Every year, my class meets Harriet Shaw through a PBS NewsHour broadcast from 1991. An economic assistant for the Bureau of Labor Statistics (BLS), she is a professional shopper who tracks prices. In the newscast, we follow her up and down store aisles until she stops to check out the hair spray. Same brand, same size, the hair spray’s price was down from $1.67 to $1.29.
I was concerned, though, that my information was from 1991. That took me to an equally fascinating 2010 Planet Money podcast. George Minichiello was also an economic assistant, doing precisely what Harriet did. Surveying a $2.99 bag of romaine lettuce hearts, he said the price was the same as last month but then realized it weighed more–the equivalent of a 27% price decrease. Soon after, at a clothing store, he records the price of a boy’s shirt.
The differences between Harriet and George? She was in Chicago, he in Brooklyn. She used a pen while he had a computer. Her information was FedExed while his is directly submitted.
So, where are we going with all of this? The accuracy of the Consumer Price Index, or CPI.
Harriet, George, and in 2010, 430 other economic assistants, collect and compare prices for the CPI. Involving something like 80,000 goods and services in 8 big categories including apparel and transportation and medical care, an urban consumer’s market basket is created by the BLS.
The results are important because the CPI is an inflation yardstick. Your grandma cares about the CPI because her Social Security increases depend on the CPI. At certain firms, the CPI affects wage negotiations. And yes, the future of quantitative easing could indeed depend on the CPI.
You can see why CPI accuracy is a concern. Reading about it, I kept thinking that it just does not seem to be a “nimble” statistic. At the BLS FAQ website (last revised August 2013), they explain that the Consumer Expenditures Survey from 2009-2010 is the basis for deciding the items in the current market basket. A smaller lag, there is a monthly wait for each new report.
The CPI lag took me to the “Billion Prices Project.” Created by MIT economists in 2008 and now overseen by a private group, a billion prices refers to the online prices that are constantly recorded and reported.
Until recently, the CPI and Billion Prices followed a similar path and then they diverged (below). WSJ hypothesizes that one reason for their most recent split was wintry weather. Unable to get to the mall, people were making more expensive online purchases that the CPI does not monitor.
What is our bottom line? While the CPI is certainly flawed, so too are a billion constantly monitored online prices because they ignore services that would include dentists and doctors and a meal at Chili’s (unless you order your food online). Is there an ideal way to calculate the inflation data we use for such important policy decisions? Your opinion? Please let us know in a comment.
Sources and Resources: H/T to Marginal Revolution for alerting me to this WSJ article (and the above graphs) on how the Billion Prices Project recently diverged from the CPI. Their information ideally complemented the PBS broadcast, Planet Money’s podcast and the BLS website.
Note: We have edited the title since the post appeared.