11 June 2015

LSE pilot study confirms that increasing new housebuilding does not drive down house prices


In two recent articles I’ve argued that the increase in UK house prices from 2000 to 2007 owed much more to strong economic conditions and the easy availability of mortgage credit than to any national or regional shortage of houses.

An interesting recent pilot study see by the Financial Times’ Kate Allen (@_Kate_Allen), led by Professor Christine Whitehead of the London School of Economics, casts some further light on the short-term impact of new housebuilding on house prices. Analysing eight new developments of nearly 300 homes apiece, built over the past five years — where in each case the new build was large compared with the existing stock of homes (so impacts were not trivial because of local scale) — the study found that in no case did local house prices fall once construction was complete. Increased new house building simply had no effect on house prices, or, in a few cases, was associated with a rise in local house prices.

That may seem like a highly counter-intuitive finding. But it shouldn’t be. Increased housebuilding will often be associated with an increase in the wealth of an area, so an increase in the willingness to pay of housebuyers and a reduction in the willingness to sell of owners of second-hand homes. Thinking more broadly, it will only very rarely be the case that housing in any one area could not be substituted by housing elsewhere — e.g. somewhere else approximately equi-distant from the place a family will commute to work. In those other areas, there will almost always be some available stock of second-hand homes available. The amount of new housebuilding will, over any short-term timescale, be largely irrelevant to the total available of houses — most of what “supply” means in the housing market is not new housing but, instead, the supply of second-hand homes by house-sellers.

There is a useful analogy to be drawn here with the prices of government bonds. Government bond prices go up or down all the time, and sometimes by quite large amounts. That is almost never a matter of how much new supply there is — i.e. of how many new bonds the government is issuing. To have an impact through sheer new issuance, governments must issue absolutely gargantuan quantities over a long period, having lost all control of borrowing. Almost always, government bond prices are driven by the willingness to buy or sell bonds of owners of second-hand bonds in bond markets, driven by factors such as their view of economic conditions.

That is how markets for second-hand assets — like bonds or houses — work.

Andrew Lilico is Chairman of Europe Economics.