NCPA - National Center for Policy Analysis

Changing an Old Idea

December 16, 1996

Critics of the proposal to fix the Consumer Price Index (CPI), our basic measure of inflation, have said that it is nothing but a gimmick to cut federal spending. Susan Dentzer of U.S. News & World Report calls it a "sneak statistical attack through the back door" on entitlements. Congressman Pete Stark (D-CA) says "hiding policy changes under the guise of technical corrections is a coward's choice."

While both say that of course the CPI should be fixed if there are legitimate problems, they say it should only be done after long, careful study. The effect would be to continue paying out billions of dollars per year in cost of-living adjustments that are simply unjustified.

Much of the discussion about adjusting the CPI implies that the problems identified by Professor Michael Boskin in his recent report to the Senate Finance Committee have only recently been discovered. In fact, they have been known for many years. Indeed, a 1981 American Enterprise Institute book by Phillip Cagan and Geoffrey Moore identified many of the same problems listed in the Boskin report. Critics also seem to imply that major changes in the CPI are unprecedented. Actually, the Bureau of Labor Statistics (BLS) continuously makes adjustments in the CPI so that it will more accurately measure the rate of inflation.

One of the biggest adjustments in the CPI occurred in 1983. Prior to that time the CPI measured the cost of housing by looking only at housing prices. But when housing prices began to skyrocket in the 1970s, it became apparent that changes in housing prices were not an accurate way of measuring the cost of housing for most people.

For one thing, the CPI implicitly assumed that everyone was buying a new house every month, when in fact few people buy new houses in any given year. Moreover, most people have fixed mortgages, so that their payments remain unchanged. And many people buy houses not just to live in, but as investments. Eventually, the BLS concluded that it was better to measure housing costs on a rental-equivalent basis to more accurately reflect changes in the true cost of housing.

The overstatement in inflation due to the mismeasurement of housing was substantial.

The official CPI figures for those years can be compared to an experimental index, known as CPI-U-X1. The experimental index measured housing costs on the rental-equivalent basis before that method was incorporated into the official CPI in 1983. It shows that in 1980 alone the official CPI overstated inflation by more than two percentage points.

Then, as now, vested interests opposed any change.

Source: Bruce R. Bartlett (Senior Fellow, National Center for Policy Analysis), December 16, 1996.


Browse more articles on Economic Issues