mouse.jpgAn unassuming research economist at the Dallas Federal Reserve Bank may influence decisions that will affect all of us in the next few months. A better measure of inflation that he has created indicates inflation is closer to 3 percent than the 2 percent rate sought by our central bank.

That could mean current interest rates may stand or increase, not decrease as many hope.

"In the grand scheme of economics, this is not rocket science. It's just something that needed to be done," Jim Dolmas said of his research project in a recent interview. A senior economist and policy adviser at the bank, Mr. Dolmas introduced a new treatment of the Personal Consumption Expenditures index, the broad inflation index the Fed prefers over the more familiar Consumer Price Index.

We'll leave exactly how his measure is done in a black box filled with the nasty little creatures statisticians deal with--- kurtosis and skewness---but the "Trimmed Mean" rate is created by eliminating the extreme changes in price measures in any month. Instead, it concentrates on the more central changes.

"A good deal of the credit goes to earlier efforts at the Cleveland Fed and elsewhere," Mr. Dolmas said.

Since most of this is pretty arcane stuff, I asked if he would explain why the Fed preferred the Personal Consumption Expenditures index over the better known Consumer Price Index.

"(The PCE) is a much broader basket of consumption goods," he said. "The CPI is a typical urban consumer index based on a basket of goods and services. The PCE goes with everything that gets consumed.

"More important, it doesn't have fixed weights. The weights (of any consumption category) vary with actual consumption. …Basically, the PCE is a much more dynamic index. If the price of gasoline goes up, people cut back. Or, if we have employee pricing for cars and people buy more cars, the weight of cars in the index rises."

While the CPI reflects a limited and fixed basket of goods and services, he pointed out, the PCE measures all consumption. It avoids the inaccuracy introduced when people move their purchases from traditional department stores to discount stores.

Why do we hear about "core inflation" when everyone consumes food and energy? I asked.

"We care about core inflation because that's what policy makers may be able to control--- they can try to control the trend in consumer prices. But more important, core (inflation) is better for forecasting future inflation. If you ask whether you want to look at the overall rate or the core rate, it turns out that you're better off looking at the core."

Since food and energy prices are highly volatile--- and always have been--- they can be misleading.

So the core rate is likely to be predictive? I asked.

"Yes. Getting a good picture of the trend is good for looking at what the rate may be in 12, 18, or 24 months. It's better than looking at the more volatile broad rate. So it's more useful for forecasting.

Unfortunately, that doesn't address the complaints people have about the specifics."

Mr. Dolmas pointed out that most of us respond to our personal experience of inflation, but that central bankers are looking for a measure that indicates our broad general well-being. Basically, they are looking for how inflation affects the average dollar. His measure, the Trimmed Mean PCE, works to include more meaningful indications of general inflation.

In late June, for instance, Mr. Dolmas found that the number of PCE index components that were rising at rates over 3 percent had risen from 33 percent to 57 percent from December 2005. He also found that index components rising faster than 2 percent had risen from 47 percent to 68 percent over the same period--- both indications of a broadening higher inflation.

Earlier this month he found that the trailing six-month annualized rate for the Trimmed Mean PCE was (1) larger than the PCE excluding food and energy and (2) rising. While both had started at a 2.4 percent trailing rate in March, the PCE excluding food and energy had risen to only 2.6 percent by August, while the Trimmed Mean PCE had risen to a 3 percent rate over the same period.

What does this all mean for you and me?

For the moment, it means that inflation is higher than the Fed would like it to be. The Fed is likely to act accordingly. That means interest rates are more likely to remain stable or increase than to decrease. On the web:

October 10, 2006: Trimmed Mean PCE inflation rate for August

August 2, 2005: The Dallas Fed introduces the Trimmed Mean PCE inflation rate

July 25, 2005: Trimmed Mean PCE Inflation paper