Measuring house prices: a comparison of different indices

Executive summary • Alternative house price indices are giving different pictures of the state of the housing market. For example, some suggest nominal UK prices had surpassed their previous 2007-08 peak by early 2014, while others do not. • In part, these differences reflect the fact that the various indices are trying to measure different things. In particular, there is some ambiguity as to whether they are trying to track the average 'price' of the stock of housing in general or the average 'price' of houses that happen to be sold over a given period. • However, much of the difference between house price indices arises because of the different ways in which the various indices attempt to overcome common problems. • The unique issue in constructing a house price index, compared with most other price indices, is that houses are sold infrequently, and as such we only know the price of houses sold in a given month. This raises two challenges. First, these houses may not be representative of whatever the index is trying to measure-namely, the housing stock or typically transacted properties. Moreover, even if they were representative, a simple average of sale prices might increase from month to month just because more expensive properties are being * The authors gratefully acknowledge funding from the Economic and Social Research Council (ESRC) through the Centre for the Microeconomic Analysis of Public Policy at IFS (grant reference RES-544-28-5001). They thank the various providers of house price indices for giving additional information on the construction of the indices, and in particular David Thorpe at Acadata for also giving very useful comments on an earlier draft. Tetlow are also thanked for commenting on this briefing note. 1 sold-even if the price of each property has remained the same. Many of the differences between indices reflect different ways of addressing these issues. • We describe five house price indices in detail. Although these indices broadly track one another over time, systematic differences between the series persist over quite long periods (say, a decade or more). • In practice, there is not one ideal index of house prices-each has its own strengths and weaknesses and may be informative about different concepts. Inasmuch as policymakers and commentators are interested in the general trend of house prices, they should look at the general trend across all the indices, analogous to a 'poll of polls' in opinion …