RBA mis-diagnoses the US housing bust

By Leith van Onselen

The Reserve Bank of Australia’s (RBA) head of financial stability, Luci Ellis, last night delivered an eye-opening speech to the Federal Reserve Bank of Atlanta 2012 Financial Markets Conference. Ms Ellis lectured the American audience about why the US housing market bubbled and then busted, and explained why the Australian housing market does not face the same risks.

The speech was “eye opening” for a number of reasons.

My main point of contention in Ms Ellis’ speech is her first point, where she argued that the responsive supply in the US was a key cause of their housing bust:

I would identify four key factors as explanations.

The first is that housing supply is quite elastic here, at least in enough parts of the country to matter. The housing boom was a construction boom as well as a price boom. As a result, by 2006 there was already a substantial overhang of excess supply (Graph 1). The inherent stock-flow interaction in the housing market means that construction booms sow the seeds of their own destruction. Prices can undershoot formerly sustainable levels.

Ms Ellis’s contention is not backed by the facts. The US markets with more responsive land/housing supply experienced smaller house price rises in the lead up to the Global Financial Crisis (GFC) and subsequently smaller price declines.

To illustrate, consider the below chart of US housing markets where land/housing supply was restricted by strong regulatory barriers (‘urban containment policies’), and the market was hindered in its ability to supply new homes quickly and cheaply in response to changes in demand:

As you can see, these markets experienced high levels of price volatility and pronounced boom/bust cycles (as well as heavy losses in home equity).

By contrast, consider some key markets operating less restrictive land-use regimes, where land/housing supply is more market orientated and supply is more easily able to respond both quickly and efficiently to changes in demand:

As you can see, in markets where land/housing supply is free to adjust to changes in demand experienced far lower price votality (booms/busts) and, by extention, smaller losses in home equity.

Ms Ellis has the US supply equation the wrong way around. Basic economics tells us that policies that restrict housing supply steepens the supply curve, which makes house prices far more sensitive to changes in demand and increases the likelihood of the housing market experiencing boom/bust price cycles as demand rises/falls. Urban containment policies and unresponsive supply do not necessarily mean that home prices will be supported; rather that price falls will be exagerated if/when demand collapses.

This is a topic that I have written about many times before, for example here and here. It is also a view supported by research. For example, a recent Journal of Urban Economics paper by Haifang Huanga and Yao Tang suggests the link between higher house price volatility and non-responsive housing supply (e.g. due to restrictive land use regulations) has become more pronounced in the US [my emphasis]:

Abstract: “In a sample covering more than 300 cities in the US from January 2000 to July 2009, we find that more restrictive residential land use regulations and geographic land constraints are linked to larger booms and busts in housing prices. The natural and man-made constraints also amplify price responses to the subprime mortgage credit expansion during the decade, leading to greater price increases in the boom and subsequently bigger losses.

And in a detailed examination of studies in this area, Harvard’s Edward Glaeser and Warton’s Joseph Gyourko also concluded that tight land/housing supply leads to greater price booms and busts:

Recent research also indicates that house prices are more volatile, not just higher, in tightly regulated markets …. price bubbles are more likely to form in tightly regulated places, because the inelastic supply conditions that are created in part from strict local land-use regulation are an important factor in supporting ever larger price increases whenever demand is increasing.

…. It is more difficult for house prices to become too disconnected from their fundamental production costs in lightly regulated markets because significant new supply quickly dampens prices, thereby busting any illusions market participants might have about the potential for ever larger price increases.

Finally, Ms Ellis seems to suggest that because Australia’s housing market is more supply restricted, it should not experience the same fate as the US. Again, Ms Ellis is incorrect in her assertion. Although Australia’s housing market is more restricted than the US as a whole – owing to a large number of US states, like Texas, where housing supply is highly responsive and there was no bubble – it is certainly not true when Australia is compared against US states where large booms and busts were experienced.

To illustrate my point, a comparison of Australia’s population growth and housing construction rates against the four key bubble markets of the US – California, Nevada, Florida and Arizona – is provided below and shows that Australia has actually built more homes relative to population growth than these states (see below charts):

The key point of all this is that responsive housing supply is a key mitigant in preventing housing bubbles/busts. Put simply, when land/housing supply is unable to respond quickly to changes in demand, the housing market becomes overly sensitive to demand shocks, resulting in greater price volatility and boom/bust cycles as demand rises/falls.

During an upswing, the extra demand (say through easier access to credit) will automatically feed into higher home prices rather than new construction. In turn, the price rises and perceived scarcity will encourage speculative demand and ‘panic buying’ from first-time buyers, which helps to drive prices up even further. The opposite holds during a downturn, where unresponsive supply will help to accentuate price falls as new housing planned years ago continues to hit the market, and/or reduced demand from credit tightening/higher unemployment causes prices to revert back to mean.

[email protected]