The following Ed Lotterman article appeared in the Sunday April 27, 2008 issue of the St. Paul Pioneer Press on page 3D.
Even economists experience sticker shock. I just bought one gallon of skim milk and one of 2 percent at the corner store. I do this regularly, but $9.14 for two gallons of milk still seems like a lot.
I'm not alone. Millions of people perceive that rising fuel and food prices are crimping their families' standard of living. No wonder they get upset when monthly inflation numbers are released and some expert notes that "core inflation" is up only slightly. It contradicts their daily personal experience.
The problem stems from the misapplication of a statistical measure - core inflation - that is useful in some situations but misleading right now.
For 80 years, the U.S. government has tabulated price indices to measure changes in general price levels. The consumer price index is the best known, as it measures goods and services that households buy. Every month, the government checks prices on thousands of items. On the whole, the CPI is an accurate indicator.
But prices of some items jump around more than others. Both food and fuel are important for most households, so they have a lot of influence on the overall index. But their prices tend to fluctuate more in the short term than those of clothing, household goods, shelter or recreation.
The fluctuations, incorporated in something tabulated on a month-to-month basis, can mislead.
Suppose overall prices are increasing at an average annual rate of 3 percent. Items other than food and fuel may increase at a rate of 2.5 percent one month and 3.5 percent the next, but follow this 3 percent trend. Then suppose food and fuel increase at a rate of 6 percent one month and again unchanged the next. Again, the long-run trend is 3 percent.
If you just consider the month that food and fuel rose at a 6 percent rate, it will appear that inflation is hot. The next month it will look like inflation is non-existent. But in the longer run, the trend in food and fuel is about the same as for other goods.
One can avoid this confusion by measuring consumer prices generally, but then making a separate tabulation that excludes volitile items. Weighing the two tabulations together gives us a better picture.
That is why the "core inflation" tabulation was introduced, to allow analysts to factor out short-run volatility that might be misleading.
A problem arises, however, when the longer-term trend for food and fuel is not the same as for other items. Suppose other items increase at rates between 2 percent and 4 percent with a trend of 3 percent. Month after month, food and fuel fluctuate between 6 percent and 10 percent annual rates with an average trend of 8 percent. Just looking at core inflation blinds one to a larger problem.
Taking comfort in a measure designed to dampen short-run fluctuations can mislead one about true prices trends. Over most of the past 25 years, food and fuel pries ahve increased more slowly than everything else. In the last two years, however, they have increased faster, and the discrepancy is widening. Citing restrained core CPI numbers as evidence of constrained inflation assumes that food and fuel will soon be dropping. REalistically, that is not in the cards.
St. Paul economist and writer Edward Lotterman can be reached at firstname.lastname@example.org