Global Farming Systems Study: Challenges and Priorities to 2030 SYNTHESIS AND GLOBAL OVERVIEW
The Global Farming Systems Study was conducted by FAO under the overall
coordination of S. Funes (Director, Rural Development Division) and the technical
leadership of J. Dixon (Senior Officer, Farming Systems, Farm Management and
Production Economics Service, Agricultural Support Systems Division) and
A. Gulliver (Economist, Investment Centre Division).The Study benefited from
the guidance of D. Forbes Watt (Director, Investment Centre Division), J. Monyo
(Director, Agricultural Support Systems Division), D. Baker (Chief, Farm
Management and Production Economics Service,AGS) and A. MacMillan (Principal
Adviser, Project Advisory Unit,TCI) in FAO and of C. Csaki (Senior Advisor/Team
Leader-Rural Strategy) and S. Barghouti (Research Advisor) of the Rural
Development Department,World Bank.
As part of this review,......
Global Farming Systems Study:
Challenges and
Priorities to 2030
S Y N T H E S I S A N D G L O B A L OV E RV I E W
J. Dixon, A. Gulliver and D. Gibbon
The Global Farming Systems Study was conducted by FAO under the overall
coordination of S. Funes (Director, Rural Development Division) and the technical
leadership of J. Dixon (Senior Officer, Farming Systems, Farm Management and
Production Economics Service, Agricultural Support Systems Division) and
A. Gulliver (Economist, Investment Centre Division).The Study benefited from
the guidance of D. Forbes Watt (Director, Investment Centre Division), J. Monyo
(Director, Agricultural Support Systems Division), D. Baker (Chief, Farm
Management and Production Economics Service, AGS) and A. MacMillan (Principal
Adviser, Project Advisory Unit,TCI) in FAO and of C. Csaki (Senior Advisor/Team
Leader-Rural Strategy) and S. Barghouti (Research Advisor) of the Rural
Development Department, World Bank.
© FAO 2001
ii
Preface
For more than a decade, the proportion of internation- sary to re-estimate and re-analyse a wide variety of
ally supported public investment directed at agricul- data relating to system characteristics, including
ture and the rural sector in developing countries has physical, social, economic, demographic and environ-
been declining. In the year 2000, World Bank commit- mental parameters. This analysis provided the neces-
ments to the rural sector reached their lowest ever sary quantitative underpinning for the central, quali-
levels, measured as a proportion of their total lending tative, task of developing expert judgements on the
portfolio. Moreover, this has occurred at a time when future evolution of farming systems and their devel-
the process of globalisation is bringing about profound opmental priorities. In all, the study encompassed the
changes in patterns of trade and investment, placing contributions of over 40 specialists in a range of disci-
agricultural producers and rural communities, more plines, both within and outside of FAO, and took
generally, under tremendous pressure to adapt to into account comments from many others. The whole
changing circumstances. Nor is there any evidence of exercise was completed in just over six months.
significant progress in reducing the incidence of Although any specific farming system embraces
hunger. In order to reinvigorate its efforts aimed at considerable heterogeneity, the diagnosis of the
poverty reduction and sustainable growth among rural dynamics, constraints and opportunities of typical
populations, the World Bank initiated in 2000 a review farm households contributes to the identification of
of its rural development strategy1. interventions to improve system performance and sus-
As part of this review, the World Bank sought the tainability. Therefore, the farming systems presented
assistance of the Food and Agriculture Organization of in this study are considered to provide an effective
the United Nations (FAO) in evaluating how farming broad framework for the prioritisation of develop-
systems might change and adapt over the next thirty ment actions and investments for accelerating agricul-
years. Amongst other objectives, the World Bank tural development, particularly in ways which can
asked FAO to provide guidance on priorities for reduce rural poverty and hunger.
investment in food security, poverty reduction, and The results of the study are summarized in a set of
economic growth, and in particular to identify promis- seven documents, comprising six regional reports and
ing approaches and technologies that will contribute the global overview contained in this volume. This
to these goals. The identification of future changes document, which synthesises the results of the six
affecting farming systems relied heavily on work regional analyses as well as discussing global trends,
undertaken in FAO over many years in monitoring cross-cutting issues and possible implementation
trends affecting agricultural production and assessing modalities, presents an overview of the complete
their likely implications for future output, productivi- study. This document is supplemented by two case
ty and nutrition levels.2 study reports of development issues of importance to
The global study commenced with the delineation farming systems globally.
and characterisation of about 70 major farming
systems encompassing all developing regions of the S. Funes,
world. As existing data systems are based, almost Director,
without exception, on national and sub-national Rural Development Division,
administrative areas, while farming systems cross Sustainable Development Department,
national and even regional boundaries, it was neces- Food and Agriculture Organization.
1
“Rural Development: From Vision to Action”. World Bank, Washington D.C., 1997.
2
Most recently in “Agriculture:Towards 2015/30.Technical Interim Report” Global Perspective Studies Unit, FAO, Rome, April, 2000.
iii
John Dixon is Senior Farming Systems Officer, Farm Management and
Production Economics Service, Agricultural Support Systems Division, FAO,
Rome, Italy.
Aidan Gulliver is an Agricultural Economist with the Project Advisory Unit,
Investment Centre Division, FAO, Rome, Italy.
David Gibbon is a Farming Systems Consultant, Sidmouth, Devon, UK.
The Study benefited from discussions at a series of FAO seminars, working
sessions and video conferences. In particular, the contributions and edition
of A. Carloni, F. Dauphin, A. MacMillan, and J. Weatherhogg (Investment
Centre), E. Kueneman (Agriculture Department), N. Nguyen and D. Tran
(Plant Production and Protection Division), J.Bruinsma (Global Perspectives
Studies Unit), S. Tanic (Subregional Office for Central and Eastern Europe)
and D. Ivory (Regional Office for Asia and the Pacific), R. Brinkman and
M. Hall (consultants). Review and comments were contributed by S. Barghouti
and C. Csaki (World Bank), S. Barraclough and B. Roitman (consultants), as
well as by numerous FAO staff including T. Bachmann, D. Baker, L. Clarke,
R. Florin, P. Koohafkan,S. Mack, J. Maki-Hokkonen, F. Moukoko-NÕDoumbe,
F. Nachtergaele, M. Porto, J. Poulisse, R. Roberts, P. Santacoloma,
A. Shepherd, J. Slingenbergh and N. Urquia (Agriculture Department), F. Egal,
Y. Lambrou, K. Stamoulis and J. Smidthuber (Economic and Social
Department), M. Gauthier and K. Warner (Forestry Department), L. Collette,
J. Dey-Abbas, P. Groppo, A. Herrera, J. Juhasz, J. Latham, P. Munro-Faure
and D. Palmer (Sustainable Development Department) and C. Bevan,
G. Evers, T. Tecle and M. Wales (Technical Cooperation Department). Any
remaining errors are the responsibility of the Study Team.
The Study Data and GIS Team, responsible for generating the farming
systems specific data and developing the GIS-based maps used in the
study, was led by C. Auricht (consultant) with the support of P. Aguilar
(WAICENT/FAOSTAT Data Management Branch), M. Zanetti (GIS Unit), L. Hein
(Investment Centre), G. Agostini, S. Accongiagico, M. Lespine and T. Rossetti
(consultants).
iv
Table of Contents
PREFACE iii
1 INTRODUCTION 1
Study Purpose 1
Poverty and Agricultural Development 1
The Concept of Farming Systems 4
Delineation of Major Farming Systems 6
Evolution of Farming Systems 8
Factors Influencing Farming Systems Development 8
Study Structure and Format 9
2 GLOBAL FACTORS INFLUENCING THE EVOLUTION
OF FARMING SYSTEMS 10
Natural Resources and Climate 10
Science and Technology 12
Globalisation and Market Development 14
Policies, Institutions and Public Goods 15
Information and Human Resources 16
3 SUB-SAHARAN AFRICA REGION 18
Introduction 18
Characteristics of the Major Regional Farming Systems 18
Regional Strategic Priorities 22
Map: Major Farming Systems 25
4 MIDDLE EAST AND NORTH AFRICA REGION 27
Introduction 27
Characteristics of the Major Regional Farming Systems 28
Regional Strategic Priorities 30
Map: Major Farming Systems 33
5 EASTERN EUROPE AND CENTRAL ASIA 35
Introduction 35
Characteristics of the Major Regional Farming Systems 35
Regional Strategic Priorities 39
Map: Major Farming Systems 43
6 SOUTH ASIA REGION 47
Introduction 47
Characteristics of the Major Regional Farming Systems 47
Regional Strategic Priorities 52
Map: Major Farming Systems 57
7 EAST ASIA AND PACIFIC REGION 59
Introduction 59
Characteristics of the Major Regional Farming Systems 61
Regional Strategic Priorities 65
Map: Major Farming Systems 69
v
8 LATIN AMERICA AND THE CARIBBEAN 71
Introduction 71
Characteristics of the Major Regional Farming Systems 72
Regional Strategic Priorities 75
Map: Major Farming Systems 79
9 GLOBAL CHALLENGES AND PRIORITIES 81
The Challenge of Contrasting Farm Characteristics 81
Global Challenges and Priorities for Coming Decades 83
Achieving Sustainable and Productive Use of
Natural Resources 83
Deploying Science and Technology 84
Exploiting Globalisation and Market Development 87
Refocusing Policies, Institutions and Public Goods 89
Enhancing Agricultural Information and Human Capital 92
10 SOME OPERATIONAL IMPLICATIONS 94
Demand-driven Approaches to Integrated Rural Development 94
Support Services and Related Institutions 96
Financing Instruments 96
Assessing Impact using Farming Systems Frameworks 98
vi
1
Introduction
STUDY PURPOSE division of the Bank has been given primary responsi-
bility for developing its own regional strategy, and a
In 1997, the World Bank issued a statement of its number of supporting studies have been commis-
global strategy for rural development entitled “Rural sioned – largely from thematic groups within the Bank
Development: From Vision to Action”. Subsequently, – to provide technical inputs to the strategy formula-
important improvements in the performance of the tion. In this context, FAO was invited to collaborate in
rural portfolio have been achieved. These include preparing a supporting study with the following
increases in quality-at-entry of rural projects, reduc- objective:
tions in the proportion of projects at risk and an
improved impact from supervision. However, the “On the basis of a determination of the principal
strategy does not appear to have achieved its key trends and issues affecting major farming
objective of reversing the declining trend in rural systems in each World Bank region over the
investment volumes within the Bank. In financial year next 30 years, propose operational strategies,
(FY) 2000, a historical low of only 38 rural develop- approaches and technologies that will contribute
ment projects were approved world-wide by the Bank, to significant and sustainable rural development
with a total value of US$1.5 billion – in comparison and poverty reduction among farming system
with US$4 billion in FY97. This is equivalent to only participants.”
10 percent of new loan commitments by value 3. Of
further concern is the assessment by the Operations
Evaluation Division of the Bank during FY99 that POVERTY AND AGRICULTURAL
only 37 percent of agricultural projects and 42 DEVELOPMENT
percent of all rural projects were sustainable. Finally,
important changes in world-wide economic, institu- Recent World Bank activities have been categorised
tional and political conditions have occurred since the according to final beneficiary location in urban space
preparation of Vision to Action and these now need to or rural space. Of an estimated total population in
be taken into account in looking at future operations. developing countries of approximately 5.1 billion in
With poverty reduction still the central goal of 1999, 3.0 billion reside in rural areas. Of these some
the World Bank and considering that a majority of 80 percent, or 2.5 billion people, are members of agri-
the world’s poor are dependent on agriculture, the cultural households 4 – including farming, pastoral,
Rural Development Strategy remains an important fishing and forestry households (see Table 1-1).
document, but needs to be updated. The revised Women constitute 44 percent of the 1.3 billion
Strategy will be more action-oriented and will have a persons in the agricultural labour force of developing
stronger regional focus than previously. Each regional countries. Despite the trend towards urbanisation, the
3
Cees de Haan and Sanjiva Cooke, 2000. FY00 Report on the Rural Portfolio. Unpublished Bank document. August 2000.
4
FAOSTAT, 2000. FAO, Rome.
1
Table 1-1: Distribution of Rural and Agricultural Populations in Developing Countries
Total Rural Agric. Agric Popn. Economically Econ. Active Female Econ.
Region Population Population Population as % of Active in in Agric. as Active as %
(million) (million) (million) Total Popn. Agriculture % of Total of Econ Active
(million) Econ. Active in Agriculture
Sub-Saharan
Africa 626 417 384 61% 176 63% 47%
Middle East/
North Africa 323 138 99 31% 35 31% 44%
E. Europe/
Central Asia 478 154 86 18% 47 36% 44%
South Asia 1 325 955 742 56% 345 59% 39%
East Asia/
Pacific 1 836 1 184 1 119 61% 654 63% 47%
Latin America/
Caribbean 505 126 110 22% 44 21% 17%
Developing
World 5 093 2 974 2 540 50% 1 300 53% 44%
Note: 1999 national statistics as reported in FAOSTAT.The definition of rural varies by country, but is often a residual after urban population numbers are
extracted. Agricultural population is usually defined as individuals employed in agriculture, fishery, forestry and hunting and their non-working dependents.The
definition of developing regions follows World Development Report 2000/2001, with the exception that Turkey is included in Eastern Europe and Central Asia.
absolute number of people living in rural areas world- this period, the number of poor decreased in East Asia
wide is growing at nearly 1 percent per annum.5 and the Pacific (strongly influenced by China) and the
It is estimated that, world-wide, 1.2 billion people 6 Middle East and North Africa region. In contrast, the
live in poverty (i.e. consume less than US$1 per day number of poor people has increased in the South
per capita) and that 790 million are under-nour- Asia, Latin America and the Caribbean, Eastern Europe
ished 7. The distribution of poor people between and Central Asia regions, with an especially large and
regions is shown in Table 1-2, for both 1987 and 1998. disturbing rise in Sub-Saharan Africa. The total popu-
A majority of the poor are found in South Asia, East lation of poor in developing countries changed little
Asia and the Pacific and Sub-Saharan Africa. During during this period.
Table 1-2: Distribution of the Poor between Developing Regions (millions)
Developing Region 1987 1998
Sub-Saharan Africa 217 291
Middle East and North Africa 9 6
Eastern Europe and Central Asia 1 24
South Asia 474 522
East Asia and Pacific 418 278
Latin America and Caribbean 64 78
All Developing Countries 1 183 1 199
Source: World Bank (1999)
5
World Bank, 2000. World Development Indicators 2000.Table 3.1.
6
World Bank, 2000. World Development Report 2000.
7
FAOSTAT, 2000. FAO, Rome.
2
Table 1-3: Rural/Urban Poverty Indicators for Selected Developing Countries
Population below Population below
Region/Country national poverty line (%) Region/Country national poverty line (%)
Rural Urban Rural Urban
Sub-Saharan Africa South Asia
Ghana 34.3 26.7 Bangladesh 46.0 23.3
Kenya 46.4 29.3 India 43.5 33.7
Nigeria 49.5 31.7 Nepal 44.0 23.0
Madagascar 77.0 47.0 Pakistan 36.9 28.0
Middle East/North Africa East Asia/Pacific
Algeria 16.6 7.3 China 7.9 reduction. The form that this growth takes, however, The functioning of a farm system is strongly influ-
it expected to have a bearing on its input on poverty enced by the external rural environment, including
reduction. Thus growth in the output in productivity markets, policy and information linkages. Not only
of small farms can be expected to have a broader are farms closely linked to the non-farm economy
effect on poverty reduction than growth on large through commodity and labour markets, but also the
mechanised holdings. It should be noted, however, rural and urban economies are strongly interdepen-
that agricultural growth alone may not be sufficient to dent. For example, it is quite common for small farm
achieve inclusive food security in most rural commu- households to derive 40 percent or more of their
nities, needing to be complemented by measures income from off-farm activities. Farm women and
which lead to broader access to food. Evidence from men are also linked to rural communities and social
empirical research also suggests that the provision of networks, and this social capital influences the man-
public goods, especially research, extension and edu- agement of farms.
cation play a central role in this process. The perfor- For the purposes of this study, farming systems
mance of the public sector is thus extremely important are defined as populations of farms that have broadly
in determining the rate and distribution of agricultur- similar resource bases, enterprise patterns, household
al growth. livelihoods and constraints, and for which similar
The challenge for developing countries is to development strategies and interventions would be
identify specific agricultural and rural needs, and to appropriate. The biophysical, economic and human
focus investment in areas where the greatest impact elements of a farm are interdependent, and thus
on food security and poverty will be achieved. This is farms can be analysed as systems from various points
made possible through developing an understanding of view. Although smallholder farms are more
of the local factors and linkages found in the wide numerous than large commercial or co-operative
range of rural locations subject to differing socio- farms, the latter provide livelihoods for a significant
economic and ecological conditions. Within this proportion of the rural population in some regions.
process, it is also extremely helpful to be able to group Regardless of their size, individual farm systems are
locations with similar development constraints and organised to produce food and meet other goals
investment opportunities. The analysis of farming through the management of available resources –
systems facilitates both the identification and the whether owned, rented or jointly managed – within
planning stages of the process. the existing social, economic and institutional envi-
ronment. They often consist of inter-dependent pro-
duction and gathering components concerned with
THE CONCEPT OF FARMING SYSTEMS crops, livestock, trees and fish farming. Thus, in this
study, farm activities and household livelihoods
Farmers typically view their farms, whether large cor- embrace fishing, pastoralism, farm forestry, hunting
porations or small subsistence units, as systems. The and gathering, as well as cropping and intensive
following diagram of a Bangladesh farm system drawn animal husbandry. Non-farm income, which makes a
by farmers (see Figure 1-1) illustrates the structural significant contribution to the livelihoods of many
complexity and interrelationships between various poor rural families, is also considered. Farm systems
components of a farm system. The diagram shows the are not only found in rural areas: there is a growing
multiple natural resources available to farm women realisation of the magnitude of the urban agriculture
and men, which often include different types of land, in many cities and towns in developing countries.
various sources of water and access to common Refugees and the landless, however, are excluded in
property resources including ponds, grazing areas this Study, although the impact of farming systems
and forest. To these natural resources may be added development on these categories of poor will be
climate and biodiversity. The resource endowment of flagged.11
any particular farm depends, inter alia, on population The analysis of farming systems constitutes the
density, inter-family distribution of resources and the core of the Farming Systems Approach, which diag-
effectiveness of institutions in determining access to noses constraints and identifies opportunities and
resources. strategic priorities for rural development. The power
11
It is understood that the landless will be covered under another study of non-farm income which also contributes to the rural development strategy formulation process. In this study the
importance of off-farm income is recognized and the linkages to farming systems development are identified, as well as the potential impact of farm growth on the landless.
4
Figure 1-1. Farmers’ view of farm system, Bangladesh 12
of the approach lies in its ability to integrate multi-dis- expanded, placing increasing emphasis on non-pro-
ciplinary analyses of production and its relationship to duction activities at the farm level, the role of the com-
resources, technologies, markets, services, policies and munity, the environment and support services. The
institutions in their local cultural context. In such current perspective, with its focus on the farm house-
analyses, bio-physical dimensions (such as soil nutri- hold as the centre of a network of resource allocation
ents and water balances) and socio-economic aspects decisions, corresponds closely to the Sustainable
(such as gender, food security and profitability) are Livelihoods Approach, promoted by DfID.
combined at the level of the farm, where most agricul- The livelihoods of practically all of the rural poor
tural production and consumption decisions are depend directly or indirectly on natural resources.13
taken. Through grouping relatively homogeneous Poor farm households manage small individual
farms into farming systems, the approach facilitates resource endowments, while artisanal fishing and
the ex-ante assessment of investment and policy actions pastoral households often utilise limited common
concerning relatively large rural populations. property/open access resources. The heavy depen-
The use of the Farming Systems Approach as an dence of poor farm households on natural assets or
analytical framework became common in the late resources, complemented by human and social
1960s and early 1970s, as a response to the failure of capital, is in marked contrast to the reliance of urban
technologically driven approaches to small holder households on physical, financial and human capital;
development. Over the past 30 years, the approach this contrast is even more accentuated for those in
has evolved markedly, as shown in Table 1-4. severe poverty.
Essentially, the scope of the analysis has gradually
12
Extracted from “Households, Agroecosystems and Rural Resources Management. A guidebook for broadening the concepts of gender and farming systems.” Lightfoot,
C., S. Feldman and M.Z. Abedin. Bangladesh Agricultural Research Institute and the International Center for Living Aquatic resources Management. Educational Series 12.
13
Idriss Jazairy, Mohiuddin Alamgir and Theresa Panuccio. 1990. The State of World Rural Poverty. New York University Press for IFAD. Rome.
5
Table 1-4: Evolution of the Farming Systems Approach14
Characteristic 1970s 1980s 1990s 2000
System Level:
Farm
Household
Groups/Community
District/Zones/Catchments or Sector
Livelihood Focus:
Crops
Crop-Livestock
Multiple Household Livelihoods
Functional Focus:
Research
Research + Extension
Research+Extension+Support Service
Multi-sectoral, incl. Non-agricultural
Stakeholder Focus:
Public
Public + Civil society
Public + Civil society + Private
Other Foci:
Gender
Household food security
Productivity+Resource mgmt
Darker squares indicate greater focus on the element in that period
DELINEATION OF MAJOR FARMING mapped in order to estimate the magnitudes of their
SYSTEMS populations and resource bases. Within each of the
broad systems, emphasis has been placed on the iden-
The delineation of the major farming systems tification of the typical farm type or household liveli-
presented in this study provides a useful framework hood pattern, and the associated trends and develop-
to determine appropriate agricultural development ment issues, thus contributing to the identification of
strategies and interventions in developing countries. broad strategic approaches to poverty reduction, food
The definition of such broad farming systems security improvement and agricultural growth.
inevitably results in a considerable degree of hetero- The general criteria used for the definition of the
geneity within any single system. However, the alter- farming systems in this study have been based on the
native of identifying discrete micro-level farming following:
systems in each developing country – which could • the available natural resource base, including water,
result in hundreds or even thousands of systems land, grazing areas and forest; climate – of which
world-wide – would complicate the debate concerning altitude is one important determinant; landscape,
appropriate regional and global strategic responses. including slope; and farm size and tenure, in
The main farming systems have, therefore, been relation to access to different resources;
14
Adapted from J. Dixon and P. Anandajayasekeram, 2000. “Status of FSA Institutionalisation in East and Southern Africa and its
Implications”, International Farming Systems Research Extension Symposium, November 2000, Santiago, Chile.
6
• the dominant farm activities and household liveli- • the dominant livelihood, e.g. root crop, tree crop,
hood pattern (e.g. crops, livestock, trees, aquacul- rice-wheat, artisanal fishing, pastoral;
ture, hunting and gathering, off-farm activities); • production intensity,
technologies and the resulting intensity of produc- e.g. intensive, extensive, sparse;
tion and integration of crops and livestock; and
• crop-livestock integration, e.g. mixed;
farm management and organisation (e.g. family,
corporate, co-operative, etc).
• location, e.g. urban based.
The spatial mapping of farming systems present-
Based on these criteria, the following seven broad ed in this study represents a compromise between the
types of farming system are prevalent, to a greater of usefulness of showing farming system areas in a
lesser degree, in the developing regions: (i) irrigated graphical manner, and the dangers of implying sharp
farming systems, embracing a broad range of food and boundaries between neighbouring systems. With a
cash crops, and of farm sizes; (ii) rainfed farming large degree of variation inevitable among individual
systems in humid high potential areas, with systems farm households within any one system, there are in
dominated by one or another crop activity (notably root reality, no sharp boundaries but rather fuzzy transi-
crops, cereals, industrial tree crops – both small scale tions. Often, one farming system gradually merges
and plantation – and commercial horticulture) and into another. In other cases, broad systems may be
mixed crop-livestock systems; (iii) rainfed farming separated by limited areas with quite distinct charac-
systems in steep and highland areas, often mixed crop- teristics (e.g. lower slopes of mountain areas), the
livestock systems; (iv) rainfed small-scale farming identification of which would not be useful in a study
systems in dry or cold low potential areas, with mixed with this purpose and on this global scale.
crop-livestock and pastoral systems which grade into Irrigation constitutes a special case in relation to
sparse, often dispersed, systems with very low current the heterogeneity of farming systems. Where irriga-
productivity or potential because of extreme aridity or tion-based production is the dominant characteristic
cold; (v) large-scale commercial farming systems, across within an area, as in the case of large-scale irrigation
a variety of ecologies and with diverse production schemes, the entire zone has been classified as an irri-
patterns; (vi) coastal artisanal fishing and mixed farming gation-based farming system. However, substantial
systems; (vii) urban-based farming systems, typically amounts of irrigation appear as small yet important
focused on horticultural and animal production. areas of otherwise rainfed farming systems, and their
Applying the above criteria and farming system implications are reflected in the analysis of constraints
groups in a pragmatic fashion, with emphasis on and opportunities. Because irrigated agriculture is so
poverty reduction and agricultural growth, resulted in different from rainfed – not only in characteristics, but
the identification of 72 farming systems, with an also in terms of priorities and strategic approaches –
average agricultural population of about 40 million substantial localised concentrations of irrigation have
inhabitants. Sometimes, sufficient differences exist been identified through cross hatching on the farming
within a farming system to justify reference to distinct system maps.
sub-types, for example, small scale farms and planta- For the purposes of this study, from three to five
tions or commercial farms, or low altitude and high farming systems were identified in each region on the
altitude areas. The names chosen for the farming basis of those judged to constitute key regional targets
systems reflect the seven farming systems types for poverty reduction in the coming three decades.
outlined above and incorporate key distinguishing The main criteria employed were; (i) potential for
attributes, notably: poverty reduction and (ii) potential for agricultural
growth. Rapid and sustained growth in a major
• water resource availability, e.g. irrigated, farming system – even one not currently associated
rainfed, dry;
with high levels of poverty – could be expected to
• natural resource extraction basis, have a significant impact on regional poverty through
e.g. forest-based, coastal;
migration and market linkages. Factors determining
• climate, e.g. tropical, temperate, Mediterranean; a system’s apparent growth potential include:
• landscape relief/altitude, e.g. highlands, upland, (i) favourable or acceptable underlying agro-climatic
lowland; and soil conditions; (ii) a relatively high ratio of land
• farm scale and structure, e.g. small scale, and other resources (water, forest) to human popula-
large scale; tion; (iii) a current low intensity of exploitation, and
7
(iv) the identification of constraints to intensification natural disasters, or policy shocks such as structural
which are now considered to be feasible to remove or adjustment. Moreover, completely new alternatives
reduce. may arise in the future, perhaps related to technology
or markets, which could not easily be foreseen at this
point.
EVOLUTION OF FARMING SYSTEMS Over decades, farming systems may differentiate
into sub-types that continue to evolve along recognis-
To achieve the study objective of identifying issues and ably different pathways. For example, in systems
strategies related to farming systems development under population and market pressure some farms
during the coming 30 years requires an understand- may successfully intensify for market production,
ing of the dynamics of farming systems.15 Both whereas others may regress to low input-low output
internal and external factors will influence the evolu- systems. Such differentiation has been observed in
tion of individual farms and, in aggregate, the farming some regions under the pressure of structural adjust-
system. Whilst internal factors centre on household ment programmes.
goals, the resource base (closely related to population In this study, an attempt is made to anticipate the
pressure) and the technologies in use, external factors co-evolution of farming systems and their environ-
are more diverse. These may include market develop- ments from the present until the year 2030, taking
ment and shifts in demand, agricultural services account of:
and policies, and the availability of market and • key trends in the farming systems, including
policy information. Moreover, relationships are recip- resource and asset patterns (natural, physical,
rocal; the farm system co-evolves with its external financial, human and social), technology and pro-
environment. ductivity, livelihoods (crops, livestock and off-farm
Often, the evolution of farm systems follows a enterprises) and outcomes (household food security
recognisable pathway. For example, a system original- and income);
ly dependent solely on the use of hand hoes may face
constraints as market-driven diversification occurs. • key trends in the socio-economic and institutional
environments, including community organisation,
This could lead to the increasing use of cattle for
markets, services and information.
draught power, replacing some manual operations
and, if land is available, an expansion of the cultivated
What is clear, however, is that no single strategy
area. Later, the intensification of crop production may
can be relied upon to respond to the needs of differ-
be driven by population expansion and shortage of
ent farms, or the needs at different times during the
land. Market-driven evolution sometimes leads to spe-
coming 30 years. Multiple support and intervention
cialisation in production and often to greater use of
strategies will be required to allow for these diverse
external inputs. Further stages may include partial
development paths, and they must be flexible enough
mechanisation of crop production, substantial market
to evolve to meet new conditions and influences that
integration and increased use of inputs. Ultimately, a
arise over time.
high degree of production intensity is likely, perhaps
with an export orientation, usually characterised by
intensive use of inputs, land aggregation and a high FACTORS INFLUENCING FARMING SYSTEMS
degree of mechanisation. In certain circumstances DEVELOPMENT
intensive mixed systems may develop. In either case,
good technical and market information is important. In order to present the analysis of farming systems
In any one location within a farming system, dif- and their future development within a framework that
ferent farms may be at different stages of evolution is broadly comparable between systems and across dif-
because of differentiated resource bases, family goals ferent regions, a number of broad sets of influences
and capacity to bear risk, or degree of market access. have been defined, within which the discussion of
Individual farm systems may also be shifted out of the issues, trends and strategies is generally presented
overall trajectory of system evolution because of for each region, as well as at a global level. These
internal or external shocks, such as family sickness, influences, described briefly below, group factors that
15
Volumes have been written on the evolution of agriculture. Boserup (1965) in “The Conditions of Agricultural Growth” analysed
the effects of population growth; Pingali and Binswanger, and later McIntyre, took market development into consideration as well.
8
are of importance to the present and future status and vance of farming systems analysis to rural develop-
development of farming systems. The categories ment is discussed, and particular attention is paid to
themselves represent, in the broad opinion of a wide describing the key trends that are expected to influ-
range of experts within the United Nations Food and ence farming system evolution over the next thirty
Agriculture Organisation, the major areas in which years, as well as their likely impact on poverty and
farming system characteristics, performance and evo- growth. This overview also presents a synthesis of the
lution are likely to be significantly affected over the six individual regional analyses, available separately,
next thirty years. and then reviews commonalities and crosscutting
issues emerging from these analyses, as well as the
Natural Resources and Climate lessons to be drawn in terms of broad priority areas
Issues and expected changes related to the availabili- which would benefit from consideration in a cross-
ty, quality, utilisation and management of natural regional context. It concludes with a brief discussion
resources, as well as possible changes in climatic of implementation modalities and other issues of rele-
parameters, such as rainfall, temperature and the vance to the implementation process.
frequency of severe weather events. The six complete regional analyses 16 provide more
detailed coverage and maps of each World Bank
Science and Technology region, and illustrate key issues, strategies or inter-
Current levels and distribution of technologies, as ventions. An initial overview of the agricultural status
well as changes and advances in their utilisation and of the region in question is followed by a brief descrip-
scientific developments in areas such as analytical tion and prioritisation of its major farming systems.
tools, biotechnology and post-harvest treatments. Historical and anticipated future trends related to
agriculture within the given region are also provided.
Globalisation and Market Development Selected farming systems from the region are then
The impact and changes related to expanding market examined in considerable detail. As a single region
infrastructure and activity in rural areas, as well as the may contain as many as 15 identified farming systems,
broader implications of reductions in barriers to trade 3-5 priority systems have been selected in each region
between countries and future patterns of demand for for this particular purpose, on the basis of the poten-
agricultural outputs. tial for poverty reduction or economic growth existing
in the system. Discussion of each priority system is
Policies, Institutions and Public Goods divided into three sections: (i) system description; (ii)
The role and impact of the state and related institu- system issues and trends, and (iii) recommended
tions on the functioning of farming systems, expressed strategies and interventions. The regional analyses
principally through policies, programmes, institu- conclude with a discussion of regionally important
tions, services and public investment in the rural issues and present proposals for overall strategic
space. priorities.
Information and Human Resources
The relevance of non-material capital to farming
systems, in terms of knowledge, information and
ability to access and utilise such knowledge.
STUDY STRUCTURE AND FORMAT
The study is documented in seven parts. This Synthesis
and Global Overview provides an outline of future
challenges, opportunities and proposed development
strategies from a world-wide perspective. The rele-
16
Comprising Sub-Saharan Africa (AFR); Middle East & North Africa (MNA); Eastern Europe and Central Asia (ECA); South Asia (SAS);
East Asia & the Pacific (EAP); and Latin America & the Caribbean (LAC).This study does not provide any analysis of farming systems
in OECD countries except in so far as they are expected to influence systems in the developing world.
9
2
Global Factors Influencing the
Evolution of Farming Systems
Farming systems have changed substantially in recent reaching 5.1 billion in the year 2000 18, of which
decades. Their evolution is directly influenced by 59 percent are classed as rural, and 85 percent of
internal factors – notably the availability of resources those as agricultural19. Over the next thirty years, rates
and population growth – as well as by external factors of population growth in developing countries are pro-
such as markets, new technologies, support services, jected to slow from their current level of 1.8 percent
policies and information. The trends affecting these per annum to an estimated 0.4 percent per annum.
forces at the global level are discussed in broad terms When combined with increased rates of urbanization
in the following section17. More detailed trends at the (from 40 percent in 2000 to 57 percent in 2030), the
regional and farming system levels are presented in rural population of developing countries is expected
the separate regional volumes and briefly summarised to start to decline after 2020.
in the relevant following Chapters of this document. The rapid population growth of past decades has
increased demand for food and other agricultural
produce. Since the early 1960s, the amount of land
NATURAL RESOURCES AND CLIMATE under cultivation in developing countries increased by
a quarter to just over 1 billion ha; and an additional
The interaction of natural resource availability, 0.1 billion ha are under permanent crops. Relative
climate and population determines the physical basis resource availability is very much a function of popu-
for farming systems. During the early stages of devel- lation – the availability of arable land per capita in
opment, increasing population generally leads to an developing countries has declined by almost half since
expansion in cultivated area and, in many cases, the 1960s. The current pressure of agricultural popu-
conflict between different land and water resources lation on arable and permanent crop land averages
users. Once the majority of good quality land is 2.3 persons per hectare – in relation to total popula-
already exploited, further population increases lead to tion the pressure averages 4.6 persons per hectare.
the intensification of farming systems. As forests come The pressure of population on land varies widely
under increasing pressure, biodiversity is threatened across regions, as shown in Box 2-1, from 0.3 persons
and there may be growing tension between develop- per hectare in East Europe and Central Asia and
ment and conservation goals. These trends may be 0.7 persons per hectare in Latin America, to 4.9
exacerbated by colonial and post-colonial forces that persons per hectare in East Asia and the Pacific.
have resulted in the concentration of indigenous or Since the 1960s, pasture and grazing land has
minority peoples on poorer quality land, thus aggra- expanded, by 15 percent globally, to about 2.2 billion
vating the degradation problem. ha in 1994. Some of this expansion came at the
In the last four decades of the 20th century, the expense of forest and woodland, which declined to
population of developing countries has doubled, about 2.3 billion ha over the same period. Annual
17
Derived from FAO staff discussions related to the Strategic Framework, Medium Term Plan and cross-departmental brainstorming
sessions for this Study, supplemented by FAO, 2000. Agriculture Towards 2015/2030.Technical Interim Report. FAO, Rome.
18
See Table 1-1.
19
Agricultural population is defined by FAO as those economically active in agriculture, fisheries or forestry and their dependants.
10
growth rates in arable areas vary considerably Despite the typically high costs involved in developing
between the regions, as shown in Box 2-2. By far the irrigation systems, irrigated land use has risen at three
highest growth rates in arable land were experienced times the rate of overall land expansion, doubling
in Latin America and the Caribbean – being 1.26 total irrigated area world-wide since 1961 to 197
percent p.a., compared with 0.18 percent p.a. in million ha. This suggests that despite global averages,
South Asia. It is worth noting that during this period the expansion phase of agricultural systems is indeed
cropping intensity rose in total only 5 percent, sug- over in many areas. However, intensification through
gesting that, in global terms, farming systems are still irrigation has its limits too. At present, irrigation
in the area expansion phase. consumes about 70 percent of the total volume of
The FAO AT 2030 report estimates that an addi- fresh water used by humans, but this proportion is
tional 1.8 billion ha of land of “acceptable” quality likely to decline during the coming 30 years. Although
remains available for exploitation, but this seemingly irrigation is expected to continue to expand in the
favourable scenario is constrained by a number of coming three decades, non-agricultural demand for
factors. Much of the land categorised as suitable for water is also expected to grow strongly 20 Despite the
agriculture is, in fact, only suitable for a very narrow fact that only 7 percent of total renewable water
range of crops (e.g. olive trees in North Africa). resources in developing country are currently exploit-
Secondly, more than 90 percent of available land is in ed, competition from other users together with the
Latin America and Sub-Saharan Africa, which means fact that much of the available water is not located in
that further expansion is a very limited option for areas of agricultural need, will slow irrigation growth
most of North Africa, Asia and the Middle East. Even in the coming decades to around 0.6 percent per
where potential for area expansion appears to exist, annum – only a third of its historical rate.
over 70 percent of these areas are estimated to suffer The expansion of agriculture plus changes in pro-
from one or more soil or terrain constraints. duction technologies has led to a reduction of agro-
Nevertheless, as shown in Box 2-2, FAO foresees some biodiversity. During the past four decades, there
continued expansion in cultivated land area to 2030 – has been a considerable reduction in the number of
but at only half the rate (0.34 percent p.a.) of the last species and an even greater contraction in the number
40 years – adding about 120 million ha to the world of cultivars of rice, wheat and maize in current use. A
total. By the year 2030, despite the addition of a similar loss of biodiversity has occurred with domestic
further 2 billion people to the global population, animals. With the expansion of plant and animal
average arable land per agricultural person is actually breeding capacity, the number of crop varieties and
expected to rise slightly, as more land is brought into animal breeds in common usage is likely to decline
cultivation and urbanisation increases. even further.
The projected rapid increase in irrigation during There is a widespread belief that increasing pop-
the same period reflects the uneven distribution of ulation pressure within individual farming systems
agricultural land suitable for rainfed production. will inevitably lead to further land resource degrada-
BOX 2-1: AGRICULTURAL POPULATION BOX 2-2: EXPANSION IN ARABLE LAND
PRESSURE ON ARABLE AND FROM 1961 TO 1997, AND PROJECTED
PERMANENT CROP LAND BY REGION TO 2030
1961-97 1995/7-
(persons/ha) 2030
Sub-Saharan Africa 2.2 (% p.a.)
Middle East & North Africa 3.1 Sub-Saharan Africa 0.73 0.65
East Europe & Central Asia 0.3 Middle East & North Africa 0.42 0.22
South Asia 3.5 South Asia 0.18 0.13
East Asia & Pacific 4.9 East Asia & Pacific 0.91 0.07
Latin America & Caribbean 0.7 Latin America & Caribbean 1.26 0.55
Source: FAOSTAT, 2000 Source: FAO 2000
20
Rosegrant, M. 1997 “Water Resources in the Twenty-First Century: Challenges and Implications for Action”, IFPRI.
11
tion, including soil erosion, mining of soil nutrients,
depletion of groundwater and disappearance of biodi- BOX 2-3: POPULATION DRIVEN
versity. However, evidence that increasing population RESOURCE ENHANCEMENT
density – of humans or animals or both – causes land
In Machakos, Kenya increasing population density led
degradation is contradictory. In fact, there is a
initially to fragmentation of land, deforestation, soil
growing body of studies that indicate that, at least in
degradation, household food insecurity and poverty.
some cases, the reverse could be true. Increased pop-
However, recent studies have identified a U-turn in
ulation density has sometimes been accompanied by
resource management. As remittances have flowed
more productive and diversified farming systems and
back into the community, farmers have invested in their
more sustainable livelihoods (see Box 2-3). This is not
land, farm forestry has expanded rapidly, and erosion
to imply that high population densities do not often
and soil fertility declines have been reversed, leading to
lead to severe resource management problems, partic-
resource enhancement, and rising household incomes.22
ularly on unstable hill slopes. However, the relation-
ship between population density, farming systems and
resource degradation is complex and dynamic, and century will become normal. Rising temperatures will
degradation is not inevitable. Inequitable distribution inevitably lead to a rise in sea levels (estimated at
of land is often a root cause of degradation, through between 0.1 and 0.9m over this century).
the alienation of better quality land by capital-inten- There is little doubt that agriculture, and food
sive, large-scale farmers and the pushing of smaller, security among rural populations, will be affected by
weaker, or migrant farmers onto marginal land. these changes. Among the impacts predicted with
The nature of resource management trends some degree of confidence by the IPCC working
differs between high and low potential areas. In group is a reduction in potential crop yields in most
intensely-farmed high potential areas, soil organic tropical and sub-tropical regions, and also in mid-lat-
matter commonly decreases, soil structure deterio- itudes if temperature increases are towards the
rates and micronutrient deficiencies and nutrient higher end of the predicted range. Water availability,
imbalances arise. In irrigated areas, ground water particularly in the sub-tropics, is expected to diminish,
tables may fall. In low potential areas, nutrient losses although some areas, such as South East Asia, may
from erosion and nutrient mining are increasing and experience increased water availability, as a result of
desertification is a major concern. increased intensity in monsoon activity. There is
The potential impact of global warming has been expected to be a widespread increase in the risk of
widely debated in the scientific community, culminat- flooding, as a result both of rises in sea level, and as a
ing in the UN Framework Convention on Climatic consequence of increased severity of precipitation
Change (UNFCCC) in 1992. Agriculture currently from storms, hurricanes and monsoons. Finally,
contributes about 30 percent of the global anthro- labour availability may be affected as a result of an
pogenic emission of greenhouse gases. Growth in the expected increase in the transmission of diseases,
production of these gases by crops is expected to slow, both vector borne (e.g. malaria), and water borne
but methane production by livestock could increase (e.g. cholera). Overall, the report states that the food
around 60 percent by 2030. Accumulated evidence 21 security position in Africa can only be expected to
now strongly suggests that impacts from global climate worsen as a result of predicted climate changes.
change will be significant: average global surface tem-
peratures are expected to rise by an estimated 1.4 to
5.8oC in the next 100 years, while the frequency of SCIENCE AND TECHNOLOGY
climatic extremes (temperatures, precipitation and
winds) is expected to increase dramatically. Models Investments in agricultural science and technology
based on the Intergovernmental Panel on Climate have expanded rapidly during the last four decades of
Control (IPCC) scenario of a one percent increase in the 20th century and have led to significant innova-
greenhouse gases per year predict, with a very high tion within the sector. In addition, major technical and
degree of statistical significance, that within 80 years institutional changes have also occurred that are
extremes that are currently experienced only once a expected to lead to long-term structural changes in
21
IPCC Working Group 2 Third Assessment Report, February 19, 2001.
22
Machakos, Kenya (see Tiffen et al), Sindalpalchowk, Nepal (Dixon pers comm).
12
the process of technology development and dissemi- productivity. As shown in Box 2-4, global average
nation. In the earlier years, the CGIAR international wheat yields under both rainfed and irrigated condi-
research system was established and national agricul- tions, are expected to increase by about 40 percent
tural research organisations (NARs) were greatly from the mid-1990s to 2030; rice yields by about 33
strengthened. More recently, the research agenda has percent during the same period; and maize yields by
moved from a focus on individual crop performance 39 percent during this period (see also Figure 2-1).
to a growing acceptance of the importance of
increased system productivity. This is viewed as result-
BOX 2-4: PROJECTED YIELD INCREASES
ing from better-managed interactions among diversi-
TO 2030
fied farm enterprises, from sustainable resource man-
agement, and better targeting of technologies towards 1995-7 2030
women farmers and poor households. Research (t/ha)
methods are being revolutionised by biotechnology. Wheat 2.46 3.44
Perhaps even more importantly in the long term, insti- Rice 3.50 4.63
tutional modalities are now shifting. From a public Maize 2.52 3.49
sector focus, largely led by the international system, Source: FAO 2000
more emphasis is now given to public-private partner-
ships driven more by the demands of clients.
The historical focus of research effort on food Average yield increases for all cereals are expected
crop production technologies, with a particular to be even greater under irrigation (3.82 t/ha to
emphasis on improved varieties, has been undeniably 5.16 t/ha) than under rainfed conditions (1.71 t/ha to
successful. Average crop yields have increased by 2.23 t/ha). Irrigation efficiency is expected to increase.
nearly three quarters (71 percent) since 1961, while Average fertiliser nutrient use in developing countries
average cereal grain yields have doubled to 2.8 t/ha. is expected to grow from around 90 kg/ha in the mid
Increased yields have contributed to increased food 1990s 23 to 107 kg/ha in 2030; and increases in fertilis-
security at all levels and have led to declining real er nutrient efficiency are also expected.
prices for food grains. It is significant that FAO pro- Investments in technology development for non-
jections to 2030 indicate a continuing increase in land cereal crops have usually received a lower priority.
Figure 2-1: Cereal Grain Yield Trends 1961-2030
5
4.5
4
3.5
3
2.5
2
1.5
1
0.5
0
1961 1970 1980 1990 2000 2010 2020 2030
Maize Rice Wheat All Cereals
23
A repetition of the rapid growth of the 1960-1990 period, from a low base of 7 kg/ha in the early 1960s, is not expected.
13
The private sector and farmers organisations have private sector and thus is likely to focus on profit-gen-
invested in research for commercially important cash erating inputs, export crops and agro-processing.
crops – examples include coffee, tea, sugar cane and Countries such as China and Argentina have sown
bananas – but many tropical staples and minor cash large areas of genetically modified cereals, oil crops
crops have received relatively little attention. and other commercially important species within the
last five years. Whilst there has been a gradual
decrease in the national and international public
BOX 2-5: IS STAGNATION OF RICE YIELDS
funding available for agricultural research and exten-
A PROBLEM? 24
sion systems, private sector biotechnology research
There has recently been increasing concern about stag- has attracted ample funding.
nating rice yields. After rapid growth in the 1970s and Substantial regional differences exist in the way in
about 80s, yield increases in the 1990s have been lower which research and extension are financed, with many
than the rate of population growth. However, a synthe- Latin American countries increasingly privatising
sis of evidence suggests that these reductions do not their extension services. With the shift in public
necessarily presage stagnation in average yields. resource allocation from agriculture to processing,
Potential productivity gains exist both in terms of marketing and service industries, farming systems
raising maximum yields and through reduction in the have often become more self-reliant, more diverse or
“yield gap” between the best and the average farmers
more integrated into the rural and rural/urban
within a region. In fact, although the annual yield
economy.
increases of about 0.8% forecast by FAO for the next
30 years are considerably slower than for earlier
decades, average global irrigated rice yields are still
GLOBALISATION AND MARKET
expected to increase from 4.35 t/ha in the mid 1990s
DEVELOPMENT
to 5.77 t/ha in 2030 (rainfed rice yields are also
expected to increase).Yields also need to be viewed in
The process of globalisation, through the reduction of
the context of the productivity of the whole farm
system over the course of the full year; in general, these impediments to international trade and investment, is
whole-farm productivity levels are rising steadily. already generating profound changes in the structure
of production at all levels, including smallholder-
farming systems in many developing countries. Not
Weed and pest pressure is increasing for many only is market development in developing country
crops and further research is needed in this area. systems accelerating, but patterns of production and
Similarly, investment in livestock research has gener- natural resource usage are also changing profoundly
ally not been commensurate with the contribution of in response to market forces. The FAO AT2030 study
the sub-sector to household income or GADP. Only assumes that this process will continue at a moderate
one Consultative Group research centre concentrates pace, but a reversal or even acceleration is always
on livestock (ILRI), although other centres have possible.
animal production programmes. In contrast, agricul- As barriers to trade between countries diminish,
tural research in industrialised countries has been rel- and subsidies to industrialised country producers are
atively well funded (albeit at much lower levels than gradually removed, those products which are compet-
for human sciences). Consequently, a greater range of itive in world markets will benefit, while those that
new technologies is available for production systems have hitherto relied on protection will be under
and crops of interest to developed countries than threat. Broad social, economic and cultural trends will
for smallholder production systems in developing also contribute to a profound reshaping of market
countries. demand, as increased urbanisation, rising incomes,
Growing investments in biotechnology are likely improved communications and the diffusion of
to increase agricultural research productivity, and cultural preferences, have their impact. Finally, the
have the potential to revolutionise production prac- availability of new production, post-harvest and trans-
tices and generate customised varieties. However, port technologies will also change demand patterns by
most biotechnology research is concentrated in the making possible the delivery of new products, or
24
For further details see Tran, D. and N. Nguyen, 2001,Trends in Rice Yields, Crop Production and Protection Division, Case Study, Global Farming Systems Study, FAO, Rome.
14
