CPI Bias

1035 Words3 Pages

The accuracy of the CPI (consumer price index) as a measure of the rate of increase in the cost of living has long been under scrutiny, with many studies showing that it overestimates figures - albeit to different extents. One of these, “Measurement error in the consumer price index: where do we stand?” (2003) by David Lebow and Jeremy Rudd, provides a comprehensive analysis of the five causes of the bias in the CPI and a new set of estimates for them. They claim to be more accurate than previous estimates as a result of procedural improvements at the BLS (Bureau of Labour Statistics), new research and alternative judgment of existing information. This essay will briefly explain the report’s argument, giving further detail about the two main sources of bias; upper level substitution and new or improved goods entering a market. It will then explain the practical implications of this for the government and the economy given the CPI’s wide range of uses, making clear the need for policy makers to pay more attention the issue.

In their report, Lebow and Rudd conclude that the USA’s CPI overestimates increases in the cost of living by 0.87%, within a range of (0.3%-1.4%). This figure can be broken down into five constituent categories of bias: new outlets, weighting, lower-level substitution, upper-level substitution and quality change or new items. The former three have a significantly smaller effect on the bias, making up only 0.2% of the 0.87% estimate. New outlet bias, said to make up 0.05%, occurs because the CPI only uses changes in quality to explain price differences between old and new outlets.This fails to take into account shifts in buying patterns which often cause price changes. Weighting bias, which had prev...

... middle of paper ...

...o make informed decisions. They should also encourage research into these areas and improvements in how the CPI is collected, for example using the PCE instead of the CEX, and calculated, for example using a geometric mean aggregation formula instead of Laspeyres.

“Measurement error in the consumer price index: where do we stand?” has provided a detailed analysis of the causes of bias in the US consumer price index and the uncertainty behind both previous and their own estimates. Of these, upper-level substitution and quality or quantity changes have proven both the most significant and the most controversial. With the figure for the bias calculated at a plausible 0.87%, it has been made clear the practical implications of their argument across the economy, and that policy makers must keep track of bias estimates and inform their choices acknowledging them.

Open Document