Economic

Research
Service
China’s Meat Consumption:
Economic
Research
Growth Potential
Report
Number 320
Fred Gale and Fengxia Dong

July 2023
RecommendedResearch
Economic citation format for this publication:
Service
www.ers.usda.gov
Gale, F., & Dong, F. (2023). China's meat consumption: Growth potential. U.S. Department of
Agriculture, Economic Research Service. ERR-320.

Cover image assets sourced from Getty Images.

Use of commercial and trade names does not imply approval or constitute endorsement by USDA. The analysis, findings, and
conclusions expressed in this report should not be attributed to IRI or NielsenIQ.

To ensure the quality of its research reports and satisfy governmentwide standards, ERS requires that all research reports with
substantively new material be reviewed by qualified technical research peers. This technical peer review process, coordinated
by ERS' Peer Review Coordinating Council, allows experts who possess the technical background, perspective, and expertise
to provide an objective and meaningful assessment of the output’s substantive content and clarity of communication during
the publication’s review.

In accordance with Federal civil rights law and U.S. Department of Agriculture (USDA) civil rights regulations and policies, the
USDA, its Agencies, offices, and employees, and institutions participating in or administering USDA programs are prohibited
from discriminating based on race, color, national origin, religion, sex, gender identity (including gender expression), sexual
orientation, disability, age, marital status, family/parental status, income derived from a public assistance program, political
beliefs, or reprisal or retaliation for prior civil rights activity, in any program or activity conducted or funded by USDA (not all
bases apply to all programs). Remedies and complaint filing deadlines vary by program or incident.

Persons with disabilities who require alternative means of communication for program information (e.g., Braille, large print,
audiotape, American Sign Language, etc.) should contact the responsible Agency or USDA's TARGET Center at (202)
720-2600 (voice and TTY) or contact USDA through the Federal Relay Service at (800) 877-8339. Additionally, program infor-
mation may be made available in languages other than English.

To file a program discrimination complaint, complete the USDA Program Discrimination Complaint Form, AD-3027, found
online at How to File a Program Discrimination Complaint and at any USDA office or write a letter addressed to USDA and
provide in the letter all of the information requested in the form. To request a copy of the complaint form, call (866) 632-9992.
Submit your completed form or letter to USDA by: (1) mail: U.S. Department of Agriculture, Office of the Assistant Secretary
for Civil Rights, 1400 Independence Avenue, SW, Washington, D.C. 20250-9410; (2) fax: (202) 690-7442; or (3) email: program.
[email protected].

USDA is an equal opportunity provider, employer, and lender.

Economic
Research
Service China’s Meat Consumption:
Economic
Research
Report
Growth Potential
Number 320 Fred Gale and Fengxia Dong
July 2023 Abstract
China is a major meat producer, consumer, and importer, but consumption changes are difficult to
assess due to inconsistencies in Chinese data. China’s population growth is slowing, meat prices are
rising, and income is growing at a slower pace than during earlier decades. Disease and other supply-
side factors constrain domestic meat supply and imports are a growing share of the supply of each type
of meat. Nevertheless, statistical models based on consumer income growth and meat prices suggest
China has potential for continued growth in meat consumption. Consumption of poultry, beef, and
mutton is growing faster than pork consumption but pork still comprises more than half of consumer
meat expenditure in China. Consumption is growing despite rising meat prices in China, and statistical
models confirm that consumption is relatively insensitive to price changes. Beef and mutton consump-
tion have risen despite sustained price increases for these meats. Poultry appears to be a substitute for
pork.

Keywords: meat consumption, China, livestock, pork, poultry, beef, mutton, Almost Ideal Demand
System

About the Authors

Fred Gale and Fengxia Dong are agricultural economists with the USDA, Economic Research Service.

Acknowledgments
The authors thank Alexandra Baych, Claire Mezoughem, and Erin Covert from the USDA, Foreign
Agricultural Service; Shawn Arita and Justin Choe of the USDA Office of the Chief Economist; Joshua
Huang from the USDA, Economic Research Service; David Ortega of Michigan State University; and
Holly Wang of Purdue University. The authors thank USDA, Economic Research Service staff Elaine
Symanski and Chris Whitney for editorial assistance, and Xan Holt of USDA, ERS for layout and
design.
Contents

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1
Conflicting Measures of Per Capita Meat Consumption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5
Inconsistencies in Chinese Meat Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .8
Factors Affecting Meat Consumption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .11
Population and Urbanization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
Regional Differences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .13
Income and Prices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
Calculating Annual Meat Prices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
Consumer Budget Shares Devoted to Meat . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
Supply Side Factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .20
Estimating Sensitivity to Personal Income and Meat Prices . . . . . . . . . . . . . . . . . . . . . . . . . . . . .23
Linear Consumption Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .24
Case Study of a Meat Supply Shock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .26
Almost Ideal Demand System (AIDS) Model for Meat . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .27
Predicted Meat Consumption Growth, 2022–31 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .30
Discussion of Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .33
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .35
Appendix 1: China’s Meat Reserve . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .39
Appendix 2: A Review of China’s Meat Statistics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .41
National Bureau of Statistics Official Livestock Production Data . . . . . . . . . . . . . . . . . . . . . . . . . 41
Ministry of Agriculture and Rural Affairs Livestock Reporting. . . . . . . . . . . . . . . . . . . . . . . . . . .43
Household Income and Expenditure Surveys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .44
Appendix 3: Meat Demand Estimates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .45
Appendix 4: Data Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .49

ii
China’s Meat Consumption: Growth Potential, ERR-320
USDA, Economic Research Service
A report summary from the Economic Research Service July 2023

China’s Meat Consumption:

Growth Potential
Fred Gale and Fengxia Dong

What Is the Issue? Economic

Research China’s Meat Consumption:
Service

Economic
Research
Growth Potential
China’s emergence as a major meat producer, consumer, and importer may have Report
Number 320

Month 2023
Fred Gale and Fengxia Dong

potential for export opportunities for the United States and other producers. China
has become the world’s largest meat importer, as growth in its own production
has diminished. Growth in China’s meat production and consumption has slowed
due to decelerating population and income growth, animal disease, scarcity of
land for feed and forage, rising production costs, and health concerns. Has China’s
consumption of meat reached a ceiling, or is there room for more growth? This study
investigates trends in China’s meat supply and household purchases, discusses data
inconsistences, analyzes population, income, and price data that influence consump-
tion, and estimates statistical models to ascertain future growth in China’s meat
consumption.

What Did the Study Find?

Chinese meat prices have risen faster than other commodity prices since the 1990s, but meat consumption is relatively
insensitive to prices.

• ERS analysts looked at beef, mutton, pork, and poultry and found that the relationship between consumption and
prices is inelastic (i.e., the demand changed proportionately less than the price).

• Poultry consumption is stimulated by rising pork prices, but substitutability between other types of meats is weak
and inconsistent across data sets and estimation techniques.

• Several supply shocks (due to swine disease outbreaks) resulted in large swings in pork prices.

• Poultry consumption decreased during outbreaks of avian influenza when cases in humans occurred.

• Consumption of beef and mutton continued rising despite large price increases.

ERS is a primary source of economic research and analysis from the U.S. Department of Agriculture, providing timely
information on economic and policy issues related to agriculture, food, the environment, and rural America.

www.ers.usda.gov
Imports comprised 9.1 percent of China’s meat supply in 2021, up from 1.0 percent in 2000. Meat imports
accounted for a growing share of supply, as production growth diminished. In 2021, net meat imports comprised
18.6 percent of China’s beef and mutton disappearance, 8.3 percent for pork, and 5.0 percent for poultry.

Data discrepancies still prevent precise analysis of meat consumption trends despite China’s revisions of production
data and survey methods. Statistics indicate that China’s per capita meat supply is about twice as large as consump-
tion reported by a Chinese national household survey. Meat supply statistics indicate an apparent plateau in China’s
meat consumption after 2014, but the household survey indicates more robust growth in meat consumption.

• Both sources indicate that poultry, beef, and mutton are gradually increasing their share of China’s meat
consumption spending.

• Pork’s share declined but still comprises more than half of consumer meat expenditures.

• For each type of meat, analysis of household survey data indicated greater sensitivity to each meat’s own price
than did analysis of disappearance data.

• Household survey data indicate a small decline in meat consumption during 2019–20, when China’s meat
supply declined sharply due to a swine disease. However, the household survey and supply data both indicate
a strong consumption recovery in 2021.

China’s household income growth averaged 7.6 percent annually during 1985–2021, but income growth fell below
this average during most years after 2014.

• Despite rising consumption and prices, meat expenditures comprise a shrinking share of Chinese household
budgets.

• Pork consumption is relatively insensitive to income, but poultry, beef and mutton consumption is more
responsive to income growth.

Based on past relationships between meat consumption, income, and prices, per capita meat consumption is
projected to rise during 2022–31 by 23 kilograms using consumer purchase data and 21 kilograms using disappear-
ance data. Pork consumption is projected to grow slower than the consumption of other meats.

How Was the Study Conducted?

This report analyzes China’s meat consumption using two official Chinese data sources: (1) disappearance calcu-
lated from meat production and net import data, and (2) household purchases of meat reported by a national
household survey. The report discusses the divergence of the two data series, changes in consumption of three major
categories of meat, growth in household income, and fluctuation in meat prices. The study estimates relationships
between meat consumption, meat prices, and household income, using two methods: (1) first-differenced linear
regressions of three categories of per capita meat consumption on inflation-adjusted per capita income and meat
prices, and (2) a first-differenced Almost Ideal Demand System for expenditure on three meat categories. Using
these estimates, the study calculates predicted meat consumption for 2022–31 based on assumed paths of income
and prices.

www.ers.usda.gov
China’s Meat Consumption:
Growth Potential
Introduction

Meat consumption in China was low until the late 20th century. After several decades of rapid growth,
China has emerged as a major producer, consumer, and importer of meat. This trend creates new opportuni-
ties for exporters in the United States and other countries, but it also poses food security challenges and envi-
ronmental impacts (Ortega et al., 2015; Yu, 2015; Ortega et al., 2022). In recent years animal disease, scarcity
of land for feed production, rising costs, and health concerns have contributed to slower growth. Has China’s
meat consumption reached a ceiling, or is there room for more growth?

An early 20th-century study of Chinese food and agriculture (Buck, 1937) found that nearly all calories were
obtained from plant sources because little land was available to support animal agriculture. Yet, China seem-
ingly overcame this constraint during the 1980s to early 2000s, as its production and consumption of meat
boomed. National data show that China’s meat output grew from less than 10 million metric tons in the
1970s to 60 million metric tons in 2000, reflecting a vast increase in meat consumption (figure 1). Data from
the United Nations Food and Agriculture Organization show that the share of calories supplied by meat in
China rose from 4 percent to 19 percent between the 1960s and the 2000s (Ortega et al., 2015).

However, the rate of growth slowed somewhat after 2000 and then abruptly ceased after reaching a peak
of 88 million metric tons in 2014. Production dropped to 77.5 million metric tons during 2019–20 due to
a severe swine disease outbreak. The data indicate a recovery to a record 89 million metric tons in 2021.
China’s Ministry of Agriculture and Rural Affairs (China MARA, 2022) projected a resumption of modest
growth in meat output that would reach 93.1 million metric tons in 2025 and 95.7 million metric tons by
2031.

The difficulties of assessing China’s meat situation—due mainly to concerns about the accuracy of China’s
data—are illustrated by a comparison with USDA estimates of pork, beef, and chicken output (figure 1;
note that USDA data exclude sheep, waterfowl, and other meats such as rabbit and deer). USDA estimates
also show a peak in meat output in 2014, but a steeper output decrease from 2014 to 2020 than the Chinese
MARA data shows. The USDA estimates show a delayed Chinese recovery from the trough in output in
2020, but output in 2031 is still projected slightly lower than the 2014 peak.

1
China’s Meat Consumption: Growth Potential, ERR-320
USDA, Economic Research Service
Figure 1
Increase in China’s meat output, 1950–2031

Million metric tons

120

100

80

60

40

20

0
1950 1960 1970 1980 1990 2000 2010 2020 2030
Year

Meat output China's Agriculture Ministry projection USDA* USDA projection

*USDA data include pork, chicken, and beef and exclude sheep, waterfowl, and minor meats such as rabbit and deer.
Note: Output of meat is a carcass weight equivalent for slaughtered hogs, cattle, sheep, and poultry.
Source: USDA, Economic Research Service, based on China National Bureau of Statistics data, China MARA (2022), and USDA data
from Production, Supply, and Distribution database and USDA International Projections to 2031.

USDA’s Production, Supply, and Distribution database reported that China was the largest producer of pork
in 2018, the second-leading producer of chicken, and the third-leading producer of beef. However, China’s
per capita consumption of meat was lower than in many middle- and high-income countries. According to
USDA’s international baseline database, China’s combined consumption of pork, chicken, and beef was 53.9
kilograms per capita in 2018 (this number excludes sheep and waterfowl that were not included in the USDA
data), which is about the same as the meat consumption level in Japan and slightly more than in Thailand
(figure 2). Some middle-income regions in Asia (such as Malaysia, Taiwan, and South Korea) had a per capita
consumption that exceeded China’s by 10 kilograms or more. The United States, Argentina, Brazil, Canada,
and the European Union also had a much higher per capita meat consumption than in China. Per capita
consumption in the United States and Argentina is about twice the level in China. Meat consumption in
China was dominated by pork—China’s per capita pork consumption was one of the world’s highest, but per
capita consumption of chicken and beef in China were low, shown in figure 2.

2
China’s Meat Consumption: Growth Potential, ERR-320
USDA, Economic Research Service
Figure 2
Per capita meat consumption, selected countries and regions, 2018

Country/region

Nigeria
Indonesia China, 53.9 kilograms
Iran
Turkey
Philippines
Vietnam
Thailand
Japan
South Africa
Malaysia
Mexico
Russia
South Korea
Taiwan
European Union
Canada
Brazil
Argentina
United States
China
0 20 40 60 80 100 120 140
Kilograms
Pork Chicken Beef

Source: USDA, Economic Research Service based on USDA international baseline database.

China relies on the global market to sustain growth in meat production and consumption. China’s meat
production created demand for animal feeds that made China the top importer of soybeans and feed grains
such as corn, sorghum, and barley (Gale, 2015; Gale et al., 2019). As meat production growth slowed, China
began to import meat and now is an important player in global meat markets. USDA estimates show that
China’s meat imports surpassed those of other top importers (Mexico, South Korea, and Japan) between 2015
and 2019 (figure 3). China was the world’s top meat importer by a substantial margin in 2020. China was
the top importer in 2022, despite a sharp drop in imports after pork production recovered rapidly in 2021.
Future meat imports could remain large or increase if consumption outpaces growth in domestic produc-
tion.1 USDA’s Baseline projections showed that China’s pork imports could reach 6.3 million metric tons in
2031, beef imports could reach 4.0 million metric tons, and chicken imports could reach nearly 1.3 million
metric tons.

1 Beckman et al. (2022) found that nontariff barriers prevented China’s pork and beef imports from growing even larger.

3
China’s Meat Consumption: Growth Potential, ERR-320
USDA, Economic Research Service
Figure 3
Leading meat-importing countries, 2000–22*

Million metric tons

10

-2

2000 2005 2010 2015 2020

Year
China Japan South Korea Mexico

*2022 is a forecast.
Note: Net imports = imports – exports. Meat is the sum of pork, beef, and chicken in carcass weight equivalent.
Source: USDA, Economic Research Service analysis of USDA Production, Supply and Distribution data.

The potential for further growth in China’s meat consumption is uncertain. While a significant portion of
China’s population still has nutritional deficiencies that indicate low meat consumption (Liu et al., 2018),
rising incidences of obesity, heart disease, and other diet-related diseases have prompted Chinese public
health experts to recommend that consumers curb meat consumption (China State Council, 2014; China
Nutrition Society, 2021). Additionally, concerns about environmental impacts of livestock production and
animal welfare are encouraging consumption of plant-based and lab-grown meats (Ortega et al., 2022).

Income growth is likely to prompt additional spending on meat, but China’s income and population growth
are slowing. Some studies found that the pace of meat consumption growth slows as consumer incomes reach
a higher level (Gale and Huang, 2007; Chen et al., 2016). Land constraints in China identified by Buck
(1937) contributed to rising feed costs and the degradation of grasslands. With constraints on Chinese supply
and growth in demand, meat prices have been rising faster than inflation. The rise in prices may offset the
effects of income growth to some extent.

The report analyzes trends and fluctuations in production, net imports, consumption of meat, meat prices,
and consumer income by using data from China’s National Bureau of Statistics and Administration of
Customs. The analysis disaggregates meat into three main components—pork, poultry, and beef and
mutton—and investigates the changing composition of China’s meat consumption. The report also discusses
inconsistencies between disappearance and household purchase data that cannot be reconciled and conducts
parallel analyses of the two data series. In addition, the study estimates the sensitivity of meat consumption
to changes in consumer income and the fluctuation in meat prices (based on historical data) and supply-
side factors constraining meat consumption are discussed. The report projects future consumption based on
moderate income growth and a continuation of the trend toward rising inflation-adjusted meat prices, to
ascertain the prospects for continued growth.

4
China’s Meat Consumption: Growth Potential, ERR-320
USDA, Economic Research Service
Conflicting Measures of Per Capita Meat Consumption

The amount of meat consumed in China is difficult to ascertain since no statistics directly measure China’s
aggregate meat consumption. Analysts indirectly estimate consumption as “disappearance” using an identity
based on the assumption that consumption equals the available supply:

Disappearance = Production + Net Imports – Inventory Change (1)

Disappearance is a residual quantity estimated by adding domestic production and net imports (imports –
exports) (shown in figures 1 and 3), plus inventory change (ending inventory – beginning inventory). The
calculation assumes that all meat available is consumed. This report assumes that inventory change is a
negligible portion of the annual meat supply since inventories of meat—a perishable product—cannot be
stored easily and China’s national meat reserves are not revealed to the public (see appendix 1, “China’s Meat
Reserve”).

China produced nearly all the meat it consumed until recent years, so meat production is the main statistical
indicator available to measure consumption, estimated annually by China’s National Bureau of Statistics. The
main components are the dressed weights of slaughtered hogs, cattle, and sheep and the final weight of poul-
try.2 The accuracy of consumption estimates is also determined by the quality of production data. Appendix
2 discusses the changes in statistical methodology and meat data revisions that raise questions about the
accuracy and consistency of these data.

Growth in Chinese per capita meat disappearance is an indicator that consumption has outpaced population
growth. Per capita disappearance estimates indicate that China’s meat consumption grew from under 10 kilo-
grams/person in the 1970s to 66 kilograms/person in 2014 (figure 4). After 2014, the growth trend reversed,
as meat disappearance declined gradually to 64.5 kilograms/person in 2019. Meat disappearance then fell
sharply to 58.7 kilograms/person in 2019, reflecting reduced meat supplies resulting from a nationwide
outbreak of African swine fever.3 Net imports grew during 2014–20 but not enough to fully offset the decline
in meat output. Official Chinese data indicate that meat disappearance rebounded in 2021, as farms reestab-
lished swine herds and production recovered. With imports added to domestic production, disappearance was
at a record-high 69.4 kilograms/person in 2021 and rose to 70.2 kilograms/person in 2022.

2 Documentation in China National Bureau of Statistics Rural Statistical Yearbooks indicates that its estimates of meat output are the weight

of carcasses after removing head, feet, internal organs, and blood. The data are estimated from rural livestock surveys and do not appear to directly
measure carcasses produced by slaughter facilities.
3 The African swine fever virus spread throughout China during 2018–19, causing a large reduction in the pork supply.

5
China’s Meat Consumption: Growth Potential, ERR-320
USDA, Economic Research Service
Figure 4
Per capita meat disappearance and household purchases per capita in China, 1952–2022

Kilograms per person

80

70

60

50

40

30

20

10

0
1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015 2020
Year
Imports Production plus imports=disappearance Meat produced Household purchases

Note: Per capita values are calculated using official population data. Household purchases are the weighted average of rural and
urban Chinese consumption of pork, poultry, beef and mutton for 1981–2012 and the national average for 2013–22, reported by China
National Bureau of Statistics.
Source: USDA, Economic Research Service, based on China National Bureau of Statistics and China Statistical Yearbook, various
years.

Another measure of meat consumption—household purchases of meat per capita—is measured by a house-
hold survey conducted by China’s National Bureau of Statistics (independent of the surveys used to develop
meat production estimates). While the household survey is the only direct measure of consumption, the
survey is incomplete because it does not capture meat consumed away from home in restaurants, cafeterias,
and institutions. Nor does the household survey fully account for meat used in processed foods or waste and
spoilage in the marketing chain. Some observers suspect that the survey sample may not be representative
of the national population and discontinuities in the data suggest that the survey may not accurately reflect
trends and fluctuations. The issues related to this survey are discussed in more detail in appendix 2.

The per capita amount of meat Chinese households purchased shown in figure 4 is not entirely consis-
tent with the trend in disappearance.4 Household purchases and disappearance were approximately equal
during the 1980s, but they diverged during the 1990s as disappearance rose at a much faster pace, as shown
in appendix figure 2. This inconsistency prompted some scholars in China to suggest that the meat disap-
pearance data were inflated by exaggerating production growth while household purchases missed rising
consumption of meat away from home (see box, “Inconsistencies in Chinese Meat Data”).

Growth in household purchases and meat disappearance were nearly parallel during the 2000s, but per capita
disappearance was more than twice as large as household purchases. After 2014, per capita disappearance
began to decline, while household purchases continued to grow. The magnitude of the decline in disap-

4 The chart shows an average of Chinese urban and rural consumption weighted by population for 1985–2012 and the national average reported

by China’s National Bureau of Statistics for 2013–22.

6
China’s Meat Consumption: Growth Potential, ERR-320
USDA, Economic Research Service
pearance during 2019 was much larger than the decline in purchases. Both disappearance and purchases
increased sharply in 2021. Household purchases grew 7.0 kilograms in 2021, slightly less than the 7.5-kilo-
gram growth in disappearance that year. In 2021, per capita meat disappearance was 69.2 kilograms, still
67.0 percent higher than household purchases that averaged 41.4 kilograms.

Detailed estimates of per capita meat disappearance show the growing importance of meat imports in China’s
meat supply. Data showed growth in meat imports partially offset falling production during 2014–20 and
facilitated record-high meat consumption in 2021 and 2022. Net imports grew 1.1 kilograms during 2000–
14, then grew from 1.6 kilograms to 4.9 kilograms during 2014–22. Net imports accounted for 2.3 percent
of disappearance in 2014, peaked at 10.9 percent in 2020, and declined to 6.9 percent in 2022. The increase
in imports from 2014 to 2022 accounted for nearly all the 3.9-kilogram increase in meat disappearance over
that period.

Table 1
Estimates of China's per capita meat disappearance, 2000–22
Meat supply = Net imports
Production Net imports*
Year disappearance share of disappearance
Kilograms Percent
2000 47.4 0.5 47.9 1.0

2014 64.8 1.6 66.4 2.3

2015 64.0 1.7 65.7 2.7
2016 62.8 3.1 65.8 4.7
2017 62.6 2.6 65.2 4.0
2018 61.8 2.7 64.5 4.2
2019 55.4 4.1 59.5 6.9
2020 54.9 6.7 61.7 10.9
2021 62.9 6.3 69.2 9.1
2022 65.4 4.9 70.2 6.9
Cumulative changes:
2000–14 17.4 1.1 18.4 1.3
2014–22 0.6 3.3 3.9 85.6

*Net imports = imports - exports, Harmonized System code 02 "meat and edible meat offal."
Note: Calendar year data. Production data reflect revisions made by China’s National Bureau of Statistics following agricultural
censuses conducted in 2007 and 2017. Per capita values are calculated by dividing national production and import values by total
population.
Source: USDA, Economic Research Service calculations using data from the China National Bureau of Statistics and China Customs
Administration.

7
China’s Meat Consumption: Growth Potential, ERR-320
USDA, Economic Research Service
Inconsistencies in Chinese Meat Data

Zhong (1997) first pointed out the apparent exaggeration of meat production statistics and their incon-
sistency with household survey data. Several studies investigated the inconsistencies and attempted to
reconcile the production and consumption data (Xiao et al., 2015). Detailed discussion of statistical
methodologies and revisions of China’s meat data are in appendix 2, Review of China’s Meat Statistics.

Fuller et al. (2000), Ma et al. (2004), and Yu and Abler (2014) attributed the inconsistencies between
meat production and consumption data to several factors: incentives to inflate livestock numbers,
failure to incorporate consumption of meat away from home, and the exclusion of waste and food
processing. Those authors used multiple data sources and assumptions to reconcile the two series.

A downward revision of official livestock data following a 1997 Chinese agricultural census seemed to
validate concerns about over-reporting of livestock data. Fuller et al. (2000) explained how inflation of
livestock data could have resulted from pressure to achieve production targets and incentives created
by the privatization of livestock production and marketing. Xiao et al. (2015) denied any evidence
existed of overstated livestock data, yet they suggested that census data revisions following the 1997
and 2007 censuses improved the accuracy of the statistics. Calculations indicate that beef production
data was revised downward again by 12.6 percent after the 2017 census, but pork and mutton were
revised upward slightly.

Shao et al. (2020) observed that declining inventories of swine—a variable related to pork output—
after 2014 was inconsistent with market conditions. As possible causes, they cited environmental regu-
lation and reduced reporting by local officials concerned that inflated data would be exposed by the
Chinese agricultural census in 2016. It is unclear, however, why no similar decline was reported by
China’s National Bureau of Statistics for beef, mutton, or poultry which were also subject to environ-
mental regulation and had also been revised downward in censuses (see appendix table 2). Shao et al.
(2020) used hog and feed prices and a machine learning algorithm to adjust monthly swine inventory
data reported by China’s Ministry of Agriculture and Rural Affairs (MARA), but the study did not
analyze pork output directly.

Failure to capture Chinese meat consumption away from home is potentially a major source of
discrepancy. A survey by Xiao et al. (2015) conducted during 2010 found that 33 percent of meat was
consumed away from home, and Bai et al.’s (2020) surveys during 2010–12 found that 28 percent of
meat was consumed away from home. Xi et al. (2016) cited several surveys showing that 12–40 percent
of poultry was consumed away from home. Yu and Abler (2014) argued that China’s National Bureau
of Statistics rural household survey understated per capita consumption by counting migrants and
students who lived and ate elsewhere as rural household members. A revision of the household survey
in 2012 appeared to address the rural migrant data issue by eliminating the rural-urban distinction in
the survey and counting migrants based on their actual place of residence.

Yu and Abler (2014) concluded that over reporting of production was the largest contributor to the
gap between reported pork supply and consumption but under reporting of consumption also played
a role. Xiao et al. (2015) attributed most of the discrepancy in meat statistics to failure to capture meat
consumed away from home.

Production data is based on carcass weights and exclude feet, organs, blood, and head of animals (30
percent or more of the liveweight) but these items are often consumed in China. The household survey
includes these items in meat purchases (see appendix 2).

8
China’s Meat Consumption: Growth Potential, ERR-320
USDA, Economic Research Service
 Some meat counted in farm production statistics never reaches consumers due to waste and spoilage,
thus accounting for part of the discrepancy between production and consumer purchase data.
Aggregate supply and demand estimates by China’s Ministry of Agriculture and Rural Affairs (2021)
included estimates of waste and spoilage equal to 4.7 percent of supply for pork, 3.5 percent for beef,
and 3.4 percent for mutton (no waste was reported for poultry). Xiao et al. (2015) cited estimates of
meat waste that ranged from 5–20 percent. The only direct measurement of China’s food waste, a
survey of restaurant waste by Wang et al. (2017), indicated that meat accounted for 17 percent of
waste (pork 8 percent, beef 3 percent, and poultry 6 percent). Based on this percentage, the Chinese
food waste estimate of 11 kilograms per person per year implied 1.9 kilograms of per capita meat and
poultry waste. That total would be less than 3 percent of meat disappearance that year, but does not
include waste in processing, distribution, and retail.

Pork has always been the predominant meat consumed in China but its consumption has also been volatile.
Pork disappearance peaked in 2014, fell dramatically in 2019, and rebounded in 2021 (figure 5 and appendix
table 5). The peak in meat disappearance in 2014 shown in figure 4 appears to reflect the peak in pork disap-
pearance while disappearance of poultry, beef, and mutton continued to grow from 2014 to 2022.

Chinese per capita household purchases of pork were constant from 2013 to 2017 (figure 6 and appendix
table 5). Household purchases did not show the decline evident in the disappearance data during those years.
Household pork purchases peaked in 2018, fell during 2019–20, then rebounded to a record high in 2021.
Purchases of poultry were generally on an upward trend and surged during 2019–20 when pork purchases
declined, consistent with Ma et al.’s (2021) finding that poultry is a strong substitute for pork. Beef and
mutton purchases were much smaller but rose gradually during 2014–21.

Figure 5
China’s per capita disappearance of pork, poultry, and beef and mutton, 2000–22

Kilograms per person

50
43.2
45
41.2
40
35
30
25
18.0
20
12.9
15
11.0
10
5
8.0
0
2000 2005 2010 2015 2020
Year
Pork Poultry Beef/mutton

Note: Disappearance = production + net imports. Net imports of pork (harmonized system (HS) codes 0203, 020630, 020641,
020649), poultry (HS code 0207), beef and mutton (HS codes 0201, 0202, 0204, 020610, 020621, 020622, 020680, 020690, 020629) are
calculated from customs data.
Source: USDA, Economic Research Service, based on China’s National Bureau of Statistics and China customs data, accessed
through the Trade Data Monitor.

9
China’s Meat Consumption: Growth Potential, ERR-320
USDA, Economic Research Service
Figure 6
China’s household purchases of pork, poultry, and beef and mutton, 2000–21

Kilograms per person

30
Peak disappearance 25.2
25
20.0
20

15
12.3
10 8.0

5
2.5
3.9
0
2000 2005 2010 2015 2020
Year
Pork Poultry Beef/mutton

Source: USDA, Economic Research Service, based on data from China’s National Bureau of Statistics.

Authors compared the Chinese disappearance and household purchase data series for each of the three meat
categories. Disappearance was generally about twice as large as household purchases, but there was a slight
tendency for the differences between the two series to narrow after 2014. In particular, pork disappearance
was 100–125 percent larger than household purchases from 2000 to 2013 (see appendix figure 2). The differ-
ence in the pork data narrowed to 55 percent in 2019 and 2021, as disappearance fell faster than household
purchases. For beef and mutton, the difference between disappearance and household purchases was even
larger, fluctuating between 150 percent and 240 percent. From 2003 to 2021, poultry had the least difference
between disappearance and household purchases. The difference for poultry declined from 75–100 percent
during 2003–09 to 45 percent in 2021.

Authors also checked the consistency of year-to-year changes in consumption indicated by the disappearance
and household purchase data. The correlation between changes in disappearance and household purchases for
pork was 0.70 but the correspondence between the two data series deteriorated after 2014 (see appendix table
9). During 2000–13, the correlation for pork data was a relatively high 0.87 but the correlation for pork data
during 2014–21 declined to a low value of 0.29. In contrast, the correlations between changes in disappear-
ance and household purchases were higher for other meat categories: 0.98 for poultry and 0.95 for beef and
mutton, and the correlations were also relatively stable between time periods.

Net imports comprised a growing share of disappearance for each of the three Chinese meat categories. Per
capita net imports in 2010 were 0.6 kilograms for pork, 0.3 kilograms for poultry, and less than 0.1 kilo-
grams for beef. These net import values were 2.0 percent or less of disappearance for each of the three meat
categories. Pork net imports rose to 2.1 kilograms per person in 2016 and 2019, then reached 3.9 kilograms
in 2020 and 3.4 kilograms in 2021 before they fell to 2.0 kilograms in 2022. Net imports comprised 11.9
percent of pork disappearance in 2020, then fell to 4.7 percent in 2022. Beef and mutton net imports rose to
19.9 percent of disappearance in 2022. Poultry net imports were generally between 1 percent and 2 percent of
disappearance until they rose to 5 percent in 2020 and 2021 and 4.1 percent in 2022.

10
China’s Meat Consumption: Growth Potential, ERR-320
USDA, Economic Research Service
Table 2
China’s net imports of pork, poultry, and beef and mutton per capita, 2010–21
Net imports per capita Net imports share of disappearance
Year Pork Poultry Beef and mutton Pork Poultry Beef and mutton
Kilograms Percent
2010 0.6 0.3 0.0 1.5 2.0 0.5
2011 0.9 0.2 0.1 2.4 1.2 0.8
2012 1.0 0.2 0.1 2.3 1.8 1.7
2013 1.0 0.3 0.4 2.3 2.1 5.3
2014 0.9 0.2 0.4 2.2 1.4 5.4
2015 1.1 0.1 0.5 2.6 0.9 6.3
2016 2.1 0.3 0.6 5.1 1.8 7.0
2017 1.7 0.2 0.7 4.2 1.1 8.0
2018 1.5 0.2 1.0 3.7 1.4 11.0
2019 2.1 0.4 1.5 6.6 2.5 15.3
2020 3.9 1.0 1.8 11.9 5.4 17.7
2021 3.4 0.9 2.0 8.3 5.0 18.6
2022 2.0 0.7 2.2 4.7 4.1 19.9

Note: net imports = imports – exports for pork (harmonized system codes 0203, 020630, 020641, 020649), poultry (HS code 0207),
beef and mutton (HS codes 0201, 0202, 0204, 020610, 020621, 020622, 020680, 020690, 020629). Per capita values are calculated
from population data. Disappearance = production + net imports.
Source: USDA, Economic Research Service calculations using data from China customs administration accessed through the Trade
Data Monitor and China’s National Bureau of Statistics.

Factors Affecting Meat Consumption

Population and Urbanization

Population growth affects aggregate consumption of meat. China’s population grew 1.0 percent to 1.5 percent
annually during the 1980s and 90s, contributing to rapid growth in aggregate meat consumption (figure 7).
Population growth decelerated to 1.0 percent in the 2000s, coinciding with slower growth in meat disappear-
ance. The one-child policy adopted in 1979 was replaced with a two-child policy in 2016 and then a three-
child policy in 2021. Population growth nevertheless fell to near zero in 2021 and during 2022, according
to official data. The United Nations projected a peak Chinese population of 1.425 billion and population
declines of -0.1 percent to -0.3 percent annually during 2025–40. The United Nations projected a decline in
China’s population to 1.38 billion by 2040.

11
China’s Meat Consumption: Growth Potential, ERR-320
USDA, Economic Research Service
Figure 7
China’s population growth, 1950–2040
Percent

3.5
3.0
2.5
2.0
1.5 China reported
1.0 first decline
in 2022
0.5
0.0
-0.5
-1.0
-1.5
-2.0

1950 1960 1970 1980 1990 2000 2010 2020 2030 2040
Year
Reported by Chinese statistics U.N. projection

Note: Reflects population data following revisions from the 2020 population census. Projections for 2022–40 are from the United Na-
tions World Population Prospects 2022.
Source: USDA, Economic Research Service calculations using data from China’s National Bureau of Statistics web site.

Urbanization of the population is another factor that contributes to rising meat consumption around the
world (Regmi and Dyck, 2001) and in China (Zheng et al., 2019). China’s urban population was less than 20
percent of the total during the 1970s. Urbanization grew to 27 percent in 1990, 36 percent in 2000, and 50
percent in 2010 (figure 8). China continued to urbanize during the 2010–20 decade, reaching 64 percent in
2020 and 65 percent in 2022. China’s 14th 5-year plan set a target of 75 percent urbanization in 2035.

Figure 8
Urban share of China’s population, 1950–2035

Percent
80
70
75
60
65
50
40
Targets set by
30 14th 5-year plan
20
10
0
1950 1960 1970 1980 1990 2000 2010 2020 2030
Year

Source: USDA, Economic Research Service, calculated from China’s National Bureau of Statistics data and China’s 14th 5-year plan
for social and economic development.

12
China’s Meat Consumption: Growth Potential, ERR-320
USDA, Economic Research Service
Per capita meat consumption is typically higher for urban than for rural residents in China—as the migra-
tion of rural Chinese workers to cities increases income, alters lifestyles, and gives migrants access to a greater
variety of foods in urban markets and in workplace cafeterias and restaurants (Zheng et al., 2019). Urban and
rural per capita household purchases of pork, poultry, and beef/mutton for 2013–21 indicate that urban-rural
differences in pork and poultry have diminished (figure 9). The urban-rural difference in consumption largely
disappeared in 2018 for pork and in 2020 for poultry. In contrast, an urban-rural difference in beef/mutton
consumption of about 2.0 to 2.3 kilograms persisted, as both values rose in parallel over time.5 Urban and
rural values of pork and poultry purchases also displayed similar year-to-year fluctuations during 2017–21.

Figure 9
Per capita purchases of pork, poultry, and beef and mutton, China urban and rural households,
2013–21

Kilograms

30

Pork
25

20

15
Poultry

10
Beef and mutton
5

0
2013 2015 2017 2019 2021 2013 2015 2017 2019 2021 2013 2015 2017 2019 2021
Year

Urban Rural

Source: USDA, Economic Research Service analysis of data from China’s National Bureau of Statistics, China Statistical Yearbook.

Regional Differences
China’s regions have distinctly different meat consumption patterns that are obscured in national aver-
ages. The patterns reflect the different types of livestock traditionally predominant in these regions as well
as income differences. The persisting differences across provinces are illustrated by fluctuations in per capita
purchases of pork, poultry, and beef/mutton in five diverse regions (Beijing, Guangdong, Henan, Sichuan,
and Xinjiang) during 2015–21 (figure 10). The national average for each year is also shown for comparison.

Per capita pork consumption is highest in southwestern Chinese provinces like Sichuan, where swine raising
has historically been widespread, even though incomes are not high in those regions. According to China’s
National Bureau of Statistics household purchase data, per capita pork consumption in Sichuan exceeded
the national average by 70 percent in 2018 (figure 10). In contrast, per capita pork consumption in Henan
Province was consistently about 40 percent less than the national average despite also being a major pork-

5 Appendix 2 notes that a change in survey methodology in 2012 may have reduced the Chinese urban-rural difference in consumption by

adjusting their classification of migrants in the household survey.

13
China’s Meat Consumption: Growth Potential, ERR-320
USDA, Economic Research Service
producing region. In Xinjiang, a region where much of the population is Muslim, pork purchases were about
80 percent below the national average. The large pork-consuming provinces of Sichuan and Guangdong had
the largest fluctuations in pork purchases during 2019–21. The fluctuation in the national average was muted
by much smaller fluctuations in regions like Beijing and Xinjiang.

Guangdong’s per capita purchases of poultry were more than double the national average. Sichuan’s poultry
purchases were also slightly above the national average. Other regions shown (Beijing, Henan, and Xinjiang)
had poultry purchases 3 to 4 kilograms below the national average. Fluctuations in the national average
appear to reflect mainly fluctuations in Guangdong’s and Sichuan’s poultry purchases.

Conversely, beef and mutton consumption was highest in regions of western China, where grasslands histori-
cally supported pastoral grazing as a primary food source and the Muslim faith proscribes consumption of
pork. Xinjiang’s beef and mutton purchases far exceeded the national average, peaking at more than 19 kilo-
grams per capita in 2018. Beijing’s beef and mutton purchases were also about double the national average.
Other regions shown in figure 10 had very low beef and mutton purchases, but the purchases tended to
increase over time. For example, Henan’s beef and mutton purchases rose from 2.1 kilograms to 3.3 kilo-
grams between 2015 and 2021. Guangdong had low beef and mutton purchases, despite having a relatively
high income, reflecting regional preferences for poultry and pork.

Figure 10
Chinese per capita household purchases of pork, poultry, and beef and mutton in selected prov-
inces, 2015–21

Kilograms
45
Sichuan
40
35
Guangdong
30
25
Xinjiang
20
15
10
5
0
2015 2017 2019 2021 2015 2017 2019 2021 2015 2017 2019 2021
Year
Pork purchases Poultry purchases Beef and mutton purchases

National Beijing Henan Guangdong Sichuan Xinjiang

Source: USDA, Economic Research Service analysis of household purchase data from the China National Bureau of Statistics.

14
China’s Meat Consumption: Growth Potential, ERR-320
USDA, Economic Research Service
Income and Prices
Many previous studies found that meat consumption tends to rise with income and decline as meat prices
rise but results varied in magnitude. Meta-analysis of 85 studies released between 1987 and 2012 found that
income elasticities of demand for pork, beef, and mutton averaged about 0.6 but elasticities varied widely
from negative values to elastic values of 1.5 (Chen et al., 2016; Zhou et al., 2020). Poultry income elastici-
ties averaged 0.85 but also ranged widely from 0.14 to 3.5. Studies found meat and poultry price elasticities
of demand that averaged -0.6 to -0.8, but they also varied from highly elastic values of -2.5 to some positive
values. Demographic characteristics, like age of the household head, had minor effects (Chen et al., 2016).
Most studies of food consumption in China analyzed cross sections of the per capita household purchase
data. Studies used data produced prior to the revision of the household survey in 2013 discussed in appendix
2. A more recent study using rural and urban data for all provinces for 2000–12 found (unconditional) price
elasticities above -0.7 for urban meat and poultry, -0.48 for rural pork, -0.38 for rural beef, and -0.28 for
rural poultry (Zheng et al., 2019). Zheng et al. (2019) noted that their study excluded data after 2012 due
to the survey revisions. Ma et al. (2021) estimated a price-flexibility model using time-series disappearance
data for 1991–2018 to assess impacts of African swine fever on China’s pork market. Their findings indicated
that pork demand is price inelastic and that pork and poultry are strong substitutes. A recent analysis of
time series data on meat demand in other Asian countries found results similar to those of the China studies
(Rathnayaka et al., 2021).

USDA, ERS estimates that growth in China’s real per capita household income averaged 7.6 percent per year
from 1985 to 2021 (figure 11). Income growth exceeded the 7.6-percent average each year from 2001 to 2014,
peaking at 13.9 percent in 2007. Income growth fell below the long-run average during 2015–20, slowing
to 6 percent in 2019 and 2.2 percent during the Coronavirus (COVID-19) pandemic year of 2020. Income
growth exceeded the long run average during 2021 (8.2 percent) but growth slowed to 2.9 percent during
2022 when many cities were under extended lockdowns to prevent spread of COVID-19. Meat consump-
tion—especially away from home—was likely curtailed by restrictions on travel and reduced consumption in
cafeterias and restaurants during intermittent city lockdowns during 2020–22. Projected Chinese per capita
gross domestic product (GDP) growth of 4 to 5 percent annually for 2022–33 suggests that income growth
may also remain at a rate well below the historical average.6

6 Income growth data for 2022 was announced after the 2022–33 projections were made.

15
China’s Meat Consumption: Growth Potential, ERR-320
USDA, Economic Research Service
Figure 11
Growth in China’s per capita disposable income
Percent

16
13.9
14
12.2
12
10.9
10
8
7.6 4.7
6
6.0
4
2
2.2 2.9
0
-2
-4
-6

1985 1990 1995 2000 2005 2010 2015 2020 2025 2030

Year
Real per capita household income growth Average, 1985-2021 Projected

Note: Chart shows annual change in per capita disposable income. Real income growth was calculated by subtracting the change in
the consumer price index each year. Projected values for 2022–33 reflected the projected per capita gross domestic product (GDP)
from the USDA, Economic Research Service, International Macroeconomic Database.
Source: USDA, Economic Research Service calculations using data from China’s National Bureau of Statistics “cha shu” database.

Average annual meat prices adjusted for inflation rose faster than the general level of prices during 1985–2021
(figure 12).7 The weighted average of meat prices largely reflected the price of pork, the predominant type
of meat produced and consumed in China. Pork prices fluctuated from year to year, with prominent peaks
occurring in 1997, 2008, and 2020. The inflation-adjusted pork price (in constant 2000 yuan) more than
doubled from 10.9 yuan in 1986 to 22.5 yuan in 2021. One of the most severe meat price increases occurred
during 2019–20, when the African swine fever epizootic reduced pork supplies. The pork price peaked at
34.82 yuan in 2020, before dropping to 22.75 yuan in 2021.

Beef and mutton prices clearly grew faster than other meat prices. The fastest growth occurred during 2007–
14, when beef and mutton prices diverged from pork and poultry prices. Beef and mutton prices grew again
during 2018–21. The average inflation-adjusted beef and mutton price in 2021 was 57.1 yuan, nearly five
times higher than the 1986 average. Poultry prices grew at a slower pace than other meat prices. Inflation-
adjusted poultry prices declined during the 1990s, then grew moderately during 2005–11 and displayed no
clear trend during 2011–21. The average inflation-adjusted poultry price in 2021 was about the same as in
1990.

7 Price data without inflation adjustment are available in appendix table 7.

16
China’s Meat Consumption: Growth Potential, ERR-320
USDA, Economic Research Service
Figure 12
China’s average meat prices, 1986–2022 adjusted for inflation

Yuan per kilogram

60

50

40

30

20

10

0
1985 1990 1995 2000 2005 2010 2015 2020
Year
All meat, weighted average Beef and mutton Pork Poultry

Note: See box, “Calculating Annual Meat Prices” for an explanation of the data. Annual averages are calculated from monthly data,
deflated to constant 2000 Chinese yuan with China’s consumer price index. Prices in current yuan in 2021 were 36.15 yuan for pork,
21.71 yuan for chicken, and 85.6 yuan for beef and mutton.
Source: USDA, Economic Research Service calculations using data posted online by the China National Development and Reform
Commission (2000–09), China National Bureau of Statistics (2010–14), and China Ministry of Agriculture and Rural Affairs (2015–21).

Calculating Annual Meat Prices

China has no continuous data series measuring meat prices over 1985–2022. USDA, ERS authors
constructed a price series for Chinese pork, chicken, beef, and mutton by combining retail price data
reports posted online by China’s National Development and Reform Commission (NDRC, 1999–
2017) and National Bureau of Statistics (NBS, 2013–17) with wholesale market prices posted by
China’s Ministry of Agriculture and Rural Affairs (MARA, 2006–22). ERS authors also calculated
average annual Chinese prices for 1986–90 from expenditures and quantities reported in the house-
hold surveys for those years (see appendix table 7 for 2000–21 price data).

Monthly series were compared to verify that levels and fluctuations were consistent during years when
multiple series were available. Annual average prices were calculated from monthly and weekly prices.
Prices from household survey data represented 1986–2000. USDA, ERS selected the NDRC data to
represent 1999–2008, NBS for 2009–2017, and MARA for 2018–22. An average retail-wholesale
margin was added to the wholesale prices for 2018–22 to approximate the retail price. The margin was
calculated as the difference between retail and wholesale prices for 2009–17 when retail and wholesale
data were both available.

An average overall Chinese meat price was calculated by weighting the price of meat in each category
by production of that type of meat. China’s consumer price index, with a fixed base of 1978, was used
to deflate prices. The CPI was adjusted to reflect constant yuan for year 2000.

17
China’s Meat Consumption: Growth Potential, ERR-320
USDA, Economic Research Service
Consumer Budget Shares Devoted to Meat
The change in the Chinese share of consumer spending devoted to meat is an indication of income elasticity
and the sensitivity of consumer purchases to price changes. Zheng et al. (2019) showed that overall food
spending’s share of household expenses declined over time. USDA, ERS estimated meat household expendi-
tures from 1986 to 2021 as the product of average household purchases and the weighted average meat price.
Estimated meat expenditure was divided by total household expenditures and household food expenditures
reported by the National Bureau of Statistics in the China Statistical Yearbook to ascertain the importance
of meat in consumer budgets. The estimates indicate that the share of meat in Chinese consumers’ overall
budgets had shrunk by half since the 1990s, despite growth in meat consumption and rising meat prices
(figure 13). Meat’s share of all household expenditures fluctuated in the range of 11 to 13 percent during the
1980s to mid-1990s, then fell to 4 to 6 percent during 2012–18. The share rebounded to near 7 percent in
2020 due to high pork prices following the African swine fever epizootic. The meat share declined slightly to
6.3 percent in 2021 when pork prices fell but quantity consumed increased sharply.

The share of meat spending in consumers’ food budgets fluctuated between 20 percent and 25 percent from
1986 to 2005. The meat share of food spending fell to 15 percent in 2014 and below 15 percent in 2017–18.
The meat share of food spending rebounded to 22 percent in 2020 due to high pork prices that year, similar
to the share during the 1990s. The share of food spending remained at a relatively high level of 21.1 percent
in 2021.

The long-term downward trend in the meat share of household expenditure is the result of Chinese house-
holds’ faster spending growth on nonfood items. Most of the shrinkage in meat’s share of the food budget
occurred from 2011 to 2018 when average inflation-adjusted meat prices were relatively steady. Temporary
increases in the meat budget share correspond to surges in pork prices during 2007–08, 2011, and 2019–20.
Each of these increases occurred after a major swine disease event. The downward trend in meat budget share
was reversed by the increase in meat prices during 2019–20. The increase in household meat consumption
during 2021 sustained the meat budget share at a relatively high level.8 Still, meat comprised just 6 percent
of all household spending in 2021. Economic theory suggests that a small budget share reduces the relative
impact of meat prices on consumer budgets.9 A smaller budget share, therefore, tends to reduce consumers’
sensitivity to meat price changes.

Chinese consumers’ shares of spending on different types of meat indicate that pork is the predominant
component of meat spending. The large share of pork reflects traditional consumption habits, the use of pork
in recipes and, in turn, the widespread raising of pigs by Chinese households in past decades. Calculations
indicate that pork accounted for 70–80 percent of Chinese consumers’ meat spending during the 1980s and
1990s (figure 14). At that time, poultry accounted for 12–15 percent of meat spending while beef and mutton
accounted for less than 10 percent.

8 Chinese household purchases increased for nearly every food item in 2021. The ratio of food spending to household expenditure was nearly
constant in 2020–21 at about 30 percent. Increased at-home purchases of all foods, combined with an unchanged overall food expenditure share,
might reflect increased consumption of food at home and less spending on food consumed away from home during COVID-19 lockdowns.
9 The “Slutsky equation” (derived in many economics textbooks) decomposes the effect of price changes on consumption into a substitution effect

and an income effect that depends on the good’s income elasticity of demand and its share in a consumer’s budget.

18
China’s Meat Consumption: Growth Potential, ERR-320
USDA, Economic Research Service
Figure 13
Estimated meat spending share of Chinese consumers’ household budget, 1986–2021

Percent
30

25

20

15

10

0
1985 1990 1995 2000 2005 2010 2015 2020
Year

Meat share of food expenditures Meat share of all household expenditures

Note: Expenditures on meat are the product of average per capita household purchases and weighted-average meat price. Budget
share is calculated as the ratio of estimated meat expenditures to per capita total household expenditure.
Source: USDA, Economic Research Service analysis of data from China’s National Bureau of Statistics, National Development and
Reform Commission and Ministry of Agriculture.

Figure 14
China meat budget shares, 1986–2021

Percent Disappearance data Percent Household purchase data

90 90
80 80
70 70
60 60
50 50
40 40
30 30
20 20
10 10
0 0
5 0 5 00 05 10 15 20 5 0 5 00 05 10 15 20
198 199 199 20 20 20 20 20 198 199 199 20 20 20 20 20

Pork Poultry Beef and mutton

Note: USDA. Economic Research Service estimated Chinese expenditures on pork, poultry, and beef-mutton as the product of per
capita disappearance and household purchases and average price for each meat. Budget shares are the ratio of each type of meat to
total meat expenditure. Shares add to 100 for each year.
Source: USDA, Economic Research Service calculations using data from China’s National Bureau of Statistics, National Development
and Reform Commission, and Ministry of Agriculture.

19
China’s Meat Consumption: Growth Potential, ERR-320
USDA, Economic Research Service
Pork’s dominance gradually declined as other meats gained popularity. Pork’s share of meat spending fell
to 60 percent during 2014–17, then fell to 57 percent in 2018, before rebounding to 63 percent in 2020.10
Shares of expenditures on beef, mutton, and poultry gradually increased. Beef and mutton’s share grew to
a peak of 24 percent in 2018 and fell to 19 percent in 2020. The growing beef-and-mutton share reflected
both rising purchases and higher prices for these meats. While poultry purchases rose in volume, the price
of poultry did not grow as fast as the prices of other meats. Consequently, poultry’s share of the meat budget
remained roughly steady at 16–20 percent during 2002–20.

Supply Side Factors

Supply-side factors indirectly affect meat consumption through their impact on production costs and prices
of livestock, and the supply of meat. Disappearance estimates show that rising Chinese imports have allowed
meat consumption to exceed domestic production. However, China’s meat consumption is not entirely inde-
pendent of its domestic supply, since growth in meat imports is limited by tariffs and nontariff barriers, which
prevent imports from rising enough to equilibrate Chinese and international prices (Beckman et al., 2022).
Consequently, sustained growth in domestic livestock prices led to rising consumer prices that constrained
growth in consumption.

China overcame its land scarcity constraints to produce more meat than Buck (1937) thought possible during
the 1980s and 90s, when rapid expansion by small-scale operators was facilitated by abundant rural labor
resources, utilization of low-cost feeds like food waste and milling by-products, and the use of crude housing
for animals. More recently, livestock supply has become constrained by rising off-farm wages and barriers to
entry that raised capital requirements. Land scarcity continues to restrain growth—reflected by rising prices
for feed, degraded grassland, and shortage of fodder crops.

Authors investigated historical events and compared Chinese trends in livestock production costs with the
rise of corresponding meat prices to gain insight about supply side factors. China Animal Husbandry and
Veterinary Yearbooks revealed that animal diseases disrupted markets in recent years. Reduced pork supplies
due to swine diseases corresponded to prominent surges in pork prices: the “blue ear disease” epizootic during
2007, the porcine epidemic diarrhea virus (PEDv) in 2011, and the African swine fever epizootic in 2018–20.
These swine disease outbreaks reduced pork supplies by killing many animals and prompting widespread
culling of additional animals to stop the spread of the diseases. These swine diseases did not directly affect
pork demand since the diseases do not spread to humans.11 China also had numerous strains of avian influ-
enza constantly circulating in the poultry population (Su et al., 2015). Widespread outbreaks in 2004, 2005,
2013, and 2014 reduced demand for poultry due to the risk of diseases spreading to humans. Poultry sales
dropped 50 to 70 percent during an outbreak in 2005 that spread to humans (Jiang, 2006). Workers Daily
(2013) said demand for poultry had “fallen off a cliff” during H7N9 avian influenza outbreaks that began in
March 2013. Xi et al. (2016) attributed reduced demand during 2013–14 to a panic that was set off by news
media reports.

Cattle and sheep diseases are a less-publicized problem that likely constrained beef and mutton supplies. In
2011 a meeting of Chinese academicians described the cattle disease issue as “grim,” with new diseases ap-
pearing and a resurgence of diseases like brucellosis, foot and mouth disease, and bovine tuberculosis (Xinhua

10 Expenditures on each of these meats accounted for a declining share of total household expenditures.
11 Government publicity campaigns during 2018 sought to assure consumers that African swine fever (ASF) is not transmitted to humans by

eating pork. Opinion polls and news articles indicated that some consumers did avoid pork after the first ASF outbreaks.

20
China’s Meat Consumption: Growth Potential, ERR-320
USDA, Economic Research Service
News, 2012).12 The meeting linked disease problems to a state of decline in the beef cattle industry and rising
beef prices. More recently, a team of China Agricultural University scholars attributed growing outbreaks
of cattle and sheep disease epizootics to scaled-up farming and growth in international trade (Wang et al.,
2020).13 These diseases included a resurgence of zoonotic brucellosis and anthrax in pastoral regions. A rising
trend in beef prices began in 2007, concurrent with a widespread outbreak of foot and mouth disease that
year, noted by Wang et al. (2020).

A public health program to reduce the incidence of schistosomiasis among humans in provinces along the
Yangtze River reduced the use of cattle and water buffalo as draft animals in several provinces (Wang et al.,
2021).14 A “substitute machinery for cattle” program that began around 2007 subsidized up to 50 percent
of purchase costs for tractors and other machinery to replace bovines that are carriers of snails that transmit
schistosomiasis to humans (Farmers Daily, 2011). Much of the beef in the region had traditionally come from
butchering cattle and water buffalo no longer fit for field work, so this program’s elimination of cattle may
have reduced beef supplies.

The 2013 edition of the Animal Husbandry and Veterinary Yearbook summarized chronic problems facing
livestock farmers in China. The yearbook cited rising prices for feed and increasing wages for hired workers
as factors that directly raise production costs. Feed is generally the largest component of production cost for
hogs and poultry and labor costs are the largest component for cattle and sheep. Gale (2017) found that feed
was the main component of rising hog costs in China. Gale and Arnade (2015) found that rising chicken
prices in China were linked to increases in corn and soybean meal prices, as well as increases in wages. The
yearbook noted that large numbers of small-scale “backyard” farmers quit the industry to pursue off-farm
work or to specialize in another enterprise, consistent with exits from the sector documented by Rae and
Zhang (2009). An influx of larger-scale commercial farms are replacing smaller farms, but the larger farms’
entry is constrained by other problems cited by the yearbook: difficulty gaining access to land and credit, and
costs of complying with environmental regulations (Gale, 2017). Beginning in 2014, a series of environmental
decrees and regulations ordered local officials to ban or limit livestock farms near residential communities,
bodies of water, scenic areas, roads, and markets, another initiative that may have reduced meat supplies.

The especially rapid growth in beef and mutton prices may be traced to concerns raised in the 2013 yearbook
about the failure of cattle and sheep production to keep pace with growth in demand for beef and mutton.
Decline in cattle numbers occurred faster than statisticians realized, reflected by sharp downward revisions of
beef cattle inventories after a 2017 agricultural census.15

Cattle and sheep were raised predominantly by extensive-margin grazing on grasslands, hillsides, and other
spaces unsuited for row-crop farming, including public spaces like roadsides and railway rights-of-way. Dur-
ing the 2000s, cattle and sheep herds were reduced by grazing bans and a series of programs that resettled
herders or paid them to remove animals from degraded grasslands.16 In major farming regions, quality pas-
ture and forage are scarce since nearly all land is used for row crops or vegetables. Industry analysts blamed
the long cycle for cattle production as a deterrent to Chinese farmers seeking quick returns on investment.

12 These diseases are spread among cattle by bacteria and viruses. Brucellosis and tuberculosis can be spread to humans, but their main impact on

beef markets is to restrain supplies through deaths and culls of infected animals.
13 The authors cite the grassland program that shifts animals from extensive grazing to fenced pasture, and confinement housing and transmission

of diseases by imported livestock or animal products.

14 Schistosomiasis is a disease spread to humans by parasitic flatworms.
15 Beef cattle numbers in revised data for 2016 were 14 percent less than estimates for the same year released in the 2017 edition of the China
Statistical Yearbook, before the census data had been incorporated. Dairy cattle numbers were also revised downward, which may have also contributed
to reduced beef supplies.
16 The “退牧还草” (reduce grazing, restore grassland) program began in pastoral provinces in 2004.

21
China’s Meat Consumption: Growth Potential, ERR-320
USDA, Economic Research Service
By 2009, a Ministry of Agriculture grassland monitoring report said animal stocking density on grasslands
exceeded capacity by 34 percent.17 A grassland protection program initiated in 2011 paid pastoral families
to reduce herds and subsidized planting of improved grass varieties, fencing of pastures, and construction of
barns for animals. Expansion of corn production during 2008–15 (incentivized by a price support for corn)
displaced grassland and fodder crop production. In 2016 the corn price support program was ended and a
grain-to-fodder program encouraged farmers to replace field corn with alfalfa and other fodder crops in pas-
toral regions.18

In 2020, an assessment of beef supply problems by China’s Academy of Agricultural Sciences raised concerns
about the long cattle cycle, outdated beef production methods, poor fodder management, and high produc-
tion costs (Tang and Zhang, 2020). China’s MARA (2021c) blamed similar problems for high beef and mut-
ton prices when MARA issued a 5-year action program for developing beef and mutton production, citing
the failure of supply to keep up with growing consumer demand.

The authors calculated the average production cost and sale price per kilogram of live hogs, chickens, cattle,
and sheep using annual farm production cost data from various editions of the Compilation of Agricultural
Commodity Production Costs and Income, published by China’s National Development and Reform Commis-
sion. These farm-level values are compared with corresponding retail meat prices in figure 15.

Inflation-adjusted production costs rose over time—to varying degrees—for each type of livestock. Hog
production cost rose 135 percent from 2000 to 2021 while sheep costs rose nearly 250 percent and beef costs
rose threefold over the same period.19 Chicken cost rose at the slowest pace: 40 percent from 2000 to 2021.
The three main components of production costs were feed, labor, and feeder stock. Feed costs are prominent
for hogs and chickens, but cattle and sheep typically graze on pastures or public areas and consume straw and
fodder gathered by farmers. Labor is the chief cost for sheep and cattle. The fluctuating price of piglets causes
hog production costs to vary cyclically.

Livestock and meat prices generally rose along with production costs. Hog prices rose above production costs
in some years—often during a disease epizootic—and fell to a point slightly above production costs in other
years. Pork prices were higher than hog prices, but fluctuations of hog and pork prices were similar. Chicken
prices tended to rise slightly faster than chicken production costs and chicken prices fluctuated less from year
to year than pork prices. Beef and mutton prices rose even faster than their production costs. Inflation-adjust-
ed beef prices rose by 300 percent during 2000–21 and mutton prices rose 230 percent over that period.

These data suggest that production costs have a role in determining the prices of meats. Prices of live animals
rarely fall below production costs, so increases in costs tend to result in higher prices. Mutton and beef (the
meats with the highest production costs) have the highest prices. Pork (the most consumed meat) has a lower
cost than beef or mutton but a higher cost than chicken. The close link between prices and costs for pork and
chicken suggests that costs are a major factor determining their final prices. However, beef and mutton prices
have risen much faster than their production costs, suggesting that growth in demand was outpacing produc-
tion and consequently driving up prices.

17 A grassland monitoring report in 2019 said grasslands were still overstocked by 10 percent.
18 In 2016, the 5-year “粮改饲” (grain to fodder) initiative began subsidizing farmers who planted fodder crops, beans, and minor grains in place
of corn.
19 This comparison excludes the 40-percent rise in hog production costs during 2020 due to a surge in feeder pig prices during the recovery from

the African swine fever epidemic.

22
China’s Meat Consumption: Growth Potential, ERR-320
USDA, Economic Research Service
Figure 15
Rising meat prices reflect farm prices and production costs for four types of meat, 2000–21

Hog-pork prices and production cost Chicken prices and production cost

Yuan per kilogram Yuan per kilogram

40 16
35 14
30 12
25 10
20 8
15 6
10 4
5 2
0 0
2000 2005 2010 2015 2020 2000 2005 2010 2015 2020

Pork price Hog price Production cost Chicken price Farm price Production cost

Cattle-beef prices and production cost Sheep-mutton prices and production cost

Yuan per kilogram Yuan per kilogram

60 60
50 50
40 40
30 30
20 20
10 10
0 0
2000 2005 2010 2015 2020 2000 2005 2010 2015 2020

Beef price Cattle price Production cost Mutton price Sheep price Production cost

Note: Data are in constant 2000 yuan, adjusted for inflation, using China’s consumer price index. Costs are a ratio of production cost
per head to weight. The farm price is from production cost data.
Source: USDA, Economic Research Service, based on data from China National Development and Reform Commission, compilation
of materials on agricultural production costs and returns, Beijing: Statistics Publishing House.

Estimating Sensitivity to Personal Income and Meat Prices

ERS authors estimated the relationship between Chinese meat consumption, income, and meat prices by
fitting two types of statistical models to annual data from 1999 to 2021: (1) a set of linear models that
regressed quantities of meat consumed on average prices and income, and (2) a meat demand system that
estimated how household meat budget shares for the three meat categories were related to prices and expendi-
ture. The models were estimated separately with disappearance data and with household purchase data.

23
China’s Meat Consumption: Growth Potential, ERR-320
USDA, Economic Research Service
Linear Consumption Models
First, authors estimated a set of linear demand equations for three meat categories using the three meat prices
and per capita disposable income as the main explanatory variables. Augmented Dickey-Fuller tests suggest
that several data series were “non-stationary”—they did not have a constant mean and variance over time.
Regressing data series that have strong time trends—for example poultry and beef/mutton consumption—on
another series with a time trend—household income—could lead to a spurious result. The models were there-
fore estimated in first differences to avoid incorrect conclusions due to such spurious relationships.

Authors estimated the following linear demand models:

Here, Qi is the quantity demanded for meat i; pj is the price of meat j with j ranging from 1 to 3 representing
pork, poultry, and beef/mutton, respectively; Y is per capita income; and Δ is the difference operator; ai , bij ,
and di are parameters to be estimated; and ei is the error term. A dummy variable IAI captures the direct
effects of major avian influenza outbreaks on poultry consumption due to fears of transmission to humans.
IAI takes a value of 1 in years when major avian influenza outbreaks occurred with reported human contrac-
tions (i.e., in years 2004 and 2013); 0 otherwise.20 Impacts of swine diseases on consumption were indi-
rect—through contractions of supply that drove up prices (see box: "Case Study of a Meat Supply Shock").
Therefore, the model does not include a swine disease variable. Similarly, no beef/mutton disease indicator is
included since impacts were mainly indirect.

The analysis used inflation-adjusted per capita household disposable income data and average annual meat
prices (see figures 10 and 11 above). Income and prices were deflated to constant 2000 currency values, using
China’s consumer price index. Based on Akaike’s Information Criteria (AIC) and the Bayesian Information
Criteria (BIC), researchers concluded that a model without a constant term is preferred over a model with a
constant term. The models are estimated by seemingly unrelated regression with robust standard errors. The
constant was excluded from the linear models based on the AIC and BIC criteria. More details about the esti-
mation and related tests are provided in appendix 3. The regression results are presented in appendix table 2.

Price and income elasticities estimated at mean are presented in table 3. Income elasticities of each type of
meat with both data sets are between 0 and 1, suggesting that the demand for each type of meat increases
(decreases) less than 1 percent when income increases (decreases) 1 percent. The inelastic response of China’s
meat consumption to income changes is consistent with the declining budget share of meat calculated above.

Pork disappearance has an insignificant relationship to income, suggesting that pork disappearance consump-
tion does not respond to an income change. Estimates using both data sets indicate that pork’s income elas-
ticity is much smaller than the income elasticities estimated for poultry, beef and mutton, suggesting that
consumption of poultry, beef and mutton is more sensitive to income growth than pork. The relatively small
income elasticities for pork suggest that this category of meat can be classified as a “necessity” that varies little
with income. This finding is consistent with pork’s role in traditional Chinese diets and the declining share of
pork in meat spending.

The magnitude of income elasticities obtained from the disappearance data, 0.41 for poultry and 0.33 for
beef and mutton, is inelastic (less than 1.0). The household-purchase income elasticities for poultry, beef

20 Major disease outbreaks were identified by consulting China Animal Husbandry and Veterinary Yearbooks and a review of the poultry industry

by Xin et al. (2016).

24
China’s Meat Consumption: Growth Potential, ERR-320
USDA, Economic Research Service
and mutton are larger in magnitude than the ones estimated from disappearance data but still inelastic. The
income elasticities suggest a slow shift in the composition of meat consumption from pork to other meats as
income rises (holding prices constant). The shift is especially slow for disappearance data.

Both data sets indicate that pork consumption is slightly sensitive to changes in its price. Analysis of both
data sets yields a negative own-price elasticity that is inelastic: -0.16 for disappearance data and -0.37 for
household purchase data. The inelastic price response suggests that a relatively large change in pork price is
needed to induce consumers to change their consumption of pork. This finding is consistent with market
events in several recent years. When swine disease reduced pork supplies, large increases in pork prices
occurred to restore equilibrium between demand and supply (see box, “Case Study of a Meat Supply Shock”).

Poultry has an insignificant own-price elasticity for both disappearance and household purchase data,
suggesting that poultry consumption is not responsive to changes in poultry price. Beef and mutton have an
inelastic negative own-price elasticity for household purchase data, indicating a decline in beef and mutton
consumption following a rise in beef and mutton price. The own price elasticity of beef and mutton for disap-
pearance data is insignificant, possibly due to the growing disappearance of beef and mutton but reduced/
restrained supply. The analysis found that the disappearance of beef and mutton had grown at a robust pace
in past years, despite large increases in prices.

Most cross-price elasticities are insignificant, indicating that consumption of meats is not very sensitive to
prices of other types of meats. However, elasticities estimated from both datasets indicate that poultry is a
substitute for pork. That is, the coefficients suggest that poultry consumption rises when the pork price goes
up. In contrast, there is no statistically significant effect of poultry price on pork consumption. The insignifi-
cant own price elasticity of poultry, but significant cross price elasticity of pork price can be explained by the
dominant role of pork in China’s meat consumption. The positive beef/mutton price elasticity for pork esti-
mated from household purchase data suggests that higher beef/mutton prices lead to higher pork consump-
tion, but this elasticity estimated from disappearance data was not statistically significant.

Table 3
Estimates of price and income elasticities for Chinese meat consumption
Demand elasticities estimated with disappearance data
Consumption of: Pork price Poultry price Beef/mutton price Income
-0.163** -0.232 0.201 0.058
Pork
(0.078) (0.239) (0.131) (0.147)
0.165*** -0.106 0.027 0.411***
Poultry
(0.039) (0.182) (0.098) (0.145)
0.021 0.030 -0.006 0.334***
Beef/mutton
(0.019) (0.092) (0.049) (0.055)
Demand elasticities estimated with household purchase data
Consumption of: Pork price Poultry price Beef/mutton price Income
-.374*** 0.136 0.183* 0.305*
Pork
(0.085) (0.268) (0.100) (0.162)
0.276*** -0.076 0.002 0.706***
Poultry
(0.057) (0.201) (0.198) (0.164)
0.078 0.219 -0.730*** 0.926***
Beef/mutton
(0.076) (0.240) (0.144) (0.206)

Note: Standard errors are in parenthesis. ***, **, and * indicate statistical significance at 1 percent, 5 percent, and 10 percent, respec-
tively. Elasticities are estimated at the mean. Regression results are presented in appendix 3, “Meat consumption regression esti-
mates.”
Source: USDA, Economic Research Service estimates based on data from China’s National Bureau of Statistics.

25
China’s Meat Consumption: Growth Potential, ERR-320
USDA, Economic Research Service
Case Study of a Meat Supply Shock
During 2018, China had a severe outbreak of African swine fever that significantly reduced the coun-
try’s meat supply during 2019 and 2020. Millions of pigs died of the disease, were culled to prevent the
spread, or were sent to slaughter early by farmers to forestall infection of their herd (Ma et al., 2021).
According to official Chinese data, China’s pork production dropped about 21 percent in 2019. Output
of poultry, beef and mutton increased, and imports of all types of meat increased, but these increases
did not fully offset the loss of domestic pork supplies. Authors’ calculations indicate that China’s per
capita output of meat declined 9.2 percent in 2018–19; disappearance of meat decreased 7.6 percent.

With less meat available, prices rose. China’s consumer price index (CPI) for meat rose 29.1 percent
during 2019, tenfold faster than the 2.9-percent increase in the overall CPI. Thus, meat prices rose
26.2 percentage points faster than the general rate of inflation that year. The increase in meat prices
net of inflation was more than three times the 7.6-percent drop in meat disappearance. The ratio of the
-7.6-percent decrease in meat quantity to the +26.2-percent increase in real meat price implies a price
elasticity of about -0.3.

Pork accounted for most of the increase in meat prices, but prices of other meats increased as well.
According to the monthly CPI report released November 2019 (when meat prices were at their 2019
peak), consumer meat prices were up 74.5 percent from a year earlier, including increases of 110.2
percent for pork, 22.2 percent for beef, and 14.3 percent for mutton (the CPI does not report poultry
prices as a separate category.) The CPI report estimated that meat prices accounted for 3.27 percentage
points of the 4.5-percent year-to-year increase in the overall CPI that month.

Household data showed a relatively small 1.7-percent decrease in per capita meat purchases in 2018–19,
much smaller than the 7.6-percent decrease in meat disappearance that year. The decrease in house-
hold purchases reflected an 11-percent decrease in pork purchases, a 20-percent increase in poultry
purchases, and a 3-percent increase in beef and mutton purchases.

The small reduction in Chinese household meat purchases seems inconsistent with the much larger
decline in pork disappearance and raises questions about the validity of the data. For example, it is
possible that the Chinese national household survey gives less weight to rural regions (figure 9) and
to regions like Sichuan and Guangdong (see figure 10) that experienced sharper declines in pork
consumption than did Beijing. The small change in household purchases might be consistent with the
change in disappearance data if nearly all the reduction in meat consumption occurred in restaurants
or food processing that is not counted by the household survey.

26
China’s Meat Consumption: Growth Potential, ERR-320
USDA, Economic Research Service
Almost Ideal Demand System (AIDS) Model for Meat
ERS authors estimated an Almost Ideal Demand System (AIDS) that is commonly used to estimate food
demand relationships to analyze changes in demand among different meats. While the linear demand models
above analyzed meat consumption and consumers’ income, this model analyzes how consumers allocate their
meat expenditure across the three categories of meat (pork, poultry, and beef/mutton), and the model assumes
that this allocation can be analyzed separately from a consumer’s overall budget decision. Figure 13 showed
that meat expenditure comprised 4 to 6 percent of household expenditure during 2012–18.

The shares of meat expenditure for the three meat categories are represented by , where E is expen-
diture on meat and pi and Qi are the price and quantity of meat category i with i = pork, poultry, and beef/
mutton. Note that . These correspond to the budget shares shown in figure 14.

The budget shares shown in figure 13 followed trends over time with no steady or constant mean. Consistent
with this observation, augmented Dickey-Fuller tests could not reject the null hypothesis that the shares of
meat consumption, wj, have a unit root, but the differenced data was stationary. ERS authors, therefore,
utilized a first-differenced AIDS to estimate meat demand relationships (Deaton and Muellbauer, 1980; Eales
and Unnevehr, 1993; Ortega et al., 2009).

Three equations were estimated with dependent variables, Δwj, corresponding to the three meat categories
with a set of explanatory variables that include each of the three meat prices and meat expenditure.

Here ln(P) is a Stone’s price index defined by ln(P) = and ∆ ln(P)= .

are parameters to be estimated. Based on AIC and BIC tests, the constant term ci was
retained in the model. All prices and expenditures were normalized by their means before taking logarithms
(Moschini and Meilke, 1989). More details about the estimation and related tests of the first-differenced
AIDS model are presented in appendix 3. Parameter estimates are presented in appendix tables 3 and 4.

The expenditure elasticities shown in table 4 were calculated from the parameter estimates. These elasticities
are “conditional” on the level of overall meat expenditure, representing the tendency to allocate expenditures
across the three meat categories, as overall meat expenditure rises. However, the presence of constant terms
in each of the first-differenced AIDS models (see appendix tables 3 and 4) means that the elasticities are also
conditional on long-term trends indicated by the constants. Since the data are expressed in differences, nega-
tive constants for pork indicate a declining trend in its budget share, while positive constants for beef/mutton
indicate rising trends. These trends in budget shares appear to be independent of the expenditure and price
variables in the model. Thus, the expenditure and price elasticities are interpreted as short-run deviations
from the long-run budget share trends.

The expenditure elasticity estimates with disappearance data show that only the expenditure elasticity for
pork is elastic, suggesting that the consumption of pork increases (decreases) more than 1 percent when the
meat expenditure grows (declines) 1 percent. Inelastic expenditure elasticities for poultry and beef/mutton
(with disappearance data) indicate that the consumption of those meats is not influenced by meat expendi-
ture growth. The household data has elastic expenditure elasticities for pork and poultry but no statistically
significant expenditure elasticity for beef/mutton. The negative trend in pork and positive trends in poultry
and beef and mutton seem to dominate the expenditure effects. Thus, the trend in shifting budget shares

27
China’s Meat Consumption: Growth Potential, ERR-320
USDA, Economic Research Service
away from pork to poultry, beef and mutton seems to be independent of expenditure growth and could reflect
some other unobserved factor, such as change in tastes.

The results shown in appendix 3 also confirm that avian influenza reduced the consumption of poultry
during years of widespread outbreaks. The two data sets indicate differing cross-commodity impacts of avian
influenza outbreaks. During years of AI outbreaks, the poultry expenditure share shifted to pork with house-
hold purchase data, while the shift was to beef/mutton with disappearance data (see appendix tables 3 and 4).

Uncompensated price elasticities reflect effects of price changes on meat consumption, including both a
substitution effect and an income effect. The first-differenced AIDS model estimates indicate that own-price
elasticities are negative for each meat category, although that for beef and mutton estimated with disappear-
ance data is not statistically significant. Estimates with disappearance data suggest that pork consumption
is more sensitive to its own price change than are the other two meats. In contrast, estimates with house-
hold purchase data suggest that poultry consumption is the most sensitive to changes in its own price. The
first-differenced AIDS model with household purchase data suggests that beef and mutton are complements
for pork. This may reflect an income effect due to pork’s large share of the meat budget (recall that pork
accounted for 60 to 80 percent of consumers’ meat budgets during 2000–20). An increase in the pork price
not only increases the price of pork relative to substitute meats, but it also tends to reduce the quantity of
all meats purchased by reducing the amount of meat that can be purchased with a given expenditure. For
example, the pork share of the meat budget rose during 2019–20, as the 120-percent increase in pork price
offset the proportionately smaller decrease in pork; the beef and mutton share of the meat budget declined,
as the available budget for them became less. Similarly, the negative uncompensated cross price elasticities for
pork with disappearance data indicate that pork consumption increases when both poultry and beef/mutton
prices go down due to an income effect and the large share of pork in consumer’s meat budgets. Lower
prices for poultry or beef/mutton may free up more funds from a consumer’s meat budget to spend on pork,
the largest component of meat spending. In contrast, the estimates indicate that beef/mutton and poultry
are substitutes. Both datasets indicate that an increase in poultry prices leads to an increase in beef/mutton
consumption and the household purchase data indicates that an increase in beef/mutton prices leads to an
increase in poultry consumption. The effect of beef/mutton prices on poultry estimated from the disappear-
ance data also indicates that the two meats are substitutes, but the elasticity is not statistically significant. The
apparent substitute relationship between poultry and beef/mutton is surprising since poultry is considered to
be a low-priced meat option and beef and mutton are considered to be more of a luxury item. The low expen-
diture elasticities for beef/mutton also contradict the perceived luxury role of these meats. The positive time
trend for the beef budget share suggests that beef consumption also has a rising trend that is independent of
overall expenditure/income and prices. It is also surprising that the estimate for the avian influenza indicator
from disappearance data also suggests that the beef/mutton budget share tends to increase during avian influ-
enza outbreaks.

28
China’s Meat Consumption: Growth Potential, ERR-320
USDA, Economic Research Service
Table 4
Estimates of uncompensated price and expenditure elasticities from the Chinese meat demand
system
Elasticities estimated with disappearance data
Consumption of: Pork price Poultry price Beef/mutton price Expenditure
-1.070*** -0.180*** -0.376*** 1.626***
Pork
(0.048) (0.023) (0.026) (0.089)
0.230 -0.371*** 0.120 0.022
Poultry
(0.155) (0.100) (0.109) (0.278)
0.051 0.103** -0.002 -0.153
Beef/mutton
(0.083) (0.051) (0.070) (0.151)
Elasticities estimated with household purchase data
Consumption of: Pork price Poultry price Beef/mutton price Expenditure
-0. 877*** -0.059 -0.141*** 1.078***
Pork
(0.052) (0.053) (0.022) (0.105)
-0.196 -1.095*** 0.259** 1.032***
Poultry
(0.201) (0.233) (0.111) (0.404)
-0.294*** 0.345*** -0.698*** 0.647***
Beef/mutton
(0.067) (0.112) (0.084) (0.134)

Note: Standard errors are in parenthesis. *** and ** indicate statistical significance at 1 percent, and 5 percent, respectively. Calcu-
lated from parameters estimated with a first-differenced Almost Ideal Demand System (AIDS) and estimated at mean. “Expenditure”
refers to expenditure on meat. These are uncompensated price elasticities that include effects of changes in relative prices and the
impact of a price change on the purchasing power of a consumer’s meat budget.
Source: USDA, Economic Research Service estimates based on data from China’s National Bureau of Statistics.

Table 5 shows “compensated” price elasticities, calculated from the first-differenced AIDS model.21 These
elasticities exclude the impact of a price change on the purchasing power of a consumer’s meat budget, solely
measuring the substitution effect. Consequently, the compensated own-price elasticities are negative and
smaller in magnitude than the uncompensated elasticities. For example, the compensated own-price elasticity
for pork (estimated with disappearance data) is -.04, compared with its uncompensated own-price elasticity of
-1.07. All compensated own-price elasticities are inelastic (absolute value is less than 1). With exclusion of the
income effect, poultry is the most sensitive to changes in its own price with both data sets. Positive compen-
sated cross-price elasticities (estimated with household purchase data) indicate substitution effects among
all three meats. Those estimated with disappearance data suggest the substitution effect between pork and
poultry, as well as beef/mutton and poultry, but the relationship between pork and beef/mutton is not statisti-
cally significant. The compensated own-price elasticities for disappearance data are smaller in magnitude than
those for household data.

21 Parameter estimates are presented in appendix table 4.

29
China’s Meat Consumption: Growth Potential, ERR-320
USDA, Economic Research Service
Table 5
Estimates of “compensated” price elasticities for Chinese meat consumption
Compensated demand elasticities, estimated with disappearance data
Consumption of: Pork price Poultry price Beef/mutton price
-0.040** 0.056*** -0.016
Pork
(0.020) (0.014) (0.011)
0.244*** -0.368*** 0.124*
Poultry
(0.062) (0.088) (0.073)
-0.046 0.081* -0.036
Beef/mutton
(0.032) (0.048) (0.048)
Compensated demand elasticities, estimated with household purchase data
Consumption of: Pork price Poultry price Beef/mutton price
-0.161*** 0.128*** 0.033***
Pork
(0.039) (0.039) (0.012)
0.490*** -0.916*** 0.426***
Poultry
(0.150) (0.185) (0.094)
0.136*** 0.457*** -0.593***
Beef/mutton
(0.050) (0.101) (0.082)

Note: Standard errors are in parenthesis. *** and ** indicate statistical significance at 1 percent, and 5 percent, respectively. Cal-
culated from parameters estimated with a first-differenced Almost Ideal Demand System (AIDS) and estimated at mean. These are
compensated price elasticities that measure pure substitution effects among meat by excluding the impact of a price change on
consumers’ purchasing power.
Source: USDA, Economic Research Service estimates based on data from China’s National Bureau of Statistics.

Predicted Meat Consumption Growth, 2022–31

ERS authors calculated predicted values from the estimated linear meat demand models to evaluate the
potential for future growth in per capita meat consumption, based on a scenario of slowing income growth
and rising meat prices.

Predicted values were calculated for years 2022–31 using projected growth in income and meat prices. Future
per capita disposable income values were calculated using per capita GDP growth rates for China from the
USDA, ERS “International Macroeconomic Database” (available online from the ERS web site). Annual
Chinese income growth was projected to fall gradually from 5.5 percent in 2022 to 4.6 percent in 2031,
reflecting growth below the historical 7.6-percent average growth rate.

ERS authors created a rising meat price scenario that assumed annual growth in real prices during 2022–31
will be the same as that in the last 10 years: 4 percent for pork, 5.8 percent for beef and mutton, and 2.4
percent for poultry. Predicted values from the disappearance and household purchase data (based on the
linear demand models) are shown for each of the three meat categories in figures 16 to 18. Predicted values
were calculated using historical data for 1999–21 for both disappearance data and household purchase data.
Predicted values for future years to 2031 were calculated using income growth and price-increase assump-
tions. The sum of pork, poultry, and beef/mutton values are shown in figure 19.

The predicted values of pork disappearance and household purchase for past years fit the actual data well.
Future predicted values indicate a steady growth that continues into future years. Future disappearance
increases roughly 9.47 kilograms/person over 11 years (2021 to 2031). Predicted household purchases of pork

30
China’s Meat Consumption: Growth Potential, ERR-320
USDA, Economic Research Service
rise a cumulative 9.81 kilograms/person between 2021 and 2031. The stronger trend in household purchases
likely reflects the larger income elasticity (0.33 versus 0.05).

Predicted values for poultry disappearance and household purchases also reflect a clear rising trend that
extrapolates past growth (figure 17). Assuming no avian influenza occurrences from 2021 to 2031, predicted
poultry disappearance rises about 7.96 kilograms/person. Similarly, predicted household purchases rise
approximately 7.21 kilograms/person from 2021 to 2031. The stronger growth in poultry consumption
reflects mainly the higher income elasticity (compared with pork), estimated from historical data. This is a
best scenario case with an assumption of no avian influenza. If intermittent occurrences of avian influenza
continue in the future, poultry consumption could be lower.

Predicted values for beef and mutton disappearance and household purchases also reveal a clear upward trend
(figure 18), although the predicted disappearance data show a weaker trend than the predicted household
purchase data during 2021–31. Predicted beef/mutton disappearance rises 3.45 kilograms/person or about 33
percent from 2021 to 2031, while predicted household purchases of beef and mutton rises 5.97 kilograms/
person or 163 percent over next 10 years.

The predicted total consumption of meat—represented by the sum of predicted pork, poultry, and beef/
mutton disappearance and purchases—rises between 2021 and 2031 (figure 19). Predicted meat disappear-
ance rises about 17 kilograms from 2021 to 2031. The sum of predicted household meat purchases rises about
19 kilograms during 2021–31. The changes result from a combination of own-price effects, cross-price effects,
and income effects.

Growth of this magnitude could make China’s per capita meat consumption among the highest in Asia.
While these data are not directly comparable, the comparison of pork, chicken, and beef presented in figure 2
suggests that a 17-kilogram gain in China’s meat consumption could put China ahead of Japan, South Africa,
and Mexico and close the 20 to 30 kilogram gap with neighbors Taiwan and South Korea. China’s meat
consumption would likely still be behind levels in the United States, Argentina, Brazil, and Canada.

Figure 16
China’s per capita pork consumption, actual and predicted

Kilograms

50
45
40
35
30
25
20
15
10
5
0
2000 2005 2010 2015 2020 2025 2030
Pork disappearance Predicted values
Pork household purchases Predicted values

Note: Assumed income values and rising meat prices for 2022–31 are described in the text.
Source: USDA, Economic Research Service based on predicted values calculated from regression coefficients, price, and income
data.

31
China’s Meat Consumption: Growth Potential, ERR-320
USDA, Economic Research Service
Figure 17
China’s per capita poultry consumption, actual and predicted

Kilograms

30

25

20

15

10

5
0
2000 2005 2010 2015 2020 2025 2030
Poultry disappearance Predicted values
Poultry purchase Predicted values

Note: Predicted values are calculated from regression coefficients, price, and income data. Assumed income values and rising meat
prices for 2022–31 are described in the text.
Source: USDA, Economic Research Service, based on predicted values calculated from regression coefficients, price, and income
data.

Figure 18
China’s per capita beef and mutton consumption, actual and predicted

Kilograms

16
14
12
10
8
6
4
2
0
2000 2005 2010 2015 2020 2025 2030
Beef/mutton disappearance Predicted values
Beef/mutton purchase Predicted values

Note: Predicted values are calculated from regression coefficients, price, and income data. Assumed income values and rising meat
prices for 2022–31 are described in the text.
Source: USDA, Economic Research Service, based on predicted values calculated from regression coefficients, price, and income
data.

32
China’s Meat Consumption: Growth Potential, ERR-320
USDA, Economic Research Service
Figure 19
China’s per capita meat consumption, actual and predicted

Kilograms

100
90
80
70
60
50
40
30
20
10
0
2000 2005 2010 2015 2020 2025 2030
Meat disappearance Predicted values
Meat purchase Predicted values

Note: Chart shows sum of predicted values for pork, poultry, beef and mutton shown in figures 16–18.
Source: USDA, Economic Research Service, based on predicted values calculated from regression coefficients, price, and income
data.

Discussion of Results

While inconsistencies in China’s meat data prevent any strong or precise conclusions, the data indicate that
China’s per capita meat consumption has potential for further growth. Consumption of each type of meat
appears to be growing, but pork has a weak relationship to growth in household income; consumption of
beef, mutton, and poultry tends to grow at a little slower pace than the growth in household income. Analysis
of Chinese meat budget shares indicates a persistent consumption shift from pork to poultry, beef, and
mutton that appears to be independent of expenditure growth and prices. Adjustments in the mix of meat
categories consumed are likely to continue.

This analysis found that Chinese household meat purchases were more sensitive to income growth than was
the disappearance of meat. This analysis seems counterintuitive, since disappearance includes away-from-
home consumption that has a higher income elasticity than at-home consumption. The rapid growth in
food service and processed food in China suggests that away-from-home marketing channels are the fastest-
growing components of meat consumption. Yet, data on household purchases of meat (which exclude away-
from-home consumption) appeared to show faster growth in meat consumption than did disappearance. The
results could have been influenced by strong growth in household purchases in the final 2 years of data. A
possible explanation is that restrictions on travel and COVID-19 lockdowns during 2020 and 2021 may have
resulted in a shift toward at-home consumption.

China’s meat consumption appears to be relatively insensitive to price changes. Meat consumption has
grown over the years despite increases in meat prices. Consumption of beef and mutton (in particular) grew
at a robust pace despite large increases in price. Short-run increases in pork prices (resulting from disease-
induced supply shocks) were proportionately much larger than the decrease in consumption, also indicating

33
China’s Meat Consumption: Growth Potential, ERR-320
USDA, Economic Research Service
price-inelastic demand. Some anecdotal evidence exists of substitution between meats. A noticeable surge in
poultry consumption was evident during 2019–20 when pork prices were record-high, but poultry consump-
tion still did not catch up with pork consumption despite the much lower price of poultry meat.

While this study focused on the potential for meat consumption to continue its past growth, supply side
constraints may prevent domestic production from rising fast enough to satisfy meat demand. For example,
initiatives to address environmental degradation impose new constraints, such as a requirement to invest
in manure collection and treatment facilities and reductions in cattle and sheep stocking ratios to restore
grasslands.

With rising Chinese demand and slower growth in meat supplies, meat prices tend to rise. The Chinese
Agriculture Ministry’s 5-year plan cited rising prices for beef and mutton, as well as generally weak interna-
tional competitiveness, among the challenges facing the sector. A 2020 China State Council “Opinion on
promoting high quality livestock development” set self-sufficiency targets for meat. The 5-year plan adopted
those targets and warned that growing meat imports led to increased uncertainty and market risk.

In view of the resource constraints on Chinese meat production, the country would not be able to sustain
growth in meat consumption without rising meat imports. ERS authors’ calculations indicate that the
imported proportion of meat in China’s disappearance rose from 2.5 percent in 2014 to 9.6 percent in 2021.
China’s per capita meat imports of 6.7 kilograms in 2021 were higher than those of many neighboring Asian
countries but were about one-third of the per capita imports of middle/high-income Asian countries South
Korea and Japan. China has opened its markets to more meat suppliers, as Chinese meat price increases indi-
cated constraints on domestic supplies. However, barriers to trade (such as stringent Chinese certifications for
foreign exporters, standards and regulations on feed additives and labeling, and bans due to disease concerns)
could restrain further growth in meat imports. Beckman et al. (2022) observed that wide spreads between
Chinese and international meat prices could be narrowed by reducing barriers to trade.

Growth in Chinese meat consumption could be slower than indicated if China’s income growth slows more
than expected. Concerns about health impacts from meat consumption and aging of the population could
also slow the growth in meat consumption. Rising prices and scarcity of raw materials for conventional meat
production in China might stimulate development of viable plant-based or artificial meat products (Ortega et
al., 2022).

34
China’s Meat Consumption: Growth Potential, ERR-320
USDA, Economic Research Service
References

Bai, J., J. Seale, and T. Wahl. 2020. Meat Demand in China: To Include or Not To Include Meat Away From
Home?, Australian Journal of Agricultural and Resource Economics 64(1):130–150.

Beckman, J., F. Gale, S. Morgan, E. Sabala, D.J. Ufer, A. Valcu-Lisman, W. Zeng, and S. Arita. 2022.
China’s Import Potential for Beef, Corn, Pork, and Wheat, ERR-310, U.S. Department of Agriculture,
Economic Research Service.

Buck, J.L. 1937. Land Utilization in China. New York: The Council on Economic and Cultural Affairs (origi-
nally published by the University of Nanking).

Chen, D., D. Abler, D. Zhou, X. Yu, and W. Thompson. 2016. A Meta-analysis of Food Demand Elasticities for
China, Applied Economic Perspectives and Policy, 38(1):50 –72.

China MARA (Ministry of Agriculture and Rural Affairs). 2001–21. “China Animal Husbandry and
Veterinary Yearbook” Beijing: China Agricultural Press.

China MARA (Ministry of Agriculture and Rural Affairs). 2020. 农业农村部畜牧兽医局关于进一步加强

畜牧业统计监测有关工作的通知 “Ministry of Agriculture and Rural Affairs Livestock and Veterinary
Bureau Notice on Further Strengthening Work Related to Livestock Industry Statistical Monitoring,”
September 2, 2020.

China MARA (Ministry of Agriculture and Rural Affairs). 2022. “China Agricultural Outlook 2022 –2031,”
Beijing: China Agricultural Science and Technology Press.

China MARA (Ministry of Agriculture and Rural Affairs). 2021a. 农业农村部关于印发 十四五全国畜牧
兽医行业发展规划的通知. Nong mu fa (2021) number 37.

China MARA (Ministry of Agriculture and Rural Affairs). 2021b. 十四五”全国畜牧兽医行业发展规划

"14th 5-Year Plan for Development of the National Livestock and Veterinary Industry,” Nong Mu Fa
(2021) number 37.

China National Development and Reform Commission. 2001–21. “Compilation of Materials on Agricultural
Product Costs and Returns,” Beijing: China Statistics Press.

China National Bureau of Statistics. 2018. 国家统计局关于7起农业普查造假弄虚作假典型案件的通报

“Notice on 7 Typical Cases of Agricultural Census Fraud and Falsification,” December 27, 2018.

China National Bureau of Statistics. 2001–22. “China Statistical Yearbook,” Beijing: China Statistics Press.

China Nutrition Society (中国营养学会). 2021. 中国居民膳食指南科学研究报告. January, 2021.

China State Council. 2014. 中国食物与营养发展纲要(2014—2020年) “China Food and Nutrition

Development Outline 2014 –20,” Guo ban fa (2014) number 3.

China State Council. 2020. 国务院办公厅关于促进畜牧业高质量发展的意见 “State Council Opinion on

Promoting High Quality Development of Livestock Industry,” Guo Ban Fa (2020) number 31.

Deaton, A., and J. Muellbauer. 1980. “An Almost Ideal Demand System,” The American Economic Review
70(3):312–326.

35
China’s Meat Consumption: Growth Potential, ERR-320
USDA, Economic Research Service
Eales, J. and L. Unnevehr. 1993. “Simultaneity and Structural Change in U.S. Meat Demand,” American
Journal of Agricultural Economics 75(2):259–268.

Economic Observer. 2009. 我国进口猪肉储备大增 通胀防范已经开始 “Our Country’s Imported Pork

Reserves Increased Greatly; Inflation Prevention Is Underway],” News report, December 11, 2009.

Farmers Daily. 2011. 实施’以机代牛’以来，血防疫区粮食产量提高了，血吸虫病发病率明显降--低’

小虫’无奈’铁牛’何 “Since Implementing ‘Substitution of Machinery for Cattle’, Grain Production
Increased in Disease Prevention Regions, Schistosomiasis Infection Rate Declined—'Little Bug’ Helpless
Against ‘Iron Bull’,” November 15, 2011.

Fuller, F., D. Hayes, and D. Smith. 2000. Reconciling Chinese Meat Production and Consumption Data,
Economic Development and Cultural Change 49(1):23–43.

Gale, F. 2015. Development of China’s Feed Industry and Demand for Imported Commodities, FDS-15K-01, U.S.
Department of Agriculture, Economic Research Service.

Gale, F. 2017. China’s Pork Imports Rise Along with Production Costs, LDPM-271-01, U.S. Department of
Agriculture, Economic Research Service.

Gale, F., and C. Arnade. 2015. Effects of Rising Feed and Labor Costs on China’s Chicken Price. International
Food and Agribusiness Management Review 18, special issue A:137–50.

Gale, F., and K. Huang. 2007. Demand for Food Quantity and Quality in China, ERR-32, U.S. Department
of Agriculture, Economic Research Service.

Gale, F., C. Valdes, and M. Ash. 2019. Interdependence of China, United States, and Brazil in Soybean Trade,
OCS-19F-01, U.S. Department of Agriculture, Economic Research Service.

Hovhannisyan, V. and B. Gould. 2011. “Quantifying the Structure of Food Demand in China: An
Econometric Approach,” Agricultural Economics 42:1–17.

Hu, Y. 1985. 胡耀邦同志关于发展畜牧业问题的谈话 “Comrade Hu Yaobang Discusses Livestock

Development Issues,” in China Agriculture Yearbook, Beijing: Agriculture Press.

Jiang, F. 2006. “Countermeasures of China’s Poultry Industry to Bird Flu,” in Ministry of Agriculture
Research Center for Rural Economy, China Rural Research Report 2006, Beijing: China Finance and
Economics Press, 242-47.

Li, Z., H. Liu, H. Xie, F. Wang. 2019. 基层畜牧业统计监测工作问题及其策略 “Problems and Strategy for
Grass Roots Statistical Monitoring Work,” Zhifu Shidai no. 11.

Liu, A., A. Han, and L. Chai. 2021. “Assessing the Nutrient Adequacy in China’s Food Supply from 1965 to
2018, Nutrients 13:2,734.

Livestock and Veterinary Bureau of China’s Ministry of Agriculture and Rural Affairs. 2020. 农业农村部
畜牧兽医局关于进一步加强畜牧业统计监测有关工作的通知 (Livestock and Veterinary Bureau of the
Ministry of Agriculture and Rural Affairs Circular on further strengthening work on Livestock Industry
Statistical Monitoring). 农牧便函(2020) no. 702.

Ma, H., J. Huang, S. Rozelle. 2004. “Reassessing China’s Livestock Statistics: An Analysis of Discrepancies
and the Creation of New Data Series,” Economic Development and Cultural Change 52(2):445–473.

36
China’s Meat Consumption: Growth Potential, ERR-320
USDA, Economic Research Service
Ma, M., H.H. Wang, Y. Hua, F. Qin, and J. Yang. 2021. “African Swine Fever in China: Impacts, Responses,
and Policy Implications,” Food Policy 102:1–14.

Moschini, G. and Meilke, K. 1989. “Modeling the Pattern of Structural Change in U.S. Meat Demand,”
American Journal of Agricultural Economics 71(2):253–261.

National Agricultural Census Office. 2000. “牲畜饲养数量，规模和结构 “Number of Livestock, Scale, and
Structure,” in 中国第一次农业普查资料开发课题研究—农村基本国情 China First Agricultural Census
Data Research Topics: National Rural Conditions,” Beijing: China Statistics Press, 13–15.

Ortega, D.L., J. Sun, and W. Lin. 2022. “Identity Labels as an Instrument To Reduce Meat Demand and
Encourage Consumption of Plant Based and Cultured Meat Alternatives in China,” Food Policy 111:1–13.

Ortega, D.L., H.H. Wang, and M. Chen. 2015. “Emerging Markets for U.S. Meat and Poultry in China,”
Choices, Second Quarter.

Ortega, D.L., H.H. Wang, and J.S. Eales. 2009. “Meat Demand in China,” China Agricultural Economic
Review 1(4):410–19.

Rathnayaka, S.D., S. Selvanathan, and E.A. Selvanathan. 2021. “Demand for Animal-Derived Food in
Selected Asian Countries: A System-wide Analysis,” Agricultural Economics 52:97–122.

Regmi, A., and J. Dyck. 2001. Effects of Urbanization on Global Food Demand, in A. Regmi, Changing
Structure of Global Food Consumption and Trade, WRS-01-1, U.S. Department of Agriculture, Economic
Research Service.

Shao, Y., T. Xiong, M. Li, D. Hayes, W. Zhang, and W. Xie. 2020. “China’s Missing Pigs: Correcting
China’s Hog Inventory Data Using a Machine Learning Approach,” American Journal of Agricultural
Economics 103(3):1082–1098.

Southern Rural Journal (南方农村报).2016.“权威解答!农业部4千猪价监测点真不是盖的 [Authoritative

Explanation! Agriculture Ministry’s 4,000 Hog Price Monitoring Points Are Not Really Covered], News
report, December 23.

Su, S., Y. Bi, G. Wong, G. Gray, G. Gao, and S. Li. 2015. “Epidemiology, Evolution, and Recent Outbreaks
of Avian Influenza Virus in China,” Journal of Virology, 89(17):8671–6.

Tang, X.F., and L.P. Zhang. 2020. “我国肉牛产业亟待全链条提速” (Our Country’s Beef Industry Urgently
Needs To Speed Up the Whole Chain). 中国科学报 China Journal of Science, December 7.

United Nations. 2022. “World Population Prospects,” Department of Economic and Social Affairs,
Population Division.

U.S. Department of Agriculture, Economic Research Service. 2022. “International Macroeconomic Data
Set.”

Wang, W., R. Bergquist, C.H. King, K. Yang. 2021. Elimination of Schistosomiasis in China: Current Status
and Future Prospects. PLOS Neglected Tropical Diseases 15(8):e0009578.

Wang, L., G. Liu, X. Liu, Y. Liu, J. Gao, B. Zhou, S. Gao, and S. Cheng. 2017. “The Weight of Unfinished
Plate: A Survey-Based Characterization of Restaurant Food Waste in Chinese Cities,” Waste Management:
3–12.

37
China’s Meat Consumption: Growth Potential, ERR-320
USDA, Economic Research Service
Wang, Y., Z. Guo, and J. Liang. 2020. 当前羊病流行情况与羊业发展 “Current Epidemic Situation of
Sheep Disease and Development of Sheep Industry,” Farmers Daily, June 23.

Wei, He. 2021. 如何做好粮食畜牧业统计调查 “How to Improve Grain and Livestock Statistical Surveys,”
中国信息报 China Information News, December 2.

Workers Daily. 2014. “H7N9冲击 家禽养殖业现“断崖式下跌” [Under Pressure from H7N9, Poultry
Farming Falls off a Cliff], news report, April 17.

Xin, X.F., Y. Zhang, and J.M Wang. 2016. 中国肉鸡产业经济问题研究综述 “Review of Research on
Economic Issues in China’s Broiler Industry,” 世界农业[World Agriculture], number 2, pages 174–78.

Xinhua News. 2012. “我国牛病防控形势严峻” [Country’s Cattle Disease Prevention and Control Situation
is Grim], news report, November 5.

Xinhua News Net. 2021. “老百姓”吃肉”需求加快升级” [Common Peoples’ ‘Eating Meat’ Demand
Rising Rapidly], March 30.

Xiao, H., Q. Chen, J. Wang, L. Oxley, and H. Ma. 2015. “The Puzzle of the Missing Meat: Food Away From
Home and China’s Meat Statistics, Journal of Integrative Agriculture,” 14(6):1033–44.

Yao, L. 2021. “如何做好畜牧业统计监测工作” [How To Improve Livestock Statistical Monitoring Work],

动物防疫[Animal Epidemic Prevention], number 1.

Yicai 第一财经. 2019. 储备肉制度40年: 历次收储投放背后玄机 “40 Years of the Meat Reserve System: The
Mystery Behind the Reserve Release,” News report, October 8.

Yicai 第一财经. 2022b. 第六批冻猪肉收储即将启动，业内预计为本轮猪周期最后一次 “The Sixth Batch

of Frozen Pork Reserve Purchases Has Begun, People in the Industry Expect This Will be the Final
Round of This Cycle,” News report, April 28.

Yu, Xiaohua. 2015. “Meat Consumption in China and Its Impact on International Food Security: Status
Quo, Trends, and Policies,” Journal of Integrative Agriculture, 14(6):989–994.

Yu, X. and D. Abler. 2014. “Where Have All the Pigs Gone? Inconsistencies in Pork Statistics in China,”
China Economic Review 30:469–84.

Zheng, Z., S. Henneberry, Y. Zhao, and Y. Gao. 2019. “Predicting the Changes in the Structure of Food
Demand in China,” Agribusiness 35(3):301–28.

Zhong, F. 1997. 关于肉类生产统计数字中的水分及其原因的分析 “Analysis of Inflation in Meat

Production Statistics and Reasons,” China Rural Economy 10:63–66.

Zhou, D., X. Yu, D. Abler, and D. Chen. 2020. “Projecting Meat and Cereals Demand for China Based on a
Meta-Analysis of Income Elasticities,” China Economic Review 59:1–15.

38
China’s Meat Consumption: Growth Potential, ERR-320
USDA, Economic Research Service
Appendix 1: China’s Meat Reserve

China maintains national, provincial, and local meat reserves in case of disaster, public health emergency,
animal disease outbreak, and for market stabilization. Reserves consist of live swine and frozen pork (beef
and mutton are held in certain regions). In 2007, China’s Ministry of Commerce released a list of more
than 200 farms and meat companies that held swine or frozen pork in reserve. Facilities and storage costs
are financed by the Chinese Government. The former Economic and Trade Commission managed reserves
until the Huashang Reserve Management Center (华商储备商品管理中心) was established by China’s State
Council in 1998.

Pork is reportedly held up to 4 months and beef and mutton 8 months (Yicai, 2019). No data are revealed on
the size of the reserve. In 2011, China’s State Council issued a document ordering large- and medium-sized
cities to maintain a 10-day supply of pork in reserves and a 7-day supply for other cities. No pork is held in
reserve for rural markets.

Reserves are also stocked with imported pork. The Chinese Economic Observer (2009) revealed that large
volumes of pork were imported during 2008 and 2009 to stock reserves. Announcements of auctions to
release pork reserves during 2020–21 stated that the reserves were composed entirely of imported pork.

Yicai (2019) cited instances of reserve releases during and after floods, an earthquake, a foot and mouth
disease outbreak, and a reserve of 40,000 metric tons during a swine disease epidemic in 2007. The reserve
was incorporated into a 2009 hog “price alert” policy, established to stabilize the market by purchasing and
releasing pork reserves, triggered mainly by the ratio of hog price to corn price. A similar mechanism was
revived in 2021.

Purchases of pork for reserves are made when pork prices are falling and the ratio of pork to corn price is low,
such as in 2013, early 2019, and 2021–22. Yicai (2022b) said 290,000 metric tons of pork purchased in 2010
were the largest purchase of reserve pork to intervene in the market. Pork reserves are sold to bolster supplies
when the ratio of pork price to corn price is high.

The first large, concentrated releases of pork reserves started in 2019 and continued into 2020 after an
African swine fever epizootic reduced pork supplies (appendix figure 1). Reserve releases were reduced in
2021, as pork output increased and prices declined sharply. Reserve releases resumed during 2022 to address
a new round of pork price increases. Annual totals of monthly releases indicate that 170,000 metric tons of
pork were released in 2019, 620,000 tons in 2020, 293,000 tons in 2021, and 650,100 tons in 2022. The
largest releases were about 1.5 percent of national pork consumption in 2020 and 1.2 percent of consump-
tion in 2022. Three releases of beef and mutton reserves of about 100 tons each were made during 2022 in
western regions Xinjiang and Tibet, and a purchase of 300 tons was made in Tibet during October 2022.

39
China’s Meat Consumption: Growth Potential, ERR-320
USDA, Economic Research Service
Appendix figure 1
China’s monthly sales of pork reserves, 2019–22
1,000 metric tons

180
160
140
120
100
80
60
40
20
0
Jan. Apr. July Oct. Jan. Apr. July Oct. Jan. Apr. July Oct. Jan. Apr. July Oct.

2019 2020 2021 2022

Source: USDA, Economic Research Service, compiled from announcements posted on the Huashang Reserve Management Center
web site.

40
China’s Meat Consumption: Growth Potential, ERR-320
USDA, Economic Research Service
Appendix 2: A Review of China’s Meat Statistics

China’s livestock production and consumption data have been based on a combination of administrative
reports, surveys, and censuses over the years. Sample surveys and decennial agricultural censuses were incor-
porated into the estimation methods beginning in the 1990s. China’s National Bureau of Statistics (NBS)
is the official source of China’s statistics. The National Animal Husbandry Station of China’s Ministry of
Agriculture and Rural Affairs (MARA) has a parallel livestock statistical reporting system. The MARA data
are used in internal reports, and some data are published in an annual yearbook and on public web sites. A
separate NBS household survey of household income and expenditures directly measures purchases of meat
and other food items.

National Bureau of Statistics Official Livestock Production Data

NBS reactivated its rural survey teams to measure agricultural output in 1978 (NBS had been suspended
during political turbulence that began in the mid-1960s), but surveys focused on the measurement of grain
and field crops. An administrative reporting system was the main mode of counting animals until livestock
surveys were launched in 1999. MARA continues an administrative reporting system that compiles data
reported by village officials responsible for veterinary affairs or bookkeeping. These reports are filed with town
officials, compiled, and passed up to prefecture, provincial, and national authorities.

Zhong (1997) called attention to the apparent inflation of Chinese meat production statistics during the
1990s. At the time, local officials had incentives to exaggerate animal numbers to show growth in production.
The incentives included meeting requirements of specific campaigns, such as the “market basket system” to
boost supplies of meat and other “nonstaple” foods and a “straw for beef” program to expand cattle produc-
tion by feeding animals stalks, straw, and other crop residues. Zhong and other authors observed a divergence
between the per capita production of meat products and direct measures of consumption reported by the
NBS household survey.

Zhong suspected that the inflation of livestock production data was a primary cause of the divergence in
the two data series. China’s first agricultural census of animal inventories and hog slaughter in 1996—the
first national direct enumeration of livestock—confirmed that livestock and meat output data was inflated.
A National Bureau of Statistics analysis of the census data estimated that per capita meat output was 36.6
kilograms, nearly 20 percent less than the 43.8-kilogram amount published in annual statistical yearbooks
(National Agricultural Census Office, 2000). The difficulty of counting widely dispersed animals was high-
lighted by the census’s finding that 135 million rural households raised hogs, 28.5 million raised sheep, 10.6
million households raised cattle, and nearly all the producers raised 10 or fewer head.

Following the census, NBS revised downward its estimates of livestock numbers and meat output. USDA,
ERS compared 1996 data published in NBS’s Rural Statistical Yearbooks for 1997 and 1998 to evaluate the
revisions. The comparison shows that Chinese aggregate meat output was revised downward by 22.5 percent.
Beef output was revised downward by 28.1 percent, mutton by 24.6 percent, and pork by 21.8 percent. NBS
did not publish poultry meat production estimates in the 2 years following the first census.22

22 NBS omitted estimates of poultry meat output from the 1998 and 1999 editions of Rural Statistical Yearbook.

41
China’s Meat Consumption: Growth Potential, ERR-320
USDA, Economic Research Service
Appendix table 1
Revisions of China's meat output data following three agricultural censuses, 1996–2016
Year Meat Pork Beef Mutton Poultry
Percent
First census 1996 -22.5 -21.8 -28.1 -24.6 NA
Second census 2006 -12.0 -10.5 -23.1 -22.5 -9.5
Third census 2016 1.1 2.4 -13.9 0.2 6.0

NA=not available.
Note: Revisions are calculated by USDA, ERS by comparing data released before and after agricultural censuses.
Source: USDA, Economic Research Service calculations using meat production data published in China’s National Bureau of Statis-
tics, Rural Statistical Yearbook.

In 2000, NBS began reporting livestock production data based on a dual statistical system. A survey moni-
tored animal inventories and slaughter for a sample of rural households, while other livestock-producing
units (presumably including state farms, collectives, cooperatives, companies, and institutional farms) had
to file regular reports of livestock data with statistical departments. NBS expanded the sample data to make
national and provincial estimates based on the results of the 1996 census.

A second agricultural census was conducted for 2006, but the report included only one table showing swine,
cattle, sheep, and poultry inventories. Meat output data reported in the 2008 Rural Statistical Yearbook
included a footnote explaining that livestock data for 2000–2006 included in the yearbook had been revised
based on results of the agricultural census. Comparison with data in the 2007 yearbook indicated that meat
output for the census year 2006 was revised down 12 percent. The pork revision was smaller (-10.5 percent)
than the 1996 revision, but beef (-23.1 percent) and mutton (-22.5 percent) were similar in magnitude to revi-
sions following the first census. Poultry meat output was revised downward by 9.5 percent.23

NBS revised meat data again following a third agricultural census conducted for 2016. Only a brief commu-
nique was issued about the census results. The 2018 Rural Statistical Yearbook included a footnote explaining
that livestock data for 2007–17 had been revised based on census results (the description of livestock data in
the appendix did not mention the census). A comparison of data in 2017 and 2018 yearbooks showed a much
smaller 1.1-percent upward revision in meat production. Poultry, pork, and mutton data were revised upward,
but beef output was revised downward by 12.9 percent. Historical revisions of beef output erased growth that
had previously been reported in yearbooks for 2008–16.

China’s agricultural censuses are also subject to many of the same problems. News reports (and at least one
article issued by NBS) revealed that some officials had filled in census forms in their offices using data from
previous censuses, other records, or their own estimates. An article by NBS cited examples of fraud detected
in the third agricultural census that included evidence of altered or fabricated data, failure of enumerators
to personally interview farmers, report forms filled in with administrative data, and inconsistencies between
report forms and data transmitted to NBS.

A circular issued by the Livestock and Veterinary Bureau of China’s Ministry of Agriculture and Rural Affairs
(2020) instructed local officials to improve livestock data collection and reporting acknowledged “prominent
problems.” These included lack of attention to the task in many localities, lack of budgetary support, incom-
plete records on farms, late or inaccurate reporting of data, and unclear responsibilities for livestock reporting
at the county level. A county statistical bureau in Guangdong Province reported that statisticians lacked

23 Explanatory text in appendices of NBS yearbooks from 2000 to 2017 described the “successful” completion of the 2006 census, but continued

to state that livestock data estimates were based on the 1996 census.

42
China’s Meat Consumption: Growth Potential, ERR-320
USDA, Economic Research Service
knowledge of the industry due to rapid personnel turnover; statisticians were not permitted to enter swine
or poultry farms to verify data due to biosecurity measures; and statisticians did not interface or exchange
data with related departments (Wei, 2021). A county bureau in Gansu Province complained that veterinary
personnel who were given responsibility for livestock statistical work were under-qualified and neglected
statistical data due to animal disease epidemics (Yao, 2021). The Gansu article said that increased attention to
counting animals to support poverty alleviation programs led to a neglect of counting other farms, including
those farms raising domesticated wildlife. The Gansu article noted that statisticians struggled to count the
growing volume of livestock transported from other counties and provinces. Statisticians collected data by
telephone from far-flung farmers, some of whom intentionally concealed livestock from statisticians.

Ministry of Agriculture and Rural Affairs Livestock Reporting

The MARA livestock reporting is based on monitoring of 4,000 fixed observation points that include a
sample of large-scale pig farms, villages, and rural households. These data do not directly measure meat
consumption and this survey is independent of the NBS survey used to develop national meat production
estimates. Authors discuss this data collection system to gain perspective on official meat production data
that comprise the main component of meat disappearance.

A Southern Rural Journal (2016) interview with a MARA official explained that the livestock reporting
system had consisted of crude annual reports filed once per year by veterinary technicians and village accoun-
tants until 2008. That year, a fluctuation in swine numbers due to a disease outbreak prompted China’s State
Council to fund monthly livestock reporting and standardize the reports. The official noted that the MARA
system duplicates NBS surveys to some degree but only NBS can report national data. The main function
of these data is to provide monthly percentage changes in swine and sow inventories that are reported on a
MARA web site (statistics on levels of swine inventories occasionally appear in news articles or presentations
made by officials but are not published). According to the Southern Rural News (2016) interview, the MARA
reporting system bypasses prefecture and county officials to prevent tampering with the data. The system
incorporates direct reporting by large-scale farms to a central data cloud platform.

Documents and news releases about an initiative to improve the MARA system during 2019–20 revealed
the difficulties of counting livestock. These problems may also affect the quality of surveys conducted by the
National Bureau of Statistics.

A group of officials at a Shanxi Province livestock and veterinary extension center observed that village-level
reporters often do not clearly understand technical livestock concepts in the forms the reporters are asked
to complete, and personnel often neglect to file reports because they receive no compensation (Li et al.,
2019). The authors observed that many farmers kept rudimentary records that did not conform to the items
requested on statistical report forms; surveyors were often prohibited from entering farms to verify data due
to biosecurity measures; computer equipment in villages was often poor; and data were vulnerable to loss or
errors because the data was stored on multiple machines or media.

Similarly, a September 2020 document issued by the MARA Livestock and Veterinary Bureau called for
rectifying persistent problems in livestock monitoring work. According to the document, some localities
had lapses in reporting, errors, and unverified data due to personnel turnover and a low priority on livestock
reporting. Moreover, poor management and conflict with other departments could result in the omission
of records, multiple sets of data, and numbers that are confused or inaccurate. The document ordered local
personnel to verify the accuracy of data, carry out spot checks of large-scale farms, and establish a record
system for small farms by the end of 2020.

43
China’s Meat Consumption: Growth Potential, ERR-320
USDA, Economic Research Service
Household Income and Expenditure Surveys
The household income and expenditure survey (described in the appendix of Gale and Huang (2006)) reports
average per capita purchases of meats, consumer expenditure, and personal income based on accounts kept by
a national sample of households that recorded income and expenditures throughout the year. This survey has
been conducted annually since the early 1980s.

According to documentation in the 2012 edition of the China National Bureau of Statistics China Yearbook
of Household Survey, parallel rural and urban surveys collected income and expenditures from 74,000 rural
households and 66,000 urban households. In 2013, the household survey was reorganized as a national
survey. The 2016 edition of the yearbook said the household survey was conducted for a sample of 160,000
households in urban and rural communities based on the households’ place of residence. The sample had
previously been chosen on the basis of the locality where they are officially registered as residents (the hukou),
thus potentially excluding migrants who did not actually live where they were registered. The 2012 yearbook
stated that one-third of the urban household sample was rotated each year and the entire rural sample was
changed every 5 years. The yearbook documentation states that the standard error for income estimates was
less than 1 percent.

According to documentation in the 2013 edition of the China Yearbook of Household Survey (applying to
data collected in 2012), urban “meat” included purchases of fresh, live, and processed products; and various
domestic livestock, poultry, and wild meat. Fresh and frozen pork, beef, and mutton included internal organs,
heads, feet, skin, bones, and blood, but not processed products. Poultry included all domestic and wild
poultry. The rural survey said it reported live, fresh, and chilled/frozen meat based on the estimated equiva-
lent weight as fresh meat, as well as canned meat and cured pork. Rural poultry was based on weight after
removal of feathers and internal organs. No details on meat items were provided in more recent editions of
the yearbook.

Beginning in 2013, the household survey counted income of rural migrants living in cities as urban income.
The documentation did not address expenditures and purchases, but USDA, ERS authors presume that
purchases of meat were also counted based on where the household resides. The pre-2013 sampling method
may have understated rural per capita consumption by counting migrants as members of rural households
even though the migrants were working or at school elsewhere most of the year and therefore did not
consume food at their registered place of residence.

Discontinuities in the meat consumption were reported separately for rural and urban households. In
2013—the year when the survey was reorganized—rural pork consumption jumped by 6 kilograms and rural
poultry consumption jumped by 1.5 kilograms. The discontinuities were not evident in the weighted averages
of rural and urban consumption analyzed in this report.

Sample households kept accounts of food expenditures at stores, markets, cafeterias, and restaurants
throughout the year. Expenditures on away-from-home food and beverage service were reported in statis-
tical tables, but the amounts of food consumed away from home were not counted. The amount of meat
in processed ready-to-eat meals is also difficult to count. With rising food consumption away from home
and in processed forms, it seems likely that these consumption statistics understated the growth in meat
consumption.

Some analysts have questioned whether the burden of keeping diaries throughout the year may prompt
higher-income and busier households to decline participation. While respondents receive a small subsidy it
may not offset the burden of recording purchases throughout the year. Annual consumption quantities are
likely to be understated if some respondents failed to record or transmit data during part of the year.

44
China’s Meat Consumption: Growth Potential, ERR-320
USDA, Economic Research Service
Appendix 3: Meat Demand Estimates

Authors conducted a stationarity test of meat demand to check if statistical properties of mean demand data
changed over time. Augmented Dickey-Fuller (ADF) tests fail to reject the null hypothesis that each meat
demand, measured by both disappearance and household purchase, has a unit root at all common signifi-
cance levels. Therefore, authors took first differences of the variables to transform them to be stationary and
estimated the following differenced linear consumption models: as described in the text.

pork, poultry, beef/mutton as described in the text.

There is a concern that prices may be endogenous in demand estimation as demand may affect prices during
the price formation procedure. Without addressing the endogeneity issue, the estimates may be biased. To
test and control for the potential price endogeneity, researchers utilized a generalized method of moments
(GMM) with instrumental variables. The two main components of production costs were used as instru-
mental variables for meat prices. These variables included cost categories “material and services” (mainly
consisting of feed cost) and labor cost for producing each kilogram of carcass weight of each type of meat.
Production cost data were obtained from the Chinese National Development and Reform Commission’s
annual production cost survey for hogs, chickens, beef cattle, and sheep. Costs per head were divided by
reported average weight of the animal to obtain cost per kilogram.

Hansen’s J statistic suggested that the instruments were uncorrelated with the error term and thus valid.
The null hypothesis that prices are exogenous cannot be rejected at all common significance levels by the C
(difference-in-Sargan) statistic. One reason for the exogeneity of prices is that consumers generally take prices
as predetermined when they make purchase decisions. A constant was not included in the models since the
model without a constant has lower AIC and BIC than the one with a constant.

The model was estimated using seemingly unrelated regression with robust standard errors. The regression
estimates from annual data for 1999–2021 household purchase data and disappearance data are shown in
appendix table 2. The parameter estimates of meat prices and income were used to calculate price and income
elasticities, which are shown in table 3. The estimation from the differenced linear demand models shows that
avian influenza reduced poultry consumption significantly.

45
China’s Meat Consumption: Growth Potential, ERR-320
USDA, Economic Research Service
Appendix table 2
Meat consumption regression estimates
Disappearance of: Household purchases of:
Pork Beef/mutton Poultry Pork Beef/mutton Poultry
-0.703*** -0.825***
IAI
(0.174) (0.150)
-0.325** 0.009 .113*** -.364*** 0.011 .109***
Δ (pork price)
(.155) (.008) (.027) (.083) (.011) (.022)
.232 -0.001 .011 0.103* -0.061*** 3.50e-4
Δ (beef/mutton price)
(.151) (.012) (.039) (.056) (.012) (.045)
-0.696 0.019 -0.110 0.199 0.048 -0.045
Δ (poultry price)
(.719) (.058) (.188) (.394) (.052) (.119)
1.88e-4 2.28e-4*** 4.59e-4*** 4.84e-4* 2.17e-4*** 4.54e-4***
Δ (income)
(4.77e-4) (0.37e-4) (1.62e-4) (2.58e-4) (0.48e-4) (1.06e-4)

Observations 22 22 22 22 22 22
R-Squared 0.45 0.59 0.56 0.73 0.64 0.54

Note: Standard errors are in parentheses. IAI was retained in the models based on AIC and BIC (Akaike’s Information Criteria and
Bayesian Information Criteria). “e-4” represents “x 10-4.” Models were estimated by seemingly unrelated regression with robust
standard errors. ***, **, and * indicate statistical significance at 1 percent, 5 percent, and 10 percent, respectively. Both data are for
1999–2021.
Source: USDA, Economic Research Service, based on data from China’s National Bureau of Statistics.

A second set of estimates was made based on Almost Ideal Demand System (AIDS) developed by Deaton and
Muellbauer (1980). ADF tests failed to reject the null hypothesis that each meat expenditure share had a unit
root at all common significance levels, with both disappearance data and household purchase data. USDA,
ERS researchers, therefore, adopted the first-differenced AIDS model (Deaton and Muellbauer, 1980; Eales
and Unnevehr, 1993; Ortega et al., 2009). Dependent variables in the equations are meat budget shares,
, where Δ represents the difference operation, pi and Qi are price and quantity of meat in cate-
gory i (i=pork, poultry, and beef/mutton) and E is meat expenditure. Note that .
The expenditure share generated from first-differenced AIDS is expressed as

where are parameters to be estimated; pj is the price of meat category j with j ranging from 1
to 3 representing pork, poultry, and beef/mutton, respectively; ln(P) is a Stone’s price index defined by
. The dummy variable IAI is included to capture the impacts
of avian influenza on meat consumption. The constant ci is included to capture the trend effect in the differ-
enced budget share equation, based on the AIC and BIC tests as the model with the constant c_i has lower
AIC and BIC than does the model without the constant. The expenditure elasticity, which measures the
percentage change in meat demand caused by 1 percent change in meat expenditure, is calculated as
. The uncompensated price elasticity is calculated as and compensated price
elasticity is the Kronecker delta .
All elasticities are calculated at mean.

Estimates of the first-differenced AIDS system of equations were made with restrictions on the parameters of
“adding up” , “homogeneity”
. Prices and the expenditure are normalized by dividing by their sample mean before the
logarithmic transformation. To avoid simultaneity problem, one-period lagged wi is used in Stone’s price
index (Eales and Unnevehr, 1993).

46
China’s Meat Consumption: Growth Potential, ERR-320
USDA, Economic Research Service
Several studies (e.g., LaFrance, 1993; Eales and Unnevehr, 1993; Dhar et al., 2003) pointed out that price
and expenditure may be endogenous in the estimation of a food demand system. To test for the presence of
endogeneity in price and expenditure, USDA, ERS researchers utilized a Durbin, Wu, and Hausman
(DWH) test for the consistency of parameter estimates (LaFrance, 1993). Researchers applied two sets of
parameter estimates from the demand system that control for the endogeneity and that do not control to the
DWH test. To control for the endogeneity of price and expenditure, price is specified as a function of supply
shifters and expenditure as a function of income (Dhar et al., 2003). The reduced-form price equations are
specified as , where Fj is the material and service cost mainly consisting of feed
cost and Lj is the labor cost for producing one unit of carcass weight of the are
parameters to be estimated. The explicit specification captures the supply side of the price formation mecha-
nism (Dhar et al., 2003). The expenditure equation is formulated as , where Y is real
household income and g0 and g1 are parameters to be estimated. Given four hypothesis that are H0: prices
and income are exogenous; H1: prices are endogenous and income is exogenous; H2: prices are exogenous
and income is endogenous; and H3: prices and income are endogenous, researchers conducted three DWH
tests, namely, H0 vs. H1, H0 vs. H2, and H0 vs. H3. Following Dhar et al. (2003), the test statistic is
computed as represents parameter estimates and
the corresponding parameter covariance matrices, respectively, under H0; and represent
those from the model that controls for the endogeneity of price and/or expenditure under Hh with h=1, 2,
and 3. The model under H0 is estimated by using iterative seemingly unrelated regression (SUR) and models
under Hh are estimated by using a full information maximum likelihood (FIML) estimation procedure. Dh is
asymptotically distributed as X2(k), where k is the number of potentially endogenous variables. The test
suggests that prices are endogenous in estimation with both disappearance and household purchase data sets.
One possible reason for price endogeneity is that prices and budget shares are affected by common factors
unobserved by the econometrician. The test values of Dh are available from USDA, ERS researchers upon
request.

The estimates from the first-differenced AIDS model (using disappearance data with a full information
maximum likelihood (FIML) method and with endogeneity taken into account) are reported in appendix
table 3. Statistically significant constant terms in each equation indicate persistent trends in budget shares
for each of the meat categories. The negative constant in the pork equation suggests a declining pork budget
share, while the positive constant terms for poultry and beef and mutton indicate a rising trend in budget
shares for those meats. These trends are consistent with those shown in figure 13. These trends are statisti-
cally significant when expenditure and price variables are included in the model, suggesting that long-term
shifts in budget shares occur independent of changes in expenditure growth and changes in relatively prices
of meats.

I AI coefficients indicate the occurrence of avian influenza reduced the budget share for poultry and induced
more consumption of beef/mutton, but there was no significant impact on pork. Similar results are obtained
by using household purchase data, except that more budget shares switch to pork consumption with occur-
rences of avian influenza and there is no trend in poultry budget share. Estimates for meat price and expendi-
ture elasticities are discussed in the main text.

47
China’s Meat Consumption: Growth Potential, ERR-320
USDA, Economic Research Service
Appendix table 3
Parameter estimation of first-differenced AIDS Model (disappearance data)
Pork Poultry Beef/mutton
-0.012*** 0.005*** 0.007***
Constant
(0.003) (0.002) (0.002)
0.004 -0.011** 0.007**
IAI
(0.006) (0.005) (0.004)
0.397*** -0.142*** -0.255***
Δ ln (expenditure/P)
(0.056) (0.040) (0.033)
0.207*** -0.056*** -0.150***
Δ ln (pork price)
(0.012) (0.009) (0.007)
0.071*** -0.014
Δ (poultry price)
(0.009) (0.011)
0.164***
Δ ln (beef-mutton price)
(0.011)

Observations 22 22 22
R-Squared 0.92 0.78 0.94

Note: Standard errors are in parenthesis. ***, **, and * indicate statistical significance at 1 percent, 5 percent, and 10 percent, respec-
tively. AIDS=Almost Ideal Demand System.
Source: Estimated by USDA, Economic Research Service, with data from China’s National Bureau of Statistics, Administration of
Customs, National Development and Reform Commission, Ministry of Agriculture and Rural Affairs.

Appendix table 4
Parameter estimation of first-differenced AIDS Model (household purchase data)
Parameters Pork equation Poultry equation Beef/mutton equation
-0.009** 0.003 0.006***
Constant
(0.004) (0.004) (0.002)
0.031*** -0.026** -0.005
IAI
(0.011) (0.011) 0.004
0.052 0.006 -0.057***
Δ ln (expenditure/P)
(0.070) (0.070) (0.022)
0.116*** -0.016 -0.085***
Δ ln (pork price)
(0.026) (0.032) (0.008)
-0.016 0.046***
Δ (poultry price)
(0.032) (0.016)
0.040***
Δ ln (beef-mutton price)
(0.016)

Observations 22 22 22
R-Squared 0.68 0.31 0.82

Note: Standard errors are in parenthesis. ***, **, and * indicate statistical significance at 1 percent, 5 percent, and 10 percent, respec-
tively. AIDS = Almost Ideal Demand System.
Source: Estimated by USDA, Economic Research Service, with data from China’s National Bureau of Statistics, National Development
and Reform Commission, Ministry of Agriculture and Rural Affairs.

48
China’s Meat Consumption: Growth Potential, ERR-320
USDA, Economic Research Service
Appendix 4: Data Tables

Appendix table 5
China's per capita disappearance of pork, poultry, beef and mutton, 2000–21
Per capita output Per capita net imports Per capita disappearance
Beef/ Beef/ Beef/
Pork Poultry Pork Poultry Pork Poultry
mutton mutton mutton
Year Kilograms
2000 31.3 9.4 4.0 0.1 0.4 0.0 31.4 9.7 4.1
2001 31.7 9.5 4.0 0.0 0.2 0.0 31.7 9.7 4.0
2002 32.1 9.7 4.1 0.0 0.2 0.0 32.1 9.9 4.1
2003 32.8 10.2 4.2 0.0 0.3 0.0 32.7 10.4 4.2
2004 33.4 10.4 4.3 -0.2 0.1 0.0 33.2 10.4 4.3
2005 34.8 11.2 4.3 -0.2 0.2 0.0 34.7 11.4 4.3
2006 35.4 10.4 4.4 -0.2 0.3 0.0 35.2 10.7 4.4
2007 32.6 11.0 4.7 0.0 0.5 0.0 32.6 11.4 4.7
2008 35.3 11.5 4.7 0.2 0.5 0.0 35.5 12.0 4.7
2009 37.0 12.0 4.7 0.0 0.4 0.0 37.0 12.4 4.7
2010 38.3 12.4 4.7 0.1 0.3 0.0 38.4 12.6 4.7
2011 38.0 12.7 4.5 0.3 0.2 0.1 38.3 12.8 4.6
2012 40.0 13.4 4.5 0.3 0.2 0.1 40.4 13.7 4.6
2013 41.1 13.2 4.5 0.4 0.3 0.4 41.5 13.4 4.9
2014 42.3 12.7 4.5 0.3 0.2 0.4 42.6 12.9 4.9
2015 40.8 13.2 4.5 0.5 0.1 0.5 41.3 13.3 5.0
2016 39.0 14.4 4.4 1.1 0.3 0.6 40.1 14.6 5.0
2017 38.9 14.2 4.5 0.8 0.2 0.7 39.8 14.3 5.2
2018 38.4 13.8 4.6 0.8 0.2 1.0 39.3 14.0 5.5
2019 30.2 15.9 4.7 1.4 0.4 1.5 31.6 16.3 6.2
2020 29.1 16.7 4.8 3.0 1.0 1.8 32.2 17.7 6.5
2021 37.5 16.8 4.9 2.5 0.9 1.9 40.0 17.7 6.9

Note: Meat output data are from China's National Bureau of Statistics; Net imports = imports - exports; Disappearance = output +
net imports. Per capita values are calculated using population data.
Source: USDA, Economic Research Service based on data from China National bureau of Statistics.

49
China’s Meat Consumption: Growth Potential, ERR-320
USDA, Economic Research Service
Appendix table 6
China's per capita household meat purchases, 2000–21

Pork Poultry Beef/mutton

Year Kilograms
2000 14.5 3.8 1.9
2001 14.3 3.8 1.9
2002 16.3 5.4 1.9
2003 16.5 5.6 2.1
2004 15.9 5.3 2.3
2005 17.6 5.9 2.4
2006 17.5 5.7 2.5
2007 15.6 6.5 2.6
2008 15.8 6.1 2.3
2009 17.1 7.3 2.5
2010 17.6 7.2 2.6
2011 17.6 7.6 3.0
2012 18.0 7.8 2.9
2013 19.8 7.2 2.5
2014 20.0 8.0 2.5
2015 20.1 8.4 2.8
2016 19.6 9.1 3.3
2017 20.1 8.9 3.2
2018 22.8 9.0 3.3
2019 20.3 10.8 3.4
2020 18.2 12.7 3.5
2021 25.2 12.3 3.9

Note: Data for 2000–11 are weighted averages of rural and urban data.
Source: USDA, Economic Research Service, based on China’s National Bureau of Statistics.

50
China’s Meat Consumption: Growth Potential, ERR-320
USDA, Economic Research Service
Appendix table 7
Average meat prices in China, 2000–21
All meat
Beef-mutton
Pork Chicken Beef Mutton weighted
average
average
Year Yuan per kilogram
2000 14.01 9.71 14.02 17.03 15.04 13.28
2001 14.14 10.07 17.18 15.57 16.61 13.64
2002 13.67 10.14 18.00 16.29 17.40 13.44
2003 13.69 10.35 19.28 17.26 18.54 13.65
2004 16.73 11.29 19.30 18.24 18.91 15.91
2005 17.01 11.70 19.39 19.71 19.51 16.22
2006 15.68 11.00 19.56 19.52 19.55 15.28
2007 22.35 13.45 23.06 23.44 23.20 20.57
2008 27.57 15.01 32.72 30.99 32.06 25.52
2009 19.43 15.46 34.38 33.64 34.10 20.56
2010 20.36 16.32 35.05 37.69 36.08 21.57
2011 27.96 18.53 39.01 46.15 41.81 27.68
2012 26.14 18.94 50.59 55.16 52.39 27.78
2013 25.96 18.57 63.04 63.03 63.04 28.88
2014 24.39 19.38 66.64 65.92 66.35 28.46
2015 26.55 19.94 67.23 62.71 65.35 29.95
2016 30.78 20.29 67.03 59.42 63.78 32.49
2017 27.45 20.05 67.39 60.03 64.26 30.50
2018 24.11 20.72 68.54 64.89 66.99 29.00
2019 35.36 23.71 76.57 74.68 75.77 38.05
2020 53.97 23.64 87.41 83.02 85.55 49.41
2021 35.28 23.24 89.90 86.79 88.58 39.32

Note: Data are not adjusted for inflation; 1991–2021 data are annual averages from monthly data.
Source: USDA, Economic Research Service, based on Chinese National Development and Reform Commission retail food prices,
2000–2008; China’s National Bureau of Statistics average food prices, 2009–17; Ministry of Agriculture and Rural Affairs wholesale
market prices plus estimated retail margin, 2018–21.

51
China’s Meat Consumption: Growth Potential, ERR-320
USDA, Economic Research Service
Appendix table 8
China's income, inflation, and population data, 2000–21
Per capita Change in Change in
Annual in-
disposable consumer real per capita Population Population growth
come growth
income price index income
Year Yuan Percent Percent Percent Million Million Percent
2000 3,721 6.8 0.4 6.4 1,267 9.6 0.76
2001 4,070 9.4 0.7 8.7 1,276 8.8 0.70
2002 4,532 11.4 -0.8 12.2 1,285 8.3 0.65
2003 5,007 10.5 1.2 9.3 1,292 7.7 0.60
2004 5,661 13.1 3.9 9.2 1,300 7.6 0.59
2005 6,385 12.8 1.8 11.0 1,308 7.7 0.59
2006 7,229 13.2 1.5 11.7 1,314 6.9 0.53
2007 8,584 18.7 4.8 13.9 1,321 6.8 0.52
2008 9,957 16.0 5.9 10.1 1,328 6.7 0.51
2009 10,977 10.2 -0.7 10.9 1,335 6.5 0.49
2010 12,520 14.1 3.3 10.8 1,341 6.4 0.48
2011 14,551 16.2 5.4 10.8 1,349 8.3 0.62
2012 16,510 13.5 2.6 10.9 1,359 10.1 0.75
2013 18,311 10.9 2.6 8.3 1,367 8.0 0.59
2014 20,167 10.1 2.0 8.1 1,376 9.2 0.67
2015 21,966 8.9 1.4 7.5 1,383 6.8 0.49
2016 23,821 8.4 2.0 6.4 1,392 9.1 0.65
2017 25,974 9.0 1.6 7.4 1,400 7.8 0.56
2018 28,228 8.7 2.1 6.6 1,405 5.3 0.38
2019 30,733 8.9 2.9 6.0 1,410 4.7 0.33
2020 32,189 4.7 2.5 2.2 1,412 2.0 0.14
2021 35,128 9.1 0.9 8.2 1,413 0.5 0.03

Source: USDA, Economic Research Service, based on China’s National Bureau of Statistics "Chashu" database.

Appendix table 9
Correlation between Chinese household purchases and disappearance of meat

Type of meat 1999–2021 1999–2013 2014–21

Correlation coefficient
Pork .705 .866 .290
Poultry .984 .940 .987
Beef and mutton .948 .866 .881

Note: Table shows correlations between year-to-year changes in per capita disappearance and household purchases for each meat
and time period.
Source: USDA, Economic Research Service, analysis of data from appendix tables 5 and 6.

52
China’s Meat Consumption: Growth Potential, ERR-320
USDA, Economic Research Service
Appendix figure 2
Differences between Chinese per capita disappearance and household purchases of meat
(2000–21)
Percent of household purchases

250

200

150

100

50

0
2000 2005 2010 2015 2020

Pork Poultry Beef/mutton

Note: This chart shows (disappearance - household purchases)/(household purchases) x 100 calculated from data in appendix tables
5 and 6.
Source: USDA, Economic Research Service, based on data from China National Bureau of Statistics.

53
China’s Meat Consumption: Growth Potential, ERR-320
USDA, Economic Research Service