2018 Demographia International Housing Affordability Survey

By Leith van Onselen

The 14th Annual Demographia International Housing Affordability Survey has been released and, once again, it ranks Australia as having one of the most expensive housing markets out of the countries surveyed.

This year’s report assesses 293 urban markets in nine countries: Australia, Canada, Hong Kong, Ireland, Japan, New Zealand, Singapore, United Kingdom, and the United States as at the third quarter of 2017.

The survey employs the “Median Multiple” (median house price divided by gross annual median household income) to rate housing affordability. This measure is widely used for evaluating urban markets, and has been recommended by, amongst others, the World Bank and the United Nations, and is used by the Harvard University Joint Center on Housing.

The Survey ranks urban housing markets into four categories based on their Median Multiple, from “Affordable” (3.0 or less) to “Severely Unaffordable” (5.1 & Over) [Table ES-1]. Average multiple data (average house price divided by average household income) is used in Japan, since data for estimating medians is not readily available.

ScreenHunter_17047 Jan. 22 16.48

According to the Survey, housing affordability remained poor across most major metropolitan markets in 2017 (i.e. with over 1 million people).

At the national level, Hong Kong has by far the most unaffordable housing, with a median multiple of 19.4. Australia is the second most unaffordable market with a median multiple of 5.9, followed closely by New Zealand (5.8), Singapore (4.8), the UK (4.5), Japan (4.2), Canada (3.9), the US (3.7) and Ireland (3.7):

As shown in the above table, all but two of Australia’s 22 markets captured in the survey are ranked as either “Seriously Unaffordable” (5) or “Severely Unaffordable” (15).

The next chart shows that Sydney is the second most expensive major housing market:

A break-down of Australia’s rankings are provided in the below table:

In 2017, the bubble epicentres of Sydney (12.9) and Melbourne (9.9) hit their highest median multiples on record:

The overall decline of housing affordability in Australia over the past few decades is also clearly evident in the above Demographia chart. Whereas all major Australian markets, except Sydney, had Median Multiples of three in the early 1980s, today all are ranked at around five or above (well above in the case of Sydney and Melbourne).

Demographia notes that housing affordability has worsened over recent decades across all of the nations surveyed:

Historically, the Median Multiple has been remarkably similar among six surveyed nations, with median house prices from 2.0 to 3.0 times median household incomes (Australia, Canada, Ireland, New Zealand, the United Kingdom and the United States). Housing affordability remained generally within this range until the late 1980s or late 1990s in each of these nations (Figure 2). In recent decades, house prices have escalated far above household incomes in many parts of the world.

One of the key contentions of the Demographia Survey is that higher land prices are the principal contributor to the rapidly increasing home prices in unaffordable markets, as well as increased speculative activity. These land prices include the cost increasing influence of land supply restrictions (such as urban growth boundaries), excessive infrastructure fees and other overly strict land-use regulations:

The key to both housing affordability and an affordable standard of living is a competitive market that produces housing (including the cost of associated land) at production costs, including competitive profit margins.

…virtually all the severely unaffordable major housing markets covered in the Demographia International Housing Affordability Survey have restrictive land use regulation, overwhelmingly urban containment. A typical strategy for limiting or prohibiting new housing on the urban fringe an “urban growth boundary,” (UGB) which leads to (and is intended to lead to) an abrupt gap in land values (Figure 16).

Contrary to expectations that higher densities would lower land costs and preserve housing affordability, house prices have skyrocketed inside the UGBs…

In fact, the higher land prices are consistent with the basics of economics, and this has led to higher house prices. Virtually across the road land value gaps of ten or more times result.61 This destroys the competitive market for land by removing the “supply vent” necessary to maintain housing affordability…

Urban containment makes it virtually impossible to build the low cost suburban tract housing that made housing more inexpensive and led to higher rates of home ownership in the decades following World War II and the democratization of prosperity…

The far higher house prices, higher cost of living and greater poverty associated with urban containment are an exorbitant price to pay for addressing problems that have been exaggerated or limiting organic urban expansion to serve such a small number of commuters…

There is an emerging understanding that more housing supply is required to address housing affordability. However, in urban containment metropolitan areas, interest in new supply has been largely limited to higher density infill, while leaving the urban periphery restrictions in place (Section 3.9).

This seriously diminishes the potential for improving housing affordability, because the lowest land prices are on the urban periphery and because there is substantial demand for the larger housing, preferred by families, that can only be economically built there. Unless urban fringe restrictions are relaxed enough to restore the competitive market for land, improved housing affordability is unlikely (Figure 15, above).

On the contrary, even with torrid high-density building rates in urban cores, such as Vancouver, Toronto and Sydney, housing affordability has continued to worsen. Even in Inner London prices have continued to increase, despite the greatest urban core densification in the high-income world…

Demographia’s contention that Australia’s rising home prices have been caused primarily by escalating land costs is supported by evidence. The below chart shows aggregate Australian housing values relative to GDP broken down by the land component and the structure component. As you can see, almost all of the growth in Australian housing values (relative to GDP) has been in rising land values:

Meanwhile, CoreLogic-HIA has recorded massive rises in the value of Australian vacant land:

With land costs ballooning at the same time as lot volumes have contracted (suggesting choked supply):

ScreenHunter_15579 Oct. 20 11.23

A key reason for this land price escalation in Australia (as well as in New Zealand, the United Kingdom, and the expensive markets of the United States and Canada) is that the market’s ability to quickly provide low priced new housing supply is being hampered by restrictive land use regulations, many of which have come into effect since the mid-1990s (Sydney has had long-standing limits on housing development on the urban fringe). Demographia describes the key features and consequences of restrictive housing markets as follows:

More Restrictive Land Use Regulation uses urban containment policy or other mechanisms (such as comprehensive plans or development limits) to such an extent that the competitive market for land is not permitted to operate on the urban fringe. More restrictive land use regulation seeks to outlaw the liberal regulation that produced middle-income housing affordability.

Urban containment strategies are the most important of more restrictive land use regulation. Generally, urban containment regulation is “plan-driven,” as planning departments and governments determine where new housing is allowed to be built. There is a “negative presumption,” with new development generally prohibited, except in limited areas where it is permitted by government plans. Typically, urban containment policies include urban containment boundaries and related variations (such as urban growth boundaries, green belts, urban service districts, “growth areas” and other strategies that substantially reduce the amount of land available for house building. Urban containment policy may also be characterized by terms such as “densification policy,” “compact development”, “urban consolidation”, “growth management” “and “smart growth.”

By severely limiting or even prohibiting development on the urban fringe, urban containment eliminates the “supply vent” of urban fringe development, by not allowing the supply of housing to keep up with demand, except at prices elevated well above historic norms.

Urban containment policies are often accompanied by costly development impact fee regimes that disproportionately charge the cost of the necessary infrastructure for growth on new house buyers. There is particular concern about the cost increasing impacts of these fees and levies, especially in Australia, Canada (Canada Mortgage and Housing Corporation), New Zealand (New Zealand Productivity Commission) and California…

By contrast, affordable housing markets, like many cities in the US, utilise open market-based land use structures whereby plentiful new housing supply is able to be built quickly and cheaply on the urban fringe, thereby preventing rapid house price escalation. Demographia describes these markets as follows:

Liberal Land Use Policy (Less Restrictive Markets) applies in markets not classified as having more restrictive land use regulation (where competitive land markets are permitted to operate on the urban fringe). In these markets, residential development is allowed to occur based upon consumer preferences, subject to basic environmental regulation. Generally, liberal land use regulation is “demanddriven” Land is allowed to be developed, except in limited areas, such as parks and environmentally sensitive areas. By allowing development on the urban fringe, liberal land use regulation allows the “supply vent” to operate, which keeps house prices affordable. Less restrictive regulation can also be called traditional or liberal regulation. In addition to lower housing costs relative to incomes, the lower population densities typical of liberal markets are associated with less intense traffic congestion and shorter average work trip journey times. Liberal land use regulation has also been called “traditional” regulation.

So, under an open market-based model (provided there are not also substantial physical barriers to housing supply), increased demand, such as from reduced lending standards and easier availability of credit, quickly leads to the building of additional low priced housing on the urban fringe, which helps keep house prices in check and reduces the likelihood of speculative housing bubbles developing. Further, highly leveraged speculators are less likely to be encouraged into open land markets, since there is little prospect of achieving strong capital gains. Investing in open land markets is, instead, more about rental yield.

I will add that restrictive urban planning structures should not be viewed as a one-way bet for house prices, with unresponsive land supply also more likely to result in higher levels of house price volatility and boom/bust price cycles – a fact also acknowledged by Demographia and evident in Ireland’s boom and bust.

Why? Because strict land-use policies (planning) steepens the supply curve, which makes house prices more sensitive to changes in demand, increasing the likelihood of the housing market experiencing boom/bust price cycles as demand rises/falls.

The full 2018 Demographia Housing Affordability Survey can be downloaded here.

[email protected]


  1. The money shot.

    “…So, under an open market-based model (provided there are not also substantial physical barriers to housing supply), increased demand, such as from reduced lending standards and easier availability of credit, quickly leads to the building of additional low priced housing on the urban fringe, which helps keep house prices in check and reduces the likelihood of speculative housing bubbles developing…”

    Its no surprise that the Australian bubble is right up there when we lead the developed world in packing in lots of fresh sardines into a tin constructed out of:

    Over the top land use restrictions


    A pathological dependence on private bank money creation sprayed with great tax policy driven intensity at a specific existing asset class.

    Stupidity by Design.

    To the extent that there has been some tweaking of the settings (Hi-rise skybox petrol clad human BBQs and Macroprudential) neither is allowed to put the bubble under any real pressure.

    • “…So, under an open market-based model (provided there are not also substantial physical barriers to housing supply), increased demand, such as from reduced lending standards and easier availability of credit, quickly leads to the building of additional low priced housing on the urban fringe, which helps keep house prices in check and reduces the likelihood of speculative housing bubbles developing…”

      it doesn’t. It may help keep “price to the moon” psychology a bit controlled but it doesn’t do more than that. Market (supply/demand curve) forces do not work on natural monopolies such is land. Examples from the past are telling us the same, almost every housing bubble was accompanied with building boom and crashed only when demand side crashed (easy money availability).
      Even in Texas (the most “open market-based model”), out-of-trend price growth in 80s ended because demand crashed (Savings and loan crisis) not because free supply met demand.
      Without “open market-based model” prices would maybe go even further up before they crashed but that’s only marginal. As someone who witnessed it first hand, psychology at the time was crazy and pure bubble madness not much different than early 2000s madness in California (also witnessed first hand).

      I have to be fair and say that none of these madness’s are not nearly as mad as the current Australian one. Would be interesting to see opinion of someone who experienced Irish madness but I have a feeling we are off the chart.

      • Yep the supply argument has always been fallacious.
        If investors wake up one Saturday and decide they are happy with half the yield they were happy with the previous week (say because CG expectations take-off due to mass hysteria) is it really credible to imagine you can have an equal and offsetting and immediate response in supply?
        No, because supply isn’t supposed to and can’t respond to changes in the asset price. In equilibrium the asset price adjusts purely to ensure the total return on housing as an asset offers the same total return as all other assets. Outside equilibrium it goes on a self reinforcing speculative spiral.
        It is supposed to and can respond to the consumer price of housing (market rent) – and has done if you measure housing affordability correctly.
        this counter argument has been made a few times at least, but still the shortage argument seems popular for some reason.
        Despite all evidence; for example the sequencing in bubbles always starts on the demand side. ie. housing gets financialised, the speculative ball starts rolling, supply responds but to no avail.
        Most bubble markets have had hyper development and over-zoning during the bubble phase and it was never the catalyst for the bust. The catalyst was always a collapse in speculative demand.
        Completely the reverse of the shortage story, which says the supply curve shifts or steepens first.
        The shortage argument rests on an almost vertical short-run supply curve and a stock standard downward sloping demand curve (like Beef in Cuba – because the evil government is meddling in the market)
        In reality a *bubble* market may be able to be shown with an upward sloping demand curve, and an upward sloping supply curve which is surprisingly flat. More qty supplied meets more qty demanded and higher prices – meaning dampening speculative demand is the only solution to the bubble.

      • doctorX,

        I am not a fan of picking the one true cause (supply v demand)- that is a pretty arid debate – plus no surprise that I have plenty of sympathies for the demand siders.

        “…it doesn’t. It may help keep “price to the moon” psychology a bit controlled but it doesn’t do more than that…”

        Well that is something quite substantial.

        It is a harder to keep the animal spirits pumping and the demand for credit strong if farm land can be readily converted to low cost subdivisions.

        And hard to spin the necessary narratives.

        And that is an important and often over looked issue.

        The costs of converting land use is as important as to whether you can do it at all. In theory there may be a great willingness to rezone farmland for suburbia but if the taxes and charges associated with it reflects a desire to milk easy credit mortgages as a revenue earner, the cost of doing so may kill demand.

        Add in large immigration programs and you have created an perfect environment for out of control private bank credit creation.

        Always nice to hear Sweeper holding forth on that subject … even if his pudding has too much egg as per usual.

      • How would you prefer the pudding? Flippant “both sides have merits” well padded Gittins style narratives which lead to nowhere, mixed in with casual accusations that the other person is a “developer apologist”?
        I think that would be an inadequate way of looking in to the topic 007. And I admit my attempt is lacking. Because it is a confusing topic. There is barely any agreement on how to model housing as an investment asset especially in a bubble market. But the ‘beef in Cuba’ story of the Shortage team doesn’t pass a prelim theory and evidence check.

      • Sweeper,

        You sound like Skippy a bit more every day.

        Not sure what point you were trying to make but it sounded a bit like you find non binary options too complicated.

        You seem to end up where you started.

        It is all about the excessive private banker created credit.

        I suppose I should encourage you for making further progress in that direction.

        I understand LA Tingle is talking about the 1937 Royal Commission with Uncle Phil on late night live tonight. Radio National.

        Seems to be a topic of some growing interest beyond “money crank” circles.

        Your crystal radio might be able to pick up the transmission if you fiddle with the coat hanger a bit.


    • Yes, a study of the world’s leading house price and social structure mix, a few oligarchs and barrons, lots of peasants to do thair bidding and a one political party fits all setup. Will have to swap the communist thingy for a right wing dictator scenario

  2. reusachtigeMEMBER

    Isn’t Demographia a front for land developers who want to destroy our city farms? Evil sh1t if you ask me. Established housing is best or if you are that type of person who likes new things then apartments are the way to go, and they’re doing very well everywhere!

  3. focusing on land use restrictions still begs the question of why that land even needs to be ‘used’ in the first place. it would not if there was no population growth.

    • That’s it … in the end it all comes back to the massive 3rd world immigration intake, politely referred to as, the Immigration Program

  4. If there is a problem on the supply side, how come rents have not ballooned?
    The only real measure of an actual housing shortage is when rents rise. Not everyone needs to buy but everyone needs to at least rent.

    • The last few decades have seen an epic rise in rents. For example in Sydney, an ordinary house that once rented for 1/3 an ordinary income, now rents for 1/2 an ordinary income. This is a 50% increase in one of life’s greatest expenses.
      For example, in 1985 a typical school teacher would have easily rented (or bought) a 3 bedroom house with yard an easy commute from their school. It was quite normal for teacher to live in the same suburb as their school. If the teacher was male, the wife need not work to pay the bills, although in 1985 plenty of women were working.
      Now the teacher and his wife both need to work, and it is quite common to hear of 40min, 60min or longer commutes being made by teachers.

      The only real measure of an actual housing shortage is when rents rise.

      Track the rent for any given house over a 30-40 year timeframe. Alternatively track the size and location that say 25% of a wage gets you. Both statistics show the epic rise in rents and epic crash in affordability caused by the extremely serious housing shortage (Sydney).

      Note: Shortage-deniers have various tricks up their sleeve.
      One of their favourites is finding that rents have fallen somewhere over the last few months. Perhaps you could find such a statistic?
      Another trick is claiming that renters still tend to pay around 30% of wage on rent and this hasn’t changed much over the decades. This deception works if you ignore quality. As if a dogbox 1hr commute from work is the same as a house with yard 20 minutes from work.

  5. How come Japan is so expensive? They’ve got a shrinking population so it can’t be unmet supply for people to live in.

    • It’s not really that expensive at all, at least by Australian standards.

      The Demographia report only looks at Tokyo and Kyoto as well, both of which have more than 10 million residents. That’s Hong Kong style population density with a price to income ratio about the same as Mackay.

      Mind you the median “house” in Tokyo or Kyoto is a very different thing to the average house in Australia so I’m not sure it’s really apples with apples. Be interesting to know how it works out on a price per square metre basis.

    • The FT ran a story last week about how a fifth of Japan’s territory is “unusable” due to feudal laws and untraceable owners. I’m guessing that much of that unusable land would be out in the sticks rather in the urban areas (can’t crack the FT paywall, the buggers must have tightened up their defences (??)). Also for a country that is stripping massive areas of forest in other countries for paper and chipboard production Japan has been extremely protective of its forests, which cover about 67% of its landmass. If you add just those two percentages together (assuming no intersection) that leaves bugger all space for the 127 million Japanese to live and work other than in densely populated urban areas.