View original document

The full text on this page is automatically extracted from the file linked above and may contain errors and inconsistencies.

VOL. 6, NO. 4
MAY 2011­­

Insights from the


Inflation Measurement
Gives Us Food for Thought
by Jim Dolmas

If excluding some
components from the
calculation of the
inflation rate produces
a measure that,
historically, tracks well
the trend in the all-items
inflation rate, such a
measure would be useful.


lobal food prices are soaring. Since February 2009, the United
Nations Food and Agriculture Organization world food price
index has risen roughly 67 percent, surpassing the previous peak in June
2008 (Chart 1). The last food price surge, from early 2007 to mid-2008,
prompted riots in many countries; the latest rise has also fueled riots and
may have been a factor in political unrest sweeping through North Africa
and the Middle East.1
To the extent that the increases have contributed to accelerating inflation rates, they also challenge a number of central banks attempting to balance the goal of stable prices with the desire to support economic recovery.
U.S. consumers have—until very recently—been sheltered from this
price jump. From February 2009 to December 2010, growth in the personal consumption expenditures (PCE) price index for food was essentially
zero. That may be changing. In January and February, the PCE food index
increased 1.4 percent, an annualized rate of roughly 9 percent.
While the behavior of food prices—with implications for world
poverty, geopolitics and monetary policy—is interesting in its own right,
food prices are also a useful context in which to think about inflation
Economists and central bankers often refer to “core” measures of inflation that exclude the frequently volatile prices of food and energy items.
When the prices of food or energy are rising rapidly, policymakers’ reliance
on such core figures can become controversial.2
More generally, the idea that one can learn more about inflation by
ignoring some of its components is certainly counterintuitive. We can illustrate this principle at work, though, even among the components of the
often-excluded food category.3

First, though, the recent surge in
food prices is useful for illustrating the
way in which increased scarcity makes
less-abundant commodities relatively
more expensive, something that, in
general, monetary policy can’t undo.
Scarcity, Relative Price Movements,
Monetary Policy
Among the causes of the recent
world food price surge are weatherrelated poor harvests of staple crops
such as wheat and some coarse grains.
Affected heavily by drought in Russia
and excessive rains in Canada and
Australia, world wheat production for
the 2010–11 marketing period is on
track for a decrease of 5 percent compared with the 2009–10 period. World
production of coarse grains (corn,
sorghum, barley, rye and oats) is on
pace to fall 2.5 percent.4 Export bans
by some countries, increased stockpiling by others and higher input costs—
especially for energy in the production
of fertilizer—likely also contributed to
diminished supplies of many agricultural commodities.5
A basic principle of economics
is that decreased supply increases a
good’s relative price—that is, its price in

terms of the other enjoyable things one
must sacrifice to acquire the good. The
number of theater tickets or MP3 downloads or haircuts one must sacrifice in
order to enjoy a steak dinner increase
when the relative price of steak increases. So too, does the number of hours
one must work—and forgo leisure—to
obtain a given amount of steak.
This is true in a world where money is used to facilitate the exchange of
goods and services and would be true
in a world without money (and thus
without monetary policy).
Over periods of a few years (and,
of course, over longer horizons), central
banks can exercise considerable control
over the rate at which the prices of
an economy’s goods and services rise
or fall, in units of money. An important point, though, is that—with a few
qualifications—monetary policy affects
money prices for goods and services in
general, not the terms at which goods
and services exchange for one another.6
Monetary policy can slow the rate
at which food prices (together with
all other money prices) rise; it cannot
make food—or any other particular
good or service—more affordable in
terms of other goods and services.

Chart 1

U.N. Food Price Index Climbs to New Highs
Index, 2002–04 = 100






’90 ’91 ’92 ’93 ’94 ’95 ’96 ’97 ’98 ’99 ’00 ’01 ’02 ’03 ’04 ’05 ’06 ’07 ’08 ’09 ’10 ’11

SOURCE: United Nations Food and Agriculture Organization.

EconomicLetter 2


Finding a Core for Food
The components excluded from
core inflation measures—often, though
not always, food and energy—typically
have volatile price changes, and volatility is sometimes given as the rationale for their exclusion.7 The point of
constructing any core measure of inflation, though, is not simply to reduce
volatility. A core measure ideally helps
us gauge, in real time, the underlying
trend of a more comprehensive inflation measure.
Chart 2 shows annualized monthto-month percentage changes in the
PCE price index for food, along with a
measure of the trend in those monthto-month changes. Note that the trend
line does not reach to the ends of
the sample. It cannot: To know with
confidence the trend in a series at any
point in time requires knowing the
series’ behavior both well before and
well after that date.
In real time, where policymaking necessarily operates, we can only
know past inflation, and not, obviously, future inflation. If excluding
some components from the calculation of the inflation rate produces a
measure that, historically, tracks well
the trend in the all-items inflation
rate, such a measure would be useful.
Core inflation measures, to greater or
lesser degrees, have this trend-tracking
We can illustrate this principle
simply by looking within food itself.
There are 22 food-related components—not including meals purchased
at restaurants and similar establishments—that go into the all-items PCE
price index but are absent from PCE
excluding food and energy.
Those 22 components are listed in
Table 1, divided in a way I find useful:
those that are less processed and those
that are more processed. The categorization involves some judgment—while
“processed fruits and vegetables,” for
example, obviously belongs in the
“more-processed” category, the assignment of a component like “other
meats” is less clear-cut.9

Why is the division a useful one?
Consider the more-processed components. For our purposes, their most
important characteristic is that, by
and large, they consist of items with
brand identities. Lay’s potato chips.
Campbell’s soup. Kellogg’s Frosted
Flakes. These items differ from the
steak at the butcher’s counter in more
than just the steps involved in their
production. Their producers typically
possess market power, and thus face
nontrivial pricing decisions.
Producers of such items also tend
to change their prices infrequently,
even though their production costs are
fluctuating, perhaps on a daily basis.
Knowing that their next price adjustment may be weeks or months away,
producers must be forward-looking
when setting prices. If they perceive a
jump in the price of some input—such
as an unprocessed food item—to be
temporary, they may choose not to
fully incorporate the higher input cost
into their current price. A faster or
slower expected rate of general inflation should also be reflected in larger
or smaller price increases.10
Movements in prices for these
sorts of items are thus apt to be more
informative about future price developments and the underlying trend in
food price inflation.11 This is, in fact,
the case. We can aggregate the two
sets of components into two different
price indexes—“more processed” and
“less processed”—and compare their
abilities to track the trend in all-items
food price inflation (the trend line
shown in Chart 2).
Monthly changes in the moreprocessed price index are, on average,
closer to the trend rate of food price
inflation than are the monthly changes
in either the less-processed index or
even the all-food-items index itself, by
nearly a full percentage point in that
case (Table 2). The more-processedfood price index would make a useful
core for food inflation.
Such a core food index would
exclude about 30 percent of food
expenditure but would give us, as a

Chart 2

Monthly Food Price Changes and Their Trend
Annualized percent change

PCE food price index

36-month moving average





’59 ’61 ’63 ’65 ’67 ’69 ’71 ’73 ’75 ’77 ’79 ’81 ’83 ’85 ’87 ’89 ’91 ’93 ’95 ’97 ’99 ’01 ’03 ’05 ’07 ’09 ’11

SOURCES: Bureau of Economic Analysis; author’s calculations.

Table 1

Separating Components of PCE Food Prices
Less-processed items

More-processed items

Beef and veal

Bakery products



Fish and seafood


Food produced and consumed on farms

Coffee, tea and other beverage materials

Fresh fruit

Fats and oils

Fresh milk

Food products, not elsewhere classified

Fresh vegetables

Mineral waters, soft drinks, vegetable juices


Other meats


Processed dairy products
Processed fruits and vegetables
Sugar and sweets

SOURCES: Bureau of Economic Analysis and author’s judgment.

result, a better gauge of the trend in
overall food price inflation.
What We’re Seeing Now
Over the first six months of 2010,
as the global food price surge accelerated, the less-processed-food index


increased at an average annualized
rate of 6.7 percent. Over the same six
months, though, our core, more-processed-food price index actually fell at
a 0.9 percent average annualized rate.
Consistent with the greater informativeness of the more-processed index, over

3 EconomicLetter

the second half of the year, the rate
of increase in the less-processed index
abated to a 1.8 percent annualized rate.
Overall food inflation, covering lessprocessed and more-processed items,
ended up at 1.2 percent for the year.
The increases in January and
February of this year are different.
Not only has the less-processed index
risen sharply—at a 19 percent annualized rate—but, in contrast to 2010, the
more-processed index has also grown
robustly, at a 5.2 percent annualized
rate. A sustained period of higher food
price inflation may be in store for U.S.
Dolmas is a senior research economist and advisor at the Federal Reserve Bank of Dallas.

Regarding the 2008 food riots, see “Across

Globe, Empty Bellies Bring Rising Anger,”
by Marc Lacey, New York Times, April 18,
americas/18food.html. For the more recent
political unrest that began in North Africa, see,
for example, “Durability of Mideast Leaders in
Doubt,” by Michael Peel, Abeer Allam and Abigail
Fielding-Smith, Financial Times, Jan. 16, 2011.

See, for example, the Wall Street Journal edito-

rial “Deja Deflation Fear,” Review and Outlook,
Feb. 18, 2011,

Note that my intent is not to defend any par-

ticular core measure—there are a number of
alternatives to “ex food and energy” produced
at several Federal Reserve Banks, including the
Dallas Fed’s trimmed mean PCE inflation rate, the
Cleveland Fed’s median and trimmed mean CPI
and the Atlanta Fed’s sticky price CPI.

“World Agricultural Production,” Foreign Agri-

culture Service Circular Series WAP-03-11, U.S.
Department of Agriculture, March 2011, www.

See “The Future of Food: Crisis Prevention,”

The Economist, Feb. 24, 2011, www.economist.



The key qualification is that to the extent some

prices are “sticky,” or sluggish to adjust, a
central bank’s actions may affect some relative

The Bank of Canada, for example, uses a core

CPI measure described explicitly as one that
excludes the eight most volatile components.

See, for example, “Comparing Measures of

Core Inflation,” by Todd E. Clark, Federal Reserve
Bank of Kansas City Economic Review, Second
Quarter, 2001,

Table 2


Average Discrepancy
Between Monthly
Changes in Food Price
Indexes and Trend Food




All-food-items index
Less-processed index
More-processed index


NOTES: Average discrepancy is the average absolute deviation between annualized one-month rates
of change in the three indexes and the annualized
trend rate of increase or decrease in the all-items
index. The trend is a centered 36-month moving
average of annualized one-month changes. The
sample period is January 1959 to December 2010.
SOURCES: Bureau of Economic Analysis; author’s

is published by the
Federal Reserve Bank of Dallas. The views expressed
are those of the authors and should not be attributed
to the Federal Reserve Bank of Dallas or the Federal
Reserve System.
Articles may be reprinted on the condition that
the source is credited and a copy is provided to the
Research Department of the Federal Reserve Bank of
Economic Letter is available free of charge
by writing the Public Affairs Department, Federal
Reserve Bank of Dallas, P.O. Box 655906, Dallas, TX
75265-5906; by fax at 214-922-5268; or by telephone
at 214-922-5254. This publication is available on the
Dallas Fed website,

The bulk of “other meats” consists of items

such as hot dogs and lunch meats.

This paragraph, in fact, describes in words

the pricing behavior underlying the so-called
New Keynesian Phillips curve. See, for example,
“The New Neoclassical Synthesis and the Role
of Monetary Policy,” by Marvin Goodfriend
and Robert G. King, in NBER Macroeconomics
Annual 1997, Ben S. Bernanke and Julio Rotemberg, ed., Cambridge, Mass.: MIT Press, 1997,
pp. 231–96.

This argument is presented in some detail in

Richard W. Fisher
President and Chief Executive Officer
Helen E. Holcomb
First Vice President and Chief Operating Officer
Harvey Rosenblum
Executive Vice President and Director of Research
Robert D. Hankins
Executive Vice President, Banking Supervision
Director of Research Publications
Mine Yücel
Executive Editor
Jim Dolmas
Michael Weiss
Associate Editor
Jennifer Afflerbach
Graphic Designer
Ellah Piña

“Are Some Prices in the CPI More Forward Looking Than Others? We Think So,” by Michael F.
Bryan and Brent Meyer, Federal Reserve Bank of
Cleveland Economic Commentary, May 19, 2010,

Federal Reserve Bank of Dallas
2200 N. Pearl St.
Dallas, TX 75201