Consumers and Food Price Inflation
Randy Schnepf
Specialist in Agricultural Policy
Joe Richardson
Specialist in Social Policy
November 4, 2009
Congressional Research Service
7-5700
www.crs.gov
R40545
CRS Report for Congress
P
repared for Members and Committees of Congress
Consumers and Food Price Inflation
Summary
The heightened commodity price volatility of 2008 and the subsequent acceleration in U.S. food
price inflation raised concerns and generated many questions about farm and food price
movements by Members of Congress and their constituents. This report responds to those
concerns by addressing the nature and measurement of retail food price inflation. For a discussion
of the relationship between farm and retail prices, and the major factors influencing retail food
prices, see CRS Report R40621, Farm-to-Food Price Dynamics.
During the 1991 to 2006 period, U.S. food prices were fairly stable—annual food price inflation,
as measured by the U.S. Bureau of Labor Statistics (BLS) Consumer Price Index (CPI) for all
food (excluding alcoholic beverages), averaged a relatively low 2.5%. However, several
economic factors emerged in late 2005 that began to gradually push market prices higher for both
raw agricultural commodities and energy costs, and ultimately retail food prices. U.S. food price
inflation increased at a rate of 4% in 2007 and at 5.5% in 2008—the highest since 1990 and well
above the general inflation rate of 3.8%.
By late 2008 the inflationary price trends had reversed. Prices for many raw agricultural
commodities had already started to decline by late spring of 2008; however, owing to lags in the
adjustment process, it was not until November 2008 that monthly food price inflation fell to near
0%, and it actually declined from February through May 2009. USDA projects that annual food
price inflation will ease into a range of 2%-3% in 2009 (and to a slightly higher range of 3% to
4% in 2010) as lower commodity and energy costs combine with weak domestic and global
economies to reduce demand-driven inflation from 2008 levels.
The all-food CPI has two components—food-at-home and food-away-from-home. The food-at-
home CPI is most representative of retail food prices and is significantly more volatile than the
food-away-from-home index. The food-at-home CPI is projected in a range of 1% to 2% for
2009, compared with a 3.5% to 4.5% annual inflation rate for food-away-from home prices. This
difference is partially explained by the larger share of farm products in the final price of retail
foods than in food-away-from home. Farm product prices are, in general, substantially more
volatile than the other marketing and processing costs that enter into retail or ready-to-eat foods.
Many wages and salaries, as well as federal programs (including several domestic food assistance
programs), are linked to price inflation through escalation clauses in order to retain their
purchasing power. For households where income and federal benefits do not keep up with price
inflation, declines in purchasing power are real and immediate. However, even for households
with escalation clauses, a time lag usually occurs between the time the price inflation is measured
and the time when the wage or program benefit is adjusted upward to compensate.
The 2008-2009 global economic crisis—which encompassed higher retail prices, unemployment,
income loss, and lower effective household purchasing power—has manifested itself in higher
participation rates in the federal Supplemental Nutrition Assistance Program (SNAP), formerly
known as the Food Stamp Program. As of August 2009, participation in the SNAP was at an all-
time high of 36.5 million persons (or one in every 10 Americans). Other domestic food assistance
programs also have seen increased participation and costs. In response to advocates’ calls for
increased federal assistance, Congress approved (in February 2009) additional support for
domestic food assistance programs totaling $20.8 billion over 10 years via the stimulus package
(the American Recovery and Reinvestment Act of 2009; P.L. 111-5).
Congressional Research Service
Consumers and Food Price Inflation
Congressional Research Service
Consumers and Food Price Inflation
Contents
Introduction ................................................................................................................................ 1
Consumer Demand...................................................................................................................... 1
Price Responsiveness ............................................................................................................ 1
Income Responsiveness......................................................................................................... 2
Tastes and Preferences .......................................................................................................... 3
Summary .............................................................................................................................. 3
The Consumer Price Index (CPI)................................................................................................. 4
Historic Price Inflation Patterns....................................................................................... 4
Overall Inflation versus Core Inflation ............................................................................ 6
Consumer Income and Expenditures ........................................................................................... 7
Food as a Share of Consumer’s Budget ................................................................................. 7
At-Home versus Away-from-Home Consumption.................................................................. 9
International Comparisons................................................................................................... 11
Recent Food Price Inflation....................................................................................................... 12
Annual All-Food versus All-Items Price Inflation ................................................................ 12
The 2009 – 2010 Forecast for Annual Food Price Inflation ............................................ 13
The Recent Monthly All-Food Price Inflation Pattern .................................................... 14
At-Home versus Away-from-Home Food Price Inflation ..................................................... 15
Price Inflation by Major Food Categories ............................................................................ 17
Effect of High Prices ................................................................................................................. 23
Federal Spending for Domestic Food Assistance Programs .................................................. 23
Supplemental Nutrition Assistance Program (SNAP, formerly Food Stamps) ................. 24
Child Nutrition.............................................................................................................. 25
The WIC Program......................................................................................................... 26
TEFAP and Older Americans Act Programs................................................................... 26
Foreign Food Aid ................................................................................................................ 27
Figures
Figure 1. Annual Price Inflation, 1915-2008: All-Items vs. All-Food............................................ 5
Figure 2. Annual Price Inflation Since 1960: All-Items, All-Food, and Energy ............................. 6
Figure 3. Comparison of Real U.S. Disposable Personal Income (DPI) Per Capita and the
Share of DPI Spent on Food, 1930-2007 .................................................................................. 8
Figure 4. Average U.S. Food Expenditure Shares: At-Home vs. Away-from-Home .................... 10
Figure 5. Monthly Food Price Inflation Since 2000: At-Home vs. Away-from-Home ................. 10
Figure 6. Annual Price Inflation: All-Items vs. All-Food, 1985-2010F ....................................... 12
Figure 7. Annual Food Price Inflation Since 1997...................................................................... 13
Figure 8. Monthly Retail Food Price Inflation Trended Down in 2007, Spiked Sharply
Higher in Early 2008, Then Plummeted into 2009 .................................................................. 14
Figure 9. Food Price Inflation Since 1997: Monthly Change vs. 11-Mo. Moving Ave................. 15
Figure 10. Annual Food Price Inflation for All, At-Home, and Away-from-Home CPI ............... 16
Congressional Research Service
Consumers and Food Price Inflation
Figure 11. Monthly Retail Food Price Inflation Since July 2007 ................................................ 17
Figure 12. Monthly Retail Price Indexes: Various Major Food Groups....................................... 19
Figure 13. Monthly Retail Price Indexes: Beef, Pork, Poultry, and Eggs .................................... 20
Figure 14. Monthly Retail Price Indexes: Dairy, Fresh Milk, Cheese, and Ice Cream ................. 20
Figure 15. Monthly Retail Price Indexes: Fruits and Vegetables................................................. 21
Figure 16. Monthly Retail Price Indexes: Sugar, Fat & Oils, and Snacks.................................... 22
Figure 17. Monthly Retail Price Indexes: Coffee, Carbonated Beverages, and Prepared
Foods..................................................................................................................................... 22
Figure 18. CPI Relative Weights by Major Category.................................................................. 32
Tables
Table 1. Retail Price Change, Mean and Variability, by Historic Time Period............................... 5
Table 2. Average Household Food Expenditures in 2007 by Income Quintiles ............................. 8
Table 3. International Comparison of Food-at-Home Budget Shares, Selected Countries,
2007 ...................................................................................................................................... 11
Table 4. Annual Food Price Inflation Since 2005 ....................................................................... 16
Table 5. The Food-at-Home CPI by Category Since 2005 .......................................................... 18
Appendixes
Appendix. Major Statistical Indicators of Consumer Expenditures and Food Price
Changes ................................................................................................................................. 28
Contacts
Author Contact Information ...................................................................................................... 36
Congressional Research Service
Consumers and Food Price Inflation
Introduction
Everyone eats. As a result, everyone is affected to some degree by food price changes. This
makes understanding food price changes and their effects on consumers an important matter for
Congress. This report provides information on the current status of U.S. food prices, measuring
their changes, and how such changes relate to U.S. consumers.
The first section of the report, “Consumer Demand,” briefly reviews the major economic
concepts underlying consumer food behavior. The second section, the “Consumer Price Index”
describes how U.S. food price inflation rates have evolved since 1915, when federal price data
collection for inflation measuring purposes began. The third section, “Consumer Income and
Expenditures,” provides information on recent history and projections for U.S. food expenditure
shares relative to total household budget, with comparisons across income quintiles, as well as
internationally. The fourth section, “Recent Food Price Inflation,” examines retail food price
inflation, including a review and discussion of the level of food price inflation registered by the
consumer price index for all-food, at-home, and away-from-home food purchases as well as for
major food groups. A fifth section, entitled “Effect of High Prices,” discusses the impact that
rapid food price inflation can have on government food programs and the more vulnerable
consumer groups. Finally, an appendix section, “Major Statistical Indicators of Consumer
Expenditures and Food Price Changes,” describes the nature of the statistical indicators used to
measure consumer food prices and expenditures.
Each section may be read independently of the others. Thus, those readers that are concerned
primarily with the status of U.S. food price inflation may proceed directly to the sections entitled
“Historic Price Inflation Patterns” or “Recent Food Price Inflation.”
Consumer Demand
Consumer demand is influenced by economic factors—own-price, the price of close substitutes,
the price of complementary items, and household income—as well as by several non-economic
factors including tastes and preferences, family size, age of family members, geographic location,
shopping behavior, and lifestyle choices. Economists attempt to study and measure the nature of
consumer behavior in response to changes in prices, incomes, and household characteristics, with
an eye for understanding the potential social welfare outcomes that may result from price and
income changes across different socioeconomic groups. Policymakers, in turn, often attempt to
use that information to design and implement policies that mitigate the more deleterious effects of
price and income changes on consumers.
Price Responsiveness
In general, consumers will use less of any good if its price increases relative to other goods
(referred to as the pure “substitution effect” by economists). However, a consumer’s price
responsiveness is a matter of degree and is subject to the potential influence of disposable income
as well as other non-price factors such as those listed in the preceding paragraph.
Under most circumstances, the availability of many close substitutes is likely to make consumers
more sensitive or responsive to price changes, because they have the opportunity to switch to
Congressional Research Service
1
Consumers and Food Price Inflation
similar alternatives. In contrast, a lack of substitutes may give the consumer little choice but to
continue to purchase the available good, even as its price rises, especially if it is deemed a
necessity. Strong ethnic or cultural tastes and preferences may endear a person to a particular food
type such that he or she will continue to purchase that food as its price rises even in the presence
of abundant substitutes (for example, ethnic groups that are accustomed to eating rice at every
meal may be reluctant to switch to bread or potatoes even if the price of rice rises relative to those
other foods).
Rapid or unexpected changes in retail food prices will impact some consumers more than others
depending on income levels and the importance of the affected food items in consumers’ budgets.
In general, if an item represents a very small portion of the consumer’s budget (for example,
consider salt), then a consumer is less likely to respond to a price change. Basic food staples such
as bread, potatoes, pasta, and rice tend to take smaller shares of the consumer’s food budget
(relative to meat, dairy products, fruits and vegetables, and more processed food products) and, as
a result, consumers are less responsive to a change in their price. In contrast, high-valued food
items such as expensive cuts of meat or seafood probably represent more costly (and infrequently
purchased) delicacies for most households. As a result, most households will tend to be far more
responsive to changes in the prices of such high-valued products than for basic staples. Often a
price change for an item within a specific food group1 may result in consumers switching to
lower-quality items within that food category—the classic example being a switch from steak to
hamburger when meat prices rise. In contrast, a widespread price rise across all food groups may
engender substantial reshuffling of consumer food budget allocations as households try to meet
their nutritional goals with their limited budgets.
Of course, the absolute size of a consumer’s disposable income is also important in determining
actual purchasing power. For households with smaller incomes, the food budget itself is likely a
larger portion of total household expenditures, and such households are likely to be more
responsive to price changes across all food categories than are higher-income households.
In summary, lower-income consumers who spend a significant share of their household budget on
food are likely to be impacted more severely by rising food prices (and are likely to be more
responsive to price changes) than high-income consumers with lower food budget shares.
Income Responsiveness
A household’s absolute level of disposable income (and, to a lesser degree, wealth) directly
affects its ability to respond to price changes. As a result, as household incomes grow, consumers
often opt for more expensive or higher-quality selections of foods than are presently in their food
budget, or may experiment by trying new or unfamiliar foods. For example, as incomes increase
in less-developed countries, it is common to see per-capita expenditures on meat and dairy
products increase. In contrast, when incomes decline, consumers tend to pull back from more
expensive options. If the income decline is severe and is perceived as permanent or long-lasting,
consumers may make substantial changes to their food budget choices.
In the aggregate, household consumption behavior in response to perceived income changes (if
persistent and widespread) may affect a country’s agricultural production or trade patterns. As a
1 Examples of food groups include meat, dairy products, bakery goods, fruits, or vegetables.
Congressional Research Service
2
Consumers and Food Price Inflation
result, it is important for policymakers to monitor household wealth and income levels and
distribution for unexpected shifts that may have important economic consequences.
Economists call the relationship between changes in consumer income and the quantity of an item
purchased an Engel curve. This relationship is used by economists to classify goods.
• For a normal good, consumers buy more of it as their incomes increase, but at a
decreasing rate such that its average budget share declines for higher income
levels.
• For a luxury good, consumers buy more of it as their incomes increase and at an
increasing rate such that its budget share increases at higher income levels.
• For an inferior good, consumers buy less of it as their incomes increase.
Of course, different goods will be classified differently by different people since tastes and
preferences also are important. However, with respect to the overall food budget, in the aggregate
certain behavioral norms are expected. Engel’s law is the idea (largely validated by data with
some minor exceptions) that food, in general, is a normal good, so that the budget share spent on
food declines as a consumer’s income rises. While Engel’s law is generally observable for
individual households, it tends to hold best in the aggregate—that is, when considering an entire
population. To the extent that this “law” holds, then the proportion of a nation’s income spent on
food serves as a good index for international comparisons of relative consumer welfare.
Tastes and Preferences
Non-economic factors such as cultural or ethnic preferences may determine both the share of a
particular food product in the household’s budget (e.g., rice represents a larger share of per-capita
expenditure in most Asian households than in most European households at similar income
levels) as well as a household’s responsiveness to a change in the price of a particular product.
Dietary needs also change with age. For example, young children and adolescents generally need
both more calories and a higher portion of protein-based calories to meet nutritional demands of
rapid physical growth and high activity levels. Populations or households with a large share of
individuals from this demographic stratum are more likely to consume larger per-capita portions
of meat and dairy products than an older, more mature and sedentary population would. As a
result, population demographics such as household composition, size, and age structure often play
an important role in consumer price sensitivity and income responsiveness. In increasingly
affluent societies, lifestyle choices, when complemented with sufficient purchasing power, can
also play an influential role in household food purchases.
Summary
For households with low disposable income levels where food expenditures are a large share of
the budget, rising food prices result in greater responsiveness and may force more difficult
budgetary tradeoffs than in higher-income households with smaller food-budget shares. Of course
the opposite effect is true during periods of falling prices. However, each household’s price and
income effects also are influenced by its particular set of non-economic characteristics.
Congressional Research Service
3
Consumers and Food Price Inflation
The Consumer Price Index (CPI)
The CPI is perhaps the most widely reported measure of U.S. price inflation.2 The CPI is used
both as an economic indicator of retail price inflation and as a means of adjusting current-period
values for inflation. The “All-Items” CPI is the index most often referred to (i.e., the headline
CPI) for representing consumer price inflation. It is generally divided into eight major spending
categories, including a “Food and Beverage” category (Figure 18).
The CPI category of “Food and Beverages” is composed of two major subcategories: “All-Food”
(which has a relative weight of 14.6% in the all-items CPI), and “Alcoholic Beverages” (1.1%).
The all-food CPI is the principal indicator of consumer food price changes. The all-food CPI can
be subdivided into the “Food-at-Home” and “Food-Away-from-Home” categories. The food-at-
home CPI reflects changes in the prices of foods consumed at home. As such it is the principal
indicator of changes in retail food prices in the United States. The food-away-from-home CPI
reflects changes in the prices of foods purchased outside of the home, primarily at eating and
drinking places such as restaurants and other eating establishments. However, it also includes
price changes for ready-to-eat foods purchased at hotels and motels, recreational places and
sporting events, vending machines, and school and work cafeterias.
Historic Price Inflation Patterns
Over time, the all-food and all-items CPIs have moved together, although the all-food CPI has
been consistently more variable than the all-items CPI (Figure 1 and Table 1). Prior to 1960, both
of these indexes exhibited higher average inflation rates and more volatility than in recent years.
During the 1914-1920 period, both price indexes recorded double-digit annual inflation. Food
inflation hit its all-time high of 28.7% in 1917 (Figure 1). All-items price inflation peaked a year
later at 18%. Just four years later retail prices entered a prolonged deflationary period, starting in
1921 with a deflationary plunge of -24.2% for all-food and -10.5% for all-items, that lasted until
1941 when war time shortages finally renewed retail price inflation.
The variability of the overall CPI and its individual components is important because uncertainty
about price changes makes planning more difficult—whether it is the investment planning of a
business, the meal planning of a household, or the policy intervention planning of a government
agency.
During the 1941-1960 period, price inflation remained extremely volatile, alternating between
spikes of inflation and steep disinflationary (i.e., deflationary) drops. It was not until 1960 that
retail prices stabilized with tolerably mild inflation. However, this proved short-lived as the 1970s
saw a return to sharp price spikes generated by an energy crisis and rapid, unexpected shifts in
global crop supply and demand. 3 By the early 1980s, retail price inflation had returned to modest
levels below 5%. Since 1983 retail prices, as measured by the all-items and the all-food CPIs,
2 For more information on statistical issues related to the measurement of price inflation, see the Appendix at the end of
this report. See also CRS Report RL30074, The Consumer Price Index: A Brief Overview, by Brian W. Cashell.
3 For a brief discussion of 1970s energy markets, see CRS Report R40187, U.S. Energy: Overview and Selected Facts
and Numbers, by Carl E. Behrens and Carol Glover. For a brief discussion of 1970s agricultural markets, see May
Peters, Suchada Langley, and Paul Westcott, “Agricultural Commodity Price Spikes in the 1970s and 1990s,” Amber
Waves, ERS, USDA, March 2009, at http://www.ers.usda.gov/.
Congressional Research Service
4
Consumers and Food Price Inflation
have been relatively low and relatively stable, except for temporary surges in 1989-1990 and
again in 2007-2008.
Figure 1. Annual Price Inflation, 1915-2008: All-Items vs. All-Food
30%
20%
All-Food
All-Items
10%
e
g
n
a
l Ch
a
0%
u
1915
1925
1935
1945
1955
1965
1975
1985
1995
2005
n
An
% -10%
-20%
-30%
Source: Department of Labor, Bureau of Labor Statistics.
Notes: The percent change is calculated from the annual average CPI for successive years.
Table 1. Retail Price Change, Mean and Variability, by Historic Time Period
(all data—mean and standard deviation (SD)—are percentages)
1914-1920
1921-1941
1941-1960
1960-1983
1983-2008
CPI
Series
Mean SD Mean SD Mean SD Mean SD Mean SD
All-Items
10.8 7.4 -1.4 4.5 3.9 4.1 5.3 3.6 3.1 0.9
Energy
— — — — — — 6.9 9.1 3.6 7.1
All-Food
11.6 9.6 -1.8 9.0 4.9 7.2 5.2 4.1 3.1 1.2
At-Home
— — — — — — 5.0 4.5 3.1 1.6
Away-from-Home — — — — — — 6.0 3.1 3.2 0.9
Corea
— — — — — — 5.1 3.3 3.2 1.1
Source: Calculations were made by CRS based on BLS CPI data.
Notes: ”—“ = not available. The mean is the average annual price change for each period. The standard
deviation (SD) is a measure of dispersion around the mean value for each period. Plus or minus one (two) SD
captures 68.2% (95.4%) of the variation around the mean value for each period. The mean and SD have been
calculated using the annual percent change data for each of the five different periods. A larger SD implies greater
variability, for example, all-food price inflation has shown a clear pattern of declining variability as the SD has
fallen from 9.6% during the 1914-20 period to 4.1% during 1960-83 (more than halving the variability), and finally
to 1.2% during 1983-2008 (a further decline of over 70% in variability).
a. The “Core” price index is the all-items CPI without the energy and all-food components.
Congressional Research Service
5
Consumers and Food Price Inflation
Overall Inflation versus Core Inflation
Many economists and policymakers believe that the food and energy components of the CPI are
volatile and subject to shocks not easily dealt with through government monetary policy. In
response, the BLS also reports another price index, referred to as the “core” index because it
removes the food and energy price components from the all-items CPI.4 The so-called core CPI is
thought to be a useful measure of underlying trend inflation in the short run. According to BLS
data, the food component of the CPI, although more volatile than the overall CPI, is still
substantially less volatile than the energy component (Figure 2 and Table 1).
Figure 2. Annual Price Inflation Since 1960: All-Items, All-Food, and Energy
30%
Energy
20%
nge
Food
10%
Core
l Cha
a
u
0%
Ann
%
1960
1970
1980
1990
2000
2010
-10%
-20%
Source: Department of Labor, Bureau of Labor Statistics.
Since 1960, the energy price index has been a more volatile component of the all-items CPI than
the food price index by a substantial margin. For example, the energy price inflation standard
deviation (SD) of 9.1% was more than double the all-food SD of 4.1% during the 1960-1983
period, and nearly six times larger since 1983 (7.1% versus 1.2%).
Since 1983 both the all-food and the all-items CPIs have been lower (in terms of average values)
and substantially more stable (in terms of SDs) than during the preceding seven decades. In
contrast, the energy price index has been substantially more volatile since 1983 (although at a
lower mean level) than it was during the preceding two decades. This is an important point
because the energy price index has seen its weighted share of the CPI gradually increase over
time and, although energy’s current weight share of 7.6% is only about half that of the food
weight share of 14.6%, energy price inflation is far more insidious than food inflation to the
extent that energy costs figure in the retail price of practically every other component of the CPI.
4 For more information, see CRS Report RS22705, Inflation: Core vs. Headline, by Marc Labonte.
Congressional Research Service
6
Consumers and Food Price Inflation
Consumer Income and Expenditures
A household allocates its available income across a range of expenditure, savings, and investment
choices. As mentioned earlier, food expenditures as a share of a household’s total budget are an
indicator of sensitivity (or vulnerability) to unexpected food price changes. At the national level,
food budget share (via Engel’s law) can be used as a general indicator of welfare among nations.
Food as a Share of Consumer’s Budget
According to Bureau of Economic Analysis (BEA) estimates, in 2007 total U.S. disposable
personal income (DPI) was $10,171 billion, or $33,707 per capita.5 Of these estimates, ERS
calculates that, on average, 9.8% of disposable personal income was spent on food.6
Figure 3 shows the evolution of the average food budget share against U.S. DPI per capita in
constant 1982-1984 average dollars. By both measures (food budget share and real DPI per
capita), U.S. consumers have seen their “well-being” improve substantially over the past 70
years. (Note that these are national averages that ignore any potential distribution issues.) The
U.S. food share of real DPI has fallen from a high of 31.4% in 1933 to about 10% since 2000,
while the average DPI per capita (in 1982-1984 dollars) has risen from $2,824 in 1933 to over
$16,000 by 2006.
When measured as a share of average total consumer expenditures of $49,638 per household
(based on Consumer Expenditure Survey (CES) data, Table 2), average food outlays of $6,133
per household accounted for 12.4% of total spending in 2007.7 As mentioned earlier, the
difference between the two estimates of food budget share (9.8% based on DPI versus 12.4%
based on CES total expenditures) is due to how disposable income and food expenditures are
calculated for each of these indicators (discussed more fully in the appendix).
The estimated food share of household expenditures (Table 2) varied across income quintiles in
the United States, in accordance with Engel’s law—that is, each succeeding higher income
quintile increased its absolute expenditures on food (in dollar terms), but at a decreasing rate such
that the food budget share declines across higher quintiles. For example, the lowest 20% of U.S.
households spent $3,035 on food, or 14.8% of their average total expenditures of $20,471 in
2007. The budget food outlay increases in absolute dollars, while the food budget share declines
across income quintiles until the wealthiest quintile, where households spent an average of
$10,444 on food, or 10.8% of their total budget of $96,752.
5 “Personal Income and Outlays: January 2009,” News Release BEA 09-06, March 2, 2009. The estimates for 2008
total and per-capita DPI were $10,645.8 billion and $34,958, respectively. However, comparable food expenditure data
for 2008 were not yet available as of March 2009.
6 The DPI and DPI-food-share estimates are for 2007 from Table 7, Food Expenditure Tables, Food CPI, Prices and
Expenditures Briefing Room, ERS, USDA, at http://www.ers.usda.gov/Briefing/CPIFoodAndExpenditures/Data/.
7 CES data are calculated “per consumer unit” which is described as “similar to a household” by BLS.
Congressional Research Service
7
Consumers and Food Price Inflation
Figure 3. Comparison of Real U.S. Disposable Personal Income (DPI) Per Capita and
the Share of DPI Spent on Food, 1930-2007
$21,000
35%
$18,000
ta
30%
pi
a
Food Budget
r c
Share as % of DPI
$15,000
25%
ta
I) pe
pi
P
a
D
(
r c
e $12,000
m
20%
I pe
o
P
c
D
l In
a
of
n
$9,000
re
o
15%
a
rs
h
e
S
P
le
$6,000
b
10%
a
Food
pos
is
Real DPI per capita
$3,000
D
5%
in 1982-84 dollars
$0
0%
1930
1940
1950
1960
1970
1980
1990
2000
Source: “Food CPI, Prices and Expenditures Briefing Room,” Food Expenditure Tables, ERS, USDA.
Notes: Real DPI is BEA nominal DPI series deflated by the al -items CPI with base 1982-1984 = 100; real food
expenditures are the ERS series of food expenditures deflated by the al -food CPI with base 1982-1984 = 100.
Table 2. Average Household Food Expenditures in 2007 by Income Quintiles
Food-Away-
Income Quintile
Total
All Food
Food-at-Home
from-Home
Expenditures per household
Highest 20%
$96,752
$10,444
$5,265
$5,179
Fourth 20%
$57,285
$7,242
$3,980
$3,262
Third 20%
$42,447
$5,682
$3,333
$2,349
Second 20%
$31,150
$4,260
$2,741
$1,519
Lowest 20%
$20,471
$3,035
$2,005
$1,030
Average Outlay ($)
$49,638
$6,133
$3,465
$2,668
Share of Expenditures: Total and (All Food)
HIghest 20%
100%
10.8% (100%)
5.4% (50.4%)
5.4% (49.6%)
Fourth 20%
100%
12.6% (100%)
6.9% (55.0%)
5.7% (45.0%)
Third 20%
100%
13.4% (100%)
7.9% (58.7%)
5.5% (41.3%)
Second 20%
100%
13.7% (100%)
8.8% (64.3%)
4.9% (35.7%)
Lowest 20%
100%
14.8% (100%)
9.8% (66.1%)
5.0% (33.9%)
Average Outlay (%)
100%
12.4% (100%)
7.0% (56.5%)
5.4% (43.5%)
Source: Table 1. Quintiles of before-tax income: Average annual expenditures and characteristics, Consumer
Expenditure Survey, 2007, BLS, Dept. of Labor, at http://www.bls.gov/cex/2007/Standard/quintile.pdf.
Congressional Research Service
8
Consumers and Food Price Inflation
Another clear pattern that emerges from the CES data (Table 2) is the propensity to spend more
of the food budget on away-from-home food consumption at higher income levels, both in
absolute dollars and as a share of the food budget (refer to the percentages in parentheses in Table
2), thereby meeting the description of a luxury good. Or, otherwise stated, lower-income U.S.
households tend to spend a larger share of their food budget on at-home consumption and are thus
more vulnerable to unexpected retail food price increases (this is discussed further in the next
section).
At-Home versus Away-from-Home Consumption
U.S. households have shown a strong propensity over time to increase their share of annual food
consumption outside of the home (Figure 4). This tendency is associated with increasing per-
capita disposable income as mentioned above. It is also associated with increasing female
participation in the labor force, more two-earner households, increased advertising and promotion
by large food-service chains, increasing time constraints on household members (e.g., longer
commutes, increased work hours and less leisure time, etc.), the smaller size of U.S. households,
and the increased availability of relatively low-cost fast food establishments.8
With the exception of a brief period following the end of World War II, the portion of the national
food budget spent on food consumption away from the home has steadily increased from 9% in
1900 to an estimated 49% in 2008. This phenomenon has important implications for consumer
responsiveness to price and income changes, as well as for household nutrition.
The prices of food-at-home purchases are significantly more volatile than are prices of food-
away-from-home purchases (Table 1 and Figure 5). ERS research suggests that away-from-home
expenditures are typically higher for single-person households and households containing
multiple adults without living-at-home children.9 By implication, households with living-at-home
children typically rely more on at-home food consumption (as a share of their budget) and are
thus more vulnerable to the normally higher price variability associated with retail food prices.
Although increased food-away-from-home expenditure is associated with higher income (both in
absolute terms and as a share of the household food budget), it is not always a luxury item. A
partial key to understanding how increasing food-away-from-home consumption may impact
consumer behavior is the extent to which such consumption is a choice (for example, made in the
evenings or on weekends during leisure hours) or more of an obligation (made during work
hours), as well as the extent to which a consumer has alternative dining choices when eating out
(for example, subsidized cafeteria meals are often available at schools or in large institutional
work settings).
8 For a discussion of this issue, see “Food Away From Home,” Diet Quality and Food Consumption Briefing Room,
ERS, USDA, at http://www.ers.usda.gov/Briefing/DietQuality/.
9 Hayden Stewart, Noel Blisard, Sanjib Bhuyan, and Rodolfo M. Nayga, Jr., The Demand for Food Away From Home:
Full-Service or Fast Food? AER No. 829, ERS, January 2004.
Congressional Research Service
9
Consumers and Food Price Inflation
Figure 4. Average U.S. Food Expenditure Shares: At-Home vs. Away-from-Home
100%
At-Home
80%
60%
40%
Away-From-Home
20%
0%
1900
1920
1940
1960
1980
2010
Source: Table 7, “Food CPI, Prices and Expenditures Briefing Room,” Food Expenditure Tables, ERS, USDA,
available at http://www.ers.usda.gov/Briefing/CPIFoodAndExpenditures/Data/.
Figure 5. Monthly Food Price Inflation Since 2000: At-Home vs. Away-from-Home
1.5%
At-Home CPI
1.0%
nge
0.5%
Cha
hly
nt
o 0.0%
M
2000
2002
2004
2006
2008
2010
%
-0.5%
Away-From-Home CPI
-1.0%
Source: BLS, U.S. Dept. of Labor.
Notes: Month-to-month inflation measured as the percent change in the monthly seasonally adjusted CPI for
each index.
Congressional Research Service
10
Consumers and Food Price Inflation
International Comparisons
The Economic Research Service (ERS) includes in its food expenditure data series a comparison
of food budget shares (based on at-home food expenditures) for over 70 countries. Table 3
includes samples of countries from the ERS database ranked in terms of their at-home food
expenditure budget shares, from smallest to largest. These data suggest that, on average, the
United States has achieved a higher level of social welfare (based on this particular indicator)
than any of the other countries in the database.
Table 3. International Comparison of Food-at-Home Budget Shares, Selected
Countries, 2007
Total Expenditures
At-Home Food
At-Home
Country
per capita
Expenditures per capita
Food Share
United States
$33,947
$1,935
5.7%
United Kingdom
$27,272
$2,351
8.6%
Canada $23,993
$2,203
9.2%
Germany $21,832
$2,497
11.4%
France $23,184
$3,176
13.7%
Italy $21,315
$3,091
14.5%
Japan $18,737
$2,742
14.6%
South Korea
$10,287
$1,589
15.4%
Poland $6,675
$1,375
20.6%
Mexico $5,579
$1,348
24.2%
Brazil $4,183
$1,028
24.6%
Russia $4,363
$1,251
28.7%
India $536
$174
32.4%
China $882
$308
34.9%
Philippines $1,136
$424
37.3%
Egypt $1,279
$496
38.8%
Nigeria $677
$273
40.3%
Algeria $1,119
$491
43.9%
Indonesia $1,138
$520
45.7%
Pakistan $634
$290
45.7%
Source: Table 97, “Food CPI, Prices and Expenditures Briefing Room,” Food Expenditure Tables, ERS, USDA,
available at http://www.ers.usda.gov/Briefing/CPIFoodAndExpenditures/Data/.
The food budget share is only one indicator of national welfare, and it ignores any unfavorable
distribution of the food expenditure share (should any exist). Referring back to Table 2 for the at-
home food budget share for U.S. income and total expenditure quintiles based on CES data, it
would appear that even the lowest 20% of U.S. households, on average, spend less than 10% of
their budgets on at-home food consumption and thus appear relatively well-off in food terms
based on this particular international standard. Readers should note that this cursory assessment is
aggregate in nature and does not exclude the possibility that there are food-deficient individuals
Congressional Research Service
11
Consumers and Food Price Inflation
within the lowest 20% quintile of the U.S. population. According to ERS, in 2007, an estimated
11.1% of U.S. households were food-insecure at least some time during the course of the year.10
Recent Food Price Inflation
This section provides a discussion of observed food price inflation in recent years based on CPI
data. It is important to remember that the various CPI categories discussed here are indicative of
price changes at the retail level in U.S. urban settings. As such, they are indicative of the prices
faced by most consumers living in the United States (approximately 87% of U.S. consumers are
covered by the CPI data collection process).
Annual All-Food versus All-Items Price Inflation
As a general rule, the all-item and all-food CPIs tend to move together. Following a relatively
tumultuous period of price inflation in the late 1980s, both price indexes entered an extended
period of relative stability. From 1991 through 2006, the all-food CPI measured average annual
inflation of 2.5%, compared with 2.7% annual average all-items price inflation.
Figure 6. Annual Price Inflation: All-Items vs. All-Food, 1985-2010F
7.5%
All-Food CPI
5.0%
All-Items CPI
nge
Cha
l 2.5%
nua
An
% 0.0%
-2.5%
1985
1990
1995
2000
2005
2010
Source: Historical data, unadjusted for seasonality, For 1985-2008, data are from the Dept. of Labor, BLS. Data
for 2009 and 2010: al -food CPI is from USDA, ERS, as of Oct. 23, 2009; al -items CPI is from Global Insights,
U.S. Economic Outlook, October 2009.
Notes: The percent change is calculated from the annual average CPI for successive years.
10 For a discussion of food insecurity in the United States, see Food Security in the United States Briefing Room, ERS,
USDA, at http://www.ers.usda.gov/Briefing/FoodSecurity/.
Congressional Research Service
12
Consumers and Food Price Inflation
However, several economic factors emerged in late 2005 that began to gradually push market
prices higher for both raw agricultural commodities and energy costs.11 These factors included
rising consumer incomes, not just in the United States but globally, which sparked demand for
meat and dairy products, food and feed grains, and raw materials ranging from minerals and
metals to coal and petroleum. In 2007, U.S. food price inflation reached 4% (Figure 7), the
highest since 1990.
In early 2008, monthly food price inflation appeared to be accelerating (Figure 8). In July 2008,
the month-to-month food price change was 0.96% (equivalent to an annual rate of 12.1% if
sustained for the entire year). Monthly retail food price inflation responds with a lag of several
months to price changes in raw commodity markets. For the entire year, 2008 food prices rose by
5.5%, well above the all-items CPI of 3.8% (Figure 6). However, by late 2008 retail price trends
had reversed and began following prices for raw agricultural commodities, which had already
started to decline by late spring of 2008. By November 2008 monthly retail food price inflation
had fallen to near 0%, then actually declined during February through May 2009.
Figure 7. Annual Food Price Inflation Since 1997
6%
5.5%
All-Food CPI
e
4.0%
4.0%
g
n 4%
a
3.4%
h
3.2%
3.0% 3.0%
l C
a
2.6%
u
2.4% 2.4%
n
2.2%
2.3%
n
2.1%
2.2%
2.0%
1.8%
2%
% A
0%
1997
1999
2001
2003
2005
2007
Forecast ranges
2009 2010
Source: Historical data (1985-2008) are from the Dept. of Labor, BLS. Forecast for the 2009 al -food CPI is
from USDA, ERS, October 23, 2009.
The 2009 – 2010 Forecast for Annual Food Price Inflation
In 2009, both price indexes, the all-food and the all-items, are forecast sharply lower (Figure 6);
however, the all-items CPI is expected to be subject to much stronger deflationary pressures than
the all-food CPI. As a result, the disparity between the two indexes is forecast to widen as the all-
items CPI falls at an annual rate of -0.4%,12 while U.S. food price inflation in 2009 is projected
11 For more information on the factors behind the sharp run-up in global commodity prices in the first half of 2008, see
CRS Report RL34474, High Agricultural Commodity Prices: What Are the Issues? by Randy Schnepf.
12 The all-items CPI forecast is from U.S. Executive Summary, Global Insights, October 2009.
Congressional Research Service
13
Consumers and Food Price Inflation
by ERS to be in the 2% to 3% range (Figure 7).13 Sharply lower commodity and energy costs are
expected to combine with weak domestic and global economies to reduce inflationary pressures
from 2008 levels for both of these price indexes. As a result, pressure on retail food prices is
expected to subside, resulting in low to moderate food price inflation in 2009. However, by 2010
both the U.S. and global economies are expected to resume growing and, as a result, are forecast
to reverse the deflationary price pattern (Figure 6).
The Recent Monthly All-Food Price Inflation Pattern
When the all-food CPI is adjusted for seasonal variations and expressed on a monthly basis, four
strong patterns can be seen to have emerged since January 2007 (Figure 8). First, a mildly
volatile, downward pattern of monthly price change persisted throughout 2007. This was
followed by a highly volatile, upward pattern of price inflation during the first nine months of
2008, at which point monthly price changes declined sharply into December, when they actually
fell below the preceding month’s level (i.e., the monthly all-food CPI deflated or become
negative) for the first time since November 2006. Monthly food price changes continued their
deflationary trend hitting bottom in May of 2009. Finally, food prices have shown an upward
trend since May 2009.
Figure 8. Monthly Retail Food Price Inflation Trended Down in 2007, Spiked Sharply
Higher in Early 2008, Then Plummeted into 2009
1.00%
All-Food CPI
0.75%
ge
n
0.50%
Cha
hly
nt 0.25%
o
M
%
0.00%
-0.25%
2007
2008
2009
2010
Source: Data, adjusted for seasonality, are from the Department of Labor, BLS.
Notes: The percent change is calculated from the CPI for successive months. Multiply any given month’s value
by 12 to approximate the annual inflation rate experienced during that particular month, without compounding.
13 ERS updates its food price forecast monthly. For the most current ERS food price forecast with related data tables
and a discussion of the issues, see “Analysis and Forecasts of the CPI for Food,” Food CPI and Expenditures Briefing
Room, ERS, USDA, at http://www.ers.usda.gov/Briefing/CPIFoodAndExpenditures/consumerpriceindex.htm.
Congressional Research Service
14
Consumers and Food Price Inflation
Month-to-month retail food price declines continued into 2009 even though farm prices had
stabilized in early 2009. This is because food processors and retailers are traditionally slow to
pass on price decreases that they experience at the wholesale level for several reasons, including
substantial inherent operating risk associated with volatile markets.14
The same monthly all-food price inflation measures (also adjusted for seasonality) are presented
in Figure 9, but for a longer time period and accompanied by their 11-month moving average
(MA).15 The MA series reveals a strong upward inflationary trend that began at the end of 2005
and persisted through June 2008. Since June 2008, the MA has moved strongly downward. Retail
food prices are clearly responding to the deflationary pressures associated with the collapse of
U.S. and international commodity prices that occurred in mid-2008, as well as to the crisis that
emerged in late 2008 in U.S. and global financial markets.
Figure 9. Food Price Inflation Since 1997: Monthly Change vs. 11-Mo. Moving Ave.
1.00%
11- month
Moving Average
0.75%
Monthly Change
All-Food CPI
0.50%
nge
a
h
C
hly
0.25%
ont
M
%
0.00%
-0.25%
-0.50%
1997
1999
2001
2003
2005
2007
2009
Source: Data, adjusted for seasonality, are from the Department of Labor, BLS.
At-Home versus Away-from-Home Food Price Inflation
As shown earlier (Figure 5), at-home food prices are substantially more volatile than away-from-
home food prices (see also Table 1, where the at-home food price SD of 1.6% compares with a
0.9% SD for away-from-home food prices since 1983). This volatility is apparent, even when
using a shorter time period (Table 4 and Figure 10). It is not surprising, then, that at-home food
prices are forecast to make a steeper decline in 2009 and a sharper reversal in 2010, than either
all-food or food-away-from-home.
14 These issues are discussed in more detail in CRS Report R40621, Farm-to-Food Price Dynamics, by Randy Schnepf.
15 Moving averages are used to reveal underlying patterns or trends that can otherwise be hidden by a substantial
amount of month-to-month variation in price movements.
Congressional Research Service
15
Consumers and Food Price Inflation
When displayed in terms of monthly price changes, the pattern exhibited by at-home food price
inflation (Figure 11) appears very similar to the pattern for all-food price inflation (Figure 8),
although the at-home food price movements are more extreme. In contrast, monthly away-from-
home price inflation is much more stable. Note that both at-home and away-from-home monthly
price inflation show a distinct downturn beginning in July 2008.
Figure 10. Annual Food Price Inflation for All, At-Home, and Away-from-Home CPI
(historical data since 1997, with forecast for 2009 and 2010)
7%
At-Home
6%
All Food
5%
ange
h
C
al 4%
nnu
Away-From-Home
3%
A
%
2%
1%
1997
1999
2001
2003
2005
2007
2009
Source: See source info from Table 4.
Table 4. Annual Food Price Inflation Since 2005
Annual % Change
Category Weights
2005
2006
2007
2008
2009F
2010F
All-Items
na 3.4 3.2 2.8 3.8 -0.4 1.4
Al -Food
100%
2.4
2.4
4.0
5.5
2.0 to 3.0 3.0 to 4.0
Food-at-Home
55.8%
1.9
1.7
4.2
6.4
1.0 to 2.0 2.5 to 3.5
Food-Away-from-Home
44.3%
3.1
3.1
3.6
4.4
3.5 to 4.5 3.5 to 4.5
Source: Historical data (unadjusted for seasonality) for 2005 to 2008 are from BLS, Dept of Labor; 2009 and
2010 forecasts are from ERS, USDA, as of Oct. 23, 2009, at http://www.ers.usda.gov/Briefing/
CPIFoodAndExpenditures/Data/cpiforecasts.htm. The al -items CPI forecasts for 2009 and 2010 are from Global
Insights, U.S. Economic Outlook, October 2009.
Notes: BLS data are as of February 2009 using 2005-2006 weights for U.S. City Average (CPI-U). Annual
percent changes are calculated from annual average indexes.
Congressional Research Service
16
Consumers and Food Price Inflation
Figure 11. Monthly Retail Food Price Inflation Since July 2007
(at-home vs. away-from-home)
1.5%
At-Home CPI
1.0%
nge
a 0.5%
h
C
Away-From-
Home CPI
hly
0.0%
% Mont
-0.5%
-1.0%
Jul-07
Jan-08
Jul-08
Jan-09
Jul-09
Jan-10
Source: Historical data (adjusted for seasonality) are from BLS, Dept of Labor.
Notes: The percent change is calculated from the CPI for successive months. Multiply any given month’s value
by 12 to approximate the annual inflation rate experienced during that particular month, without compounding.
Price Inflation by Major Food Categories
The relatively high food price inflation of 2007 and 2008 was not felt evenly across all food
groups, but varied widely in terms of both the timing and the relative magnitude of inflation. In
2008, the inflation rate for at-home food was 6.4% (Table 5); however, only the “fruits and
vegetables” category experienced a similar level of price inflation (6.2%). Several product
categories experienced substantially higher inflation, including egg prices which were up 14% in
2008 after having soared by 29% in 2007; fats and oils were up 13.8%, cereals and bakery
products up 10.2%, snacks up 8.1%, and dairy products up 8%. In contrast, the broad price index
of meats, poultry, and fish was up only 4.2%, and non-alcoholic beverages (including carbonated
and non-carbonated drinks and juices) were up 4.3%.
The demand-side influences of income growth (and decline) and the global financial crisis that
emerged in late 2008 have already been discussed briefly. On the supply side, food price inflation
is the result of dynamic forces that occur both at the farm where the raw agricultural ingredients
for retail food items are produced, and along the marketing chain as the farm output is
transformed and moved to the retail customer. An array of costs are layered on top of the price of
the raw agricultural commodity, including handling, transportation, storage, and processing, as
well as the insurance, financing, and advertising costs necessary to move the product to the retail
customer. The relative importance of these marketing costs varies widely for different retail food
products depending on the degree of processing and transformation (i.e., cleaning, packaging,
shipping, advertising, etc.). As a result, economic forces such as higher energy costs or increased
labor rates do not impact all food categories equally.
Congressional Research Service
17
Consumers and Food Price Inflation
Table 5. The Food-at-Home CPI by Category Since 2005
Weightsa
Annual Percent Change
Category
%
% 2005 2006 2007 2008 2009F
2010F
Food
at
Home
100%
1.9 1.7 4.2 6.4
1.5-2.5 2.5-3.5
1-Cereal & Bakery Products
14.1%
1.5
1.8
4.4
10.2
3.0 to 4.0
2.5 to 3.5
Cereals and productsb
4.5%
0.3
0.3
3.9
10.1
Na Na
Bakery productsc
9.4%
2.0
2.6
4.5
10.3
Na
Na
2-Meats; Poultry; & Fish
21.9%
2.4
0.8
3.8
4.2
1.0 to 2.0
1.5 to 2.5
Beef & veal
6.8%
2.6
0.8
4.5
4.5
-1.0 to 0.0
1.0 to 2.0
Pork
4.2%
2.0
-0.3
2.0
2.3
-1.0 to 0.0
0.5 to 1.5
Poultry
4.1%
2.0
-1.8
5.2
5.0
2.0 to 3.0
1.0 to 2.0
Fish & seafood
3.7%
3.0
4.7
4.6
6.0
4.0 to 5.0
4.0 to 5.0
3-Eggs
1.3%
-13.7
4.9
29.2
14.0
-16 to -15
2.0 to 3.0
4-Dairy & Products
11.2%
1.2
-0.6
7.4
8.0
-7 to -6
2.5 to 3.5
Milk
3.8%
1.6
-1.2
11.7
6.0
Na
Na
Cheese
3.6%
1.4
-1.4
5.9
12.1
Na
Na
Ice
Cream
1.8% -0.4 0.9 2.3 5.1 Na
Na
Other
2.0% 1.6 0.6 6.1 7.7 Na
Na
5-Fruits & Vegetables
14.6%
3.7
4.8
3.8
6.2
-1.0 to 0.0
2.5 to 3.5
Fresh
fruits
5.6% 3.7 6.0 4.5 4.8
-2.0
to-1.0
3.0
to
4.0
Fresh vegetables
5.6%
4.0
4.6
3.2
5.6
-1.0 to 0.0
2.0 to 3.0
Processed fruits & veg.
3.5%
3.3
2.9
3.6
9.5
6.0 to 7.0
3.0 to 4.0
6-Non-alcoholic Beverages
12.0%
2.9
2.0
4.1
4.3
2.5 to 3.5
2.5 to 3.5
Juices & non-alc. drinks
8.0%
1.9
2.3
4.2
4.4
Na
Na
Coffee, tea, & other
4.1%
5.0
1.6
4.0
4.2
Na
Na
7-Sugar & Sweets
3.7%
1.2
3.8
3.1
5.5
5.0 to 6.0
3.5 to 4.5
8-Fats & Oils
3.0%
0.0
0.2
2.9
13.8
2.5 to 3.5
4.0 to 5.0
8-Other Foodsd
18.2%
1.5 1.4 1.8 5.2
4.0
to
5.0
2.5
to
3.5
Froz./freeze-dried
foods
3.8% 0.5 0.4 1.9 4.3 Na
Na
Snacks
3.8% 2.8 1.5 2.0 8.1 Na
Na
Source: Historical data for 2005 to 2008 are from BLS, Department of Labor, for the U.S. City Average (CPI-
U). The 2009 forecasts are from ERS, USDA, as of October 23, 2009.
a. Weights are “as a percent of total at-home food expenditures.”
b. Flour and prepared flour mixes, breakfast cereals, rice, pasta, and cornmeal.
c. Bread, fresh biscuits, rolls, muffins, cakes, cupcakes, cookies, and other bakery products.
d. Includes soups, spices, seasonings, condiments, sauces, baby food, and other miscellaneous foods.
Congressional Research Service
18
Consumers and Food Price Inflation
In 2009, at-home food prices are forecast to rise a modest 2% (within a range of 1.5% to 2.5%).
However, eggs and dairy products are forecast to fall sharply from their 2008 levels—egg prices
are forecast to fall by 15.5% (in a range of -16% to -15%), and dairy products are forecast down
by 6.5% (in a range of -7% to -6%).
Annual averages can cloud over substantial inter-year price movements. As a result, it is
worthwhile to glance over the monthly price indexes for the past two years to get a better sense of
the general pattern of retail food price movements across the various food groups. Monthly price
indexes (Figure 12) for the four principal food groups—cereals and bakery products; meats
(including beef, pork, poultry, and seafood); dairy products (including milk, cheese, ice cream,
and other); and fruits and vegetables (including fresh as well as processed)—for the period from
January 2007 through August 2009 reveal very different patterns of price movement.16
Figure 12. Monthly Retail Price Indexes: Various Major Food Groups
115
Cereals & Bakery
Meat
110
Dairy
Fruit & Vegetables
105
100
95
90
January 2008 = 100
85
2007
2008
2009
2010
Source: Seasonal y adjusted monthly CPI data, BLS, Department of Labor.
The meat price index was the most stable over this period, while the cereals and bakery product
price index showed the strongest increase, rising nearly 18% between January 2007 and January
2009. The dairy price index showed a strong rise through 2007 and into 2008, then declined
slightly through the second half of 2008 before falling sharply in early 2009 (down 8.8% since
December 2008). The fruit and vegetable price index was the most volatile, as a general upward
pattern over the two-year period was punctuated by significant deflationary movements during
both 2007 and 2008.
Perhaps the most dramatic and volatile of the individual foods was eggs (Figure 13), which rose
35% from January 2007 to January 2008, then fell over 20% by July 2009. A year-over-year
production decline from 2006 to 2007 coupled with strong exports tightened U.S. egg supplies
16 Note that these statistics (based on the change in monthly price indexes) differ from the statistics reported in Table 5,
where the inflation rates are calculated using the difference from annual averages rather than monthly averages.
Congressional Research Service
19
Consumers and Food Price Inflation
and pushed prices sharply higher in 2007. Stronger egg production by mid-2008 coupled with the
global economic crisis dampened prices starting in the later half of 2008.
Figure 13. Monthly Retail Price Indexes: Beef, Pork, Poultry, and Eggs
110
100
90
80
Beef
Pork
Poultry
January 2008 = 100
Eggs
70
2007
2008
2009
2010
Source: Seasonal y adjusted monthly CPI data, BLS, Department of Labor.
Figure 14. Monthly Retail Price Indexes: Dairy, Fresh Milk, Cheese, and Ice Cream
110
100
90
80
Dairy
Fresh Milk
Cheese
Ice Cream
January 2008 = 100
70
2007
2008
2009
2010
Source: Seasonal y adjusted monthly CPI data, BLS, Department of Labor.
The components of the dairy group (Figure 14) followed distinctly different patterns, particularly
the price index for fresh milk, which showed a sharp escalation in early 2007 (up 20% on the
year), followed by a sharp drop-off in 2008 including a 22% fall from July 2008 to August 2009.
Congressional Research Service
20
Consumers and Food Price Inflation
The milk price pattern for the 2007-2008 period was very similar to the egg price pattern for that
same period, and for the same principal reasons—initially tight supplies and expensive feed costs,
followed by increased supplies and a sharp drop in prices and demand. U.S. milk production
expanded through 2007 and into 2008, while the global economic crisis weakened demand,
especially from international markets. In contrast, highly processed ice cream showed a fairly
steady upward rise from mid-2007 before slowing in early 2009.
Similarly, the price index for processed fruits and vegetables (Figure 15) rose steadily through
2008, while the price indexes for fresh fruit and fresh vegetables exhibited volatile, slightly
upward patterns into mid-2008 before declining into 2009.
Figure 15. Monthly Retail Price Indexes: Fruits and Vegetables
120
Fruit & Vegetables
Processed Fruits & Veg.
Fresh Fruit
Fresh Vegetables
110
100
January 2008 = 100
90
2007
2008
2009
2010
Source: Seasonal y adjusted monthly CPI data, BLS, Department of Labor.
The price index for highly processed snacks (Figure 16) rose slowly in 2007, then accelerated
upward during 2008 before falling back somewhat in 2009. The sugar and sweets price index had
a similar pattern, although with a somewhat slower rise during 2008. The fats and oils price index
rose rapidly during 2007 and during the first nine months of 2008, then leveled off as the global
market for oils became over-supplied, in part due to the fall-off in demand (related to the global
financial crisis) in lesser-developed countries, where fats and oils are still treated as luxury goods.
Similar to other highly processed food products, the prepared-food group (which includes frozen
and freeze-dried prepared foods) and the carbonated beverages index both rose steadily through
2008 (Figure 17). However, both the prepared foods and carbonated beverages price indexes
showed signs of leveling off in early 2009. In contrast, coffee prices rose sharply during the first
half of 2008 before falling sharply in response to a decline in demand following the global
financial crisis.
Congressional Research Service
21
Consumers and Food Price Inflation
Figure 16. Monthly Retail Price Indexes: Sugar, Fat & Oils, and Snacks
120
Sugar & Sweets
Fats & Oils
Snacks
110
100
January 2008 = 100
90
2007
2008
2009
2010
Source: Seasonal y adjusted monthly CPI data, BLS, Department of Labor.
Figure 17. Monthly Retail Price Indexes: Coffee, Carbonated Beverages, and
Prepared Foods
112
Carbonated Beverages
Coffee
Prepared Foods
106
100
January 2008 = 100
94
2007
2008
2009
2010
Source: Seasonal y adjusted monthly CPI data, BLS, Department of Labor.
In sum, evidence from recent years suggests that highly processed foods more consistently adhere
to steady, stable upward price trends. In contrast, prices for less-processed retail food products—
such as eggs, milk, and fresh fruits and vegetables—respond far more quickly to changes in both
farm commodity prices and economic conditions and have followed farm prices downward in the
early months of 2009.
Congressional Research Service
22
Consumers and Food Price Inflation
Effect of High Prices
The surge in agricultural commodity prices as well as the rise in costs of raw materials, energy,
and transportation that persisted from mid-2005 through early 2008 translated into higher retail
prices for food and other household products. For a given level of income, higher prices mean
lower effective purchasing power, since the same household budget will now acquire a smaller
volume of products.
Many wages and salaries, as well as federal programs (including several domestic food assistance
programs), are linked to price inflation through escalation clauses in order to retain their
purchasing power. For households where income does not keep up with price inflation, declines
in purchasing power are both real and immediate. However, even for households with escalation
clauses that adjust incomes or benefits for price inflation, there is a time lag between the time the
price inflation is measured and the time when the wage or program benefit is adjusted upward to
compensate. As a result, for households with incomes or federal benefits linked to price inflation
escalators, higher prices can cause a short-term decline in real purchasing power. This is most
meaningful when prices are accelerating. When prices are falling, as during a deflationary period,
consumers with fixed incomes realize gains in real income (provided that they are not subject to
wage cuts or layoffs).
Although commodity prices peaked in early 2008 and have been declining since, most retail
prices have been slow to reflect wholesale commodity price declines, and it has only been since
early 2009 that retail prices have significantly retreated for most foods. The negative aspects of
high retail prices have been magnified by the global financial crisis that emerged in 2008. The
economic downturn has manifested itself in a decline in household wealth due to sharply lower
real estate values, tighter business and consumer credit, and rising unemployment numbers.
Federal Spending for Domestic Food Assistance Programs
The federal government operates several domestic food assistance programs targeted to low-
income households, as well as schoolchildren and nutritionally vulnerable groups such as
pregnant and/or lactating mothers.17
Food price inflation leads to more spending on domestic assistance efforts. Increasing prices
encourage those who are eligible, but not participating, to enroll. They also translate directly
(albeit with a time lag) into higher benefit payments and per-meal subsidies for “entitlement”
programs in which benefits are indexed to food-price inflation. These entitlement programs
include the Supplemental Nutrition Assistance Program (the SNAP, formerly the Food Stamp
Program) and programs supporting meals served to children in schools and other venues.
Increasing prices also place pressure on appropriators to provide more funding to support
caseloads for “discretionary” programs like the Special Supplemental Nutrition Program for
Women, Infants, and Children (the WIC program).
17 For details on the individual programs, see “Programs and Services,” Food and Nutrition Service, USDA, at
http://www.fns.usda.gov/fns/services.htm. Also see CRS Report R40397, Child Nutrition and WIC Programs:
A Brief Overview, by Joe Richardson. For program funding information, see the “Domestic Food Assistance” section
(by Joe Richardson) of CRS Report R40000, Agriculture and Related Agencies: FY2009 Appropriations, coordinated
by Jim Monke.
Congressional Research Service
23
Consumers and Food Price Inflation
The 2008-2009 global economic crisis—with its higher unemployment, income loss, and lower
effective household purchasing power—following on the heels of higher retail prices, has brought
on higher participation rates and greater costs for domestic food aid programs like the SNAP,
child nutrition meal-service programs (e.g., the School Lunch program), the WIC program, The
Emergency Food Assistance Program (TEFAP), and Older Americans Act nutrition programs. As
of August 2009 (the latest available information to date), participation in the SNAP was at an all-
time record high of 36.5 million persons (or about one in every 10 Americans). SNAP monthly
benefit costs have grown from $2.8 billion in January 2008 to $4.9 billion in August 2009, with
average per-person monthly benefit spending rising from $100 to $133.18 Other domestic food
assistance programs also have seen increased participation (and costs). The number of lower-
income children receiving free or reduced-price school lunches has risen from 31.0 million
schoolchildren in February 2008 to 31.4 million in February 2009.19 WIC participation has grown
from 8.6 million women, infants, and children in January 2008 to some 9 million in January
2009.20
Responding to calls for increased federal assistance, the 2009 American Recovery and
Reinvestment Act (ARRA; P.L. 111-5) provided additional support for domestic food assistance
programs: an estimated $11.5 billion for FY2009-FY2010 and $20.8 billion through FY2019.21
The SNAP is the primary recipient of this new money, most of which will be used to pay for
added benefits, loosened eligibility standards, and administrative costs.
Supplemental Nutrition Assistance Program (SNAP, formerly Food Stamps)
The SNAP is the largest of the federally supported domestic food assistance programs. SNAP
benefits normally are indexed annually (each October) for changes in the cost of USDA’s least
costly food plan, the “Thrifty Food Plan” (TFP). For a number of years and well into 2006,
annual increases in the cost of the TFP typically ranged between 1.5% and 2.5%. However,
starting in late 2006, food prices reflected in the cost of items in the TFP began to increase at a
much faster rate. For example, basic benefits were increased by 4.6% in FY2007 and by 8.5% in
October 2008. While these were substantial increases, they lagged by three months in reflecting
rising food costs—they were (by law) based on prices from the immediately previous June. Thus
there is a three-month gap between the calculation of the price inflation index in June and its use
to adjust SNAP benefits in October.
In recognition of the lag in the inflation index for SNAP benefits, increased food needs, and
reduced income, the ARRA included a major across-the-board boost in SNAP benefits, effective
in April 2009. The ARRA provisions mandate that the TFP dollar amounts used for basic SNAP
benefits be 13.6% above the June 2008 TFP cost level; this effectively boosts each recipient
household’s monthly benefit by an amount equal to 13.6% of the maximum (TFP) benefit for its
household size. For a one-person household, the added benefit is $24 a month; for two persons,
18 USDA, Food and Nutrition Service (FNS), Supplemental Nutrition Assistance Program, “Monthly Data -- National
Level: FY 2006 through August 2009,” at: http://www.fns.usda.gov/pd/34SNAPmonthly.htm.
19 USDA, FNS, Program Data—Child Nutrition Tables, “National Level Monthly Data: FY 2005 through August
2009,” at: http://www.fns.usda.gov/pd/36slmonthly.htm.
20 USDA, FNS, WIC Program, “Monthly Data -- National Level: FY 2006 through August 2009,” at:
http://www.fns.usda.gov/pd/37WIC_Monthly.htm
21 For more information, see CRS Report R40160, Agriculture, Nutrition, and Rural Provisions in the American
Recovery and Reinvestment Act (ARRA) of 2009, coordinated by Jim Monke.
Congressional Research Service
24
Consumers and Food Price Inflation
$44 a month; for three persons (the typical household), $63 a month; for four persons, $80 a
month; and for larger households, higher amounts. This change adds to the already effective 8.5%
increase in benefits under pre-existing law. For example, inflation-indexed monthly benefits for
four-person households went up by $46 in October 2008 (the 8.5% increase mentioned above);
the ARRA adds another $80 a month. However, the structure of the ARRA increase is such that
the new, higher benefit levels will continue until normal annual indexing rules provide benefits
that surpass the value of the new add-on. As a result, it is unlikely that there will be another
SNAP benefit increase related to food costs until October 2010 or later, unless food prices
increase dramatically.22
The impact of benefit increases on SNAP costs also depends on participation. The number of
Americans receiving SNAP assistance has climbed steadily over the past eight years from a low
of about 17.2 million in 2000 to an all-time record high of 36.5 million in August 2009. The
previous food stamp participation peak was in 1994, when about 27.5 million Americans received
assistance.
For FY2008, the regular benefit increase noted above (combined with estimated growth in
enrollment) yielded a likely $2 billion cost attributable to adjustments for food price increases
(out of total spending of $36.7 billion), about double the $1 billion that would have occurred
based on pre-2007 price increases. Costs are expected to increase even more in FY2009.
Child Nutrition
Federally supported child nutrition programs and initiatives reach more than 40 million children.
In FY2008, federal spending on these programs totaled over $15 billion, the second-largest
federal commitment to domestic food assistance. The basic goals of federal child nutrition
programs are to improve children’s nutrition, increase lower-income children’s access to
nutritious meals and snacks, and help support the agricultural economy.
Federal payments for meals and snacks served to children now range as high as $2.80 per meal,
including the value of USDA commodity donations. Cash per-meal payments are indexed every
July to food-price changes reflected in the food-away-from-home component of the CPI over the
12-month period ending each May. Commodity support (now some 21 cents per meal) also is
indexed annually based on the Bureau of Labor Statistics’ Producer Price Index for five major
food components (cereal and bakery products, meats, poultry and fish, dairy products, processed
fruits and vegetables, and fats and oils).
Inflation-indexed subsidy rates (those paid for the majority of school meals that are served free or
at a reduced price to children from lower-income families) have increased by over 35 cents per
meal between the 2005-2006 school year and the 2008-2009 school year. The annual increase in
subsidies has gone from 2.9% for the 2005-2006 school year to 3.3% for the 2007-2008 school
year and well over 4% for the 2008-2009 school year; in FY2008, this increased federal support
by about $300 million above spending if earlier food price increases had prevailed. This trend is
expected to continue into FY2009.
22 As of February 2009, the cost of the TFP was just 1% above the base June 2008 TFP.
Congressional Research Service
25
Consumers and Food Price Inflation
While inflation indexing of child nutrition subsidies lags behind actual increased costs to
participating meal providers like schools, no change has been made to the annual July indexing
cycle.
The WIC Program
Unlike the SNAP and child nutrition programs, which receive mandatory funding, the WIC
program is funded from discretionary sources. Spending depends on annual appropriations, based
largely on estimates of participation and the cost of the food packages that are purchased with
WIC vouchers. The value of benefits is not indexed, per se. Rather, WIC vouchers are redeemable
at whatever the participating retailer charges for the items covered by the vouchers, which differ
according to the type of recipient (e.g., pregnant mother, infant, child). As a result, the cost of
WIC vouchers reflect food price changes without the time lag built into other inflation-indexed
nutrition programs. Just as important, WIC vouchers are highly specific as to the food items they
cover and have a relatively heavy emphasis on certain types of food—for example, dairy items
and infant formula are major components.
In recent years, the cost of WIC food vouchers has varied a great deal, largely because of changes
in dairy-related food prices. The average per-participant monthly cost of vouchers has ranged
from $34.80 in FY2002 to $39.15 in FY2007. However, the annual percentage increase has been
very small for some years (1% or less for FY2003, FY2005, and FY2006) and more substantial
for other years (6.6% for FY2004 and 5.6% for FY2007). Most recently, monthly per-participant
WIC food costs averaged just under $45 for the first three months of FY2009, compared to about
$42 for the same period last year. Given this significant volatility, it is difficult to produce specific
estimates of the effect of food price inflation on WIC program costs.
Although WIC spending is discretionary, Congress has historically shown a willingness to
appropriate whatever amounts are necessary to meet costs imposed by increased participation or
food costs. Most recently, it provided a $400 million contingency reserve to meet unexpected
costs in FY2009 and FY2010 as part of the ARRA.
TEFAP and Older Americans Act Programs
The Emergency Food Assistance Program (TEFAP) and meal service programs under the Older
Americans Act also provide key food assistance support for vulnerable groups. Like the WIC
program, they are discretionary, and rising need and higher food prices have placed pressure on
appropriators to add to federal funding. TEFAP, which provides federally donated food
commodities and supports distribution of privately donated commodities, is budgeted at $450
million in FY2009 ($350 million for commodities and $100 million for distribution costs); as a
result of new funding provided by the ARRA, this is $150 million above the amount previously
available. Older Americans Act nutrition programs (“meals-on-wheels” and meals served to
seniors in congregate meal settings) also have received added funding under the ARRA; $100
million in new FY2009 federal support was added to about $800 million already available.
Congressional Research Service
26
Consumers and Food Price Inflation
Foreign Food Aid
USDA’s international activities are funded by discretionary appropriations (e.g., foreign food
assistance under the Food for Peace Act (P.L. 480) and by using the borrowing authority of the
CCC (e.g., export credit guarantees, market development programs, and export subsidies).23
Because foreign food aid is a budget value and not a food volume, its effective “purchase power”
is diminished by food price hikes without additional appropriations. Unlike some domestic
nutrition programs, foreign food aid is not adjusted to account for changing costs.
Food aid usually takes the form of basic food grains such as wheat, sorghum, and corn, and
vegetable oil—commodities critical to developing-country diets. Since there is very little value
added for these commodities, shifts in prices translate directly into higher prices for food-insecure
countries or reduced food aid contributions per dollar spent. Also, higher energy costs have
increased shipping costs for both food purchases and food aid.
23 For more information, see CRS Report RL33553, Agricultural Export and Food Aid Programs, by Charles E.
Hanrahan.
Congressional Research Service
27
Consumers and Food Price Inflation
Appendix. Major Statistical Indicators of Consumer
Expenditures and Food Price Changes
Several government agencies produce useful statistical indicators for measuring and monitoring
consumer expenditure behavior and food prices in the United States. The monthly Consumer
Price Index (CPI) produced by the Department of Labor’s Bureau of Labor Statistics (BLS) is the
most widely recognized indicator of consumer price inflation. It is described in detail below along
with a brief discussion of the Personal Consumption Expenditure (PCE) price deflator prepared
by the Department of Commerce’s Bureau of Economic Analysis (BEA).24
In contrast to the CPI, which is largely unique as a statistical indicator of consumer price changes,
several different indicators of both consumer income and food expenditures are available, due
primarily to important differences in how food expenditures and personal income are defined by
the different data series. As a result, each of these data sources has certain strengths and
weaknesses depending on how it is to be used.
The BLS, in addition to producing the CPI, is also responsible for producing and reporting on the
annual Consumer Expenditure Survey (CES). The CES is the only federal survey that allows
users to associate the complete range of household expenditures with the major characteristics
and incomes of those households. The CES is discussed in more detail below.
Two alternate sources of information on personal income are produced by the Department of
Commerce’s U.S. Census Bureau and the BEA.25 The Census Bureau collects data and reports
annually on both household income and characteristics.26 However, Census income data do not
include information on household expenditures.
The BEA produces a set of national income and product accounts (NIPAs) that provide
information on the value and composition of output produced in the United States by major sector
during a given period, and on the distribution and uses of the income generated by that output.27
The BEA complements the aggregate NIPA data with population estimates from the Census
Bureau to derive per-capita estimates of personal income and outlays by major category.
However, NIPA data are aggregate national account indicators, not household indicators.
The U.S. Department of Agriculture’s Economic Research Service (ERS) combines data from the
above sources with other information to conduct economic analyses across a range of consumer
food price and expenditure topics.28 Four noteworthy research products include:
24 The PCE price index is a measure of the change in prices paid for goods and services by the personal sector in the
U.S. national income and product accounts (NIPA).
25 Information on both of these bureaus can be obtained by visiting http://www.commerce.gov/.
26 For more information on the Census Bureau’s income data, visit http://www.census.gov/hhes/www/income.html.
27 For more information on NIPAs, see Measuring the Economy: A Primer on GDP and the National Income and
Product Accounts, BEA, U.S. Dept. of Commerce, Sept. 2007, at http://www.bea.gov/national/pdf/nipa_primer.pdf.
28 ERS undertakes economic research across a broad range of agricultural policy topics not limited to consumer food
price and expenditure issues. Information related to food prices and expenditures is accessible at the Food CPI, Prices,
and Expenditures Briefing Room, at http://www.ers.usda.gov/Briefing/CPIFoodAndExpenditures/.
Congressional Research Service
28
Consumers and Food Price Inflation
• Food CPI Forecasts—short-term (12- to 18-month) forecasts of U.S. food price
inflation for all food, food-away-from home, and food-at-home, with the latter
category broken out into 15 selected food subcategories. 29
• Food Expenditure Tables—aggregate food expenditure estimates produced
from national data that measure current sales or receipts by each type of store that
sells food.30 ERS methods provide a comprehensive measure of the total value of
all food expenditures by all final purchasers including government agencies,
businesses, and nonprofit organizations.
• Food Baseline Projections—projections for the U.S. agricultural sector 10 years
ahead, including crop production, agricultural trade, farm income, and consumer
food prices and expenditures.31 Projections are a description of what would be
expected to happen under the 2008 farm bill (P.L. 110-246), with very specific
external circumstances.
• Price and Income Response Elasticities—detailed estimates of consumer price
and income responsiveness, and food expenditure behavior, across different
income groupings both within the United States32 and for low-, middle-, and
high-income countries.33
Consumer Expenditure Survey (CES)
The CES is a nationwide, annual household survey that collects information on expenditures for
goods (including food items) and services used in day-to-day living.34 In addition, the CES
collects information on household income and household characteristics such as the number, age,
race, education, and gender of household members, housing tenure status, and household assets
and liabilities.
The CES is designed to represent the total U.S. civilian non-institutional population. It has been
collected annually by the U.S. Census Bureau—under contract with BLS—since 1979.35 Prior to
1979, the CES was conducted every 10 to 12 years.
The CES uses two main survey components, an interview and a diary, to collect information.
29 ERS food price inflation forecasts are available at http://www.ers.usda.gov/Briefing/CPIFoodAndExpenditures/Data/
cpiforecasts.htm.
30 For a list of the food expenditure tables, visit http://www.ers.usda.gov/Briefing/CPIFoodAndExpenditures/Data/.
31 “Table 30—Consumer food price indexes and food expenditures,” at http://www.ers.usda.gov/Briefing/Baseline/.
32 For information on entire demand systems, see “Food Demand Analysis,” and for information on specific food
groups, see “Who Eats What, Where, and How Much,” both at the Diet Quality and Food Consumption Briefing
Room, ERS, USDA, at http://www.ers.usda.gov/Briefing/DietQuality/whoeats.htm.
33 James Seale, Jr., Anita Regmi, and Jason Bernstein, International Evidence on Food Consumption Patterns,
Technical Bulletin No. 1904, ERS, USDA, Sept. 2003, and Anita Regmi, ed., Changing Structure of Global Food
Consumption and Trade, Agriculture and Trade Report WRS-01-1, ERS, USDA, May 2001.
34 For more information, see the U.S. Dept. of Labor, BLS, BLS Handbook of Methods, Chapter 16, Consumer
Expenditures and Income, at http://www.bls.gov/opub/hom/pdf/homch16.pdf.
35 The BLS releases data from the CES in several formats including an annual report, news releases, and articles in the
Monthly Labor Review and other journals. For more information on the CES, see the BLS website for “CES Frequently
Asked Questions,” at http://www.bls.gov/cex/faq.htm.
Congressional Research Service
29
Consumers and Food Price Inflation
• The interview survey is designed to obtain data on the types of expenditures
respondents can recall for a period of three months or longer. In general, these are
relatively large expenditures such as those for property, automobiles, or major
appliances, or they occur on a regular basis such as rent, utilities, or insurance
premiums. Approximately 14,000 addresses are contacted each calendar quarter
of each year. Usable interview data is obtained from approximately 7,100
households each quarter.
• The diary survey is designed to obtain expenditure data on small, frequently
purchased items such as food and beverage expenditures (both at home and in
eating places), housekeeping supplies and services, nonprescription drugs, and
personal care products. About 12,000 addresses are selected each year to
complete a diary survey of household expenditures for two consecutive one-week
periods. Usable diary data is obtained from approximately 7,100 households.
Major Uses of CES Data
An important use of CES data is for periodic updating of the Consumer Price Index (CPI) market
basket of goods and services, as well as to determine the relative importance of CPI components
and to derive new cost weights for the market basket items.36
Another important use of CES data is to allow data users to relate the expenditures and income of
consumers to key characteristics of those consumers. For example, CES data are used by both
government and private agencies to study the welfare of different segments of the population such
as the elderly, low-income families, urban families, and those receiving food stamps. CES data
are used by policymakers to evaluate the potential effects of policy changes on levels of living
among diverse socioeconomic groups. Also, market researchers may use CES data to analyze the
demand for various goods and services by different demographic groups. Because of the large
survey sample size and the representative nature of sample selection, the CES data also allow for
regional comparisons.
Household-Level Expenditures versus National Average Disposable Income
Clearly, only the CES household-level data allow analysts to evaluate food expenditure behavior
while controlling for specific household characteristics. However, ERS has shown (as discussed
below) that national accounts data can provide a more comprehensive evaluation of national food
expenditure behavior by capturing food industry and other non-household food behavior that is
otherwise omitted from a household survey. These differences are briefly discussed here.
National Accounts Data: Focus on Food Expenditure Trends
For monitoring national food expenditure trends and patterns (e.g., food budget shares or food-at-
home versus away-from-home consumption patterns) food expenditure estimates based on
aggregate account data generally provide the most useful information. In this regard, the ERS
food expenditure series is perhaps the most qualified and easily accessible of available data
36 CPI data, once released, are never revised because of their widespread use for adjusting wages to cost-of-living
adjustments. However, the components of the market basket, their respective weights, and the formula for calculating
price changes are revised on occasion for future calculations.
Congressional Research Service
30
Consumers and Food Price Inflation
series. ERS uses primarily Census Bureau data (as classified under the North American Industry
classification System (NAICS)) to construct very detailed and comprehensive food expenditure
series based on specific industry and business food sales.37
Unlike NIPAs, ERS food expenditure calculations exclude pet food, ice, and prepared animal
feeds, as well as nonfoods such as drugs and branded supplies, in estimating at-home food
purchases.
Unlike CES data, ERS food expenditure estimates include non-household purchases and
acquisitions—for example, government food purchases for domestic military personnel, the value
of school meals (including “free lunches” distributed under the National School Lunch Program),
and the value of food purchased by airlines for serving during flights.
CES Household Data: Focus on Individual Consumer Behavior
For evaluating household purchases and consumer behavior, household-level expenditure data (as
available from the CES) are generally preferable to aggregate income measures such as the
disposable personal income (DPI) data available from either the Census or the BEA. There are
two primary reasons for this. First, the CES expenditure data are collected at the household level,
thus permitting household comparisons rather than relying on the national averages available
from the NIPA data. Second, household expenditures are generally more stable than household
income measures—even during periods of unemployment by household members, the household
will continue to make purchases of routine items such as food. Thus, household expenditure data
are less subject to the gyrations of the general economy.
Household expenditure data can be more reliably measured by survey data than can household
income data. As a matter of fact, it is not uncommon for CES data for household annual
expenditures to exceed reported income before taxes, particularly for households from lower-
income groups. Why? Some CES interview respondents may be reluctant to answer inquiries
concerning their income. Some households whose members may experience a spell of
unemployment may draw on their savings to maintain their expenditures. Self-employed
consumers may experience business losses that result in low or even negative incomes for a
period, but are able to maintain their expenditures by borrowing or relying on savings. Students
may get by on loans while they are in school, and retirees may rely on savings and investments.
The Consumer Price Index (CPI)
The CPI is a composite measure of the average change over time in the prices paid by U.S. urban
consumers for a defined market basket of goods (including food) and services that people buy for
day-to-day living.38 Since its focus is on consumer expenditures, it intentionally excludes
government expenditures and the cost of inputs used by manufacturing and agriculture.
37 For details, see “Measuring the ERS Food Expenditure Series,” Food CPI, Prices, and Expenditures Briefing Room,
ERS, at http://www.ers.usda.gov/Briefing/CPIFoodAndExpenditures/measuringtheersfoodexpendituresseries.htm.
38 The BLS CPI website with supporting documents and news releases is at http://www.bls.gov/cpi.
Congressional Research Service
31
Consumers and Food Price Inflation
Is the CPI’s Food Market Basket Representative?39
The BLS is responsible for collecting monthly price data for the market basket of consumer
expenditure categories from across the United States, then calculating and reporting the monthly
price indexes (or CPIs) for those items. The mix of goods and services making up the market
basket is based on spending patterns established by the annual CES, and is updated every two
years. The prices of market basket items are drawn from representative urban settings across the
United States. According to BLS, the consumer groups included in the sample pool (referred to as
the U-population and used to create the CPI-U index) represent 87% of the total U.S. population
including professionals, the self-employed, the poor, the unemployed, and retired people.40 Not
included in the CPI are spending patterns of people living in rural non-metropolitan areas, farm
families, people in the Armed Forces, and those in institutions.
Figure 18. CPI Relative Weights by Major Category
Source: U.S. Department of Labor, BLS.
Note: As of December 2008 (based on the 2005-06 CES).
CPI Categories and Weights
The BLS collects prices each month in 87 urban areas across the United States for about 80,000
items from approximately 23,000 retail establishments (including department stores, filling
39 For more information, see the BLS report, The Consumer Price Index—Why the Published Averages Don't Always
Match An Individual’s Inflation Experience, at http://www.bls.gov/cpi/cpifact5.htm.
40 A subset of the U-population, the W-population consists of the U-population for whom 50% or more of household
income comes from wages and clerical workers’ earnings. The W-population is used to create its own price index (CPI-
W). For specific details refer to “Calculation of Price Indexes,” BLS Handbook of Methods, Chapter 17, The Consumer
Price Index, BLS, Dept. of Labor, p. 2, June 2007, at http://www.bls.gov/opub/hom/pdf/homch17.pdf.
Congressional Research Service
32
Consumers and Food Price Inflation
stations, supermarkets, hospitals, etc.).41 All taxes directly associated with the purchase and use of
items are included in the index. The data are further combined, based on population size, into 38
different urban geographic areas.42 The data on goods and services are aggregated into 211
categories. As a result, basic price indexes are available for 38 x 211 = 8,018 area-item
combinations. Higher-level indexes are produced by calculating a weighted average across the
relevant basic indexes. The weights are derived from reported expenditures from the CES. BLS
updates the expenditure weights every two years.
The “All-Items” CPI is the index most often referred to (i.e., the headline CPI) for representing
consumer price inflation, and represents a weighted index based on the indexes from all areas and
for all categories of goods and services whose prices are collected. The all-items CPI is generally
divided into eight major spending categories, including a “Food and Beverage” category (Figure
18Error! Reference source not found.), which accounts for 15.8% of average consumer
expenditures. This compares with the Housing category’s dominant share of 43.4% of annual
consumer expenditures and the Transportation category’s 15.3% share.43
What About Seasonality Patterns in the Data?
Some monthly price data are subject to seasonal influences that normally occur at the same time
and in about the same magnitude every year, independent of market phenomena or consumer
income shifts, or household characteristics. For example, meat purchases generally increase in the
summer months, when the “grilling” season is open. As a result, the meat market experiences a
general surge in demand during the summer months that is independent of incomes, household
characteristics, or other economic or demographic factors.
For most price indexes, BLS reports two series—a seasonally adjusted index that removes
systematic seasonal influences from the data, and the original unadjusted data.44 The seasonally
adjusted data, including the all-items index levels, are subject to revision for up to five years after
their original release. In contrast, the unadjusted data are not subject to revision, even when the
index formula is itself altered. This is because many wage contracts and government program
benefits are indexed via an escalation agreement to the unadjusted CPI. A revision of historical
CPI data would necessarily imply a revision of all wage and benefit payments made under the
previously unrevised CPI series.
For analyzing general price trends in the economy, the BLS recommends using seasonally
adjusted price indexes since they eliminate price movements that result from seasonal phenomena
such as changing climatic conditions, production cycles, model changeovers, holidays, and sales.
In contrast, BLS recommends using unadjusted price indexes when the primary interest to
consumers is the price actually paid. Unadjusted price indexes are used extensively for escalation
purposes in collective bargaining agreements or pension plans.45
41 “How BLS Measures Changes in Consumer Prices,” BLS, Dept of Labor, at http://www.bls.gov/cpi/cpifact2.htm.
42 For specific details refer to “Calculation of Price Indexes,” BLS Handbook of Methods, Chapter 17, The Consumer
Price Index, BLS, Dept. of Labor, pp. 3-4, June 2007, available at http://www.bls.gov/opub/hom/pdf/homch17.pdf.
43 All CPI data referenced in this report refer to the CPI-U data series.
44 This discussion follows from “A Note on Seasonally Adjusted and Unadjusted Data,” Consumer Price Index:
February 2009, monthly news release, BLS, U.S. Department of Labor.
45 For more information, see CRS Report RL34168, Automatic Cost of Living Adjustments: Some Economic and
Practical Considerations, by Brian W. Cashell.
Congressional Research Service
33
Consumers and Food Price Inflation
Is the Food-at-Home CPI a Good Measure of Retail Food Price Change?
As mentioned above, the food-at-home CPI is the principal indicator of changes in retail food
prices in the United States. Because of its central role as a measure of retail food price inflation,
the food-at-home CPI is routinely subject to public scrutiny by food market participants including
industry groups, consumer and hunger advocacy groups, and policymakers. A chief concern is
whether and to what extent there are any persistent biases in the food-at-home CPI. If the CPI’s
market basket of items and their respective weights fail to adjust in step with consumer
expenditure behavior or if the method for calculating price changes is itself faulty, then the
process would likely produce a biased statistical measure of price change.
Indeed, research by BLS and ERS analysts suggests that the food-at-home CPI has tended to
overstate retail food price inflation. In particular, they identified four specific problems in the
CPI’s construction that have imparted an upward bias in the food-at-home CPI.46
1. Consumers’ spending patterns change, albeit gradually, over time. For example,
an aging population (as in the United States) tends to consume less meat and
alcohol, but larger shares of cereals and bakery products. Also, food budget
shares have declined over time with rising incomes, and an increasing share of
food spending has gone for away-from-home consumption.
2. Prices for some food items—particularly fresh fruits and vegetables and fresh
meats—may vary widely on occasion. If not accounted for, these fluctuations
create an inflationary drift by attaching greater weight to price increases than to
decreases. For example, a $0.50 increase on a $1.00 price is a 50% increase,
whereas a return to the $1.00 from the $1.50 price is only a 33% price decrease.
3. Price indexes must keep up with the introduction of new food items and new
stores to fully represent consumer price alternatives. For example, the CPI has
been slow to incorporate the rapid expansion of discount stores that often sell
food items at substantially lower prices than traditional grocery stores.47
4. Fixed-weight price indexes are vulnerable to a substitution bias.48 A fixed-weight
index does not allow for the likelihood that consumers respond to relative price
changes by changing the quantity or quality of their food purchases. For example,
price rises due to higher quality should not count as inflation, since the consumer
is getting more product value for the higher price.
BLS has responded to these problems over time by implementing a series of changes to its
methods for incorporating and weighting price information from new stores and for new products
starting in 1981. In addition, starting in 2002, BLS began updating the expenditure weight shares
every 2 years. Prior to 2002, such revisions were more episodic, occurring every 10 to 12 years. A
continual rebalancing of the market basket to include new products and more current expenditure
46 This discussion follows from “CPI Bias,” Food CPI, Prices, and Expenditures Briefing Room, ERS, USDA, as of
March 26, 2007, at http://www.ers.usda.gov/Briefing/CPIFoodAndExpenditures/cpibias.htm.
47 For a discussion of this issue see The Impact of Big-Box Stores on Retail Food Prices and the Consumer Price Index,
Economic Research Report No. 33, Ephraim Leibtag, ERS, USDA, December 2006.
48 Empirical studies conducted in the 1980s and 90s suggested that substitution bias resulted in an upward bias of about
0.4 percentage points per year in using changes in the CPI to measure changes in the cost of living. M. Boskin, E.
Dulberger, R.J. Gordon, Z. Griliches, and D.W. Jorgenson, “Consumer Prices, the Consumer Price Index, and the Cost
of Living,” Journal of Economic Perspectives, Vol. 12, No. 1, Winter 1998, pp. 3-26.
Congressional Research Service
34
Consumers and Food Price Inflation
weights helps to ensure that the CPI more accurately reflects how consumers allocate their
spending.
BLS also has introduced new methods of index construction (in particular, a geometric mean
formula for within-item averages) to minimize problems associated with sharply fluctuating
prices and possible inflationary drift. However, this adaptation does not capture substitution
across items, only within items (thus mitigating lower-level substitution bias). In other words, it
would not capture a price-related shift from steak to hamburger, only a shift between differently
priced steaks. BLS began to publish a new chained index (the C-CPI-U) in August 2002, to more
closely approximate the substitution across expenditure categories that consumers undertake
when relative prices change.49 However, the traditional all-items CPI (CPI-U) remains the most
widely quoted price index and still is subject to the higher-level substitution bias.
Why Not Use the PCE Price Index to Measure Consumer Price Inflation?
An alternate measure of inflation—the Personal Consumption Expenditure (PCE) price index—is
produced by the Department of Commerce’s Bureau of Economic Analysis (BEA).50 The PCE
price index measures the change in prices paid for goods and services by the personal sector in
the U.S. national income and product accounts (NIPAs).
Despite their apparent similarity, the CPI and PCE price indexes are constructed differently, they
behave differently over time, and they have different purposes and uses.51 The PCE price index is
based on national accounts data and is primarily used for macroeconomic analysis and
forecasting. In contrast, the CPI is based on household survey data and measures the change in
prices paid by urban consumers for a market basket of consumer goods and services; it is
primarily used as an economic indicator and as a means of adjusting current-period data for
inflation. Because of this later use, historical CPI data are never revised (unlike PCE data).
Analysts at BEA and BLS have identified four major differences between the CPI and the PCE
price indexes. First, they are calculated using different formulas.52 Second, each formula’s set of
weights is different (the CPI uses base-year weights versus the PCE’s current-year weights).
Third, the CPI measures “out-of-pocket” expenditures based on individual household survey data,
whereas the PCE price index measures the aggregate outlay for all goods and services purchased
by households and nonprofit institutions as drawn from national accounts. Finally, the BEA and
BLS analysts cite an accumulation of other minor differences in seasonality adjustments, pricing,
and a residual difference due to the underlying data used for each index.
Empirical tests conducted by BEA and BLS analysts suggest that differences in the formulas and
the weights account for most of the difference in how the two indexes measure inflation. The CPI
uses a formula whereby the weights for goods and services are determined by household
49 For more information on the C-CPI-U, see CRS Report RL32293, The Chained Consumer Price Index: How Is It
Different, and Would It Be Appropriate for Cost-of-Living Adjustments? by Brian W. Cashell.
50 NIPA data tables including PCE price indexes are available at http://www.bea.gov/national/nipaweb/index.asp.
51 For a detailed discussion of their differences, see Clinton P. McCully, Brian C. Moyer, and Kenneth J. Stewart,
“Comparing the Consumer Price Index and the Personal Consumption Expenditures Price Index,” Survey of Current
Business, November 2007, at http://www.bea.gov/scb/pdf/2007/11%20November/1107_cpipce.pdf.
52 The CPI is calculated using a “modified” Laspeyres index, while the PCE price index is calculated using the Fisher-
Ideal Index. See source information from the preceding footnote 51 for details.
Congressional Research Service
35
Consumers and Food Price Inflation
expenditures in the base year. In successive years, the same fixed “base” of weighted goods and
services is valued using the new year’s prices. A shortcoming of this approach is that, because the
weights for individual goods and services are fixed, consumer substitution due to relative price
changes is not captured by the index. In contrast, the PCE price index uses current-year
expenditures to determine the weights. As a result, the PCE price index fully captures consumer
substitution among detailed items as the relative price of those items changes.
However, PCE price index methodology is difficult, if not entirely impractical, to implement
using household expenditure data. Household survey data are available with a substantial lag due
to collection and preparation time, and thus are generally not available for the most current
period. This precludes calculating current-year weights as per the PCE formula. As a result, the
CPI’s base-year weighting appears to be the most timely and practical (although perhaps not the
most theoretically sound) method for calculating a price index.
Author Contact Information
Randy Schnepf
Joe Richardson
Specialist in Agricultural Policy
Specialist in Social Policy
rschnepf@crs.loc.gov, 7-4277
jirichardson@crs.loc.gov, 7-7325
Congressional Research Service
36