I wrote:
> I'd bet that the CPI isn't calculated correctly, but it makes sense to me to measure the housing cost of owner-occupied housing using rental-equivalents rather than asset prices.<<
Marvin Gandall wrote:
This seems counter-intuitive to me, Jim, and European statisticians have apparently abandoned the practice. Can you explain further? <
I didn't know that the European economists had changed, but the logic of using rental equivalents is that they measure the actual cost of living in a house -- as opposed to the asset price, which represents the (discounted) expected benefits of living in the house for years and years. The asset price also tends to go up steeply (as in recent years) and also fall steeply. More generally, a CPI is supposed to measure the cost of receiving flows of benefits at any one time rather than the cost of buying assets. That doesn't mean that asset inflation (and deflation) isn't relevant. It just says that the CPI shouldn't be saddled with it. Among other things, asset inflation helps solve the "Monetarist conundrum": why is it that CPI-type prices and nominal measures of output have stopped rising in rough step with measures of the money supply? (M2 velocity, the ratio between M2 and nominal GDP fell drastically starting in 1997.) Because the increased money supply goes to buy non-money financial assets, including real estate. -- Jim Devine / "In the Soviet Union, capitalism triumphed over communism. In this country, capitalism triumphed over democracy." -- Fran Lebowitz
