Meet the Fed's Elusive New Inflation Target

Inflation ain't what it used to be.

That's long been true in the sense that the inflation rate is now a lot lower than it has been for most of the decade.

But it's newly true in a different sense: The Fed has a new way of measuring inflation.

In its most recent Humphrey-Hawkins report, released last Thursday, the Federal Open Market Committee said it was changing its primary measure of inflation.

Instead of forecasting the rate of increase in the Consumer Price Index, the FOMC is now forecasting the rate of increase of an inflation measure contained in the GDP report. That measure is called (brace yourself) the chain-type price index (or deflator) personal consumption expenditures, or PCE.

PCE is the amount of money individuals spend on goods and services. It's the single largest component of the GDP. For example, the government's latest estimate of fourth-quarter GDP, released this morning, is $9.1 trillion, of which PCE accounts for $6.1 trillion. The chain-type price index for PCE measures the inflation rate for those expenditures.

The FOMC forecasts that the PCE deflator will rise 2% from the fourth quarter of 1999 to the fourth quarter of 2000, the same amount it rose from the fourth quarter of 1998 to the fourth quarter of 1999.

The committee didn't deep-six CPI. Even though it will henceforth forecast inflation in terms of the PCE deflator, the Humphrey-Hawkins report says: "The FOMC will continue to rely on a variety of aggregate price measures, as well as other information on prices and costs, in assessing the path of inflation."

So why the change? The report says (at the end of Section 1, if you want to look it up) that "the chain-type price index for PCE draws extensively on data from the consumer price index but, while not entirely free of measurement problems, has several advantages relative to the CPI."

The PCE deflator is better than the CPI, the Fed says, mainly because it better accounts for the fact that, as prices of goods and services change, people's spending habits change. For example, if apple prices increase, people tend to buy fewer apples but more oranges. (Improvements to the CPI in recent years have moved it in this direction, but it continues to track a fixed basket of goods and services, while the PCE deflator tracks a variable basket.)

This way of measuring inflation "avoids some of the upward bias associated with the fixed-weight nature of the CPI," the Humphrey-Hawkins report says. Indeed, as this chart shows, the chain-type price index for PCE has generally trailed the CPI over the last 10 years. In the last two years, the gap has been particularly wide.

Rising Prices Annual changes in two measures of consumer inflation
Source: Labor Department; Commerce Department

The same is generally true for core inflation, which excludes volatile food and energy prices, as the next chart shows. Interestingly, however, while core CPI fell in 1999, the core PCE chain-type price index rose.

Rising Core Prices Annual changes in two measures of core (ex- food and energy) consumer inflation
Source: Labor Department; Commerce Department

Additionally, the PCE chain-type price index is based on a broader set of expenditures than the CPI. And historical data can be revised more easily than CPI historical data, leading to "a more consistent series over time," the FOMC says.

The wisdom of the shift, however, is debatable. A recent paper by the Kansas City Fed concluded that, despite its shortcomings, the CPI is better than the PCE deflator, mainly because it measures prices more accurately.

Whatever. The shift has been made. The more immediate problem is that the Fed's targeted inflation rate is pretty well concealed. Not from professional economists, but from the rest of us. That will change, government statisticians say.

The monthly CPI report clearly specifies the year-on-year pace at which consumer prices are rising.

By contrast, the PCE deflator's rate of increase has to be mined out of the monthly personal income and consumption report, which comes out half a month later.

Table 4 of the monthly release shows the index value (calling it the implicit price deflator), but the report doesn't calculate the year-on-year rate of change for the purposes of comparing it with the Fed's forecast. One has to do that oneself by digging historical values out of Bureau of Economic Analysis archives.

That will change, probably not in time for the next personal income and consumption release on Monday, but hopefully with the February report, slated for release March 31, BEA senior economist Larry Moran says.

The quarterly GDP report started offering a bit more guidance with today's release of revised numbers for the fourth quarter. The text of the release calculated the year-on-year change in the PCE deflator for the quarter. Moran says the report's tables will be altered to incorporate the year-on-year trend in the PCE deflator as soon as next month, and certainly by the time its annual revision to the GDP figures rolls around this summer.

Historically, the report has calculated only the seasonally adjusted annual rate of change from the preceding quarter. If that sounds like so much palaver, it really isn't. As this next chart shows, the annualized quarterly change sometimes varies widely from the year-on-year rate.

Too Many Statistics Three different ways of measuring consumer inflation.
Source: Labor Department; Commerce Department

Another issue: While the core PCE deflator appears in the BEA's archives, it isn't included in either the personal income or GDP releases. By contrast, the change in the core CPI is perhaps the most closely attended item in that monthly release.

Moran says that isn't likely to change, for the simple reason that the FOMC limited its focus to the overall measure of inflation.