A Comparative Study on the Effectiveness of Housing Allowance on Tenant Households in the UK, the Netherlands and Sweden

A Comparative Study on the Effectiveness of Housing Allowance on Tenant Households in the UK, the Netherlands and Sweden Min Ah Jung PhD Student, Department of Urban Studies University of Glasgow E-mail: lucy9683@gmail.com +44 (0)141 330 2751 The purpose of this study is to examine the effectiveness of housing allowance system as a policy instrument of social policy and housing policy. With the consideration of the design of the national housing allowance system in the UK, the Netherlands and Sweden, the changes in household s Residual Income after rents and Rent to Income Ratio before and after housing allowance are studied in terms of tenant groups with different incomes, family types, tenant tenure and employment status. Overall, it seems that housing allowances are likely to be effective to maintain household income after housing costs and reduce financial burden for housing costs. However, the clear relationship between the housing allowance impact and its design has not been found. Introduction In most welfare states, different ways of national support for housing have been implemented for decades for low-income households. As a part of housing support, the housing allowance could reduce the low-income household s burden on housing costs payment and therefore it helps the low-income household maintain their living above a certain level. Also it is noted that where housing allowances are carried out, varied national policy aims are crossed: one is to secure income maintenance as a social policy objective and another is to enhance housing affordability as a housing policy objective. Therefore, when it comes to the housing allowance evaluation, the main research question would be to what extent the housing benefit system is effective to achieve income maintenance and housing affordability for its beneficiaries. Moreover, the structure of housing allowances varies according to countries and even within one county s housing allowance system, the design of housing allowance could make a difference in the achievement of policy goals for varied household groups. To account for this, the study will adopt a comparative approach using three case countries, i.e. the UK, the Netherlands and Sweden and explore the effectiveness of housing allowances on its recipients with regard to these two policy objectives and different housing allowance systems in three countries. 1

Housing Allowances in Welfare States It is widely accepted that housing expenditure has a significant impact on household finance, in particular for low-income households economy, because it is the largest component of their budget (Freeman et al, 1999; Kemp, 2000; Alcock, 2006). Therefore housing has been considered as a subject the consumption in market, most welfare states have intervened in the housing system by employing various policy instruments. The national intervention is categorised into two strands: one is so called bricks and mortar subsidies, i.e. supply-side or producer subsidies; another is known as demand-side or consumer subsidies (see Howenstine, 1975; Quigley, 1979; Davison, 1999; Olsen, 2000; Priemus and Kemp, 2004). In housing studies, it is suggested that the major housing subsidy until the 1970s was bricks and mortar subsidies, i.e. the provision of new dwellings by the state or lowered housing price in market through financial assistances to housing builders and landlords. However demand-side housing subsidies have been increased or introduced in most European countries since the 1970s (Kemp, 2000; Priemus and Kemp, 2004). For the reasons of this policy shift, Kemp (2000) suggested that the emerging of income-related housing allowances in the 1970s is associated with the shifts in other policy areas and explained four main factors as follows: the changing notion of housing policy from public sector provision to market or quasi-market housing; the increased interest on consumer choice for the low-income households; the reduction in severe housing shortages and the growing concern over income problems to be tackled; the better targeting of assistances with housing expenditure for those who are in need, from the point of the selective view to welfare provision (Kemp, 2000, p.46). In practice, welfare states have taken into account the effect of housing costs on household s financial burden and provided financial supports for housing through parts of existing social assistance benefits or the introduction of separate income related housing allowances (see Eardley et al, 1996). And through this process, housing policy in welfare states became more tightly woven with other policy areas, in particular, social security system aiming at poverty protection and income maintenance. The Evaluation of Housing Allowances In any policies related to housing allowances, it is said that the aim of housing allowance is related to the household s living condition after housing costs payments. In practice, housing allowances have functioned as maintaining recipients circumstances and affecting the level of living after housing consumption (Freeman et al, 1999). And this function is related to different policy object, i.e. securing income maintenance for social policy and housing affordability for housing policy. In order to evaluate the effectiveness of housing allowances, two different measures are mainly used, one is residual income after housing costs and another is housing expenditure to income ratio. The concept of residual income, i.e. the amount of income left after housing costs payment takes into account three (or four) components as follows: income, housing costs, housing allowance (and minimum costs of living). Although these elements are not expressed in the same way as seen in table 1, the common issue of residual income measures is concerned with the amount of remaining income after paying housing cost that would be used for living consumption and whether the minimum standard of living would be secured or not. When it comes to housing allowance, the 2

presence and absence of housing allowance affect considerably the burden for housing costs payment and therefore it links to the problem of the income maintenance of households in the lower end of income distribution. From this point of view, the residual income measurement could be used in the evaluation of the income maintenance effect of housing allowance schemes in terms of social policy although it has been employed in the housing affordability measurement in housing literatures. Table 1. The Measurement of Residual Income in Literature Research Randolph (ed) (1992) Chaplin et al (1994) Freeman and Whitehead (1995) Freeman et al (1999) Marshall et al (2000) The Calculation of Residual Income Net weekly household income + estimated Housing Benefit (estimated household applicable amount + Housing Benefit earned income disregard) *1.2 weekly basic rent weekly eligible service charge Net income + Housing Benefit earned income disregard + Housing Allowance Rent Income Support applicable amount Net income + Housing Benefit earned income disregards Rent Income Support applicable amount Net income + Housing Benefit Minimum housing expenditure Minimum nonhousing expenditure Income Rent Income Support + Housing Benefit Source: Randolph (ed), 1992; Chaplin et al, 1994; Freeman and Whitehead, 1995; Freeman et al, 1999; Marshall et al, 2000. In the current housing literature, housing affordability is measured as the proportion of housing expenditure to household s income and it is said that the housing affordability problem occurs when a household pays more than a certain percentage of its income to obtain adequate and appropriate housing (Maclennan and Williams, 1990; Bramley, 1990; Kearns, 1992; Hancock, 1993; Chaplin et al, 1994; Marsh and Riseborough, 1995; Freeman, et al, 1999). While the concept of housing affordability is based on the ratio of housing expenditure to household income, the calculation varies depending on the inclusion or exclusion of housing allowance and the definition of income (net income or gross income) and the scope of housing expenditure (only rent or mortgage payments and housing service charges). For example, the housing affordability for tenant is defined variously as follows (Freeman et al, 1999, p.12): 1) Housing affordability ratio = dwelling rent / net income 2) Housing affordability ratio = (dwelling rent housing allowance) / net income 3) Housing affordability ratio = dwelling rent / (net income + housing allowance) 4) Housing affordability ratio = dwelling rent / gross income 5) Housing affordability ratio = (dwelling rent housing allowance) / gross income However there are practical problems when we measure housing affordability with this ratio standard. Firstly, it is difficult to decide the proportion of housing costs to income as a standard. Secondly, the measurement of housing affordability by using one single ratio is not sensitive to address the variation in households incomes (Kearn, 1992). Moreover the ratio of housing costs to income should be set out at the point where the socially acceptable standard of housing and non-housing goods are consumed (Hancock, 1993). However this ratio measurement does not provide any information about whether the household consumes housing with an acceptable standard. Nevertheless, this ratio measurement is the most widely used measure of affordability due to its simplicity and data availability (Freeman et al, 1999, p.12). In 3

practice, different ratios are employed according to countries and they vary from 15 to 30 per cent of household s income (Joseph Rowntree Memorial Trust International Housing Finance Colloquium, 1989, cited in Kearns, 1992, p.529-530). The research In this study, the impact of housing allowance on income maintenance and housing affordability will be analysed in the UK, the Netherlands and Sweden. Although tenants and homeowners are eligible for housing allowance in the UK and Sweden, the scope of this research is restricted to tenant households because they are likely to be worse off than owner-occupier households and the proportion of housing costs to household s budget and the effectiveness of the housing allowance receipt might be sizable than those for owner-occupier households. Family Resources Survey 2006 of the UK, the Netherlands Housing Research (WoON) 2006 and Household Finances (HEK) 2006 of Sweden are used. For the analysis, two main indicators, i.e. Residual Income after rents and Rent to Income Ratio are employed. Firstly, two residual incomes (after rents) before and after housing allowances are adjusted by OECD modified equivalence scale and the change in households' residual income by housing allowances is analysed. With the result of residual income changes, we will evaluate the impact of housing allowances on income maintenance, i.e. the extent to how the housing allowance contributes to sustain tenant households income after housing costs payment. Secondly, two Rent to Income Ratios (RIR) are compared, one is the ratio of gross rent to household income (i.e. RIR without housing allowance) and another is the ratio of net rent to household income (i.e. RIR with housing allowance). By comparing these two ratios, we could examine how much the housing allowance receipt has an effect on the improvement of housing affordability. Before the use of RIR, we need to decide the standard RIR to judge if a household has a housing affordability problem. In this research, housing affordability problem is defined as follows: if RIR is below 20, there is no housing affordability problem; if RIR is between 20 and 30, there is a housing affordability problem; if RIR is over 30, there is a sever housing affordability problem. That is to say, the higher RIR means the less housing affordability, i.e. the more problems. Adding to indicators, the ratio of housing allowance to gross rent is useful because this shows to what extent the housing allowance could contribute to reduce rent payment after housing allowances. Therefore this ratio is considered with regard to the research result. Finally, the results are compared by using four variables, i.e. income quintiles, household composition, tenant tenure and employment status. While the same categories are used for income quintiles, household composition in three countries, there are variances in categories of tenant tenure and employment status depending on the original national data. The National Housing Allowance System Although housing allowances are means-tested benefits for housing cost payment for low-income households in the UK, the Netherlands and Sweden, there are variances in the structure of housing allowances according to countries. Firstly, while both tenants and homeowners are eligible for housing allowance receipt in the UK and 4

Sweden, only tenant households are eligible for it in the Netherlands. Secondly, as for tenants households, housing allowances are paid for rent plus heating costs in Sweden, it is not paid for heating in the UK and Netherlands. Thirdly, housing allowances are computed differently in three countries. It is more straightforward in the Netherlands than other countries. For example, if the rents are higher than around 200 1, the housing allowances are paid for above it by applying 100%, 75% and 50% subsidy rates. In the UK, if a household receives social security benefit, i.e. Income Support, Jobseeker s Allowance and Pension Credit, housing allowance is paid for 100% of their eligible housing expenditure. However, a household does not receive social security benefit, 65% of the difference between household income and applicable amount are deducted from their housing expenditure and the remains are their housing allowances 2. The calculation of the Swedish housing allowance is more complex than former two countries because there are varied housing allowance schemes according to family types, i.e. family with children, young people (aged 18-28) without children and pensioner. When it comes to non-pensioner households, housing allowances consist of 50-75% of housing expenditure (a) and children grant (b). And if their household incomes are over income threshold, 20% (family with children) or 33% (young people without children) of the difference between household income and income threshold are deducted from (a+b). However there are some common features in three countries housing allowance system. Firstly, the current housing allowance system is advantageous to lower income households than others. For example, this aspect is related to the benefit rule that households receiving social security benefit are entitled to the maximum amount of housing allowance in the UK. Also in the Netherlands, the less personal contribution is applied to lower income households than others. Secondly, the certain types of families are preferred by the housing allowance receipt. For instance, in the UK system, the applicable amount used for the reduction in housing allowance is advantageous to large family rather than single or household without children. In the Netherlands, there are different standards of personal contribution and the subsidy rates according to household composition and these standards are more generous to large family, the elderly and the disabled than others. In Sweden, households with children are advantaged in the benefit rates because the benefit amount are related to the number of children and the standard of eligible rent ceiling, the subsidy rates, taper and income limits are more generous to them than households without children. Also there is another housing costs support for the Swedish pensioner, called special housing supplement, which is additionally paid if disposable income (including housing allowances) of pensioners is smaller than the standard of living. Considering these variances and similarities, there could be hypothesised expectations with regard to the actual function of housing allowances based on the structural characteristics of the housing allowance system between and within countries. Therefore, the design of housing allowance system should be minded when we understand the research result. 1 This price is called norm rent ( 199.52), regarded as personal contribution in the Dutch housing allowance system. 2 Since the introduction of Local Housing Allowance in 2008, Local Reference Rent is applied instead of the eligible rent for private tenants. 5

Table 2. Housing Allowance System in the UK, the Netherlands and Sweden UK Country Benefit Scheme Eligible Housing Expenditure Eligible Tenure Benefit Calculation Housing Benefit Rents and some service charges recipients of IS, JAS(IB), PC non-recipients Local authority Tenants Registered Social Landlord Tenants HB=100% of eligible rent HB=eligible rent 0.65*(assessed income applicable amount) Local Housing Allowance Private Tenants LHA= full standard allowance Income Support for Mortgage Cost -Mortgage interest payment -Interest payments of loans for repairs and improvements -Other housing cost Owner-occupiers The standard rate of interest is applied. LHA= eligible rent ceilings (Local Reference Rent) 0.65*(assessed income applicable amount) Netherlands Rent Allowance Rents Tenant 1) Rents below the norm rent ( 199.52 per month): no housing allowance 2) Rents between the norm rent and the quality discount limit ( 339.08 per month): 100% covered by housing allowance 3) Rents between the quality discount limit and the deduction cut limit: 75% covered by housing allowance - the deduction cut limit: for three or multi-person households: 520.12 per month and for smaller households: 485.33 per month 4) Rents between the deduction cut limit and the liberalisation limit ( 615.01): 50% covered by housing allowance for single-person households, three or multi-person households, elderly people and the disabled. Sweden Housing Allowance for families with children Housing Allowance for young people (18-28) without children Housing Supplement for Pensioners -Rents including heating costs -Annual fees to the cooperative, a part of interest expense on loan -Property tax, 70% of the site leasehold fee, heating and other running expenses 70% of interest expense on housing loan - Tenants - Co-operative/Tenantowned housing residents -Homeowners HA=assessed housing expenditure(a)(75% or 50% of housing expenditure) + child grant (b) If the claimant s estimated income is below an income threshold, (a+b) is provided, or not, 0.2*(estimated income-income threshold) would be reduced from (a+b). HA=assessed housing expenditure (a)(75% or 50% of housing expenditure) If the claimant s estimated income is below an income threshold, (a) is provided, or not, 0.33*(estimated income-income threshold) would be reduced from (a). The difference between the maximum housing supplement and income deduction is paid 6

The Effectiveness of Housing Allowances on Tenant Households Income Maintenance: Changes in Residual Income after Rents by Housing Allowance The result has shown that the housing allowance receipt could make a difference in residual income of housing allowance recipients in three countries: their residual incomes after rents payment have been improved relatively 70% in Sweden, 61% in the UK and 21% in the Netherlands by the housing allowance receipt. With regard to household income, it is generally anticipated that the housing allowance system would be more advantageous to lower income households for benefit receipt. However, it is not always the lowest income group that the support from housing allowance is most helpful. When we look at the housing allowance to rent ratio showing to what extent the housing allowance contribute to the reduction of housing costs, while housing allowances are more helpful for lowest and lower income groups in the UK, it is more helpful for middle lower income groups in the Netherlands and Sweden. Furthermore, rather lowest income group, these groups mentioned above have shown the better improvement in residual income in three countries. It might be because not only income level, other factors such as household types and rents are interrelated to the total housing allowance calculation. In terms of family types, household composition is presumably associated with the more generous housing allowance rates in three countries. As stated previously, there is the less reduction of housing allowance by the advantageous applicable amount for some family types in the UK; varied eligible rent ceilings, personal rent contributions and subsidy rates are applied in the Netherlands; there are different child grants, subsidy rates and tapers according to family types and additional support for pensioners in Sweden. In particular the benefit might be more generous to households with children in terms of higher benefit rates, lower taper (in Sweden), higher ceiling of eligible rent (in Sweden and the Netherland) and other rules than households without children. Therefore, we might expect that certain family types would be more favoured by the housing allowance system than others. However while households with children are likely to receive more housing allowances in three countries and pensioner households are also likely to have more housing allowances in the Netherlands and Sweden, the housing allowance is not always more helpful for households with children and pensioners than others. Considering the housing allowance to rent ratio, the housing allowance receipt is more helpful to reduce financial burden for housing costs for households without children and single pensioners in the UK, single households without children and all pensioners in Sweden but households with children and single pensioner in the Netherlands. That is to say, although the housing allowance is likely to be helpful for some pensioner households as expected, there are differences in the result of non-pensioner households between countries. Moreover, it seems that the effectiveness of housing allowance is likely to be effective for single households than couple households: the improvement in residual income is better for single households excluding pensioners in the UK, all single households in the Netherlands, single pensioner, single without children in Sweden, although the result of couple with children is also good in Sweden. Therefore it is noted that all households with higher ratio of housing allowance to rent have not experienced the more improvement in residual income and it seems that the generosity of housing allowance design for family types is not directly related to the result of housing allowance on income maintenance. 7

In the three countries studied here, the same rule has been applied to tenants within country. Although there are varied rules in the current UK system according to tenure types, i.e. private or non-private, the UK data used here has been collected before Local Housing Allowances rolling-out across country. Therefore it is expected that the housing allowance system might deal with varied tenant types equally. However, with regard to housing allowance to rent ratio between tenants tenure, while the housing allowance receipt is more helpful for non-private housing tenants in the UK, the extent of housing allowance support are at similar level regardless of tenant tenure in Sweden and the Netherlands. Also the UK housing allowance is likely to be more effective for non-private households in terms of income maintenance than private housing tenants. Moreover there are differences in residual income improvement between tenant tenure in the Netherlands and Sweden, the result is better for private housing and housing association housing tenants in Sweden and the Netherlands. From this, we could say that although there are no varied rules according to tenant tenure, non-private tenants are likely to be more favoured by the UK system and also there are differences in income maintenance between tenant types in three countries. Relating to the effectiveness of housing allowance on households with different working status, it is generally expected that the housing allowance system might be more generous to workless households than working households. This is because their working status is related to their income condition (and social security benefit receipt) and family features and it would result in the different benefit receipt. And this anticipation has been linked to the result. As well as the ratio of housing allowances to rent, the improvement of residual income by housing allowance take-up are better for workless households in three countries (although it is not observed a statistically significant difference in the UK households) and part-time working than other fulltime working households in the Netherlands and Sweden. 8

Table 3. Improvement in Residual Income by Housing Allowance (%, relative term) UK Netherlands Sweden Tenant with HA (N=3,376) (N=7,124) (N=1,022) Average (%) 61.0 21.3 70.1 Income Quintiles high 1st Q (110.7) 3rd Q (24.1) 3rd Q (108.3) 2nd Q (92.1) 2nd Q (23.0) 2nd Q (93.7) 3rd Q (53.9) 4th Q (21.7) 1st Q (75.5) 4th Q (36.1) 1st Q (18.8) 4th Q (60.0) low 5th Q (12.2) 5th Q (18.7) 5th Q (12.7) (F=48.111***) (F=52.506***) (F=8.325***) Household Composition high Si without children (85.9) Pensioner Si (26.4) Pensioner Si (111.0) Lone-parent (71.4) Si without children (23.8) Co without children (83.5) Pensioner Si (50.7) Lone-parent (20.0) Si without children (60.7) Co with children (48.3) Co without children (15.9) Pensioner Co (43.5) Co without children (32.1) Pensioner Co (15.6) Lone-parent (28.4) low Pensioner Co (28.2) Co with children (14.5) Co with children (16.8) (F=9.461***) (F=193.008***) (F=8.923***) Tenant Tenure high HA (74.2) Private (22.6) Private or HA (77.9) Public (61.9) HA or Coop (21.2) Coop (34.5) low Private (30.1) Public (16.3) - (F=14.507***) (F=4.770**) (t=-.5.566***) Employment Status high Si/Co-all out of work (63.1) Si/Co-all out of work (22.7) Si out of work (96.4) Si/Co-1 or more in PT (46.9) Si/Co-1 or more in PT (18.5) Co-2 out of work (54.0) Co-1 in FT/ 1 out of work (31.0) Si/Co-all in FT (14.6) Si in work (13.9) Si/Co-all in FT (25.9) Co-1 in FT/ 1 out of work (12.3) Co-only 1 in work (13.9) Low Co-1 in FT/ 1 in PT (14.3) Co-1 in FT/ 1 in PT (11.0) Co-2 in work(8.0) (F=1.556) (F=97.273***) (F=9.974***) Source: Family Resources Survey 2006 (UK), Netherlands Housing Research (WoON) 2006, Household Finances (HEK) 2006 (Sweden), the author calculations. Note: * p<.05 ** p<.01 *** p<.001 Housing Affordability: Changes in the Rent to Income Ratio by Housing Allowance If low-income tenants do not receive their housing allowances, they would face very serious housing affordability problems 3, i.e. high financial burden for housing costs. As table 4 indicates, their Rent to Income Ratio (RIR) before housing allowance takeup are over 40% in the UK and the Netherlands and over 50% in Sweden. It means that these low-income households should spend around half of their income on housing costs and it would decrease non-housing consumption and their level of living. However, these problems have been considerably eased when housing allowances are paid: tenants with housing receipt would not have housing affordability problems in the UK; while they still have housing affordability problems, households financial burden for housing costs are improved by more than 30% (relative term) in the Netherlands and Sweden. 3 The author defines housing affordability problem as follows: if RIR is below 20, there is no housing affordability problem; if RIR is between 20 and 30, there is a housing affordability problem; if RIR is over 30, there is a sever housing affordability problem. 9

Among tenants with housing allowance receipt, housing costs burden is more serious for lowest-income tenants than other tenants in all three countries and they are still placed in the highest rank of housing affordability problem even after housing allowance receipt. In terms of household income groups, the higher housing allowance to rents ratio, i.e. getting more help from housing allowance for rents payment is closely related to the better improvement in RIR by housing allowances: they are lower income groups in the UK and Sweden and the middle three lower income groups in the Netherlands. Also the UK result presents that the housing allowance is likely to be more effective for tackling housing affordability problems than others. As the result presented, while there are no housing affordability problems after housing allowances in the UK across all income quintile groups, the lowest income groups in both countries (and the 4th income quintile in Sweden) still have a severe housing affordability problem. Moreover, the improvement in housing affordability by housing allowances is less for the Dutch and Swedish tenants with housing allowances than the UK tenants. It is noted that households with children have more housing affordability problems than others in three countries. In particular, couple with children have more housing affordability problems than other families, even after they received housing allowance. In the UK and Sweden, this might be related to the result that housing allowance to rent ratio is lower for these households than others. The results of housing affordability improvement between family types are complicated. The RIR changes are larger for household without children and single pensioner in the UK; lone-parent and single pensioner in the Netherlands; all pensioners in Sweden. However, the result is similar to that of household income, these households have faced better housing allowance to rent ratio than others, i.e. the higher housing allowance to rent ratio, the better result. Adding to this, we could look at the difference in the result between countries: while there are no housing affordability problems in the UK households except for couple with children, all family types still have housing affordability problems in the Netherlands and Sweden. As mentioned before, housing allowance receipt is likely to be less helpful for private housing tenants because their housing allowance to rents ratio is smaller than other tenants in the UK. In practice, here are much more differences in RIR between tenant tenure in the UK and the housing allowance is likely to be less effective to deal with housing affordability problems for them than non-private housing tenants in the UK. However, compared to the UK, there are not significant differences in RIR as well as RIR improvement between tenant tenure in the Netherlands and Sweden. Here again, we could find that these two countries system are prone to treat varied tenant groups equally with the reflection of their design. The presence or absence of worker in a family is related to the housing allowance rule such as the eligibility for the best allowance rate or lower deduction rate of housing allowances relating to their income. In practice, the housing allowance is likely to be helpful for workless households than working households in terms of housing allowance to rent ratio. Therefore it is expected that the current benefit system might be more beneficial for workless households than working households. Of course, RIR of workless households are not always lower than all working households even when they received housing allowances, except for Sweden. However in terms of RIR improvement by housing allowances, the anticipated result has been founded in three 10

countries. The RIR improvement is better for workless than working households than working households in all three countries. This might be related to the result that unemployed households have shown higher housing allowances to rent ratio than others with regard to the design of housing allowance. Table 4. Rent to Income Ratios of Tenants Receiving Housing Allowance (%) UK Netherlands Sweden Tenant with HA (N=3,376) (N=7,124) (N=1,022) RIR Gross R IR Net R IR Gross R IR Net R IR Gross R IR Net R IR Average (%) 45.2 9.7 44.0 27.7 54.7 31.0 Income Quintiles high 1st Q (74.3) 1st Q (12.9) 1st Q (53.2) 1st Q (33.9) 1st Q (61.2) 1st Q (36.7) 2nd Q (53.4) 3rd Q (11.4) 2nd Q (45.2) 2nd Q (28.1) 3rd Q (59.6) 4th Q (33.7) 3rd Q (45.3) 4th Q (10.0) 3rd Q (43.4) 3rd Q (26.7) 2nd Q (56.7) 5th Q (29.5) 4th Q (36.9) 2nd Q (9.8) 4th Q (40.5) 4th Q (25.5) 4th Q (55.7) 3rd Q (28.9) how 5th Q (16.2) 5th Q (4.3) 5th Q (37.5) 5th Q (24.5) 5th Q (40.3) 2nd Q (26.4) (F) (261.678***) (16.945***) (348.701***) (291.034***) (41.133***) (11.499***) Household Composition High Co with children Co with children Co with children Co with children Co with children Co with children (62.5) (20.6) (57.5) (36.0) (66.1) (44.2) Lone-parent (56.1) Lone-parent (13.1) Lone-parent (47.5) Co without children (30.0) Pensioner Co (61.2) Lone-parent (36.0) Co without children (53.3) Si without children (8.6) Co without children (45.6) Pensioner Co (28.4) Pensioner Si (54.8) Co without children (33.0) Si without children (47.7) Pensioner Co (7.8) Pensioner Co (44.3) Lone-parent (27.9) Lone-parent (53.2) Si without children (28.0) Pensioner Si Co without Pensioner Si Si without Si without Pensioner Co Low (32.2) Pensioner Co (31.3) children (6.5) Pensioner Si (5.6) (40.0) Si without children (39.4) children (26.2) Pensioner Si (25.0) children (48.6) Co without children (44.7) (27.9) Pensioner Si (26.3) (F) (61.238***) (31.399***) (303.389***) (210.247***) (16.202***) (28.827***) Tenant Tenure high Private (65.3) Private (30.5) HA or Coop (44.0) Public (28.8) Private or HA (56.8) Private or HA (31.8) HA (44.9) HA (8.1) Private (43.5) HA or Coop (27.7) Coop (45.2) Coop (27.8) low Public (39.4) Public (4.5) Public (42.7) Private (27.5) - - (F) or (t) (91.978***) (405.106***) (0.509) (0.374) (-7.313***) (-2.786***) Employment Status high Co-1 in FT/ 1 out of work (70.7) low Si/Co-1 or more in PT (56.6) Si/Co - all in FT (55.9) Si/Co-all out of work (43.8) Co-1 in FT/ 1 in PT (38.4) Co-1 in FT/ 1 out of work (25.5) Co-1 in FT/ 1 in PT (24.8) Co-1 in FT/ 1 out of work (49.9) Co-1 in FT/ 1 out of work (32.9) Co-1 in FT/ 1 in PT (30.7) Si/Co-1 or more in PT (27.9) Si/Co-all out of Co-1 in FT/ 1 in PT (45.1) Si/Co- 1 or more Si/Co-1 or more in PT (23.8) in PT (44.1) Si/Co-all in FT Si/Co-all out of (18.0) work (44.0) work (27.5) Si/Co-all out of Si/Co-all in FT Si/Co-all in FT work (8.1) (36.8) (25.0) Co-2 out of work (64.2) Co-only 1 in work (44.5) Co-only 1 in Co-2 in work work (60.7) (41.8) Si out of work Si in work (35.7) (54.6) Co-2 in work Co-2 out of (49.0) work (34.2) Si in work (46.6) Si out of work (27.3) (F) (11.531***) (46.864***) (34.078***) (32.063***) (16.537***) (26.223***) Source: Family Resources Survey 2006 (UK), Netherlands Housing Research (WoON) 2006, Household Finances (HEK) 2006 (Sweden), the author calculations. Note: Gross R refers to gross rents before housing allowances and Net R refers to net rents after housing allowances (i.e. gross rents minus housing allowances). In Sweden, the sum of rent and heating costs is used as rents in order to reflect the benefit rule. * p<.05 ** p<.01 *** p<.001 11

Table 5. Improvement in RIR by Housing Allowance (%, relative term) UK Netherlands Sweden Tenant with HA (N=3,376) (N=7,124) (N=1,022) Improvement Average (%) 77.5 36.0 40.5 Income quintiles high 1st Q (81.8) 3rd Q (37.9) 2nd Q (50.5) 2nd Q (80.6) 2nd Q (36.9) 3rd Q (48.8) 3rd Q (76.4) 4th Q (36.3) 1st Q (40.2) 4th Q (75.6) 1st Q (34.3) 4th Q (36.2) low 5th Q (73.1) 5th Q (34.3) 5th Q (26.6) (F=7.692***) (F=21.811***) (F=22.366***) Household Composition high Co without children (82.9) Lone-parent (40.1) Pensioner Co (57.2) Pensioner Si (80.2) Pensioner Si (36.8) Pensioner Si (48.5) Si without children (79.9) Co with children (36.6) Si without children (40.4) Lone-parent (74.6) Pensioner Co (34.9) Co with children (31.2) Pensioner Co (73.2) Co without children (33.5) Lone-parent (28.9) low Co with children (71.2) Si without children (32.6) Co without children (21.7) (F=6.979***) (F=57.191***) (F=21.595***) Tenant Tenure high Public (86.6) HA or Coop (36.0) Private or HA (41.0) HA (77.0) Private (35.5) Coop (38.0) low Private (48.2) Public (32.3) - (F=304.758***) (F=1.554) (t=-1.219) Employment Status high Si/Co-all out of work (79.8) Si/Co-all out of work (36.5) Co-2 out of work (47.1) Co-1 in FT/ 1 out of work (63.6) Si/Co-1 or more in PT (35.3) Si out of work (46.8) Si/Co-all in FT (62.5) Co-1 in FT/ 1 out of work (33.1) Co-only 1 in work (23.7) Si/Co-1 or more in PT (56.3) Co-1 in FT/ 1 in PT (31.2) Si in work (22.8) low Co-1 in FT/ 1 in PT (36.5) Si/Co-all in FT(31.1) Co-2 in work (14.7) (F=37.295***) (F=18.976***) (F=36.560***) Source: Family Resources Survey 2006 (UK), Netherlands Housing Research (WoON) 2006, Household Finances (HEK) 2006 (Sweden), the author calculations. Note: * p<.05 ** p<.01 *** p<.001 Conclusions This study has focused on the effectiveness of housing allowances on income maintenance and housing affordability of low-income tenants. The result has presented that the housing allowance plays a role in the improvement of income maintenance and housing affordability by reducing housing costs burden actually. Also it seems that the result is related to some features of housing allowance in each country. Social security benefit recipients in the UK are eligible for maximum housing allowance receipt and this could result in the better improvement in lowest income households. In terms of family type, the housing allowances are likely to be more effective for pensioner households as expected, in particular in the Netherlands and Sweden where the benefit system is more generous for them. Moreover, while the housing allowance system does not treat tenant tenure differently, the UK private housing tenants are less advantaged in housing allowance receipt and housing allowances is less effective for them than non private housing tenants. Commonly, it is said that housing allowance is more effective for workless household than working households. This might be because the employment status and the absence of earnings 12

are related to the social security benefit entitlement and it might make their benefit receipt better for them in three countries. However, while the result has indicated the effectiveness of housing allowances on income maintenance and housing affordability and some disadvantaged households might have more help through housing allowance receipt as expected, it is hard to say that there is the clear relationship between the result and the feature of national housing allowance system between countries. There are differences between the expectation based on the housing allowance design and the actual support from housing allowance take-up. Also while some household types would be favoured by the current system in one country, the interaction of household's living condition and the rule of housing allowances within country could cause different result. Clearly, the housing allowances own the positive function for income security and housing affordability for low-income tenant households in three countries. However, it is arguable which aspect of the national housing allowances are more closely related to these functions and the further study should be required in order to understand the relationship or interaction more clearly. Acknowledgment This paper is based on the author s PhD thesis. I would like to thank my supervisors, Mark Stephens and Kenneth Gibb. References Alcock, P. (2006) Understanding Poverty, 3 rd Macmillan. Edition, Hampshire: Palgrave Bramley, G. (1990) Access, Affordability and Housing Need, Paper presented at ESRC Housing Studies Conference, University of Surrey, September 1990. Chaplin, R., Martin, S., Yang, J. H. and Whitehead, C. M. E. (1994) Affordability: Definitions, Measures and Implications for Lenders, Discussion Paper No. 45, Department of Land Economy, University of Cambridge. Davison, A. (1999) Alternative Model of Social Housing: Tenure patterns and Costrenting in New Zealand and Sweden, Housing Studies, Vol. 14(4), pp.453-472. Department for Work and Pensions (2002) Building Choice and Responsibility: A Radical Agenda for Housing Benefit, London: HMSO. Department for Work and Pensions (2005) Receiving the LHA: claimants early experiences of the LHA in the nine Pathfinder areas. http://webarchive.nationalarchives.gov.uk/+/http://www.dwp.gov.uk/housingbenefit/l ha/evaluation/2005/lha6-receiving.pdf, accessed on 14 June 2010. Department for Work and Pensions Housing Benefit Site. http://www.dwp.gov.uk/housingbenefit, accessed on 14 June 2010. 13

Department for Work and Pensions (2007) Housing Benefit Local Housing Allowance Guidance Manual. http://webarchive.nationalarchives.gov.uk/+/http://www.dwp.gov.uk/housingbenefit/l ha/lha-guidance-manual-amd1.pdf, accessed on 14 June 2010. Department for Work and Pensions (2008) Family Resources Survey 2006-07, accessed via the UK Data Archive. https://www.data-archive.ac.uk/findingdata/sndescription.asp?sn=6079&key=frs, accessed on 14 June 2010. Doling, J. (1997) Comparative Housing Policy: Government and Housing in Advanced Capitalist Countries, London: Macmillan. Eardley, T., Bradshaw, J., Ditch, J., Gough, I. and Whiteford, P. (1996) Social Assistance in OECD Countries: Synthesis Report, Department of Social Security Research No. 46, London: HMSO. Freeman, A. and Whitehead, C. M. E. (1995) Affordability Revisited in English Housing Association New Let Accommodation 1992-1995, Discussion Paper No. 88, Department of Land Economy, University of Cambridge. Freeman, A., Holmans, A. and Whitehead, C. M. E. (1999) Evaluating Housing Affordability: Main Research Report for the CIH, LGA and NHF, Property Research Unit, Department of Land Economy, University of Cambridge and London School of Political Economics. Hancock, K. E. (1993) Can t Pay? Won t Pay? or Economic Principles of Affordability, Urban Studies, Vol. 30(1), pp.127-145. Howenstine, E. J. (1975) The Changing Role of Housing Production Subsidies and Consumer Housing Subsidies in European National Housing Policy, Land Economics, Vol. 51(1), pp.86-94. Kearns, A. (1992) Affordability of Housing Association Tenants: A Key Issue for British Social Housing Policy, Journal of Social Policy, Vol. 21(Part1), pp.525-549. Kemp, P. A. (2000) The Role and Design of Income-related Housing Allowances, International Social Security Review, Vol. 53(3), pp.43-57. Maclennan, D. and Williams, R. (1990) Housing Concerns and Affordability in Britain and the United States, in Maclennan, D. and Williams, R. (eds) (1990) Affordable Housing in Britain and America, York: Joseph Rowntree Foundation. Marsh, A. and Riseborough, M. (1995) Making Ends Meet: Older People, Housing Association Costs and the Affordability of Rented Housing, London: National Federation of Housing Associations. Marshall, D., Lyall Grant, F. and Freeman, A. (2000) Getting Rents Right? : The Place of Affordability in the Rent Setting Process: A Summary Report, Dataspring 14

Discussion Paper No. 2, Cambridge Housing and Planning Research, University of Cambridge. OECD Benefits and Wages: Statistics, OECD Tax-Ben country specific files, the Netherlands (2005). http://www.oecd.org/dataoecd/38/12/39676279.pdf, accessed on 14 June 2010. OECD Benefits and Wages: Statistics, OECD Tax-Ben country specific files (2006) the UK and Sweden. http://www.oecd.org/dataoecd/15/44/41579073.pdf, accessed on 14 June 2010. http://www.oecd.org/dataoecd/16/22/41579130.pdf, accessed on 14 June 2010. Olsen, E. O. (2000) The Cost-Effectiveness of Alternative Methods of Delivering Housing Subsidies, University of Virginia. http://www.virginia.edu/economics/workshops/papers/olsen/cesurvey2009.pdf, accessed on 14 June 2010. Priemus, H. and Kemp, P. A. (2004) The Present and Future of Income-related Housing Support: Debates in Britain and the Netherlands, Housing Studies, Vol. 19(4), pp.653-668. Quigley, J. M. (1979) Housing Allowances and Demand Oriented Housing Subsidies, Policy Studies Journal, Vol. 8(2) Special No. 1, pp. 300-306. Randolph, B. (1992) Housing Associations After the Act, Associations Research Report No. 18, London: National Federation of Housing Associations. Statistics Sweden, Household Finances (HEK) 2006, http://www.scb.se. The Ministry of Housing, Spatial Planning and Environment (VROM) international homepage, http://international.vrom.nl/, accessed on 14 June 2010. The Ministry of Housing, Spatial Planning and the Environment (VROM), Netherlands Housing Research (WoON) 2006, accessed via the Netherlands Data Archiving and Networked Services, http://easy.dans.knaw.nl/dms, accessed on 14 June 2010. The Swedish Social Insurance Agency (2005) Social Insurance. http://www.forsakringskassan.se/sprak/eng/engelska.pdf, accessed on 14 June 2010. The Swedish Social Insurance Agency (2007) Social Insurance in Sweden 2006. http://www.forsakringskassan.se/filer/publikationer/pdf/sfb06-e.pdf, accessed on 14 June 2010. The Swedish Social Insurance Agency homepage, Help with the cost of housing. http://www.forsakringskassan.se/nav/64ba8021e59365afde8ca2a89cd23f34, accessed on 14 June 2010. 15