Allocating housing assistance equitably: a comparison of in-kind versus cash subsidies in New Zealand.
The 1992 housing reforms replaced a mixed system of state housing at income-related rents and a cash subsidy for those with high housing costs (the Accommodation Benefit) with a uniformly applied cash subsidy, the Accommodation Supplement. The housing reforms in the year 2000 reversed this change, reintroducing income-related rents. The key issues this paper addresses are whether the 1992 reforms increased or decreased equity by introducing the Accommodation Supplement, and the implications for equity of the reintroduction of income-related rents in 2000. In the analysis, two types of equity are considered--horizontal and vertical. Horizontal equity requires that those with the same housing need receive the same housing assistance, while vertical equity means that those with more need should receive more assistance. The equity of each housing assistance policy hinges on whether the two groups of housing assistance recipients exhibit the same level of housing need. This paper concludes that state house tenants had higher housing needs in 1992 and that, as a result, the 1992 reforms failed to enhance horizontal equity and instead decreased vertical equity.
In 2000 the newly elected Labour government introduced legislation reinstating income-related rents for New Zealand state house tenants. The change represented a partial overturning of earlier National government reforms that abolished income-related rents and introduced a cash subsidy (the Accommodation Supplement) as the primary form of housing assistance.
This reversal in 2000 of the 1992 housing reforms was criticised by the National opposition on the basis that it would inequitably discriminate between state and non-state tenant recipients of housing assistance by providing the former with more assistance than the latter. Similarly, the National government had argued in 1992 that cash assistance was necessary to address this horizontal inequity, and accordingly replaced income-related rents in 1992 with the Accommodation Supplement. The National government clearly outlined the horizontal equity aims of the reforms:
The objective of the Government's housing reform is clear. That is to provide greater equity in the distribution of state housing assistance. The aim is to ensure that families in similar circumstances with similar needs receive the same level of support. (Shipley 1993)
As Campbell (1999) notes, this rhetoric "cast opponents of the new policy as advocates of `special privileges' for state tenants, rather than supporting a `fair go' for all."
This paper investigates whether New Zealand's abolition of income-related rents and the introduction of the Accommodation Supplement in 1992 did, in fact, enhance horizontal equity as its advocates claimed. Horizontal equity, or equal treatment of equals, is defined here as providing equal assistance to "those in equal need" (Musgrave 1976, Musgrave 1990). Vertical equity is similarly interpreted to mean the provision of more assistance to those in more need (Young 1994). If state and non-state recipients of housing assistance were substantively different populations in the sense that their needs differed, then providing both populations with the same assistance does not a priori satisfy horizontal equity. Furthermore, if one group were in more housing need than the other, then reducing the assistance to that group would reduce vertical equity.
TWO HOUSING ASSISTANCE SYSTEMS: TWO APPROACHES TO ACHIEVING EQUITY?
Prior to the 1992 reforms, housing assistance was delivered through a mixed system comprising housing with income-related rents and a cash subsidy for certain beneficiaries. Recipients in "serious housing need" (see Appendix 1 for definition) were able to rent a Housing Corporation of New Zealand (HCNZ) state house at a rent determined by their income (income-related rent). Those with high housing costs, who were not deemed to be in serious housing need, were eligible for cash assistance in the form of an Accommodation Benefit.
The system was based on the idea that it was possible to target the most vulnerable members of society and allocate substantial assistance to them, thereby achieving vertical equity. (2) Horizontal equity would not be affected by this as long as the targeting ensured that those given such special treatment were in more housing need than others who did not qualify.
As mentioned above, the National government's 1992 housing reforms were largely based on the notion that the state housing policy was actually horizontally inequitable because it targeted a certain group and gave them preferential treatment over so many others that were seen to be in similar situations of housing need. As the Social Policy Agency put it:
The system was seen as inequitable because the different conditions applying to the three forms of assistance resulted in significantly more support being given to those receiving subsidised rents than to those receiving cash grants, while other low-income groups received no assistance at all. (Social Policy Agency 1994)
The National government housing assistance policy considered state house tenants to be in no more housing need than those previously eligible for the Accommodation Benefit, despite the fact that they had been previously selected because of their serious housing needs via the points system. (3) State tenants lost eligibility for a greater amount of assistance, the money for housing was spread more thinly over both the public and private sectors, and a fiscal saving resulted. (4)
The government introduced the Accommodation Supplement, under which state house tenants and former Accommodation Benefit recipients received a similar cash subsidy, since subsidies were determined by the same eligibility criteria. (5)
EQUITY HINGING ON HOUSING NEED
Table 1 provides a way to assess how in-kind or cash subsidies affect horizontal and vertical equity. For example, the National government's argument that the higher subsidies to HCNZ tenants were horizontally inequitable falls in cell C. Table 1 shows that establishing whether state house tenants were in more need than other housing assistance recipients (i.e. Accommodation Benefit recipients) is critical to establishing which approach was more equitable.
The next step is therefore to find a way to measure whether state house tenants and other assistance recipients were in the same level of housing need or not. It is possible to measure housing need directly or indirectly. "Direct measurement" methods assess whether recipients are in need of housing assistance by looking at their current housing situation; for example, housing conditions, such as whether the roof leaks, are a direct measure of housing need. Indirect measures, on the other hand, identify recipient circumstances, such as family circumstances, average housing costs and incomes, which suggest prima facie that a recipient will be in need of assistance to meet their own housing needs.
Since housing need is about both housing outcomes and the ability to meet housing needs, it seems necessary to employ both direct and indirect measures. This is also consistent with international approaches. Out of four countries--Canada, Australia, the United Kingdom and the United States--not one has attempted to measure housing need using a purely indirect approach. All have used direct indicators, sometimes by themselves and sometimes in combination with indirect ones.
MEASURING HOUSING NEED USING INDIRECT INDICATORS
Indirect indicators of housing need provide some idea of whether state house tenants and Accommodation Benefit recipients were in the same level of housing need prior to 1992. (6) However, a number of things should be kept in mind when interpreting the data on housing need. The first is that the Accommodation Benefit was allocated on the basis of core economic units while HCNZ houses were allocated to households. (7) To make a definitive comparison it would be necessary to standardise units of analysis to control for the different unit sizes. However, because the available information is historical summary data rather than primary figures, it has not been possible to do this.
In spite of the above reservations, the data indicate that the family circumstances of HCNZ tenants and Accommodation Benefit recipients tended to differ. Accommodation Benefit recipients were much more likely than HCNZ tenants to be single, male, and have no children (Ministry of Housing 1991:11, 1993a:10, Housing Corporation of New Zealand 1988:80). This is probably at least partially attributable to the fact that state houses were targeted at families in need. For example, in 1991/92 49% of state house allocations were to sole parents and only 17% to single people (Ministry of Housing 1993a).
Average Housing Costs
Although it is not possible to standardise measures of housing costs across the two populations because of data limitations, there is some indication that HCNZ clients had, on average, higher pre-subsidy accommodation costs than Accommodation Benefit recipients (Department of Social Welfare 1990, Ministry of Housing 1991). In particular, since HCNZ tenants were more likely to have families than Accommodation Benefit recipients, it also makes sense that the former would need larger, more expensive houses and consequently be in greater housing need.
On average HCNZ tenants had higher incomes than Accommodation Benefit recipients. In 1992 the majority of HCNZ tenants (68%) had weekly incomes between $201 and $275 and 93% received less than $350/week (Ministry of Housing 1993a). By contrast, in 1992, 96% of Accommodation Benefit recipients in November 1991 had net incomes ranging between $108 and $246 per week depending on their benefit (Department of Social Welfare 1992a). The finding that HCNZ clients had higher incomes than Accommodation Benefit recipients might appear to indicate that the former group of recipients was in less financial need than the latter, negating any need for a higher subsidy. However, rather than proving the horizontal inequity of the existing system, as the National government claimed in 1992, this finding highlights the shortcomings of relying solely on indirect measures of need.
On closer examination it is by no means clear that a higher income (or benefit) translates into a diminished need for housing assistance. In 1991 approximately 84% of HCNZ clients (Ministry of Housing 1991) and over 99% of Accommodation Benefit recipients were beneficiaries (Department of Social Welfare 1991, 1992a). This means that the amount of income each group received depended largely on the benefit that most people in each group received.
It should not be surprising that Accommodation Benefit recipients had lower total incomes than HCNZ tenants, since most of the former received lower benefits than the latter (Ministry of Housing 1991, Department of Social Welfare 1991, 1992a). For example, Accommodation Benefit recipients were more likely to be Unemployment beneficiaries, while HCNZ tenants were more likely to be receiving National Superannuation.
Although HCNZ clients had higher incomes (benefits) than Accommodation Benefit recipients, this was to meet their greater overall need, as implicitly acknowledged by the structure of the benefit system. Ironically, the Accommodation Benefit and Accommodation Supplement (since it was based on a very similar formula) explicitly provided those in more need with a lower level of assistance, since income was one of the criteria on which it was assessed. This is because they labelled those on higher benefits as less needy where housing was concerned (see Appendix 2). However, earlier sections in this article have argued that HCNZ tenants were in greater housing need because of their family circumstances and subsequent need for larger housing. This suggests that, given its resources, the housing assistance regime prior to the 1992 reforms was highly effective in targeting state houses to those most in need--HCNZ tenants.
ASSUMPTIONS UNDERPINNING THE 1992 REFORMS
Overall, these data suggest that the introduction of the Accommodation Supplement in 1992, which entailed an equalisation of the housing assistance to both groups, could not, as promised, ensure horizontal equity since the groups were not in the same housing need. Instead, by taking away income from a group successfully targeted as those in greater need of assistance (state house tenants), it reduced vertical equity. In terms of Table 1, this suggests that the combination of in-kind assistance through targeted state houses and income-related rents achieved the outcome in cell A of enhanced vertical equity without compromising horizontal equity.
By contrast, the reforms introducing the Accommodation Supplement as the primary allocation mechanism in 1992 resulted in introducing vertical inequity without improving horizontal equity--i.e., the cell B outcome. Thus, the claim the government made in 1992, that the reforms would increase horizontal equity--i.e., enhancing the horizontally inequitable situation in cell C by moving to cell D--is not supported by the evidence.
MEASURING HOUSING NEED USING DIRECT INDICATORS
The above discussion illustrates how indirect measures of housing need only provide one side of the picture. In contrast, the direct approach involves an effort to identify housing outcomes such as quality and affordability, which are precursors to any societal judgement about the acceptability of housing outcomes. Comparing direct measures of housing need provides a means to assess the relative impacts of the introduction of the Accommodation Supplement on Housing Corporation and Accommodation Benefit populations. This section assesses how the 1992 reforms affected housing affordability and quality for these two populations.
Income-Related Rents Versus Accommodation Benefit Affordability
Housing affordability can be measured in a number of ways (Ministry of Housing 1994, Waldegrave et al. 1996, Krishnan 2001). (8) The measure most commonly employed in New Zealand, and the one to be used here, is the outgoings-to-income ratio. This ratio expresses the proportion of income an economic unit spends on housing costs (or outgoings). (9) To assess affordability, threshold levels or benchmarks are necessary. The threshold with the most evidence to support it is 25%. (10) This benchmark was used as the affordability benchmark for income-related rents in New Zealand.
Therefore, prior to the 1992 reforms state house tenants only paid 25% of their income in rent. Accommodation Benefit recipients, while in need of some assistance, were considered in less need than HCNZ tenants and so received smaller subsidies. As a result HCNZ tenants had lower outgoings-to-income ratios than Accommodation Benefit recipients renting in the private sector, successfully ensuring affordability for state house tenants. (11) Nonetheless, owing to the need to ration scarce houses, many Accommodation beneficiaries who qualified for a state house in terms of serious housing need were left in unaffordable situations. (12) Prior to, and even following, the reforms, state house tenants were, on average, in better affordability positions than housing assistance recipients in the private sector.
Nonetheless, housing affordability for both became much closer to equal after the reforms. Since the Accommodation Supplement formula required that tenants paid more than 25% of their income in rent, access to affordable housing no longer existed for either state house or former Accommodation Benefit recipients after the reforms. Poverty measurement research and data from social agencies and food banks also provide evidence of a decline in affordability (Department of Social Welfare 1993, Waldegrave et al. 1996, Waldegrave 2001).
Table 2 summarises the basic changes to the outgoings-to-income ratios for each group in the early 1990s. The initial impact of the reforms, according to these figures, is an increase in the proportion of state house tenants with outgoings-to-income ratios of over 30% and 40%, and a corresponding decrease in the proportion of HCNZ tenants with outgoings-to-income ratios less than 30%.
Had HCNZ clients and former Accommodation Benefit recipients been similar in terms of their housing need, this would have constituted an improvement in horizontal equity, as the government claimed. However, as the earlier part of this paper showed, since state house tenants had been carefully targeted for their greater level of housing need, horizontal equity was not achieved and vertical inequity created instead.
THE 1992 REFORMS AND HOUSING QUALITY
The term housing quality "covers matters of `adequacy' and `appropriateness' with regard to a range of housing attributes" (King 1994:42). There is considerable agreement in New Zealand and internationally regarding the aspects of housing quality necessary for a definition of housing need (Waldegrave 2001). These are
* condition of building/house amenities;
* location (for example, access to public amenities);
* occupancy (suitability of the house to the space requirements of its occupants); and
* tenure security, which ranges from no tenure (literal homelessness) to complete security of tenure.
The notion of serious housing need employed by HCNZ prior to 1992 incorporated these categories. As a result, estimates of serious housing need before and after the reforms provide the best available information on housing quality.
The findings of a study on serious housing needs in 1992-1993, when compared with the national estimate carried out by the National Housing Commission in 1988, suggested that the numbers of households experiencing serious housing need more than doubled by the end of the housing reforms (Waldegrave and Sawrey 1994). During the implementation of the housing reforms, (13) the incidence of serious housing need increased by 22% (Waldegrave and Sawrey 1994). The Ministry of Housing estimated that between 1994 and 1998 there was a further increase in serious housing need, from approximately 27,000 households to 37,000 households (14) (Preston 1998).
Despite the striking correlation between the change in housing assistance regime and the increase in serious housing need, there is no way to explicitly link the increase in serious housing need (indicative of deteriorating housing quality) to the changes in housing assistance regime.
While it is not possible to explicitly link the housing quality data to the abolition of income-related rents in 1992, it is possible to investigate the effect of the reforms on one aspect of housing quality--tenure security. Tenure security is fundamentally influenced by whether an individual finds housing in the marketplace or through state-provided in-kind assistance.
The Race Relations Office (1991) cites a number of reasons why the market might not produce as secure tenancies for those on low incomes as the state, including:
* the mismatch of demand and supply, giving greater market power to landlords;
* a diminishing rental stock; and
* the inherently adversarial relationship between the tenant (with the need for shelter and a focus for social and work activities) and the landlord (interested in financial rewards above all else) (Race Relations Office 1991:4).
The significant increase in the number of tenants vacating their Housing Corporation tenancy from 7,438 in 1991 to 10,594 in the year ending March 1993 (Department of Social Welfare 1994a:9) indicates less tenure security for those in state houses after the reforms. This change was encouraged via cash incentives to relocate. There is evidence supporting increased urban to rural migration after the reforms. (15) Furthermore, more than five times as many state house tenants moved into the private sector, where tenure is less secure, between 1991 and 1994 compared with the pre-reform period (Waldegrave and Stuart 1997:25).
The institutional change of the HCNZ to Housing New Zealand (HNZ), a quasi-state-owned enterprise, may also have increased tenure insecurity. The requirement that HNZ operate as a business meant that tenants applying for HNZ houses were subject to "normal prudential tenancy requirements". This meant that if they had insufficient funds to pay two weeks rent as bond and two weeks rent in advance, they were simply not allocated a house (New Zealand Council of Christian Social Services 1994:8). The New Zealand Council of Christian Social Services asserted that this meant that some of the poorest members of society could be asked to produce up to $750 before being housed in areas such as Auckland. At the same time, supplementary benefits only provided a maximum of $200 to assist with bond, rent in advance and moving costs (New Zealand Council of Christian Social Services 1994:8).
As well as restrictions on obtaining a house, the requirement that HNZ operate as a business meant that tenants risked eviction if they owed rent. As the New Zealand Council of Christian Social Services noted (1994):
Housing workers report that tenants are suffering considerable anxiety about whether they will be forced to move, especially as many are struggling to pay the rent, and Housing New Zealand issues 90 day eviction notices if rent is even a few days in arrears.
Finally, the reforms allowed HNZ to sell properties when they wished, with no legal obligation to re-house the tenants.
In sum, the change to market rents for state house tenants, and the institutional changes from a socially minded HCNZ to a primarily business-focused HNZ, appear to have resulted in reduced tenure security for state house tenants. This is mainly because many were effectively forced to choose between entering the private sector or facing income shortfalls as a result of the increase in rents. As found in the affordability section, this equalised the circumstances of those in state houses With those in the private sector--by ensuring that tenure was as insecure for the former as for the latter. This change simultaneously failed to enhance horizontal equity and compromised vertical equity.
Summarising the Effect of the Two Policies on Housing Quality
It is important to note the lack of robust, detailed information on housing quality in New Zealand. (16) Bearing this in mind, the information on serious housing need points to a considerable decrease in housing quality over the reform period generally. Every indicator of poor housing quality (except overcrowding) increased. This correlation should be interpreted with caution, however, as the data do not permit an analysis of whether and how the reforms might have contributed to the increase in serious housing need.
The overarching aim of this paper is to establish which of the two approaches to housing assistance was more vertically and horizontally equitable. This was done using both indirect measures of housing need (income, housing costs and family circumstances) and more direct measures (affordability and housing quality).
The direct measures of housing need show recipients of in-kind assistance (through state houses) to be in better situations of housing affordability and quality than former Accommodation Benefit recipients prior to (and, by a small margin, even after) the reforms. However, the indirect measures show that state house tenants were more likely than non-state house tenants to be female, have children, be sole parents, have higher housing costs and be in receipt of National Superannuation--all characteristics usually correlated with a greater need for housing assistance.
The direct and indirect findings therefore provide seemingly contradictory results. While direct indicators show that state house tenants were better off than Accommodation Benefit recipients prior to the reforms, this was the result of the in-kind subsidy offered to the HCNZ tenants, not because the HCNZ tenants were in less need of assistance than Accommodation Benefit recipients. This conclusion is further reinforced by the rapid deterioration of the relative position of state house tenants after the 1992 reforms.
For state house tenants, the reforms increased outgoings-to-income ratios and food bank usage. These changes suggest that state house tenants were probably not in need prior to the reforms largely because the existing housing assistance regime targeted them effectively and reduced their housing need. While the problem of scarce resources did reduce the horizontal equity of the in-kind assistance because not all those in similar situations were able to enter state houses, the 1992 reforms did not solve this problem either.
The evidence suggests that state and non-state house recipients of housing assistance were not in equal need before the reforms, so equalising the subsidies received by both groups did not enhance horizontal equity. Instead, the 1992 reforms actively caused the deterioration of housing affordability and possibly housing quality for a group earlier effectively targeted because of their vulnerability. In other words, the 1992 reforms caused a noticeable deterioration in vertical equity by reducing assistance to a group of people seriously in need of housing assistance. The corollary is that the return to income-related rents for state house tenants in 2000 will not violate horizontal equity as long as it is effectively targeted to those in the most serious housing need (people that have higher housing needs than others). Furthermore, as long as the income-related state houses are effectively targeted in this way, the policy will result in a considerable improvement in vertical equity.
1. Definition of Serious Housing Need
The Ministry of Housing defined serious housing need as the number in "inadequate" housing plus the number in "unaffordable" housing (Preston 1998). Inadequate housing was defined as including those in the following categories:
* the literally homeless;
* those living in crowded dwellings;
* those living in private temporary accommodation;
* those living in dwellings lacking basic amenities such as hot water;
* the disabled living in inappropriate accommodation; and
* those living in seriously substandard dwellings, i.e. those constructed with non-durable materials or which endanger the health of the occupants (Preston 1998:3).
Unaffordability was defined (without justification) as an outgoings-to-income ratio of 50% or more (Preston 1998:3). The estimate of inadequate housing is primarily based on Census data, while the estimate of unaffordable housing is based on Department of Social Welfare tapes for the Accommodation Supplement (Preston 1998:3).
2. Accommodation Benefit/Supplement Formula
The following represents the formula used to determine the level of assistance received under the Accommodation Benefit and then the Accommodation Supplement. The formula, through the income abatement, clearly reduces the level of entitlement by a proportion of "income", which was, in fact, usually benefit.
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]
a = "co-payment" rate (eg. 0.5 means Government pays half, recipient pays half)
O (tenant) = actual rent
O (home owner) = (actual mortgage payment (interest + principal) of owner) + rates (actual) - insurance (dwelling) + maintenance (capped)
b (tenant) = 0.25
b (homeowner) = 0.3
B = relevant benefit or income earned up to the cut-off threshold for the invalids benefit
n = net
g = gross
Y = gross income
CA = cash assets (except accounts, shares etc.)
T = $2,700 single people or $5,400 sole parents, couples and families.
Note: If Y > $8,100 for single people, then there is no entitlement; if Y = $16,200 for sole parents etc. then there is no entitlement.
3. Bank Standard Income Requirements
Banks estimate affordability thresholds to ensure that when they lend, they are likely to get their money back: they serve, to put it bluntly, to ensure that the people receiving bank loans are "good risks". Like the OTI concept, banks use incoming-to-outgoing ratios (ITOs) when determining the affordability and sustainability of paying off a mortgage. Table 3, a survey of private-sector lending guidelines, shows that, at least according to banks, when lending to anyone (not just people with low incomes), the affordability percentage should never be more than 34% (Department of Social Welfare 1992b:2). This might be used as something of an upper affordability percentage over which housing for those on low incomes can most definitely be considered unaffordable.
The reason this is a conservative benchmark is that people on low incomes have to spend a greater proportion of income on essential goods and services than people with higher incomes. The ITO benchmark is designed for people on mid- to high incomes, whereas OTI benchmarks are usually specified for those on very low incomes.
Table 1 Equity of Housing Assistance and Housing Need Extra assistance Accommodation for state house Supplement tenants through provides same income-related rents assistance for everyone State house tenants in A. Vertically B. Vertically more housing need than equitable, no effect inequitable other assistance on horizontal equity recipients State house tenants in C. Horizontally D. Horizontally same need as other inequitable equitable, no assistance recipients effect on vertical equity Table 2 Comparison of Rent/Payment Ratios Pre- and Post-Accommodation Supplement Outgoings-to-income ratios <30% 30%-40% 40%-50% 50%-60% >60% Total Pre-Supplement State Tenants % 47 47 4 1 1 100 Private Tenants % 22 45 24 6 4 100 Post-Supplement State Tenants % 33 48 11 4 3 100 Private Tenants % 23 46 21 6 4 100 Source: Ministry of Housing (1993a:32). Table 3: Survey of Private Sector Lending Guidelines as at 4 February 1992 Lending ITO normal Comments institution range % ANZ Bank 25 ITO of 30% or more for joint incomes only; 25% for lower incomes down to about $27,000 to $28,000; variations at manager's discretion. Auckland 30 No overrides Savings Bank Bank of 25 ITO of 30% or more for joint incomes only; New Zealand 25% for lower incomes down to about $27,000 to $28,000; variations at manager's discretion. Countrywide 30 No overrides Postbank 25 Minimum income determined by residual income; implied lower income about $28,000; any variation of guidelines at manager's discretion. Westpac Bank 25 No overrides; with 25% ITO, implied lower income limit about $27,000 to $28,000 depending on area; flexible loan allows customer to alter loan repayments to a minimum of interest-only ITO of $30 or more for joint incomes only; 25% for lower incomes down to about $27,000 to $28,000; variations at manager's discretion. Source: Department of Social Welfare (1992b:10).
(2) A Cabinet Policy paper submitted by Phil Goff, the Labour government Minister of Housing at the time, put forward both horizontal and vertical equity criteria as the way to judge alternative housing assistance options (Goff 1987).
(3) In its 1989 report to the House of Representatives, HCNZ observed that "Eighty-nine percent of customers housed met the criteria for serious housing need, exceeding its `Serious Housing Need' targets in all major outputs" (Housing Corporation of New Zealand 1989:5). By October 1990 and continuing through to April 1991, 94% of those newly allocated rental units were in serious housing need under the new definition, which excluded those with 50+ points, but who were able to help themselves (Housing Corporation of New Zealand 1990).
(4) In 1991 Cabinet noted the estimated fiscal cost of the post-reform regime was $500 million, indicating net fiscal savings of $220-$250 million when fully implemented (Department of Social Welfare 1992a).
(5) This assumes market rents for state house tenants were set correctly, a result that is refuted by Waldegrave and Stuart's (1997) findings.
(6) Ideally, the measurement of housing need might include indirect indicators such as incomes, housing costs, family size, assets and distribution around New Zealand. Given the lack of available data on the last two, only family circumstances, average housing costs and incomes are analysed.
(7) The notion of a household consists of "one person who usually resides alone or two or more people who usually reside together and share facilities (such as eating facilities, cooking facilities, bathroom and toilet facilities, a living area)" (Statistics New Zealand 1998:80). By contrast, core economic units refer to a family that has some means of supporting itself economically, usually via income. They comprise "either a couple, or at least one parent role/child relationship, or both" (Statistics New Zealand 1998:80).
(8) Proportional measures estimate the appropriate proportion of income that should be spent on housing in order to have sufficient income left over to purchase other necessities.
(9) Housing costs include such things as mortgage payments, property rates, rent and board payments (Krishnan 2001:6).
(10) The 25% figure was based on observations from over 100 years ago that workers tended to pay a week's wages for a month's rent (Meek 1994:20). The United States, Canada and Australia have all used 30% of gross income, or approximately 25% of net income, as their affordability benchmarks (Landt and Bray 1997:8). The 25% benchmark also receives support as a benchmark from an unexpected direction--the standard income requirements for most banks (see Appendix 3).
(11) While 69% of HCNZ tenants had "outgoings to income" ratios of less than 50%, only 53% of Accommodation Benefit recipients did.
(12) By 1990 both the Department of Social Welfare and the Minister of Social Welfare's Office received "considerable correspondence complaining about the limited level of assistance available under accommodation benefit" (Department of Social Welfare 1990d:8). Feedback from the district offices of the Department also indicated that "people on low incomes were unable to afford a reasonable accommodation from the income and are experiencing hardship" (Department of Social Welfare 1990d:8). In any case, available evidence on public opinion at the time suggests that the Accommodation Benefit did not ensure a level of affordability that the public considered acceptable.
(13) The housing reforms were largely implemented between March 1992 and August 1993.
(14) These estimates were adjusted in a number of ways to avoid the potential problems arising out of integration of a number of different aspects of housing need. Both 1994 and 1998 serious housing need figures were, for example, adjusted downward by 5% to allow for double counting.
(15) Waldegrave and Stuart (1997) showed a 100% increase in beneficiaries moving from Wellington to the Wairarapa compared to any previous year. When asked why they moved, 35% of respondents cited cheaper housing or land as a reason for leaving (1997:25).
(16) In future this should be less of a problem as SNZ is currently developing a comprehensive long-term Survey of Housing Need to fill this considerable knowledge deficit.
Campbell, S. (1999) "Restructuring New Zealand housing policy, 1990-1998: An institutional analysis" M.A. Thesis, University of Canterbury, Christchurch.
Department of Social Welfare (1990) "Report of the Accommodation Benefit policy review" DSW File B&P 3/1/5: Vol. 6, Department of Social Welfare, Wellington.
Department of Social Welfare (1991) Statistical Summary, Department of Social Welfare, Wellington.
Department of Social Welfare (1992a) Statistical Report, Department of Social Welfare, Wellington.
Department of Social Welfare (1992b) "Tenure Protection" DSW File IM/7/120/01/PT1, Department of Social Welfare, Wellington.
Department of Social Welfare (1993) "Affordability position of existing Housing Corporation/Te Puni Kokiri/Department of Social Welfare clients" DSW File IM/7/120/01/PT3: Vol. 3, Department of Social Welfare, Wellington.
Department of Social Welfare (1994a) "HNZ accommodation mismatch" DSW File IM/7/120/07/PT1, Department of Social Welfare, Wellington.
Department of Social Welfare (1994b) Statistical Information Report, Department of Social Welfare, Wellington.
Goff, P. (1987) "Objectives, criteria for evaluating housing policies and current housing problems" Cabinet Policy Committee Paper P (87) 57, Wellington.
Housing Corporation of New Zealand (1988) The Final Submission of the Housing Corporation of New Zealand to the Royal Commission on Social Policy, Housing Corporation of New Zealand, Wellington.
Housing Corporation of New Zealand (1989) Report of the HCNZ for year ended 31 March 1989 to the House of Representatives, Appendices to the Journals, Wellington.
Housing Corporation of New Zealand (1990) Statement of Intent for the Three Years Commencing I July 1990, Housing Corporation of New Zealand, Wellington.
King, A. (1994) Towards Indicators of Housing Stress in Australia, Australian Government Publishing Service, Canberra.
Krishnan, V. (2001) "The cost of housing and housing support" Social Policy Journal of New Zealand, 16:109-127.
Landt, John and Rob Bray (1997) "Alternative approaches to measuring rental housing affordability in Australia" Discussion Paper 16, NATSEM (National Centre for Social and Economic Modelling), University of Canberra.
Meek, William Thomas (1994) "The New Zealand housing reforms of the 1990s: the economic and social consequences for the masses" BCom Research Essay, University of Auckland.
Ministry of Housing (1991) "HCNZ client profile survey" Ministry of Housing File 8092, Ministry of Housing, Wellington.
Ministry of Housing (1993a) Post-Election Briefing to the Minister of Housing 1993, Ministry of Housing, Wellington.
Ministry of Housing (1993b) "Post-election briefing to the Minister of Housing 1993" Ministry of Housing File 8301, Ministry of Housing, Wellington.
Ministry of Housing (1994) "Summary of Affordability Studies" Ministry of Housing File ST94/0334, Ministry of Housing, Wellington.
Musgrave, R. (1976) "ET, OT and SB" Journal of Economics, 6:3-16.
Musgrave, R. (1990) "Horizontal equity, once more" National Tax Journal, 43:113-122.
New Zealand Council of Christian Social Services (1994) Government housing policy: An overview, New Zealand Council of Christian Social Services, Wellington.
Preston, N. (1998) Report on Serious Housing Need, Ministry of Housing, Wellington.
Race Relations Office (1991) No Longer Available: A Study of Racial Discrimination and Private Rental Accommodation, Race Relations Office, Wellington.
Shipley, J. (1993) "Reforms aim at tenant equity" The Dominion, 27 August:10.
Social Policy Agency (1994) Foodbanks in New Zealand: Patterns of Growth and Usage, Social Policy Agency, Wellington.
Statistics New Zealand (1998) New Zealand Now: Housing, Statistics New Zealand, Wellington.
Waldegrave, C., and R. Sawrey (1994) The Extent of Serious Housing Need in New Zealand, 1992 and 1993, The Family Centre, Lower Hutt.
Waldegrave, C., R. Stephens, and P. Frater (1996) "Most recent findings in the New Zealand Poverty Measurement Project" Social Work Review, 8(3):22-24.
Waldegrave, C. and S. Stuart (1997) "Out of the rat race: the migration of low income urban families to small town Wairarapa" New Zealand Geographer, 53(1):22-29.
Waldegrave, S. (2001) Allocating Housing Assistance Equitably, M.C.A.Thesis, Victoria University of Wellington.
Young, H. Peyton (1994) Equity: In Theory and Practice, Princeton University Press, Princeton, New Jersey.
The author wishes to thank Arati Mascarenhas and Charles Waldegrave for their valuable contributions to this paper. Thanks also to Dr Karen Baehler and Bob Stephens for their guidance in the production of the thesis on which this paper is based. While the author is grateful for their help, they are not responsible for any part of the text of this paper. The views expressed in this article are solely those of the author.
Stephen Waldegrave Sector Policy Group Ministry of Social Development