Philly Fed: Don’t Rely On Core Inflation

"We find that food and energy prices are not the most
volatile components of inflation
and that depending on which inflation
measure is used, core inflation is not necessarily the best predictor
of total inflation…"
  (emphasis added)

Core Measures of Inflation as Predictors of Total Inflation


You may have overlooked a recent research piece out of the Federal Reserve Bank of Philadelphia and the Wharton School at University of Pennsylvania. Don’t.

The research team essentially found that Core Inflation is an erroneous way to measure ongoing price increases.

Both of the main rationales offered for policymakers€™ for their focus on core measures of
inflation do not survive close scrutiny, argues a group of researchers.

Their main findings were that:

-Other components of Inflation are more Volatile than Food & Energy;
-Core Inflation is less Valuable as Inflation Forecastor;
-Combining CPI and PCE inflation measures can lead to more accurate forecasts

Its worth noting that the shift in focus from total inflation to core inflation was a Greenspan era "innovation."

A quick excerpt:

"We find that core inflation, which omits food and energy prices, is
less volatile than total inflation, but the reduced volatility comes
from omitting the energy components. Several components of the CPI
exhibit higher volatility than food prices. And an index that omits
food and energy prices demonstrates slightly more volatility than a
measure that omits only the energy components and retains the food

Perhaps most important, we find that including PCE inflation when
forecasting CPI inflation and including CPI inflation when forecasting
PCE inflation significantly improves the accuracy of the forecasting
model for horizons up to one year. This suggests that each measure of
inflation provides independent information that can be exploited to
yield statistically significantly more accurate forecasts."

Now you have some reading for the holiday weekend . . .


Core Measures of Inflation as Predictors of Total Inflation
     Theodore M. Crone, Swarthmore College
     N. Neil K. Khettry, Murray, Devine & Company
     Loretta J. Mester, Federal Reserve Bank of Philadelphia and the Wharton School, University of Pennsylvania
     Jason A. Novak, Federal Reserve Bank of Philadelphia
Federal Reserve Bank of Philadelphia, May 2008

Download wp08-9.pdf

Print Friendly, PDF & Email

What's been said:

Discussions found on the web:
  1. Jim Haygood commented on May 22

    In his June 2008 Investment Outlook, Bill Gross presents a chart showing that headline and core inflation have diverged for nearly 3-1/2 years now. And he’s not confident that they will EVER converge again — as they should, if core inflation is merely a less-volatile version of overall inflation.

    Mathematically, there are several ways of suppressing volatility without excluding critical inputs such as energy prices. The Cleveland Fed publishes a median CPI. In its own Beige Book forecasts yesterday, the Fed employed trimmed means (excluding the three highest and three lowest of 17 forecasts).

    Given these superior techniques, one has to question the stated agenda of reducing spurious volatility. Frankly, it’s a lie. The agenda is to suppress the inflation rate, on an ongoing basis. Bill Gross says they’ve suppressed it by over a percentage point. I concur.

    As the demise of the former Soviet Union showed, ultimately those chickens come home to roost. The estimated GDP of the Soviet Union ended up being slashed nearly in half, after the true state of its “value subtraction” economy became apparent in the early 1990s.

    The United States is not that far along, although it accumulated a great deal of malinvestment during the Greenspan Bubble era, and not only in real estate. However, the federal statistics mill is one of the value subtraction sectors of the U.S. economy, along with Congress and the overblown financial sector. The longer the sham continues, the larger the step-function drop later. Lies, lies; tell me sweet, sweet lies!

  2. Mari commented on May 22

    I am no economist – but wanted to make sure. I thought the food component was excluded not only it exhibits “volatility” but is influenced by non-human controlable events like “rain and drought” and thus monetary policy has non-significant impact. Please correct me if I am wrong. Thanks.

  3. RW commented on May 22

    Setting aside conspiracy theories for the moment (but not too long as I know my tinfoil hat protects me from more general harm), I suspect one the problems is that cost of living and inflation are not necessarily the same thing and, in different economic scenarios, a single metric attempting to measure both could wind up more accurately measuring one or the other or possibly even neither.

  4. Garry commented on May 22

    Dear Mr Barry Ritoltz have you changed something with the feedburner (RSS) for firefox ? i’don’ receive nothing in my player , I do not get more articles on your site so interesting?

    sincere greetings


  5. Steve Barry commented on May 22

    I personally have had enough…I’m not going to take it anymore…I will short the market to a smoldering crater and not shed one tear.

  6. Steve Barry commented on May 22

    Hey Kudlow,

    How’s your dry shipping rate looking now for the economy?

    Look at DRYS babay.

  7. Steve Barry commented on May 22

    DRYS down 20% off its high in 3 days…oil may have hit the magic number. Hey Fed, cut rates again for 150 oil.

  8. John Borchers commented on May 22

    The market is bubble hunting now. They found some too.

    Look at FCX.

  9. John Borchers commented on May 22

    The good news is after the market deflates the Fed can cut rates.

  10. Jim Haygood commented on May 22

    Mari is right that food costs are affected by climatic influences. But in a globalized economy, those spurious influences have less impact than they used to, when food supply was more localized.

    If the California strawberry crop fails, we buy strawberries from Chile. If the Florida orange groves freeze, we buy oranges from Brazil. And so forth.

    A worldwide grain price spike caused, in part, by government policy of burning corn for fuel, might exaggerate price pressures. But such extreme spikes rarely occur except in times of inflationary easy money (1973 being another example). So I think food price spikes are still informative, even if they overshoot in magnitude. A median or trimmed mean procedure can prevent a food price overshoot from unduly raising the composite inflation estimate.

  11. Cory commented on May 22

    If its anything like Canada, keeping the core readings down ensures muted cost increases for social benefit outlays.
    Given the demographics, the oldest countries need to find ways of reducing the burden.

    Massage the core readings and reduce outlays.

  12. Cam Hui commented on May 22

    The Dallas Fed’s trimmed mean PCE might be a better measure than core, see

    I do agree with you about the “inflation ex-inflation” concept though as most of the inflation that is occurring is in commodity prices and later feeding into the overall inflation rate.

  13. Estragon commented on May 22

    Mari & Jim Haygood,

    Whether food prices are impacted by weather, phases of the moon, or a elvish conspiracy isn’t relevant for measuring inflation.

    In a non-inflationary environment, money spent buying more expensive oranges, whether Floridian or Brazilian, means less money spent on something else (and downward pressure on the price of that something else).

    In an inflationary environment, the value of money declines, and this mitigates the decline in nominal prices that would otherwise occur.

    The idea behind excluding food and energy is to avoid using a medium term tool (monetary policy) in reaction to shorter term price movements. It has nothing to do with why those movements happen.

  14. AGG commented on May 22

    After a thorough and complete study of the CPI and all its’ provisions and effects on the economy and well being of Republicans, we at the BLS have decided to make sure the CPI reflects reality starting in February, 2009.

  15. DonKei commented on May 22

    Well said Estragon…repeat after me everyone…inflation is everywhere and always a monetary phenomenon.

    Best way to measure if the money is declining in value (i.e., if you have inflation)? Check how it looks against real things that don’t change over time, like say, commodities.

    Forget the CPI, the core and all that nonsense. Look at internationally traded commodities (a broad basket of them), and there you will find whether your money is losing or gaining value.

    If the price of pretty much all of them is rising in your currency, then you got yourself an inflation, like now. See, its always the money.

  16. Winston Munn commented on May 22


    I would make a slight change to your statement lest we confuse issues: inflation is always a monetary and debt phenomenon.

    A rising money stock is not required – if fact, money supply can be dropping while debt is rising and you get the same inflationary outcome.

  17. Investment Advice, Personal and Individual Insurance, Investment commented on May 22

    Death By Debt

    I decided to republish this information in light of the current economic crisis this country is facing now. This article first appeared on an older version of my website in January of 2007. Here we are a little more than a year later, and a little wors…

Posted Under