Kass: Not-So-Little Lies in Inflation Data

Then there is the owners equivalent rent calculation within the CPI, which has dramatically understated and misrepresented the historical cost of housing. And how about the devastating impact that mortgage-rate resets (from the teaser loans of the last three years) will have on homeowners? It, too, is nowhere to be found in the government's measures of inflation, which, in the main, represent a true work of fiction. Even New York Mets baseball tickets for the 2007 season have been goosed by as much as 20%.

Measuring the general level of prices has become a controversial exercise. Often, nonmonetary events -- such as geopolitical crises, weather and financial accidents (particularly of a hedge fund king) -- interfere with and distort the trend-line rate of inflation. The solution, adopted by monetary and government authorities, has been to look at CPI by taking out two volatile or high-noise components (food and energy).

By contrast, I have encouraged readers to look at the measure of inflation calculated by either the Federal Reserve Bank of Dallas (trimmed mean PCE rate) or the Cleveland Fed (weighted median CPI).

Of the two, it is my belief that the Cleveland Federal Reserve Bank's model (created by economist Dr. Michael Bryan) appears to present the most accurate picture and forecast of the general level of inflation, and it has a higher correlation with past monetary growth than other inflation measures. As such, it is far superior to either the CPI, excluding food and energy, or the all-items CPI.