Inside the CPI

Thanks to Rdan for the welcome.

I’m with Barry Ritholtz in admiring the NYT’s interactive consumer price inflation mosaic, which the print edition didn’t do justice to. As someone who occasionally gets a bee in the bonnet over the successes and failures of economic measurement methods, the graphic also helps illustrate a couple ways in which CPI can be unfairly maligned. As I read it, the graphic deals best with one class of arguments for why CPI is screwed up:

ZOMG the price of [CPI component in the news] is skyrocketing! Lo, prices of goods that headlines would lead you to believe should be measured as skyrocketing by-and-large are — +26% for gasoline, +48% for heating oil, etc. Gasoline is even a big enough component of average expenditures for the price increases to be legitimately painful for anyone without good substitution opportunities. But it’s still only so much of a representative budget; cheese, bread, and eggs are much less so. And ample sums are spent on other things where there isn’t much or even any inflation at all. (How these all should be treated for any given policy purpose is a separate issue.) Meanwhile, the gorilla in the room is housing, as housing rentals and equivalents account for 30% of expenditures.

However, that leads to a second bad argument:

ZOMG the biggest component of CPI is imputed! Indeed, Brad DeLong recently observed Michael Mandel suggesting that personal consumption expenditures be calculated ex-imputations — of which owner-occupied housing is the biggest. That wasn’t the sharpest thing that’ll appear in BusinessWeek this year. Catch is, it’s not like homeowners don’t (a) consume housing services and (b) incur opportunity costs doing so whether or not they (like my family among other leveraged real estate investment partnerships) actually shell out a lot of cash for house payments. We consume non-marketed goods and services all the time. Missing markets in almost everything, I like to say.

The details of the imputation of housing expenditures for owners certainly aren’t sacred. For one thing, in much of the country, the rental and owner-occupied housing stocks resemble each other very little. For another, I’d like to hear why it would be unreasonable to calculate the imputed rent at a ‘rational’ rate given what the owners paid, which goes to the issue of whether housing expenditures above imputed market rents should be treated as consumption or investment. Anyway, between vacancies and the supply of foreclosed properties, I doubt shelter inflation will be a big problem in the near future.

So, I don’t go to bed worrying about inflation, or that we’ve been in an especially protracted recession for years and apparently don’t know it (an implication of claims you can find out there that CPI and like measures understate “true” inflation by very large amounts). Employment and the distribution of such growth as we have had is another matter.