How long to turn a commencement into a completion?

By Leith van Onselen

It’s a question that I have mulled for the better part of two years without much success: how long on average does it take to turn a dwelling commencement into a completion? Yesterday, the ABS provided some answers:

This article examines the average completion times (in quarters) for new houses, townhouses and flats, units or apartments from 2006-2016.

Completion times are measured as the period (in quarters) between the commencement and completion of construction for a project creating new dwellings. National data is presented to show changes in the average completion times of new houses, townhouses and flats, units or apartments. Regional data is presented in five year periods to allow for broader comparisons between the states and territories.

The data presented is from the Australian Bureau of Statistics (ABS) quarterly publication Building Activity, Australia (cat. no. 8752.0). ‘New houses’ are defined as detached buildings used for long term residential purposes, consisting of only one dwelling unit and are not a result of alterations or additions to a pre-existing building. ‘Townhouses’ are dwellings with their own private grounds and no separate dwelling above or below. They are either attached in some structural way to one or more dwellings or are separated from neighbouring non-residential buildings by less than 500 millimetres. ‘Flats, units or apartments’ are blocks of dwellings that don’t have their own private grounds and usually share a common entrance, foyer or stairwell. For further information refer to Functional Classification of Buildings, 1999 (Revision 2011) (cat. no. 1268.0.55.001).

For the data relating to houses and townhouses, dwellings that took more than three years to complete or were constructed in groups of 10 or more were excluded. As a result, approximately 2.5% of completed houses and townhouses were excluded.


Australian average quarterly completion times

Graph 1 illustrates the Australian average completion times, in quarters, for new houses and townhouses from September quarter 2006 to June quarter 2016.

The main difference between the two types of residential dwellings is that new houses had a lower average completion time than new townhouses. Over the 2006-2016 period, new houses had an average completion time of 2.27 quarters compared with an average completion time of 2.89 quarters for new townhouses.

Average completion times for new houses remained fairly steady over the period, averaging between 2 and 2.5 quarters to complete. Average completion times for new townhouses were more volatile over the same period, varying between 2.5 and 3.4 quarters.

Graph 1: Average completion time of new houses and new townhouses, Australia

ScreenHunter_15430 Oct. 12 19.39

Graph 2 illustrates Australian average completion times, in quarters, for new flats, units or apartments from September quarter 2006 to June quarter 2016. Completion times for flats, units or apartments are substantially higher than new houses and townhouses.

Average completion times for new flats, units or apartments have also been relatively volatile, ranging from a low of 5.1 quarters to 6.8 quarters over the period.

Graph 2: Average completion time of new flats, units or apartments, Australia

ScreenHunter_15431 Oct. 12 19.39

Average completion times for new houses

Graph 3 illustrates the five year average completion times for new houses over a 10 year period for Australia and the states and territories.

Average completion times for new houses declined in New South Wales, Victoria, Queensland, South Australia, Western Australia and Northern Territory in the 2011-2016 period compared to the 2006-2011 period. In contrast, Tasmania and Australian Capital Territory recorded a slight increase in average new house completion times. Western Australia recorded the largest decrease of 0.24 of a quarter, while the Australian Capital Territory recorded the largest increase of 0.32 of a quarter.

Graph 3: Completion time of new houses, five year averages, states, territories and Australia

ScreenHunter_15432 Oct. 12 19.39

Average completion times for new townhouses

Graph 4 illustrates the five year average completion times for new townhouses over a 10 year period for Australia and the states and territories.

Australia, New South Wales, Victoria and the Australian Capital Territory all recorded increases in average completion times for new townhouses in the 2011-2016 period, compared to the 2006-2011 period. Queensland, South Australia, Western Australia, Tasmania and the Northern Territory all recorded decreases. The Australian Capital Territory recorded the biggest increase of 0.37 of a quarter, while Western Australia recorded the biggest decrease of 0.47 of a quarter.

Graph 4: Completion time of new townhouses, five year averages, states, territories and Australia

ScreenHunter_15433 Oct. 12 19.39

Average completion times for new flats, units or apartments

Graph 5 illustrates the five year average completion times for new flats, units or apartments over a 10 year period for Australia, New South Wales, Victoria and Queensland. These states account for the large majority of new flats, units or apartments under construction.

Average completion times between the two periods are fairly similar with New South Wales recording the biggest change with an increase of 0.51 of a quarter in the 2011-2016 period.

Graph 5: Completion time of new flats, units or apartments, five year averages, Australia, New South Wales, Victoria and Queensland

ScreenHunter_15434 Oct. 12 19.39

Given where national dwelling approvals and commencements are at currently (see next chart), the huge growth of high-rise apartment construction, and applying a six quarter lag between commencements and completions, Australia’s dwelling completions may not peak until early 2018.

ScreenHunter_15406 Oct. 12 14.20

[email protected]

Unconventional Economist
Latest posts by Unconventional Economist (see all)


  1. 2018 for completion and probs 2 more yrs for panic button. Proper homes correction still a half a decade away, at best.
    I guess Reusachtige is a smart qgirl. Not quite sarcastic or ironic as her comments seem most of the times.
    I am contemplating to buy in South Africa.
    Better return and with my deposit for median Sydney home I can outright buy 2-3 apartments, double that with 50% prepaid. Currency is not overly bad performer and rent return can be 10pc easily.

    • Then you would become the foreign investor pricing the locals out. Anyway you need to know that African governments do have a history of nationalism where they will hand assets to the people. South Africa may still be heading for the bottom where the other countries are climbing back out again.

  2. The actual thing that is most damaging on the downside when the time comes, is not the ‘length of the pipeline” from commencement to completion of structures, it is the length of the pipeline from the time of acquisition of sites in anticipation of “planning gain’, to the completion of structures.

    All the “asset value” volatility is in the land. That is what does the economic damage. The factor of inflation in the land when house price median multiples have risen from 3 to 9, is likely to be “20 to 30 times”. This inflation applies to YEARS worth of “land supply”, AND to the land underneath all existing property.

    The reason that Spain had a colossal bust in 2007, but the UK never has a colossal bust, just a continued chronic rising land price trend with shorter-term fluctuations that always quickly resume inflating from a higher and higher “trough” each time, is this. The UK has such a strangulatory planning system, that the “supply of land” being “gamed” by speculators is always small. All greenfields land subject to the painful, drawn-out development permission process, truly is going through this process for the reason that it will be developed just as soon as it becomes possible. The housing shortages are chronic, and the planners assumptions that “redevelopment to higher densities” will provide a healthy supply of housing, is an illusion (but that is a subject for another essay). The actual land price inflation over the decades, relative to benchmark comparison cities in the USA without planning constraints, has reached a factor of 100 to 900 (the high end in London’s case). However, this is the result of a very long trend, not a one-cycle spike, the quantities of land traded are small, and the extent to which people are mortgaged or locked into renting and/or cohabiting, has found its own level.

    Spain was marked by a planning system that was a culpable participant in the “gaming” process for a pipeline of greenfields land supply that has been estimated by some economists to have been 7 years long. Local government ability to extract fee income, is leveraged by the creation of high levels of “planning gain” in a de facto quota racket. However, on top of this 7 years apparent pipeline of acquired sites (at inflated prices) could be added the land already being speculated in in anticipation of future rezoning.

    Unfortunately, it seems that politically and intellectually, the likely policy responses to experiencing a Spain-type bust, will be ones that move closer to the UK model, rather than reversion to the automobility-based competitive land supply model that worked so well for decades in most of the first world.

    • Interesting idea. However, if you’re going to talk about comparisons between Spain & UK, you should also bring in Ireland which busted at exactly the same time and Spain. Let’s not forget that unlike the other two the UK has its own currency which allowed massive BoE intervention in the credit crisis in a way not available to the other two. Anyway, I’d like to know how the Irish planning situation compared to the UK’s. BTW, the planners know full well that their job is to help the speculators, developers and banksters rob home buyers and renters. It’s a criminal cabal and is co-ordinated by parliament, make no mistake, This is why I ignore the Greens, they are very much part of this criminal enterprise.

      • Thanks, Jacques. Ireland was a not so severe version of Spain. The prices did not go as high. Irish cities have already had decades of suburban sprawl. The potential for a land-supply / development boom and bust is probably greater when you are starting from denser, more compact city cores. There is absolutely dynamite data available on what happened in land values and property development in the USA in the 1920’s, it was very like Spain in the 2000’s. I can post some quotes if you like.

        The 1930’s Depression was probably as bad as it was because of this, rather than because of the stock market thing – Phillip J Anderson and others argue this, and I am convinced of it. Spain now is like the US in the 1930’s, which tends to support this argument. Again, I have quotes I can post.

      • Yes, Spain does have dense cities and towns.. People tend to live in apartments. This seems to be in part of Franco’s legacy. Ironically, under him, housing costs were kept affordable to encourage home ownership. From what you say it wasn’t a shortage of land, it was land banking that was the issue. This was deliberate and clearly the major parties, the PP and PSOE were equally to blame Planners were irrelevant.

      • Phil, when I said the planners were “irrelevant” I really meant that planning was just doing the bidding of their political masters in strangling supply. Please post link anyway.

      • Sorry it has taken me a long time to remember to look at this thread again. I am working on an essay that will incorporate a lot of quotes from material that is not online, or is hard to obtain. The best goldmine of data and explanation is in Louis Winnick (1953) “Wealth Estimates for Residential Real Estate, 1890-1950”.

        A later volume with two co-authors, Grebler, Blank and Winnick (1956) “Capital Formation in Residential Real Estate”, includes the following nugget:

        “…Another phase of land development has been of some historical importance….the use of real resources and capital funds in the development of premature or defunct subdivisions. These are subdivisions which were laid out without either construction of buildings at the time, or such construction later. Resources were expended for plotting, grading, utilities, streets and so forth. Large amounts of equity and borrowed funds were sunk into these developments. Financial intermediaries were involved through mortgages and other debt instruments. The frenzy of speculation in more or less improved lots during real estate booms in the nineteenth century and through the twenties in (the twentieth) century at times attracted perhaps an even wider nonprofessional participation than the speculation in stocks…

        “…Neither the recovery of building during the late thirties nor the high volume of construction since the end of World War Two has produced speculative subdivision activity of the magnitude and character observed in comparable earlier periods…

        “…Some of the premature subdivisions have been used for housing since. However, suburban development has in most cases jumped over these subdivisions because of the multiplicity of small-lot ownership, abandoned property, clouded titles, accumulated tax arrears, and outmoded street patterns have made land assembly for new building operations difficult if not impossible. In cases where defunct subdivisions were used it has often taken twenty to twenty-five years before the use has occurred…”

        Grebler, Blank and Winnick note the following supporting references:

        A.M. Sakolski (1932) The Great American Land Bubble
        Homer Hoyt (1933) One Hundred Years of Land Values in Chicago
        Ernest M. Fisher (1933) “Speculation in Suburban Lands” (American Economic Review, March 1933 Supplement)
        Philip H. Cornick (1938) Problems Created by Premature Subdivision of Urban Lands in Selected Metropolitan Districts
        Helen C. Monchow (1939) Seventy Years of Real Estate Subdividing in the Region of Chicago
        Homer V. Vanderblue (1927) “The Florida Land Boom” (Journal of Land and Public Utility Economics, May 1927)

        Phillip J. Anderson (2008), “The Secret Life of Real Estate and Banking” (P247) suggests:

        “…The causes and effects of the 1920’s stock market boom and bust are still much debated today….Historians and market forecasters would do better to turn their attention to the concomitant real estate boom and bust….A land price collapse has far wider implications than a mere stock market crash. They are not in the same league…”

        Anderson quotes from Charles Kindleberger (2005), “Manias, Panics and Crashes” (p.103):
        “…Real Estate Loans, not failed stock broker accounts, were the largest single element in the failure of 4800 banks in the years from 1930 to 1933…”

        Hope this helps make the case!

  3. Anyone know why QLD build times are so much shorter than NSW/VIC? For units, I figure it could be lowers numbers of units per site, but that doesn’t explain the difference for detached houses.

  4. Apartment commencements peaked in March quarter, so presumably that implies that late 2017 is when we should (roughly) expect things to start rolling off?
    But bearing in mind that completions always seem to undershoot commencements towards the end of the cycle, conceivably the peak could in fact be somewhat sooner?

    It would be an interesting exercise to compare time from peak commencements to peak completions for past cycles (bearing in mind that the shift to apartments will extend this out for the current cycle).