Wednesday, January 16, 2008

The CPI said what?

This morning, the Department of Labor reported that consumer prices rose 4.1 percent last year, the largest December-to-December increase since 1990. The consumer price numbers were in line with producer price figures that were reported earlier this month. Inflation was moderately high even after food and energy prices are taken out of the mix; the core rate rose 2.4 percent for the year.

One surprising component of the consumer price figures was that they showed shelter costs increasing at 3 percent for the year. The shelter figures are based mainly on changes in actual rents, which increased 4 percent last year, and imputed rents for owner-occupied housing, which were estimated to increase by 2.8 percent.

For nearly 25 years, the shelter component of the consumer price numbers has been based on this imputed rent figure rather than direct measurements of mortgage, house price, and property tax figures. The rent figure is intended to get at the consumption component of housing and to abstract from the investment element. However, there are times when rents diverge from these other fundamentals, leading to questionable results.

Shelter costs account for nearly one-third of the index. Raise your hands if you really believe that prices for owner-occupied houses increased last year. If we instead assume that there is a mistake in the CPI formula and that the change in shelter costs was essentially zero, the inflation rate drops by a full point down to 3.1.


Jeffrey Sykes said...

Dave: Good post. I'd love to see more of these posts on how these indexes are calculated.

Is there a way (or a place) to learn what factors are contributing to inflation and what impact items such as the national debt have on inflation?

Dave Ribar said...

The CPI is calculated and published by the Bureau of Labor Statistics (BLS). The BLS has web-pages dedicated to each of its data programs; the one for the CPI is

Near the bottom of the page, you will find links to detailed monthly reports and the methodology.

The determinants of inflation are complicated. Within a domestic economy, the short answer is that the causes are either creating too much money (expanding money supply) or creating too much demand for goods, such as by expanding the deficit. An introductory macroeconomics textbook would cover this. There is substantial disagreement though regarding how much weight to attribute to different factors.