May 18 2006 2:52PM | Permalink | Email this | Comments (0) |
Blog This! using: Blogger.com | LiveJournal |

The Consumer Price Index reports that the cost of owner occupied housing increased 0.4% in April from March, twice the monthly change reported for most of last year, and is now 2.9% higher than a year ago? Don’t believe it. This defies common sense.

Unfortunately, this makes the entire CPI suspect since owner occupied housing is 22% plus of the total index. The CPI is underestimating how much the recent steep rises in housing costs have reduced consumer spending power. That said, this error is at least partially offset by other measurement problems that tend to overestimate inflation. While imperfect, the accuracy of the CPI has been improved in the last ten years by other measurement changes. But it would be nice if they could get housing costs right.

The Bureau of Labor Statistics has been studying how to correctly measure housing costs for fifty years. The issue is not likely to be resolved soon. The current practice is to use changes in rental housing rents to infer changes in how much a homeowner is foregoing by not renting the house to someone else. Theoretically, this seems to be a good approach but it has not worked.

The National Association of Realtors reports that the sale prices of existing homes were 10.3% higher in the first quarter than a year ago and that home prices declined steadily from October through March. These people know a lot about home prices. A summary of the mortgage statistics from Freddie Mac (they know a lot about mortgages) also shows a 10.3% increase in home prices from Q4 2004 to Q4 2005. No report yet for the 1st quarter. And interest payments alone increased 3-4% over the last year with mortgage rates rising 20% and more than 20% of owner occupied homes being sold, refinanced or having their variable interest rate reset.