This paper discusses concepts of housing affordability and policies for increasing housing affordability. Most current measures of housing affordability used in Australia are based on the acquisition costs of housing. They define housing costs in nominal rather than in real terms and include mortgage repayments that are properly regarded as savings. Moreover, they do not account for household choice over type of house or household composition. The first part of this paper proposes that measures of housing affordability should be based on real housing user costs or rents. Turning to policy issues, the paper argues that housing affordability is essentially a household income problem made worse by government restrictions on housing supply. High housing costs do not reflect housing market failures. To reduce housing costs, Government should allow more housing in established and greenfield areas. Government may also improve housing affordability by subsidising housing for low income households. However, subsidies to urban infrastructure nearly always raise the price of land rather than reduce the price of housing. Also, the Australian Government’s proposed national housing and rental affordability funds are poorly defined and likely to be ineffective.
Keywords: Housing costs. Market failures. Housing policies.
JEL: H20, H12, R31, R38.
* The author is grateful for the comments of two referees on the draft manuscript. Any errors or omissions remain the author’s responsibility.
Concern about the high cost of housing is perennial but peaks from time to time. In Australia, concerns peaked in the mid-1970s and led to the establishment of the National Indicative Planning Council for Housing. They peaked in 1988/89 and there followed a national study of housing costs (Applied Economics and Travers Morgan, 1992; Abelson 1994). They peaked in 2002/03 and there followed a major report on housing affordability by the Productivity Commission (2004). Concerns peaked again in 2007/08 and led to a raft of Commonwealth policy proposals for the home owning and renter sectors.
A common feature of most of the concerns about housing costs and the related studies is their focus on house prices and home ownership rather than rents. However whereas the first three peaks cited above reflected a rapid rise in both nominal and real house prices in the preceding three or so years, the decline in housing affordability in 2007-08 reflected a 40 per cent increase in mortgage rates over three years when house prices in many parts of Australia were falling in real terms. However, it may also be observed that despite these periodical crises over housing costs, the home ownership rate in 2006 (at 69.8 per cent) was higher than in 1971, 1991 or 2001 and only marginally lower than in 1981 when it was 70.1 per cent (ABS, 2007).
Despite the volume of comments in the media and the regular publication of housing affordability indices by the Real Estate Institute of Australia, the Commonwealth Bank / Housing Industry Association and Bis-Shrapnel, housing affordability is poorly understood. There is no consensus about what housing affordability means and much of the discussion about affordability is based on questionable definitions and measures. Moreover, current Australian government proposals to reduce housing costs appear to be ad hoc responses to perceived problems rather than solutions based on an analysis of market and government failures in the housing sector and the major cause of low housing affordability (low household income).
This paper contains two main parts. The first part reviews the main definitions of housing affordability used in Australia and suggests how affordability can be assessed in a more meaningful way. The second part discusses policies for increasing housing affordability.
This section outlines the main concepts of housing affordability used in Australia, provides a critique of these concepts, and then provides alternative measures of housing affordability based on real user costs or rents.
Housing affordability introduced
The Australian Housing, Local Government and Planning Ministers (2006) defined affordable housing as:
● housing that is appropriate for the needs of a range of low and moderate income households; and
● priced so that households are able to meet other essential basic living costs.
This is doubtless a generally acceptable high level definition, but it lacks specific or operational guidelines.
Perhaps the simplest definition of housing affordability is based on the ratio of house prices to income. For example, the Productivity Commission (2004) estimated that for Australia as a whole the ratio of the median house price to average per capita income rose from about 6 in the mid-1990s to about 9 in 2004. Yates (2008) estimates a fourfold growth in this ratio between 1960 and 2006. Note that this measure of affordability relates house prices to individual rather than to household income. Also it is a relative measure. Neither of these reports defines a house price to income ratio that is affordable.
The UK Office of the Deputy Prime Minister (2005) defined affordability in a similar way but focused more on lower income households as “the ratio of lower quartile house price to incomes”. The report (Appendix 1) estimates that this ratio was around 8 in London and the South East compared with just below 6 in the Midlands and about 4.5 in the North of England. These ratios had risen substantially in all regions since the mid-1990s, typically by about 50 per cent but by more in London than in the North. However, the report does not prescribe a house price to household income ratio that is not affordable.
A major weakness of the house price – income ratio, as the Productivity Commission (2004) acknowledged, is that it ignores the cost of housing finance. The approximately 40 per cent fall in real interest rates in Australia between the mid-1990s and 2004 was a significant driver of the increase in house prices (Abelson et al. 2005). It also largely offset the cost effect of the higher house prices.
Measuring housing affordability by the ratio of housing costs (defined in various ways) to household income allows for both house prices and the cost of housing finance. In their survey of housing affordability, Marks and Sedgwick (2008) state that “using a standard measure of housing stress – a household is defined as being in housing stress if housing costs exceed 30 per cent of income”. The Australian Council of Social Services (ACOSS et al., 2007) adopted the same definition.
Of course, housing costs and income need to be defined.1 For renters, housing costs are rental payments. For home owners, housing costs are widely taken to be mortgage repayments plus interest payments. However, as we will see, there are several problems with this definition of housing costs. Turning to household income, Marks and Sedgwick (2008) define this as equivalised disposable household income. This definition of income seems conceptually appropriate (Abelson, 2008) but equivalence requires judgments about household equivalence and there may be practical problems in estimating equivalised disposable household income.
Moreover, as Marks and Sedgwick (2008) acknowledge, the judgment that housing costs in excess of 30 per cent of household income represent “housing stress” is arbitrary. Indices of housing affordability avoid this arbitrary judgment but can provide only measures of relative housing affordability over time and place. The three most common such indices in use in Australia are:
1. The Real Estate Institute of Australia and AMP Home Loan Affordability Index: the ratio of median family income to average new loan repayments.
2. The Commonwealth Bank of Australia and Housing Industry of Australia Housing Affordability Index: the ratio of average disposable income per household to the income required to meet repayments on a 25 year loan for 80 per cent of a median price home purchased by first home buyers.
3. The Bis-Shrapnel Home Loan Affordability Index: the proportion of full-time male earnings needed to meet the mortgage repayments on a ‘typical’ housing loan.
Another index that is sometimes used is the deposit gap. This is the difference between the price of a typical dwelling and the maximum loan that could be repaid from average household income expressed as a proportion of that income (see Yates, 2008, Figure 2 for some estimates).
Observations on the housing affordability measures
These (home-ownership) affordability indices share several weaknesses. They are based on the costs of acquiring housing rather than on estimates of real housing user costs. They also ignore the alternative of rental costs. As we will see, these home ownership indices are biased measures of housing cost and affordability.
The interest payments in these affordability indices are based on nominal rather than real interest rates. Holding house prices constant, affordability as defined above always declines with rises in nominal interest rates. However, the component of the interest payment that reflects inflation is in effect a repayment of debt (or investment in equity). Equivalently, there is no allowance for capital gains in house prices. Suppose that a household purchases a house for $400,000 with 100 per cent debt and pays 8 per cent interest with inflation running at 4 per cent. The nominal interest payment would be $32,000, but (ignoring any real changes in house prices) the house price would appreciate by $16,000 to $416,000. The difference between the nominal interest payment and the capital appreciation is equal to the real interest rate of 4 per cent. At the end of the year, debt on the house has fallen from 100 per cent to 96.2 per cent of its market value.
All the affordability indices noted above include the principal component of mortgage repayments. However, as the ABS (2007, p.20) observed, these repayments are a “form of saving rather than a recurrent housing cost”. Mortgage repayments are not included in the consumer price index. Financial liability falls as debt falls. It is true that mortgage repayments can present a short-term liquidity problem and that current household consumption may be constrained if households cannot borrow against the increased equity in the house. However, future consumption rises as a result of this saving.
It may also be noted that all the affordability measures are short-term. Interest rates may be at high end of the business cycle and expected to fall. Short-term issues are important, but less important than long-term ones.
Nor do the housing affordability measures recognize the role of household choice. As the ABS (2007) reports, between 1994-95 and 2005-06, average dwelling size in Australia increased from 2.88 bedrooms to 3.06 bedrooms per dwelling while the average household size fell from 2.69 to 2.51 persons per dwelling. This surely implies increasing housing affordability!
The price that households pay for a house reflects their preferences over house size and location relative to other budget choices. For any given level of income and housing type, some households choose to spend more in housing. Others choose inferior housing and pay less for housing. It is not clear that the former group is worse off than the latter one.
It may also be noted that none of the affordability measure allows for travel costs. Some households spend more on housing in order to reduce travel costs. Other households reduce housing costs but spend more on travel. Households that spend less on housing but more on travel may be no better off than those who spend more on housing but less on travel.
Moreover, households make choices over household size and composition. Thus household income is also a matter of choice, at least for some households.
The role of household choice is apparent from close consideration of the three main indices.
● For the REIA Home Loan Affordability Index, new loan repayments represent a choice over gearing as well as house.
● For the CBA-HIA Housing Affordability Index, the median price of dwellings purchased by first home owners is also a choice. Between the March quarter 2005 and March quarter 2007, this median dwelling price rose by 21 per cent, which was significantly higher than the rise in the median price for all dwellings. In many areas house prices did not rise by this amount.
● For the Bis-Shrapnel Home Loan Affordability Index, a ‘typical’ home loan represents a choice over gearing as well as house.
All this is not to deny that housing costs are often a serious problem for lower income households. However, they are a problem for most low income households not just for households that choose to spend a high proportion of disposable income on housing.
Real housing user costs
For consumption of a durable good such as housing, the real cost is the cost of consuming housing services (as owner or renter) over a period such as a quarter or year. This is the amount that it costs to occupy a house or apartment with no change in a household’s real net worth over the period.2 Equivalently, the real cost of housing is the amount of other goods and services foregone as a result of use of the house. This cost includes the loss of after-tax income from a homeowner’s own equity in a house. The two main ways to estimate these costs are the user cost approach and the rental payment or equivalence approach.
As we have observed, the real user cost of housing includes only the real component of interest rates. It does not include that part of interest rates that represents inflation. This component is fully offset by the increase in the nominal market value of the home and in the homeowner’s net equity in the home.
However, it is necessary to allow for any real depreciation or appreciation in the value of the building over the relevant period. Typically, without maintenance expenditures structures may depreciate. But with maintenance they may hold their value.3 Thus maintenance expense can be substituted for depreciation. On the other hand, land value may increase with scarcity of land. Thus, there may be some real appreciation of house prices inclusive of land values values in any period. The historic evidence is that, after allowing for maintenance expenses, real house prices appreciate by some 1 per cent per annum in Australia (Abelson et al, 2005).
Accordingly, ignoring taxes and assuming that households face a unique real interest rate, real housing user costs (RHUC) are a function of real interest payments and real maintenance expenses less any real increase in house values (if any).
RHUC = P0r + P0 m/(1+π) – [(P1)/1+π) - P0] (1)
P = house price,
r = the real rate of return on capital,
m = nominal maintenance costs per annum as a percentage of house price,
π = the rate of inflation per period such as a year, and
Subscripts 0 and 1 denote house prices at the beginning and end of the period
Equation (1) shows housing user costs in terms of prices prevailing at the start of the period. The final term is the change in real house prices over the period. Suppose that the value of someone’s house rises by 4 per cent over a year from $400,000 to $416,000 and that this represents 3 per cent general inflation and approximately 1 per cent real increase in value. In terms of purchasing power (or real asset value) the house at the end of the period is worth $416,000/1.03 = $403,884 compared with $400,000 at the start of the year. Thus the homeowner has increased his or her net worth and is better off by $3884 over the year.
Actually, the net worth of the homeowner may increase by more than this. Suppose that the owner has a mortgage of 70% of the house value ($280,000) at the start of the year and that he or she makes no loan repayments during the year. At the end of the year, the real cost of the mortgage is $280,000/1.03 = $271,844. The homeowner is a further $8155 better off. Thus the net worth of the homeowner has risen by $12,039 in terms of the prices prevailing at the start of the year.
Another way of expressing this is that the equity in the home has risen from $120,000 ($400,000 - $280,000) to $136,000 ($416,000 - $280,000). In terms of prices at the start of the year, the value of the equity at the end of the year is $136,000 / 1.03 = 132,039. Thus the value of equity has risen by $12,039 in real terms (which is equivalent to the calculation above).
This increase in homeowner’s equity depends only partly on the assumed real house price increase of 1 per cent. If there were no increase in the real house price and the house price rose to $412,000 simply with inflation, the owner’s equity would still have risen from $120,000 to $132,000. In real terms, the latter would be worth $132,000 / 1.03 = $128,155. In other words, the owner would still be better off by $8,155 even with no loan repayments and no real increase in house prices!
This occurs because the nominal component of an interest payment (which reflects inflation) is implicitly a loan repayment. Thus if the mortgage rate is 7 per cent and the inflation rate is 3 per cent, the real rate of interest is 4 per cent and the nominal component is 3 per cent. The real user cost of housing depends only on the real component of interest rates. The nominal component is offset by the increase in the market value of housing.
Various complications may now be introduced. Specifically we allow below that homeowners may use their own equity to finance part of the cost of the home, that the gross nominal rate of return on own investments differs from the mortgage rate, that they pay tax on nominal returns on their investments but get no tax benefit from interest payments, and that they may also pay some property taxes. Real housing user costs in any period may now be expressed as follows:
RHUC = P0αi(1-t)/(1+π) +P0(1-α)mr /(1+π) + P0 m/(1+π)+P0lt/(1+π) –[(P1)/1+π) - P0] (2)
α = percentage of owner’s equity,
i = the nominal pre-tax return available on other uses of the homeowner’s equity,
t = the marginal income tax rate of the home owner,
mr = the mortgage rate,
lt = land taxes and other property taxes per annum as a percentage of house price,
and other terms are as before.
Again, equation (2) shows housing user costs in terms of prices prevailing at the start of the period. These housing user costs include the costs of the owner’s own equity, recurrent expenses, and changes in asset value. The equation is usually estimated initially in nominal terms because tax is based on nominal values. Allowance can be made for the rate of inflation over a year (or any relevant period).
Note that the real cost of housing does not include loan repayments. It is true that a household that makes a loan repayment foregoes other goods in the short run. However, a loan repayment enables a household to reduce mortgage debt and to consume more goods in the following periods. A loan repayment reconfigures a household’s balance sheet. It does not change the net worth of a household.
To illustrate real housing user costs, suppose that P0 = $400,000; a = 0.3, i = 6.0%, t = 0.3, mr = 8.0%, m = 0.015 and lt = 0.005, P1 = 416,000 and π = 0.03. Suppose also that all costs are incurred at the end of the year. Then,
RHUC = [(400,000 × 0.3 × 0.06 × 0.7) + (400,000 × 0.7 × 0.08) + (400,000 × 0.02) + (400,000 × 0.01)] / 1.03 – (416,000 / 1.03 – 400,000)
So RHUC = [5040 + 22400 + 6000 + 2000] / 1.03 - 3884 = $28,584
In this example real housing user costs equal 7.15 per cent of the house price at the start of the year.
Clearly, the assumptions drive the results. However, generally the higher the homeowner’s own equity, the lower the real housing user cost because the after-tax return on own equity is lower than the pre-tax mortgage rate.
Of course, to determine housing affordability it would be necessary to identify the costs for particular types of housing that may be deemed suitable for particular groups of households and to compare these costs with household incomes. This is discussed briefly below.
Rental measures of housing affordability
For renters, housing rents provide a ready measure of housing costs, assuming that maintenance costs and taxes are paid by the landlord. However, for any given housing type, gross rents tend to exceed homeowner housing user costs. This occurs because investor landlords will aim to equate the after–tax return on rental property with the after-tax return on other investments including their own property but they pay taxes on net rental income and on realized capital gains in rental properties. Thus landlords need to obtain a higher gross return from rental property than from their own housing.
On the other hand, imputed rents for home owners should equate to their housing user costs. If imputed rents were higher (lower) than housing user costs, homeowners would increase (reduce) their consumption of housing. Notwithstanding the differences between actual and imputed housing rents, imputed rents may be inferred from market rents for similar housing after subtracting expenses that would be incurred by a landlord in providing that income and after making various assumptions about gearing and tax rates.
Concluding observations on housing affordability
Indices based on real housing user costs or rentals provide more credible information on affordability than do indices based on housing acquisition costs. However, estimates of housing affordability require value judgments as well as estimates of housing costs.
Determination of housing stress involves four main tasks.
(i) Definition of basic acceptable housing standards for Australian households of various kinds in different locations. A critical issue here is the trade-off between housing quality and proximity to the centre of cities. Housing may be affordable in large parts of Australia but not in the centre of large cities, since inner city locations are a form of luxury good.
(ii) Estimates of the real housing user costs or rents for each type of basic accommodation in different locations.
(iii) Estimates of the equivalised disposable household income that could afford these housing costs or rents and leave each kind of household with adequate income for other goods.
(iv) Estimates of the numbers of households in each region that can not afford adequate housing as defined in (i). This would doubtless include many households who spend less than 30 per cent of their income on housing.
Tasks (i) and (iii) involve critical value judgments. Also, the four tasks collectively would involve a significant amount of work. However, this approach is practical and would be a significant improvement on the rhetoric in much of the present discussion of housing affordability.
Clearly, to develop housing policies estimates of the low cost housing needed in the owner and rental markets by location are required. However if it is accepted, following the analyses in Marks and Sedgwick (2008) and Yates (2008), that there is an affordability problem especially for low income renter households, what policies are required?
To answer this question, the causes of the problem need to be identified. Is housing unaffordable because of market or government failures or because of low household incomes? If there are no market or government failures in the housing (or land) market, the market will produce the housing that households want at least cost. It follows that the major issue is low household income rather than inefficient housing supply.
The view of this paper is that the Australian housing market is generally competitive and efficient. There are many firms who can produce the housing that households want and there is a great deal of information about housing. There are some housing market failures due to externalities, which justify land use planning policies and in some cases slum area improvements. However these externalities are not a central issue in affordable housing.
It follows that housing affordability issues arise principally because of lack of income and to some extent because of complex housing needs. Low incomes arise from low productivity and complex needs arise most often from disability. It is a misunderstanding of the meaning of market failures to attribute the low incomes or complex needs to housing market failures. The theory of efficient markets says nothing about the distribution of income. The efficiency and equity of markets are separate issues.
However housing costs are almost certainly increased by government restrictions on housing supply. State governments around Australia restrict the supply of land for housing and urban densities (by restricting the amount of capital that can be applied to create housing on a given lot of land). Evidence of these restrictions is found in the high value of land with development rights compared with land without these rights. Also, fiscal distortions in favour of homeowners make rental properties more expensive.
There are many possible ways to reduce housing costs. Some are macroeconomic policies such as reducing interest rates generally (although this may increase house prices) or restricting immigration. We review below nine policies targeted more directly at reducing housing costs. We start with the two policies that follow most directly from the analysis above, discuss five other possible policies, and conclude with a discussion of recent Australian Government initiatives. These nine policies are:
1. Increasing the supply of housing via less restrictive land use planning regulations.
2. Subsidising low-income households in owned or rented housing.
3. Subsidies for land development and urban infrastructure.
4. Providing more subsidized public housing.
5. Providing real interest-only loans for housing.
6. Reducing housing demand and rents by reducing subsidies for home owners.
7. Modifying or abandoning negative gearing provisions.
8. The Australian Government’s Housing Affordability Fund.
9. The Australian Government’s National Rental Affordability Scheme.
Increasing land supply for housing by less restrictive land use planning regulations
Several studies of house prices have found that real house prices are responsive to the quantity of the housing stock (see Girourd et al., 2006). Abelson et al. (2005), one of the cited papers, found substantial responsiveness: an elasticity of -3.6 for Australia as a whole. This indicates that a 1% increase in the total housing stock reduces real house prices by 3.6%. Most other findings reported in Girourd et al (ibid.) are of a similar magnitude or a little lower.
Further, there is little doubt that state government restrictions on both land supply for housing and the application of capital to land are limiting the supply of housing. The Productivity Commission (2004) found that housing land values on the urban fringes of Australian cities significantly exceeded the opportunity cost of non-residential land. In the United States, Glaeser and Gyourko (2003) found that restrictions on building in established areas are major causes of house price inflation in U.S. cities. In Sydney, three-quarters of new dwellings including units are in established areas and the demand for housing is strongest in these areas. Thus, subject to environmental constraints, to reduce house prices governments should allow higher density housing (a greater application of capital to land) in established areas.
Subsidising low-income (equivalised) households
Given this paper’s position that housing affordability is principally a household income problem, albeit made worse by government failures over land supply, subsiding housing payments for low income homeowners or renters is an appropriate policy. Subsidies should be provided directly to households in greatest need and the households can in turn choose the form of housing that best suits their needs.
Of course, to be fully effective the housing subsidies would bring about an increase in housing. If the housing supply is inelastic, subsidies will simply increase aggregate demand for housing and hence house prices. However, the subsidies will still benefit the low income households that most need housing assistance, albeit at the expense of households that would pay more for housing. Moreover, as has been stressed, in many parts of Australia government has the power to increase housing and builders will respond by increasing housing,
Subsidies for land development and urban infrastructure
Another policy that is popular with the housing industry is subsidies for land development and/or subsidized public infrastructure. However, it is highly questionable whether this will reduce house prices or housing costs.
New house prices are determined essentially by (i) the supply of houses and (ii) by the services provided by new houses relative to the services provided by the existing housing stock. On the other hand, in a competitive housing market, as in Australia, the price of serviced land is the price of a new house less the cost of building and selling. This serviced land price does not determine the price of housing. Unless housing supply increases, subsidizing land development costs (by reducing developer and infrastructure charges) increases the price of unserviced land; it does not reduce the price of new housing (Abelson and Joyeux, 2007).
Increasing public housing
It is sometimes argued that long waiting lists for public housing indicate excess demand for public housing and are a result of market failure. It follows that government is expected to respond by providing more public housing. Three main reasons given in a recent consultant’s report4 are:
● Externalities: an adequate level of housing has a range of external benefits and contributes to a healthier and more productive workforce.
● Complex needs: the market does not supply adequate housing to households in crisis, vulnerable groups or high risk groups.
● Shortages of affordable housing in the rental market.
Only the first point relates to market failure. The other two points are based on a misunderstanding of the meaning of market failure. They are not market failures as described in mainstream economic texts.
Moreover, the view that there is excess demand for public housing is fundamentally flawed. In New South Wales (and doubtless elsewhere) there is a queue for public housing because public housing contains an ongoing subsidy of about $5500 a year per tenancy compared with a private rental subsidy of under $2000 a year which is also conditional on continued low income. Thus, anyone who is eligible for a housing subsidy is a potential consumer of public housing. In such circumstances it is not meaningful to talk about a demand for public housing. If households were given a rental subsidy of $5500 per annum for housing they might well choose to take this instead of public housing.
The basic policy issue is what level of housing subsidy should be provided to households who find housing “unaffordable”? This depends on political judgments of needs, not on the spurious use of the concept of the demand for public housing.
In any case, if the housing market is efficient, there is no case for a separate public housing policy except where private firms fail to provide any suitable housing. This may indeed occur for some households with complex needs or disabilities who find it difficult to maintain independent lives and housing.
Providing real interest only loans for housing
Households can face up-front liquidity costs when they are required to make mortgage repayments and pay interest in nominal interest terms. Nominal housing costs could be significantly reduced in the short run if households were required to pay only real interest rates on their mortgages. These real interest rates would increase in absolute terms each year as the nominal house value increased (as there would be no loan repayments). In the absence of mortgage repayments, mortgages would rise in nominal terms and be a constant proportion of house prices. Such a policy would raise housing affordability as currently defined by several authorities.
However, if markets are unwilling to make such loans, it would be hard to argue that governments should step in and make them. The administrative costs would be considerable. There could be a lack of transparency on the one hand and increased public risk if property prices fell on the other. Subsidising low income households directly and transparently, including possibly via interest rate subsidies, seems preferable.
Reducing demand (subsidies) for owner-occupied housing
Australian homeowners receive significant net subsidies, including un-taxed imputed rent and capital gains. Abelson and Joyeux (2007) estimate that these subsidies are equivalent to 8 per cent of imputed housing rentals and increase house prices overall by an estimated 2 per cent. However, the effect on house prices is small because the subsidy does not apply to rental housing and it is assumed that housing supply responds to the increase in prices and so constrains the price increases. This implies in turn that the impact of these subsidies on rents is not very large.
Of course, if home owner subsidies were reduced, real homeowner user costs would rise. Thus home owning would become less affordable. Overall, reducing owner-occupied subsidies would be an indirect and not very effective way to improve housing affordability for renters.
Disallowing tax benefits from negative gearing policies
Another similar policy that is discussed from time to time is disallowing (or quarantining) tax deductions for negative gearing. This means that investors would not be allowed to claim tax deductions for interest payments when a loss is incurred on a rental property investment. If the deductions are quarantined, investors can claim a tax deduction for interest payments only after there is a cumulative net surplus (this is the UK system). Currently, about a half of all Australian residential property investors claim tax deductions for losses on residential properties. On the other hand, Australian residential property investors are required to pay tax on only 50 per cent of realized nominal capital gains, so there is an asymmetry in the tax arrangements that favour residential investors.
However, Abelson and Joyeux (2007) show that the net benefits of the Australian tax system to residential investors are small and certainly much smaller than the tax benefits provided to owner-occupiers. The authors concluded that any significant change to the negative gearing provisions would reduce the supply of rental housing and increase housing costs for renters, but have little impact on prices in the larger owner-occupied market.
Housing Affordability Fund
Under the Housing Affordability Fund (HAF), the Australian Government aims to reduce the cost of home ownership for up to 50,000 low to moderate income new home buyers by lowering infrastructure and regulatory costs and increasing the supply of housing. The Fund will target greenfield and infill developments where high dwelling demand exists or is forecast for the next five years. Applicants for the funds, likely to be mainly state and local governments, must show how the cost savings to developers (due mainly to reduced charges) will be passed to new home buyers. Priority will be given to entry-level or moderately priced housing.5
However, the policies for the HAF are not clearly defined. For example, it is not clear whether new home buyers must be first time purchasers or what a moderate income is. It is not clear how high dwelling demand will be determined. Many first home buyers may prefer inner-city locations that are generally better served with infrastructure (and generally will be because of economies of scale) rather than locations with lower density living. Nor is it clear how applicants for HAF funding will be able to show that the price of housing produced in 2 to 5 years time will be lower by the amount of the subsidy (say $10,000 per house) than it would be without the subsidy. Moreover, if someone does obtain a house at $10,000 below market price, what is to prevent them from re-selling the house a few years later at the market price?
The reality is that subsidizing developers generally has no underlying impact on house prices. For any given supply of housing, house prices rise with the quality of associated infrastructure. On the other hand, so long as the price of land for housing exceeds the value of land in alternative uses, reducing infrastructure charges will increase the price of unserviced land but not reduce house prices. Only increasing housing supply (or producing inferior housing) will reduce house prices.
The Australian Government also proposes to provide funds to local governments that reform planning processes to reduce the cost of a new home. This will include funding for a national electronic development assessments and on-line tracking system. While a more efficient development approval (DA) system is certainly desirable, the key issue for low cost housing is whether this will increase the supply of housing as well as reduce the cost of the DA process. Any efficiency dividend will reduce house prices only if housing supply is increased and it is not clear how the Government will validate how planning improvements affect housing supply.
In conclusion, it is important to improve supply side efficiencies, to expedite planning and DA processes, and to have efficient (non-distorting) charges for infrastructure. However, to assist low income first home buyers, subsidies to them would facilitate home purchase, enable them to exercise their preferences over housing, introduce fewer distortions into the housing market, and be simpler and fairer to administer than the HAF.
National Rental Affordability Scheme
The Australian Government also proposes to increase the supply of affordable rental properties by 50,000 over five years.6 The Government proposes to subsidise institutional investors who provide new housing at 20 per cent or more below the market rental for 10 years to tenants receiving the maximum rate of Commonwealth Rent Assistance and paying at least 30 per cent of gross income in rent.
The scheme appears generous. The Commonwealth has pledged an annual subsidy of $6000 per dwelling and expects the States and Territories to provide a further $2000 in direct assistance. This total subsidy of $8000 per dwelling considerably exceeds the current subsidy of about $5500 per dwelling unit of public housing in Sydney and just under $2000 per household for a private rental subsidy (Abelson and Joyeux, 2007). Comparatively few renters will benefit and this appears to represent an inefficient use of public funds.
There are several other issues. It is not clear how “institutional investors” will be defined and chosen, whether ownership of housing assets can be transferred, or how this arrangement can be monitored. There will also need to be a mechanism for establishing market rents. It is not clear what will happen if a tenant’s incomes increase and the tenant(s) become ineligible for the maximum rate of Commonwealth Rent Assistance. Sub-letting of housing is an efficient response to housing problems. However, sub-letting will presumably be banned. This in turn creates a monitoring problem.
Nor is it clear why tenants should be supported only if they pay at least 30% of gross income in rent. Hardship is generally defined by disposable equivalised household income that allows for different kinds of household. It would be unfair to penalise a household with a given equivalised income that chooses to spend less that 30% of income on housing.
The rental housing proposal does not address housing efficiency. Rather it simply introduces a supply side subsidy that appears to be expensive. A demand-side subsidy would directly benefit the recipients in greatest need, be a more transparent subsidy, and almost certainly be more cost-effective and simpler to administer than the Commonwealth proposal.
Most current Australian measures of housing affordability are based on the acquisition costs of housing rather than on real housing user costs or rents, which are more reliable measures of housing costs. The acquisition cost indices define costs in nominal rather than in real terms and include mortgage repayments although these are savings rather than expenses.
Also, the affordability measures do not allow for household choice. Households choose the type of house, its size and location and make choices over household size and composition and thus over household income. Households who choose to spend a relatively high proportion of their income on housing are not necessarily suffering stress. Such definitions of stress are arbitrary. Some households spend a high proportion of their income on foreign holidays. But we would not describe these households as suffering from foreign holiday stress.
Housing affordability is essentially a household income problem made worse by government restrictions on the supply of housing. The housing market has few market failure features. This is not simply an academic debate about the nature of market failures. It has fundamental practical policy implications.
In summary, the paper argues that government should
(i) Base a housing affordability strategy on sound and measurable concepts of housing user costs or rentals and of the numbers and types of households who should receive housing support and how much, especially though not only renters.
(ii) Promote efficient housing land supply in established and greenfield areas. This should not require or involve subsidies to land development.
(iii) Provide housing income support or rental subsidies to households requiring support albeit that in many cases this would be a de facto income support grant.
(iv) Provide public housing in limited circumstances for households that cannot cope with independent living.
On the other hand, the Australian Government’s housing and rental affordability funding programs appear weakly defined and likely to be costly and ineffective.
Abelson, P., 2008, Public Economics, Principles and Practice, McGraw-Hill, Sydney
Abelson., P. and R. Joyeux, 2007, ‘Price and efficiency effects of taxes and subsidies for Australian housing’, Economic Papers, 26, 147-69.
Abelson, P., Joyeux, R., Milunovich, G. and D.Chung, 2005, ‘Explaining House Prices in Australia: 1970 to 2003’, Economic Record, 81, pp.S1-S8.
Australian Bureau of Statistics, 2007, Housing occupancy and Costs, Cat. No. 4130.055.001, ABS, Canberra.
Australian Housing, Local Government and Planning Ministers, 2006, (to be provided).
Australian Council of Social Service and others, 2007, National Affordable Housing Summit.
Girouard, N., Kennedy, M., van den Noord, P., and C. Andre, 2006, Recent House Price Developments: The Role of Fundamentals, Economics Department Working Paper, No. 475. OECD, Paris.
Glaeser, E.L. and J. Gyourko 2003, ‘The impact of building restrictions on housing affordability’, Economic Policy Review, June, 21-39.
Marks, G.N. and S. T. Sedgwick, 2008, ‘Is there a Housing Crisis? The Incidence and Persistence of Housing Stress 2001-2006’, Australian Economic Review, 41, 2, 215-12.
Productivity Commission, 2004, First Home Ownership, Report No. 28, Melbourne.
Yates, J., 2008, ‘Australia’s Affordability Crisis’, Australian Economic Review, 41, 2, 200-14.
1 It is not clear how Marks and Sedgwick (2008) or ACOSS (2007) define housing costs.
2 Arguably, the cost of housing should also include costs of commuting to work as this increases as house costs fall with distance from employment centres.
3 Of course, with renovations buildings may increase in value, but this is not relevant to the considerations in this paper.
4 This report cannot be formally cited.
5 For a description of the Housing Affordability Fund, see Australian Department of Families, Housing, Community Services and Indigenous Affairs, 2008, Housing Affordability Fund, Consultation Paper and Housing Affordability Fund: The Guidelines (see www.fahcsia.gov.au).
6 Australian Department of Families, Housing, Community Services and Indigenous Affairs, 2008, National Rental Affordability Scheme (www.fahscia.gov.au).