Housing supply determines price volatility

ScreenHunter_01 Feb. 17 18.55

By Leith van Onselen

For years I have argued that markets where land/housing supply is unresponsive (inelastic) – via planning constraints or geographical barriers – are far more prone to suffer from more expensive housing, higher house price volatility, and bigger boom and bust cycles than markets where land/housing supply is relatively responsive (elastic) to changes in price.

These dynamics were explained in detail in last year’s presentation to a mortgage risk roundtable in Melbourne (available for download here).

Last night, the Wall Street Journal published a great article explaining how differences in the elasticity of housing supply across the US housing market has affected housing market outcomes, with bigger booms and busts experienced in supply-restricted markets (e.g. California, Florida, Arizona, and Nevada), and the more risky profile of home builders working in those markets:

The crucial difference between housing markets in places such as San Jose, Calif., and Austin, Texas—more important than considerations like quality of life—is the availability of land…

Each metro area [has been given] an “elasticity score”—the higher the number, the easier it is to put up a home. San Jose scored 0.76, ranking it the 10th most constrained among the 100 largest areas by population. Austin, scoring 3.0, ranked a decidedly easier 88th. The range for all 100 was 0.6 to 5.5… the scores correlate with other constraints, such as zoning and taxes…

ScreenHunter_1433 Feb. 28 07.18

Tapping this data, economists Atif Mian at Princeton University and Amir Sufi at the University of Chicago’s Booth School of Business have shown that more constrained areas saw bigger booms in the housing bubble—but also bigger busts on the way down.

While house prices in constrained areas have bounced around more than in less-restricted areas, they also have done better over the long haul. Consider two gauges of home prices, one made up of the 10 most elastic areas in 20 separate S&P/Case-Shiller metropolitan home-price indexes, the other of the 10 least-flexible areas. The latter is more volatile but up 65% since January 2000. The former is up 39%…

Constrained builders, conversely, must navigate markets where securing land can be tricky and there can be sudden price declines. But when prices rise, they can do very well, making them more suited for investors with a broadly bullish view on U.S. property…

At the other end of the spectrum… [unconstrained builders]… should offer investors relative insulation if house prices crack. In periods of growth, they should generate steady profits as they buy land, build on it and sell. Their biggest risk may be that they operate in markets where barriers to entry are lower.

The above article summarises why it is so important to liberalise the supply-side of the housing market. Not only does it improve overall housing affordability, but it also reduces price volatility and the potential for boom/bust asset prices as demand changes.

Housing markets where strict regulatory barriers are in place – such as urban growth boundaries, restrictive planning/zoning requirements, minimum lot sizes, and upfront development taxes – are incapable of quickly and efficiently supplying low-cost housing. These supply constraints thereby ensure that increases (decreases) in housing demand feed primarily into higher (lower) prices instead of changes in new construction. The perceived land/housing shortages and rising prices during the upswing also encourages speculative demand and ‘panic buying’ from first-time buyers, which assists in driving home prices up even further. However, when the economy and sentiment sour, such as in the wake of the financial crisis in the US, the slump in housing demand can cause prices to collapse.

By comparison, in housing markets with lighter-touch land-use regulations and planning, low-cost housing is more able to be built quickly and efficiently in response to rising demand. The rapid supply response thereby prevents prices from rising dramatically, which also reduces speculative intent, since there is little prospect of achieving strong capital gains. And with house prices remaining relatively stable, ‘panic buying’ from first home buyers is generally less prevalent. Affordability is also generally much better.

[email protected]


12 Responses to “ “Housing supply determines price volatility”

  1. melbourneguy says:

    Very cool UE and ……

    Egg on my face. I remember challenging this supply restrictions -> elasticity -> price volatility issue (to be clear I was challenging the volatility argument not price which is obvious) with a few of the commenters here and clearly the empirical analysis shows that it is true.

    However, the question is (and a bit of an academic one) is this best explained by a neo-classical framework? Or … it would it better explained through a behavioural argument.

    Ask any average punter that religiously believes that property investments is the undeniable guaranteed path to riches and one of their mantras (along with the old price doubles every 7 years or whatever) is that if you buy in an area where there is restricted supply you can NEVER LOSE.

    As in it is always a good time to buy IRRESPECTIVE of current price levels…… you see, it’s all “demand and supply”…. i.e. doesn’t matter if you’re paying $5m for a studio.

    That sort of flawed/irrational thinking clearly encourages speculation/bubbles/booms and busts.

  2. Explorer says:

    So rational property owners in low constraint states and cities who have low mobility because of the cost of housing in high constraint cities and states will campaign to turn their cities and states into high constraint jurisdictions. They will do this in spite of it making it more difficult for existing non-owners and future generations to buy a house.

    Conversely, non-owners in high constraint jurisdictions will campaign for those jurisdictions to become low constraint jurisdictions no matter how much harm it would cause to some existing owners

  3. speculator says:

    This map is either wrong or it doesn’t prove the point. Few examples:

    - Phoenix (huge bubble during 2000s) and Dallas (no bubble during 2000s) have the same supply elasticity

    - Santa Anna with huge bubble during 2000s (200% up than 66% down) has much better elasticity than Pittsburgh where prices changed with inflation during 2000s

    - very similar cities: Austin (high elasticity) had mini bubble while Houston (lower elasticity) had none.

    • Nice obfuscation of the debate. The map might very well be wrong on Phoenix, because it had similar land-use constraints to Vegas (i.e. most of the developable land was either owned by government or Indian reservations, and kept off limits to development).

      Obviously, supply elasticity isn’t a perfect gauge of price volatility, as demand-side factors matter too (e.g. Austin experienced more rapid population growth than, say Pittsburgh). But the general rule holds, despite your protest.

      Loads of studies that include the supply-side (rather than focusing on demand-side factors only) have concluded the same. Maybe it’s time you took the blinkers off – house prices are set by both supply and demand.

      • speculator says:

        ” house prices are set by both supply and demand.”

        I agree with this, this is truth when speculative demand is not significant portion of total demand.

        But, I want to point out that speculative demand alone can be responsible for house prices bubble even when supply or real demand (demand for homes to live in) is not an issue (just take a look into some of the declining cities in UK or USA north east).

        This is especially true in cities where certain geographically areas are desirable in eyes of public (city centre or coastal areas). From Manhattan home price perspective, it doesn’t matter how easy and cheap is to built home in a suburb on Long Island just 30km further away, or even how easy and cheap is to built new unit blocks on Manhattan itself. People pay more because they think they can make big buck and because they can borrow more.

        Just look at Melbourne inner city unit market. Supply is huge, many units vacant; it’s not hard to get approval for new units, … still prices are rising fast.

        So, I agree with you, that supply is important under normal market conditions when price is set between real demand and supply. But supply almost doesn’t matter when people start behave irrationally supported by banks willing to lend as much as they want.

        Market is not perfect, especially in case on natural monopolies such land.

      • PhilBest says:


        Yes, highly distinctive locations like Manhattan can be expected to have their own dynamic.

        Yet, note that Cheshire and Hilber found in their famous study of office rents, that almost every city in the UK had higher CBD office rents than Manhattan. This is what anti-growth planning can do to the cost of space in even quite small and unimpressive cities.

        Even the Manhattan land market likely has a “vent” in the space just across the Hudson River, which also has a vent in the space another few miles inland; which also has a vent, and so on ad infinitum. The fact that Manhattan and the New York urban area has this feature but the UK’s cities do not, is certainly the reason for the absurd cost of rents in the UK’s cities.

        In fact urban FRINGE land in most US cities, as soon as development permission is secured, is inflated over and above true rural values by a factor of some 400 to 700. You will find locations around the New York urban area where rural land is being converted to urban use at a price inflation factor of orders of magnitude less than this; this makes a systemic difference.

        And New York is still more inelastic in its supply than most US cities.

        I suggest Aussie cities are probably closer to the UK’s disastrous example than to New York. I mean, using New York as a model to explain central Melbourne’s RE market behaviour??? Come ON – this is Aussie try-hard syndrome, not reality. Try comparing it to the UK’s cities – (not London, either) instead and it makes a lot more sense.

    • PhilBest says:

      The problem with that map is this. The “supply elasticity” figures are taken from a study by Albert Saiz which only included data up to 2006.

      If this study was re-done, it would most certainly provide elasticity figures that are more consistent with the local housing market outcomes SINCE then, that the WSJ is talking about.

      Phoenix became “supply constrained” when developers securing sites based on growth projections came up against government land holdings that were not for sale. Any new study will have to show Phoenix elasticity much lower today than what it was in the period leading up to 2006.

    • PhilBest says:

      It is so ridiculous what basic essential factors can be missed and missed and missed by academics, one wonders whether there is willful obfuscation going on.

      Leith Van O leaves all these people for dead.


      “……despite there being ample developable land within 40 miles of the Las Vegas city centre to accommodate this population growth, the actual quantity of land available for development was heavily restricted by the federal government Bureau of Land Management (BLM), which owns around 90% of the land in Clark County, which contains the entire Las Vegas urban area. The BLM heavily restricted sales of land to the market in an effort to maximise revenues, causing builders and developers to bid-up land price in period auctions to ensure their supply of land for construction (called ‘land banking’)…..

      “…….An article published in late 2006 in USA Today explains the supply situation in Las Vegas:
      The valley that Las Vegas and 1.8 million residents call home is nearly built out. Mountains, national parks, military bases, an Indian community and a critter called the desert tortoise have Sin City hemmed in. At the current building pace in the USA’s fastest-growing major metro area, available acreage will be gone in less than a decade, developers and real estate analysts say……Developers who 15 years ago paid less than $40,000 an acre are paying more than $300,000 today. In an auction of public land that went on the market last year, a developer paid $639 million for 2,655 acres…..”

  4. md says:

    UE, while I understand your call for more supply to cope with demand, what about about focusing more on the immigration policy and the foreign investment which is creating the demand?

    I know you talk about these things every day, and you are a great ambassador for affordable housing, and highlighting all the factors creating unaffordable housing, but it seems to me that urban sprawl is just coping with the issue (but always behind). We all know the “Big Australia” policy is there to provide more “greater fools” to buy into the Ponzi scheme, and surely there needs to be some debate at government levels as to how many people we actually want.

    Having said that, keep up the good work, UE – we are all hoping people with a voice like yours will make a difference, and most, if not all of us at MB are right behind you.

    • I have written loads of articles protesting against the Australian population ponzi. But I can’t see the government changing course, so we need to accommodate the growing population.

    • PhilBest says:

      Bear in mind that it is possible, with sufficiently rigid growth containment urban planning, to have very expensive urban land rent and unaffordable housing in spite of small average house size, even with a shrinking population. Welcome to Liverpool, UK.

      With sufficient freedom of urban growth, it is possible for a city to grow its population 20% in one decade while house prices remained affordable and the average house size grew. Welcome to Houston, TX.