Housing shortage? What do the numbers say?

Most property investors & bullish news articles would have you believe that we have a housing shortage in Australia, however some figures that I've put together tonight would indicate that we may actually have an oversupply.

The figures indicate there is an oversupply of 430,000 dwellings (calculated using 20 years data), enough to fit an extra 1.1 million people. Even taking into account the recent 4 year shortage (where demand has outstripped NEW supply) there is no way that I can see we have a shortage of housing in Australia.

The table doesn't take into account everything, like housing that's demolished and not replaced by new and also assumes there wasn't a shortage in 1990, but other than that would love to see you guys challenge the figures with hard evidence (or agree that there is no shortage :p)!


Note: Apart from dwelling approvals, the numbers represent # of people.
# provided for figure calculated by multiplying approvals by .77 (assuming 15% replacement of old stock, 8% 2nd homes), multiplied by 2.6 (average person per dwelling). Compiled using ABS data. Final half of 2009 (pop growth) calculated by doubling first half as ABS figures not yet available.
 
Last edited:
Have a read of Housing and the Economy from the RBA.

So how do we square up the seemingly contradictory evidence that overall investment in housing is relatively high, yet there seems to be a shortage of dwellings? I think four factors help to explain this contradiction.

First, Australians are on average spending a lot more on each new dwelling. Real expenditure on each new dwelling built is now 60 per cent higher than it was around 15 years ago (Graph 5). This is due to improvements in quality and increases in size.

Second, a high proportion of dwelling investment is in the form of alterations and additions – i.e. upgrading existing houses rather than building new ones. Almost half of all dwelling investment has been accounted for by alterations and additions in recent years.

Third, a higher proportion of the new houses built are simply replacing existing houses that have been demolished. We estimate that between 2001 and 2006, around 15 per cent of new houses built replaced houses that had been demolished; 10–15 years earlier, that figure was less than 10 per cent.

Fourth, a significant proportion of dwelling investment appears to have gone into holiday homes or second homes. Census data show that the number of dwellings built has exceeded the increase in the number of households by a large margin. As a result, the ratio of the number of dwellings to the number of households has been rising over time; as at 2006, there were 8 per cent more dwellings in Australia than there were households. Presumably, most of this surplus reflects holiday houses and second houses.

(There are a couple of graphs in the original).

It makes a couple of relevant points, that your spreadsheet doesn't allow for.
 
keithj, if you read the bottom section in grey I multiply dwelling approvals by .77 to account for those exact figures.
Aah.... maybe you could state your assumptions/calcs in the main text.... it's really hard (for some of us) to read small italics text on a grey background in an image.

I'd question your assumption of 2.6 occupants per approval. What proportion of new dwellings have been higher density studios or 1 bedders with a lot less than that many occupants ?

Do all approvals translate to new dwellings ?

Isn't household density more relevant than a straight approvals/population growth absolute figure ?
 
Fair call re text, I've popped it underneath as well. Interesting point regarding density, I think the ABS has figures on number of bedrooms per dwelling and if I recall it's on the increase which would not support the theory of a larger percentage of high density (I'll see if I can dig up figures tonight)...
 
Hobo-jo, You need to take into consideration the fall in persons per dwelling during that time...



Persons per dwelling has recently ticked back up again (not shown on the chart above). I think the reason it started moving back up is because there are no longer enough new houses for people to continue spreading out at the rate at which they have wanted to for the past twenty years.

We also need to consider where the houses have been built. Have they been built where they are needed, and are they the right type of dwelling?

But I agree the shortage is a recent phenomenon, and is not present everywhere. I think the shortage is most pronounced in Sydney, and doesn't exist at all in other cities. Most commentators when discussing the shortage, are really talking about its implications for the future - i.e. if things continue as they have done for the past few years (record high population growth with low construction activity), then the shortage (in the places where it already exists) will just get worse.

For example, here is a recent chart from ANZ...



Now, personally I don't think things will be as bad as this ANZ chart suggests, because as most people know I am expecting a construction boom.
 
Besides, I know of a couple of small very rural villages and towns where there is an oversupply of houses - but no jobs, not close to anywhere that jobs can be found. Then compare this too a place like canberra, where there is a less then 1% vacancy rate and a new phenonenom called the 'working homeless' (those people who employed but unable to find housing for themselves and their families - they often live out of their cars).
 
I think the extremely tight rental market is a key indicator that there aren't as many dwellings relative to demand as previously. That could be just re-emphasizing Shadow's point about the location of housing.

In the last four to five years there looks to be a population boom without a corresponding boom in housing construction. This could just be a lag effect.

However I think one of the constraints is the shortage of tradies to build stuff. Many trades have as much work as they want, they can charge slightly more which means they have no particular need or inclination to do overtime.
 
I looked at the ABS data in this field mid last year.

Some points:

- total occupied and unoccupied dwelling stock in ABS 06 census was 8,606,000.
- around 150,000 new dwellings are completed each year
- demolitions are estimated by dept of families and housing at less than .002% pa. = ~25,000
- 2010 estimated stock will be around 9,106,000
- there are over 500,000 foreign students in Australia.

My views:
- I don't think unoccupied dwellings is a high number. What % of the population can afford an unoccupied dwelling in a low rental vacancy precinct?

- people living longer contributes significantly to the housing shortage. i.e. empty nesters and single elderly people occupying family sized homes close to major employment centres. This is exacerbated by a lack of retirement villages and their high relative cost.

- cumulative net overseas migration has contributed a growing % towards a total population growth of ~2% pa since 2004.

- there is a definite supply side constraint on the release of new dwellings, as indicated in the chart below.




Governments need to consider this when setting population growth policy. Building constraints are a combination of green policy restricting dwelling density in metro areas (preservation of std of living), lack of land and tax dollars to fund infrastructure on metro outskirts, lack of skilled labor, competition for labor from mining sector.

- Sydney and Melbourne have had significant demand for private dwellings from foreign students.


Nevertheless, when you divide estimated resident population by cumulative
dwelling commencements, you get a new dwelling for every 1.6 people. This seems like more than enough if the average household is 2.6 people.




However, the answer to whether there's a shortage or not, won't necessarily be unmasked by average household size......but rather, where exactly is the shortage, which is best represented by rental vacancy rates.
 
infill developments aren't taken into account either.

take one house on large block. put a triplex development on it. that's three houses, not four, because the old was demolished.

if the oversupply were the case, that says that there's a great number of houses across australia sitting empty.

i know there's barely ANY houses in Perth sitting empty....
 
Hi,

If vacancy rates are anything to go by and you can roughly accept the figures provided by SQM research then

Perth City Vacancy 7%
Melbourne City 3.4%
Sydney City 3.3%
Brisbane Inner 3.4%

Not the best figures if there is meant to be a rental shortage. There will always be a shortage of affordable places, just like I think there is a shortage of affordable new Porsche 911's.

http://www.sqmresearch.com.au/graphs/graph_vacancy.php?region=qld%3A%3AInner+Brisbane&t=1

As to the number of occupants per household, I will assume that this goes down in prosperous times and up when times get tough.

So current economic conditions could determine through number of occupants whether there is a short or not.

Interesting times we live in. Still in a bit of disbelief that anyone would consider that the shortage could be a myth. Certaintly would not have engaged any discussion in the past.

Cheers
 
there's certainly some contrary evidence floating around.

trouble is - the minute someone finds isolated pocket evidence, it's bandied about like it's median national values.
 
there's certainly some contrary evidence floating around.

trouble is - the minute someone finds isolated pocket evidence, it's bandied about like it's median national values.

A prime example of that is our last weekend papers here in the West.
I actually had to laugh at most of the property reports.
 
I am not going to get caught up in statistics because explaining housing shortage is futile because it OBVIOUSLY doesn't paint the whole picture.

I will briefly list out a few points why such statistics are either misleading or irrelevant.

1. Who cares if there is an oversupply of homes (even if it is true) when people quite frankly don't want to live there (i.e. regional towns, city fringe suburbs, high-rise apartments)

2. Does not take into account land bankers, people leaving houses vacant (for whatever reason)

3. Tax fraud (the house is considered vacant to the tax man, yet there are tenants there so landlord is effectively collecting tax free rental income)

4. Holiday homes

More correctly, there is a shortage of properties in places people want to live in and this will only get worst. Such areas are finite, and likewise, such properties are finite. Human history has shown that population increases each and every year (sometimes exponentially) and I doubt Australia will go against this trend (unlike Japan, but that is an entire different topic of discussion altogether). Also, I disagree with town planners who believe the answer is to build high density monstrosities near the CBD (or wherever people want to live). Firstly, I believe it would turn the streetscape into a piece of sh*t. Secondly, there would be immense red tape and/or community backlash especially with 'save our suburbs' groups. Thirdly, developers are only there to make money and will still charge the average 2 bedroom apartment in the $600K's which is out of reach for many Australians anyway (otherwise, why else would they need to live in Werribee, Melton, Cranbourne, Wallan etc...?). For that budget, they could get something closer, unless of course it is their dream to live in Werribee, Melton etc.... (of which I would have no argument, haha). Lastly (only because I cannot be bothered to explain further, because I just realised it wasn't brief afterall), heaps of people won't buy these high rise apartments anyway because of a combination of potential oversupply, anti-family oriented living, historically poor capital growth assets, high ownership costs (body corp) and general lack of character. I work in the property and construction industry and am immensely familiar with these apartment developments, OTP's and existing ones and there are many that are either vacant (been on the market for months), had their asking prices slashed, unsold during the OTP process or developers needing to spice things up with 'free extras' to entice purchasers.

All in all, if population does double in Melbourne in say, the next 20-30 years (or whatever the prediction is), then all I can say is that you don't want to be poor because otherwise, you would be living in the oldest of crappy apartments in the CBD or some random shed in Lara (because Werribee would be too expensive, haha - probably the only time I have said something positive about Werribee). The rich gets richer!!!
 
I will briefly list out a few points why such statistics are either misleading or irrelevant.

1. Who cares if there is an oversupply of homes (even if it is true) when people quite frankly don't want to live there (i.e. regional towns, city fringe suburbs, high-rise apartments)
Do you have evidence to prove that the empty homes are in these areas or are you simply speculating?

2. Does not take into account land bankers, people leaving houses vacant (for whatever reason)
I think you should look up the definition of land banker before coming back to this discussion, it has nothing to do with this discussion

3. Tax fraud (the house is considered vacant to the tax man, yet there are tenants there so landlord is effectively collecting tax free rental income)
What percentage of the 430,000 extra dwellings is subject to this? Do you have stats? If not how can you claim my stats are misleading or irrelevant?

4. Holiday homes
Assuming you mean second homes/holiday homes, they are taken into account, read the first post.
Had a good chuckle at some of these points, thanks DeeHwa.


I think WinstonWolfe has perhaps touched on one of the more important conflicts to the table and that is the large number of foreign students in Australia, which I had not considered. A couple of points regarding that, I imagine many students are living with existing residents either through boarding rooms, house sharing, staying with families, etc and I am trying to find confirmation whether student housing (e.g. oncampus & student accommodation like unihouse in Adelaide) is actually considered residential or commercial because this will have a bearing if those staying in such sites should be considered.
 
a point made earlier - but i wanted to repeat - approvals does not equate to building. i personally has 4 townhouses approved in january 2008, but not looking to start building then until later this year. with the really big boys in town the process from approval to building takes even longer - so there is a lag during which population increases.

also, my townhouse build is going from a house occupying 4 people (family) to properties occupying 8 people, assuming a couple or share renters move in ... some of them may be single renters ... so i don't significantly increase the number of people per dwelling space.

according to your figures (not sure how you got 430,000 from an net oversupply of 111,000?) the oversupply is merely around 5.3%. this is about normal vacancy rate when balance is met (5% vacancy is considered a balanced rental market). if one takes into effect an area that has a vacancy rate of 8% - such as broken hill having 60 rentals advertised on realestate.com alone, dubbo has 84, parkes has around 50 - then somewhere else would have a lower vacancy rate to compensate.

i also feel that your assumption of 2.8 people per dwelling is overstated considering the recent increase in single person dwellings, smaller families and the elderly staying in their homes for longer.
 
Last edited:
.... and I am trying to find confirmation whether student housing (e.g. oncampus & student accommodation like unihouse in Adelaide) is actually considered residential or commercial because this will have a bearing if those staying in such sites should be considered.
Add caravan parks with permanent residents, and grey nomads to your list to find out about too :).

I think the result your spreadsheet is trying to produce is affected mostly by the constantly changing persons per household.

As others have pointed out household densities have been falling for decades.... more divorces, more elderly living alone.

Household density increases in tougher times as people share, recently departed kids move back home, kids getting older but can't afford to move out. There is elasticity in demand. So using a constant 2.6 people per house isn't going to reflect reality.
 
a point made earlier - but i wanted to repeat - approvals does not equate to building. i personally has 4 townhouses approved in january 2008, but not looking to start building then until later this year. with the really big boys in town the process from approval to building takes even longer - so there is a lag during which population increases.
I took the idea of approvals from an RPData blog that was looking at supply/demand and just further analised the figures:
http://blog.rpdata.com/?p=227

according to your figures (not sure how you got 430,000 from an net oversupply of 111,000?) the oversupply is merely around 5.3%.
No it's not, unless you are calculating against the entire housing supply which is not the right way to look at it. I don't have the spreadsheet with me to get an exact figure, but as a percentage of the new dwelling approvals the 430,000 figure is closer to 13-15% of the new inventory.

It's 430,000 dwellings, deprived by dividing number of extra people we can accomodate for (1.1 million) divided by 2.6 per dwelling.

i also feel that your assumption of 2.8 people per dwelling is overstated considering the recent increase in single person dwellings, smaller families and the elderly staying in their homes for longer.
It's 2.6 and not an assumption, it's an ABS statistic.

Add caravan parks with permanent residents, and grey nomads to your list to find out about too :).
I think the result your spreadsheet is trying to produce is affected mostly by the constantly changing persons per household.
As others have pointed out household densities have been falling for decades.... more divorces, more elderly living alone.
Household density increases in tougher times as people share, recently departed kids move back home, kids getting older but can't afford to move out. There is elasticity in demand. So using a constant 2.6 people per house isn't going to reflect reality.
People living in caravan parks would add to my argument that there is an oversupply as it draws people out of normal residential homes...at least that's how I see it?

All the changes in household density are taken into account with average household size...it's varied between 2.5-2.8 over the last 20 years, not enough to make a large change to results.
 
Had a good chuckle at some of these points, thanks DeeHwa.


I think WinstonWolfe has perhaps touched on one of the more important conflicts to the table and that is the large number of foreign students in Australia, which I had not considered. A couple of points regarding that, I imagine many students are living with existing residents either through boarding rooms, house sharing, staying with families, etc and I am trying to find confirmation whether student housing (e.g. oncampus & student accommodation like unihouse in Adelaide) is actually considered residential or commercial because this will have a bearing if those staying in such sites should be considered.

Hobo-Jo

1. I don't need evidence to prove that people prefer not to live on city-fringes and regional towns (property values and abundance of land dictates this) and city high-rise apartments (know a lot of developers with many apartments less than 50% sold OTP and a lot of apartments for sale which have been on the internet for months - just do a search)

2. Land banking has everything to do with this discussion. What do you think a lot of overseas investors are doing? Obviously I don't know the % of people who keep them empty, but it all contributes to the fact that it leaves 1 less house that 'could' be occupied but is not.

3. Er, if the ATO doesn't have stats on tax fraud, what makes you think I do?

4. For the purposes of argument, 'holiday homes' can also be treated as land banking and leaving house vacant (for whatever reason)

The overall point I am trying to make is that housing shortages (or oversupply) is not as critical of an issue as people make it out to believe and the abovementioned points I made is really just an illustration of HOW the oversupply statistics could be misleading and misrepresented. I don't need to prove anything.
 
Top