The Australian housing supply myth

ABSTRACT

Australia's expensive housing market is often claimed to be primarily the result of a shortage of supply due to town planning constraints, leading to political pressure on councils and state governments to remove planning regulations, regardless of their planning merit. We argue that this supply story is a myth and provide evidence against three key elements of the myth. First, there has been a surplus of dwellings constructed compared to population demand, rather than a shortage. Second, planning approvals typically far exceed dwelling construction, implying that more approvals or changes to planning controls on the density and location of development cannot accelerate the rate of new housing supply. Third, large increases in the rate of housing supply would have small price effects relative to other factors, like interest rates, and come with the opportunity cost of forgone alternative economic activities. Indeed, if the story were true, then property developers would be foolishly lobbying for policy changes that reduce the price of their product and the value of the balance sheets, which mostly comprise undeveloped land.

KEYWORDS:

• housing supply
• home prices
• planning
• housing policy

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 Yglesias (2012) and Glaeser and Gyourko (2018) have been particularly influential at popularising this view in the United States, while in the United Kingdom many think-tanks and government agencies continue to argue this position, such as DCLG (2017), Philp (2017), and Wilson and Barton (2018), amongst many more summarised in Mulheirn's (2019) critique of the supply story.

2 The crisis had been many years in the making. In 2014, the 21-storey Lacrosse apartment tower in Melbourne's Docklands district suffered a major fire due to numerous design defects, including the use of flammable aluminium composite cladding (Hanmer 2019). Audits of the use of flammable cladding in New South Wales and Victoria have found that hundreds of buildings have potentially non-compliant flammable cladding (Victorian Cladding Taskforce 2019). High profile cases of poor construction quality at Opal Tower and Mascot Towers in Sydney have resulted in buildings remaining unoccupied indefinitely (Casben 2019). A review of 212 residential building defect reports found that ‘85% of all the buildings had at least one defect across multiple locations’, with 14 line-item defects found per building on average (Johnston and Reid 2019, 21). Most Australian States have set up inquiry processes to examine the extent of the building quality crisis (ACT Legislative Assembly 2018; NSW Parliament 2019).

3 The method used to construct this ratio is to divide the change in private housing stock by housing turnover from Table 6 of ABS (2018b). Shaded periods are when smoothed annual price growth exceeded 2% for at least four consecutive quarters, which also applies to subsequent Figures.

4 This is higher than the 20% home price decline because land makes up 65% of the home values (Kendall and Tulip 2018), amplifying the effect of home price reductions on the land values. A 20% home price reduction where land makes up 65% of the price is a 31% land price reduction. In lower value areas, the amplification effect will be more substantial. Land values in Brisbane are around half dwelling values, meaning a 20% home price decline would result in a 40% land value decline.

5 As compiled in Table 2 of Gurran and Phibbs (2013), with updates from 2012 to include 1. AHURI's Housing Supply Responsiveness report in 2017, 2. Housing and Employment Land Supply Program – Greater Adelaide 2012, 3. NHSC Housing Supply and Affordability Issues 2012-13, and 4. Out of reach? The Australian housing affordability challenge 2015.

6 Sydney housing stock data is only available for the most recent census year of 2016.

7 This contrasts with OECD data that shows a decline in dwellings per 1000 persons aged over 21 from 411 in 2010 to 401 in 2015, which has been relied upon to claim a shortage of supply is leading to large price effects (Daley et al. 2019, 67). ABS data confirms this pattern, with 547 dwellings per 1,000 adults over age 21 in 2001 and 536 in 2015, a 2% decrease. However, it also shows that in 2018 that this measure was 541 dwellings compared to 539 dwellings in 1996, a slight increase.

8 Actual demand is the increase in households from December 2007 to December 2017, the closest dates at which reliable estimates are available, and actual supply is the increase in private dwellings over the same ten year period (ABS 2018a).

9 Prices in the previous ten year period to December 2008 increased 93%.

10 Counting building construction and apportioning half of the building services employment to housing and its related construction needs.

11 One could argue that there is a macroeconomic stimulus effect from higher construction that would reduce unemployment. However, again, there is a cost of forgone alternative non-housing stimulatory investments that could have instead been made.