Friday, 30 January 2015

House price indices vs average house prices: Another little mystery solved.

What has always puzzled us is how HM Land Registry can confirm on the one hand that the average price paid for a home in the UK in December 2013 was £250,000, but on the other hand, the headline figure in their January 2014 report is 'only' £177,766 (pdf).

(If you download their price paid data for 2014 and take a simple, mathematical average of all recorded selling prices for England & Wales in that year, it is in fact £259,000).

For a very detailed explanation of how the various indices and averages are calculated, see this fine summary by Acadata.

To cut a long story short, the ones called 'Index' (Nationwide, Halifax and HM Land Registry) are just that; they have a hypothetical starting point adjusted up or down for observed price changes.

Or as HM Land Reg put it so succinctly:

1.5 Calculating the standard average price

The standard average house price is calculated by taking the average (geometric mean) price in April 2000 and then recalculating it in accordance with the index change back to 1995 and forward to the present day.


The key word here is 'geometric mean' (which is two words). This is always lower than the mathematical average because of the way it is calculated; and the wider the gap between values at the top and the bottom, the bigger the difference between geometric mean and mathematical average.

So while the geometric mean is a useful indicator, falling some way between the median and the mathematical average, it is completely useless if you are trying to work out the total value of all UK housing. For that you just need the mathematical average.

Hence and why, Savill's recent estimate of the total value of all UK housing of £5,750 billion (average £205,500 x 28 million units) looks perfectly reasonable, even though their estimate of the average value of a unit of social housing (£79,252) looks on the low side to me.

(All of which backs up my assertion that if you wanted to raise £200 billion a year in LVT, that averages out at 3.5% on the potential 2014 selling price of each home.)

3 comments:

James Higham said...

I did read it, Mark but it did my head in. How can you cope with figures like that?

Random said...

Guardian supporting nonsense that is HB:
http://www.theguardian.com/housing-network/2015/jan/31/reduced-benefit-cap-force-large-families-out-south-east#comment-46899911

Mark Wadsworth said...

JH, force of habit, practice etc.

R, splendid. HB is negative LVT, so I support LVT and oppose HB. Far better to get rid of HB and have higher basic welfare payments.

We get the Poor Widow argument in all its guises whether we propose LVT or withdrawing HB.