U.S. Inflation Measure May Be Rotten at the Core

[FRED]

U.S. Inflation Measure May Be Rotten at the Core
September 15, 2006 (Caroline Baum - Bloomberg)

U.S. policy makers have long defended the practice of using a core inflation measure as a check on how they're doing. The public, of course, sees this as just another gimmick the government uses to pull the wool over its eyes.

The concept of stripping out historically volatile food and energy prices from inflation indexes is ``an issue of trying to forecast more effectively the overall inflation rate,'' Federal Reserve Chairman Ben Bernanke explained in the Q&A following an Aug. 31 speech in Greenville, South Carolina.

A second reason for targeting core inflation ``has to do with policy making,'' he said. In order to offset the immediate effect of an increase in energy prices, the Fed would have to ``force down wages and other prices quite dramatically to keep the overall price level from rising,'' he said. The alternative is to allow ``first-round effects to pass through'' and try to ensure that the energy-price spike doesn't pass through to other prices and wages.

It sounds nice, but when you think about it, it doesn't make sense. Taken at face value, Bernanke is saying the Fed would be happy to see inflation rise as long as the core rate doesn't budge from its comfort zone. Huh?

AntiSpin: Caroline's got this partly right. To economists, the term "inflation" is an abstract term that's used to determine policy. There are lags between the time inflation starts to show up in input costs, such as energy, and the time inflation starts to show up in final goods prices, and later when it starts to show up in wages, at which point it tends to get built into contracts and become self-reinforcing. The Fed doesn't want that but also tries not to react to rises in input costs that can lead to inflation if they stand a chance of going away on their own, else the Fed will choke of economic growth unnecessarily.

That said, the Bureau of Labored Statistics, that generates the PPI, CPI and other inflation measures, has other constituencies besides the Fed, notably the Treasury Department. TIPS and other inflation-adjusted government securities are price-adjusted based on the BLS' inflation numbers, and difference of a fraction of a point represent billions of dollars in government liabilities one way or another. Given the stakes, one can be forgiven for wondering if the separation between the inflation accounting wolves and the treasuries hen house is always perfect.

In any case, as normal, beer drinking, food eating, car driving folks, we experience "inflation" in non-abstract ways. The chart below shows how we experience the 300% rise in CPI inflation since 1978.

Depending on how much you as a person or a household spend on, say, consumer electronics, cars and clothes versus college tuitions and medical care, you are going to experience the inflation very differently, as the former are declining in real terms and the latter rising. Throw into the mix the fact that the rising cost of medical care arrives in the form of rising insurance premiums, which are purchased out of income, whereas the TVs are put on the credit card, which costs rise and fall with interest rates, and the actual experience of inflation versus the abstract CPI aggregate explains why most people think economists are full of it.

[jk]

Re: U.S. Inflation Measure May Be Rotten at the Core

i heard on the radio this morning that the projected cost of a college education at a private university for a child born this year is...... ready?........$435,000.

[Finster]

Re: U.S. Inflation Measure May Be Rotten at the Core

U.S. Inflation Measure May Be Rotten at the Core

Of course it is. The rationalization usually offered for citing so-called "core" inflation is that the food and energy components are volatile; "core" therefore gives you a truer reading of the underlying trend.

Poppycock and balderdash. Every statistician and mathematician knows if you want to eliminate noisy volatility, you take a moving average.

Trouble is, merely taking a moving average of the broad series, you wind up with something that tenaciously runs above what you get if you merely remove the things that tend to go up the most. This puts the lie to the whole "eliminating volatility" story. If that was all you really wanted to do, you'd just take a moving average.

So-called "core" inflation is a hoax upon a hoax. Even the so-called "headline" inflation smooths, lags, and understates. The US dollar is losing value faster than officialdom will admit.

The problem of determining what inflation really is eventually led me to write to a respected publisher of financial indices urging them to publish an independent measure of the value of the dollar, much like they do with stocks, commodities, bonds, and other securities. Failing that, I ultimately resorted to producing one of my own. The message is not quite the same as what we hear from the sponsor of that security - it's been losing value considerably faster than it admits.

Over the past year, according to the "core" CPI, the dollar has lost 2.7 % of its value (2.8% increase in "core" CPI), according to the "headline" CPI, it’s lost 3.8% (3.9% CPI inflation). By the more comprehensive and unbiased FDI, however, over the same period the dollar had lost 4.2% of its value (4.4% inflation). And this is with the benefit of avoiding the pronounced lag embedded in consumer prices which is tending to exert and upward bias due to their capture of even higher past inflation. A year ago, for example, inflation was running as much as four percent higher.

Rotten to the "core" all right.

[Uncle Jack]

Re: U.S. Inflation Measure May Be Rotten at the Core

I'm so tired of inflation being stated in terms of price I can hardly stand to listen/watch those morons on financial entertainment television any more. Let's have some serious deflation, which I think has already started and is baffling the bonars out of those talking heads, so we can really have some fun.