When the Consumer Price Index is reported every month, the media highlight two numbers: the change in the total index and the change in core prices. The traditional measure of core inflation excludes from the total the volatile components of food and energy. In June, seasonally adjusted core prices for all urban consumers rose 0.1 percent, following a 0.2 increase in May. This was widely interpreted to mean underlying inflation had cooled.
But did it? An alternative and in many respects preferred measure of core inflation says otherwise.
First some background. Measures of core inflation attempt to discern the underlying trend in consumer prices and as such are important to economic policymakers, notably the Federal Reserve, as well as to investors and consumers themselves. In the overall index, trends are often obscured by temporary or erratic fluctuations in some components, which has led economists to devise measures to eliminate the “noise” and capture the longer-term movement of prices. A measure that reveals the underlying trend in consumer prices serves as an important predictor of future inflation.
Economists have criticized the traditional measure of core inflation on the grounds excluding food and energy throws out useful signals with the noise. The excluded components constitute a substantial part of the total price index and can exhibit trend movements of their own, not just temporary fluctuations. Moreover, other price components in the traditional core measure are often more volatile than food and energy. If reducing short-term volatility is the aim, there are better ways to do it.
An argument for the traditional measure is that food and energy prices are subject to supply shocks, such as droughts and production cutbacks by oil cartels, and so should be excluded. This makes the index more relevant for monetary policy, it is argued, since the Federal Reserve has its impact on the economy mainly through its influence on aggregate demand. Critics reply not all food and energy price changes are the result of supply shocks and that other prices included in the traditional measure are also subject to supply disturbances.
Proponents of the traditional core measure also argue it’s well known and accepted and that its simplicity makes it more readily understandable by the public. However, it’s not necessary to sacrifice accuracy for transparency and an alternative core measure need not be complex.
Among the alternative core inflation measures proposed and tested by economists, two in particular stand out. One measure trims the Consumer Price Index to exclude a specified percentage of extreme positive and negative monthly price changes. The particular components excluded may vary each month. The resulting core measure is called a trimmed mean index. Even though it does not exclude the same components every month, as the traditional core measure does, the trimmed mean index still tosses out potentially useful information on trend inflation.
A second alternative measure of core inflation is the median monthly change in the Consumer Price Index, that is, the value at which half of all changes are larger and half are smaller. This measure doesn’t exclude components of the overall index yet has the virtue of reducing the influence of extreme price changes.
Both of these refined core inflation measures are well accepted and regularly estimated each month by central banks around the world. Economists have tested alternative core indicators for their ability to track current trend inflation and to predict future inflation. For the most part, the test results favor the trimmed mean or the median over the traditional U.S. core measure, although all the measures have their shortcomings.
The Federal Reserve Bank of Cleveland regularly calculates a median inflation rate each month based on the official consumer price data and reports the result on the same day the Labor Department’s price data are released. The Bank’s news release reports percent changes in its median index from the previous month and from a year ago. The information is easily accessible on the Cleveland Fed’s Web site and is a valuable supplement to the traditional core inflation number. The news release can also be obtained by e-mail. Contact: linsey.molloy@clev.frb.org. Media take note.
In June, according to the Cleveland Fed, the median price index rose by 0.2 percent, or at the same rate as in May. In contrast to the reduced increase in the traditional core index, widely reported in the press, the median index showed no deceleration in the underlying rate of inflation in June.
Over the last 12 months the Cleveland Fed’s core measure rose 21/2 percent, more than a half point above the rise in the traditional core rate. On a 12-month change basis, both measures show a pattern of increase since the beginning of the year. The message does not make for complacency.
Alfred Tella is former Georgetown University research professor of economics.
Please read our comment policy before commenting.