Bias in the Consumer Price Index: What Is the Evidence?

Abstract

Recent research has suggested that the upward bias of the U.S. consumer price index may be significant, and correcting the biases would have important long-run effects on the federal budget deficit. The author describes the sampling procedures used in constructing the consumer price index, and gives simple examples of formula bias and quality adjustment. He then reviews the empirical evidence, attempting to show which biases are reliably estimated and which estimates of bias are based on extrapolation and guesswork. The author discusses possibilities for further research leading to potential improvements in the consumer price index.

A Bureau of Labor Statistics Perspective on Bias in the Consumer Price Index

“…Although the effect of going to a geometric mean formula at the substratum level from the pre-1995 version of the CPI formula appears to have reduced the rate of growth of the index by perhaps 0.5 percent per year, BLS research during the past year suggests that only about half of this difference is accounted for by formula bias (see McClelland, 1996, and unpublished estimates by Smedley and Gallagher, cited in Moulton, 1996). BLS made several changes to correct formula bias in January 1995 and made two more changes in June and July 1996 that should effectively have eliminated the problem.…”

“…Brent Moulton has further pointed out how the introduction of new goods or new types of outlets at systematically lower (or higher) quality-adjusted prices extends the issue of quality change from situations of item substitution to sample rotation as well (Moulton, 1996). That is, at present, BLS makes quality adjustments when an item disappears from its CPI sample and is replaced by another good with different characteristics.…”

Biasing Factors of the Consumer Price Index

“…Naturally many initiatives from within (Bureau of Labor Statistics) have seen the light of day (Gordon, 1995, Griliches, 1995, Reinsdorf, 1998, Moulton, 1996, Triplett, 1999. In 1994 university professors and statisticians created the Ottawa group in order to coordinate work concerning the Consumer Price Index.…”

The Boskin Commission Report and its Aftermath

“…The Commission did not give adequate attention to quality deterioration , BLS, 1997, Abraham et al 1998 2. The BLS already makes lots of quality adjustments, which the Commission did not adequately credit , BLS, 1997, Moulton, 1996 Moulton and Moses, 1997); 3. The Commission made too many back-of-the-envelope calculations and was too willing to generalize from research on one item to research on related items, both with respect to quality change and with respect to the desirability of geometric means to deal with lower level substitution bias (Abraham et al 1998, Nordhaus 1998 4.…”