sector was lower
than net expenditure among owner-occupiers,4 the difference in disposable household
income meant that the average net expenditure-to-income ratio in the rental sector was much higher (24 per cent) than in the owner-occupied sector (16 per cent). As average disposable household income in the rental sector was 56 per cent lower than that of owner-occupiers, the lower average expenditure on utilities and local levies in the rental sector
0 1 2 3 4 5 6 7 8 9 10 1990-1994 1994-1998 1998-2002 2002-2006 % change Inflation
Net housing expenditure
Total housing expenditures Disposable income household
Figure 6. Annual changes (%) in components of expenditure ratios in owner-occupation compared to inflation between 1990 and 2006. Sources: CPB (2008); Rigo Research en Advies (2004); WBO
(2002) (revised) and WoON (2006), (OTB/TU Delft calculations).
–2 0 2 4 6 8 10 1990-1994 1994-1998 1998-2002 2002-2006 % change
Inflation Net rent Total housing expenditures Disposable income household
Figure 5. Annual changes (%) in components of expenditure ratios in renting compared to inflation, in four periods between 1990 and 2006. Sources: CPB (2008); Rigo Research en Advies (2004);
WBO (2002) (revised) and WoON (2006), (OTB/TU Delft calculations).
put a heavier burden on the average disposable income in the rental sector. The total ratio amounted to 36 per cent for tenants in 2006 and 25 per cent for owner-occupiers. This difference in total ratio was greater in 2006 than in 2002.
These differences may come as a surprise, considering that the relatively large social rental sector of the Netherlands caters for large sections of the population. The question then arises whether this picture of catering for large segments of the population may slowly be becoming outdated. In other words, are these differences between the tenures increasing incidentally or structurally and in the latter case is the rental sector increasingly catering for the lowest-income groups? The following section speculates about such a development. First, there will be a look at the size and average income of different income groups in both tenures. Next, there will be an analysis of the movements of households between tenures and their presumed effect on average income for each group and on household characteristics, as well as the development of income inequality between tenants and owner-occupiers.
Policy Groups: Their Size and Income
The Dutch national government distinguishes two target groups for their housing policy. One group is referred to as the low-income group in Table 5. It consisted of almost 1.8 million households in 2006, which can in principle apply for housing allowances in the rental sector due to their income. This group could be called the policy target group for housing allowances. The second target group consists of this group plus the middle-income group. Both groups together consist of over 3.2 million households and include households
Table 5. Households (absolute numbers and percentages) in different income groups in the rental and owner-occupied sectorsaand their average yearly disposable incomes, in 2002 and 2006
Rental sector Owner-occupied sector Total 2002 2006 2002 2006 2002 2006 Absolute numbers (x 1,000) Low-income groupb 1193 1331 367 467 1560 1798 Middle-income groupb 736 723 635 734 1371 1457 High-income group 1105 950 2590 2595 3695 3545 Total 3035 3004 3592 3797 6627 6800 Percentage Low-income groupb 39 44 10 12 23 26 Middle-income groupb 24 24 18 19 21 21 High-income group 36 32 72 68 56 53 Total 100 100 100 100 100 100
Average disposable household income per year
Low-income groupb 12,500 13,300 15,200 17,600 13,100 14,480
Middle-income groupb 17,800 18,600 21,700 23,500 19,600 21,100
High-income group 32,500 32,300 43,400 44,000 40,100 40,900
Total 21,100 20,600 36,700 36,800 29,500 29,600
Notes:aOver 300,000 households not in independent homes are not included in the analysis.
bThe policy target group for housing allowances is designated the low-income group and the remainder of
the policy target group for social renting as middle-income group.
Sources: WBO (2002) (revised) and WoON (2006), (OTB/TU Delft calculations).
which, on the basis of their income, would qualify for a social rental dwelling. Together they can be referred to as the policy target group for social renting.
Between 2002 and 2006, the low-income group in the rental sector and in the owner-occupied sector increased by 138 000 households and 100 000 households respectively. At the same time, the absolute number of households with a middle income or higher income increased in the owner-occupied sector and decreased in the rental sector. The rental sector decreased in size in itself. These findings would appear to provide evidence that greater numbers from the lowest-income groups were being housed in the rental sector.
The average disposable household income for these groups in nominal terms is also presented in Table 5. The larger proportion of low-income households in renting and owner-occupation will also have had an effect on the development of the average income for each sector. In the rental sector in particular, the effect of the changing composition of households on the decrease in average disposable income of households is visible.
Policy Groups: Their Mobility
The average incomes of those in the rental sector shown in Table 5 indicate that the only group for which the average disposable household income did not increase in nominal terms in the period under review was that of the high-income group in the rental sector. This fact could also be regarded as evidence of structural changes if the average ‘stable’ income for all tenants was the result of these households moving out of the rental sector. The subject of this sub-analysis is whether households on higher incomes seem to be leaving the rental sector. The movements of households in and out of the sectors are analysed in the two years up to 1 January 2006, the reference date of the survey. The low, middle and high-income groups are further subdivided into three income subclasses: lower (bottom 25 per cent of households), middle, and higher (top 25 per cent of households).
Table 6 shows that the number of households moving out of the rental sector among the high-income group was twice as high as those moving into the rental sector. In the other two income groups, moves into the rental sector outnumbered the moves out of the rental sector. For homeowners, it was the other way around. In the low-income group, more households moved out of the sector than moved into it, while in the two other income groups, households moving in outnumbered households moving out. The next section will show that this shift is associated with an increase in older households with less prospect of future income growth. The averages show a higher annual income for those households moving out of the rental sector (e29 500) and a lower income for households moving into the rental sector (e19 200). Again, for the owner-occupied sector this is the other way around: households moving into the owner-occupied sector have a higher income (e33 000) than households moving out (e22 100). The same relationships between the incomes of households moving in and out also applies to each income group within each sector. These figures imply an ongoing structural change in housing the lower income groups in the rental sector, particularly if the average income of the tenants moving out (e29 500) is compared to the overall average income of tenants (e20 600; Table 5).
In the low-income group, the number of tenant households moving in (51 400) is much larger than the group moving out (25 200) for the households in the bottom 25 per cent of that income class. The same difference can be observed for the households in the
Table 6. Household moves into and out of the rental sector and the owner-occupied sector in 2004 and 2005 and their disposable incom e a (in sub-classes and on average) per year in 2005 Low-income group Middle-income group High-income group Total Move in Move out Move in Move out Move in Move out Move in Move out Rental sector Number of moves 123,000 59,000 76,000 54,000 62,000 119,000 262,000 233,000 Lower sub-class 51,400 25,200 23,200 8300 28,400 16,400 103,000 49,900 Middle sub-class 55,100 19,800 45,400 26,100 25,700 77,000 126,200 122,900 Higher sub-class 16,600 14,000 7400 20,100 8200 25,700 32,200 59,800 Average disposable income 11,700 13,000 18,400 22,200 35,200 40,900 19,200 29,500 Owner-occupied sector Number of moves 34,000 43,000 68,000 35,000 154,000 34, 000 256,000 112,000 Lower sub-class 4100 14,100 2800 11,400 18,400 13, 600 25,300 39,100 Middle sub-class 11,000 20,700 37,200 18,200 105,600 15, 100 153,800 54,000 Higher sub-class 18,700 7800 27,800 5200 30,300 5400 76,800 18,400 Average disposable income 18,400 12,400 23,000 18,800 40,600 37, 800 33,000 22,100 Notes: a Income at the moment of questioning and not at the moment of the move; we have designated the policy target group for housing allowances as the low-income group and the remainder of the policy target group for social renting as the middle-income group ; the sub-classes contain 25%, 50% and 25% of households respectively. Source: WoON (2006) (OTB/TU Delft calculations).
income group. In the top 25 per cent of the low-income group, the difference between both groups of movers is much smaller.
Characteristics of Households: Greater Differences Between Tenants and Owner-occupiers
It has been shown that an income gap is opening up between tenants and owner-occupiers. The question then becomes whether income is the only differentiating variable or whether households will also differ increasingly in other characteristics.
Schutjens et al. (2001) find that this is largely the case for a variable as such as household type. According to the results of their logistic regression, families were decreasingly likely to live in a social rental dwelling between 1980 and 1994, while this likelihood increased for singles, one-parent families and two-person households.
Boumeester et al. (1997, pp. 34 – 35) also found selective movements of households into and out of renting and owner-occupation in 1994. They signalled a possible sorting out of certain households in the rental sector: it was found that younger singles, older couples, older singles, more low-income households and more households with an income not derived from employment (retired people or those on social security benefits) were increasingly remaining in the rental sector.
If the effects of households moving through their lifecycle within the rental sector are disregarded, the composition of the group of tenants nowadays clearly differs from that of the group of owner-occupiers, as can be seen in Table 7. In relative terms, the rental sector includes more younger heads of households and, more particularly, more households in which the head is 65 years of age or older. Furthermore, almost half of all tenants are single, while only one in five homeowners is single. Among owner-occupiers, almost three in four heads of households are in paid work versus 47 per cent of tenant heads of households. The other half of
Table 7. Characteristics of households (percentages) in the rental and owner-occupied sector, in 2006
Rental sector Owner-occupied sector Total Age group head household
under 23 years of age 3 0 1
aged 23 – 44 years 37 40 39
aged 45 – 54 years 16 23 20
aged 55 – 64 years 15 19 17
aged 65 years or older 29 17 23
Household composition
single 49 20 32
couple 20 36 29
couple with child(ren) 32 44 39
Labour market position head household
paid work 47 73 62
(early) retirement 32 21 26
other social security benefits 20 5 12
no income 1 1 1
Total 100 100 100
Source: WoON (2006) (OTB/TU Delft calculations).
tenant households derive their income from retirement benefits and other social security benefits. In comparison to the observations from the mid-1990s (Boumeester et al., 1997), the polarisation of both tenures in terms of household characteristics seems to have become more pronounced.
Income Distribution: Greater Differences Between Tenants and Owner-occupiers There is evidence of polarisation between the two forms of tenure, not only in terms of households’ characteristics, but also in terms of longer-term income trends. The Theil coefficient was used as a measure of income inequality in households within the rented and owner-occupier sectors, as well as between the two sectors. The Theil coefficient is equal to the average of the log of all relative income shares. Complete income equality will produce a value of zero (Van den Brakel-Hofmans, 2007). Table 8 shows the results of these calculations.
For those periods in which the Theil coefficient for between-group inequality increases over time (1981 – 2006), income inequality between the tenants and owner-occupiers increases, as can be seen in the last column in Table 8. The Theil coefficient for inequality between groups in the 21st century is, at 0.038, almost 10 times higher than the coefficient for 1989. The development of the Theil coefficient within the two groups between 1989 and 2006 can also be interpreted as supporting this observation. These coefficients seem to decrease slightly over time, indicating that both groups (but tenants at 0.126 more so than the owner-occupiers at 0.134) are becoming more homogeneous in terms of disposable income. These findings underpin the evidence presented so far of the continuing marginalisation of the rental sector in terms of income.
The results also offer some partial support for the findings of Schutjens et al. (2002) that the proportion of lower-income households (households in the first four deciles of household income) in the social rented sector increased between 1981 and 1994. The income gap between social tenants on the one hand and owner-occupiers and private tenants on the other, as calculated using logistic regression, increased relatively rapidly during the 1980s, and slowed during the early 1990s.
The Theil coefficients also demonstrate a more rapid increase in income inequality between tenants and owner-occupiers during the early 1980s. The groups became more
Table 8. Theil coefficients of inequality for disposable household incomeaby tenure, 1981, 1985,
1989, 2002 and 2006 Inequality within the groups
Rental sector Owner-occupied sector Inequality between the groups
1981 0.136 0.145 0.001
1985 0.130 0.135 0.002
1989 0.142 0.146 0.004
2002 0.142 0.136 0.036
2006 0.126 0.134 0.038
Notes:aThe definition of disposable income varies throughout the years. In the 1980s, it is household
income from the survey while this century it is the income registered with the fiscal authorities. Sources: for 1981, 1985 and 1989 Conijn & Lamain (1994); for 2002 WBO (2002) (revised) and for 2006 WoON (2006) (OTB/TU Delft calculations); all years except 2002 first published in Heylen & Haffner (2009).
differentiated in terms of income. If we take what happened between 1981 and 1985, a period of economic recession, as an indication for what may happen in the near future, then the conclusion must be that if the current economic downturn that began in 2007 affects the housing market, tenants and owner-occupiers will become even more homogenous as groups. In general, household incomes may begin to fall and pressure on the rental sector—and the social rental sector in particular—will increase.
Tentative Conclusion
The data suggest that the income gap between the households in the rental sector and those in the owner-occupied sector has widened during the past 10 years. It is reasonable to view the differences in the development of average income and net expenditure-to-income ratio found between 2002 and 2006 as an outcome of that process and not only as a result of the ‘dotcom crisis’. Of course, the process is a gradual one and differences are therefore fairly small. Moreover, only cross-sectional data could be used. However, the shift towards increasing lower-income groups (and fewer households with a middle income) in the rental sector, would seem to be an accurate conclusion. This process was also detected during the 1990s by Schutjens et al. (2002) and Boumeester et al. (1997).
Effects of Increasing Gap in Income on Housing Affordability
This section discusses whether the increasing gap in incomes between tenure types is affecting the affordability of housing in certain ways. Table 9 shows for tenants that the lower the disposable income, the higher the expenditure-to-income ratios were in 2006.
Table 9. Average housing expenditure and disposable income in the rental sector, in e per month, for the different income groups in 2006
Low-income group Middle-income group High-income group Total Gross rent 383 393 465 411 2 Housing allowances 96 11a 5a 47 ¼ Net rent 287 381 461 365 Net ratio 28% 26% 18% 24%
þ Necessary incidental expenditure 172 166 193 177
¼ Total housing expenditure 459 548 654 542
Total ratio 44% 37% 26% 36%
Disposable income households 1105 1547 2695 1714
Number (x 1000) of households 1331 723 950 3004
Notes:aHousing allowance received partly in the year before questionnaire, based on lower income level
at that time.
The policy target group for housing allowances is designated as the low-income group and the remainder of the policy target group for social renting as the middle-income group.
Source: WoON (2006), (OTB/TU Delft calculations).
Engel’s Law of Consumption (the lower the income, the greater the proportion that must be spent on ‘food’) or perhaps even more correctly Schwabe’s Law of Renting (the lower the income, the greater the proportion that will be spent on rent) seem to hold true (Hulchanski, 1995). The net ratio ranges from 28 per cent to 18 per cent across the three groups from low to high income; the total ratio from 44 per cent to 26 per cent. The conclusion must be that households with a relatively low average income do not have relatively low housing costs, despite the fact that a system of housing allowances is in place. The relatively low incomes of tenants, in combination with relatively high energy expenses, produce this result.
This phenomenon can also be observed in the owner-occupied sector, although it is most probably less intense because the sector is growing and becoming more mature. The growth of the sector may contribute to a greater variety of occupants, including among those households that have repaid their mortgage loan, resulting in lower housing expenditure-to-income ratios, as well as the differences in incomes between both tenures. Net ratios range from 26 per cent to 12 per cent across the four income groups and gross ratios from 40 per cent to 16 per cent (see Table 10). The biggest differences in gross ratios between the two tenures can be found in the middle and higher-income groups.
The fact that the households with the lowest incomes, the tenants, have to spend the largest share of their income on housing, on average, is regarded as a result of the fact that differences in incomes are increasing. However, it is also a result of the way that expenditure components developed between 2002 and 2006 for the various income groups. On the basis of net rent and net housing expenses, the average difference in the net expenditure-to-income ratios between income groups changes little, either for tenants or for owner-occupiers. Housing allowances and the tax regime have had effects. For both tenants and owner-occupiers, the approximately equal expenditure on energy accounts for a larger average proportion of income for lower-income households than for higher-income households. Because average higher-incomes among tenants are lower, tenants are hardest hit.
Paradox: A More Effective Housing Policy and Worse Affordability in Rental Sector The results suggest that there is a change in household income within the rental sector. The ‘dotcom crisis’ of 2001 had contributed to that process. However, it seems that there is also a more structural change due to the fact that more special groups, such as the elderly, singles and lower-income groups find a home in the rental sector. Policy instruments in the rental sector, such as rent regulation, housing allowances and the allocation of social rental housing, can help to explain this finding, since they facilitate the entry of low-income groups. The net expenditure-to-income ratios of these households are broadly comparable to the ratios of tenants with higher incomes. Expenditures on energy however create different ratios for the various income groups within the rental sector. So, when in the future housing policy creates a rental sector that caters more for special policy groups, this could also reduce the affordability of housing for tenants. A factor that falls outside housing policy—necessary incidental expenditure—will contribute to this development to a large extent.
Table 10. Average housing expenditure and disposable household income in the owner-occupied sector, in e per month, for the different income groups in 2006 Low-income group a Middle-income group a Higher income group Highest income group Total Gross housing expenditure 467 469 478 1,007 602 2 Tax allowances 99 128 105 396 173 ¼ Net housing expenditure 369 341 373 611 430 Net ratio (%) 26% 18% 14% 12% 16% þ Necessary incidental expenditure 229 217 231 283 239 ¼ Total housing expenditure 598 558 604 894 669 Total ratio (%) 40% 30% 23% 16% 25% Disposable income households 1465 1961 3072 6035 3063 Number of households 467 734 2119 476 3797 Notes: aThe policy target group for housing allowances is designated as the low-income group and the remainder of the policy target group for social renting as the middle-income group. Because of the size of the ‘higher’ income group in the owner-occupied sector, it is split in this Table. Source: WoON (2006) (OTB/TU Delft calculations).
Conclusions
The expenditure-to-income ratios calculated for households in the Netherlands confirm that affordability of housing has declined for tenants and homeowners as well as for most income groups between 2002 and 2006. This paper has analysed which components in household expenditure have contributed to these higher ratios. The fact that average disposable household incomes have either fallen (tenants) or remained stable (homeowners) during the four-year period is one of the most important factors here. The question arose of whether this income development can be considered a once-only reaction to the ‘dotcom crisis’ that began in 2001 or whether it signals that structural changes have taken place in the housing market.
Compared to earlier periods, the 2002 – 2006 period appears to be special in terms of income development, indicating that a more structural development may be taking place because household disposable income for tenants (and a range of subgroups) is much lower than for homeowners. The evidence also suggests that the income profile of households in the various forms of tenure is changing, thereby increasing the income gap between renting and owner-occupation. Some selective movement of households into and out of the rental sector seems to be taking place. The analysis suggests that households with higher incomes are leaving the rental sector, while households with lower incomes are moving into the sector. Some analyses of the development of income inequality over time also suggest that both tenants and homeowners have become more homogeneous as groups, and that the two groups are becoming more differentiated from one another. This development is reflected in the household characteristics of the two groups. Distinct differences between tenants and homeowners confirm structural changes in household incomes in the rental sector. The indicators that were analysed here thus confirm the idea that tenants are becoming a more ‘selective’ group.
These developments are resulting in decreasing affordability within the rental sector, even though various components of housing expenditures are lower in the rental sector than in owner-occupation. The average expenditure-to-income ratio for tenants is higher and rose more rapidly than that of homeowners between 2002 and 2006. With general lower household incomes in the rental sector, the ratios turn out to be highest for the tenants with a low income.
The analyses thus show that more special groups, such as the elderly, singles and lower-income groups live in the rental sector, and that this is not simply the result of the economic downturn during the first few years of this century. Rather, this seems to be a structural change, a widening income gap between the forms of tenure which is in line with the direction in which the decisions of households are being swayed by policy instruments. Policy instruments in the rental sector, such as rent regulation, housing allowances and the allocation of social rental housing, facilitate the entry of low-income groups. In the owner-occupier sector, meanwhile, the higher the income, the bigger in absolute terms the mortgage interest deduction. This effect may certainly sway households with higher incomes to choose homeownership. It is thus more than likely that housing policy is largely responsible for the gap between renting and owning since it incentivises renting for low-income households, particularly when entering the housing market for the first time. As a factor outside housing policy, expenditure on energy reinforces the widening of the gap between both forms of tenure, again because of the difference in income. The rental sector will thus in the future tend to cater more for special policy groups, but this will affect its affordability.
Notes
1 The second well-known indicator of affordability is residual income, which is income left for non-housing needs after having paid for non-housing.
2
House prices in nominal terms rose by almost 30 per cent in the period from the end of 2000 to the end of 2005 (Kadaster, http://www.kadaster.nl), while household incomes could not keep up the pace (Ministry of VROM, 2007). The average nominal house price reached almost e223 000 in 2005 (e165 000 for apartments; e370 000 for a detached single-family dwelling).
3
The policy aim is to reduce scarcity on the housing market by 2010 to 1.5 per cent of housing stock (Ministry of VROM, 2004/2005).
4
These are averages for all homeowners including the almost 15 per cent of households that have repaid their mortgage. For homeowners that moved the average net rate in 2006 is 22.5 per cent for first-time buyers and 20 per cent for other movers.
References
Baer, W. C. (1976) The evolution of housing indicators and housing standards: some lessons for the future, Public Policy, 24, pp. 361 – 393.
Boelhouwer, P. J. (2005) The incomplete privatization of the Dutch housing market: exploding house prices versus falling house-building output, Journal of Housing and the Built Environment, 20, pp. 363 – 378. Boumeester, H., Elsinga, M. & Priemus, H. (1997) De klant in beeld. Marktperspectieven voor
woningcorporaties (Almere: Aedes).
CBS (2009) Renteverlagingen ECB leiden doorgaans tot lagere hypotheekrente, Conjunctuurbericht, donderdag 20 augustus.
Claessens, Stijn, M., Kose, A. & Terrones, M. E. (2008) What happens during recessions, crunches and busts? Working paper (Washington DC: International Monetary Fund).
Conijn, J. B. S. & Lamain, C. J. M. m. m. v. A. A. A. Marie¨n (1994) Individuele woonuitgaven in de huur- en koopsector 1981 – 1989 (Zoetermeer: Ministerie van VROM).
CPB (2008) Macro Economische Verkenningen 2009 (Den Haag: CPB).
DNB (2008) Kentering op de Europese woningmarkten, DNB Kwartaalbericht, September, pp. 41 – 46. Gabriel, M., Jacobs, K., Arthurson, K. & Burke, T. with Yates, J. (2005) Conceptualising and Measuring The
Housing Affordability Problem. Melbourne, Australian Housing and Urban Research Institute, Research Paper 1 in National Research Venture 3.
Haffner, M. & Heylen, K. (2008) Affordability of housing. Flanders and the Netherlands compared in the short-term and long-short-term. Paper presented at the ENHR-conference, Dublin, 6 – 9 June.
Haffner, M. & de Vries, P. (2009) Dutch house price fundamentals. Paper presented at Australian Tax Research Foundation Housing and Tax Symposium, Melbourne, Australia, 11 February.
Haffner, M., Boumeester, H., Dol, K., Goetgeluk, R. & Neuteboom, P. m. m. v. Cor Lamain & Gust Marie¨n (2008) Woonuitgaven 2002 – 2006 in beeld (Delft: Onderzoeksinstituut OTB). Available at http://www. tudelft.nl/live/pagina.jsp?id¼05797765-b837-4727-b1e6-0ce55c7f871d&lang¼nl
Haffner, M. E. A., Hoekstra, J. S. C. M., Oxley, M. & van der Heijden, H. M. H. (2009) Bridging the Gap Between Market and Social Rented Housing in Six European Countries? (Amsterdam: IOS Press).
Hancock, K. E. (1993) ‘Can pay? Won’t pay?’ or economic principles of ‘affordability’, Urban Studies, 30(1), pp. 127 – 145.
Heylen, K. & Haffner, M. (2009) Inkomensongelijkheid en woonuitgaven in Vlaanderen en Nederland, Tijdschrift voor Sociologie, 30, pp. 143 – 171.
Hulchanski, D. J. (1995) The concept of housing affordability: six contemporary uses of the housing expenditure-to-income ratio, Housing Studies, 10, pp. 471 – 491.
IMF (2009) World Economic Outlook. April 2009. Crisis and Recovery (Washington DC: International Monetary Fund).
Maclennan, D. & Williams, R. (1990) Affordable Housing in Britain and the United States (York: Joseph Rowntree Foundation).
Ministerie van VROM (2004 – 2005) Woningbouwafspraken 2005 – 2010 (Den Haag: Ministerie van VROM), Tweede Kamer, Vergaderjaar 2004 – 2005, 27562, nr. 6.
Ministerie van VROM (2006) Huurgerelateerde parameters huurtoeslag 1 juli 2006 t/m 30 juni 2007. Circulaire aan colleges van burgemeesters en wethouders, d.d. 21 maart (Den Haag: Ministerie van VROM).
Ministerie van VROM (2007) Cijfers over Wonen 2006. Feiten over mensen, wensen, wonen (Den Haag: Ministerie van VROM).
Moore, E. & Skaburskis, A. (2004) Canada’s increasing housing affordability burdens, Housing Studies, 19, pp. 395 – 413.
Ouwehand, A. & Van Dalen, G. (2002) Dutch Housing Associations. A Model for Social Housing (Delft: DUP Satellite).
Quigley, J. M. & Raphael, S. (2004) Is housing unaffordable: why isn’t it more affordable? Journal of Economic Perspectives, 18, pp. 191 – 214.
Reinhart, C. M. & Rogoff, K. S. (2008) This time is different: a panoramic view of eight centuries of financial crisis. Working paper 13882 (Cambridge, MA: NBER).
Reinhart, C. M. & Rogoff, K. S. (2009) The aftermath of financial crises. NBER Working Paper Series, working paper 14656.
Rigo Research en Advies (2004) Betaalbaarheid van het wonen (Den Haag: Ministerie van VROM). Schutjens, V. A. J. M., van Kempen, R. & van Weesep, J. (2002) The changing tenant profile of Dutch social
rented housing, Urban Studies, 39, pp. 643 – 664.
Stone, M. E. (2006a) What is housing affordability? The case for the residual income approach, Housing Policy Debate, 17, pp. 151 – 184.
Stone, M. E. (2006b) A housing affordability standard for the UK, Housing Studies, 21, pp. 453 – 476. Van den Brakel-Hofmans, M. (2007) Meten van inkomensongelijkheid. Methoden en definities Inkomen en
bestedingen (Voorburg/Heerlen: Centraal Bureau voor de Statistiek (CBS)).
Van der Heijden, H. & Haffner, M. E. A. (2000) Housing expenditure and housing policy in the West European rental sector, Journal of Housing and the Built Environment, 15, pp. 71 – 92.
Van Ewijk, C., Konink, M., Lever, M. & de Mooij, R. (2006) Economische effecten van aanpassing fiscale behandeling eigen woning (Den Haag: CPB).
Whitehead, C. M. E. (1991) From need to affordability: an analysis of UK housing objectives, Urban Studies, 28, pp. 871 – 887.
Whitehead, C. & Yates, J. (2007) Increasing affordability problems—a role for shared equity products? Experience in Australia and the UK—an update, Housing Finance International, 2, pp. 16 – 20.
