| Core of the Problem |
24 February 2003
|
CPI: Leaving Out the Inconvenient Facts
Inflation numbers as issued by the US government are supposed to measure price changes in the overall economy so that the Fed can set interest rates appropriately. The Producer Price Index gives a clue to what businesses face, and the Consumer Price Index provides a similar measure of what John Q. Public pays per month. However, the "core" CPI, which excludes energy and food prices, is a silly notion that should simply vanish.
The argument for keeping the core number states that food and energy prices are volatile and, therefore, they tend to distort the trend of inflation or disinflation (careful not to say deflation -- that's a whole different problem). In other words, by throwing out noise, one gets a better read of the signal, and say the statisticians, one gets a better understanding of the economy.
The hole in that logic is wide enough for a battleship. Every month people consume food, and they most often cook it using energy. They heat their homes, and they take hot baths and showers. Most of the time, peoplein the developed world are consuming food and energy on an almost hourly basis. Taking these figures out of the measure does not decrease the noise-to-signal ratio -- it diminishes the signal.
If Mr. Public sees food and energy prices rise sharply, while core CPI holds steady, he still feels inflationary pressures that will affect his spending habits (presuming a rational consumer, which economists tend to do too often). Dining out less, cutting back on more expensive foods, driving less, switching to public transportation are all very real effects of higher food and energy prices that will affect the economy in following months. Core CPI misses this, and it would be more accurate to ignore any measure that drops fundamental factors to make statisticians happy.