A to Z Index  |  FAQs  |  About BLS  |  Contact Us    

Beyond BLS

February 2016

A price index that matches perceptions of inflation

Richard Hernandez

Do your inflation expectations align with the inflation rate calculated from the Consumer Price Index (CPI)? If not, you’re not alone, and two economists believe they have found the reason why.

In the working paper titled “Capturing the inflation that people experience: the Everyday Price Index vs the Consumer Price Index” (American Institute for Economic Research working paper no. 004, June 2015), economists Polina Vlasenko and Steven R. Cunningham explore why inflation expectations often don’t match up to reality and show how their Everyday Price Index (EPI) is better than the CPI at reflecting perceived inflation.

To explain the difference between perceptions of inflation and the reality of inflation, the authors use the concept of frequency bias, which they explain is “the idea that people tend to weigh prices of more frequently purchased items more heavily [than prices of less frequently purchased items] when forming their view of inflation.” The authors created the EPI with the goal of measuring price changes for goods and services consumers purchase on a regular basis (defined by the authors as purchased at least once a month).

The study finds that from 1987 to 2001 the CPI and the EPI had similar movements, rising 57 percent and 54 percent, respectively. Starting in 2001, a divergence began: the EPI increased by an average of 3.1 percent a year, compared with a 2.3-percent average annual rise in the CPI. By December 2014, the CPI had increased 35 percent and the EPI had risen 46 percent. The authors found that the EPI is more volatile than the CPI because of the different weights each index gives to purchased items. The EPI puts an emphasis on food and beverages and, in contrast with the CPI, excludes housing costs. After 2001, food items, as well as gasoline, tended to be more volatile than housing.

To test how the EPI is better than the CPI at representing the public’s view of inflation, the authors used the Granger causality test. Analyzing data from the Survey of Consumers, the Consumer Confidence Survey, and the Survey of Professional Forecasters, the authors arrived at the following: the CPI is better at predicting the professional forecasters’ expectations of inflation, but the EPI provides a better prediction of inflation expectations of the general public.

Because the EPI focuses on items bought on a repetitive basis, such as food, the EPI can be seen as being more relevant for everyday price changes affecting consumers. Its relevance may be especially high for individuals who have contractually fixed housing costs and who tend to have a fixed basket of goods they purchase on a regular basis. The EPI, therefore, can be especially helpful for budget and planning purposes of people on a fixed income, such as the elderly. The authors caution, however, that the EPI is not an inflation index for overall price levels, and so an EPI should be constructed along with the CPI.

Download PDF »