The ‘housing shortage’ in historical context

The Department of Infrastructure and Transport released a fascinating report late yesterday looking at the last 150 years or so of population and housing stock development. The report is mostly quite balanced and is well worth your time. It begins with a long term view of current population versus housing stock ratios:

Australia’s recent levels of housing supply have been lower than population growth levels for the most sustained period in a century (Figure 2-24). The four previous periods over the past 150 years when population growth levels were significantly greater than increases in dwelling stocks were the 1890s Depression, World War I, the Great Depression and World War II.

Then canvasses occupancy rates:

Household occupancy rates rose steeply from 1860 to 1889 as cities struggled to cope with population growth rates up to four times higher than those of today (Frost and Dingle 1995). Occupancy rates fell just as quickly after the 1890 crash before stabilising. There was a further fall during World War I with troops in Europe and the Middle East and a virtual
cessation in migration.

From 1920, occupancy rates again started to fall, punctuated only by World War II. They stopped falling in the mid-2000s and have even begun to rise slightly, reflecting the growing gap between population growth and dwelling supply (Figure 2-25). Somewhat paradoxically, Australia finds itself in a position of also having an increasing number of unoccupied houses. Reasons for this increase are not fully understood, particularly considering the shifting trend in occupancy rates since the mid-2000s.

And the reasons for the compressed rate:

From 1996 to 2011, households with one or two people (couples without children, lone person households) grew as a proportion of households by three per cent. In the same period, the proportion of couple families with children declined by around four per cent, while the proportion of one-parent families grew slightly (Figure 2-27).

If household occupancy rates are increasing at the same time as the proportion of smaller households is rising, the increase in occupancy rates must be occurring in family households.

The study also notes that:

Occupancy rates respond to the relationship between the number of additional dwellings required by population increase and the number of these new dwellings actually built. Small shifts in occupancy rates can significantly alter the number of dwellings required to house a population. As an example, Sydney’s current population of 4.6 million occupies 1.6 million dwellings, resulting in an occupancy rate around 2.8 persons per household. Theoretically, if the occupancy rate rose to 2.9 persons per household then more than 50,000 dwellings would no longer be required. Sydney’s current population growth rate demands roughly 20,000 new dwellings per year, so one impact of a theoretical increase of just 0.1 in Sydney’s occupancy rate would be that no additional dwellings would need to be constructed for two and a half years.

And it concludes:

The current gap between housing supply and the population increase we are experiencing is the largest and most sustained in a century and this is the only significant period of undersupply relative to population growth that has occurred outside the Depressions or World Wars since at least 1860.

Since 1996, we have seen the most acute and sustained increases in house prices in at least 130 years, although there has been some moderation since the Global Financial Crisis. Outright home ownership has decreased from 60 to 46 per cent, between 1996 and 2011.

Rising house prices appear to be closely related to rising land prices, driven by declining lot production per capita, falling lot sizes and sharply rising costs per square metre. Also associated with price rises is an increasing premium for living near city centres. Dwelling construction costs is not the largest contributor to price increases.

In response to the gap between population growth and housing supply since the mid-2000s, occupancy rates have stopped falling and have begun to rise again, mainly in households comprising families with children. With government housing having declined proportionally, demand for rental properties remains high. Rental income has continued to increase, weekly rents have redistributed towards higher prices and rental vacancy rates have remained low across all capitals.

My summary has not done the report justice. It is an excellent piece of work that describes pretty fully the cross-currents at work in housing.  The report may strike you as bullish for housing and indeed it is in some measure. But my own view is that although some  pick-up in construction can be reasonably expected in the post-mining boom economy, it will be unable to sustain the income and credit growth needed to satiate our desire to live in ever smaller households. Although global authorities have done a good job of preventing a rerun of the Great Depression, the dynamics of an epochal develeraging are the same. We are no different, especially since we rely on global funding for our high house prices. In this context, the ‘housing shortage’ will more likely be resolved by some new construction but mostly by the occupancy rate rising as asset prices continue to fall over the long term.

2012 05 Infra1360 Mcu Soac Chapter 2 Web Fa


  1. Thanks HnH and whoever linked to this yesterday in the comments on another piece. Some very interesting material. I find the comparison to Depression and the World Wars quite telling. Perhaps in several decades time we will see current economic changes as equally momentous.

    I do have a quibble with this part of the paper ” theoretical increase of just 0.1 in Sydney’s occupancy rate would be that no additional dwellings would need to be constructed for two and a half years.” It’s a pretty daft argument. Occupancy rates are an average based on millions of data, so a small increase in the mean can only occur if a lot of above average results are introduced. It’s not like households just decide to increase occupancy by 0.1 persons.

      • Yeah agreed. My quibble is with the way the authors are looking at this particular metric. They should observe that it requires a large number of households increasing their occupancy by one person to achieve a small change in the average occupancy rate.

        • well, the numbers are right there… it requires 50,000 people to decide to be in a share house instead of their own 1 bed. There was an interesting statistic I saw a while ago, about the number of empty bedrooms – that might be something you could watch go up or down more directly than the occupancy rate.

          • Mining BoganMEMBER

            Economically worthless anecdote time.

            My fellow bogans who are single own apartments. Thet’re not home much so can lock them up when away at work or in Bali. Also, apartments are usually close to the amazingly deep and fulfilling pastimes that bogans are wont to do.

            The last year or two they’ve often spoke about renting the spare room out on the cheap to ‘lend a mate a hand’. This is where all those empty bedrooms will be sucked up when TSHTF.

            It’s called survival. You have to huddle together to keep yourself warm in the winter.

    • But the calculations look good. In new Zealand’s case a 0.1% increase in the dwelling density over the 4.4 mio of us would see us have an oversupply of about 70,000 homes! Given that we are losing population ( about 4,000 net people per annum) and have been now for 4 years, I wonder at those who see a ‘housing shortage’ here.

  2. This is the best government housing report that I have seen for a long time. The fact that so few homes have been built in the face of rapid house price growth between 1996 and 2010, as well as strong population growth since the mid-2000s, is testament to how messed up Australia’s housing market has become.

    Put simply, Australia needs lower land prices, which requires a more responsive (and liberal) planning system, as well as improved funding and provision of housing-related infrastructure. Broad-based land taxes in place of inefficient transaction taxes would also help.

    • Exactly UE. Can anyone take a punt at the possible reasons for the rapid increase in vacant dwellings from 1950-1970?

      • House price appreciation. It looks remarkable similar (but with different intersect) to the chart of real house prices in Sydney and Melbourne.

        As house prices rise, holding onto vacant houses becomes profitable, so it happens more often. As prices fall, watch the flood of properties come onto the market… This is a massive reservoir of houses, and is the elephant in the room, completely missed by this report.

        • All in the wrong place though Nathan.
          Would you buy a holiday home in your current suburb, or would it be by the beach and far away?

          A weekend getaway 5 minutes down the road just doesn’t have that get away from it all romantic feel.

          • Unfortunately a bit late in my reply, but I’ll say it anyway – this is absolute rubbish. The stats are all there, for everyone to see, at the ABS census website.

            You can draw maps of the areas with highest unoccupied %, and while the holiday locations top the list (often above 50%), in the major cities the inner-city areas are the worst, at around 15! It then drops gradually until you get to the outer suburbs, where (not including unsold housing estates), the unoccupied rate is about 2%.

            Seriously, dig through the stats and you’ll see that this is why the “shortage” will suddenly disappear as prices continue to drop.

        • Good point, Nathan. “DrHousingBubble” has a term, “shadow inventory”, that he tries to track over the years in California.

      • Can anyone take a punt at the possible reasons for the rapid increase in vacant dwellings from 1950-1970?
        Yes – holidays.
        The biggest jump coincided with the Whitlam governments increase in the standard holidays to 4 weeks per year (early 70’s).
        Another factor since 1970 is cheap airfares and the boom in OS travel. Now many rich people can keep a place in 2 or more locales.

  3. There has been a very slight reversal in the rate of divorce lately (can’t remember where I read that) but unless there is a significant reversal in the divorce rate, we will continue to see households split and the number of homes needed increase.

    It’s not just the household formation rate, but the reverse also increases the need for housing.

    “For richer or poorer or until I can find a good divorce lawyer”

    I will read the report in depth later this morning.

    • A decrease in STD’s (sexually transmitted debt).
      Actually I reckon financial stress is right up there as a big cause of divorce.

      • Well Jack it certainly is after the divorce. The best way to lose a fortune is to give it to your lawyer in a court battle over something that you probably no longer want.

        A quote that I saw the other day was –

        “You spend 25% of your wages getting her, 42% of your wages keeping her, and 108% of your wages getting rid of her”

    • Mining BoganMEMBER

      So the housing industry needs divorce proceedings to continue unabated to keep a floor under housing demand?

      I’ve seen you bag folk for wishing misery on others when they’ve voiced their wishes for a price drop. Which misery is worse?

  4. This is one more bias writing report. They started with the conclusion about “chronic shortage that drives home prices up”. They found some historical data and used it as “evidence”.

    The only thing that this data proves is that we were building less in 2000s that in 1990s or some other periods. It says nothing about real need, and real supply (including overbuilding over the previous period). Similar conclusion can be made based on USA construction rate in 2000s. It was very low in historical terms, but more than enough to provide excess supply that helped the housing market crash.

    Sure, we built less in 2000s than before but there is no proof whatsoever that we build less homes than needed. Same population growth with different demographic dynamics creates different household growth and requires different supply.
    It is naive to think that population growth of 2% in 1970s created the same increase in number of households as population growth in 2000s. The main driver of adult population growth in 2000s was immigration, especially foreign student population growth. It is foolish to think that immigrant families and foreign students create the same housing demand as Australians in their 20s who are single.

    Under normal circumstances, household size dynamic has little to do with housing supply or cost. There are many places with similar demographic, high home supply and low housing cost that recorded household size increase. These places, like Texas, recorded high family immigration (similar to us). That immigration pushed household size up and reduced housing demand per new adult resident. As a consequence, they needed to built less homes for the same increase in adult population to meet demand.

    With the same “scientific credibility” as used in this report (or lack of it), construction slowdown in 2000s can be attributed to huge overbuilding in 1990s. We could say that constructed so many homes in ‘90s, more than enough to offset slower construction in 2000s. Reality is that we had higher demand for homes per new adult resident as a consequence of little baby-boom in early 1970s and slow immigration.

    Bottom line is that this report just adds to the “shortage spruik” enviroment that benefits construction and RE industry and doesn’t help young people at all.

    • +1 . It is a rehash of the same old underlying data and the same old methodologies to arrive at an “undersupply”. It brings nothing new to the table.

      It is paradoxical that the housing industry keeps arguing that we have an “undersupply” at the same time as their members, including 100 year old building companies that survived the Great Depression, WWII, 1992 recession and the GFC, are going bust.

      There is a lot of information asymmetry in this area, possibly deliberate. But Mirvac cancelling its unit development near the Brisbane CBD and HRH Triguboff, having run out of China stimulus, now begging for local stimulus are a good sign that we don’t have a shortage, either on the outskirts or next to the CBD.

      • actually mav, it’s a good sign that we will have a shortage even if we don’t have one now.

        What I don’t understand about you and rav is why do you insist on shooting all your toes off?

        • What I don’t understand about you and rav is why do you insist on shooting all your toes off?

          That is a very intelligent and incisive on-topic comment from you, Pete./sarc

          But I expected nothing better from a merchant of debt other than desperate hand waving to detract attention away from the subject matter.

      • +1 Mav

        We don’t have an undersupply of housing in Bris, we have an undersupply of buyers willing to pay $800,000 for a jerry-built two bedroom dogbox.

        • Oh come on now, that just a blatantly inaccurate made up piece of misinformation designed to hide the truth from the sunlight.

          If you are going to make things up please try to make them vaguely plausible.

          In Brisbane $800K buys a very very nice home well situated in a suburb with all the normal amenities and in easy drive to the CBD. Anyone can do a search on RPData to confirm that, why didn’t you?

        • Although PF sounding a bit like Oddball from Kelly’s Heroes, he’s right..

          Please refrain from hyperbole The Patrician – as Peter shows, has some very nice, large homes close to the city for that price:

          Same cant be said for Melbourne – probably add another 100-200K…

          • well played Pat. You forgot to add it starts at $590K….still a lot of money, dont get me wrong I think property in Bris is still 20-40% overvalued (but up to 15% overvalued is ok for a homeowners premium vs. an investment)

            But FFS, what is it about the comments section that brings out the pedants?

            I’m really looking forward to my holiday break next week.

          • Thanks Prince,

            Apology accepted 😉

            I never said you couldn’t get other 2brm dog boxes for less.

            My point (using this real example)is that a significant amount of the current supply is priced outside any sensible market and the billionaire owners are not motivated to sell due to our redundant tax settings.

      • Mav, it is worth noting that in the UK, each cycle of the housing market over the decades, has had a weaker and weaker supply response and a more and more volatile price inflation response. This is in the face of rising population and serious aging of the housing stock.

        And there is serious overcrowding in homes in parts of London, and the average ago of first home buyers has risen to 39.

        Meanwhile, building firms have been steadily going out of business and employment in the sector is about 1/4 of what it was 40 years ago.

        So where are the causes and where are the effects, in this situation? Strangled supply due to regulations.

        Of course building firms go bust even when there is a serious “undersupply”, when the whole market in which they are competing, has been made high-risk, riddled with “gaming”, and with bidding wars to the death for available land.

        • Of course building firms go bust even when there is a serious “undersupply”, when the whole market in which they are competing, has been made high-risk, riddled with “gaming”, and with bidding wars to the death for available land.

          I am not talking about land bankers aka developers. Builders.. Phil.. Builders. They don’t go around bidding for available land.

          Developers, who do the bidding and the banking, have been very profitable. In fact, all evidence suggest that “supply constraint” has led to large developers making all the planning gains by gaming against smaller players and made huge windfall profits. But now, even they cannot game the demand side (which you choose to always ignore) beyond a point, leading to the recent softness as cited above (Mirvac, HRH).

          • Mav, growth-constraint urban planning, aka government enforced land rationing, forces anyone who wants to build anything, into bidding wars for their land. Even if it is just one lot for a spec house.

            If builders as well as developers do not want to indulge in bidding wars and/or high risk gaming, when there is rationing of the supply of urban land, they simply have to shut up shop. Even if they don’t, they frequently end up out of business anyway because of being squeezed between outbidding competitors for sites (and/or paying the price demanded by incumbent land owners), and what the buyers of the finished product can afford to pay, especially during a “down” phase.

            I agree with you about the smaller players being the first to be forced out of the industry altogether. This is certainly the case in the UK. But even the bigger players who might survive for a while, are somewhat like Politburo members during Stalinism. Riding the tiger might be fun while it lasts, but…….

    • +1

      I have been skeptical of the idea that Australia had somehow failed to build enough houses in the middle of the greatest housing bubble in history.

      Historically, a bubble left useful products in its wake. In the US, the housing bubble left millions of new houses. The dot com bubble left information super highways and new industries. During the railroad bubble back in the 19th century, it left 250000km long rail roads.

      If it turns out that Australia had indeed failed to build enough houses in the middle of the greatest housing bubble in history, that will raise serious questions as to why the Australian version of capitalism is so much inferior to the American one.

      • See my comment above about housing cycles in the UK. The construction industry in the UK has been building less and less houses in the face of higher and higher price increases in each cycle of the property market, for decades. The cause is the urban planning system, and Australia is heading the same way if reform is not imposed.

  5. Some of the cross currents over the last 30 to 40 years include:
    1. Divorce increasing from the mid 70’s providing demand for 1 and 2 bedroom apartments
    2. Increased wealth increasing number of holiday apartments
    3. Anglo and mediterranean populations reproduction rate falling below replacement but later arrivals having high reproduction rate
    4. Some housing stock reaching end of economic life or being of such limited appeal (fashion, configuration, local amenity) due to being another 30 to 40 years older
    5. Kids’ average marriage age blowing out from early 20’s to late 20’s deferring household formation and often leading to flats in garages, large houses being effectively dual occupancy, partly because of education or travel debts limiting ability to purchase.
    6. Shape of the population pyramid changing as baby boomers age leading to demand for a different dwelling mix and including need for assisted care accommodation.

    • Some of the cross currents over the last 30 to 40 years include:
      Exactly. It is a complex situation. The real shortage is evident by looking at high prices and high rents.
      The real suffering is evident by opening one’s eyes.

  6. I think the flexibility of the occupancy rate is seriously underestimated. It can and will rise if houses continue to be priced at a level that doesnt incentivise people to move out.

    This is a ridiculous situation for the least-densely populated country in the world to find itself in.

    Over time, mega-trends like IT and homeworking will continue to change the picture and confound the mathematics and assumptions of housing analysts.

    • +1.

      “This is a ridiculous situation for the least-densely populated country in the world to find itself in.”

      It is also very immoral for the said underpopulated country to have done this to its young people. In the past, “doing the right thing for the young people” was regarded as a moral obligation. Now they are just expected to be the “greatest suckers” in the great Australian Housing Ponzi.

      The “well meaning” advice peddled by older people who were put on the pig’s back in their own time by pro-housing policies, is sickening. It is more ignorance than malice, but the outcome in bankruptcies and divorces is just as painful regardless.

  7. It’s hard to argue with this data.

    Particularly fig 2.24 shows (as we all know) that ever increasing debt since the 1980s has has done nothing to increase housing stock.

    For those of us that like our kids to touch the grass when they play, this data probably even over-states the quality of the housing stock available.

    For mine, this shows two things:

    1. The massive failure of government policy over the last 30yrs – both on population/infrastructure and regulation of existing housing speculation;

    2. Peter Fraser is possibly right – there is a latent lag of housing that could feed back into price bubble conditions very easily if the credit setting/mood is right.

    Building zillions of dog-boxes in the boon docks is not the answer to this problem. The answer is about getting the settings right on our policy so we improve the utility of existing stock and getting the dead-weight passive speculators out of the markets.

    The punters should pay attention – it was the politicians from both parties that got us here. The have shown they don’t have the integrity or courage to make good policy.

    • I offer anecdotal evidence that there is no shortage, at least in Perth where there is supposed to be one. I walk for an hour in the evenings, both in Cottesloe (expensive) and a northern Perth suburb (blue collar). I have noticed up to 1 in 10 houses is unoccupied — permanently; usually at least one or two per street.

      Make of that what you will, possums.

      • Simple; it is speculators who have been “holding out” for capital gains and possibly have even taken the properties off the market because they are shunting themselves about it looking “flooded”. They possibly even know that there are thousands of young people who could actually buy those homes if the price was 40% lower. And worse, they expect the young people to get desperate enough, and the finance sector stupid enough, to actually play the “greater sucker” role; and they probably expect the politicians and the RBA to assist.

    • For those of us that like our kids to touch the grass when they play, this data probably even over-states the quality of the housing stock available.

      Forget price. To improve the lot of our children we need to produce more than one extra decent house for each extra family (raised here or immigrant).

      Are we doing this? No.

  8. Zero supply response in the face of rapidly rising prices is proof that the housing market is dysfunctional and the supply-side is stuffed, irrespective of whether you believe that there is a shortage.

    • Zero price response in the face of rapidly falling demand is proof of collusive price-fixing in the housing market, particularly when supply has been constant for the last 15+ years

  9. Have said several times and will say it again, you cannot take headline population stats at face value as significant qualitative issues have impacted growth.
    The definition was changed in 2006 with application of the 12/16 months rule whereby temporary residents are included e.g. international students, temp workers, 2nd year WHVs and their dependents, and at times Australian citizens and permanent residents if more returning than leaving…..