• TABLE OF CONTENTS
HIDE
 Front Cover
 Title Page
 Table of Contents
 List of Tables
 List of Illustrations
 Foreword
 Acknowledgement
 Summary
 Introduction
 Cereal feed situation in 1980
 Trends in cereal feed use
 Growth in cereal feed use in Taiwan...
 Dynamics of demand for food and...
 Projections of cereal feed use...
 Conclusions and policy implica...
 Appendix 1. Data and methodolo...
 Appendix 2. Classification of countries...
 Appendix 3. Supplementary...
 Bibliography
 Back Cover














Group Title: Research report
Title: Cereal feed use in the Third World
CITATION THUMBNAILS PAGE IMAGE ZOOMABLE
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/UF00085374/00001
 Material Information
Title: Cereal feed use in the Third World past trends and projections to 2000
Series Title: Research report
Physical Description: 68 p. : ill. ; 26 cm.
Language: English
Creator: Sarma, J. S
Publisher: International Food Policy Research Institute,
International Food Policy Research Institute
Place of Publication: Washington D.C
Publication Date: 1986
Copyright Date: 1986
 Subjects
Subject: Feed industry -- Forecasting -- Developing countries   ( lcsh )
Grain trade -- Forecasting -- Developing countries   ( lcsh )
Animal industry -- Forecasting -- Developing countries   ( lcsh )
Meunerie -- Prévision -- Pays en voie de développement   ( rvm )
Céréales -- Commerce -- Prévision -- Pays en voie de développement   ( rvm )
Industrie animale -- Prévision -- Pays en voie de développement   ( rvm )
Genre: bibliography   ( marcgt )
statistics   ( marcgt )
non-fiction   ( marcgt )
 Notes
Bibliography: Includes bibliographical references (p. 67-68).
Statement of Responsibility: J.S Sarma.
General Note: "December 1986."
 Record Information
Bibliographic ID: UF00085374
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: oclc - 15017840
lccn - 86027355
isbn - 0896290581 (pbk.)

Table of Contents
    Front Cover
        Page 1
        Page 1a
    Title Page
        Page 2
        Page 3
    Table of Contents
        Page 4
    List of Tables
        Page 5
    List of Illustrations
        Page 6
    Foreword
        Page 7
    Acknowledgement
        Page 8
    Summary
        Page 9
        Page 10
        Page 11
    Introduction
        Page 12
        Page 13
        Page 14
    Cereal feed situation in 1980
        Page 15
        Page 16
        Page 17
        Page 18
        Page 19
    Trends in cereal feed use
        Page 20
        Page 21
        Page 22
        Page 23
        Page 24
        Page 25
        Page 26
        Page 27
        Page 28
        Page 29
    Growth in cereal feed use in Taiwan and the Republic of Korea
        Page 30
        Page 31
        Page 32
        Page 33
        Page 34
    Dynamics of demand for food and feedgrains in selected developing countries
        Page 35
        Page 36
        Page 37
        Page 38
        Page 39
        Page 40
    Projections of cereal feed use in developing countries
        Page 41
        Page 42
        Page 43
        Page 44
        Page 45
        Page 46
    Conclusions and policy implications
        Page 47
        Page 48
        Page 49
        Page 50
        Page 51
        Page 52
    Appendix 1. Data and methodology
        Page 53
        Page 54
        Page 55
    Appendix 2. Classification of countries selected for the study
        Page 56
        Page 57
        Page 58
        Page 59
    Appendix 3. Supplementary tables
        Page 60
        Page 61
        Page 62
        Page 63
        Page 64
        Page 65
        Page 66
    Bibliography
        Page 67
        Page 68
    Back Cover
        Page 69
        Page 70
Full Text
RESEARCH REPORT 57
i57





CEREAL FEED USE IN THE
THIRD WORLD: PAST TRENDS
AND PROJECTIONS TO 2000

J. S. Sarma


Deeme 198
INTRNTINA
FOOD
POIC
REEAC















The International Food Policy Research
Institute was established in 1975 to identify
and analyze alternative national and inter-
national strategies and policies for meeting
food needs in the world, with particular em-
phasis on low-income countries and on the
poorer groups in those countries. While the
research effort is geared to the precise ob-
jective of contributing to the reduction of
hunger and malnutrition, the factors involved
are many and wide-ranging, requiring analy-
sis of underlying processes and extending
beyond a narrowly defined food sector. The
Institute's research program reflects world-
wide interaction with policymakers, adminis-
trators, and others concerned with increasing
food production and with improving the
equity of its distribution. Research results
are published and distributed to officials and
others concerned with national and inter-
national food and agricultural policy.

The Institute receives support as a con-
stituent of the Consultative Group on Inter-
national Agricultural Research from a number
of donors including the United States, the
World Bank, Japan, Canada, the United
Kingdom, Australia, the Ford Foundation,
Norway, the Federal Republic of Germany,
Italy, the Netherlands, India, Switzerland,
the Philippines, the People's Republic of
China, and the International Development
Research Centre (Canada). In addition, a
number of other governments and institu-
tions contribute funding to special research
projects.


Board of Trustees


Dick de Zeeuw
Chairman, Netherlands
Ralph Kirby Davidson
Vice Chairman, U.S.A.
Eliseu Roberto de Andrade Alves
Brazil
Yahia Bakour
Syria
Lowell S. Hardin
U.S.A.
Ivan L. Head
Canada
Nurul Islam
Bangladesh
Anne de Lattre
France
James R. McWilliam
Australia
Philip Ndegwa
Kenya
Saburo Okita
Japan
Sukadji Ranuwihardjo
Indonesia
Samar R. Sen
India


Leopoldo Solis
Mexico
Charles Valy Tuho
Ivory Coast
John W. Mellor, Director
Ex Officio, U.S.A.



















CEREAL FEED USE IN THE
THIRD WORLD: PAST TRENDS
AND PROJECTIONS TO 2000

J. S. Sarma
















Research Report 57
International Food Policy Research Institute
December 1986















Copyright 1986 International Food Policy
Research Institute.
All rights reserved. Sections of this report may
be reproduced without the express permission
of but with acknowledgment to the International
Food Policy Research Institute.


Library of Congress Cataloging
in Publication Data

Sarma, J. S.
Cereal feed use in the Third World.

(Research report; 57)
Bibliography: p. 67.
1. Flour and feed trade-Developing coun-
tries-Forecasting. 2. Grain trade-Developing
countries-Forecasting. 3. Animal industry-
Developing industry-Developing countries-
Forecasting. I. Title. II. Series: Research report
(International Food Policy Research Institute) ; 57.
HD9052.D44S27 1987 338.1'7325' 86-27355
ISBN 0-89629-058-1










CONTENTS



Foreword
1. Summary 9
2. Introduction 12
3. Cereal Feed Situation in 1980 15
4. Trends in Cereal Feed Use 20
5. Growth in Cereal Feed Use in Tai-
wan and the Republic of Korea 30
6. Dynamics of Demand for Food and
Feedgrains in Selected Developing
Countries 35
7. Projections of Cereal Feed Use in
Developing Countries 41
8. Conclusions and Policy Implications 47
Appendix 1: Data and Methodology 53
Appendix 2: Classification of Coun-
tries Selected for the
Study 56
Appendix 3: Supplementary Tables 60
Bibliography 67










TABLES


1. Estimated sources of livestock feed
in the world
2. Distribution of cereal feed use, by
region, 1980
3. Output of meat, milk, and eggs in
livestock output units and cereal
feed use, 1980
4. Distribution of cereal feed use by
region and commodity, 1976-80
average
5. Trends in the feed use of grains,
by region and commodity, 1966-
70 and 1976-80
6. Annual growth rates of use of brans
as feed and output of cereals, by
region, 1966-70 and 1976-80
7. Regional distribution and trends
of output of livestock products,
1966-70 and 1976-80, and distri-
bution of population, 1980
8. Cereal feed use, classified by level
and growth of per capital income
and growth of livestock output,
1966-70 and 1976-80
9. Rate of growth of use of cereals
for food and feed, by region, 1966-
70 and 1976-80
10. Exports, imports, and net trade in
coarse grains, 1966-70 and 1976-
80 averages
11. Ratios of prices received by farmers
for poultry and pigmeat to maize,
selected countries, 1967-80
12. Average production and trade of
cereals and livestock products,
feed use, per capital income, and
population, Taiwan, selected years
13. Per capital consumption per year,
Taiwan, selected years


14. Average production and trade of
14 cereals and livestock products,
Republic of Korea, selected years
17 15. Per capital consumption per year,
Republic of Korea, various years
16. Three scenarios for income elas-
18 ticities of demand
17. Projected growth in food and feed
use of cereals, 1980-2000 and
22 average income elasticity of de-
mand for food and feed, selected
countries
22 18. Cereal feed use projections under
alternative assumptions, by region,
1990 and 2000
23 19. Projected cereal feed use in 75
selected countries, by per capital
income groups, 1990 and 2000
20. Classification of developing coun-
23 tries by per capital income, income
growth, and livestock output
growth
21. World cereal feed use and total
26 output of cereals by type, 1980
22. Cereal feed use by type, 1980
27 23. World livestock numbers, 1980
24. Trends in cereal feed use, by re-
gion, 1966-70 and 1976-80
27
2 25. Cereal feed use of grains and brans
and by-products, by region and
commodity, 1966-70 and 1976-80
29
26. Cereal feed use and total livestock
output, by region and subregion,
1966-70 and 1976-80

32 27. Production of compound feed, se-
lected developing countries, 1974-
32 76 average and 1980










ILLUSTRATIONS


28. Income elasticity of demand for
cereals and livestock products, se-
lected countries, 1980 and 2000
29. Ratio of foodgrains to food and
feedgrains, selected countries,
1980 to 2000
30. Selected developing countries with
cereal feed use exceeding 500,000
metric tons, 1980
31. Cereal feed use projections under
method 1, 1990 and 2000


1. World distribution of cereal feed
use by crop, developed and develop-
64 ing countries, 1980
2. Trends in feed use of cereals in the
study countries, by crop and region,
65 1966-70 and 1976-80
3. Growth rates of cereal feed use and
livestock products output, by region,
65 1966-70 and 1976-80
4. Trend cereal feed use in 1980 and
66 projections to 1990 and 2000, by
region, based on projected livestock
output










FOREWORD


Use of cereals for livestock feed is grow-
ing rapidly in developing countries and is
reaching significant proportions. This fact
has profound implications for food security
of low-income people, employment, and the
size of markets available to exporters of ce-
reals. However, developing countries vary
greatly in their pace of growth in the use
of cereals for feed, even within similar in-
come groups. J. S. Sarma analyzes not only
the underlying causes of the rapid growth
in cereal use for feed, but also the extent
of participation among countries and some
of the reasons for the variation. From this
analysis comes an understanding of the
forces driving future usage and a basis for
understanding and modifying the projection
of future feed use, which he provides.
Sarma's work continues a stream of
IFPRI research concerned with the analysis
and projection of global forces of supply and
demand for food. It is a sequel to Research
Report 49, Livestock Products in the Third
World: Past Trends and Projections to 1990
and 2000, which provides the background
on livestock supply and demand in develop-
ing countries essential to understanding the
complexities of feed utilization. Food in the
Third World: Past Trends and Projections
to 2000, Research Report 52, also recently
published, analyzes past trends and projec-
tions of the supply and demand for basic
food staples. Because of the rapidly growing
importance of the use of basic food staples
as livestock feed, and much greater uncer-
tainty about the key variables determining
feed supply and demand, Sarma's in-depth
analysis is of particular importance.
An intimate knowledge of basic data
sources and their problems is essential in
dealing with the particularly imprecise data
on livestock and feed sectors in developing
countries. Sarma brings to this task an im-
portant set of skills. He was formerly chief
executive officer of the National Sample
Survey Organization in India and, prior to


that, head of the Directorate of Economics
and Statistics, Ministry of Agriculture, Gov-
ernment of India. He has also served as a
member of the FAO Statistics Advisory
Committee.
Particular note should be taken of the
period 1966-80, which is used as the data
base for growth rate projections in this re-
port. Questions have been raised on whether
this is an appropriate base period, for, by
past standards, most of this period was one
of high growth in per capital incomes in
developing countries, which in turn led to
a rapid rise in livestock consumption. Sub-
sequently, however, because of recession
and protectionism in the Organization of
Economic Cooperation and Development
countries and mounting debt repayment
problems, economic growth in the develop-
ing countries has slowed down. On the one
hand, using data from the period of rapid
growth could lead to overestimation. On
the other hand, one may reasonably argue
that the development process in developing
countries should accelerate as trained people
and institutional capacity expand, thus lead-
ing to economic recovery and better eco-
nomic performance in the coming years.
Thus projections based on the 1966-80
trends would be more likely to reflect the
effects of overcoming current structural
problems and returning to the underlying
process of growth.
Finally, the importance of livestock feed
issues to traditional global concerns about
food security and employment have not
been generally recognized. We at IFPRI
hope that Sarma's seminal piece will stimu-
late a large amount of additional analysis
and hence a rapidly improving basis for
policy in this crucial area.

John W. Mellor
Washington, D.C.
December 1986










ACKNOWLEDGMENTS



The author gratefully acknowledges the
advice and encouragement received from
John W. Mellor and the valuable discussions
he had with him on the subject of this study.
In particular, he is indebted to him for his
views, which contributed greatly to the is-
sues discussed in Chapter 6 of the report.
The author is also grateful to Leonardo
Paulino, who made useful suggestions on
an earlier draft, and to Peter Oram and
Romeo Bautista, whose comments greatly
benefited this report. The help received
from Ulrich Koester, Patrick Yeung, Ashok
Parikh, and Rafael Celis on Chapter 6 is also
gratefully acknowledged.
The author wishes to thank Darunee
Kunchai for her competent statistical assis-
tance in the analysis of the data presented
in the report.









1


SUMMARY


Use of cereals for livestock feed in de-
veloping countries has increased much faster
than direct food use during the last two
decades. Cereal feed also grew faster than
production and consumption of livestock
products, such as meat, milk, and eggs; and
projections to 1990 and 2000 indicate that
with rising per capital incomes, these trends
are likely to intensify. Many of the devel-
oping countries are meeting the increased
demand for livestock products through ac-
celerated domestic production. Much of this
increase in production has occurred outside
the traditional livestock sector, which in
turn has resulted in a rapid rise in the de-
rived demand for cereal feed.
Moreover, livestock (including poultry)
production is an important source of rural
income and employment in the Third
World. When it is based on labor-intensive
technology, its development helps meet
equity objectives by contributing cash in-
come to small farmers and improving their
nutrition. Because an assured supply of feed
is an essential prerequisite for this develop-
ment, the use of cereals for feed will prob-
ably continue to grow rapidly.
In 1980, about 680 million metric tons
of cereals, including by-products, were used
as livestock feed in the world. The de-
veloped countries used about three-fourths
and produced nearly two-thirds of the global
livestock output in that year. The remaining
one-fourth was fed to livestock in develop-
ing countries (including China). Among the
cereals used as feed, three-fourths of the
world total was coarse grains, especially
maize and barley. About 82 percent of
coarse grains produced in developed coun-
tries was fed to livestock, whereas only
about 46 percent went to feedgrains in de-
veloping countries.
In 1980, Third World countries (exclud-
ing China) used about 100 million metric
tons of cereals as livestock feed, of which


less than two-thirds was coarse grains and
the balance was divided between paddy and
wheat. About 85 percent of paddy and 77
percent of wheat were in the form of by-
products, mainly bran. For cereal feeds as
a whole, by-products constituted a little
more than one-third of the total.
Among the developing regions, Latin
America used the most cereals for feed-
about 43 million metric tons in 1980, of
which 86 percent was coarse grains. Asia
accounted for 30 percent, nearly half of
which was paddy and rice by-products. Sub-
Saharan Africa used the least, only 4 percent
of the Third World total.
Between 1966-70 and 1976-80, the use
of cereal grains and by-products as feed in
the countries studied increased by 57 per-
cent, an average growth rate of 4.6 percent
a year. During the same period output of
meat, milk, and eggs increased at an average
annual rate of 3.4 percent. The rate of
growth in feed use outstripped growth of
livestock output in all regions.
The increase in cereal feed use was par-
ticularly rapid in Taiwan and the Republic
of Korea, both newly industrializing econ-
omies, where income was growing rapidly
and per capital incomes were moderately
high. In these economies, where consump-
tion of livestock products increased rapidly,
domestic production based on imported feed-
grains rose to meet the demand. The amount
of feed used per unit of livestock output
rose continuously in these two economies,
largely because more pigs and poultry were
raised, requiring intensive grain feeding.
The growth in demand for livestock prod-
ucts and for feedgrains in these economies
is indicative of what is likely to happen in
Third World countries that achieve rapid
economic growth. In many of these coun-
tries, however, the increased demand for
feedgrains might be met by increased do-
mestic production.









For the 104 countries studied, imports
of coarse grains increased from an average
of 5 million metric tons during 1966-70 to
about 16.5 million metric tons during 1976-
80. The Food and Agriculture Organization
of the United Nations estimates that more
than three-quarters of the coarse grain im-
ported by developing countries in 1981
were used for feed. Taking into account ex-
ports of coarse grains, the developing coun-
tries have gone from being net exporters of
about 6 million metric tons of grain in the
late 1960s to net importers of 3 million'
metric tons in the late 1970s.
As per capital incomes increase, the in-
come elasticity of demand for cereals for
direct consumption as food declines. The
much higher income elasticity of demand
for feedgrains may also decline but less
rapidly. In some developing countries, such
as Brazil, the amount of cereals used for
feed already exceeds that used for food. As
this trend becomes more widespread, the
aggregate demand for cereals in developing
countries will continue to increase, the de-
cline in consumption as food being more
than compensated for by the rise in feed use.
Alternative projections of the demand
for cereal feed to 1990 and 2000 in the
Third World in this study are based on past
trends in cereal feed use and in livestock
output. Five alternative methods result in
projections ranging from 151 to 163 million
metric tons of cereals used for feed in 1990
and from 245 to 286 million metric tons in
2000. The demand for cereal feed is pro-
jected to grow at rates varying from 4.7 to
5.5 percent per year to sustain an average
growth in livestock products projected for
1980-2000 at about 3.7 percent.
Projections of the International Food
Policy Research Institute (IFPRI) of the food
deficit in the Third World, based on the
continuation of past trends in production
and per capital income, show that the de-
veloping countries (excluding China) would
have an overall net food deficit of 76 million
metric tons of basic food staples, including
cereals and other major food crops, by the
end of the century. This estimate is based
on about 245 million metric tons of cereals
used for feed in 2000, which is the lower


level of demand projected in this report. If
the higher level of demand should material-
ize, feed use could exceed that level by
about 40 million metric tons, pushing the
total food deficit to about 116 million metric
tons, other assumptions remaining the same.
A recently completed IFPRI analysis of
livestock products in the Third World also
shows that if past income trends continue,
trend projections of meat and milk output
could fall considerably short of the projected
demand at constant relative prices by the
end of the century. The domestic output of
meat and milk would have to increase 50
percent faster than past output trends to
avoid significant pressure on prices or in-
creases in imports. Under the alternative
assumption of slower income growth, which
is 25 percent less than trend growth, live-
stock production would still have to expand
at rates higher than the 1961-77 trends to
meet the projected demand. Thus, the de-
rived demand for cereal feed could be even
larger than projections based on trend esti-
mates of output.
Because value added in the production,
processing, and manufacture of meat, eggs,
and dairy products is high, many of the de-
veloping countries may attempt to accelerate
domestic production by introducing trade
policies that favor imports of feedgrains
rather than of meat and dairy products and
by encouraging research and investment in
intensive livestock production. Taste pref-
erences for fresh meat, eggs, and milk over
frozen or packaged meat, egg powder, or
dry milk powder may reinforce such a pol-
icy. With economic growth and improved
provision of refrigerated storage, transport,
and marketing facilities, constraints on mar-
keting and processing of perishable products
may be eased, encouraging more rapid ex-
pansion in production and consumption of
livestock products in the coming years. And
because assured supply of livestock feed is
an essential prerequisite for the development
of livestock output, the use of cereals for
feed would continue to grow rapidly.
Where gaps between the projected de-
mand and supply of cereals used for feed
cannot be met by imports, market forces
could lead to higher meat and cereal prices,









which might worsen the food situation for
the poor. Therefore, policies concerning ce-
reals and livestock need to be reviewed and
modified. For example, in countries with a
surplus of rural labor a conscious decision
should be made to adopt labor-intensive live-
stock production strategies to meet equity
objectives: increased employment, improved
nutrition, and food security for low-income
people in developing countries. Improve-
ments in the yields of feedgrains and in feed
efficiency through emphasis on research
and development of feed technology could
also partly alleviate the situation. In addi-
tion, the scope for substituting noncereals,
such as cassava supplemented with pro-
teins, for cereals in compound and mixed
feeds needs to be explored. Increased pro-
duction of fodder also should be considered.


However, as growing demand for feeds and
fodder increase competition for land, it will
have to be used more intensively. And large
allocations of research resources will have
to be made to development of new sources
of feed, greater use of by-products and agri-
cultural wastes, and of compound and mixed
feeds.
A basic problem with the analysis of feed
trends is the lack of reliable data on their
use by type of feed and by category of live-
stock output. Developing countries should
take steps for collection and publication of
data on livestock feed and fodder, including
data on compound and mixed feeds collected
from existing feed manufacturing industries
and data on cereals used directly as feed out
of farmers' home-grown or purchased grains.









2


INTRODUCTION


A recent analysis of past trends in live-
stock products' in the developing countries
shows that if these trends continue into the
future, domestic demand for meat and milk
could grow faster than indigenous produc-
tion, leading to the emergence of large gaps
between the two by the end of the century.2
If the rapid income growth attained during
1966-77 continues, the consumption of
meat may increase 5.0 percent a year,
whereas milk consumption may rise 3.9 per-
cent a year during the 1977-2000 period.
Production is projected to grow by 3.4 per-
cent a year for meat and 2.9 percent a year
for milk during the same period. As a group,
the 104 Third World countries included in
the study, which had a surplus of 300,000
metric tons of meat in the mid-1970s, have
already become net importers: by 1981
their net imports had risen to more than
1 million tons.3 These countries imported
nearly 9 million tons, net of exports, of milk
and milk products in whole milk equiva-
lents in the mid-1970s, and their net im-
ports rose to about 16 million tons by 1981.4
The deficits in these products could widen
further by 2000, even if the rate of growth
of output achieved during 1961-77 con-
tinued into the future. For eggs, the pro-
jected supply and demand would nearly
balance, provided production by 2000 is
three to four times the level attained in the
mid-1970s. Even for meat and milk the pro-
jected outputs depend upon a twofold ex-
pansion of the levels of the mid-1970s. Even
if per capital incomes increase less rapidly,
that is, a quarter less than the 1966-77 rates,


the projected shortfalls of meat and milk
would still be large compared to the past,
unless the output growth rate is accelerated.
Livestock is an important source of in-
come and employment in the rural economy
of Third World countries. It supplies essen-
tial inputs to crop production through the
provision of draft power and manure for
improving soil fertility. Development of live-
stock production can also help to meet the
equity objective of rural development through
the sector's contribution to cash incomes
of small farmers, particularly in areas where
a labor-intensive rural livestock develop-
ment strategy is adopted. Increased income
and employment are also realized through
the value-adding industries associated with
these products. Additional employment is
generated by slaughterhouses, transport ser-
vices, and the marketing of livestock prod-
ucts and feeds. Because livestock products
contain high quality protein, their increased
consumption could lead to improved nutri-
tion in both urban and rural areas where
protein intakes are particularly low. The
livestock sector also contributes to foreign
exchange earnings, especially in Latin
America. For all of these reasons developing
countries need to pay increasing attention
to the development of the livestock sector
in the coming decades. And, to sustain rapid
growth in livestock production, large in-
creases in livestock feed supplies will be
essential.
Livestock feed is the most important in-
put in livestock industry, no matter in what
size farm or enterprise it is organized. Based


SHereafter in this report the term "livestock" includes poultry.
2 J. S. Sarma and Patrick Yeung, Livestock Products in the Third World: Past Trends and Projections to 1990 and
2000, Research Report 49. Washington, D.C.: International Food Policy Research Institute, 1985.
3 This study includes data for 104 countries referred to in this report as "the Third World countries." China and
24 very small developing countries are excluded. The term "developing countries," however, includes all of the
developing economies.
All tons referred to in this report are metric tons.
4 Imports of milk include those under food aid.









on the source of origin and end use, live-
stock feed may be classified into feedgrains,
other staple crop feeds, high protein meals,
agriculture and household by-products, and
forage and fodder crops. Essentially, animals
convert plant materials into edible products.
Feeds provide the metabolic energy (in
calories), proteins, and other nutrients, and
these vary from feed to feed. Feed require-
ments can be differentiated according to the
quantities needed for maintenance of the
animals and those required for conversion
into products. The standard requirements
of different types of feed for different species
of animals are usually drawn up in feed
schedules designed to suit different environ-
mental conditions for given levels of produc-
tion. Such schedules, which often take into
account the costs of feed and of output, are
generally adopted by commercial or semi-
commercial farmers or government re-
search stations and farms. But a majority of
the rural smallholders and traditional
graziers feed their livestock with locally
available materials, including farm by-prod-
ucts, supplemented occasionally with con-
centrate feeds. These unbalanced and often
inadequate diets, the traditional breeds of
animals raised, and the prevalence of dis-
eases explain the low productivity in output
per animal on small farms in many develop-
ing countries.
Livestock feed can also be classified into
the following groups: cereals; pulses; roots
and tubers; oil seeds, cakes, and meals, and
other protein meals including fish meal;
molasses and other crop residues and agri-
cultural and household by-products; culti-
vated fodder crops and grasses; pastures and
other grasses; and agricultural wastes. The
cereal feedgrains include coarse grains (such
as maize, sorghum, millet, barley, oats, rye,
and other cereals not elsewhere specified),
wheat, and paddy (rice in husk), and their
products and by-products. Cereal products


are processed grains such as milled and bro-
ken rice and flours, and by-products are
brans and cakes.
Feedgrains and roots and tubers provide
a large share of the energy requirements of
poultry and intensively fattened animals.
Large ruminants such as cattle and small
ruminants such as sheep and goats depend
on forages and grass supplemented by feed-
grains and protein meals for the bulk of
their energy. Agricultural by-products such
as straw and fodder stalks, which can be
stored for the dry season, are also an impor-
tant source of animal feed, particularly in
developing countries. Feedgrains form 3.3
percent of the total estimated sources of
livestock feed by weight and 16.6 percent
in terms of energy in the world (Table 1).
The major difficulty in the analysis of
past trends of livestock feeds is the nonavail-
ability of reliable time-series data on their
production and use in a number of develop-
ing countries. In many countries, separate
annual data on grains used as feed are not
systematically collected. Often such data are
derived as a fixed percentage of production
or supply, and hence the annual variations
in feed use based on the published data for
some countries do not reflect the true vari-
ations.5 In countries where the commercial
livestock sector is important, data on feed
utilization are obtained and used to arrive
at the total feed use for the year. The data
used in this study are based on information
obtained from the Food and Agriculture Or-
ganization of the United Nations.6 (See Ap-
pendix 1 for additional discussion of the
data and methods used.)
Another major difficulty is that informa-
tion on feed use by species of livestock or
type of livestock product is often not avail-
able. Apart from meat, milk, and eggs, draft
power is one of the important uses of live-
stock in developing countries, but separate
data on feed use by draft animals are not


5 FAO's food balance sheets for 1975-77 specifically caution, "Since the quantities for feed and waste shown in
the food balance sheets are often estimated from various sources, the percentage figures should be seen in
conjunction with the balancing equation for the commodity in question in the food balance sheets and should
not be judged on its own" (Food and Agriculture Organization of the United Nations, Food Balance Sheets,
1975-77, Average and Per Caput Food Supplies [Rome: FAO, 1980]).
6 Food and Agriculture Organization of the United Nations, "Agricultural Supply/Utilization Accounts Tape, 1981"
Rome, 1982.









Table 1-Estimated sources of
livestock feed in the world

Feed In Tonnage' In Energyb
(percent)
Grain 3.3 16.6
Protein meal or cakes 0.5 2.3
By-products 1.7 6.2
Forages and others 94.5 74.9
Total 100.0 100.0

Source: R. O. Wheeler, et al., The World Livestock
Produce Feedstuff and Food Grain System
(Morrilton, Arkansas: Winrock International,
1981).
a Based on projected tonnage for 1985.
b Based on energy estimates for 1977/78.




collected.7 Thus it is difficult to relate di-
rectly the feed use data with those for live-
stock output. As the grains produced within
the country and those imported are both
used as food and feed, it is difficult to build
up a time series of production of feedgrains.
Thus the trend analysis in this study is con-
fined to feed use of grains only and not to
their output. In view of the growing impor-
tance of feedgrains to sustain a rapidly grow-
ing livestock output in the coming years,
the study attempts to analyze the available
data despite these limitations.8
The analysis in this report covers the
same 104 developing countries (excluding
China) that are covered in Livestock Prod-
ucts in the Third World.9 Of these, 21 are
in Asia, 19 in North Africa/Middle East, 40
in Sub-Saharan Africa, and 24 in Latin


America. Each of these regions is subdivided
into two or three geographical subregions.
For a classification of the countries, see Ap-
pendix 2, Table 20. China is excluded be-
cause no consistent and comparable time-
series data on livestock products or feed use
for mainland China, based on government
series, are available.
The analysis of past trends is generally
based on the averages for 1966-70 (the late
1960s) and 1976-80 (the late 1970s).10
Chapter 3 discusses the current situa-
tion for the use of cereal feed in 1980 and
gives the corresponding figures on livestock
numbers and output in 1980 to serve as
background. Chapter 4 considers trends in
aggregate cereal feed use and livestock out-
put between the late 1960s and 1970s by
regions and subregions and other typologies
based on per capital income levels, per capital
income growth, and growth in livestock out-
put in the different countries. Chapter 5
examines developments in Taiwan and the
Republic of Korea, where feed use has ex-
panded rapidly in the last two decades.
As incomes rise beyond certain levels,
per capital consumption of cereals directly
as food declines and consumption of live-
stock products increases, resulting in in-
creased derived demand for feedgrains. The
dynamics of the combined use of cereals for
food and feed are examined in Chapter 6,
with illustrations from selected countries.
Chapter 7 gives the main projections for
feed use for 1990 and 2000 based on alter-
native methods and assumptions. The main
conclusions of the study and their policy
implications are given in Chapter 8.


7 Moreover, in some developing countries a clear distinction is not maintained between animals used for draft
power or for milk and those slaughtered for meat.
8 This study is part of a series on past trends and projections prepared by the International Food Policy Research
Institute, and its findings are derived in part from two earlier studies. See Leonardo Paulino, Food in the Third
World: Past Trends and Projections to 2000, Research Report 52 (Washington, D.C.: International Food Policy
Research Institute, 1986); and Sarma and Yeung, Livestock Products in the Third World
9 FAO publications do not give separate data for Taiwan but include them under China. Thus Taiwan is also not
included in the 104 countries or in the Asia region. However, the case study on Taiwan discussed in Chapter 5
is based on data taken from Republic of China, Council for Agricultural Planning and Development, Taiwan Food
Balance Sheets, 1935-1980, (Taiwan: Republic of China, 1981). These are not strictly comparable with FAO data
and hence could not be combined with those for the 104 countries.
'o Time series data on annual domestic utilization of cereals for food, feed, seed, and other uses were available
from 1966 only.










3


CEREAL FEED SITUATION IN 1980


About 680 million tons of cereals, in-
cluding by-products, were used as livestock
feed in the world in 1980. This constituted
about 44 percent of the global cereal output
(Appendix 3, Table 21). Among the cereals,
coarse grains were the most important feed-
grains, accounting for about three-fourths
of the total. Wheat and its by- products were
nearly one-fifth, the balance comprising
paddy and its by-products (Figure 1).
Seventy percent of world coarse grain pro-
duction was used as feed, whereas only 8
percent of paddy and its by-products were
fed to livestock. Nearly one-third of the


global output of wheat (including by-prod-
ucts) was used as feed.
In the aggregate, developed countries
used three times more grain to feed their
livestock than the developing countries.
About three-fourths of total cereal feed was
used in the developed countries, which pro-
duced nearly two-thirds of the world's live-
stock output in 1980. Coarse grains-the
major feedgrains-provided more than
three-fourths of the total grains fed in the
developed countries: about 82 percent of
coarse grains produced was fed to livestock.
Although coarse grains are also the pre-


Figure 1--World distribution of cereal feed use by crop, developed and
developing countries, 1980


Feed Use
(Million Metric Tons)
600 -


500 -


400


300


200


100


coarse grams


vvneat


Other developing
countries
Study countries
Developed
countries













Paddy


Source: Food and Agriculture Organization of the United Nations, "Agricultural Supply/Utilization Accounts Tape,
1981," Rome, 1982.
Note: Other developing countries include China and 24 small countries that are excluded from the study countries.









ferred feed in developing countries, about
46 percent of their domestic production is
used as feed.
In 1980, about one-sixth of the cereals
used as feed in the world was actually by-
products, mostly bran. At 83 percent, the
proportion of by-products was the highest
for paddy. Only 5 percent of coarse grains
fed was in the form of by-products; the pro-
portion for wheat was about 37 percent.
Again there were significant differences be-
tween the developed and developing coun-
tries in the use of by-products as feed-the
respective shares being 9 percent in the
former and 36 percent in the latter.
In the 104 study countries, about 100
million tons of cereal grains and by-products
were fed to livestock in 1980-nearly 15
percent of the world feed use and 60 percent
of their use in developing countries (includ-
ing China). Of this quantity, two-thirds was
coarse grain and the balance was divided
almost equally between wheat and paddy
(Table 2). Eighty-five percent of paddy, 77
percent of wheat, but only 10 percent of
coarse grains were fed in the form of by-
products. The share of by-products of wheat,
paddy, and coarse grains in total feed use
was a little more than one-third in the Third
World countries in 1980. Feed use of cereals
by type of product or by-product in 1980 is
given in Appendix 3, Table 22.
As one would expect, Latin America,
with its predominant share in the output of
meat, milk, and eggs in the Third World,
leads in the feed use of grains. All but one
country (Haiti) in this region are in the mid-
dle- or high-income range. Meat is a staple
food in some of these countries, especially
Argentina and Uruguay. In 1980, about 43
million tons of cereals were fed to animals
in this region. The share of coarse grains
was nearly 86 percent. Asia followed with
30 million tons consumed as feed, but
nearly half of this was in the form of paddy,
rice bran, and other rice products and by-
products. This is understandable because
Asia produced 88 percent of the paddy in
the study countries. Sub-Saharan Africa,
which produced the smallest quantities of
meat, milk, and eggs, also used the smallest
amount of feedgrains, mainly coarse grains.


More than two-thirds of the livestock feed
in North Africa/Middle East was also in the
form of coarse grains, particularly barley
(9.2 million tons). About 14 million tons
each of rice and wheat bran and other by-
products were fed to animals in the four
regions, taken together. Of the 65 million
tons of coarse grains used as feed in the
study countries, a little less than 57 percent
was in Latin America, followed by about 24
percent in North Africa/Middle East.
Of the coarse grains, 39.1 million tons
were maize, 10.4 million tons were barley,
11.7 million tons were sorghum, and the
rest were other grains such as millets, oats,
and rye. Sorghum was mainly used as feed
in Latin America. In the case of wheat and
its by-products, North Africa/Middle East
is reported to have used about 40 percent
of the Third World total for feeding its live-
stock. Two-thirds of this was wheat bran.
About half a million tons of wheat bran were
used as feed in Sub-Saharan Africa.
More than half of the cereal feed used
in Latin America in 1980 was in the Upper
South America subregion. In North Africa/
Middle East, 70 percent was used in West-
ern Asia; in Sub-Saharan Africa, nearly half
was used in Eastern and Southern Africa;
and in Asia, 57 percent was used in East
and Southeast Asia.
As explained earlier, China is not covered
by the analysis of past trends and projections
because of a lack of adequate time-series
data on feed use. According to recent FAO
estimates, however, nearly 70 million tons
of cereals were fed to livestock in China,
including Taiwan, in 1980-slightly more
than 10 percent of world cereal feed use.
Coarse grains constituted 63 percent of the
feed used in China.
Because the details of cereals fed to dif-
ferent species of livestock are not available,
it is not possible to relate feed use to the
number of animals in the different regions.
However, Appendix 3, Table 23 shows that
although nearly two-thirds of the world's
large and small ruminants and 50-60 per-
cent of pigs and poultry are in the develop-
ing economies (including China), they use
only about a quarter of the cereal feed. This
indicates a larger use of pastures and grazing








Table 2-Distribution of cereal feed use, by region, 1980
Share of
Coarse By-Products
Region Grains Paddy Wheat Total (IncludingBran)
(million metric tons)
Asia 9.4 15.0 5.7 30.2 20.6
North Africa/Middle East 15.5 0.3 7.2 23.0 5.4
Sub-Saharan Africa 3.2 0.3 0.5 4.0 2.5
Latin America 37.0 1.5 4.7 43.2 6.1
Total 65.1 17.1 18.1 100.3 34.6

Source: Food and Agriculture Organization of the United Nations, "Agricultural Supply/Utilization Accounts Tape,
1981," Rome, 1982.
Note: Parts may not add to total due to rounding.


lands as well as low-cost agricultural wastes
and by-products for direct consumption as
feeds in these countries." Although at the
regional level the proportion of ruminants,
large and small, was higher in Asia than in
Latin America, the proportion of feed use
of cereals in the latter was higher, perhaps
because Latin America has the largest num-
ber of pigs and poultry.
The developed economies, with almost
one-quarter of the world population, pro-
duced nearly two-thirds of the meat, three-
fourths of the milk, and three-fifths of the
eggs in 1980, which shows the wide dispar-
ity in per capital production between the
developed and developing economies. The
104 countries covered by this study shared
roughly one-fifth of each of these products
as compared to their share of 50 percent of
global population in the same year. Among
the regions, Latin America produced nearly
half of the meat, one-third of the milk, and
nearly two-fifths of the eggs in the Third
World. Asia led in the production of milk
with about 40 percent of the aggregate for
the study countries. The output of all three
products was extremely low in Sub-Saharan
Africa, much lower than its 15 percent share
of the population (see Table 3).
Table 3 also gives quantities of cereals


fed to livestock in 1980. Because the
weights of the three products are expressed
in different terms, it would not be appropri-
ate to add the output of each of the three
products, ton for ton, and compare the total
with the livestock feed use given in the
table. Each of the outputs needs to be con-
verted to a common denominator before
they are aggregated to obtain the feeding
ratio. Milk is the main problem because
fresh milk contains 87-88 percent water.
Often international trading is done in dry
milk powder or in condensed and evapo-
rated milk, which has a much lower percen-
tage of water. One method of aggregation
is to reduce milk to its dry weight by divid-
ing by 10 (an average figure). The results
are given in brackets below the fresh milk
figures. When the weights of meat and eggs
are added, the total is expressed as livestock
output units (LOUs). In this method no cor-
rection factor is applied for meat or eggs,
which also contain water.
Another method of aggregating the out-
puts would be to combine them on an en-
ergy basis because cereal feeds primarily
provide energy to the animals. The energy
content of various types of livestock prod-
ucts is expressed in physiological fuel values
(PFVs).12 The feeding ratio would represent


'' Agricultural wastes and by-products are also used in the developed countries but mostly as ingredients of
compound feeds.
12 The individual PFVs expressed in 1,000 calories per kilogram of fresh product weight are: buffalo and cattle
meat (carcass weight), 2.31; sheep and goat meat (carcass weight), 2.00; pigmeat, 4.20; poultry meat, 1.40;
cow's milk, 0.62; buffalo milk, 1.00; sheep milk, 1.12; goat milk, 0.75; eggs, 1.50 (Winrock International, Role
of Ruminants in Support of Man [Morrilton, Arkansas: Winrock International, 1981)).









Table 3--Output of meat, milk, and eggs in livestock output units and cereal
feed use, 1980

Output
Livestock Cereal Feeding
Country Group/Region Meat Milk" Eggs Output Units Feed Use Ratio
(million metric tons)
World 137.8 470.0 28.7 213.5 681.4 3.2
(47.0)
Developed economies 88.0 362.3 18.0 142.2 506.9 3.6
(36.2)
Developing economies
(includingChina) 49.8 107.8 10.8 71.4 174.6 2.4
(10.8)
104 study countries
(excludingChina)d 27.0 99.3 5.3 42.2 100.3 2.4
(9.9)
Asia 5.8 43.8 1.6 11.8 30.2 2.6
(4.4)
North Africa/Middle East 3.5 16.4 0.8 6.0 23.0 3.8
(1.6)
Sub-Saharan Africa 3.5 5.9 0.5 4.5 4.0 0.9
(0.6)
Latin America 14.2 33.3 2.3 19.9 43.2 2.2
(3.3)

Sources: All figures are derived from data of the Food and Agriculture Organization of the United Nations (FAO).
World data are from FAO, Production Yearbook 1982 (Rome: FAO, 1983); country data are from FAO,
"Production Yearbook Tape, 1981," Rome, 1982; and the cereal feed use data are from FAO, "Agricultural
Supply/Utilization Accounts Tape, 1981," Rome, 1982.
Notes: In this table, the output of meat is expressed in carcass weight, milk in whole fresh milk equivalents,
and eggs in fresh weight. Livestock output units are weights of meat, milk, and eggs converted to a
common denominator to obtain a total for livestock products. Milk is reduced to its dry weight by
dividing by 10; then weights of meat and eggs are added. Meat includes beef, veal, buffalo meat, mutton
and goat meat, pig and poultry meat; and milk includes cow, buffalo, sheep, and goat milk; eggs include
hen eggs and other eggs. Parts may not add to total due to rounding.
a The figures in parentheses refer to dry milk equivalents.
b Feeding ratio is the average quantity of cereals and by-products used to produce one unit of livestock products.
c These countries are grouped according to the classification system of the Food and Agriculture Organization of
the United Nations.
d These countries are grouped according to the classification system of the International Food Policy Research
Institute.


the cereal feed use per unit of PFV.
Whereas there would be no difficulty in
using PFVs by type of meat and milk in the
analysis of past trends of specific com-
modities, using individual energy values to
obtain the aggregate of livestock products
is difficult, particularly when output projec-
tions for the future are based on the aggre-
gates for meat and milk, and not by the
types of meat and milk. For cereal feeds,
although the energy content of each cereal
grain is different, the quantities of all the


grains are aggregated ton for ton. It is also
possible to convert the cereals into energy
equivalents and to work out a ratio of the
energy equivalents of cereal feeds and live-
stock outputs, but as a feeding ratio this
coefficient has limited use.
Another method takes into considera-
tion the input/output ratios of each of the
livestock products. But these vary consider-
ably from country to country and over time
for the same species of livestock depending
upon the composition of the feed, techniques








of production, and so forth. In general the
input/output ratio for milk varies from 0.3
to 0.4:1. For meat, it varies widely between
types. For example, the ratio is 2 to 3:1 for
intensively fed chicken meat, 4 to 6:1 for
intensively fed pigmeat; and even higher
for intensively fed beef. For traditionally
raised, home-produced livestock products
the ratios are much lower. On average, the
feeding ratio may be taken to be around 3
to 4:1 for meat and eggs. In other words,
the grain feed required to produce one unit
of fresh milk is about one-tenth of that re-
quired for meat or eggs. This also suggests
that dividing the milk output by 10 for con-
verting into LOUs is reasonable, though this
understates the importance of milk from a
nutritive point of view, but not from a feed
use point of view.
Although this method of dividing milk
output by 10 and adding it to the output of
meat and eggs is somewhat rough, it seems
to be a better indicator of total livestock
production than the one obtained by adding
meat, milk, and eggs ton for ton for the
limited purpose of working out the feeding
ratios. Thus, the output of meat, milk, and
eggs is converted into LOUs in Table 3.
The total cereal feed use in the different
regions still cannot be compared with the
total output of livestock products because
the feed data include quantities fed to draft
animals and those fed to other animals such
as young calves and cows not in milk, for


which no separate data are available. In in-
terpreting the data on cereal feed use and
livestock output in the different regions in
Table 3, these limitations must be kept in
mind. It will be seen that the average ratio
of feed used for one unit of livestock out-
put-the feeding ratio-comes to 2.4 in
1980 for the study countries as a whole.
Despite the limitations of the data, they
show that the feeding ratio in developing
countries is about two-thirds of that for de-
veloped countries. The lower ratio in de-
veloping countries suggests that they rely
more on nongrain feeds, such as pasture,
straw, and agricultural wastes for feeding
cattle, which, though cheaper, reduce the
output per animal. Among the study coun-
tries, the ratio is high in North Africa/Mid-
dle East and low in Sub-Saharan Africa. The
composition of meat by type and the live-
stock production technology and feeding
practices adopted also partly explain the re-
gional differences in the ratios.13
To the extent that improved livestock
production technologies are more feed-
intensive, these technologies increase the
demand for feedgrains. Thus, in areas where
poultry, pig, and dairy production have ex-
panded through commercial or semicom-
mercial enterprises, the manufactured feed
industry has also developed, thereby in-
creasing the demand for cereal feeds. This
increased demand has been met by imports
where domestic production is inadequate.


13 A disaggregated analysis of meats by type would have thrown more light on the variation in the ratios, but
data on feed use by type of meat are not available.








4


TRENDS IN CEREAL FEED USE


Use of cereals for feed in the Third
World countries increased faster than the
output of meat and milk. The quantity of
cereal grains and by-products fed to live-
stock rose from 57 million tons a year in
1966-70 to 90 million tons in 1976-80, an
average rate of 4.6 percent a year (see Ap-
pendix 3, Table 24 and Figure 2).
Among the developing regions, Latin
America is a major livestock producer, and
this is reflected in the amount of cereals
used for feed. In the late 1970s, this region
held a 42 percent share of total feed use in
the Third World, and it also accounted for
44 percent of the increase between the two
periods (Table 4). The 5 percent-a-year rate
of growth in feed use in this region is thus
a little higher than the average growth rate
for the 104 study countries taken together.
In North Africa/Middle East, feed use in-
creased from 12.5 million tons in the late
1960s to 20.5 million tons in the late 1970s.
At 3.1 percent a year, growth of feed use
in Sub-Saharan Africa was the lowest among
the four developing regions; this seems con-
sistent with the relatively slow growth rate
for output of livestock products in this re-
gion. In Asia, which used about 31 percent
of total feed, growth in the use of cereal for
feed was less than the average for Third
World countries, largely because of the
slower growth in South Asia and in the use
of by-products such as rice bran for feed.
Reaching 56.6 million tons in the late
1970s, feed use of coarse grains increased
at a rate slightly faster than that of total
cereal feeds between 1966-70 and 1976-
80. The growth rates were even higher in
Asia and Latin America. Quantities of wheat
and paddy used as feed were nearly the
same at 17 million tons each in the late
1970s, but the growth in the use of wheat
at 5.5 percent was much higher than the
3.0 percent in paddy. Seventy-five percent
of the wheat fed was in the form of bran.


Consumption of wheat bran increased from
7.5 million tons in 1966-70 to 12.5 million
tons in 1976-80, an average annual increase
of 5.3 percent. Nearly 90 percent of the
paddy fed was in Asia. Again, 85 percent
of total paddy feed was in the form of rice
bran and cakes in 1976-80, and use of these
by-products rose more slowly at 2.8 percent
a year. In the case of coarse grains, bran
and cakes formed only about 10.5 percent
of feed use in the late 1970s, which is less
than their 13.6 percent share in the late
1960s.
Among the coarse grains, maize was the
preferred feedgrain in Latin America and
Asia. In North Africa/Middle East, barley
was more important than maize, although
maize's share was increasing. In Latin
America, maize accounted for three-fourths
of the coarse grain fed in the late 1960s.
This share declined to about two-thirds in
the late 1970s, when about 21.3 million
tons of maize were used as feed. The use
of sorghum for feed in Latin America in-
creased relative to maize. In Asia, maize's
share of coarse grains increased from about
60 percent to nearly 75 percent, largely
from imported maize. In North Africa/Mid-
dle East in the late 1970s, 8.2 million tons
of barley and 3.5 million tons of maize were
utilized out of a total cereal feed use of 20.4
million tons. The corresponding figures for
the late 1960s were 5.3 million tons of bar-
ley and 1.4 million tons of maize out of a
total feed use of 12.5 million tons.


Use of Cereal Grains
and By-Products
The growth rates and importance of
whole grains and by-products varied be-
tween regions and commodities. In the ag-
gregate, nearly 64 percent of the cereal feed
used in the late 1970s was in the form of









Figure 2-Trends in feed use of cereals in the study countries, by crop and
region, 1966-70 and 1976-80
Million Metric Tons
100 -
90 -
80 -
70 -
60 -
50 -
40 -
30 -
20
10
0


Coarse grains


Wheat


Paddy


Total


Million Metric Tons
100 -
90 -
80 -
70 -
60 -
50 -
40 -
30 -
20 -
10 -
0
Asia


1966-70 average
1976-80 average


a North Africa/ Sub-Saharan
Middle East Africa


Latin Study countries
America


Source: Food and Agriculture Organization of the United Nations, "Agricultural Supply Utilization Accounts Tape,
1981," Rome, 1982.


grains (including grain products) and 36 per-
cent was in by-products, mainly bran and
cakes. In the late 1960s grains were 60
percent and by-products 40 percent.
Thus cereal feeds, excluding by-prod-
ucts, increased from about 34 million tons
in the late 1960s to 57 million tons in the
late 1970s, a growth rate of 5.3 percent a
year, which is higher than 4.6 percent for
total cereal feed (Table 5 and Appendix 3,
Table 25). Latin America held a 56 percent
share of this feed use in the late 1970s. Asia


exhibited the most rapid growth at 6.3 per-
cent a year. Coarse grains constituted 88
percent of the total in the late 1970s. The
use of wheat and wheat flour, however, rose
most rapidly at about 6 percent a year, while
paddy use increased by 4 percent a year.
Of the total of 33 million tons of cereal
by-products, mainly bran, used as livestock
feed, rice and wheat held shares of 43 and
38 percent, respectively, in the late 1970s.
The coarse grains were fed largely as grains,
and their by-products constituted about one-









Table 4-Distribution of cereal feed use by region and commodity, 1976-80
average

Regional Distribution Commodity Distribution
Region Coarse Grains Wheat Paddy Total Coarse Grains Wheat Paddy Total
(percent)
Asia 15 31 89 31 30 18 52 100
North Africa/Middle East 24 40 2 23 66 32 2 100
Sub-Saharan Africa 5 3 2 4 81 12 7 100
Latin America 56 27 8 42 85 12 3 100
Total 100 100 100 100 63 19 18 100

Source: Food and Agriculture Organization of the United Nations, "Agricultural Supply/Utilization Accounts Tape,
1981," Rome, 1982.
Note: Parts may not add to total due to rounding.


tenth of the total. Asia used most of the rice
bran-about 12.6 million tons out of 14.2
million tons of feed. Wheat bran feed use
was distributed more evenly among the three
developing regions of Asia, North Africa/
Middle East, and Latin America (Appendix
3, Table 25).
Because brans are by-products, growth
in their use as feeds depends in part on
growth in output of the grains from which
they are derived. For example, paddy output



Table 5-Trends in the feed use of
grains, by region and
commodity, 1966-70
and 1976-80

Region/ Grains Used for Feed
Commodity 1966-70 1976-80 GrowthRate
(million metric tons) (percent)
Asia 4.54 8.35 6.3
North Africa/
MiddleEast 9.41 15.61 5.2
Sub-Saharan
Africa 0.98 1.40 3.6
Latin America 19.29 31.99 5.2
Total 34.21 57.35 5.3
Coarse grains 30.22 50.68 5.3
Wheat 2.37 4.26 6.0
Paddy 1.62 2.41 4.0

Source: Food and Agriculture Organization of the
United Nations, "Agricultural Supply/Utiliza-
tion Accounts Tape, 1981," Rome, 1982.
Note: Parts may not add to total due to rounding.


in the developing countries increased 2.8
percent a year between the late 1960s and
1970s, which is the same as the growth
rate in the use of rice bran as feed during
the period. On the other hand, wheat bran
use increased by 5.3 percent a year, which
was higher than the 4.8 percent-a-year
growth in the output of wheat in Third
World countries. This is largely because of
the increase in net imports of wheat into
these countries, either through trade or food
aid. It is significant that rice is traded in its
milled form, whereas wheat is mostly traded
as whole grain. Thus rice bran is available
for use in the exporting countries, while
wheat bran is more likely to be used in the
importing countries. These differences are
evident in Table 6, which shows the annual
growth rates in the use of wheat and rice
brans as feed and the corresponding growth
rates in the output of wheat and paddy dur-
ing the two periods.
In Asia, however, the use of wheat bran
as feed grew more slowly than the output
of the whole grains, especially in South Asia.
Including the use of bran from coarse grains,
the use of total cereal by-products increased
3.6 percent during the period, compared to
4.6 percent for all cereal feeds. Moreover,
because growth in the use of rice bran was
constrained by growth in the output of
paddy, any further acceleration in the over-
all use of cereal feed, particularly in Asia,
would have to come from growth in feeds
other than rice bran.








Table 6-Annual growth rates of use
of brans as feed and output
of cereals, by region,
1966-70 and 1976-80

Rateof Growth
Wheat Rice
Bran Wheat Bran Paddy
Region Feed Output Feed Output
(percent/year)
Asia 5.9 6.7 2.8 2.8
North Africa/
Middle East 5.5 3.7 -0.4 0.8
Sub-Saharan
Africa 6.3 0.7 3.1 2.9
Latin America 4.1 2.8 3.9 3.7
Total 5.3 4.8 2.8 2.8

Source: Food and Agriculture Organization of the
United Nations, "Production Yearbook Tape,
1981," Rome, 1982; and Food and Agriculture
Organization of the United Nations, "Agricul-
tural Supply/Utilization Accounts Tape,
1981," Rome, 1982.


Trends in Output Growth
of Livestock Products
When the production of meat, milk, and
eggs in each country was converted into
LOUs, an average growth rate of 3.4 percent
was recorded between the periods 1966-70
and 1976-80. Meat output grew 3.1 per-
cent; milk, 3.0 percent; and eggs, 5.7 per-
cent at annual compound rates. The output
of these products increased from 28.3 mil-
lion tons in the late 1960s to 39.4 million


tons in the late 1970s (Table 7 and Appen-
dix 3, Table 26). Less than half of the output
during the latter period was from Latin
America, although its share of the popula-
tion was only one-sixth. Asia, with its 58
percent population share, produced only 28
percent of Third World output. Except in
Sub-Saharan Africa, where output growth
was about 2.2 percent, output in the other
three regions grew about 3.5 percent a year.
The annual growth rate in cereal feed
use in developing countries outstripped that
of the output of livestock products by more
than 1.2 percent a year (see Figure 3). In
Latin America, feed use growth was 1.5 per-
cent ahead of livestock output growth. The
difference between the two growth rates
was smallest in Asia, where feed use in-
creased at 4.2 percent a year. These differ-
ences are more marked at the subregional
level.


Subregional and Selected
Country Trends
Countries of the Upper South America
subregion, which is largely in the tropics,
accounted for a little more than half of the
cereal feed use in Latin America, as well as
a little less than half of the region's livestock
output in the late 1970s. The most rapid
growth in cereal feed use, however, was in
the Central American subregion, which in-
cludes Mexico and the Caribbean countries.
There livestock feed more than doubled,


Table 7-Regional distribution and trends of output of livestock products,
1966-70 and 1976-80, and distribution of population, 1980
Output in Livestock Output Units
AnnualAverage 1980
Region 1966-70 1976-80 Growth Rate Population
(million metric tons) (percent)
Asia 7.73 10.94 3.5 58.2
North Africa/Middle East 3.88 5.58 3.7 11.2
Sub-Saharan Africa 3.42 4.24 2.2 14.9
LatinAmerica 13.24 18.59 3.5 15.7
Total 28.27 39.35 3.4 100.0

Source: Food and Agriculture Organization of the United Nations, "Production Yearbook Tape, 1981," Rome,
1982; and United Nations, Department of International Economic and Social Affairs, "World Population
Prospects as Assessed in 1982," New York, 1983 (computer printout).








Figure 3-Growth rates of cereal feed use and livestock products output, by
region, 1966-70 and 1976-80


Percent/Year
6r-


Livestock output units
Cereal feed use


Asia North Africa/
Middle East


Sub-Saharan
Africa


Latin Studycountries
America


Sources: Food and Agriculture Organization of the United Nations, "Production Yearbook Tape, 1981," Rome,
1982; and Food and Agriculture Organization of the United Nations, "Agricultural Supply/Utilization
Accounts Tape, 1981," Rome, 1982.
Note: Livestock products are expressed in livestock output units derived as meat + 1/10 milk + eggs.


rising from 5.1 million tons in the late
1960s to 10.4 million tons in the late 1970s
(see Appendix 3, Table 26). In comparison,
livestock output in this subregion increased
only about 60 percent. In temperate Lower
South America, growth rates in the use of
feed and in livestock output were low during
this period-1.5 percent and 1.4 percent.
This subregion held a one-fifth share of Latin
America's feed use but nearly 30 percent
of its livestock output in the late 1970s.
Cereal feed use in Latin America was high-
est in Brazil-15.3 million tons a year on
average during 1976-80 compared to 9.6
million tons during 1966-70. In Mexico,
feed use doubled from 4.0 million to 8.1
million tons during the same period. Argen-
tina showed a slower growth rate of about
1.5 percent a year, rising from 5.4 million
tons to 6.3 million tons during this period.


Of the 28.3 million tons of feed used in
Asia in the late 1970s, South Asia accounted
for about 13.3 million tons or less than half,
whereas this subregion produced about 55
percent of the livestock output. The East
and Southeast Asia subregion, which com-
prised several newly industrializing coun-
tries with relatively rapid growth in income,
used more than half of Asia's feed. This was
more than 70 percent higher than in the
late 1960s. The growth rate of feed use was
1.3 percent higher than that of livestock
output in this subregion, and nearly double
that of South Asia. The Republic of Korea
recorded a 13.5 percent yearly increase in
feed use. It rose from about 800,000 tons
in the late 1960s to 2.8 million tons in the
late 1970s. Cereal feed use in India was
estimated at about 10 million tons in the
late 1970s, although this represented an









increase of only 2.8 percent a year com-
pared to the late 1960s. Hong Kong, In-
donesia, Malaysia, Mongolia, the Philip-
pines, and Singapore all recorded growth
rates of 5 percent or more.
In North Africa/Middle East, Turkey
used the largest amount of cereal feed-8.3
million tons in the late 1970s, mostly coarse
grains. Turkey's rate of growth in feed use,
however, was only 3.2 percent a year. The
Western Asia subregion, which included
Turkey, had a 70 percent share of the re-
gion's feed use and a 65 percent share of
livestock output. The growth rate for feed
in this subregion was two-thirds that in the
Northern Africa subregion. A number of
countries in the latter subregion exhibited
growth rates exceeding 5 percent a year,
including Algeria, Egypt, Libya, Morocco,
Sudan, and Tunisia. Feed use in Saudi
Arabia increased rapidly, by 29 percent a
year, increasing from 42,000 tons in the
late 1960s to more than half a million tons
in the late 1970s. In many of these coun-
tries, rapid growth in per capital incomes
through oil revenues resulted in larger de-
mand for livestock products, part of which
was met from increased domestic produc-
tion.
In the late 1970s, more than half of the
cereal feed use in Sub-Saharan Africa was
in the Eastern and Southern Africa subre-
gion. Central Africa's share was less than
10 percent. These shares are similar to their
shares of livestock output. The growth rate
for feed use in Eastern and Southern Africa
was about 3.0 percent a year, and that for
Central Africa was 4.3 percent. In West
Africa, feed use increased by 3.1 percent a
year. Nigeria used roughly 578,000 tons of
cereal feed in the late 1970s, which ac-
counted for about 40 percent of West Africa's
feed utilization, as compared to its share of
48 percent of livestock output.


Feed Use by Per Capita
Income, Income Growth,
and Livestock Output Growth
The utilization of cereals as livestock
feed is high in high-income countries. To


study this relationship, the countries were
classified into four groups: those with a per
capital income at 1980 prices of less than
$250, $250-499, $500-1,999, and more
than $2,000 per year in 1980. Cereal feed
use was then aggregated for each group of
countries. Feed data were also grouped by
two other criteria: per capital income growth
during 1961-80 (less than 1.0 percent, 1.0
to 2.9 percent, 3.0 to 4.9 percent, and 5.0
percent and more per year) and by rate of
growth in livestock products (less than 1.0
percent, 1.0 to 2.9 percent, 3.0 to 4.9 per-
cent, and 5.0 percent and more) between
1966-70 and 1976-80. The relevant data
are given in Table 8.
Of the 104 developing countries included
in the study, 20 high-income countries,
with average per capital incomes exceeding
$2,000 a year, used nearly 43 percent of
the feed consumed in developing countries
in the late 1970s; this is about the same as
their share of livestock output. These coun-
tries accounted for only 15 percent of the
Third World population. At the other end
of the scale, 19 very-low-income countries
accounted for about 44 percent of the total
population, but their share of feed use was
less than one-sixth, and they contributed
18 percent of the total livestock output. The
growth rate of feed use in these countries
was also low-a little more than half of the
growth rate for the developing countries as
a whole.
When the countries were grouped
according to per capital income growth,
countries with medium growth had a 36
percent share of feed use and those with
rapid growth, a 27 percent share in the late
1970s. Their respective shares of livestock
output were 27 and 21 percent. The growth
in feed use in the rapid-income growth
countries was a little less than 6 percent a
year, which is about 30 percent higher than
the overall growth rate. In 27 countries,
with one-sixth of the total population, per
capital incomes rose less than 1 percent.
These countries used only about 8 percent
of the total cereal feed and produced only
13 percent of livestock products in the late
1970s.
Finally, the countries were grouped










Table 8-Cereal feed use, classified by level and growth of per capital income
and growth of livestock output, 1966-70 and 1976-80
Growth Rate
1966-70 to
1966-70 1976-80 1976-80
Percentof Livestock Livestock Livestock
Numberof Population Output Output Output
Country Group Countries in 1980 Unit Feed Unit Peed Unit Feed
(million metric tons) (percent)
Level of per capital
income, 1980
Less than $250 19 44 5.65 11.70 7.12 14.94 2.3 2.5
$250-$499 19 16 2.99 3.32 4.08 5.24 3.2 4.7
$500-$1,999 46 25 8.07 18.38 11.70 30.86 3.8 5.3
$2,000 or more 20 15 11.57 23.82 16.45 38.99 3.6 5.1
Total 104 100 28.27 57.21 39.35 90.03 3.4 4.6
Growth rate of per capital
income, 1961-80
Less than 1 percent 27 16 3.97 5.91 5.07 7.45 2.5 2.3
1 percent-2.9 percent 36 48 11.89 18.63 15.23 25.84 2.5 3.3
3 percent-4.9 percent 23 25 7.19 19.23 10.61 32.86 4.0 5.5
5percentormore 18 11 5.22 13.44 8.45 23.88 4.9 5.9
Total 104 100 28.27 57.21 39.35 90.03 3.4 4.6
Growth rate of livestock
output, 1961-80
Less than 1 percent 13 5 2.14 1.73 2.13 1.88 ... 0.8
1 percent-2.9 percent 23 15 7.37 11.71 8.82 16.00 1.8 3.2
3 percent-4.9 percent 44 62 14.51 32.74 20.83 48.89 3.7 4.1
5percentormore 24 18 4.26 11.02 7.58 23.26 5.9 7.7
Total 104 100 28.27 57.21 39.35 90.03 3.4 4.6

Source: Food and Agriculture Organization of the United Nations, "Agricultural Supply/Utilization Accounts Tape,
1981," Rome, 1982; Food and Agriculture Organization of the United Nations, "Production Yearbook
Tape, 1981," Rome, 1982; and World Bank, "Gross National Product by Country Data Tape, 1961-80,"
Washington, D.C., 1981.
Notes: Parts may not add to total due to rounding. Livestock output unit (LOU) = meat + 1/10 milk + eggs.


according to the rates of growth of their
livestock output. Between the late 1960s
and late 1970s livestock output increased
between 3 and 5 percent a year in 44 coun-
tries with a combined population share of
62 percent. In another 24 countries with a
population share of 18 percent, the rate of
growth exceeded 5 percent per year. The
former group of countries had more than a
54 percent share of feed use, whereas the
latter group accounted for about 26 percent.
Between them their share of feed use was
a little more than 80 percent. These coun-
tries accounted for 72 percent of livestock
output and 80 percent of population. Coun-
tries where the output of livestock rose more
than 5 percent a year recorded a growth of
about 8 percent per year in feed use.


This analysis indicates that in the past
the share held by high-income countries in
total livestock output was high; conse-
quently, their share of feed use also was
high. Many of these countries, particularly
in East and Southeast Asia, had poor grazing
lands and relied more on poultry and pigs
for increased domestic meat production.
This also contributed to the rapid growth
in the use of cereal feeds.


Trends in Food and
Feed Use of Cereals
A recent IFPRI study of the utilization of
cereals for food and feed shows that between
1966-70 and 1976-80, feed use increased









more rapidly than direct consumption as
food. For the Third World countries as a
whole, livestock feed use rose 4.6 percent,
whereas food use increased at an average
rate of 3.1 percent per year (Table 9). Mean-
while, population grew 2.5 percent and live-
stock output 3.4 percent a year. In Latin
America, feed use rose almost one-and-a-
half times as fast as direct consumption of
food. In the three other regions as well, feed
use rose faster than food.
Consumption of livestock products rose
faster than that of cereal foods, and, as a
result, the derived demand for cereals for
feed increased more rapidly than direct food
demand. This is discussed in more detail in
Chapter 6.


Imports of Coarse Grains
Some developing countries met their in-
creased demand for feedgrains, largely
coarse grains, through imports. For the 104
study countries taken together cereal im-
ports increased from an average of 4.9 mil-
lion tons during 1966-70 to 16.5 million
tons during 1976-80 (Table 10). Although
part of these imports must have been for
direct food use, FAO estimates that animal
feed accounted for more than three-quarters


Table 9-Rate of growth of use of
cereals for food and feed,
by region, 1966-70 and
1976-80

Region Food Use Feed Use
(percent/year)
Asia 3.0 4.2
North Africa/Middle East 3.6 5.0
Sub-Saharan Africa 2.6 3.1
Latin America 3.3 5.0
Total 3.1 4.6

Source: Food and Agriculture Organization of the
United Nations, "Agricultural Supply/Utiliza-
tion Accounts Tape, 1981," Rome, 1982.




of developing-country imports of coarse
grains in 1981.14 Among the major net im-
porters of coarse grains for feed were the
Republic of Korea, Malaysia, Iran, Egypt,
Saudi Arabia, Cuba, Mexico, Brazil, and
Venezuela, each of which had net imports
of more than half a million tons on average
during the 1976-80 period. The 104 study
countries taken together, which had net ex-
ports of coarse grains of about 5.8 million
tons in the late 1960s, became net impor-
ters of 3.2 million tons in the late 1970s.


Table 10-Exports, imports, and net trade in coarse grains, 1966-70 and
1976-80 averages
1966-70 Average 1976-80 Average
Region Exports Imports NetTrade Exports Imports NetTrade
(million metric tons)
Asia 1.86 2.08 -0.22 2.76 4.05 -1.29
North Africa/Middle East 0.42 0.80 -0.38 0.34 3.95 -3.61
Sub-SaharanAfrica 0.85 0.71 0.14 0.52 1.69 -1.17
Latin America 7.59 1.29 6.29 9.67 6.80 2.87
Total 10.71 4.87 5.83 13.28 16.49 -3.21

Source: Food and Agriculture Organization of the United Nations, "Agricultural Supply/Utilization Accounts Tape,
1981," Rome, 1982.
Note: A minus sign indicates net imports.


14 Food and Agriculture Organization of the United Nations, Livestock Development in Developing Countries and
Implications for Consumption and Trade of Feeds (CCP: GR 82/83), 1983.









Meat/Feed Price Ratios
When world food prices are high, as in
the early 1970s, the feeding of cereals to
livestock is reduced in the developed coun-
tries, and the grains released become avail-
able for food use. When prices are low,
larger proportions of cereals, particularly
coarse grains, are diverted to feed use. Al-
though in the short run, "livestock in fact
provided a cushion that served to dampen
world market shocks by 'stretching' the
grains that were diverted from indirect to
direct use," the long-term consequences of
food-feed competition are more complex.15
For a systematic study of the relationship
between the prices of cereal feed and the
demand for it, several other factors, such as
prices of different components of feed and
of different types of livestock output, feed
efficiency, and price regulations if any on
feed ingredients and livestock products,
would have to be considered. This is beyond
the scope of the present study. However,
to give a broad picture of the price trends
in feed, annual ratios of producer prices of
poultry meat/maize and of pigmeat/maize
were worked out for the period 1967-80
for selected countries and are presented in
Table 11. This ratio indicates the units of
maize that could be purchased in exchange
for one unit of meat. Where ratios are high
or increasing, livestock production becomes
more profitable, leading to increased feed
use and vice versa.
The actual data given in the tables do
not, however, show any uniform and consis-
tent trends among the selected developing
countries, except that, as mentioned earlier,
when world food prices were high in the
years 1973-75, the meat/feed price ratios
were mostly low. In Brazil and Kenya, the
ratios were generally falling during the


1967-80 period. On the other hand, in Tur-
key, Egypt, and India the ratios were rising.
In other countries, the behavior was
mixed.


Manufactured Feed
The increased demand for livestock
products in some developing countries was
met through livestock, poultry, and dairy
enterprises organized on commercial or
semicommercial lines, while in others in-
creased production was sought through
small-scale rural enterprises, developed on
a cooperative or other basis. In either case,
these efforts were supported by feeds, either
compound or mixed, largely manufactured
in the private sector. The available data in-
dicate rapid growth in the output of these
industries. According to information col-
lected by FAO from national governments
and feed and feed-ingredient associations,
the production of compound feeds in de-
veloping countries (excluding China) dou-
bled over a period of five years from an
average of about 20 million tons during
1974-76 to 40 million tons in 1980. Latin
America produced more than half of the
1980 output, 21.6 million tons. Brazil alone
produced 11.4 million tons of compound
feed in 1980. Available data for selected
developing countries for the years 1974-76
(average) and 1980 are given in Appendix
3, Table 27.
Detailed data on the composition of
these feeds are not available, but cereals
constituted an important ingredient, al-
though the proportion of cereals varied from
type to type. This increase in the output of
manufactured feed also contributed to the
rapid growth in cereal feed use. These
trends are likely to continue into the future.


's The linkage of the two markets, grains for direct consumption and those for animal feed, and the policy
implications of the food-feed competition are discussed in Pan A. Yotopoulos, "Middle Income Classes and Food
Crisis: The 'New' Food-Feed Competition." Economic Development and Cultural Change 33 (No. 3, 1985):
463-483.







Table 1 -Ratios of prices received by farmers for poultry and pigmeat to maize, selected countries, 1967-80

Ratio/Country 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980


Poultry meat/maize price ratio
Argentina 9.50 10.80 8.90 9.50 12.18 7.70 8.14 10.12 19.48 13.03 12.83 11.96 16.29 14.14
Brazil 15.27 15.28 14.02 14.92 15.00 13.31 15.59 12.97 12.41 11.27 12.39 10.27 11.18 9.18
Egypt 11.26 14.59 13.33 13.37 13.32 13.40 12.36 12.95 13.39 15.53 14.89 17.69 18.39 11.64
India 5.97 7.34 6.84 8.13 7.92 6.85 6.70 5.64 6.97 10.24 8.78 9.67 10.45 10.86
Indonesia 34.76 23.66 12.99 13.66 17.91 17.54 20.64 20.55 17.08 13.71 20.98 27.94 19.81 18.70
Kenya 12.67 14.53 16.25 19.87 16.70 14.29 14.99 14.81 11.76 12.22 11.72 15.10 13.38 13.78
Korea, Republic of 16.75 16.68 15.86 17.78 16.84 12.86 12.74 12.67 14.71 17.49 17.76 18.68 15.56 11.29
Mexico 12.45 12.80 13.15 12.89 12.89 13.06 11.53 9.93 8.60 8.29 10.80 9.27 9.35 7.57
Nigeria 17.55 17.39 17.50 17.50 17.47 17.53 17.56 17.50 16.61 15.31 15.31 15.31 11.45 13.45
Philippines 10.24 11.12 10.76 9.44 8.70 8.22 8.65 8.74 8.67 8.75 9.78 10.52 13.40 13.34
Thailand 11.56 18.92 17.61 15.19 20.08 14.30 10.74 9.11 10.46 11.66 12.39 12.83 10.56 11.10
Turkey' 13.73 14.46 14.21 14.81 19.73 22.03 18.31 13.10 13.03 16.17 18.18 23.17 43.60 38.86
Pigmeat/maize price ratio
Argentina 8.11 12.29 10.76 9.45 11.34 10.91 10.10 12.52 12.61 11.72 12.25 13.62 14.79 10.99
Brazil 6.09 6.05 5.60 5.88 6.35 5.54 6.75 5.06 5.23 4.54 5.11 4.45 5.48 4.74
India 4.57 5.80 5.34 6.27 6.31 5.34 4.70 3.81 4.71 6.91 5.93 6.53 7.06 7.33
Indonesia 10.27 4.58 5.45 5.57 5.33 5.06 6.05 6.01 12.89 14.52 21.43 29.41 20.86 19.68
Korea, Republic of 7.33 9.82 8.42 9.96 10.47 7.59 9.16 7.32 9.01 10.88 9.79 13.52 9.78 8.87
Mexico 9.66 9.95 11.81 12.22 12.33 13.44 12.97 12.67 10.19 9.19 10.08 10.30 11.33 9.05
Nigeria 11.84 11.74 11.83 11.76 11.72 11.72 11.78 11.76 11.22 10.31 10.31 10.31 7.70 9.05
Philippines 7.55 7.76 6.89 7.44 7.28 6.42 6.24 6.11 5.66 5.46 8.89 10.56 10.43 10.51
Thailand 9.59 17.28 16.77 11.71 13.49 11.08 8.46 9.05 10.58 10.87 12.83 11.16 10.52 11.60

Source: Food and Agriculture Organization of the United Nations, "Producer Prices Tape, 1985," Rome, 1985.
a For Turkey, barley prices were used instead of maize prices.









5


GROWTH IN CEREAL FEED USE IN TAIWAN
AND THE REPUBLIC OF KOREA


The use of cereal grains as live
has increased sharply in Taiwar
Republic of Korea, both of which
industrializing economies with rap
ing per capital incomes and moder
income levels. Changes in their
tion of food and feed are indicate
tential developments in other d
countries if they also achieve rapid
growth.


Taiwan
In Taiwan, cereal feed use
dramatically from 1950-52 to 19


Period
(average)
1950-52
1960-62
1970-72
1980-82


Feed Use
(1,000 metric tons)
69
96
1,302
3,521


This phenomenal growth in feed
consequence of the changes in
sumption patterns that followed
remarkable economic progress. D
period, per capital income increase
age of 6 to 7 percent a year and tt
tion growth rate declined signific
about 3.5 percent in the early 19
present 2.0 percent per year.


Index of Per
Period Capita Income
(average) (1976 = 100)
1950-52 n.a.
1960-62 35
1970-72 72
1980-82 135


Po
(n


n.a. means not available.


stock feed In 1982, the population of Taiwan stood at
Sand the 18.3 million with an average per capital in-
are newly come of U.S. $2,340.
>idlygrow- During the second half of the 1930s,
ately high when Taiwan was under Japanese colonial
consump- rule, the average food consumption was less
ive of po- than 100 kilograms of cereal and 15 kilo-
eveloping grams of meat per year. These foods, along
economic with other components of the food basket,
such as starchy roots, pulses, fruits and veg-
etables, milk, and fish, supplied 1,865 cal-
ories of energy and 45 grams of protein per
day. By the second half of the 1940s, the
average per capital cereal consumption had
increased increased by 22 percent, but the intake of
80-82. meat had declined to nearly half of 1935-39
levels. Steady recovery began in the early
Feeding 1950s, when the per capital consumption
Ratio of cereals rose to 140 kilograms and that of
meat to 11 kilograms per year. Per capital
0.6 consumption reached 2,068 calories and
0.9 protein 47 grams per day. Cereal consump-
2.8 tion peaked in the early 1970s at about 165
4.0 kilograms a year but then declined to 134
kilograms in 1980-82. The consumption of
use was a livestock products continued to increase
food con- steadily, reaching 26 kilograms of meat in
Taiwan's 1970-72 and 41 kilograms in 1980-82. Per
during this capital consumption of milk and eggs also
d an aver- increased dramatically. Total calorie intake
ie popula- rose to 2,700 calories per day, while intake
antly from of proteins increased to 77 grams by 1980-
50s to the 82. The proportion of animal proteins in
the total protein intake also increased from
22 percent in 1950-52 to about 42 percent
in 1980-82.
pulation Far-reaching changes took place on the
millions) supply side also. In the second half of the
.4 1930s, cereal production in Taiwan con-
8.4 sisted almost entirely of rice, nearly 50
14.9 percent of which was exported, mostly to
17.9 Japan. By the early 1950s, rice production
reached 1.5 million tons, of which about
87,000 tons were exported, but the coun-









try imported about 70,000 tons of wheat
to augment domestic cereal supplies. From
then on, imports of wheat for direct con-
sumption increased regularly, reaching
nearly 300,000 tons by the early 1960s and
about 670,000 tons by 1980-82. Production
of rice also increased to 2 million tons by
1960-62. Thereafter, growth in rice produc-
tion slowed; rice output hovered between
2.4 and 2.6 million tons in the 1970s. Tai-
wan, which normally produces only a small
amount of coarse grain, increased its output
from less than 40,000 tons in 1960-62 to
about 125,000 tons in 1980-82. However,
coarse grain imports increased from 18,000
tons to 3.6 million tons during the same
period, mostly for use as livestock feed. As
food consumption habits shifted toward a
diet oriented to meat and other livestock
products, imports of wheat showed only a
small increase between 1970-72 and 1980-
82. The country started exporting large
quantities of rice in the late 1970s, reaching
430,000 tons in 1979. From a net exporter
of cereals in 1950-52, Taiwan became a net
importer, with net imports as a percentage
of total domestic utilization increasing from
13 percent in 1960-62 to 58 percent in
1980-82. As stated, the bulk of imports in
the 1980s were for livestock feed.
The increased demand for meat, milk,
and eggs in Taiwan was largely met from
domestic production. Meat output increased
steadily from less than 100,000 tons in
1950-52 to 728,000 tons in 1980-82. More
than 60 percent of this increase was in pig-
meat, which rose from 80,000 to 470,000
tons during this period. Impressive increases
were recorded in the output of milk and
eggs as well. Milk output rose from less than
1,000 tons to about 52,000 tons, while out-
put of eggs increased elevenfold to about
148,000 tons in 1980-82.
In LOUs, the output of livestock prod-
ucts increased from 108,000 tons in 1950-
52 to 881,000 tons in 1980-82. Thus the
feeding ratio-that is, the use of cereal feed
per unit of livestock output-increased from
0.6 to about 4.0 during this period. Selected
data on production and trade of cereals and
livestock products are given in Table 12.
Consumption data are given in Table 13.


Republic of Korea
Rapid increases in the output of live-
stock products also occurred in the Republic
of Korea. Korea had an average per capital
income growth rate of 7 percent a year dur-
ing the last two decades, and the average
per capital income was $1,910 in 1982.


Period Per Capita Income
(average) (1980 U.S. $)


1961-65
1969-71
1979-81


523
840
1,567*


Population
(millions)

26.7
31.3
38.5


*In current prices.


Between the 1960s and the end of the
1970s, the output of meat increased at a
rate exceeding 10 percent a year, rising
from a little less than 100,000 tons during
1961-65 to about 516,000 tons during
1979-81. The growth in fresh milk produc-
tion was even more phenomenal; it rose
from about 7,000 tons to 452,000 tons. The
output of eggs rose 5.5 times (Table 14).
To sustain this production growth, the
quantity of cereals used as livestock feed
increased from 229,000 tons in the early
1960s to more than 3 million tons during
1979-81. Of the latter, coarse grains-
mostly imported-accounted for about 90
percent. Thus the feeding ratio increased
from 1.6 to 3.7 during the period.


Period Feed Use
(average) (1,000 metric tons)


1961-65
1969-71
1979-81


229
725
3,052


Feeding
Ratio


1.6
2.4
3.7


On average, the Republic of Korea pro-
duced 4.6 million tons of cereals during
1961-65, of which two-thirds was milled
rice, about 30 percent was coarse grains,
and the balance was wheat (Table 14). By
1969-71, rice production rose to 3.6 million
tons, an increase of 16 percent over seven









Table 12-Average production and trade of cereals and livestock products,
feed use, per capital income, and population, Taiwan, selected years

Variable 1950-52 1960-62 1970-72 1980-82
(1,000 metric tons)
Cereal production
Rice (husked) 1,492 2,007 2,376 2,419
Wheat 17 44 3 3
Coarse grains 13 39 79 125
Total 1,522 2,089 2,457 2,547
Cereal exports
Rice 86.9 63.9 38.2 213.6
Wheat ... 2.0 1.8
Coarse grains
Total' 86.9 65.9 40.0 213.6
Cereal imports
Rice ... 46.1 2.4 4.7
Wheat 70.2 293.5 628.5 670.2
Coarse grains ... 18.2 1,208.1 3,585.5
Total" 70.2 357.8 1,839.0 4,260.4
Total available supply of cereals" 1,430.2 2,297.8 4,659.8 7,009.4
Net imports of cereals -16.7b 291.9 1,788.9 4,046.8
Net imports as a share of total
supply (percent) ... 12.70 38.39 57.73
Output of livestock products
Meat 94 145 398 728
Milk 1 5 19 52
Eggs 13 19 64 148
Total in livestock output units 108 164 465 881

Source: Council for Agricultural Planning and Development, Taiwan Food Balance Sheet, 1935-1980 (Taiwan:
Republic of China, July, 1981). Data from 1981 onward were obtained from the Council separately.
Notes: Meat includes beef, buffalo, mutton, goat, pig, and poultry meat, all in carcass weight equivalents.
Production and consumption were more or less the same, as the net trade in total meat use was almost
negligible. Parts may not add to total due to rounding.
a Total available supply = production + changes in stock + imports exports.
b This figure represents net exports.


years. Average output of rice reached 4.3
million tons during 1979-81, showing an
average growth rate of 2.0 percent per year
over the entire period. Coarse grains pro-
duction rose much faster during the first
period, but it declined thereafter. The aver-


age production during 1979-81 was even
lower than that in 1961-65. The overall pro-
duction of cereals, which increased at an
average rate of 3.1 percent a year in the
first period, declined by 1979-81, reducing
the overall growth rate to 1.1 percent. This


Table 13-Per capital consumption per year, Taiwan, selected years

Years Cereals Meat Milk Eggs Calories Protein
(kilograms) (per day) (grams/day)
1950-52 140 11.2 0.5 1.6 2,068 47
1960-62 161 12.0 1.3 1.6 2,345 58
1970-72 165 26.3 2.5 4.3 2,653 73
1980-82 134 40.6 5.8 8.2 2,706 77

Source: Council for Agricultural Planning and Development, Taiwan Food Balance Sheet 1935-1980 (Taiwan:
Republic of China, July, 1981). Data from 1981 onward were obtained from the Council separately.








Table 14-Average production and trade of cereals and livestock products,
Republic of Korea, selected years

Variable 1961-65 1969-71 1979-81
(1,000 metric tons)
Cereal production
Rice 3,077.5 3,566.7 4,339.2
Wheat 169.8 213.0 64.0
Coarse grains 1,387.7 1,961.0 1,229.0
Total 4,635.0 5,740.8 5,632.2
Cereal exports
Rice 17.4
Wheat ... 2.7
Coarse grains 0.5 4.8 1.0
Total 17.9 7.6 1.0
Cereal imports
Rice 20.9 657.9 1,005.4
Wheat 523.5 1,541.3 1,848.0
Coarse grains 190.0 288.9 2,827.0
Total 734.4 2,488.0 5,680.4
Total domestic utilization
of cereals' 5,268.4 7,621.4 10,933.4
Net imports of cereals 716.5 2,480.5 5,679.4
Net imports as a share of total
domestic utilization (percent) 13.6 32.5 51.9
Output of livestock products
Meat 98 163 516
Milk 7 53 452
Eggs 46 132 256
Total in livestock output units 144 300 817

Sources: Food and Agriculture Organization of the United Nations, (FAO), "Agriculture: Toward 2000 Tape,"
Rome, 1981; FAO, Food Balance Sheets, 1979-81 Average (Rome: FAO, 1982); and FAO, "Agricultural
Supply/Utilization Accounts Tape, 1981," Rome, 1982; and FAO, "Production Yearbook Tape, 1981."
Rome, 1982.
Notes: The ellipses indicate nil or negligible. Parts may not add to total due to rounding.
a Total domestic utilization = production' + net imports + change in stocks.


was nearly half of the population growth of
2.2 percent per year.
The increased demand for cereals both
for domestic consumption as food and the
derived demand for feed had to be met
through imports. These imports increased
eightfold from about 734,000 tons in 1961-
65 to 5.7 million tons in 1979-81. About
half of the imports during the latter period
were for livestock feed and consisted of coarse
grains. Thus net imports of cereals as a per-
centage of total domestic utilization increased
from about 14 percent to 52 percent.
The per capital consumption of cereals
also followed the expected path (see Table
15). Use of cereals for food rose from 178
kilograms per year in 1961-65 to 207 kilo-


grams in 1969-71, but in the next decade
it declined marginally to 205 kilograms. On
the other hand, the per capital consumption
of livestock products increased fourfold over-
all. Total consumption also increased during
this period-by about 35 percent in calories
and 50 percent in protein. The rapid in-
crease in livestock production led to a sub-
stantial increase in the derived demand for
cereal feed. Putting the food and feed uses
of cereals together, total per capital use of
cereals increased from 186 kilograms dur-
ing 1961-65 to 284 kilograms in 1979-81,
an increase of more than 50 percent. The
share of cereal feed in total grain use also
increased from 4 percent to 28 percent dur-
ing the same period.








Table 15-Per capital consumption per year, Republic of Korea, various years
Years Cereals Meat Milk Eggs Calories Protein
(kilograms) (per day) (grams/day)
1961-65 178 3.6 2.7 1.5 2,147 55
1969-71 207 5.2 4.6 3.8 2,539 67
1979-81 205 14.7 11.7 6.1 2,895 82

Sources: Food and Agriculture Organization of the United Nations, (FAO), "Agriculture: Toward 2000 Tape,"
Rome, 1981; FAO, Food Balance Sheets, 1979-81 Average (Rome: FAO, 1982); and FAO, "Agricultural
Supply/Utilization Accounts Tape 1981," Rome, 1982.


Conclusions

To summarize the conditions favorable
to a rapid rise in livestock output and the
derived demand for cereal feed based on
the experiences of Taiwan and Korea, first,
both the economies had rapid growth in per
capital incomes and their current income
levels were in the upper middle range. Sec-
ond, the per capital consumption of livestock
products until the early 1960s was low and
the income elasticity of demand for these
products was high, which led to rapid growth
in their aggregate demand. Third, both had
potential for the development of livestock
products, particularly pork and poultry
meat, milk, and eggs. They imported new
production technology from the developed
countries and invested in capital-intensive
commercial or semicommercial livestock
enterprises. Fourth, neither Taiwan nor the
Republic of Korea had large permanent pas-
tures or grazing areas, and hence they had
to depend on concentrate feeds. In both
countries land is a constraint: domestic
shortages in feedgrains were met through
large-scale imports, and the feeding ratio
increased rapidly. The infrastructure needed


for processing, refrigerated storage, transport,
and marketing of these perishable products
was available in both. The two economies
had rapid trade growth; thus foreign exchange
was not a constraint. They preferred to im-
port feeds and meet the increased demand
through domestic production rather than
importing livestock products from outside.
It is true that all these conditions may
not be replicated in other developing coun-
tries. For example, in some of the medium-
and high-income countries of Latin America,
where current levels of per capital livestock
consumption are relatively high, the income
elasticity of demand for these products is
lower, though still higher than that for ce-
reals. Consequently, future growth in per
capital demand for livestock products may
not be as fast. Moreover, in some of the
developing countries in Latin America and
Sub-Saharan Africa where land is not scarce,
the higher demand for feedgrains may be
met from increased domestic production.
What the Taiwanese and Korean experiences
indicate is that with rapid increases in per
capital income and high income elasticity of
demand for livestock products, the derived
demand for feedgrains grows rapidly, and
the feeding ratio also rises rapidly.








6


DYNAMICS OF DEMAND FOR FOOD AND
FEEDGRAINS IN SELECTED DEVELOPING
COUNTRIES


Data presented in Chapter 4 show that
the use of cereals for feed rose faster than
food use between the late 1960s and late
1970s in Third World countries. Projections
of demand for cereals and other food staples,
based on income elasticities and the con-
tinuation of past income trends, suggest that
this trend is likely to continue. The case
studies of Taiwan and the Republic of Korea
indicate that the derived demand for feed
could grow even faster than past trends
suggest. In general, as per capital incomes
rise in developing countries, consumption
of livestock products rises much faster be-
cause the income elasticity of demand for
these products is much higher than that for
direct consumption of cereals. This could
lead to a rapid rise in the derived demand
for feed, particularly in countries where the
increased domestic demand for livestock
products is met through accelerated domes-
tic production. In addition, the feeding ratio,
that is, the average quantity of cereals used
to produce one unit of livestock products,
increases because of changes in the compo-
sition and structure of livestock products
and feeds.
The proportion of poultry and pigmeat
in the total output of meat has increased
significantly in recent years. In developing
countries, poultry, pork, and dairy enter-
prises are increasingly organized along com-
mercial or semicommercial lines. They use
more compound and mixed feeds, which
have cereals as a major component. These
enterprises are no longer backyard opera-
tions that rely on farmyard and household


wastes and by-products. To meet the in-
creased demand for compound and mixed
feeds, the feed manufacturing industry is
developing rapidly, leading to larger de-
mand for feedgrains and consequently to
higher average feeding ratios in livestock
production.
Further, the income elasticity coeffi-
cient for livestock products, particularly
poultry and eggs, remains relatively stable
at one or a little above one in many develop-
ing countries (see Appendix 3, Table 28).
At low incomes, the per capital consumption
of livestock products is small, and hence the
derived demand for cereals for feed is small
compared with that for direct food consump-
tion. As incomes rise, the elasticity coeffi-
cient for cereals for direct food consumption
declines; the much higher income elasticity
of derived demand for cereals arising from
livestock products may also decline but less
rapidly. Although feedgrains' share of the
total domestic utilization of cereals is still
small in many developing countries, it
seems likely that it will grow rapidly, leading
to a decline in the share of cereals used for
food. Thus the behavior of the average in-
come elasticity of demand for grains re-
quires careful study. When the weighted
average elasticity is high, especially in Third
World countries with rapid per capital in-
come growth and moderately high popula-
tion growth, aggregate demand for grains
would increase rapidly. Where this demand
is not matched by increased domestic pro-
duction, cereal imports could increase sub-
stantially.16 Here the conditions under


16 Mellor, John W. "Third World Development: Food, Employment and Growth Interactions," American Journal
of Agricultural Economics 64 (May 1982): 304-311; and John W. Mellor and Bruce F. Johnston, "The World
Food Equation: Interrelations Among Development, Employment, and Food Consumption," Journal of Economic
Literature 22 (June 1984): 531-574.








which the combined demand for grains
would increase rapidly are examined analyt-
ically and tested with empirical data.


Equations (2), (3), and (4) can be substituted
into equation (1) to give17


D = PoCoI[(1 + rpt)t(l + gtryt)t]


Total Demand for Grains
The total demand for grain in year t, Dt,
is the product of population, Pt, and per
capital consumption, Ct:

Dt = PtCt. (1)


Further,


where rpt and ryt are the rates of growth in
population and per capital income, and rgt
is the income elasticity in period t.
The income elasticity for grain, Tig, is a
weighted average of the elasticities of food
(-qf) and feed ([N) with p and (1 p) as
weights, where 3 represents the share of
food and (1 P) is the share of feed in total
grain use,18 or


rig = PTIf + (1 P)h.,


Pt = Po(1 + rp)',


Ct = Co(l + rc)t, and (3)

re = (dc/dt)/c = [(dc/dy)(y/c)]
[(dy/dt)(l/y)] = -gry, (4)
where

Po = the population in base period 0,
Co = the per capital consumption in base
period 0,
rp = the rate of growth of population,
r, = the rate of growth of per capital con-
sumption,
c = per capital consumption,
y = per capital income,
rTg = the income elasticity of demand for
grain, and
ry = the rate of growth of per capital income.


The demand for feedgrains for, say, meat
production is equal to the demand for meat
(g3) multiplied by the feeding ratio (a) or
the amount of feed required to produce one
unit of meat. Then the income elasticity of
demand for feed equals the sum of income
elasticities of meat (Tim) and the relative
change in the feeding ratio to the percentage
change in per capital income (rl J:'9


"-% = 1lm + "a-


If equations (6) and (7) are combined,
the income elasticity of demand for total
grains equals

Sig = p-if + (1 p)(im + (l). (8)

Example 1
This example and the ones that follow
merely illustrate the behavior of these vari-


17 If T1g and r, are not constant over time, every time -ri or r, or both change, rc will also change. Thus the rate
of growth in per capital consumption over time period t is given by r, = sI, x r,, where ,gp and ry also relate
to time period t.
18 The total demand for grain, g, is the sum of demand for food (gi) and feed use (g2). Thus g = g, + g2; (dg/g) =
(gl/g)dgl + (g2/g)dg2 = pdg, + (1 p)dg2. Dividing both sides by (dy/y), r, = p x T, + (1 p) x I,.Total
domestic utilization of grain also includes its use for seed, nonfood industrial uses, and allowance for wastage.
These are not considered here for the sake of simplicity.
19 Thus, g2 = ag3, dg2/g2 = dg3/g3 + da/a. Dividing throughout by (dy/y), r,q = Tm, + 9,.
Here meat is taken to represent the livestock output; milk and eggs are not considered.








ables and do not necessarily represent the
actual conditions in any particular country.
Scenario I in Table 16 shows the behavior
of ng as -qf declines for different values of
p and for the constant value of 1"m = 1.0,
assuming that T1, = 0 and hence ,2 = 1.20
The combined elasticity first decreases
and then increases. The combined elasticity
increases rapidly when the ratio of feed to
food plus feed increases (or the value of p
declines).

Example 2
The values of qf and p are assumed to
remain the same in two other cases, sce-
narios 2 and 3. In scenario 2, %, and 'im
are taken to equal 1 and hence % = 2 (con-
stant). This implies that, associated with a
1 percent increase in per capital income, the
feeding ratio also increases by 1 percent. In
scenario 3, -qi = 0 and n', decreases. The
computed values of rg are given in Table 16.
In scenario 2, where = 2 (constant),
ig increases continuously. In the third sce-
nario where T1 and 'm decrease, -ig first
decreases, then increases, and finally de-
creases. Declining elasticities of meat occur
when meat consumption is already high, as
in parts of Latin America.


The two examples given above show
that apart from the values of f, 9im, and
7il, the behavior of p also influences the
changes in -lg.
The magnitude of the change in p de-
pends on the initial values of p, T~f, and
the rate of change in per capital income
(dy/y), as equation (10) shows:

P = gl/g; dp = (dgl/g) (gldg/g2). (9)


Thus


dp/p = (dgl/gi) (dg/g)
= (-f qg)(dy/y).21 (10)

Example 3
For values of p = 0.95, 'if = 0.22, and
T, = 1.00, the decrease in p associated
with a 4 percent increase in per capital income
will be 0.0015 per year. The same figure
works out to be about 0.0095 per year
where p = 0.50, ,f = 0.05, and 'r = 1.00.
The changes in -qg with changes in per
capital income are given by the inequality


drig/dy 0.


Table 16-Three scenarios for income elasticities of demand
Constant Scenario 1 Scenario 2 Scenario 3
M i r?1 q 7 < 9<

0.22 0.95 1.00 0.259 2.00 0.309 1.00 0.259
0.15 0.90 1.00 0.235 2.00 0.335 0.80 0.215
0.10 0.75 1.00 0.325 2.00 0.575 0.64 0.235
0.05 0.50 1.00 0.525 2.00 1.025 0.40 0.225


20 The assumption that i. = 0 implies that the average feeding ratio remains unchanged throughout the projection
period.
21 Dividing both sides of (9) by (dy/y), one obtains
(d3/p) + (dy/y) = [(dgl/gl) (dy/y)j [(dg/g) + (dy/y)j = lf g.








Differentiating equation (6) with respect to y,

dqg/dy = (dp/dy)rqf + P(dyf/dy)
+ [d(l p)/dy] ,
+ (1 p)(dI,/dy)
= (dp/dy)(r- )
+ P(dqf/dy)
+ (1 P)(d-,/dy). (12)

From empirical evidence, in the develop-
ing countries generally >rlqf; and hence
-{f -T, is negative. As incomes increase, P
usually decreases, and hence dp is negative.
Thus the first term is positive. q, is con-
stant or decreases. The second term is zero
or negative. Further, if 9, is constant, d,9
is zero, and hence the third term is zero.
Then, qg decreases or increases if

|P(d9f/dy)| -| (dp/dy) (rf- ) |. (13)

When -, decreases, then %g decreases or
increases if

IP(dy,/dy)l + 1(1 p)(dT,/dy) I
S|I(dp/dy)(Tf- 1)|. (14)

Thus the behavior of -qg is determined
by the inequalities (13) and (14).

Example 4
To determine the precise effects of growth
components on overall growth is complex
and varies from country to country. To illus-
trate, the grain demand for feed use was
projected for 12 selected countries assum-
ing that the income elasticity of meat is a
proxy for feed elasticity. Data on the pro-
jected growth rates of utilization of cereals


for food and feed based on this assumption
are given in Table 17.
Table 17 shows that if past income trends
continue, the overall demand for grains in
Kenya is projected to grow between 1980
and 2000 at 5.1 percent a year and in Nigeria
at 4.7 percent a year. In Kenya and Nigeria,
population growth is the major contributing
factor. In Indonesia, demand is expected to
grow less than 2 percent a year. In all the
other countries, demand would grow be-
tween 2.0 and 4.0 percent a year. In Turkey,
the combined food and feed elasticity is high
at about 0.5.
Growth in Foodgrain Use. The income
elasticity of demand for foodgrains, ~qf, is
low and is declining in a number of develop-
ing countries. The average implied income
elasticity during the 1980-2000 period is
negative in 5 out of the 12 countries: In-
donesia, Thailand, Malaysia, Turkey, and
Mexico.22 In India and the Philippines qf
lies between 0.20 and 0.24, and in the Re-
public of Korea it is very small. In the re-
maining four countries, r, is positive and
less than 0.20 (also see Appendix 3, Table
28). Taking into account the effects of pop-
ulation and income growth, the projected
growth in food use of cereals exceeds 4
percent per year in Kenya and Nigeria, and
it lies between 2 and 3 percent in India,
Egypt, the Philippines, and Brazil. In the
remaining six countries growth in the de-
mand for cereals for food use is projected
at less than 2 percent per year. Even in the
five countries where ilf is negative, food
demand would still increase because of pop-
ulation growth. And the combined elasticity
for food and feed is positive in these coun-
tries.
Growth in Feed Use. The income elastic-
ity of demand for feed use with that for
meat as a proxy is more than unity in 3 out
of the 12 countries.23 In Indonesia and
Nigeria, the projected feed growth exceeds
8 percent per year. In India, however, al-


22 Subsequent evidence shows that the income elasticity for cereals is found to be positive in Indonesia and higher
than that based on FAO data.
23 The income elasticity coefficients given in Appendix 3, Table 28 represent the values by type of meat, whereas
those in Table 17 relate to implied elasticities derived from the total projected demand for meat.






Table 17-Projected growth in food and feed use of cereals, 1980-2000 and average income elasticity of
demand for food and feed, selected countries

Per Capita Feed and Average Average Average
Per Capita Population GNP Feed Food Food Feed/Feed Feed Food Combined
Selected Incomes Growthb Growthc Growth Growth Growth and Food Elasticit' Elasticity' Elasticity
Country 1980 1980-2000 1980-2000 1980-2000 1980-2000 1980-2000 1980d 2000e 1980-2000 1980-2000 1980-2000
(U.S. $) (percent)
India 223 1.68 1.51 3.42 2.05 2.20 0.10 0.13 1.13 0.24 0.34
Kenya 393 4.25 2.68 6.74 4.65 5.06 0.17 0.23 0.89 0.14 0.29
Indonesia 441 1.53 5.14 8.16 0.66 1.92 0.09 0.29 1.27 -0.17 0.07
Egypt 522 2.32 3.87 5.84 2.41 3.52 0.26 0.40 0.89 0.02 0.30
Thailand 688 1.78 4.35 3.61 1.50 2.06 0.23 0.31 0.41 -0.06 0.06
Philippines 696 2.21 3.38 4.66 2.93 3.32 0.20 0.26 0.71 0.21 0.32
Nigeria 925 3.55 4.55 8.31 4.32 4.72 0.07 0.14 1.01 0.16 0.25
Turkey 1,524 2.18 3.52 5.63 1.00 3.90 0.53 0.73 0.96 -0.33 0.48
Korea, Republic of 1,696 1.31 6.00 4.15 1.31 2.31 0.29 0.42 0.47 0.00 0.16
Malaysia 1,560 2.00 4.61 5.01 1.61 3.03 0.34 0.50 0.64 -0.08 0.22
Brazil 2,061 1.98 6.00 4.61 2.48 3.80 0.57 0.67 0.43 0.08 0.30
Mexico 2,623 2.29 3.60 4.41 1.92 3.25 0.48 0.60 0.58 -0.10 0.26

Sources: Calculated from computer printouts used in Leonardo A. Paulino, Food in the Third World: Past Trends and Projections to 2000, Research Report 52
(Washington, D.C.: International Food Policy Research Institute, 1986). GNP data are from World Bank, "Gross National Product by Country Data Tape,
1961-80," Washington, D.C., 1981; and population data are from United Nations, Department of International Economic and Social Affairs, "World
Population Prospects as Assessed in 1982," New York, 1983 (computer printout).
SPer capital income is based on the 1961-80 trend of GNP in real terms (1979-81 = 100).
b Population growth is based on United Nations medium-term population projections, as assessed in 1982.
c Per capital GNP is based on 1966-80 trend growth in per capital income. An upper limit of 6 percent was imposed on per capital income growth in projecting
it for 1980-2000.
d 1980 data relate to trend values.
e This is based on projected values of food and feed in 2000. Other uses of cereals, namely seed and waste, are not considered.
f The average elasticity of demand for feed is the same as the average elasticity of demand for meat, worked out on the basis of the implied growth rate in
projected per capital meat consumption between 1980 and 2000.
SThe average elasticity of demand for food is based on the implied growth rate in projected per capital cereal consumption between 1980 and 2000.








though 7., is high, the rate of growth in per
capital income is low; hence feed growth is
about 3.4 percent. In Turkey and Kenya,
where 9, is a little less than unity, the pro-
jected rates work out to 5.6 and 6.7 percent
respectively. In the high-income countries
of Brazil and Mexico, -% is less than 0.6.
In spite of Brazil's high income growth and
Mexico's high population growth, feed pro-
jections show 4.6 and 4.4 percent growth
rates respectively. Because of low popula-
tion growth in the Republic of Korea, feed
growth works out to 4.2 percent, despite
the high income growth and 0.5 feed elas-
ticity.

Behavior of P Ratios
Appendix 3, Table 29 gives the ratios
of the amount of grains for food use to the
combined amount for food and feed use in
the 12 selected countries, calculated on the
basis of the elasticities referred to above.
These indicators confirm the earlier obser-


vation that in Turkey, Brazil, and Mexico,
where the ps were relatively low in 1980,
they will decline much faster by 2000, and
that in countries where the initial ps are
high, the decline is much slower.



Conclusions
This analysis shows that projecting the
behavior of the overall demand for grains
is quite complex and that the conclusions
are sensitive to the assumptions made with
regard to the various growth elements. As-
sunfing that meat elasticity is a proxy for
feed elasticity implies that the feeding ratio
is constant, which is not the case in some
Third World countries. Feeding ratios do
change rapidly, as in Taiwan and the Repub-
lic of Korea, for example. For this reason,
the resulting projections probably underes-
timate the potential growth rates of demand
for feedgrains.









7


PROJECTIONS OF CEREAL FEED USE
IN DEVELOPING COUNTRIES


Any assessment of the total future de-
mand for cereals for livestock feed should
ideally take into consideration the desirable
composition of feed in terms of cereals, non-
cereal feeds, fodders, forages, and crop by-
products. The projected livestock output
should also be considered by type, and the
feed required to produce it in terms of meta-
bolic energy, protein, and total digestible
nutrients based on average feeding sched-
ules. In addition, prices of alternative feeds,
scope for improving the efficiency of feed
use, and the relationship between input and
output prices have to be kept in view.
One approach for estimating feed re-
quirements is to use a regression framework
that defines cereal feed use as a function of
a number of variables and to determine the
relevant coefficients quantitatively based on
cross-section or time-series data. The var-
iables might include per capital income;
prices; size, composition, and output of the
livestock sector; and availability of pastures.
However, the poor quality of the data on
feed use and some of the other relevant
variables in a majority of the developing
countries precluded the adoption of this ap-
proach. Alternative rough and ready meth-
ods, based largely on projected livestock
outputs, feeding ratios, or extrapolation of
past trends, where appropriate, were adopted
instead.
In IFPRI's study of livestock products,
the output of meat, milk, and eggs were
each projected to 1990 and 2000, based on
the continuation of past trends in the output
for each country during 1961-77.24 These
projected outputs were then converted into
LOUs, as explained in Chapter 3, and aggre-
gated for each subregion. These subregional
aggregates multiplied by the corresponding


feeding ratios give the requirements of ce-
real feed in 1990 and 2000. The projected
feeding ratios for the years 1990 and 2000
were obtained as extrapolations of the linear
trends fitted to the annual data on feeding
ratios, which in turn were calculated by di-
viding the actual feed use (as reported in
FAO's "Agricultural Supply/Utilization Ac-
counts") in each year by the corresponding
livestock output units for the period 1966-
80. This is the first method.
In developing countries in the past, ag-
gregate use of cereal feeds expanded faster
than output of livestock products in 10 out
of 12 subregions. The second method is
based on the assumption that this difference
in growth rates will continue, provided ob-
served changes in the composition and
structure of meat and milk output and cur-
rent production techniques continue. To re-
flect these changes, the projected growth
rate implied by the livestock output projec-
tions for 1980-2000 was first calculated for
each subregion. Next, the difference be-
tween the 1966-80 trend growth rate of
cereal feed use and that of livestock output
(expressed in LOUs) in the subregion were
computed. The difference was added to the
1980-2000 growth rate of livestock output.
The adjusted growth rate for cereal feed
thus obtained was then applied to 1980
trend estimates of cereal feed use in each
subregion to obtain the projected feed use
in 1990 and 2000.
Another approach, which involved trend
projection of time-series data on cereal feed
use for each country during the period 1966-
80, was also tried. In the methodology for
trend projections adopted by IFPRI, output
projections are based on an extrapolation
of past trends, assuming unchanged relative


24 Sarma and Yeung, Livestock Products in the Third World.









prices. This method was adopted both for
staple food crops and livestock products.
Because the requirements of cereal feeds
represent the derived demand for their con-
version into livestock products, a similar
approach was adopted for projecting feed
use. A close examination of the fitted expo-
nential trend equations showed deficiencies
in several countries. Either the R2 was too
low or the coefficients were not significant,
particularly in countries that used only small
quantities of feed. In some countries, espe-
cially in Sub-Saharan Africa, growth rates
were extremely high mainly because of data
problems; extrapolation of these trends to
1990 and 2000 would have led to unrealis-
tically high estimates unless upper limits to
future growth were arbitrarily fixed. In order
to surmount these difficulties, two modifi-
cations were adopted. First, the data were
grouped by subregions, and trend growth
rates were calculated based on the annual
aggregate feed use in each subregion for the
period 1966-80.25 These growth rates were
extrapolated to give the projected feed use
in 1990 and 2000, under the third method.
It was further noted that in 29 out of
104 countries cereal feed use exceeded
500,000 tons in 1980, and aggregate feed
use in these countries formed 90 percent
of the Third World total (Appendix 3, Table
30). In the fourth method, cereal feed use
in each of these 29 countries was individu-
ally projected. The remaining 75 countries
were grouped by level of per capital income
into four classes: less than $250, between
$250 and $499, between $500 and $1,999,
and $2,000 and more. The cereal feed use
in each group was projected separately. The
projected values were added to give the
aggregate projected feed use for the 104
countries.
In both the third and fourth methods,
the period 1966-80 is used as a data base
for growth rate projections. Questions have


been raised on whether this is an appro-
priate base period. It has been argued that
per capital incomes in developing countries
rose very rapidly during most of this period,
and the consumption of livestock products
also increased rapidly. If economic growth
slows, the demand for these products may
not grow as quickly.26 It is true that as the
result of improvements in agricultural tech-
nologies and institutions, increasing adop-
tion of outward-looking industrial policies,
and favorable external conditions, the 1966-
73 period witnessed high economic growth
in some developing countries. Subsequently,
however, the oil shocks of 1974 and 1979-
80, the ensuing recession and protection-
ism in the countries belonging to the Or-
ganization for Economic Cooperation and
Development (OECD), and mounting debt re-
payment problems slowed economic growth
in the developing countries. There are, how-
ever, indications of economic recovery. The
average economic growth during 1966-80
seemed to be one that developing countries
could achieve through the structural adjust-
ments that are now being undertaken in
several developing countries. These promise
improvements in the efficiency with which
resources are allocated and used, thus en-
hancing the prospects for better economic
performance in the future.
Moreover, growth in the use of cereal
feed is related to growth in the output of
livestock products, particularly poultry, pig-
meat, eggs, and milk. As long as trends in
these products continue, the derived de-
mand for feed will also continue to grow
rapidly. Analysis of more recent trends in
the output of livestock products in develop-
ing countries (excluding China) during the
period 1973-75 to 1982-84 shows further
acceleration in the growth rate of meat,
milk, and egg output.27
The fifth method is that adopted in the
IFPRI research report by Leonardo Paulino,


25 This is based on the assumption that statistical errors tend to be compensated for in aggregates.
26 These questions are discussed more fully in Sarma and Yeung, Livestock Products in the Third World
27 The average output of meat increased from 22.1 million tons in 1974-76 to 29.6 million tons in 1982-84-an
average annual rate of 3.7 percent. For milk, the average increase was 2.8 percent a year from 87.1 million tons
to 108.4 million tons over the same period. Production of eggs also accelerated, growing 6.1 percent a year and
rising from 4.3 million tons to 7.0 million tons.









Food in the Third World In estimating the
projected animal feed use in 2000, the in-
come elasticity of demand for meat was used
as a proxy for that of feed. However, because
the report gave the projected feed use of
major food crops as a group, the cereal com-
ponent was calculated for 1990 and 2000,
using the same method. Thus trend esti-
mates of per capital use of cereal feed in
1980 were obtained, and these were pro-
jected to 1990 and 2000, using trend per
capital income growth and income elasticity
of demand. This approach to projecting the
derived demand for animal feed essentially
assumes that feed use in a country would
grow as fast as the demand for meat. Pro-
jected feed use would be overestimated in
countries where meat consumption growth
is faster than growth of production and
underestimated in countries where produc-
tion growth rates are faster. Further, to the
extent that income growth during the early
1980s is lower than the 1966-80 trend and
is not compensated for by faster economic
growth during the remainder of the decade,
consumption of livestock products may not
grow as fast as projected; thus the calculated
demand for feed based on the elasticity of
meat may also be delayed in reaching pro-
jected levels.
It was seen in Chapter 6 that estimating
feed requirements using the income elastic-
ity of meat as a proxy for feed elasticity
implies that the feeding ratio would be con-
stant during the projection period. If the
feeding ratio increases, as it did in Taiwan
and the Republic of Korea, the projected
estimates using this method would be too
low. This method also assumes that income
elasticity of meat represents the combined
elasticity of meat, milk, and eggs. Thus, al-
though the fifth method has the advantage
of easy conceptualization and computation,
the projections under this method have
their limitations as well. Details of these
five methods are given in Appendix 1.
Table 18 and Figure 4 show trend esti-
mates for cereal feed use in 1980, calculated
as the aggregate of subregional trend values,
and the projected cereal feed use in 1990
and 2000, based on the five alternative
methods.


If historical trends in the output of live-
stock products continue, developing coun-
tries are projected to produce 51 million
tons of meat, 178 million tons of milk, and
15 million tons of eggs, a little more than
83 million tons in LOUs by the end of the
century. Using the first method, the average
projected feeding ratio for 2000 works out
to a little more than 2.9. Thus the estimated
level of cereal feed use for the Third World
would amount to about 245 million tons,
which is roughly two-and-a-half times the
level reached in 1980. This implies an an-
nual growth rate of about 4.7 percent. The
estimate for 1990 indicates a 50 percent
increase over 1980. Nearly 46 percent of
estimated cereal feed use in 2000 would be
in Latin America, 28 percent in Asia, 23
percent in North Africa/Middle East, and
the balance in Sub-Saharan Africa.
Among the subregions, rapid growth in
feed use exceeding 6 percent per year is
projected for North Africa and for Central
America (which includes Mexico and the
Caribbean) if the existing trends in livestock
output are maintained (see Appendix 3,
Table 31). Lower South America at 1.5 per-
cent and South Asia at 2.0 percent would
have the slowest growth. In both these sub-
regions the growth in livestock output be-
tween 1980 and 2000 is projected to be
relatively low, and the rate of growth in
feed use during 1966-80 was less than the
corresponding growth in livestock output.
If, instead of dividing milk by 10 and
adding meat and eggs to arrive at the aggre-
gate livestock production in LOUs, the alter-
native procedure of converting all livestock
products into Physiological Fuel Values dis-
cussed in Chapter 3 is adopted, and the
same methodology is adopted for projecting
feed requirements, the projected levels of
feed use come to 153 million tons in 1990
and 251 million tons in 2000. The latter
represents a 2.4 percent increase over the
estimate under the former procedure for
2000.
Using the second method, the projected
levels of livestock output represent an an-
nual compound growth rate of 3.7 percent
between 1980 and 2000. Adding to this the
difference between LOUs and feed growth









Table 18-Cereal feed use projections under alternative assumptions, by
region, 1990 and 2000

Rate of Total 104 NorthAfrica/ Sub-Saharan Latin
Method Year Growth" Countries Asia Middle East Africa America
(percent) (million metric tons)
Trend base 1980 n.a. 98.1 30.7 21.7 4.0 41.7
Method 1 1990 4.4 150.6 44.3 33.8 5.8 66.7
2000 4.7 244.9 68.4 56.3 8.5 111.6
Method 2 1990 5.0 160.2 46.4 36.9 5.9 71.0
2000 5.2 269.5 72.1 63.7 8.6 125.2
Method 3 1990 5.0 159.2 47.2 35.3 5.5 71.1
2000 5.1 265.5 73.9 58.2 7.5 125.9
Method 4 1990 5.2 162.5 n.a. n.a. n.a. n.a.
2000 5.5 285.6 n.a. n.a. n.a. n.a.
Method 5 1990 4.8 156.9 47.8 39.6 6.7 62.9
2000 4.7 244.6 72.6 69.9 11.7 90.4

Sources: Food and Agriculture Organization of the United Nations, "Production Yearbook Tape, 1981," Rome,
1982; Food and Agriculture Organization of the United Nations, "Agricultural Supply/Utilization Accounts
Tape, 1981," Rome, 1982; J. S. Sarma and Patrick Yeung, Livestock Products in the Third World: Past
Trends and Projections to 1990 and 2000, Research Report 49 (Washington, D.C.: International Food
Policy Research Institute, 1985); World Bank, "Gross National Product by Country Data Tape, 1961-80,"
Washington, D.C., 1981; and computer printouts used in Leonardo A. Paulino, Food in the Third World:
Past Trends and Projections to 2000, Research Report 52 (Washington, D.C.: International Food Policy
Research Institute, 1986).
Notes: Parts may not add to total due to rounding, n.a. indicates not applicable. Method I is based on projected
values of the feeding ratio in 1990 and 2000 applied to the projected value of livestock output units
(LOUs) in 1990 and 2000. Method 2 is an extrapolation of 1980 trend value of feed use, using the
growth rate of LOUs in 1980-1990/2000 adjusted for the difference between the feed use growth and
LOU growth during 1966-80. Method 3 is an extrapolation of the 1980 trend values of feed use in each
subregion, based on 1966-80 trend growth. Method 4 is an extrapolation of 1980 trend values of feed
use in 29 individual countries and the rest of the countries grouped by per capital income level. Method
5 uses the income elasticity of demand for meat as a proxy for the income elasticity of derived demand
for feed.
a This is the rate of growth between 1980 and 1990/2000.


rates, the use of cereal feed is estimated to
increase at about 5.2 percent a year, reach-
ing about 270 million tons in 2000 com-
pared with 245 million tons under the first
method. The projected level in 1990 is
about 160 million tons, which is about 10
million tons higher than in the first method.
The regional distribution of feed use is simi-
lar under the first and second methods.
Subregional trend projections proposed
under the third method indicate that the
estimated feed use in 1990 and 2000 would
be 160 and 265 million tons respectively.
The projections are close to those under the
second method. At 5.1 percent a year, the
projected growth rate between 1980 and
2000 is slightly less.

28 Paulino, Food in the Third World


The trend projections of feed use for the
29 individual countries and the grouped ag-
gregates for the other countries envisaged
under the fourth method would put the es-
timated feed use in 2000 at 286 million
tons, implying an average growth rate of
5.5 percent a year. This is the highest among
the five estimates. The 29 countries would
have a total feed use of 261 million tons in
2000. The rest of the countries would share
25 million tons by the end of the century,
as shown in Table 19.
Under method 5, which is based on the
IFPRI food gap study,28 the projected cereal
feed demand is 157 million tons in 1990
and 245 million tors in 2000, implying
growth rates of 4.8 and 4.7 percent respec-









Figure 4-Trend cereal feed use in 1980 and projections to 1990 and 2000,
by region, based on projected livestock output



Asia


i 1980 trend value
1990 projection
2000 projection


North Africa/
Middle East



Sub-Saharan
Africa



Latin America



Study
countries


Million
Metric
I Tnn,


0 50 100 150 200 250
Cereal feed use

Sources: Calculated from Food and Agriculture Organization of the United Nations, "Production Yearbook Tape,
1981," Rome, 1982; and Food and Agriculture Organization of the United Nations, "Agricultural Supply/
Utilization Accounts Tape, 1981," Rome, 1982.


Table 19-Projected cereal feed use
in 75 selected countries,
by per capital income
groups, 1990 and 2000

PerCapita Rateof Projected Cereal
Income Growth, Feed Use
Group 1966-80 1980 1990 2000
(percent/ (million metric tons)
year)
Less than $250 0.3 1.50 1.55 1.59
$250-$499 3.8 1.98 2.88 4.20
$500-$1,999 5.1 4.26 7.01 11.52
$2,000 oror more 6.3 2.26 4.16 7.66
Total75 countries 4.7 9.99 15.59 24.97

Source: Food and Agriculture Organization of the
United Nations, "Agricultural Supply/Utiliza-
tion Accounts Tape, 1981," Rome, 1982; and
World Bank, "Gross National Product by Coun-
try Data Tape, 1961-80," Washington, D.C.,
1981.
Note: Parts may not add to total due to rounding.


tively over the trend value in 1980. Al-
though the projected Third World demand
for feed in 2000 under methods 1 and 5 is
very nearly the same, the regional distribu-
tion of this demand is different. Under
method 5, Latin America's share of the total
comes to about 37 percent, nearly 9 percent
lower than under the first method. North
Africa/Middle East accounts for less than
29 percent, which is about 6 percent higher
than under the first method.
Among the 29 countries for which feed
use is individually extrapolated under method
4, the projected levels of feed use are quite
high in Mexico, the Republic of Korea, and
Venezuela compared to those under the fifth
method, whereas in Turkey the projected
level is much lower. Moreover, in four coun-
tries, namely Colombia, the Republic of
Korea, Saudi Arabia, and Venezuela, where


I _


|V.lull








1966-80 trend growth of cereal feed ex-
ceeded 10 percent a year, a constraint of
10 percent was imposed in extrapolating
for 1990 and 2000. The aggregate projected
demand depends upon the level at which
the constraint is placed.
Thus, these five methods of projection
give a range of feed demand varying from
151 million to 163 million tons in 1990
and from 245 million to 286 million tons
in 2000. The results under methods 1 and
5 are close in 2000. These imply an average
growth rate of 4.7 to 5.5 percent a year
between 1980 and 2000. The correspond-
ing feeding ratio would range from 2.9 to
3.4, as compared to 2.4 in 1980.


Comparison with FAO Results
The FAO global study Agrculture: To-
ward 2000 also covered livestock products,
but apart from providing a broad indication
of projected growth in cereal feed use for
90 developing countries, it does not go into
details on feed projections.29 The estimates
are given under two scenarios: an optimistic
scenario A and a medium growth scenario
B. According to these indications cereal feed
is projected to grow at annual rates of 5.7
and 4.8 percent between 1980 and 2000,
under scenarios A and B respectively. These
compare with IFPRI projections of 5.5 and
4.7 percent annual growth under methods
4 and 1.


29 Food and Agriculture Organization of the United Nations, Agriculture: Toward 2000 (Rome: FAO, 1981).










8


CONCLUSIONS AND POLICY IMPLICATIONS


IFPRI's projections of the Third World
food gap based on the continuation of 1961-
80 trends in production and 1966-80 trends
in per capital income show that the develop-
ing countries (excluding China) could have
an overall net food deficit of 76 million tons
of basic food staples, including cereals and
other food crops, at the end of the century.30
They also indicate that although Asia could
have a net surplus, the other three regions
are likely to have net deficits-sizable defi-
cits in North Africa/Middle East and Sub-
Saharan Africa. These estimates are based
on the projected total domestic utilization
of 874 million tons of cereals and 172 mil-
lion tons of other staple food crops. Of these,
the projected demand for livestock feed is
estimated at about 245 million tons of ce-
reals and 23 million tons of other staple
food crops.31 The analysis in the preceding
chapter has shown that if past trends in
livestock production and use of cereals for
feed continue, the derived demand for ce-
reals for feed could rise to about 285 million
tons under method 4, an increase of 40
million tons.


Implications of Accelerated
Livestock Production
There are, however, some factors that
may cause cereal feed use at the end of the
century to be even larger than projected
under method 1, which is related to the
projected output of livestock products in
2000. For example, IFPRI's analysis of live-


stock consumption in the Third World
shows that if per capital incomes in the de-
veloping countries continue to rise at rates
attained during 1966-77, trend projected
output of meat and milk would fall consid-
erably short of demand at constant prices
by the end of the century, even though out-
put increased at 1961-77 trend growth
rates. Meat and milk output growth rates
would have to be raised 50 percent higher
to balance the demand in the developing
countries.32 Even under an alternative
scenario that projects a 25 percent slower
growth in per capital income than the 1966-
77 trend, the gaps between projected pro-
duction and consumption of meat and milk
would still be large, and domestic produc-
tion would have to expand at rates higher
than the 1961-77 trends to meet the pro-
jected demand.33 In either case, the derived
demand for cereal feed could be large, lead-
ing to even larger cereal deficits.
Between the early 1960s and mid-
1970s, the 3.2 percent growth in consump-
tion of meat was faster than production in
developing countries, which rose at an aver-
age rate of 2.9 percent a year. The increase
in domestic demand is reflected in the net
exports of meat by Third World countries,
which declined from 760,000 tons in the
early 1960s to 300,000 tons in the mid-
1970s. Developing countries had already
become net importers of meat by the end
of the 1970s. Using accelerated domestic
production to substitute for imports would
result in more rapid growth in feed use.
Value added in the processing and manufac-


30 Paulino, Food in the Third World.
31 Ibid. This study projected the output of major food crops as a group, not separately for cereals and noncereals.
Hence separate estimates of projected deficits in cereals are not available.
32 Sarma and Yeung, Livestock Products in the Third World
33 Ibid. Part of the gap could be filled through imports from developed countries, but some of the developing
countries might decide to meet the increased demand through increased domestic production. The basis for the
two income trend scenarios is discussed in Chapter 7 of that report. Further, these projections of demand and
supply assume constant relative prices.









ture of meat, eggs, and dairy products is
high; thus developing countries may wish
to make this substitution.34 Taste prefer-
ence for fresh meat and eggs, instead of
imported frozen or packaged meat and egg
powder, may also influence this decision.


Consequences of Easing
Marketing and Price
Constraints
Livestock products are perishable and
require refrigerated storage and transport
to facilitate their marketing, particularly in
tropical countries.35 Well-organized slaughter-
houses are often located in or around urban
areas, and the rural population does not have
easy access to them. Thus in those develop-
ing countries where infrastructure facilities
are not developed, output growth, and con-
sequently consumption growth, have been
constrained.
Taking, for example, the basic data con-
sidered earlier for the early 1960s and mid-
1970s, consumption of meat increased 3.2
percent a year in the 104 study countries
taken as a whole. During the same period,
population rose 2.6 percent, resulting in a
per capital increase in meat consumption of
0.6 percent a year. As per capital incomes
grew at 3.4 percent, the implied overall
demand elasticity came close to 0.2. Cross-
section studies show, on the other hand, a
much higher average income elasticity of
demand for meat. Without the supply con-
straints, consumption would have grown
even faster. The story is similar for milk and
eggs. With increasing per capital incomes
and improved infrastructure development,
either through public or private investment,
marketing constraints could be eased, lead-
ing to more rapid growth in output and con-
sumption of livestock products in the future.


In countries where the higher demand
for livestock products resulting from rising
per capital incomes and increasing urbaniza-
tion cannot be met by increased domestic
supplies, prices of these products rise. This
happens particularly when domestic supplies
cannot be fully augmented through imports
because of limitations of trade in these com-
modities, such as trade barriers, perishabil-
ity, and storage and handling problems. A
recent study in the Philippines found signif-
icant negative own-price elasticities for
meat. Moreover, demand of lower-income
groups was much more price responsive
than that of higher income groups.36 If the
constraints on trade were not operating, the
growth in consumption would have been
more rapid.


Improvements in Feeding
Efficiency and the
Feeding Ratio
The assessment of expected growth in
cereal feed use in 1990 and 2000 under
alternative assumptions in the preceding
chapter was beset with difficulties, partly
because inadequate data precluded a better
understanding of the processes through
which increases in derived demand for feed
take place. With the adoption of improved
breeding and scientific feeding practices,
feeding efficiency increased and feeding
ratio declined over time in some of the de-
veloped countries with relatively more ad-
vanced livestock production systems. On
the other hand, in the developing countries,
aggregated data show increases in the feed-
ing ratio if only feedgrain is taken into con-
sideration.
Improved breeding and scientific feed-
ing have reduced the feed input-output ratio
for each species within a specific production


34 When world prices of milk powder are low, many developing countries may not have comparative advantage
in encouraging domestic production of milk. Still, several countries do so, and the reasons for this should be
studied further.
35 The problem of perishability of livestock products may be partly overcome through trading in live animals;
however, such transport over long distances and across international borders raises another set of problems.
36 B. Ragalado, "The Distributional Impact of Food Policies on Human Nutrition in the Less Developed Countries-
The Case of the Philippines," (M.S. thesis, University of the Philippines at Los Bafios, 1984).







system. However, modern livestock produc-
tion systems based on intensive grain feeding
are gradually replacing traditional scaveng-
ing systems in developing countries. More-
over, there has been a gradual shift in the
output mix away from beef production to-
ward poultry, pork, and dairy products.
These two factors have gradually led to
higher feeding ratios. In the long run, how-
ever, feed efficiency could improve in the
developing countries also. This will tend to
reduce the prices of livestock products, thus
increasing the demand for them and also
the derived demand for feed. Therefore,
even if the direct effect of improving feed
efficiency is to reduce the overall feed de-
mand, the indirect effect will be offsetting.
In developed countries more cereal
grains are used for feed than directly for
food. It is clear that with rising per capital
incomes, this phenomenon will extend to
developing countries, as has already occurred
in a number of Third World countries. An
assessment of the extent of this change and
how long it will take is difficult. The small
proportion of feed use in total domestic utili-
zation of cereals at present holds down the
weighted average income elasticity in sev-
eral of the developing countries. However,
the weighted average elasticity, based on
empirical data for the countries discussed
in Chapter 6, does not allow for changing
feeding ratios. When these feeding ratios
increase rapidly, either because of changes
in the composition of meat or changes in
feed components and their relative prices,
the weighted average elasticity will also rise.
In some countries of Latin America
where current levels of meat consumption
are already high, the income elasticity of
demand for meat is relatively low, but higher
than that for cereals for food, because the
income elasticity for cereal feeds declines
less rapidly than that for food.37 In these
countries, the combined (weighted) elastic-
ity of cereal demand for food and feed is
not very high. On the other hand, in coun-
tries where per capital income growth, in-
come elasticity of meat, and feeding ratios


are all high, rapid expansion in cereal feed
use can be expected, as in Taiwan and the
Republic of Korea. The projected cereal feed
use in 2000 would give a feeding ratio rang-
ing from 2.9 to 3.4 for all the Third World
countries taken together, as compared to a
developed-country average of 3.6 in 1980.

The Relationship Between
Livestock Development
Strategy and Feed Demand
The derived demand for feed depends
upon the composition of livestock output
and the strategy for livestock development
that is adopted. If poultry and pig raising
and the dairy industry are encouraged, and
thus grow more quickly, the demand for
concentrate feeds with grain components
could be even larger. Also, if more commer-
cial and semicommercial enterprises develop
for livestock industry expansion, the growth
in grain-based feed demand may be much
more rapid than it would be if development
is organized through rural cooperatives or
other forms of labor-intensive production
strategies. In the latter case, greater use could
be made of locally available agricultural by-
products and wastes, leading to reduction
in costs and hence to larger consumption
of livestock products.
Coarse grains are often grown on poor
soils by small farmers, and the higher de-
mand for such grains could result in in-
creased incomes for the poor.


Larger Use of By-Products
and Wastes
It is likely that many agricultural by-
products and wastes, though available, are
not at present being utilized as livestock
feed, partly because their bulkiness and the
scattered nature of agricultural production
result in high collection costs, and partly
because of lack of knowledge regarding their
suitability for feed use. However, the poten-
tial for their increased use in the future


7 John W. Mellor, "Third World Development: Food, Employment and Growth."







exists. In some cases, these products may
need further processing for which the tech-
nology may not now be available. Though
the compound feed industries in developing
countries "generally follow the technology
evolved in developed countries, there are
no technical reasons to prevent the use of
larger amounts of locally available noncereal
feed compounds."38 Developing countries
may need to spend more resources on devel-
opment of technology to enable greater use
of such materials. In addition, governments
may have to offer price and other incentives
to encourage their use. These questions
need further study in the context of rapidly
growing feed demand.


Cereal Feed Use and
Food Security
Although IFPRI projections are based on
constant relative price assumptions,
realized feed demand will depend upon a
host of other factors including food prices,
feeding ratios, and so forth. For developing
countries, the use of grains as livestock feed
has food security implications. In several of
these countries, feedgrains are diverted to
food use in drought years, adversely affect-
ing livestock numbers and output. This has
implications for short-term food security,
medium-term nutrition, and income and
employment, particularly for small farmers
for whom livestock is a major source of
livelihood. Livestock is treated as a store of
wealth and a source of savings in African
countries. It is a hedge against crop failure,
though in drought years losses in livestock
are more widespread, and it may take two
to five years for farmers to recoup their
losses.
Even if part of Third World deficits in
meat and milk are met with imports from
developed countries, the derived demand
for cereal feed in the developed countries
would increase, reducing the overall quan-


titles available for direct consumption. As
was explained in Livestock Products in the
Third World, coarse grains are directly con-
sumed as food in developing countries, and
their increased feed use could mean a diver-
sion from food use. Competition between
the two uses may result in higher prices,
which would cause hardship to the most
vulnerable population groups. Moreover, a
food-feed competition for the use of grains
may develop not only between the rich and
the poor people within a country, but also
between the richer and poorer countries.39
As Yotopoulos observes, "In the traditional
food-feed controversy, animals competed
with people for feed versus food. The mod-
em variant has retained this aspect through
competition for land. But another dimen-
sion has been added: people who compete
with people for the indirect versus direct
consumption of cereals. In this competition
between the rich and poor, relative afflu-
ence for the first time has become one of
the great claimants on world food supplies."40



Need for Research
and Improved Data
In the middle and high-income develop-
ing countries, with fewer foreign exchange
constraints, deficits in cereals for feed could
be met from imports, but in other countries
special measures are needed to meet the
shortage by increasing domestic production
of feedgrains or by substituting other feeds.
Food, feed, and livestock policies, including
incentives and infrastructure development,
need to be reviewed and modified where
necessary. Improvements in yields of feed-
grains and feed efficiency through emphasis
on research and development of feed tech-
nology could also partly alleviate the situ-
ation. Promoting cultivation of feedgrains,
especially coarse grains, can be justified
from an equity standpoint because coarse


38 Food and Agriculture Organization of the United Nations, Changing Patterns and Trends in Feed Utilization
(Rome: FAO, 1983).
3O Sarma and Yeung, Livestock Products in the Third World.
40 Yotopoulos, "Middle-Income Classes and Food Crises."








grains are usually grown on poorer lands
by poor farmers. In addition, the scope for
substitution of noncereals, such as cassava
supplemented by proteins, for cereals in
concentrate feeds needs to be explored. For
example, in some of the developing coun-
tries where cassava is grown, such substi-
tution becomes feasible with higher yields
and lower unit costs of cassava production.
Nevertheless, demand for feeds and fodder
is likely to increase rapidly in the coming
decades, resulting in increased competition
for land. Land will have to be used more
intensively, and more attention will have
to be given to studies on new sources of
feed, greater use of by-products and agricul-
tural wastes, and compound feeds. This
means that more resources will have to be
allocated to research on feeds and fodder at
both national and international levels.
A basic problem in analyzing feed trends
is the lack of reliable data on feed use. It is


therefore important that developing coun-
tries take steps--perhaps in cooperation with
FAO and with their technical assistance
where necessary-to collect and publish
regular data on livestock feed and fodders.
In the case of cereal feeds, two distinct feed
sectors must be included: the first is grains
used in feed manufacturing industries, both
small- and large-scale. It should be relatively
easy to collect these data from feed manufac-
turers' associations or other related organi-
zations. Second and more difficult are data
on grains used directly as feed by farmers
out of their home-grown produce or local
purchases. These estimates have to be based
on household or farm business surveys.
Another area in which available data are
inadequate is the utilization of feeds by type
of livestock output. In view of the rising im-
portance of livestock feed, concerted efforts
are needed to fill these data gaps quickly.









APPENDIX 1:

DATA AND METHODOLOGY


The data on use of cereals for livestock
feed and on livestock numbers and output
used in the study are taken from the inter-
national data base of FAO. The annual data
on cereals used as feed for the years 1966-
80 are from the FAO "Agricultural Supply/
Utilization Accounts Tape" for 1981. Sta-
tistics of livestock numbers for 1980 are
from the FAO Production Yearbook, 1982.
Other data on livestock output are obtained
from the FAO "Production Yearbook Tape,
1981." Projections of output and consump-
tion of livestock products to 1990 and 2000
are based on IFPRI Research Report 49, by
J. S. Sarma and Patrick Yeung, Livestock
Production in the Third World: Past Trends
and Projections to 1990 and 2000, 1985.
The sources of other data, such as trade and
prices, are quoted under their respective
tables.


Past Trends
The methodology approach taken in this
report is first to analyze the past trends of
the variables, which are based on annual
averages for the two five-year periods 1966-
70 (late 1960s) and 1976-80 (late 1970s)
for individual countries or subregions. These
averages are used for absolute measures of
change and relative distributions. The an-
nual growth rates calculated for this period
represent the compound growth rates be-
tween the mid-years of the two quinquennia.
In addition, trend growth rates of cereal
feed are also calculated using the exponen-
tial growth equation fitted to the annual
data on feed use; trend values for 1966 and
1980 are worked out for each country and
subregion. The trend growth rates for the
regions and other aggregates are based on


the aggregate trend values at the beginning
and end of the period, using the formula

Yt = Yo(1 + r)t,

where Yo and Y, are the trend estimates of
the variable in the base period (o) and the
end year (t), and r represents the annual
growth rate of the variable.



Projections of Feed Use
Projections of feed use to 1990 and
2000 are made using five alternative meth-
ods.

Method 1
This method is based on the assumption
that the projected feed use of cereals in
1990/2000 equals the projected output of
livestock products in 1990/2000 multiplied
by the corresponding projected feeding
ratio.41 Output of meat, milk, or eggs is pro-
jected as an extrapolation of 1961-77 trends
of each of these products in each of the 104
developing countries included in the study.
These data are taken from Livestock Prod-
ucts in the Third World at the country level
and have been aggregated at the subregional
level.
The projected outputs are converted into
Livestock Output Units (LOUs), using the
formula: meat + 1/10 of milk + eggs for
each subregion.
Annual values of feeding ratios for the
period 1966-80 are computed, and linear
trends are fitted to them. Projected trend
values of these ratios for 1990 and 2000
are then obtained. The product of the pro-


41 Hereafter, the term feed use relates to cereal feeds only; trends relate to exponential time trends.









jected output of livestock products and the
corresponding feeding ratio gives the pro-
jected feed use.

Method 2
This method assumes that the difference
between the growth rate in feed use and
that in livestock output in each subregion
will continue to 1990 and 2000. The growth
rates implied in the livestock output pro-
jections for 1990/2000 over the 1980 trend
level were first calculated for each subregion.
Then the difference between the 1966-80
trend growth rate of feed use and that of
livestock output in the subregion during the
same period was computed. This difference
was added to the 1980-1990/2000 growth
rate of livestock output, and the adjusted
growth rates in feed use were obtained. The
trend value of feed use in 1980 was pro-
jected to 1990/2000 using the adjusted
growth rate for each subregion.

Rcf 1980-1990/2000 = rcf 1966-80
rip 1966-80 + Rp 1980-1990/2000,


where

Rcf = the adjusted growth rate in feed use
over 1980-1990/2000,
rcf = the growth rate in feed use for 1966-
80,
rp = the growth rate in livestock output for
1966-80, and
Rp = the projected growth rate in livestock
output for 1980-1990/2000.

Projected feed use in 1990/2000
= Trend value of feed use in 1980
x (1 + Rcf)10/20.


This is done for each subregion, and the
data for the 10 subregions are then aggre-
gated.

Method 3
This method assumes that the trend
growth rate for feed use in each subregion
over the 1966-80 period is extrapolated to
1990 and 2000.

Projected feed use in 1990/2000
= 1(subregions) Trend value of
feed use in each subregion in
1980 x (1 + rcf)lo/20

where rcf has the same meaning as above.

Method 4
In this method, trend growth rates in
feed use in each of the 29 countries in which
feed use exceeds 500,000 tons in 1980 are
extrapolated. For the remaining 75 coun-
tries, these are grouped into four classes
according to the level of per capital income,
and the trend growth rate of feed use in
each of the four classes is used in projecting
the feed use. These two sets of figures are
aggregated to give the projected feed use in
1990/2000.42

Method 5
In method 5, the income elasticity of
demand for meat is used as a proxy for that
of cereal feed. Thus the projected feed use
takes into account the rise in population
and per capital income growth; the latter is
based on the continuation of 1966-80 trends
in per capital income.

Projected feed use in 1990/2000
= Trend value of cereal
feed use in 1980
(1 + r)10/20 x (I +ryX qm)10/20,


42 An upper limit of 10 percent has been imposed in extrapolating the individual country growth rates in feed use.









where All these projections assume constant
relative prices. These are computed for each
r = the rate of population growth, country and aggregated for the 104 develop-
ry = the rate of per capital income growth,43 ing countries. This is the method adopted
and in the IFPRI research report Food in the
Tim = the income elasticity of demand for Third World
meat.













































43 A lower limit of 0.5 and an upper limit of 6.0 percentwere imposed on the growth rates of per capital income.






0 APPENDIX 2:

CLASSIFICATION OF COUNTRIES SELECTED FOR THE STUDY


Table 20-Classification of developing countries by per capital income, income growth, and livestock output
growth

1961-80 Growth of GNP
1980 GNP Per Capita Per Capita Growth of Livestock Output
Less 1.0- 3.0- 5 Per- Less 1.0- 3.0- 5Per-
Less 2,000 Than 2.9 4.9 cent Than 2.9 4.9 cent
Than $250- $500- or I Per- Per- Per- or 1 Per- Per- Per- or
Region/Subregion/Country $250 $499 $1,999 More cent cent cent More cent cent cent More








North Africa/Middle East
Northern Africa
Algeria x x x
Egypt x x x
Libya x x x
Morocco x x x
Sudan x x x
Tunisia x x x
Western Asia
Afghanistan x x x
Cyprus x x x
Iran x x x
Iraq x x x
Jordan x x x
Kuwait x x x
Lebanon x x x
Oman x x x
Saudi Arabia xx xx
Syria x x x
Turkey x x x
Yemen Arab Republic x x x
Yemen People's
Democratic Republic x x x
Sub-Saharan Africa
West Africa
Benin x x x
Burkina Faso x x x
Chad x x x
C6te d'lvoire x x x
Gambia x x x
Ghana x x x
Guinea x x x
Guinea-Bissau x x x
Liberia x x x
Mali x x x
Mauritania x x x
Niger x x x
Nigeria x x x
Senegal x x x
Sierra Leone x x x
Togo x x x







Table 20-Continued

1961-80 Growth of GNP
1980 GNPPer Capita Per Capita Growth of Livestock Output
Less 1.0- 3.0- 5 Per- Less 1.0- 3.0- 5 Per-
Less 2,000 Than 2.9 4.9 cent Than 2.9 4.9 cent
Than $250- $500- or I Per- Per- Per- or 1 Per- Per- Per- or
Region/Subregion/Country $250 $499 $1,999 More cent cent cent More cent cent cent More

CentralAfrica
Angola x x x
Burundi x x x
Cameroon x x x
Central African Republic x x
Congo x x x
Gabon x x x
Rwanda x x x
Zaire x x x
Eastern and Southern Africa
Botswana x x x
Ethiopia x x x
Kenya x x x
Lesotho x x x
Madagascar x x x
Malawi x x x
Mauritius x x x
Mozambique x x x
Namibia x x x
Reunion x x x
Somalia x x x
Swaziland x x x
Tanzania x x x
Uganda x x x
Zambia x x x
Zimbabwe x x x
Latin America
Central America and the Caribbean
Costa Rica x x x
Cuba x x x
Dominican Republic x x x
El Salvador x x x
Guatemala x x x








Haiti x x x
Honduras x x x
Jamaica x x x
Mexico x x x
Nicaragua x x
Panama x x x
Trinidad and Tobago x x x
Upper South America
Bolivia x x
Brazil x x x
Colombia x x x
Ecuador x x x
Guyana x x x
Paraguay x x x
Peru x x x
Surinam x x
Venezuela x x
Lower South America
Argentina x x x
Chile x x x
Uruguay x x x

Total 19 19 46 20 27 36 23 18 13 23 44 24

Source: World Bank, "Gross National Product by Country Data Tape, 1961-80," Washington, D.C., 1981; Food and Agriculture Organization of the United
Nations, "Production Yearbook Tape, 1981," Rome, 1982; and United Nations, Department of International Economic and Social Affairs, "World Population
Prospects as Assessed in 1982," New York, 1983 (computer printout).
Notes: The income groups are determined by the GNP per capital calculated in U.S. dollars, based on the 1961-80 trend of real GNP. Livestock output is
expressed as livestock output units obtained by adding meat plus one-tenth of milk plus eggs. Then the countries are classified according to the 1966-70
to 1976-80 average annual growth rate of this livestock output.










APPENDIX 3:

SUPPLEMENTARY TABLES


Table 21-World cereal feed use and total output of cereals by type, 1980

Region/Feed Use/Output Coarse Grains Paddy Wheat Total Cereals

(million metric tons)
World
Feed use 506 32 143 681
Total output 720 399 446 1,565
All study countries
Feeduse 65 17 18 100
Total output 161 232 96 488
Other developing countriesa
Feeduse 48 13 13 74
Total output 82 143 55 280
Developed countries
Feeduse 393 2 112 507
Total output 477 24 295 796

Sources: Food and Agriculture Organization of the United Nations, Production Yearbook, 1982 (Rome: FAO,
1983); and Food and Agriculture Organization of the United Nations, "Agricultural Supply/Utilization
Accounts Tape, 1981," Rome, 1982.
Note: Parts may not add to total due to rounding.
a This category includes China and 24 small developing countries that were excluded from this study.
b Countries are grouped according to the classification system of the Food and Agriculture Organization of the
United Nations.


Table 22-Cereal feed use by type, 1980

Commodity Feed Use Commodity Feed Use
(1,000 meteric tons) (1,000 metric tons)
Wheat 3,985 Rye brana 55
Wheat flour 108 Oats 827
Wheat brana 13,995 Rolled oats 1
Subtotal for wheat 18,088 Millet 1,113
Paddy 2,317 Millet bran' 1,200
Milled rice 274 Sorghum 10,802
Broken rice 17 Sorghum flour 85
Rice bran' 13,972 Sorghum brana 812
Rice bran cakea 525 Buckwheat 30
Subtotal for paddy 17,105 Buckwheat brana 4
Barley 10,396 Canary seed 59
Maize 35,124 Mixed grain 145
Maize flour 3 Grain not elsewhere specified 285
Maize brana 3,879 Flour not elsewhere specified 20
Maize cake' 68 Bran not elsewhere specified" 69
Maize gluten b Subtotal for coarse grains 65,141
Rye 164
Rye flour b Total 100,334

Source: Food and Agriculture Organization of the United Nations, "Agricultural Supply/Utilization Accounts Tape,
1981," Rome, 1982.
aThis is a by-product.
b Less than 500 metric tons.









Table 23-World livestock numbers, 1980

Cattle and Sheep and
Country Group Buffalo Goats Pigs Poultry
(millions)
Worlda 1,325 1,578 795 6,607
Developed economies 425 552 340 3,159
Developing economies 899 1,026 455 3,448
104 study countries 829 840 132 2,578
Asia 367 223 48 829
North Africa/Middle East 62 265 ... 354
Sub-Saharan Africa 131 209 12 433
Latin America 268 143 72 961

Source: Food and Agriculture Organization of the United Nations, Production Yearbook, 1982(Rome: FAO, 1983).
Notes: Parts may not add to total due to rounding. The ellipsis indicates a nil or negligible amount.
'Countries are grouped according to the classification system of the Food and Agriculture Organization of the
United Nations. China is included.
b Countries are grouped according to the classification system of the International Food Policy Research Institute.
China is excluded.


Table 24-Trends in cereal feed use, by region, 1966-70 and 1976-80


Region Coarse Grains Paddy Wheat Total Cereals
(million metric tons)
Asia
1966-70 4.83 11.09 2.92 18.84
1976-80 8.39 14.73 5.20 28.32
Growth rate (percent) 5.68 2.88 5.94 4.16
North Africa/Middle East
1966-70 8.42 0.34 3.81 12.54
1976-80 13.57 0.32 6.66 20.46
Growth rate (percent) 4.88 -0.40 5.73 5.02
Sub-Saharan Africa
1966-70 2.32 0.21 0.23 2.76
1976-80 3.05 0.27 0.43 3.75
Growth rate (percent) 2.76 2.60 6.50 3.10
Latin America
1966-70 19.39 0.77 2.87 23.03
1976-80 31.62 1.30 4.49 37.40
Growth rate (percent) 5.01 5.34 4.56 4.97
Total 104 countries
1966-70 34.96 12.41 9.84 57.21
1976-80 56.63 16.63 16.78 90.03
Growth rate (percent) 4.94 2.97 5.48 4.64

Source: Food and Agriculture Organization of the United Nations, "Agricultural Supply/Utilization Accounts Tape,
1981," Rome, 1982.









Table 25-Cereal feed use of grains and brans and by-products, by region
and commodity, 1966-70 and 1976-80

Grains Brans and By-Products
Region/Commodity 1966-70 1976-80 1966-70 1976-80
(million metric tons)
Asia
Coarse grains 2.77 5.79 2.06 2.60
Wheat 0.25 0.45 2.67 4.75
Paddy 1.52 2.11 9.57 12.62
Total 4.54 8.35 14.30 19.97
North Africa/Middle East
Coarse grains 7.98 13.04 0.44 0.53
Wheat 1.43 2.57 2.38 4.09
Paddy ... .. 0.34 0.32
Total 9.41 15.61 3.16 4.94
Sub-Saharan Africa
Coarse grains 0.89 1.29 1.44 1.76
Wheat ... ... 0.23 0.42
Paddy 0.09 0.11 0.12 0.16
Total 0.98 1.40 1.79 2.34
Latin America
Coarse grains 18.58 30.57 0.80 1.05
Wheat 0.69 1.23 2.18 3.26
Paddy 0.02 0.19 0.75 1.10
Total 19.29 31.99 3.73 5.41
All study countries
Coarse grains 30.22 50.68 4.74 5.94
Wheat 2.37 4.26 7.46 12.52
Paddy 1.62 2.41 10.78 14.21
Total 34.21 57.35 22.98 32.67

Source: Food and Agriculture Organization of the United Nations, "Agricultural Supply/Utilization Accounts Tape,
1981," Rome, 1982.
Note: The ellipses indicate a nil or negligible amount.









Table 26-Cereal feed use and total livestock output, by region and subregion,
1966-70 and 1976-80

Cereal Feed Use Livestock Outpute
Growth Growth
Region/Subregion 1966-70 1976-80 Rate 1966-70 1976-80 Rate
(million metric tons) (percent/ (million metric tons) (percent/
year) year)
Asia 18.84 28.32 4.16 7.73 10.94 3.53
SouthAsia 10.05 13.27 2.82 4.50 6.04 2.99
East and Southeast Asia 8.79 15.05 5.52 3.23 4.90 4.25
North Africa/Middle East 12.57 20.56 5.04 3.88 5.58 3.70
NorthernAfrica 3.11 5.90 6.62 1.41 1.94 3.23
Western Asia 9.46 14.65 4.47 2.47 3.64 3.96
Sub-Saharan Africa 2.76 3.75 3.10 3.42 4.24 2.17
WestAfrica 1.05 1.42 3.07 1.20 1.50 2.24
CentralAfrica 0.20 0.31 4.31 0.24 0.36 4.15
Eastern and Southern Africa 1.51 2.01 2.96 1.98 2.38 1.86
Latin America 23.03 37.40 4.97 13.24 18.59 3.45
Central America 5.10 10.41 7.40 2.72 4.27 4.62
Upper South America 11.41 19.43 5.47 5.82 8.93 4.37
Lower South America 6.53 7.56 1.48 4.71 5.39 1.39
Total 104 countries 57.21 90.03 4.64 28.27 39.35 3.36

Sources: Food and Agriculture Organization of the United Nations, "Agricultural Supply/Utilization Accounts
Tape, 1981," Rome, 1982; and Food and Agriculture Organization of the United Nations, "Production
Yearbook Tape, 1981," Rome, 1982.
a Livestock output is expressed in livestock output units derived as meat + 1/10 milk + eggs.



Table 27-Production of compound feed, selected developing countries,
1974-76 average and 1980

Country 1974-76 Average 1980
(1,000 metric tons)
Argentina 1,100 1,850
Brazil 6,900 11,400
Colombia 290 1,100
Egypt 150 1,640
India 700 1,150
Iran 100 200
Korea, Republic of 1,070 3,500
Malaysia 319 475
Mexico 3,167 4,600
Morocco 80 210
Nigeria 135 400
Peru 320 467
Philippines 747 1,150
SaudiArabia 15 300
Thailand 700 1,100
Tunisia 130 450
Turkey 484 1,449
Venezuela 1,233 1,700
Other developing countries 2,675 6,562
Total developing countries
(excluding China) 20,315 39,703

Source: Food and Agriculture Organization of the United Nations, Changing Patterns and Trends in Feed Utilization
(Rome: FAO, 1983).









Table 28-Income elasticity of demand for cereals and livestock products,
selected countries, 1980 and 2000

Cow's Milk
Country/Year Cereals Beef Mutton Pigmeat Poultry (Whole) Eggs


Brazil
1980
2000
Egypt
1980
2000
India
1980
2000
Indonesia
1980
2000
Kenya
1980
2000
Korea, Republic of
1980
2000
Malaysia
1980
2000
Mexico
1980
2000
Nigeria
1980
2000
Philippines
1980
2000
Thailand
1980
2000
Turkey
1980
2000


0.15
0.05

0.04
0.02

0.25
0.23

0.29
-0.40

0.35
0.02

0.09
-0.04

0.14
-0.21

-0.10
-0.10

0.17
0.15

0.22
0.20

0.06
-0.13

-0.05
-0.47


0.58 0.29
0.45 0.25

0.80 0.90
0.80 0.90

1.20 1.20
1.20 1.20

1.50 1.60
1.50 1.60

1.00 0.90
1.00 0.90

0.80 1.00
0.42 1.00

0.49 0.32
0.49 0.32

0.59 0.69
0.50 0.57

1.20 1.30
1.20 1.30

1.20 0.90
1.20 0.90

0.56 0.50
0.41 0.50

0.80 1.00
0.80 1.00


Source: Food and Agriculture Organization of the United Nations, "Parameters of the Demand Functions, Fifth
Run," Rome, April, 1978 (computer printout).









Table 29-Ratio of foodgrains to food and feedgrains, selected countries,
1980 to 2000

Country 1980 1985 1990 1995 2000

Brazil 0.43 0.40 0.38 0.35 0.33
Egypt 0.74 0.71 0.67 0.64 0.60
India 0.90 0.89 0.89 0.88 0.87
Indonesia 0.91 0.89 0.85 0.79 0.71
Kenya 0.83 0.82 0.80 0.79 0.77
Korea, Republic of 0.71 0.67 0.64 0.61 0.58
Malaysia 0.66 0.63 0.59 0.55 0.50
Mexico 0.52 0.49 0.46 0.43 0.40
Nigeria 0.93 0.92 0.90 0.88 0.86
Philippines 0.80 0.78 0.77 0.75 0.74
Thailand 0.77 0.75 0.73 0.71 0.69
Turkey 0.47 0.43 0.37 0.32 0.27

Source: Calculated from computer printouts used in Leonardo A. Paulino, Food in the Third World: Past Trends
and Projectons to 2000, Research Report 52 (Washington, D.C.: International Food Policy Research
Institute, 1986).



Table 30-Selected developing countries with cereal feed use exceeding
500,000 metric tons, 1980

Country Feed Use Country Feed Use
(million metric tons) (million metric tons)
Algeria 1.06 Malaysia 1.09
Argentina 5.10 Mexico 9.52
Bangladesh 1.73 Morocco 1.41
Brazil 19.37 Nigeria 0.67
Burma 0.98 Pakistan 1.41
Chile 1.22 Peru 0.75
Colombia 1.24 Philippines 1.72
Cuba 1.03 Saudi Arabia 0.98
Egypt 3.13 Syria 1.30
India 9.50 Thailand 2.71
Indonesia 2.48 Tunisia 0.57
Iran 2.78 Turkey 8.81
Iraq 0.74 Venezuela 2.08
Korea, Democratic Vietnam 1.02
People's Republic of 2.58 Total for 29 countries 90.40
Korea, Republic of 3.42 Total for 104 countries 100.33

Source: Food and Agriculture Organization of the United Nations, "Agricultural Supply/Utilization Accounts Tape,
1981," Rome, 1982.









Table 31-Cereal feed use projections under method 1, 1990 and 2000

TrendValue Projected Cereal Feed Use
Region/Subregion 1980 1990 2000
(million metric tons)
Asia 30.69 44.33 68.45
South Asia 14.10 16.67 20.92
East and Southeast Asia 16.59 27.66 47.53
North Africa/Middle East 21.70 33.85 56.33
Northern Africa 6.60 11.80 21.19
Western Asia 15.10 22.05 35.14
Sub-Saharan Africa 4.03 5.77 8.49
WestAfrica 1.50 1.77 2.46
CentralAfrica 0.33 0.55 0.89
Eastern and Southern Africa 2.20 3.45 5.14
Latin America 41.66 66.69 111.63
Central America 12.32 22.36 41.79
Upper South America 21.19 34.99 58.94
Lower South America 8.15 9.34 10.90
Total 98.08 150.64 244.89

Sources: Food and Agriculture Organization of the United Nations, "Production Yearbook Tape, 1981," Rome,
1982; Food and Agriculture Organization of the United Nations, "Agricultural Supply/Utilization Accounts
Tape, 1981," Rome, 1982; and J. S. Sarma and Patrick Yeung, Livestock Products in the Third World:
Past Trends and Projections to 1990 and 2000, Research Report 49 (Washington, D.C.: International
Food Policy Research Institute, 1985).
Notes: Method 1 is based on projected values of feeding ratios in 1990 and 2000 applied to projected output
of livestock products expressed in terms of livestock output units, which are obtained by adding projections
for meat + 1/10 milk + eggs. Parts may not add to total because of rounding.








BIBLIOGRAPHY

China, Republic of, Council for Agricultural Planning and Development. "Taiwan Food
Balance Sheets 1935-1980." Taiwan, Republic of China, July 1981.
Food and Agriculture Organization of the United Nations. "Agricultural Supply/Utilization
Accounts Tape, 1981." Rome, 1982.
Agriculture: Toward 2000. Rome, FAO, 1981.
Changing Patterns and Trends in Feed Utilization. Rome: FAO, 1983.
Food Balance Sheets, 1975-77 Average and Per Caput Food Supplies, 1961-65
Average, 1967-77. Rome: FAO, 1980.
Food Balance Sheets, 1979-81 Average. Rome: FAO, 1983.
Livestock Development in Developing Countries and Implicationsfor Consumption
and Trade of Feeds. CCP document GR 82/3.
"Producer Prices Tape, 1985." Rome, 1986.
"Production Yearbook Data Tape, 1981." Rome, 1982.
"Production Yearbook Data Tape, 1983." Rome, 1984.
Hrabovszky, J. P. "Feeding of Draft Animals." World Animal Review 46 (April-June 1983):
47-48.
Mellor, John W. "Third World Development: Food, Employment and Growth Interactions."
American Journal of Agricultural Economics 64 (May 1982): 304-311.
Mellor, John W., and Johnston, Bruce F. "The World Food Equation: Interrelations Among
Development, Employment and Food Consumption." Journal ofEconomic Literature
22 (June 1984): 531-574.
Paulino, Leonardo A. Food in the Third World: Past Trends and Projections to 2000.
Research Report 52. Washington, D.C.: International Food Policy Research Institute,
1986.
Regalado, B. "The Distributional Impact of Food Policies on Human Nutrition in the Less
Developed Countries-The Case of the Philippines." M.S. thesis, University of the
Philippines at Los Bafios, 1984.
Sarma, J. S., and Yeung, Patrick. Livestock Products in the Third World-Past Trends and
Projections to 1990 and 2000. Research Report 49. Washington, D.C.: International
Food Policy Research Institute, 1985.
United Nations, Department of International Economic and Social Affairs. "World Population
Prospects as Assessed in 1982." New York, 1983 (computer printout).
Unnevehr, Laurian J., and Nelson, Gerald C. "Structural Transformation in Philippine
Livestock and Corn Markets." Los Bafios, January 1986 (mimeographed).
Wheeler, R. O., et al. The World Livestock Product Feedstuff and Food Grain System.
Morrilton, Arkansas: Winrock International, 1981.
Winrock International. The Role of Ruminants in Support of Man. Morrilton, Arkansas:
Winrock International, 1981.








World Bank. "Gross National Product by Country Data Tape, 1961-80." Washington, D.C.,
1981.
Yotopoulos, Pan A. "Middle-Income Classes and Food Crises: The 'New' Food-Feed Com-
petition." EconomicDevelopmentand Cultural Change33 (No. 3,1985): 463-483.









IFPRI RESEARCH REPORTS (continued)
30 INSTABILITY IN INDIAN FOODGRAIN PRODUCTION, May 1982, by Peter B. R. Hazell
29 GOVERNMENT POLICY AND FOOD IMPORTS: THE CASE OF WHEAT IN EGYPT, December 1981, by Grant M.
Scobie
28 GROWTH AND EQUITY: POLICIES AND IMPLEMENTATION IN INDIAN AGRICULTURE, November 1981, by
J. S. Sarma
27 AGRICULTURAL PRICE POLICIES UNDER COMPLEX SOCIOECONOMIC AND NATURAL CONSTRAINTS: THE
CASE OFBANGLADESH, October 1981, by Raisuddin Ahmed
26 FOOD SECURITY IN THE SAHEL: VARIABLE IMPORT LEVY, GRAIN RESERVES, AND FOREIGN EXCHANGE
ASSISTANCE, September 1981, by John Mclntire
25 INSTABILITY ININDIANAGRICULTURE IN THE CONTEXT OF THE NEW TECHNOLOGY, July 1981, by Shakuntla
Mehra
24 THE EFFECTS OF EXCHANGE RATES AND COMMERCIAL POLICY ON AGRICULTURAL INCENTIVES IN
COLOMBIA: 1953-1978, June 1981, by Jorge Garcia Garcia
23 GOVERNMENT EXPENDITURES ON AGRICULTURE IN LATIN AMERICA, May 1981, by Victor J. Elias
22 ESTIMATES OF SOVIET GRAIN IMPORTS IN 1980-85: ALTERNATIVE APPROACHES, February 1981, by Padma
Desai
21 AGRICULTURAL PROTECTION IN OECD COUNTRIES: ITS COST TO LESS-DEVELOPED COUNTRIES, December
1980, by Alberto Vald6s and Joachim Zietz
20 IMPACT OF IRRIGATION AND LABOR AVAILABILITY ON MULTIPLE CROPPING: A CASE STUDY OF INDIA,
November 1980, by Dharm Narain and Shyamal Roy
19 A COMPARATIVE STUDY OFFAO AND USDA DATA ON PRODUCTION, AREA, AND TRADE OFMAJOR FOOD
STAPLES, October 1980, by Leonardo A. Paulino and Shen Sheng Tseng
18 THEECONOMICSOFTHEINTERNATIONAL STOCKHOLDING OF WHEAT, September 1980, by Daniel T. Morrow
17 AGRICULTURAL RESEARCH POLICY IN NIGERIA, August 1980, by Francis Sulemanu Idachaba
16 A REVIEW OF CHINESE AGRICULTURAL STATISTICS, 1949-79, July 1980, by Bruce Stone
15 FOOD PRODUCTION IN THE PEOPLE'S REPUBLIC OF CHINA, May 1980, by Anthony M. Tang and Bruce Stone
14 DEVELOPED-COUNTRY AGRICULTURAL POLICIES AND DEVELOPING COUNTRY SUPPLIES: THE CASE OF
WHEAT, March 1980, by Timothy Josling
13 THE IMPACT OF PUBLIC FOODGRAIN DISTRIBUTION ON FOOD CONSUMPTION AND WELFARE IN SRI
LANKA, December 1979, by James D. Gavan and Indrani Sri Chandrasekera
12 TWO ANALYSES OF INDIAN FOODGRAIN PRODUCTION AND CONSUMPTION DATA, November 1979, by
J. S. Sarma and Shyamal Roy and by P. S. George
11 RAPID FOOD PRODUCTION GROWTH IN SELECTED DEVELOPING COUNTRIES: A COMPARATIVE ANALYSIS
OF UNDERLYING TRENDS, 1961-76, October 1979, by Kenneth L. Bachman and Leonardo A. Paulino
10 INVESTMENT AND INPUT REQUIREMENTS FOR ACCELERATING FOOD PRODUCTION IN LOW-INCOME
COUNTRIES BY 1990, September 1979, by Peter Oram, Juan Zapata, George Alibaruho, and Shyamal Roy
9 BRAZIL'S MINIMUM PRICE POLICY AND THE AGRICULTURAL SECTOR OF NORTHEAST BRAZIL, June 1979,
by Roger Fox
8 FOODGRAIN SUPPLY, DISTRIBUTION, AND CONSUMPTIONPOLICIES WITHIN DUAL PRICING MECHANISM:
A CASE STUDY OF BANGLADESH, May 1979, by Raisuddin Ahmed
7 PUBLIC DISTRIBUTION OF FOODGRAINS IN KERALA-INCOME DISTRIBUTION IMPLICATIONS AND EFFEC
TIVENESS, March 1979, by P. S. George
6 INTERSECTORAL FACTOR MOBILITY AND AGRICULTURAL GROWTH, February 1979, by Yair Mundlak
5 IMPACT OF SUBSIDIZED RICE ON FOOD CONSUMPTION AND NUTRITION IN KERALA, January 1979, by
Shubh K. Kumar
4 FOOD SECURITY: AN INSURANCE APPROACH, September 1978, by Panos Konandreas, Barbara Huddleston, and
Virabongsa Ramangkura
3 FOOD NEEDS OF DEVELOPING COUNTRIES: PROJECTIONS OF PRODUCTION AND CONSUMPTION TO
1990, December 1977
2 RECENT AND PROSPECTIVE DEVELOPMENTS IN FOOD CONSUMPTION: SOME POLICY ISSUES, July 1977
1 MEETING FOOD NEEDS IN THE DEVELOPING WORLD: LOCATION AND MAGNITUDE OF THE TASKIN THE
NEXT DECADE, February 1976




J. S. Sarma came to IFPRI in 1978 as a visiting research fellow. Since
1980 he has been a research fellow on IFPRI's staff.





IFPRI RESEARCH REPORTS
56 THE EFFECTS OF TRADE AND EXCHANGE RATE POLICIES ON AGRICULTURE IN ZAIRE, November 1986, by
Tshikala B. Tshibaka
55 THE EFFECTS OF TRADE AND EXCHANGE RATE POLICIES ON AGRICULTURE IN NIGERIA, October 1986, by
T. Ademola Oyejide
- 54 WEATHER AND GRAIN YIELDS IN THE SOVIET UNION, September 1986, by Padma Desai
53 REGIONAL COOPERATION TO IMPROVE FOOD SECURITY IN SOUTHERN AND EASTERN AFRICAN COUN-
TRIES, July 1986, by Ulrich Koester
52 FOOD IN THE THIRD WORLD: PAST TRENDS AND PROJECTIONS TO 2000, June 1986, by Leonardo A. Paulino
51 DETERMINANTS OF AGRICULTURAL POLICIES IN THE UNITED STATES AND THE EUROPEAN COMMUNITY
November 1985, by Michel Petit
50 GOVERNMENTEXPENDITURESONAGRICULTUREAND AGRICULTURAL GROWTHINLATINAMERICA, October
1985, by Victor J. Elias
49 LIVESTOCK PRODUCTS IN THE THIRD WORLD: PAST TRENDS AND PROJECTIONS TO 1990 AND 2000, April
1985, byJ. S. Sarma and Patrick Yeung
48 RURAL HOUSEHOLD USE OF SERVICES: A STUDY OF MIRYALGUDA TALUKA, INDIA, March 1985, by Sudhir
Wanmali
47 EVOLVING FOOD GAPS IN THE MIDDLE EAST/NORTH AFRICA: PROSPECTS AND POLICY IMPLICATIONS,
December 1984, by Nabil Khaldi
46 THE EFFECTS ON INCOME DISTRIBUTION AND NUTRITION OF ALTERNATIVE RICE PRICE POLICIES IN
THAILAND, November 1984, by Prasarn Trairatvorakul
45 THE EFFECTS OF THE EGYPTIAN FOOD RATION AND SUBSIDY SYSTEM ON INCOME DISTRIBUTION AND
CONSUMPTION, July 1984, by Harold Alderman and Joachim von Braun
44 CONSTRAINTS ON KENYA'S FOOD AND BEVERAGE EXPORTS, April 1984, by Michael Schluter
43 CLOSING THE CEREALS GAP WITH TRADE AND FOOD AID, January 1984, by Barbara Huddleston
42 THE EFFECTS OF FOOD PRICE AND SUBSIDY POLICIES ON EGYPTIAN AGRICULTURE, November 1983, by
Joachim von Braun and Hartwig de Haen
41 RURAL GROWTH LINKAGES: HOUSEHOLD EXPENDITURE PATTERNS INMALAYSIA AND NIGERIA, September
1983, by Peter B. R. Hazell and Ailsa Roell
40 FOOD SUBSIDIES IN EGYPT: THEIR IMPACT ON FOREIGN EXCHANGE AND TRADE, August 1983, by Grant
M. Scobie
39 THE WORLD RICE MARKET: STRUCTURE, CONDUCT, AND PERFORMANCE, June 1983, by Ammar Siamwalla
and Stephen Haykin
38 POLICY MODELING OF A DUAL GRAIN MARKET: THE CASE OF WHEAT IN INDIA, May 1983, by Raj Krishna
and Ajay Chhibber
37 SERVICE PROVISION AND RURAL DEVELOPMENT IN INDIA: A STUDY OF MIRYALGUDA TALUKA, February
1983, by Sudhir Wanmali
36 AGRICULTURE AND ECONOMIC GROWTH IN AN OPEN ECONOMY: THE CASE OF ARGENTINA, December
1982, by Domingo Cavallo and Yair Mundlak
35 POLICY OPTIONS FOR THE GRAIN ECONOMY OF THE EUROPEAN COMMUNITY: IMPLICATIONS FOR
DEVELOPING COUNTRIES, November 1982, by Ulrich Koester
34 EGYPTS FOOD SUBSIDY AND RATIONING SYSTEM: A DESCRIPTION, October 1982, by Harold Alderman,
Joachim von Braun, and Sakr Ahmed Sakr
33 AGRICULTURAL GROWTH AND INDUSTRIAL PERFORMANCE IN INDIA, October 1982, by C. Rangarajan
32 FOOD CONSUMPTION PARAMETERS FOR BRAZIL AND THEIR APPLICATION TO FOOD POLICY, September
1982, by Cheryl Williamson Gray
31 SUSTAINING RAPID GROWTHNINDIA'S FERTILIZER CONSUMPTION: A PERSPECTIVEBASED ON COMPOSITON
OF USE, August 1982, by Gunvant M. Desai
(continued on inside back cover)


Wahigtn D.C 20036 US




University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs