Sunday, January 17, 2010

Population growth and the residential property market

I have been asked to develop further my ideas on population growth and residential property.  I hope to make it clear that arguments using population growth as a cause of future house price growth are probably misleading.

The first chart (above) shows the rate of population growth (RHS) and the the growth in the ABS capital city price index (LHS).  There are two important things to take away from this chart.
1. The rate of population growth can change very rapidly, and extrapolating past trends will always miss changes to this rate.
2. There is no significant relationship between these two figures over the period.

In fact, this chart supports the ideas espoused (for example here and here) that population growth and home prices are simultaneously determined by other factors, such as the health of the economy as a whole.  We should really put to rest all arguments centred around population growth and the apparent housing shortage.  I always wonder how this population growth occurs without new development anyway?  Apart from small changes due to increased occupancy rates,  population growth and development reinforce each other.

Even if population growth did strongly explain home price growth, one must imagine that an unexpected future decline in the rate of population growth would lead to price declines.  Does that mean that to sustain prices we need to sustain this rate of population growth indefinitely?

Either way you take it, invoking population growth to explain house prices is inherently misleading - why not simply price level and population level?

The second chart (below) show the ABS capital city price index in real terms, and the NSW median household income in real terms.  After the 1980s boom, the house price index did not recover its real 1989 peak until 1998 - a period of 9 years with no real capital gain.

What the data misses is that during this 9 year stagnation, the quality of homes, in terms of size and location, was probably improving.  A hedonic index (such as the RPdata-Rismark index), had one been around at the time, may have shown and even lower real return, or in effect, a longer period of stagnation.

What is also interesting is the significant change in the ratio of prices to income since the late 1990s, even though Australian mortgage interest rates have been relatively stable since about 1996 (between 6 and 8%).

So what does it all mean?

Essentially, we are in for either a long period of stagnation in residential property prices, or, if my predictions of a W shaped financial crisis are correct, potentially a severe short term decline in prices with a very slow recovery.  Of course, the RBA has plenty of interest rate ammunition should another crisis precipitate.

Time will tell, but my advice (if you want to know) is that potential home buyers, especially first home buyers, should be in no rush to get into the market.


  1. As the population grows faster than houses are being supplied, it begs the question 'Where are they going to live ?'

    Will the demand for houses coupled with the current severely restricted supply cause prices to rise or fall or stagnate ?

    To attempt to correlate population growth & house prices on a single graph is an oversimplification.

  2. Thanks for the comment. I guess my starting point for answering your more detailed question is "Where do they live now?". Remember, all population growth statistics show how to we got to the present, not how we will get to the future (although this is one figure that is prone to ill-informed extrapolation). Future people will live in houses currently available and those supplied IN THE FUTURE. If we have experienced some kind of 'shortfall' in construction it will translate into higher occupancy rates, as we have seen. If this higher occupancy rate is likely to break out as housing demand in the future, so be it. But the question remains - what is going to change to make those people who are occupying together because of high prices suddenly be able to express their demand and purchase new dwellings when prices are even higher? How will all these people suddenly and dramatically increase their incomes above without the incomes of others also rising?

    One of my main points here is that population growth in a region is driven by the broader economy, including the demand for labour (which itself is driven by the supply of new homes, leading to a self-reinforcing cycle).

    The second important point is that if you believe population growth drives prices, then a sudden decline in the rate of growth will lead to a price crash. The first chart shows how quickly the rate of population growth can change, and in 2009 we were at record rates of growth.

    On your final point, if you believe population growth causes house price growth, I would be interested to see some evidence from the 57 countries listed here who had higher population growth in 2008 than Australia did in 2009.

    Apart from the role of population growth, probably one of the main disagreements I would have with your comment is that supply is somehow constrained. Does that mean you believe that if land holders could build any type of building on their property without restriction that prices would be significantly lower? And that is town planning controls where 'tightened', that it would lead to higher prices? Maybe that's a topic for another time, but as you would have guessed, I also disagree with that.

    As I said, time will tell.

  3. Where do they live now ?

    You made that point that 'Apart from small changes due to increased occupancy rates'. I asked where could they be living ? in caves ? caravan parks ? motels ?

    I feel that people tend to stay with parents, take in boarders, share houses in bad economic times, etc.... and when the future is a little more stable they feel confident enough to decrease household densities.... and there's more demand for houses.

    Try offsetting the ave house price growth by +2 periods - can you see any pattern now ?

    I'm not sure what you point about rising incomes is ? ....another blog post perhaps ?

    I believe popln growth is a long term factor in house price growth. Over shorter periods (<5 yrs?) other factors (eg interest rates, economic growth, regional factors, etc) cloud the issue.

    I believe population growth causes house price growth in Australia for 2 reasons.

    1) Supply is restricted both by recent financing issues and for the long term by planning laws

    2) Desirable land (CBD, close to amenities & jobs) is restricted. We've got lots of undesirable land (deserts, Western Sydney, etc). As popln grows more people want to live near desirable locations - ergo prices must rise.

    If land holders could build any type of building on their property without restriction then prices wouldn't really be significantly lower - the land & building cost are roughly 50% each. If land cost only half as much (not likely IMO) then prices would be only 25% lower.

  4. Cameron - if your graph covered the 2009 year you would note that house prices snapped upwards to more closely follow population increases. I suspect that some continued upside is to be expected, a;though personally I would prefer a flat market for an extended period.