One of the minor obsessions of the new British government is the desire to include housing in the measure of the inflation target given to the Bank of England. In his reply to Mervyn King’s letter explaining why inflation was too high, George Osborne brought the subject up on Tuesday. He wrote:
“As we have discussed, over the longer term I would welcome your views on how we might accelerate the process of including housing costs in the CPI inflation target.”
Of course, housing is included in the CPI already for those who rent property, but not for owner-occupiers. This was followed by the commitment in yesterday’s coalition agreement and programme for government:
“We will work with the Bank of England to investigate how the process of including housing costs in the CPI measure of inflation can be accelerated.”
This ambition is often championed by the Eurosceptic wing of the Conservative Party because it would replace the use of a nasty harmonised European inflation measure as the Bank of England’s inflation target with a measure designed here in Blighty.
The Eurosceptics have a point – not because Britain might ditch something European – but because it is widely recognised that the Harmonised Index of Consumer Prices (the UK CPI) includes rents but nothing for the housing costs of owner occupiers. And that is crackers.
Eurostat is well aware of the shortcoming and is looking into ways of bringing the costs of owner-occupied housing into European CPI, but it is a glacial process and countries have every right to devise their own indices. This is not anti-European. Britain still continues with its retail prices index which has a pretty poor stab at including housing costs. But Britain is not alone in Europe in having cost of living indices which include housing in rather more sophisticated ways. Countries in the euro also have their own indices including Germany, France, the Netherlands and Finland.
But before I get all gushy about the Bank of England targeting an inflation measure including owner-occupied housing costs I think we should recognise that it might make almost no difference to the inflation figures or the operation of monetary policy.
How can that be, given the huge swings in house prices we have seen?
The answer is that most countries that do include housing, do it on a rental equivalence basis – the US CPI is a good example of this. The theoretical justification is that owner-occupiers have done two things: bought a capital asset and rented it to themselves at a zero rent. Changes in capital asset prices do not go into inflation, but the rent should be included, the thinking goes.
And if you go down that route, the effect on inflation is very small if rents and house prices have diverged. In Britain, they have. The following chart shows the inflation rate of rents and house prices – taken from the RPI rent and depreciation components – and you can see that significantly increasing the weight of rent (even private sector rents) will make little difference to the overall inflation figures, because its relationship to house prices has been very limited.
The upshot is that while George Osborne and the new government is making quite a bit of fuss about the merits of including housing in the inflation target, the reality might well have a huge “what’s the point” element, particularly if we ditch European practices and follow the US.
