“There is an ever-growing divergence between the price of Australian properties and their underlying costs of construction. Importantly, this does not appear to be a manifestation of natural constraints on the supply of land, but rather a product of regulatory restrictions that artificially inflate the price of housing. Viewed differently, these limits on dwelling dispersion and the release of greenfield sites act as a burdensome tax on new building, which in turn leads to a mismatch between the accommodation needs of Australian households and the stock of available properties… The overall objective here [must be] to accelerate the approval and land release process so as to promote private sector investment in the production of affordable housing.”Joye et al, 2003 Report to the Prime Minister’s Home Ownership Task Force.
The Federal Housing Minister, Tanya Plibersek, has announced over the weekend that the FHOG will be capped from 1 January 2010 at a level determined by the States. According to the SMH, this means that first time buyers will face caps of $1m in Queensland, $750k in NSW, and $600k in Victoria.
I guess I have a few gut reactions to this. First, the proposed caps will have little-to-no impact on the vast bulk of all first timers. Less than 8% of all property sales in Australia are for homes worth more than $800k (see figure below). A tiny 4.4% of all sales are accounted for by homes worth more than $1m. Of course, first timers are heavily skewed towards homes at or below the median price nationally of around $400k (in all metro and non-metro areas).
Note: Click to enlargeSecond, as I argued in The Australian earlier this year, a far better way to deal with the FHOG, which was meant to be a one-off subsidy to compensate for the introduction of the GST, would be to convert it from the current cash handout into an interest-free loan. Now who can honestly complain about getting an uber-long-term, interest-free loan from the government. Importantly from a policy perspective, this would have the profoundly beneficial attribute of transforming the existing grant from a $1.5-$2 billion expense sliced directly off the Budget's bottom-line into an asset on the government’s balance-sheet. It should, therefore, generate massive annual savings for the Commonwealth, which is going to become an increasingly important issue as time passes.
To the extent the pollies were concerned about political sensitivities, they could actually increase the existing grant from the baseline $7k figure today to, say, $10k given the savings that they would capture via the shift from a cash expense item to a loan.
Practically speaking, the loan would rank behind any traditional mortgage debt held by the borrower and would only be due for repayment on sale of the property. If the Commonwealth was somewhat more ambitious, they could easily link the loan to the value of the property. Based on some rough numbers, this would make the grant self-funding within about seven years. In The Australian, I summarised the proposal as follows:
"Most policymakers...agree that there is little economic sense in maintaining the $1 billion a year first-home owner's grant, which has long outlived its original purpose of compensating buyers for the GST. And why should taxpayers give away $7000 to non-means or asset-tested buyers with above-average incomes who can afford to purchase a home of any value?My next observation is that policymakers are missing the point every time they try and employ blunt demand-side measures, such as higher interest rates, new taxes (like those that will be inevitably canvassed by the Henry Review), or reducing the FHOG, in a superficial attempt to influence the cost of housing.
Yet once a subsidy is put in place, it is hard to remove
One solution is to convert the grant into an interest-free loan. Rather than taxpayers making a gift to every new owner, the subsidy could be adjusted into a loan that would be repaid when the property is sold. This is similar to the way the Higher Education Contribution Scheme works for higher education. A better solution would be to index the loan to the lesser of inflation and the home's capital gains, ensuring taxpayers avoid real losses."
Why? Because as I first outlined in my 2003 report to the Prime Minister’s Home Ownership Task Force, the most powerful long-term affordability solutions are to be found on the supply-side of the housing market equation. While this is accepted wisdom today, it was an apparent innovation at the time, and was bizarrely rejected out-of-hand by key technocrats such as the RBA (which media commentators forget has only very, very recently begun to publicly acknowledge Australia’s supply-side problems). For a long-time the RBA was beholden to the idea that the supply-side was largely a non-meaningful participant in the affordability debate. This is perhaps understandable vanity given its one policymaking tool is interest rates. As I have noted here before, in 2004 the RBA claimed in its submission to the Productivity Commission inquiry (which was established to investigate the findings of our 2003 report) that:
“At the macro level there is not much evidence to suggest that the growth in house prices has been due to a persistent shortage of supply of houses relative to underlying demand for new housing.”Prior to the 2003 Prime Minister’s Home Ownership Task Force, virtually all explanations for the increase in the real cost of housing in Australia during the late 1990s and early 2000s had been limited to a demand-side lens: eg, the significant reduction in real interest rates; the expansion in the availability and flexibility of housing finance due the to the rise of non-bank lenders and the new securitisation markets that funded them; the seemingly secular decline in unemployment combined with the period of sustained economic growth that came to be known as the Great Moderation, and so on.
In fact, I recall doing a media search at the time trying to identify public references to “housing supply” while writing up the supply-side chapter with the help of Professor Edward Glaeser at Harvard (starts page 267) and could find almost none. It just did not feature on the policy radar screen with the complaints of the building industry being consigned as a rent-seeking interest group.It took a further conference call with the Prime Minister’s chief of staff, Arthur Sinodinos, and his economic advisor, Peter Crone, to explain to them the ramifications of these insights—not only was the cost of housing being heavily influenced by the supply-side, but this was a State not Federal Government matter. You don’t need to be a member of Mensa to work out why they then established the Productivity Commission Inquiry a month after I delivered my final report with terms of reference that concentrated mostly on supply-side considerations.
Today most policymakers are in violent agreement that we need to urgently work to stimulate new private-sector investment in housing supply given the burgeoning demand-side requirements.
If one accepts this very credible objective, surely the worst thing in the world one could then do is to place direct downward pressure on housing demand and the expected returns to those that we are desperately trying to convince to reallocate capital to the production of new supply? All other things being equal, higher interest rates, higher taxes, and reduced subsidies will only further diminish the incentives builders have to invest in new stock. While on the one hand the Rudd Government agrees with the logic—as demonstrated through the $500 million plus they are committing to boost developer returns via the NRAS program—policymakers run the risk of undermining the collective focus on stimulating new supply with distractions such as the raft of tax changes being countenanced by the Henry Review, and the unhelpful and erroneous media claims that the RBA will crush house price inflation (even though the RBA has recently emerged as the most vocal advocate of the need to elastify the supply-side).
While creating significant uncertainty for builders and developers as to the future returns that they can expect from their capital investments, these dynamics are also presumably spooking highly risk-averse lenders that have, to date, been criticised for rationing credit to the sector.
A salient point here is that this policy confusion appears to at least in part derive from the mistake people frequently make when they mix up questions around "ability to pay" with the "cost of housing". Ability to pay, which is often loosely referred to as "affordability", is determined by a range of factors that have nothing to do with the costs of producing new housing, including interest rates and household disposable income growth. In my 2003 report to the PM, I noted:
“A great deal of confusion tends to reign in the emotive affordability debate. Above all, combatants make the mistake of attempting to judge the costs of home ownership in relation to the income levels of prospective acquirors…Good public policy does not…obfuscate issues that cause high house prices with those that contribute to depressed incomes. As such, we concentrate our efforts on schemes that improve the affordability of housing by relaxing regulatory restrictions that have the potential to propagate price rises. We do not recommend responding to a housing ‘crisis’ by lavishing low-income occupiers with taxpayer-funded subsidies. To reiterate, the two problems are distinct.If we accept that interest rates are largely the domain of the RBA, and policies surrounding the distribution of income having nothing directly to do with the housing market, the most legitimate question for discussion is whether there are any factors that have artificially raised the cost of new housing that policymakers can seek to mitigate. In 2003 I posited three candidates:
One of the fundamental implications of modern economics is that in an open, competitive, and unregulated market, the price of a commodity should not be greater than the marginal cost of producing it. If such an inequality did emerge, suppliers would have strong incentives to manufacture more of the goods in question. Sooner or later, competition amongst agents would ensure that prices converged with their marginal costs. This logic holds as strongly for Australian houses as it does for meat pies”
“First, in cities such as Sydney, there may be limitations on the availability of land, which can place considerable pressure on prices. Second, there could be an upward shift in the underlying costs of production, with much the same effect. (Both these factors feature prominently in the classical urban economics literature.) Third, it is conceivable that in spite of sufficient land, housing is expensive because of regulatory constraints on development created by sluggish release programs, zoning, antediluvian approval processes and other government imposed transaction costs. Adopting this argument, the advent of new construction could reduce the cost of housing towards its production baseline in high priced areas.”For the wonks out there, I have enclosed below our findings, which are even more relevant today than they were five years ago:
“When thinking about the cost of supplying new housing, economists like to identify two broad components: the physical construction charges and everything else. Historically, building-related expenses (bricks and mortar, wood etc) have accounted for the lion’s share of supply costs in Australia and the US. To get a better feel for this dynamic, we examine the time path of dwelling and building material prices, where the established (project) house price index includes (excludes) the cost of land (see Figures 70 through 76).
Prior to the asset price inflation of the late 1980s, all three lines hugged one another quite closely. Since that point, there has been a striking wedge between the price of established homes and the cost of the inputs used to build them in every city except Hobart. Disturbingly, this disjunction has become increasingly large over the past one and a half decades, with unusually rapid growth during the last five years. As one would expect, the price trajectory of project homes (which does not include the value of the land on which they sit) has mirrored that of building material prices throughout most of the sample period. Nevertheless, there has been a noticeable breach in the past few years. This is explained in part by the introduction of the goods and services tax in July 2000, which precipitated a sudden jump in the price path at that time. It is also possible that manufacturers have capitalised on Australia’s prosperous economic conditions by expanding margins somewhat. But this trend should abate once conditions moderate, as was the case in the early 1990s…Note: Click to enlarge
The ever-widening gap between the prices of established properties and their costs of supply could be seen as both a warning and an opportunity. It is a warning in the sense that unless something is done soon, real estate in Australia will become progressively less affordable for low- to middle-income households. To make matters worse, this demographic is responsible for contributing the bulk of ‘essential workers’ to the domestic economy. In cities such as London, New York and San Francisco, these indispensable members of the community have been slowly priced out of the market, undermining crucial services like education, health and policing (a similar effect is said to be emerging in Sydney).
On the other hand, the housing needs-production mismatch may be viewed as an opportunity. If the price inflation manifest throughout the 1990s simply reflected rising costs of construction (e.g., bricks, labour and wood), then our ability to temper it would be rather limited. Put differently, how can one expect to make housing more affordable when the principal inputs have become increasingly expensive?
Thankfully, the high cost of home ownership in Australia has little to do with swelling construction prices, as the figures above clearly demonstrate. No, this phenomenon is an artefact of something else, which might be loosely referred to as the ‘extrinsic’ cost of land. Here it is useful to distinguish between market-based valuations that recognize control rights, and intrinsic measures of worth that make no attempt to incorporate such. Ultimately, a property’s costs of production will be determined by three factors: the physical characteristics of the dwelling structure, the innate value of the turf on which it was built, and land use regulations that interfere with the market’s estimate of the latter. These distortions may take the form of specific rights that attach to the lot in question (i.e., zoning), or holistic supply-side strategies that dictate the release of greenfield and brownfield sites.235 As we shall see, the soaring cost of owneroccupied housing in this country has much more to do with government restrictions of this type than, say, a shortage of exploitable land…
In light of the issues above, it makes sense to embark on a more thorough decomposition of the housing needs-production mismatch. We saw earlier that there was a growing discrepancy between the price of established properties and the costs associated with their construction process. One might reasonably infer that this disjunction is explained by growth in the extrinsic value of land. We undertake two experiments to assess the veracity of this conjecture. In the first, we quantify the real differential between new house prices and the value of approved private sector dwellings over time. This facilitates a more accurate comparison of the price of a property with its developer-estimated costs of production (which include all margins, taxes and related charges), where the disparity between the two should reflect the market value of land. The ABS figures also strip out the cost of alterations and additions, which in view of the historical shift towards larger homes, could be used to rationalise the time-series changes in the cost-price gap that were previously discerned (see Appendix 8.11). The figure below presents the results.Note: Click to enlarge
Although there will inevitably be some degree of lumpiness in the series because of the nonsynchronicity between the reported number of approvals and their corresponding value, the overall trend is quite clear. In June 1985, the land component of the median Australian dwelling was valued at $30,058. In constant dollar terms, today’s equivalent figure is three times higher at $103,306—a phenomenal increase in anyone’s books. Importantly, this growth has been strong right across the board. In Sydney (Melbourne), land prices have risen by 360 percent (418 percent) over the last one and a half decades. Maybe there is something to be said for decoupling ownership rights to the physical dwelling structures and the lots on which they are built; but that story will have to wait for another day! Table 22 provides a nominal dissection of the data and plainly shows that a considerable proportion of the housing costs in this country can be ascribed to the extrinsic value of land. In Sydney, 66.5 percent of the median dwelling price is attributable to this factor. At the other end of the spectrum are smaller, and presumably less constrained, conurbations like Perth, where only 38.0 percent of the value of residential real estate is accounted for by terra firma.Note: Click to enlarge
In a second test of the housing needs-production mismatch, we draw on a higher resolution set of cost data. This was kindly supplied by the HIA and consists of the value of building approvals over time combined with the total number of dwelling units and their estimated floor size area. The figure below illustrates the time-path of construction prices per square metre for five of the major capital cities. In accordance with the trends documented earlier, the basic costs of erecting homes in this country have not increased much during the past decade or so. Indeed, production prices in Sydney have almost declined when measured in real terms. Based on this data alone, it does not seem unreasonable to suggest, as several developers do in Chapter 4.4, that the domestic housing industry has done a good job of controlling its expenses.Note: Click to enlarge
Now the question is what do these new numbers tell us about the value of land? Applying roughly the same technique that was used before, we contrast changes in median house prices over time with the costs of production per square metre multiplied by the average floor size area. This furnishes us with a comparable set of results against which one can benchmark the earlier findings (see Figure 79 below). Although the method is somewhat different, the time-series dynamics are strikingly similar. In Sydney (Melbourne), there has been a 272 percent (475 percent) increase in land prices since 1988. Once again, minor metropolises such as Adelaide and Perth register even more robust growth.
So our basic conclusion is straightforward: the high cost of home ownership in Australia appears to be a function of growth in the extrinsic value of land. Furthermore, this is a disease that is rapidly spreading throughout our largest urban centres. And unless radical action is taken, there seems to be no respite in sight."
