-->Fortsetzung zum Dauerthema US-Statistiken


Anyone Think Inflation Is (Gasp) Understated?: Caroline Baum

Nov. 19 (Bloomberg) -- No one much cares about official price measures anymore because inflation is low enough so as not to be a factor in business or consumer decision-making.

That, of course, is the Federal Reserve's standard definition of price stability. While the rest of us might benefit from an actual numerical equivalent, the Fed has guided us to the Promised Land without a formal inflation target, and it's not about to institute one as long as Alan Greenspan is at the helm.

Nowadays inflation is discussed in terms that make it sound like some odd vestige of the 20th century, with the 21st century plagued by the risk of too little, not too much.

The monthly consumer price index still has some curiosity value. Because the Fed has abandoned any kind of market-price- based indicators of inflation or inflation expectations in favor of a state of Missouri attitude -- ``show me'' -- the CPI is the sole guide for measuring the Fed's progress in raising inflation to a more comfortable level.

So what is the CPI telling us? The CPI rose 2 percent in October from a year ago while the core index, which excludes food and energy, rose 1.3 percent. The year-over-year core increase ticked up 0.1 percentage point from September's four-decade low but is hardly conclusive proof that inflation is on the rise.

`Distortion'

There are some anomalies suggesting the core CPI may be understating inflation. In the past 12 months the annual increase in the core CPI has fallen by 0.9 percentage point. Exactly half of that decline can be attributed to one component: the imputed rental value of a home, otherwise known as owners' equivalent rent (OER).

OER is the largest single component in the CPI, accounting for 22.2 percent of the index and 28.2 percent of the core CPI. It is derived from a survey of rental units -- the same survey that is used to calculate another CPI housing component, rent of primary residence -- which is then weighted accordingly, depending on whether the unit is in an owner-dominated or renter- dominated area.

When utilities and fuel prices are rising rapidly, as they are now -- fuels and utilities prices rose 7.3 percent in the past year -- it distorts OER. (The Bureau of Labor Statistics eschews that term. It prefers to say that when utilities costs are rising, it makes OER go up more slowly than primary rents.) Utilities are subtracted from OER to arrive at a measure of pure rent (homeowners pay utilities separately), which has the effect of depressing the rise in OER.

Acknowledgement

The Fed is aware of the bias. Fed governor Ben Bernanke addressed it in a Sept. 4 speech.

``In periods when energy prices and utility costs are rising, as in the first half of 2003, the BLS procedure may overstate the deceleration in rents net of utilities and hence in owners' equivalent rent,'' Bernanke said. ``As a result, on this particular count, the slowdown in the CPI inflation may have been slightly overstated in the first half of 2003.''

In the past 12 months, primary rents rose 2.8 percent compared with the 2.1 percent increase in owners' equivalent rents.

Another oddity in the CPI is the plunge in used car and truck prices as calculated by the BLS. The three-month annualized decline in used vehicle prices of 27 percent is the largest in 25 years and looks suspicious when compared with the Used Vehicle Index compiled by Manheim Auctions, the largest and highest volume wholesale automobile auction company in the world.

Divergence

Manheim's October report showed new vehicle prices rising for the sixth consecutive month, with the year-over-year change turning positive for the first time since April 2001.

``Used auto prices account for 2.6 percent of the core CPI but (the 3 percent decline in October) subtracted almost 0.1 percentage point off the core,'' said Ian Shepherdson, chief U.S. economist at High Frequency Economics in Valhalla, New York.

On its Web site, Manheim compares its used vehicle index to the BLS measure of used vehicles in the CPI. Guess which one gets the higher grade?

The BLS uses a limited sample of used vehicles to determine price changes each month. Manheim incorporates all sales transactions -- over 5 million a year. The BLS uses used vehicle prices from a guidebook. Manheim uses actual sales. The methodology differs as well, with the BLS using an ``expected depreciation'' factor.

Then there's the whole gnarly issue of ``quality adjustment,'' by which the BLS subjectively determines how much of a price increase is improved quality, not a price change.

Assumptions

``The BLS has to make a lot of assumptions to turn (the Manheim) price increases over the last few months into huge declines,'' said Joe Carson, head of global economic research at Alliance Capital Management.

That it does. Policy makers are comfortable with them, however. And that's all that matters right now.

If the CPI anomalies start shifting to the other side of the boat, you could see the Fed resorting to some fancy oarsmanship to keep it on course.
Last Updated: November 19, 2003 00:01 EST