Jim D wrote, >at some point, economists decided on a conventional definition of inflation >as referring only to increasing prices of newly-produced goods and services. >Given that convention, inflation in housing prices only counts when it >affects apartment rents (or "imputed" rent on owner-occupied housing) and >other expenses of using housing, rather than the hike in the price of housing >as an asset.
Correction: The part about "newly-produced goods and services" does apply to GDP but does _not_ apply to CPI. I know Jim knows this but he was just typing faster than he should have. The difference between consumption and investment is what is critical. According to the BLS, the purchase of a home (new or used) is the purchase both shelter and of a capital asset that might lead to a capital gain in the future. As BY DEFINITION the CPI is concerned ONLY with the cost of buying current consumption, the capital asset component of housing is eliminated from the calculation of the CPI. But while this makes sense for a definitional point of view, it does not necessarily make sense from the point of view of someone who needs to pay a big portion of this monthly income to fund their buying of a capital asset. This squeezes their ability to fund current consumption of items in the CPI market basket. Eric
