Note: Where available, the PDF/Word icon below is provided to view the complete and fully formatted document
Some observations on the cost of housing in Australia: address to 2008 Economic and Social Outlook Conference, Melbourne.

Download PDFDownload PDF

Some Observations on the Cost of Housing in Australia1

Anthony Richards

Head of Economic Analysis Department

Address to 2008 Economic and Social Outlook Conference The Melbourne Institute Melbourne - 27 March 2008

In recent years, there has been concern in many sections of the community about the cost of housing in Australia. This is not the first time that such concerns have surfaced: the 2003-04 Productivity Commission Inquiry on First Home Ownership was preceded by studies into housing issues in 1977-78 and 1990-92.2 But the current episode of high housing costs has been quite prolonged, and there are few indications that any amelioration is in prospect in the near term.

In this talk, I will provide an update on some trends in housing prices and affordability, and discuss some of the factors that have influenced these outcomes.

I will begin with a graph of housing prices and some housing-related ‘fundamental’ factors (Graph 1). One clear fact is that in the 35 years since 1972, nationwide house prices have risen significantly faster than average household incomes, house-building construction costs, and average rents. Most of the increase in real house prices occurred in two episodes, in the late 1980s boom and the subsequent boom in the late 1990s and into this decade. Growth in prices has been broad-based across the different states and territories. The run-up in prices is likely to mostly reflect an increase in the price of land.3

The increase in housing prices has been a mixed blessing for Australians. At one level, rising housing prices have made many people feel wealthier and have contributed to higher levels of consumer spending than might otherwise have occurred. But they have also resulted in concerns about housing affordability.

The difference in views reflects the fact that housing is not just an asset but also a consumption item. When housing is thought of purely as a consumption item, it would seem that in aggregate we would be better off if its price were lower. Because we all need to consume some level of

housing services, either rented or purchased, a higher level of housing prices and rents allows less spending on other items.

But housing is also a long-lived asset, and there are distributional aspects to changes in housing prices and rents. Renters will be worse off when housing prices rise whereas those who own rental property will be better off. Owner-occupiers may be largely unaffected, since they can be thought of as being ‘hedged’ against increases in the cost of housing. There are also generational differences. Younger people who have not yet bought homes will be hurt by higher housing prices. Older owner-occupiers may benefit from an increase in prices if they are intending to extract part of the increased value of their homes. Of course, if older people pass on some of their increased wealth to younger relatives, the gains and losses of these two age groups will be reduced. Indeed, the biggest difference may be between those who benefit from transfers from older relatives and those who do not. Both home ownership and ownership of rental property tend to rise with incomes (Graph 2), so it is lower income households that tend to suffer from rising housing prices and higher income households that tend to gain.

But although there are significant distributional effects across the age and income structure, one can make the case that the population in aggregate does not benefit from increases in housing prices.4 5

This discussion of housing as a consumption item leads us to the issue of housing affordability.6

For renters, affordability is typically measured by the ratio of rent paid to household income (Graph 3). Survey data show that the proportion of income being allocated to rent payments has risen over the past two decades for renter households of all income levels.7 In addition, the share of lower income households in ‘housing stress’ has tended to rise.8

In the case of home buyers, concerns about affordability are typically about the accessibility of home ownership, or the ability of younger households to gain access to home ownership for the first time. The standard measures of accessibility show an improvement when average household income is growing faster than housing prices, or when mortgage interest rates are falling so that the borrowing power of households is increasing. Such measures suggest that there are cycles in affordability, but that it was at low levels by historical standards at the end of 2007 (Graph 4).9

But the existing measures of housing accessibility have a few shortcomings. Most importantly, they tend to focus on the average income level for all households rather than focusing on households in the age groups that are typically looking to purchase homes.

Accordingly, we have calculated an alternative measure which represents an estimate of the proportion of all dwellings (both houses and apartments) transacted in any year that would have been accessible to a typical household in the prime home-buying years, based on certain assumptions about bank lending behaviour.10 We focus on households headed by persons aged between 25-39 years as potential home buyers. The estimates suggest that in four of the major capitals, around 30-35 per cent of transacted dwellings (houses and apartments) would have been accessible to the median household in the home-buying age groups in 2006/07 (Graph 5). Perth was the exception, where only around 10 per cent of dwellings would have been accessible.

Taking account of accessibility outside the capital cities, we estimate that on a nationwide average basis around 33 per cent of transacted dwellings would have been accessible to the median young household in 2006/07, compared with a longer-run average of around 45 per cent. Of course accessibility would have been much lower for many lower-income households.11


An alternative way of looking at affordability for younger households is to consider trends in the real income that this group would have had after servicing a mortgage of a given size (Graph 6).12 The data indicate that real residual income of this group would have fallen between the early 1980s and early 1990s, but then would have increased through to 2006/07. For the 25-year period from 1982/83, expenditure on servicing a mortgage would have grown faster than income. But the real residual income available for other goods and services would nevertheless have grown, by around 0.5 per cent per annum. So the increase in housing prices has not in aggregate terms resulted in a fall in real spending on other goods.

I will turn now to the determinants of housing affordability and housing prices.

In any discussion of whether there is an affordability problem, we should remember that the run-up in housing prices in Australia has not occurred in isolation, with many other countries also experiencing housing booms. Admittedly there are problems of comparability when looking across countries, but the data suggest that Australia’s median house price to income ratio is quite high by international standards (Graph 7).

An additional perspective on this issue comes from the fact that the standard accessibility measures are driven by three variables: housing prices, household incomes and mortgage interest rates. So we can consider whether the relatively low levels of affordability in Australia over the past five or six years are more a function of housing prices being high relative to incomes, or to mortgage interest rates being relatively high. The data show that the recent period when affordability measures have been at low levels has been a period when the housing price to income ratio has been well above its average level for the low inflation period (Graph 8). Mortgage rates will of course fluctuate, but they could not be said to have contributed significantly to the persistently low levels of housing affordability over the past five or six years.

So looking either at the standard inputs into accessibility ratios or at an international comparison, it appears that the low level of housing accessibility in Australia can be thought of mostly as a reflection of the persistently high level of average housing prices. Given that houses and

apartments can be either rented or owner-occupied, this high level of housing prices affects both home buyers and renters.

Of course, we must recognise that housing prices are not set exogenously, but reflect the interaction of demand and supply.

Certainly there are many well understood factors (the long economic expansion, the fall in inflation and interest rates, developments in the financial sector, etc) that have contributed to the household sector choosing to spend more money on housing.13 Indeed, the experience of the past couple of decades suggests that, for a significant part of the population, housing may have been something of a ‘superior good’, that is the type of good to which consumers devote an increasing share of their income as incomes rise. To some extent, the recent experience might also suggest that in the earlier era of high interest rates and a regulated financial system, households were unable to spend as much on housing as they might otherwise have chosen.

In addition, a number of demographic and social trends have increased the demand for housing in the economy. And the effect of these ‘fundamental’ factors have probably been added to by increased demand for housing as an asset, due to aspects of the tax system and a broader shift in attitudes about housing.

It should not be surprising that these demand-side factors have boosted housing prices, especially the price of more favourably located housing which is in limited supply.14 Indeed, within our largest cities, house prices have increased more in closer-in suburbs than in more distant ones in recent decades (Graph 9).15 In four of the five major capitals, average annual growth in house prices within five kilometres of city centres has been about 2 percentage points higher than for houses close to the edge of the cities: the exception is Adelaide where the difference is smaller. In addition, waterfront suburbs have had annual price growth around ½-1 percentage point higher than similarly proximate non-waterfront suburbs.16 The greater run-up in closer-in and waterfront suburbs suggests that as the income and borrowing power of households has risen, there has been greater competition for housing that is viewed as more desirable.

So demand factors have played an important role in the run-up in housing prices. However, developments on the supply side should have worked to dampen the impact of demand pressures somewhat. As the value of land rises there is an incentive to increase the intensity of its use, for example by building townhouses on land that was previously used for single family houses or building high-rise apartments on land previously used for small blocks of units.

The demand factors discussed above would be expected to have contributed to some increase in real housing prices even far from the city centres. However, supply-side factors should have a much greater influence on prices towards the fringes of cities, where land is less scarce and accounts for a smaller proportion of the total dwelling price. In principle, the price of housing there should be close to its marginal cost, determined as the sum of the cost of new housing construction, land development costs, and the cost of raw land. And in the absence of any restrictions on supply, the price of raw land on the fringes should be tied reasonably closely to its value in alternative uses, such as agriculture. So unless there has been a marked increase in the value of this land when used for other purposes, the availability of additional land towards the edges of our cities should have limited increases in the cost of housing there.

However, the evidence in Graph 9 provides only limited support for the proposition that real housing prices should not have risen significantly in suburbs far from the CBD. While price growth has been strongest in the most desirable suburbs, the fact is that real price increases in the outer suburbs have been quite large as well. Hence, in 2006/07 median prices of houses in suburbs in the outer parts of the capitals were typically in the $250,000-$300,000 range, except in Perth where prices were slightly higher (Graph 10).

So if we are looking for explanations why housing is not as affordable as we might like, it may be necessary to look at factors on the supply side as well. One obvious place to start is the cost of land for building new houses near the edges of our cities. To shed some light on this, in late 2007 Reserve Bank staff looked at newspaper advertisements and websites for new housing developments in each of the five major capitals, focusing on the least expensive land available in major developments. These were typically lots around 400 square metres in size, ranging from around 25 kilometres (Adelaide) to around 40 kilometres (Sydney, Perth and Melbourne) from the CBD. Entry-level lots ranged from around or a little below $100,000 in Melbourne and Adelaide to around $200,000 in Sydney and Perth (Graph 11). The implied per hectare price of the developed land ranged from a little over $2 million in Melbourne to around $5 million in Sydney and Perth. These were representative prices for low-cost land, with the average cost of building sites noticeably higher.

Copyright | Disclaimer | Judicial Notice | Privacy Statement | Online Action Plan | Accessibility

© Reserve Bank of Australia, 2001-2008. All rights reserved.