Indicating Housing

by Ken Gibb

The Economist had a snappy and eye-catching piece on international housing markets today on Twitter. Essentially, the story is based around looking at national housing market performance and linking these to two key indicators: the deviation (and direction of deviation) from long run equilibrium price to income and price to rent ratios. In other words, does country X have a house price to income ratio above or below its long run position and by how much; similarly, is the house price to rent ratio above or below its long run position and to what extent? It then becomes possible to build an economic narrative about market strength, overheating, or, continuing weakness, etc.  The Economist highlighted the notion of the contrast between a common surge in housing markets before 2008-09 but an uneven recovery now. It also singled out Canada as having likely future problems (as well as Eurozone members: Spain, Italy and France).

I found this analysis  certainly interesting, but one that raises more questions than it answers. First of all, the use of price to income ratios is a commonplace in housing economics but there has undoubtedly being growing reluctance to talk about a standard or long run relationship, both because of volatile and disequilibrium features of recent housing market experience, but also because of a much greater interest in the distributional nature of the housing market. In mass home ownership societies, it is how these ratios operate at different points on the income scale, how housing wealth relates to income and life cycle, that are of increasing interest and importance. It also cannot be sensibly detached from mortgage lending and that linked or joint market.

A second issue is that the price to income ratio and in particular the concept of the equilibrium ratio is perhaps an elusive one. Economists may be right to attach importance to the fundamentals and the way market forces embodied in the elements of ratios such as price to income drag values back to a stable long run position – but the more complex the market and the sector, the less recognizable may be the transmission, the path and the end point. It is all the more likely that at intermediate stages other forces will drive us off course needing other corrections in any case. One such force is the mortgage market – there is a strong case for indicators around credit tightness, mortgage lending and market transactions. If, as is surely right, we think the housing market makes large monetary and real economic contributions, these value/volume indicators of scale are critical to our macro understanding.

A third point is of course that looking at countries’ ratios (i.e. aggregate analysis) is not perhaps as meaningful as looking at better defined housing market systems, often at regional or metropolitan levels. Contrasting Greater London with Paris or Manhattan or Tokyo, or indeed with other regions in the UK would be more useful and comparable.

Fourth, the price to rent ratio is a relatively recent innovation that reflects a broader financialization of housing. The indicator is supposed to create a variable similar to a stock market price to earnings ratio on the basis that house prices should have a long-term multiplicative relationship with market or imputed rents: too big and it is over-priced and heading for a correction. I can see that this plays to the Economist’s readership and indeed there are contemporary housing economics papers using this sort of concept. It may well be a useful indicator but it rest on key assumptions. These include that we have securely identified the long-term equilibrium price: rents number (or has it been calculated from trend data only?). We also must assume that there is a well-defined equilibrium relationship between owner housing markets and renter housing markets that holds in the same way over time across different countries. I have my doubts that this is true. The UK has a growing but immature rental market, which has a counter-cyclical relationship with home ownership and a different, competitive relationship with social housing. That is quiet different from the inter-tenure relationships that exist in Germany, the Netherlands, the USA and Australia, each of which arguably have different, distinctive relationships between owning and renting. The indicator used may show the deviation from a trend or long run ratio but the base may mean quite different things.

We clearly cannot expect a universal battery of well worked through and updated local housing models that can be calibrated, contrasted and policies consequently based on robust findings. Indicators are not models but  have a place and I agree they should be combined to provide economic narratives but they have to be at the right scale and require sufficient indicators (though still only a parsimonious number in total) to capture the essence of the market. I had a small role in the 1990s World Bank housing indicators programme which was a fascinating attempt to do just this – working at a city/region scale to try to first conceptualise and then apply a small set of clustered indicators that complemented each other and for which the sum told a useful story. Later on (2008-11) working with colleagues at the Universities of Manchester and Ulster, we developed a housing and neighbourhood monitor and a user-friendly GIS interface that allowed analysis of local housing markets for the Joseph Rowntree Foundation, very much in the same spirit  as the earlier World Bank approach (http://www.jrf.org.uk/publications/housing-neighbourhoods-monitor).

I would suggest, finally, that there are other indicators that need to be included in this small bundle of essential measures of the housing market. These would include those in the Economist article, suitably caveated, the aforementioned indicators of mortgage credit tightness (and this would be a macro variable), as well as a standardized measure of transactions volumes. In addition, countries like the UK must have a measure of regional long run supply elasticity, since this is at the heart of our housing market volatility problems. Steve Malpezzi at Madison, Wisconsin has also suggested constructing an index of regulatory tightness for land release and planning for housing – a difficult and contentious task but something worth thinking about further for the UK?

Advertisements