Recent measures of inflation by the chain-weighted Consumer Price Index and Personal Consumption Expenditures index show some progress towards the Federal Reserve’s 2% objective, with the former on a slightly faster march towards the target, economists at the Cleveland Fed said Tuesday.
In a research note posted on the bank’s website, Vice President Todd Clark – in charge of monetary policy and macroeconomic research – and Senior Research Analyst William Bednar also cautioned against using the regular CPI to assess how close the central bank is to meeting its price stability mandate.
The measure of inflation preferred by the policymaking Federal Open Market Committee is the PCE price index, which – with the core index up 1.5% year-over-year in June – the Fed economists say shows that inflation “remains somewhat below the desired level.”
But, “the chain CPI, like the PCE, shows some recent progress toward the FOMC’s longer-run inflation goal of 2%, with chain CPI inflation a little closer to 2% than PCE inflation has been,” they added.
Inflation as measured by the core chained has been running “consistently a little higher” than core PCE inflation, they said, but following a very similar monthly pattern. At present, core chain CPI inflation is a little below 1.8%, while core PCE inflation is 1.6%.
On the other hand, recent CPI data paint a different picture regarding where the Fed is in relation to a 2% target. Core CPI inflation increased to 2.0% in May, and as of July, was at 1.9%, Clark and Bednar wrote, so “evaluating current inflation levels with the CPI could lead us to think that we have been right on target over the past few months.”
However, they argued “it isn’t quite right” to use regular CPI measures of inflation to evaluate how close the economy is to hitting the FOMC’s inflation target.
Substitution bias – when consumers change their purchasing habits due to the price of one good increasing vs. another, spending more on the now cheaper item – is a major factor in the discrepancy between CPI and PCE inflation rates, they said.
“The PCE measure of inflation does a better job of capturing this kind of substitution than does the CPI,” The Cleveland Fed economists said.
They added that the weights used in the construction of the PCE index are updated each month based on the composition of consumption in that month, while those used in the CPI are adjusted at a lower frequency. “As a result, the CPI does not do as good of a job as the PCE price index at capturing substitution, which causes CPI inflation to (on average) exceed PCE inflation,” Bednar and Clark said.
However, the chain-weighted CPI does try to account for substitution bias in a manner similar to the PCE price index, they said.
In the chained CPI, weights for each individual category are adjusted on a monthly basis in order to keep up with changes in consumption, they said, eliminating the substitution bias in the CPI.
“Over time, the average rate of inflation in the chained CPI is similar to the average rate of inflation in PCE prices (both overall and the PCE excluding food and energy),” the Cleveland Fed economists wrote, “as a result, for comparing CPI inflation to the FOMC’s 2% inflation goal, the chained CPI is more appropriate than the more widely publicized, simple CPI.”