« Midwest economy still expanding, but is the rest of the country slipping into recession? | Main | Nonfarm payroll employment up by 128,000 »
August 31, 2006
Core inflation lower than expected--but how much does rounding matter?
After reading the FOMC minutes, I became convinced that it would take a spike in inflation (particularly the core measure) to get the Fed to resume the rate hikes in September. Well let me revise my priors upward once again. Today's inflation figures make it more likely that the Fed will keep rates where they are. But it does little to ease my concern that more increases are on the way in the months ahead. The year-on-year measure is still higher than many would like. The longer it stays that way, the harder it is to get it down (unless you get lucky like we did in the mid-'90s).
So the expectation was for the core to rise by 0.2% and it only rose by 0.1%. Reuters has the story.
WASHINGTON (Reuters) - Core U.S. inflation rose less than expected in July, but the year-over-year pace remained elevated, according to a government report on Thursday that the inflation-wary Federal Reserve will examine.
Core consumer prices rose a less-than-expected 0.1 percent in July, but the year-on-year rate of nonfood, nonenergy inflation remained at 2.4 percent, the highest since September 2002, a Commerce Department report showed.
Analysts polled by Reuters were expecting a 0.2 percent gain in core consumer prices, an indicator that will be a factor in the Fed's decision whether to hold interest rates steady.
For the Federal Reserve "this is another reason to not increase," said Robert MacIntosh, chief economist at Eaton Vance Management in Boston.
Or another reason to go to three decimals as the BLS is doing for the Consumer Price Index. Today's report is on the PCE which, presumably, will remain a one decimal series. (UPDATE: See comments.)
Not that it makes a whole lot of difference in the long run. If there is a lot of rounding down this month, it will just be more likely that it will round higher next month. But it does make one cautious about interpreting the difference between 0.2% and 0.1% in one month's change.
But even if it were a 0.2% in the PCE price index, it probably wouldn't be enough to move rates at this moment in time. The more incoming data we receive, the more it looks like the Fed will simply pause until the recession threat subsides, even if core inflation remains at current levels. They remain committed to preventing further increases in core inflation, but are not rushing towards "disinflation". If we get some relief on energy prices, we might just get lucky, just like in the '90s.
The march of data continues tomorrow with nonfarm payroll employment.
UPDATE: Macroblog links to the Dallas Fed's trimmed mean PCE inflation rate and is not ready to "declare victory" over inflation.
Posted by William Polley at August 31, 2006 11:20 AM
Trackback Pings
TrackBack URL for this entry:
http://www.williampolley.com/cgi-bin/mt-tb.cgi/581
Comments
(Assuming I’m reading the report correctly…)
Although the change is only reported to one decimal place, the raw deflator appears to be reported to 3 decimal places. I was actually concerned about just the issue that you cite here, but looking at the change in the raw deflator, it appears that the growth rate was almost exactly 0.10%.
Posted by: knzn at September 1, 2006 7:20 AM
You are correct about the raw deflator. In my haste I looked only at table 11 and not table 9 in the report.
When you do look closer at these numbers, you find that there is a bit of a difference between the PCE ex. food and energy and the "market based" PCE ex. food and energy. The latter is, as you say, close to 0.10% but the former is about 0.136%. That would mean that implicit prices are rising more than explicit prices. Interesting.
Anyway, the BLS will continue to report 1 decimal percentage changes for the CPI. As they state, "In general, the sampling error associated with published CPI percent changes suggests that continuing to publish percent changes to one decimal place remains appropriate." That all depends on the interpretation of the data. (The burden is on us.) Annualized, the difference between 0.1% and 0.2% is quite large (about 1.2% yearly). And so we caution ourselves (and others) not to put too much into one month's data point. Or at the very least calculate another decimal place before attributing too much to the figure. Thanks for the correction.
Posted by: William Polley at September 1, 2006 5:51 PM