I was disappointed with The Economist’s recent special reports on housing. Whilst the reports captured the psychology and the demand-drivers of bubbles well (discussed in an earlier post), they failed to adequately capture the role of supply-side constraints. As far as I could see, the only reference to the supply-side of the housing market were the below paragraphs from the article When the roof fell in:
Work on the relationship between housing supply and bubbles by Edward Glaeser of Harvard University and Joseph Gyourko and Alberto Saiz of the Wharton School suggests that places with relatively elastic supply have fewer bubbles, of shorter duration, than those where the supply is more restricted.
In many respects the recent boom bears this out. Differences in supply constraints can explain much of the striking disparity between American states, from the modest run-up in prices in Texas, where land is easily available, to the huge surge in places like Nevada, where land-use regulations are tighter.
But elasticity is not always a good thing. When the housing market can respond to demand by adding to supply, there is a greater risk of overbuilding. In theory, booms in elastic markets do not last for long because as new housing becomes available it puts pressure on prices, puncturing expectations of further appreciation and popping the bubble. For the 1996-2006 cycle in America Messrs Glaeser, Gyourko and Saiz find that places with more developable land did have shorter booms.
The Economist’s first three paragraphs on housing supply are largely correct. However, in relation to comment that “elasticity is not always a good thing… there is a greater risk of overbuilding”, I would argue that a few thousand empty homes is a small price to pay for avoiding a damaging housing bubble/bust whereby an entire housing market gets drastically over-inflated and then crashes back down to earth.
Consider Atlanta, Georgia. It built too many homes but it’s Median Multiple (median house price divided by median household income) reached a peak of only 3.0 during the recent US housing bubble, before dropping back to 2.3 in 2010. Now compare Atlanta to the supply-restricted markets of Phoenix, Arizona and Las Vegas, Nevada. Their Median Multiples reached 4.7 and 5.9 respectively, before crashing back down to 2.7 and 2.6 respectively in 2010.
In my opinion, the losses of huge amounts of housing equity in the supply-restricted states are far more damaging than the building of excess homes in supply-responsive states with only small losses in overall housing equity.
After making some worthy observations on the relationship between housing supply and bubbles, The Economist then undoes its good work by citing the experience of Spain and Ireland – two markets that both overbuilt and experienced painful housing bubbles/busts – and use these to discredit the supply-side argument:
But when it comes to the biggest house-price bubble in history, theory does not get you very far. In some places the boom was big enough and irrational enough to suppress price signals from lots of new supply. Instead, availability of land simply fed speculative activity, which has made the popping of the bubble much more painful.
In a report last December the Bank of Spain reckoned that the country has a glut of 700,000-1.1m unsold homes, which will continue to weigh on prices this year. Bernstein Research estimates that these unsold houses will take four to five years to clear, and even that may be too optimistic given high unemployment, the threat of a sovereign-debt crisis and fewer immigrants. It could have been worse: Spanish banks have repossessed huge amounts of land that had not yet been built on, and residential-mortgage standards are rather conservative. But the oversupply means that prices will keep falling. They have dropped by only 16% from their peak in real terms, and Bernstein reckons the eventual fall will be more like 30%.
Ireland’s building boom also went over the top. The government relaxed planning laws, the banks threw money at anything involving cement, and investors gobbled up houses in the expectation that prices could only go up. An Irish government report last October into the country’s “ghost estates” identified more than 2,800 housing developments where construction had been started but not completed. Between them these estates had planning permission for 180,000 units, which roughly translates into a new residence for one in every 25 Irish people.
The Economist couldn’t have chosen two worse examples to discredit the argument that more responsive (‘elastic’) supply helps mitigate the formation of housing bubbles. Had the Economist undertaken a little research, it would have discovered that both Spain and Ireland operated highly restrictive land-use regimes that prevented their markets from quickly responding to increasing levels of demand (from easier credit, population growth and rising incomes) with construction of low-cost/good quality housing where people actually wanted to live. Instead, low quality/high cost ‘dormatory-style’ housing was built in undesirable locations.
A recent paper from the Spanish Institute of Economics explains how the planning system of Spain is highly restrictive and subject to high degrees of political interference and corruption:
Land use regulations in Spain adhere to an extremely interventionist and highly rigid system. A key characteristic is that, although an individual might own the land, the government is empowered to control and implement all processes of urban development. Landowners are not permitted to develop their land without the prior agreement of the local administration…
Municipalities draw up a ‘General Plan’, which provides a three-way land classification: built-up land, developable land (the areas of the community where future development is allowed), and non-developable land (the rest of the territory – agrarian and other uses, where the development process is strictly prohibited, at least until a new plan is accepted). The existence of a ‘development border’ [i.e. urban growth boundary], a line between plots of land on which developers are allowed to build and plots where development is banned, is a key feature of Spain’s land regulation system. In periods of high demand this border creates a rent differential which might fuel the bribes developers are willing to pay to local politicians in exchange for shifting this border to their advantage…
A further feature of this border is that once it has been fixed it is very difficult to remove. The reason for this is that the establishment of the border creates certain rights for the landowners, and the only way to change these would be for the government to acquire the land…
In short, the system of land use regulation in Spain is characterized by its high degree of interventionism and its extremely discretionary nature… The profits of land developers, therefore, depend on many largely discretionary political decisions – the primary one being the fixing of the ‘development border’, the line separating land on which development is allowed from that where it is prohibited – and this provides strong incentives for this lobby to offer bribes to politicians.
And the effect of Spain’s restrictive urban planning structure was summarised as follows in the RICS European Housing Review:
Municipalities where there are few competing development localities have an incentive to restrict land supply in order to encourage increases in local land and property prices and, so, maximise their tax take. There have been complaints that some municipalities, especially in the areas of greatest shortage around the major cities, were adept at this activity in recent years and held back land supply. One consequence was the leapfrogging of development to more distant locations and resultant long commuting times.
It was a similar situation in Ireland where the Government granted planning permits too late in response to rising demand, resulting in the building of large numbers of standardised, small, poor quality homes in satellite locations far away from the major cities. A 2005 paper by the UK’s Policy Exchange explained the Irish supply situation well:
According to Dr Stevenson of UCD Dublin, development for new housing actually took off much too late. “In the early years of the boom, we did not see much building in Ireland”… Because supply was late to meet demand, by the time construction activity actually took off it was too late to deal with the backlog in a reasonable way. All that planners and developers could do was try to satisfy the huge pent-up demand quickly. The result was a quick fix, not a thoroughly reasoned solution…
First, large numbers of flats – something the Irish were not used to – went up, in the form of large apartment-blocks. Second, whole new housing colonies were built, often consisting of hundreds of virtually identical semi-detached or terraced houses lacking any individual character. However, these were often far away from existing amenities, hardly provided any amenities themselves, were poorly built, and served as dormitory towns for existing cities (mainly Dublin). They were also much smaller than comparable houses built twenty or thirty years ago…
To sum up, in the words of Dr Brendan Williams: “The quantity of our supply is very, very good. The quality leaves a lot to be desired”…
Much of the hasty development that Ireland has seen over the past few years could have been avoided if supply had reacted earlier and in a more flexible way to rising demand. But in the early years of the boom, central government had still not realised how important it was to encourage housing development, and local councils, left alone without an independent tax base and lacking the incentives as well as the means to engage in proactive planning, did not plan for enough new houses.
Then, when it was almost too late and prices had already skyrocketed, the government realised the dangers of this situation and encouraged building. But, as Ronan O’Driscoll put it, “the government only thinks in numbers and units” – thus failing to understand that good, flexible and strong supply means more than just “throwing in a few hundred two-storey, three bedrooms semis” (Liam O’Donnell). And this “thinking in numbers” was passed down from central government to local planners, who could basically declare that they had done their jobs properly if they could only show that their numbers had gone up. The quality of these developments does not appear in the statistics and is hard to quantify anyway.
And the outcome of Ireland’s restrictive urban planning regime was summarised by Michael Warby in the March issue of Quadrant:
In Ireland, local government areas with the highest vacancy rates were the most liberal in releasing land, while the areas with the lowest vacancy rates were the most restrictive. This led to the worst of both worlds, as housing estates were built where they were permitted, rather than where the demand was, while quantity controls in the more desirable areas still drove prices up. The post-bust result is empty estates of unwanted housing.
Market stock versus bubble stock:
The Economist appears to have made the classic economist’s mistake of focussing only on the quantity of homes built relative to the population, and ignored the cost and responsiveness of supply, as well as qualitative factors such as the quality, size and location of the new housing. In doing so, it has assumed that the supply-side responses in Spain and Ireland were adequate, and wrongly suggested that responsive (‘elastic’) supply is not particularly effective in preventing housing bubbles.
It’s important at this juncture to distinguish between ‘market stock’ and ‘bubble stock’. As explained by Hugh Pavletich, co-author of the Annual Demographia Housing Affordability Surveys, new ‘market stock’ is housing construction of: (a) the type the market wants; (b) where it wants it; and (c) within sound Development Ratios (e.g. 20% land cost and 80% construction cost). By contrast, new ‘bubble stock’ is: (a) where people don’t want it but the planners force it to be (e.g. high rise poky apartments within the urban boundaries or poorly located homes in satellite cities outside of the urban boundaries); (b) of the type real buyers – not speculators – don’t want; and (c) outside sound and normal Development Ratios (e.g. 50% land cost and 50% construction cost).
Texas and Georgia, which have consistently provided family homes for $150,000 on the urban fringe of their major cities (comprising $30,000 land cost; $120,000 construction cost) are clear examples of ‘market stock’. By contrast, the expensive poorly-located construction in Ireland, Spain, Miami, etc are examples of ‘bubble stock’.
The project times for high rise ‘bubble stock’ can also take many years, which impedes the ability of supply to respond adequately to demand. By contrast, normal ‘market stock’ in the form of single family homes is typically much quicker and more responsive, provided there is a ready supply of serviced lots on the fringes (not the case in Australia).
Little wonder then that due to the long lead times inherent in high rise ‘bubble stock’, much of the new supply was completed just as the global housing bubble burst. Miami was a classic example of this. The Gold Coast in Australia looks to be playing out the same way.
Ultimately, the question of whether there is adequate housing supply has little to do with the physical number of homes built relative to population growth. Rather, what is important is the relative gap between new home prices and the cost of construction. Whenever you have a major gap between the two – as exists in Australia where small land lots typically cost around $200,000 – there must be regulatory or physical barriers inflating land values and driving a wedge in the market.