One of the endless sources for discussion and debate in economics is whether key macroeconomic indicators actually measure what they are designed to measure. Inflation is certainly high on that list. The usefulness of the consumer price index as an inflation measure is a critical question, since inflationary pressures are pivotal to much of the developed world's monetary policy.
Most complaints about CPI centre on whether the basket of goods and services used to compile the measure is a fair representation of how consumers experience price inflation. And more particularly, there is an ongoing objection in some quarters to the use of so-called "core" CPI inflation measures that exclude the most volatile goods – typically, food and energy – in order to remove perceived distorting noise from the underlying trend of inflationary pressure, yet dismiss massive swaths of day-to-day costs to which household consumers feel keenly sensitive.
But researchers at the Bank of Canada – which itself has a keen interest in such matters, as it relies on inflation targeting as its sole guide for setting interest-rate policy – are asking an even more fundamental question: Is CPI inflation truly measuring rising prices of goods and services? Or might it largely be measuring "quality" shifts – i.e. consumers gradually and continually opting for higher-quality purchases that simply cost more?
This is the central premise in a study released by the bank this week. The paper, authored by Oleksiy Kryvtsov, aimed to identify how much of Canada's CPI inflation can be accounted for by consumers upgrading to higher-quality products. Mr. Kryvtsov found that there is such a "quality bias" in Canadian non-shelter CPI (i.e. excluding housing, where quality upgrades are a large and obvious element) – but it's not big enough to undermine the indicator.
Using newly available Statistics Canada data for prices of goods and services posted by Canadian retail outlets from 1998 to 2006, Mr. Kryvstov identified new-and-improved products that replaced previous products as part of the CPI calculation. He determined that about one-third of price increases for new-model products reflected a quality increase; the remainder was price inflation. And since new-model upgrades represent only a smallish slice of all products in the CPI calculation in a given year, this quality bias has a relatively small impact on overall CPI numbers.
"This implies overall price inflation [is] close to inflation measured by the official index. I conclude that … the quality bias is not an important source of potential mis-measurement of CPI inflation in Canada," he wrote.