Front Cover
 Title Page
 Table of Contents
 Acronyms in text and reference...
 Historical and theoretical...
 Food and agricultural policy
 Smallholder farming
 Technical change
 Trade, marketing, credit,...
 Migration, employment, and equity...
 Synthesis and implications for...
 List of MSU development papers
 Back Cover

Group Title: MSU international development papers, no. 1
Title: Research on agricultural development in sub-Saharan Africa
Full Citation
Permanent Link: http://ufdc.ufl.edu/UF00086762/00001
 Material Information
Title: Research on agricultural development in sub-Saharan Africa a critical survey
Series Title: MSU international development papers
Physical Description: xi, 335 p. : ; 28 cm.
Language: English
Creator: Eicher, Carl K
Baker, Doyle C ( Doyle Curtis )
Publisher: Dept. of Agricultural Economics, Michigan State University
Place of Publication: East Lansing Mich
Publication Date: 1982
Subject: Agriculture -- Economic aspects -- Africa, Sub-Saharan   ( lcsh )
Agriculture -- Africa, Sub-Saharan   ( lcsh )
Agriculture -- Aspect économique -- Afrique noire   ( rvm )
Agriculture -- Afrique noire   ( rvm )
Rural conditions -- Africa, Sub-Saharan   ( lcsh )
Conditions rurales -- Afrique noire   ( rvm )
Genre: bibliography   ( marcgt )
non-fiction   ( marcgt )
Bibliography: Includes bibliographical references (p. 263-335).
Statement of Responsibility: by Carl K. Eicher and Doyle C. Baker.
 Record Information
Bibliographic ID: UF00086762
Volume ID: VID00001
Source Institution: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: oclc - 08649106
lccn - 82622612
issn - 0731-3438 ;

Table of Contents
    Front Cover
        Front Cover 1
        Front Cover 2
    Title Page
        Page i
        Page ii
        Page iii
        Page iv
    Table of Contents
        Page v
        Page vi
        Page vii
        Page viii
        Page ix
    Acronyms in text and references
        Page x
        Page xi
        Page 1
        Page 2
        Page 3
        Page 4
        Page 5
        Page 6
        Page 7
        Page 8
        Page 9
        Page 10
        Page 11
        Page 12
        Page 13
        Page 14
        Page 15
        Page 16
    Historical and theoretical perspectives
        Page 17
        Page 18
        Page 19
        Page 20
        Page 21
        Page 22
        Page 23
        Page 24
        Page 25
        Page 26
        Page 27
        Page 28
        Page 29
        Page 30
        Page 31
        Page 32
        Page 33
        Page 34
        Page 35
        Page 36
        Page 37
        Page 38
        Page 39
        Page 40
    Food and agricultural policy
        Page 41
        Page 42
        Page 43
        Page 44
        Page 45
        Page 46
        Page 47
        Page 48
        Page 49
        Page 50
        Page 51
        Page 52
        Page 53
        Page 54
        Page 55
        Page 56
        Page 57
        Page 58
        Page 59
        Page 60
        Page 61
        Page 62
        Page 63
        Page 64
        Page 65
        Page 66
    Smallholder farming
        Page 67
        Page 68
        Page 69
        Page 70
        Page 71
        Page 72
        Page 73
        Page 74
        Page 75
        Page 76
        Page 77
        Page 78
        Page 79
        Page 80
        Page 81
        Page 82
        Page 83
        Page 84
        Page 85
        Page 86
        Page 87
        Page 88
        Page 89
        Page 90
        Page 91
        Page 92
        Page 93
        Page 94
        Page 95
        Page 96
        Page 97
        Page 98
        Page 99
        Page 100
        Page 101
        Page 102
        Page 103
        Page 104
        Page 105
        Page 106
        Page 107
        Page 108
        Page 109
        Page 110
        Page 111
        Page 112
    Technical change
        Page 113
        Page 114
        Page 115
        Page 116
        Page 117
        Page 118
        Page 119
        Page 120
        Page 121
        Page 122
        Page 123
        Page 124
        Page 125
        Page 126
        Page 127
        Page 128
        Page 129
        Page 130
        Page 131
        Page 132
        Page 133
        Page 134
        Page 135
        Page 136
        Page 137
        Page 138
        Page 139
        Page 140
        Page 141
        Page 142
        Page 143
        Page 144
        Page 145
        Page 146
        Page 147
        Page 148
        Page 149
        Page 150
        Page 151
        Page 152
        Page 153
        Page 154
        Page 155
        Page 156
        Page 157
        Page 158
        Page 159
        Page 160
        Page 161
        Page 162
        Page 163
        Page 164
        Page 165
        Page 166
        Page 167
        Page 168
        Page 169
        Page 170
        Page 171
        Page 172
        Page 173
        Page 174
        Page 175
        Page 176
        Page 177
        Page 178
    Trade, marketing, credit, and consumption
        Page 179
        Page 180
        Page 181
        Page 182
        Page 183
        Page 184
        Page 185
        Page 186
        Page 187
        Page 188
        Page 189
        Page 190
        Page 191
        Page 192
        Page 193
        Page 194
        Page 195
        Page 196
        Page 197
        Page 198
        Page 199
        Page 200
        Page 201
        Page 202
        Page 203
        Page 204
        Page 205
        Page 206
        Page 207
        Page 208
        Page 209
        Page 210
        Page 211
        Page 212
        Page 213
        Page 214
        Page 215
    Migration, employment, and equity issues
        Page 216
        Page 217
        Page 218
        Page 219
        Page 220
        Page 221
        Page 222
        Page 223
        Page 224
        Page 225
        Page 226
        Page 227
        Page 228
        Page 229
        Page 230
        Page 231
        Page 232
        Page 233
        Page 234
        Page 235
        Page 236
        Page 237
        Page 238
        Page 239
        Page 240
        Page 241
        Page 242
        Page 243
        Page 244
        Page 245
        Page 246
        Page 247
        Page 248
        Page 249
        Page 250
        Page 251
        Page 252
        Page 253
        Page 254
    Synthesis and implications for research and development strategies in the 1980s
        Page 255
        Page 256
        Page 257
        Page 258
        Page 259
        Page 260
        Page 261
        Page 262
        Page 263
        Page 264
        Page 265
        Page 266
        Page 267
        Page 268
        Page 269
        Page 270
        Page 271
        Page 272
        Page 273
        Page 274
        Page 275
        Page 276
        Page 277
        Page 278
        Page 279
        Page 280
        Page 281
        Page 282
        Page 283
        Page 284
        Page 285
        Page 286
        Page 287
        Page 288
        Page 289
        Page 290
        Page 291
        Page 292
        Page 293
        Page 294
        Page 295
        Page 296
        Page 297
        Page 298
        Page 299
        Page 300
        Page 301
        Page 302
        Page 303
        Page 304
        Page 305
        Page 306
        Page 307
        Page 308
        Page 309
        Page 310
        Page 311
        Page 312
        Page 313
        Page 314
        Page 315
        Page 316
        Page 317
        Page 318
        Page 319
        Page 320
        Page 321
        Page 322
        Page 323
        Page 324
        Page 325
        Page 326
        Page 327
        Page 328
        Page 329
        Page 330
        Page 331
        Page 332
        Page 333
        Page 334
        Page 335
    List of MSU development papers
        Page 336
        Page 337
    Back Cover
        Page 338
Full Text

I. I Iri D v 3o e gp ars

Car K. Eihe

I* ~ Dol C.t BakerS* ~ .

1982 Eas Lasig Mihia 482


Eric W. Crawford, Carl K. Eicher and Carl Liedholm, Co-Editors

The MSU International Development Paper series is designed to further

the comparative analysis of international development activities in Africa,

Latin America, Asia and the Near East. The papers report research findings

on historical, as well as contemporary, international development problems.

The series includes papers on a wide range of topics, such as alternative

rural development strategies; non-farm employment and small-scale industry;

housing and construction; farming and marketing systems; food and nutrition

policy analysis; economics of rice production in West Africa; technological

change, employment, and income distribution; computer techniques for farm

and marketing surveys; and farming systems research.

The papers are aimed at teachers, researchers, policy-makers, donor

agencies, and international development practitioners. Selected papers

will be translated into French, Spanish, or Arabic.

Individuals and institutions in Third World countries may receive single

copies free of charge. See inside back cover for a list of available papers

and their prices. For more information write to:

MSU International Development Papers
Department of Agricultural Economics
Agriculture Hall
Michigan State University
East Lansing, Michigan 48824-1039




Carl K. Eicher** and Doyle C. Baker***


*This paper has been prepared and published with financial assistance from The
Woodrow Wilson International Center for Scholars, Washington, D.C.; the Department of
Agricultural Economics, Michigan State University; the Sahel Secretariat and Documenta-
tion Center Grants: (AID/afr-G-1261) and (AID/afr-0929-G-SS-2011-00), Bureau for
Africa, Agency for International Development; and the Alternative Rural Development
Strategies Cooperative Agreements: (AID/ta-CA-3) and (AID/DAN-1190-A-00-2069-00),
Office of Multi-Sectoral Development, Bureau for Science and Technology, Agency for
International Development, Washington, D.C.

**Professor of Agricultural Economics, Michigan State University, East Lansing,

***Department of Agricultural Economics, Michigan State University, East Lansing,

MSU is an Affirmative Action/Equal Opportunity Institution

ISSN 0731-3438
All rights reserved by Michigan State University, 1982.

Michigan State University agrees to and does hereby grant to the United States
Government a royalty-free, nonexclusive and irrevocable license throughout the
world to use, duplicate, disclose, or dispose of this publication in any manner
and for any purpose and to permit others to do so.

Published by the Department of Agricultural Economics, Michigan State
University, East Lansing, Michigan 48824 U.S.A.


This survey of literature on the rural economies of sub-Saharan Africa has been
prepared at the request of the American Association of Agricultural Economics (AAEA).
It will be published in 1983 along with similar reviews on Latin America and Asia in
Volume IV of the AAEA's literature surveys.- The survey is being published by Michigan
State University as an MSU International Development Paper in advance of publication of
the AAEA volume in order to make it quickly available to students, teachers, and
researchers in universities and research institutes in Africa.
Our initial mandate was to review the literature on agricultural development by
agricultural economists but we have broadened this mandate. First, we review research
by economic historians on the precolonial and colonial development experiences. Second,
we present a brief overview of the technical literature on farming and livestock systems
for the benefit of social scientists. Third, we go beyond the agricultural sector to
appraise the literature on the rural nonfarm economy, including small-scale industry,
fishing, processing, storage, migration, income distribution, and other topics. Fourth, we
present some of the key research findings of political scientists, anthropologists,
sociologists, geographers, and technical scientists.
We have had two major goals in preparing this survey. The first is to present a
critical review of the major theoretical and policy debates and empirical findings on the
development of Africa's rural economies. The second is to identify the major research
gaps and research directions for the 1980s and 1990s. During the 1970s, Africa began to
receive the attention of scholars and donor agencies which was noticeably lacking in the
1950s and 1960s. Because of lagging food production and widespread poverty, Africa is
likely to receive even more attention in the coming decades than it has in the past
relative to Asia and Latin America and it is important that research resources be wisely
and effectively used.
This project began several years ago when the authors were a Resident Scholar and a
Research Associate, respectively, at the Woodrow Wilson International Center for
Scholars, in Washington, D.C. We gratefully acknowledge the support of the Wilson
Center, its staff, and the opportunity we had to draw on the resources of the Library of
Congress. The bulk of the research and preparation of the survey has been carried out in
the Department of Agricultural Economics at Michigan State University. The project
would never have been realized without the support of Larry Connor and Lester

-Lee Martin, ed. A Survey of Agricultural Economics Literature: Vol. IV. Agricul-
ture in Economic Development. Minneapolis, University of Minnesota Press, 1983.

Manderscheid. Laura Wilson helped with the tedious task of checking and rechecking our
bibliographical references. Joe Lauer, Director, MSU Sahel Bibliographic Center,
provided valuable advice on citations for many difficult references. Lucy Wells and
Jeanette Barbour were tireless in typing and retyping a number of drafts and redrafts.
We also would like to acknowledge the invaluable help from the following reviewers:
Vincent Barrett, Sara Berry, Derek Byerlee, Enyinna Chuta, Eric Clayton, John Cohen,
Mike Collinson, Eric Crawford, Bob Deans, Christopher Delgado, W. Doppler, John
Erikson, B. Falusi, Don Ferguson, Pascal Fotzo, Russell Freed, Donald Heisel, Lane
Holdcroft, Francis Idachaba, R.W. Palmer-Jones, M.C. Latham, Uma Lele, Carl Liedholm,
A.R.C. Low, R.E. McDowell, K. Meyn, Isaac Minde, M. Miracle, Wilford W. Morris,
W. Mwangi, David Norman, O. Ogunfowora, Kenneth Robinson, Stephen Sanford, Tjaart
Schillhorn, Kenneth Shapiro, John Staatz, Martin Upton, William Whelan, and David
We would like to especially express our gratitude to the Editor of the Survey
Volumes, Lee Martin of the University of Minnesota. Lee Martin's insightful comments
have helped us revise the manuscript and bring it up to his high standards. The support of
our wives, Shirley Eicher and Kathy Baker, has been invaluable. Finally, we would like to
thank the innumerable researchers who are responsible for the research on which this
survey is based. They have contributed to our understanding of African rural economies.
We, of course, take full responsibility for the interpretations and views presented in these

Carl K. Eicher and Doyle C. Baker




Scope of Review 5
Standard References 8
Overview of Agricultural Systems 9
Farming Systems 9
Ecological Zones 10
Intensity of Cultivation 11
Major Crops 12
Livestock Systems 14

Historical Perspective 17
Precolonial Economic Activity: 1800 to 1880s 17
The Colonial Period: 1880-1960 19
Transition to Independence 23
Theoretical Perspectives 24
Economic Behavior of Farmers and Traders 26
Target Income/Backward Bending Supply Curve of Labor 28
The Cultural Barrier Hypothesis 29
Synthesis 30
Western Development Models 30
Vent-for-Surplus 31
Labor Surplus 33
Land Surplus 34
Assessment 35
Political Economy and Radical Perspectives 35
Dependency Theory 36
Micro-Marxists 38
Assessment 40

Agrarian Capitalism and Socialism 41
Planning and Agricultural Sector Modelling 44
Large Versus Small Farms 47
Smallholder Farming: A Descriptive Overview 47
Large-Scale Farming: An Overview 49
Small-Scale and Large-Scale Farming in Nigeria 50
Assessment 51
Marketing Boards and Food Grain Boards 53
Marketing Boards 53
Food Grain Boards 55
Agricultural Prices 57
Rural Development Programs 59
Community Development and Animation Rurale 60
Integrated Rural Development 61
Accelerated Food Production Campaigns 63
Synthesis 66


Overview of Institutions Carrying Out Farm Level Studies 67
Methodological Issues in Rural Surveys 72
Issues in Survey Design 72
References on Survey Design 76
Data Processing 77
Applications of Analytical Techniques 79
Budgeting 79
Regression Analysis 82
Farm Production Functions 82
Determinants of Resource Use 86
Supply Functions 87
Programming Models 90
Research on Selected Topics 94
Social and Cultural Issues 95
Class Analysis 95
Social Impact Analysis 96
Indigenous Knowledge 97
Land Tenure 97
Labor Use 99
Supply of Labor 102
Allocative Efficiency 105
Returns to Land and Labor 108
Land Intensification 110

Historical Perspective 113
Agronomic Research 116
Crop Improvement 117
Sorghum and Millet 117
Maize 119
Rice 120
Wheat and Triticale 121
Cotton 122
Legumes 123
Root Crops 123
Soil Fertility and Fertilizer 124
Soil Resources 124
Fertilizer Use 125
Yield Response 125
Profitability of Fertilizer 127
Demand for Fertilizer 128
Research on Managerial Practices 129
Intercropping 129
Plowing 130
Crop Rotation 131
Sowing Dates and Planting Density 131
Synthesis 132


Irrigation 133
Potential for Irrigation 134
Country and Regional Experiences 135
Economics of Irrigation 137
Synthesis and Research Direction 138
Irrigation References 139
Animal and Tractor Mechanization 139
Methodological Issues 140
Animal Traction 140
History of Animal Traction 141
Empirical Results 142
Tractor Mechanization 146
Synthesis 149
Agricultural Extension 150
Historical Perspective 151
Alternative Extension Approaches 152
National Extension Programs 152
Farmer Training Centers 154
Commodity Parastatals 155
Training and Visit 156
Diffusion Studies 156
Farming Systems Research (FSR) 158

Behavior of Herders: Empirical Evidence 164
Livestock Production: Major Issues for Researchers 168
Improving the Data Base 169
Herders or Ranches? 170
Animal Breeding 171
Cattle Feeding and Animal Nutrition 172
Cattle Diseases and Parasites 173
Range Management and Land Use Planning 174
Small-Scale Fattening Schemes 176
Small Ruminants 177
Research Direction 177
Livestock References 178

International Agricultural Trade 179
Export Linkages 179
Intra-African Trade 180
Regional Integration 180
Special Trade Agreements 181
Food Aid 181
Local Trade and Food Marketing 182
Studies by Geographers and Anthropologists 183
Competitiveness and Efficiency of Local Trade 184
Livestock Marketing 188
Fish Marketing 189
Transportation 191
Processing 191


Storage 194
Methodological Issues 194
On-Farm Storage 195
Off-Farm Storage 196
Storage Costs 197
Storage Losses 198
Improvements in Traditional Granaries 198
Credit 198
Theoretical Framework 199
Informal Lenders 200
Government Credit Institutions 201
Should Interest Rates Be Subsidized? 201
Credit Components of Technical Packages 202
Rural Savings 203
Policy Direction 203
Research Direction 204
Cooperatives 204
Consumption 207
Theoretical Issues 207
Methodological Issues 208
Empirical Findings 209
Synthesis 211
Nutrition 212
Seasonal Hunger 213
Impact of Cash Crop Expansion on Nutritional Status 213
Strategies for Alleviating Malnutrition 215

Income Distribution and Inequality 216
Historical Perspective 216
Methodological Issues 217
Empirical Evidence 218
Research Agenda 220
Population 221
Migration 223
Theoretical Perspectives 224
Methodological Issues 226
Seasonal and Rural to Urban Migration 227
Who Migrates? 227
Rates of Migration 228
Determinants of Migration 229
Remittances 230
Level of Urban Unemployment 231
International Migration 232
Who Benefits From Migration? 233
Synthesis and Research Direction 234
Rural Employment 235
Women in Development 237
Empirical Results 239
Needed Research 241



Rural Small-Scale Industry 241
Theoretical Issues 242
Empirical Results 242
Policy Direction 244
Research Direction 246
Fisheries 246
Small-Scale (Artisan) Fishing 247
Large-Scale Fishing 248
Fisheries Management 249
Fish Ponds 250
Research Direction 250
Recovery of the Sahel 251
Research Findings 252
Climate 252
Desert Encroachment 252
Demography 253
Overcoming Food Dependency 253
Assessment 254





ACE = American Council on Education, Washington
ADC = Agricultural Development Council, New York
AID = U.S. Agency for International Development, Washington
AMIRA = Groupe de Recherche pour I'Amelioration des Methodes d'Investigation
en Milieu Rural Africain (Research Group on the Improvement of Survey
Methods in Rural Africa), Paris
BRALUP = Bureau of Resource Assessment and Land Use Planning, University of
Dar es Salaam
CARDAN = Centre d'Analyse et de Recherche Documentaires pour l'Afrique Noire
(Documentary Analysis and Research Center for Africa), Paris
CEEMAT = Centre d'Etudes et d'Experimentation du Machinisme Agricole Tropicale
(Center for Studies and Experimentation on Tropical Agricultural Machinery),
Antony, France
CGIAR = Consultative Group on International Agricultural Research, Washington
CILSS = Comite Permanent Inter-Etats de Lutte Contre la Secheresse au Sahel
(Permanent Inter-State Committee of Drought Control in the Sahel),
CIMMYT = Centro Internacional de Mejoramiento de Maiz y Trigo (International Maize
and Wheat Improvement Center), Mexico D.F.
CNRA = Centre National de Recherches Agronomiques (National Center for Agronomic
Research), Bambey, Senegal
CODESRIA = Council for the Development of Economic and Social Research in Africa,
CRED = Center for Research on Economic Development, University of Michigan,
SAnn Arbor
ECA = United Nations, Economic Commission for Africa, Addis Ababa
ECOWAS = Economic Community of West African States
FAO = Food and Agriculture Organization of the United Nations, Rome
IADS = International Agricultural Development Service, Washington
IAR = Institute for Agricultural Research, Ahmadu Bello University, Zaria, Nigeria
IBRD = International Bank for Reconstruction and Development (World Bank)
ICRISAT = International Crops Research Institute for the Semi-Arid Tropics, Hyderabad
IDEP = Institut Africain de Developpement Economique et de Planification (African
Institute for Economic Development and Planning), Dakar
IDRC = International Development Research Centre, Ottawa
IDS = Institute for Development Studies, University of Nairobi, Nairobi
IEMVT = Institut d'Elevage et de Medecin Veterinaire de Pays Tropicaux (Institute
of Livestock and Veterinary Medicine for Tropical Countries), Maisons-
Alforet, France










Institute d'Economie Rurale (Institute of Rural Economy), Bamako
International Fund for Agricultural Development, Rome
International Fertilizer Development Center, Muscle Shoals, Alabama
International Food Policy Research Institute, Washington
International Institute of Tropical Agriculture, Ibadan
International Labor Office, Geneva
Institute National de Recherche Agronomique (National Institute of Agronomic
Research), Niamey
Institute National de la Statistique et des Etudes Economiques (National
Institute of Statistics and Economic Studies), Paris
Institute de Recherches Agronomiques Tropicales et des Cultures Vivrieres
(Institute of Agronomic Research on Tropical Crops), Paris
Institute Senegalais de Recherches Agricoles (Senegalese Institute of Agricul-
tural Research), Dakar
Institute of Statistical, Social, and Economic Research, University of Ghana,
Massachusetts Institute of Technology, Cambridge, Massachusetts
Nigerian Institute of Social and Economic Research, Ibadan
Operation Arachide et Cultures Vivrieres (Groundnuts and Food Crops
Operation), Mali
Organization of African Unity, Addis Ababa
Overseas Development Council, Washington
Overseas Development Institute, London
Organisation for Economic Cooperation and Development, Paris
Office de la Recherche Scientifique et Technique Outre-Mer (Overseas
Scientific and Technical Research Office), Paris
Swedish Agency for Research Cooperation With Developing Countries
Societe d'Etudes pour le Developpement Economique et Social (Economic
and Social Development Studies Corporation), Paris
United Nations Development Program, New York
United States Department of Agriculture, Washington
West African Rice Development Association, Monrovia
World Health Organization, Geneva



Carl K. Eicher and Doyle C. Baker
Department of Agricultural Economics,
Michigan State University, East Lansing, Michigan


Sub-Saharan Africa is a vast subcontinent comprised of 41 countries- with
heterogeneous endowments of resources, colonial histories, and levels and opportunities
for development (Figure 1). The population of Africa is presently about 350 million but
Nigeria with 80 million people has one-fourth of the population and produces 46 percent of
the gross national product of the region. Population densities in Africa are extremely low
relative to Asia. The Sudan, for example, is two-thirds the size of India but it has only 18
million people as compared with 670 million in India. The Republic of Zaire (formerly the
Belgian Congo) is five times the size of France but only 5 percent of its arable land is
estimated to be under cultivation.
Although the density of population as a whole is low relative to Asia, the
distribution is uneven and there are areas that are near maximum capacity given present
agricultural technology and knowledge of how to deal with soil erosion and environmental
problems. Moreover, according to recent estimates by the United Nations, sub-Saharan
Africa is the only region in the world where the rate of natural growth of population
increased over the 1960-80 period (U.N., 1981). The population growth rate increased
from 2.5 percent in the 1960s to an estimated rate of 2.7 to 2.9 percent in the 1970s-/ and
is projected to be 3.0 percent per year over the 1980-85 period (World Bank, 1980).
In Africa, 60 to 90 percent of the people are in the agricultural sector, in contrast
to 30 to 50 percent of the population in most Latin American nations. Agricultural and
mineral exports dominate most African economies as they do in many Latin American

!iThe definition of sub-Saharan Africa normally includes 40-46 countries depending
on the number of off-shore islands--e.g., Seychelles, Mauritius--one wishes to include.
We have excluded the Republic of South Africa and some of the off-shore islands and
arrived at an arbitrary list of 41 countries to include in this survey (see Table 1, page 4).

!/The World Bank (1981b) estimate is 2.7 percent and the UN (1980) estimate is 2.9
percent. The population picture has a bright side, however. The life expectancy at birth
has increased from an estimated 38 years in 1950 to almost 50 years in 1980--and the
crude death rate has fallen from an estimated 27 per 1,000 in 1950 to 18 per 1,000 per
year in 1980.

Figure 1. Countries in Sub-Saharan Africa
Included in Literature Review

countries. But unlike much of Latin America, the size of the industrial sector is
extremely modest in most African countries and agriculture is claimed to be the
foundation of nearly all national development programs. Although the rate of growth of
African cities is currently the highest in the world, sub-Saharan Africa continues to be the
least urbanized region. Overall, the population was about 70 percent rural in 1980
compared with about 80 percent in 1960, but the percentage varies from 54 percent rural
in Ghana to around 90 percent in Tanzania and Upper Volta. It is clear that for the next
25 to 50 years the majority of the population of most countries will continue to live in
rural areas.
Sub-Saharan Africa is the poorest part of the world's economy. The World Bank's
World Development Report, 1981 shows, for example, that 22 of the 36 of the low-income
countries in the world are sub-Saharan African countries.-/ Moreover, the rate of
economic growth of sub-Saharan Africa has been bleak relative to the other regions of the
world over the past ten years. For example, the average rate of growth of GNP per
capital for the 22 low-income sub-Saharan African countries was -0.4 percent compared to
1.1 percent for low-income countries in Asia for 1970-80 (World Bank, 1981a, p. 3).
Table 1 reveals that 7 countries had a negative average annual growth rate of GNP per
capital during the 1970s.
Turning to the performance of the agricultural sector, of the 22 countries in the
World Bank's category of low-income countries in sub-Saharan Africa, 5 experienced
negative rates of growth of agricultural output over the 1970-79 period (see Table 1).
Three middle-income countries--Angola, Ghana, and Nigeria--also experienced negative
agricultural growth during the 1970s. In only 11 of 31 countries for which data are
available did the average annual growth rate of agriculture exceed the growth rate of
One of the most disturbing trends over the last 20 years has been declining per
capital production of food crops (FAO, 1978). The USDA (1981) points out that sub-
Saharan Africa is the only region of the world where per capital food production declined
over the past two decades. As a result, the average per capital calorie intake was below
minimum nutritional levels in most countries during this period and many countries which
were formerly self-sufficient in food have increased the ratio of food imports to total
food consumption. The declining per capital food production being experienced in most
countries has led to what many observers are now calling an "agrarian crisis" or "Africa's
food crisis" (see Eicher, 1982).

I/The World Development Report, 1981 defines low-income countries as countries
with per capital incomes of $370 or less (in 1979 dollars).

Table 1. Social and Economic Indicators for Sub-Saharan Africa

Average Index
Average Asmnal of Food Life
Population Average Annual Growth Rate of Average Annual Production Expectancy Adult
Size Growth Rate of GNP GNP Per Capita Growth Rate (%) Per Capita at Birth Literacy
1980 Population (W) Per CaPta (Rea (%) of Agriculture (1969-71 100) (Years) Rate (%)
Country (Millions) a 1970-1979 1979 1970-1979 1970-1979 b 1977-1979 b 1979 b 1976 b

Angola 7.1 2.3 440 -2.1 -10.2 83 42
Benin 3.5 2.9 250 0.6 .. 97 47 11*
Botswana .8 2.0c 660 13..
Burundl 4.2 2.0 180 2.1 1.8 103 42 25
Cameroon 8.4 2.2 560 Z.5 3. 110 47
Cape Verde .3 1.9c 260d
Central African Republic 2.3 2.2 290 0.7 2.4 102 44
Chad 4.5 2.0 110 -1.. 0.7 91 41 13
Congo 1.5 2.3 630 0.9 0.1 81 47 50e
Equatorial Guinea .4 2.2c
Ethiopia 31.5 2.1 130 1.3 0.4 84 40 13
Gabon .6 I.1c 3370d 3.c
Gambia .6 3.1c 180d 2.9 .
Ghana 11.7 3.0 400 -0.8 -0.2 82 49 30e
Guinea 5.0 2.9 280 0.3 3.0c 86 44 20
Guinea Bissau .6 1.6c 160d 2. ..
Ivory Coast 8.0 5.5 1040 2.4 3.4 102 47 20
Kenya 16.3 3.4 380 Z.7 3.4 92 33 45
Lesotho 1.3 2.3 340 6.0 1.8 100 51 32
Liberia 2.0 3.3 500 t.6 5.0 101 53 30
Madagascar 8.7 2.5 290 -0.4 0.1 94 47 50
Malawi 6.2 2.8 200 2.9 4.1 100 47 25
Mali 6.9 2.6 140 1t. 4.2 88 43 10
Mauritania 1.6 2.7 320 1.9 -1.4 75 43 17
Mozambique 10.5 2.5 250 0.1 -1.8 75 47
Namibia 1.0 2.8c 1160d .6C
Niger 3.3 2.8 270 -1.3 -1.3 89 43 8
Nigeria 77.1 2.3 670 3.7 -0.3. 87 49
Rwanda 4.8 2.8 200 1.5 .. 107 47 23
Senegal 5.7 2.6 430 -0.2 3.6 88 43 10
Sierra Leone 3.5 2.3 250 0.4 2.3 87 47 1f
Somalia 4.6 2.3 130d 3tc 2.7 85 44 60
Sudan 18.4 2.6 370 0.6 2.7 105 47 20
Swaziland .6 2.3c 80d ...F
Tanzania 17.9 3.4 260 2.3 4.9 94 2 66
Togo 2.6 2.4 350 1.1 0.3 81 47 18
Uganda 13.2 *3.0 290 -.2 3.8 90 54
Upper Volta 6.9 1.6 180 0.3 -3.3 93 43 5
Zaire 28.3 2.7 260 0.7 1.2 90 47 13
Zambia 5.S 3.0 500 0. 2.3 99 49 39
Zimbabwe 7.4 3.3 470 0.8 -0.5 100 33

Source a'United Nations, World Population Prospects, 29 April 1981.
bWorld Bank, World Development Report, 1981.
CFigures are for 1970-78, World Bank, World Development Report, 19BS

dFigures are for 1978, World Bank, World Development Report, 1980.
eFlgures are for 1975, World Bank, World Development Report, 198.
Key.. Not available.

The prospects for economic development in sub-Saharan Africa in the 1980s are not
encouraging. Nearly all African nations are overwhelmingly dependent upon agriculture
for the bulk of their national income, employment, and foreign exchange but only a few
countries such as Malawi, Zimbabwe, and the Ivory Coast have harnessed their agriculture
as an "engine of growth" of the overall economy. The few countries where economic
progress is reasonably optimistic are oil exporting, Nigeria and Gabon. Only a few
countries are not facing a food crisis. But this bleak assessment should not overlook the
major achievements in many countries during the past three decades. A few countries--
Kenya, Malawi, and the Ivory Coast--have achieved respectable rates of economic growth
over the past 10-20 years. The achievements in education have been dramatic in a few
countries such as Tanzania and Somalia. For example, in Tanzania, literacy rates
increased from 10 percent in 1960 to 66 percent in 1975. But adult literacy is below 25
percent in most countries in sub-Saharan Africa.-
Scope of Review
This survey focuses on the rural economy which is broadly defined to include the
agricultural sector plus rural nonfarm activities such as small-scale industry, trade,
processing, and fishing. In addition, we review research on migration, employment, and
income distribution in order to understand linkages between the agricultural and nonagri-
cultural sectors. We have adopted a broad rural economy perspective for our review
rather than focusing on the agricultural sector because research has shown that rural
households in sub-Saharan Africa typically allocate from 25 to 40 percent of their labor to
nonfarm activities.
We begin by identifying standard references on African agricultural development
and presenting a descriptive overview of agricultural systems in sub-Saharan Africa. In
Part II, we examine research on food and agricultural policy in historical perspective
starting with the precolonial period (1800-1880s) and moving to the colonial period which
covered the 1880s-1960 period. We also discuss the key theoretical perspectives on
agricultural development which have had an important influence on policy makers and
scholars working in Africa. We then move in Part III to policy debates during the
postindependence period of the sixties and seventies. In the 1960s, most African
governments stressed industrialization and large-scale farming, including what a World

I/Compared with other regions of the world, the cost of education per pupil as a
percentage of GNP per person is the highest of any region of the world. For example, this
ratio in West Africa is more than twice as high for primary education as compared with
Asia and five times as high for secondary education as for Asia (World Bank, 1980). The
high cost of education per student is largely a function of teachers' salaries which
typically account for 75 percent of educational cost (Hanson, 1980).

Bank mission to Tanzania called the "transformation approach" to modernizing African
agriculture. Despite numerous experiments with large-scale farming and ranching
schemes during the colonial and postindependence periods, the bulk of the land and labor
in sub-Saharan Africa is still devoted to small-scale farming and pastoral systems of
livestock production.
In the late 1960s and early 1970s, many policy makers, planners, and foreign donors
shifted their attention to small-scale farming and small-scale livestock projects. In light
of this shift in policy, we devote substantial attention in Parts IV to VI to micro research
on smallholder crop and livestock production. In Part VII, we shift to research on food and
agricultural distribution systems, reviewing research on topics such as international trade,
marketing, processing, and storage; and research on credit, cooperatives, consumption,
and nutrition. Finally, in Part VIII, we move to research on equity and employment issues
such as income distribution and inequality, population, migration, women in development,
small-scale industry, and fishing.
Several caveats are in order. First, throughout this review, we shall emphasize
research completed since 1970. The literature of the 1960s has already been covered in a
number of surveys, including McLoughlin (1967) for East Africa and Eicher (1970) for West
Africa. Second, we have limited our references as much as possible to journal articles and
books and have deemphasized unpublished reports and theses.
Third, we have found it impossible to provide uniform coverage of the literature by
topic and for each of the 41 countries. We are struck by the unevenness of the research
coverage and the number of publications by country. This unevenness is dramatized by
comparing the number of publications in the references for Niger, Nigeria, Botswana, and
Chad. For example, Sims and Kagan (1976) report that there were 4 American and
Canadian Doctoral dissertations and Master's theses on Niger and 714 for Nigeria over the
1886-1974 period. Shirley Eicher (1981) cites 1,280 publications in her bibliography on
rural development in Botswana from independence in 1966 to 1981. If some countries such
as Chad, Niger, and Rwanda seem to be undercited in our survey, it is partially because of
language barriers and political instability, and partially because the research industry like
the fashion industry is trendy. For example, almost every Western researcher tried to
touch down in Tanzania sometime during the last 15 years to gain some first-hand
impressions of the Ujamaa experiment and the villagization program. The fascination
with Tanzania has produced a vast amount of literature. For example, Kocher and
Fleisher (1979) cite 761 publications in their bibliography on rural development in
Tanzania from independence in 1960 through 1979. For this reason, one can say that
Tanzania has been vastly overstudied relative to countries such as Somalia, the Gambia,
and Rwanda.

Another caveat about the danger of generalization about African agriculture comes
from one of our reviewers--Dr. Francis Idachaba--of the University of Ibadan. Dr.
Idachaba observed that he could find an exception to almost every generalization in the
first draft of this survey. To quote:
Nigeria is an important exception to the recurring treatment of the prices of
cotton and groundnuts relative to the prices of food grains (guinea corn and
millet). It has been the Nigerian experience in recent years for farmers to
shift from cotton and groundnut production to food grain production in
response to higher relative prices of food grains. Nigeria is also an important
exception to the issue of rural underemployment, unemployment, and rural
employment generation. Labour shortage at economic rural wages is the most
important resource constraint facing Nigerian farmers today.l/
Dr. Idachaba's points are well taken. One could find exceptions to every generalization
about research findings or research gaps when one includes 41 countries in a survey. We
agree that there is no such thing as a typical African economy and that it is dangerous to
advance generalizations for a sub-continent four times larger than the United States.
A final caveat is in order when the authors are from the same sub-discipline--agri-
cultural economics. Since agricultural development is a technical, social, and political
process, the literature of any sub-discipline such as agricultural economics will fall short
of capturing the complexity of the development process. Albert Hirschman sums up this
problem in his analysis of the rise and decline of development economics over the past 25
years as follows:
Development economics started out as the spearhead of an effort that was to
bring all-around emancipation from backwardness... By now it has become
quite clear that this cannot be done by economics alone. It is for this reason
that the decline of development economics cannot be fully reversed: our sub-
discipline had achieved its considerable luster and excitement through the
implicit idea that it could slay the dragon of backwardness virtually by itself
or, at least that its contribution to this task was central. We now know that
this is not so; a consoling thought is that we may have gained in maturity what
we have lost in excitement. (1981, p. 23)

IPersonal communication, February 27, 1981.

Standard References
Until the 1960s, social science research on Africa was dominated by anthropologists,
historians, and geographers. Anthropologists--mostly Europeans-were noted for their
ethnographic studies which were largely financed by colonial offices in the 1930s, 1940s,
and 1950s. A handful of agricultural economists and economists started to pursue
research in sub-Saharan Africa beginning in the 1950s and the number greatly expanded in
the 1960s. But the technical and social science knowledge base for agriculture continues
to be sparse and uneven. Except for a few countries such as Nigeria and Kenya,
agricultural research is fragmentary and the scientific knowledge base--especially for
food crops--is 10 to 20 years behind most Latin American and Asian countries.
Despite the generally weak data base, a number of studies over the last 20 years
have been extremely useful to policy makers, scholars, and donor agencies. Foremost is
the two-volume study coordinated by John de Wilde for the World Bank. De Wilde et al.
(1967) drew on information from thirteen agricultural projects; five in Kenya, two each in
Uganda and Mali, and one each in Tanzania, Upper Volta, Chad, and the Ivory Coast. The
first volume, the synthesis, contains information about the distinguishing features of
agriculture, response of farmers to incentives, labor allocation, mechanization, land
tenure, extension, credit, and marketing institutions. Volume two includes case studies.
A classic of the 1960s that unfortunately has received little attention is Jurion and
Henry's (1969) Can Primitive Farming Be Modernized? This book summarized the
extensive research of the Belgian scientists at the INEAC research station which was
established in northern Zaire (formerly Belgian Congo) in the 1930s.
Two important books based on many years of micro-economic research experience in
Africa are Michael Collinson's (1972) Farm Management in Peasant Agriculture and
Martin Upton's (1973) Farm Management in Africa. Collinson draws on many years of
farm level research experience in Tanzania to show how practical farm management
studies can contribute to the needs of extension workers and local planners. Upton's text
stresses the application of production economics to the study of farming systems.
Standard references of the mid-1970s include John Cleave's (1974) African Farmers,
the authoritative volume on labor use in African agriculture, and Uma Lele's (1975) The
Design of Rural Development which summarizes problems encountered in 17 major rural
development projects in Eastern and Western Africa. Valuable sources on the diverse
cropping systems and technical problems are Benneh (1972); Leakey and Wills (1977); and
Morgan (1978). Particular attention should be given to the classic by the late Hans
Ruthenberg, Farming Systems in the Tropics (1980).

Agricultural development strategies and policy issues are covered in edited volumes
by Bunting (1970); McLoughlin (1970); Amann (1973); and Ofori (1973). Recent books with
a policy orientation include a comparative study by Anthony et al. (1979) Agricultural
Change in Tropical Africa; a collection of papers edited by Bates and Lofchie (1980)
entitled Agricultural Development in Africa; Rice in West Africa by Pearson, Stryker,
Humphreys et al. (1981); and a collection of 13 case studies edited by Heyer, Roberts, and
Williams, Rural Development in Tropical Africa (1981).
There are four basic reports that are indispensable for examining the crisis in food
and agriculture in sub-Saharan Africa. The first is FAO's Regional Food Plan (FAO, 1978)
which outlines the background to Africa's food crisis and steps to meet it. The Food Plan
was endorsed by the Organization for African Unity (OAU) in Arusha in 1978 and in
Monrovia in 1979. The second is the Lagos Plan of Action (OAU, 1980) which was
endorsed by the African heads of state when they met in Lagos in April 1980. The Lagos
Plan calls for massive increases in foreign assistance and measures to increase food
production. The third basic document is the USDA's Food Problems and Prospects in Sub-
Saharan Africa (1981). The USDA report discusses trends in the demand and supply of
food over the 1960-81 period. The fourth basic reading is the World Bank's Accelerated
Development in Sub-Saharan Africa: An Agenda for Action (1981).
Overview of Agricultural Systems
There are numerous methods of classifying agricultural systems in sub-Saharan
Africa. Ruthenberg (1980) identifies six approaches, including (1) type of rotation,
(2) intensity of cultivation, (3) water supply, (4) cropping pattern and animal activities,
(5) implements used for cultivation, and (6) the degree of commercialization. Benneh
(1972) argues that since the "pivot of every system of agriculture is the technique used to
restore soil fertility," the method of fertility restoration should be the basis for
classifying agricultural systems. Morgan (1978) classifies systems on the basis of purpose
(commercial versus subsistence), type of management, and enterprise class (single staple
dominant, two or more staple dominant, mixed crops and livestock, and livestock). In this
review, we shall distinguish farming systems and livestock systems since there is little
mixed farming in Africa. While many farmers raise poultry and small ruminants, the large
ruminant livestock economy is based on herding as a distinct activity from farming.

Farming Systems
We identify farming systems on the basis of three characteristics: ecological zone,
intensity of rotation, and major crops.-

-The division into small and large farms is frequently used in the literature and will
be used in this survey.

Ecological Zones
Knowledge of soil and water resources in various ecological regions is crucial for
understanding farming and livestock systems. Ecological zones in West Africa can be
divided into three to seven classifications. A three-tier classification which is commonly
sub-divided into several sub-zones includes the Sahelian, Sudanian, and Guinean zones.-/
The Sahelian zone, also referred to as the sub-desert wooded savanna, has less than
500 mm of rainfall annually and the rainfall tends to be irregular in amount and
distribution. In the northern part of the Sahelian zone, the rainy period is frequently less
than 55 days, the minimum period needed for settled rainfed farming. While the northern
Sahel is unsuitable for farming, herders migrate throughout the zone searching for pockets
of grass which grow in riverbeds after the sporatic rains. Throughout the Sahelian zone,
the open water evaporation rate generally exceeds 2,000 mm per year, in part attributable
to desiccating winds such as the harmattan which originate in the Sahara in the dry season
and move southward.
The Sudanian zone (or arid wooded savanna) is characterized by open grassland and
an annual average of 500 to 800 mm of rain. This zone has favorable climate and soils for
farming, particularly the southern part of the zone which is intensively farmed. Areas
with rainfall exceeding 650 mm are considered a transition zone from the arid to humid
wooded savannas and are often referred to as the northern Guinean zone.
The Guinean zone proper has rainfall exceeding 1,000 mm per annum with the
rainfall distributed over a 120- to 190-day period. This zone is the most extensive in West
Africa and it has high agricultural potential. The Guinean zone includes the sub-arid
wooded savanna, sub-humid savanna, and the humid forest zone. The southern part of the
Guinean zone, commonly referred to as the sub-humid tropics or forest zone, is
characterized by heavy rainfall during most of the year and reliance on swamp rice and
root and tuber crops (Leakey and Wills, 1977). Parts of the Guinean zone are further
classified as derived savanna. The derived savanna refers to former forest land which has
been irreversibly transformed into a semi-wooded landscape as a result of farming
The above classification of ecological zones applies to a wide band across West
Africa south of the Sahara. There is not a comparable pattern of ecological zones in
Eastern and Central Africa where the zones are greatly influenced by variation in

-The three-tier classification scheme has been refined by several researchers, most
notably by Phillips (1959) who uses a seven-zone classification: Desert-Southern Saharan
Fringe, Sub-Desert Wooded Savanna, Arid Wooded Savanna, Sub-Arid Wooded Savanna,
Mild Sub-Arid Wooded Savanna, Sub-Humid Wooded Savanna, and the Derived Savanna.
The northern Guinea is the most intensively farmed zone of West Africa because rainfall
is more reliable than more northern zones and the zone remains free of tsetse.

altitude. The semi-arid belt of the Sahelian and Sudanian zones in West Africa does
continue into southern Sudan and south through Kenya and Tanzania, but the distribution
of rain and its seasonal character changes (Ahn, 1977). In many of the highlands of East
Africa and Rwanda and Burundi, soils are of volcanic origin and rainfall is generally
sufficient to grow a wide range of crops. In contrast to West Africa where there is one
rainy season from May to September, in the East African savanna rainfall has a bi-modal
distribution which is variable in amount and length.

Intensity of Cultivation
The intensity of cultivation can be used to distinguish among shifting cultivation,
fallow, and permanent cultivation systems. A simple criterion for classifying the
intensity of a system is the relationship between crop cultivation and fallowing.
Following Ruthenberg (1980), the index of land use intensity (R) is defined as the years of
cultivation multiplied by 100 and divided by the sum of the number of years in cultivation
plus the number of years in fallow. A value of 0 is given for land in permanent fallow and
a value of 100 for continuous cultivation.
Shifting cultivation- is a system of growing crops for a few years on selected fields
and then allowing the land to rest and regenerate for a lengthy period--often exceeding
20 years (Ruthenberg, 1980). Shifting cultivation is predominantly practiced in humid and
semi-humid climates such as in the forests of Central Africa and in the lowlands of West
Africa. Migration is an integral feature of systems of shifting cultivation and nomadic
pastoralism as households move periodically to more fertile land.
Fallow systems are generally defined as those in which one-third to two-thirds of
the land is cultivated each year (the land use intensity (R) ranges between 33 and 66).
Landholdings are usually clearly defined in fallow systems and the rural households have
permanent housing or they move occasionally over short distances. The economic
importance of fallow farming systems is far greater than that of shifting cultivation
systems. A modified form of fallow cultivation systems found throughout sub-Saharan
Africa is the ley system. Ruthenberg defines a ley system as one in which several years
of arable cropping are followed by several years of grass and legumes for livestock.
Regulated ley systems, which may include both pasture management and planting of grass
leys, are found on large farms in Zambia, Zimbabwe, the Kenyan highlands, and in some
settlement schemes in other countries.

/"Shifting cultivation" is a particular type of swidden agriculture in which
homesteads are frequently moved so as to remain close to fields which are only cultivated
for a few years. Standard references on shifting cultivation are De Schlippe (1956), Nye
and Greenland (1960), and Allan (1965).

In systems of permanent cultivation, arable crops are only rarely interspersed with
fallow or leys. A land use system is considered permanent cultivation when the land use
intensity (R) exceeds 70. The major problem of permanent cultivation systems is
maintaining soil fertility. Efforts to maintain yields in the face of higher land use
intensity have forced farmers to modify their land use practices. One of the most widely
noted responses to a reduction in soil fertility is to farm fields closer to the compound
more intensively, using household wastes and animal manure to enrich the soil.

Major Crops
The heterogeneity of ecological conditions in sub-Saharan Africa is reflected in the
diversity of cropping patterns. In general, one or two staple food crops dominate cropping
systems in each ecological zone and can account for as much as 70 to 80 percent of the
area cultivated. After giving priority to planting their dominant staples, African farmers
rely on a strategy of diversified production, incorporating food crops, cash crops, and non-
farm activities. While ecological conditions, principally the amount and distribution of
rainfall, determine the dominant crop or crops in any region, there is variability in the
combinations of secondary crops grown in any village, by given farmers.
Cereal grains are the most important crops grown in the semi-arid regions of sub-
Saharan Africa, both in terms of acreage and calories. The major cereal grains are
sorghum, millet, maize, rice, wheat, and teff (in Ethiopia). In most parts of Africa, grains
are intercropped with secondary crops such as cowpeas, beans, and vegetables. Intercrop-
ping (mixed cropping) entails growing two or more crops at the same time on the same
Sorghum and the various millets are the most widely grown crops. Sorghum is
concentrated in a belt south of the Sahara where rainfall ranges from 600 to 1,000 mm. a
year. In northern Nigeria, for example, sorghum and millets are grown on over 50 percent
of the area cultivated (Beeden et al., 1976). In the Savanna region of West Africa,-
sorghum is generally grown in a bush fallow land use system and is frequently intercropped
with millet. Sorghum is generally dominant in areas susceptible to both drought and
flooding while millets, which are extremely drought resistant and tend to store better
than other cereal grains, play a major role in zones receiving low and uncertain rainfall.
Even in cases where sorghum and millet are not intercropped, farmers often grow several
varieties of each crop on different fields in the same year.

-The term the "Savanna" zone or region encompasses the semi-arid regions of West
Africa, including the Sahelian zone, the Sudanian zone, and the northern parts of the
Guinean zone where cereals are the main staple.

2Some advantages of growing multiple varieties are: (1) security against crop
failure, (2) diet diversity, and (3) spreading seasonal labor requirements.

Maize is the staple grain in many parts of Southern and Eastern Africa. Maize is the
most important food crop grown in Zambia, Tanzania, Malawi, and Kenya and it is
increasing in West Africa as a replacement for sorghum and millet (Taylor and Bailey,
1979). Maize is also grown as an important secondary crop in many areas, particularly
forest zones, and is eaten fresh. Under favorable conditions, maize tends to outyield
other cereal grains and is less susceptible to attacks by birds than sorghum and millet.
Maize has been found to do particularly well in West Africa in Cameroon, Togo, and
While sorghum, millet, and maize continue to be the primary staple food grains, rice
and wheat are increasing in importance throughout Africa. Madagascar is Africa's largest
rice producing nation. Rice is also pervasive in West Africa where it is the staple food in
Liberia, Sierra Leone, and parts of Ghana. For example, a 1974/75 survey in Sierra Leone
revealed that rice was grown on over 97 percent of the small farms (Spencer and Byerlee,
1976). Rice is grown in four major cropping systems: upland or rainfed, in paddies, in
mangrove swamps, and on bottomlands (bas fonds). Upland rice, which accounts for 75
percent of the rice cultivated in West Africa, is generally grown in a bush fallow system
and is frequently intercropped with maize, millet, and cassava. Most rice is grown as a
subsistence crop. Rice imports have increased dramatically over the last decade in
response to growing urban demand. In 1980, for example, West Africa imported 1.6
million tons of rice. A number of studies are now focusing on the reasons for the shortfall
in production in a region where land is abundant and unused (Pearson et al., 1981).
The major wheat producing areas are in Ethiopia, Kenya, Tanzania, Lesotho, Zim-
babwe, and the Sudan. In these countries, wheat is often the major staple in highlands
above 2,100 meters. The majority of the wheat is grown on large farms and is sown and
harvested mechanically. In lower areas, there has been little expansion of wheat acreage
because of low and unreliable yields. Wheat production is being rapidly expanded on
government-directed irrigation schemes in the Sudan, northern Nigeria, Zambia, and in a
number of other countries, partly for political reasons and partly because foreign aid
encourages wheat production in countries such as Zambia and Tanzania (Freeman, 1980).
In humid forest zones, farmers rely on plantains, roots, and tubers rather than grain
for most of their calories. Cropping systems in forest zones are usually dominated by one
or two crops. Plantains are the major staple food grown in parts of Uganda, Tanzania,
Ghana, Ivory Coast, and Zaire. Plantains may be planted in pure stands but often are
intercropped as shade for coffee or with food crops such as maize or beans.
Root crops, primarily cassava (manioc) and yams, are widely grown throughout sub-
Saharan Africa. Cassava is growing in importance because it requires less labor than
yams and cereals, can be grown in a wide range of climatic conditions, has low soil
nutrient requirements, is resistant to pests, and is regarded as a "famine crop" because

some varieties can be left in the ground for several years with no significant deterioration
in quality. Many high-yielding varieties of yams are being grown in West Africa,
particularly in Nigeria where 90 percent of the West African output and two-thirds of the
world output of yams is grown. Yams are labor-intensive and declining in importance in
some regions because of rising wage rates.
The major grain legumes are groundnuts (peanuts), cowpeas, and pigeon peas. Cow-
peas, which are the most commonly eaten legume, are generally intercropped with cereal
grains. Groundnuts are an important cash crop throughout the semi-arid regions but they
are also eaten boiled or roasted. Soybeans are a minor crop but are of increasing
importance. There are problems of consumer resistance to soybeans and in storing
soybean seeds in the dry season in semi-arid regions.
Since the colonial period, African farmers increasingly have produced certain crops
primarily for sale (cash crops). Groundnuts and cotton are important cash crops in the
semi-arid regions. Cotton plays a major role in Uganda, Tanzania, Malawi, the Sudan,
Mali, Upper Volta, Senegal, and northern Nigeria. The most important perennial cash
crops are oil palm, cocoa, coffee, tea, pyrethrum flowers,- and rubber. Oil palm is
important in Zaire and in the rain forest of West Africa, particularly in the Ivory Coast
and Nigeria. Cocoa is an important crop in Ghana, Nigeria, Cameroon, and the Ivory
Coast. Coffee is grown by small farmers in Uganda, the Ivory Coast, Cameroon, Kenya,
and Ethiopia. Coffee in West Africa is mostly of the Robusta type which is of lower
drinking quality and is primarily used for making instant coffee. Most of the coffee in
Kenya, Uganda, and Ethiopia is Arabica and is considered to be some of the best quality
coffee produced anywhere in the world. Kenya is the center of a flourishing smallholder
tea industry.
Farmers grow vegetables such as okra, onions, tomatoes, peppers, and leafy greens
on small garden plots to supplement the staples in their diets. Vegetables may also be
intercropped with staples. Vegetables are frequently grown on small plots along riverbeds
during the dry season. While labor inputs are high for vegetable production, the returns to
labor are about the same as most staple crops because of the high value of vegetables.
Some vegetables are air freighted to Europe during winter months (FAO, 1976c).

Livestock Systems
Livestock is a major industry in Mauritania, Mali, Niger, and Chad in West Africa;
Ethiopia and the Sudan in Eastern Africa; and Botswana in Southern Africa.- Cattle are

-An environmentally safe but highly toxic pesticide which is grown mainly by small
farmers in Kenya, Tanzania, Rwanda, and Zaire.

2/References and bibliographies on livestock are R. E. McDowell (1972); Dahl and
Hjort (1979); Ergas (1979); and ILCA (1978, 1979a,b,c,d, 1980a,b).

the most important type of livestock in terms of milk and meat consumption but small
ruminants (goats and sheep) are dominant in a few countries. For example, small
ruminants account for about one-half of red meat consumption in rural Mali (Delgado,
1980) and in the humid zone of Nigeria (ILCA, 1979a).
Sheep and goat production are attractive to small farmers because of low initial
investment (relative to cattle) and the corresponding small financial loss with the death of
individual animals. Small ruminants can be fed roughages and crop by-products, managed
by women and children, and sold to pay school fees and to purchase grain and cloth. Until
recently, sheep and goats were neglected by both technical researchers and social
scientists (McDowell and Bove, 1977; ILCA, 1979a).
Donkeys and camels are primarily used for transportation and traction in the semi-
arid and arid regions. Camels are an important source of milk in some countries. There is
relatively little swine production in Africa, particularly in Moslem areas. Poultry are
common in rural households and are increasingly produced in specialized units around
large cities.
There are many arbitrary methods of classifying livestock production systems. We
shall use the following: (1) nomadism or total pastoralism, (2) semi-nomadism cr semi-
pastoralism, (3) sedentary pastoralism, and (4) ranching--private, group, association,
cooperative, and state.
Nomadic households are defined as those which do not have a permanent place of
residence or practice regular cultivation. Nomadic and semi-nomadic households have
historically migrated through arid and semi-arid regions in search of water and grazing
land and have played a major role in regional trade--especially in West and East Africa
(Salzman, 1980). Pastures are traditionally considered to be available to all members of
an ethnic group or are rented under land use agreements. The priority of nomads is to
feed their families a continuous supply of milk and meat from their herds. Nomadic
pastoralism is presumed to be declining-1/ in most countries because of (1) the loss of
prime grazing areas to semi-nomadic herders and to farmers, (2) arid livestock zones have
become relatively overpopulated, forcing a reduction in distance covered by transhu-
mance,- and (3) recurring drought.

/But Stephan A. Sanford of the O.D.I. points out that it is unwise to generalize
because to test this proposition one needs data on the (a) area used by nomads, (b) number
of nomads, (c) number of livestock, and (d) length of nomadic moves. Moreover, the
length of moves and the area of land used by nomads vary enormously from year to year
depending on rainfall (personal communication, March 17, 1981).

Z/Transhumance refers to the pattern of regular movement of cattle and herders in
search of grazing land. For a classic study of transhumance in northern Nigeria and
Niger, see Stenning (1957).

Semi-nomadic pastoralism is a system in which households establish a permanent
place of residence which is kept for several years. Crops are cultivated as a
supplementary food source but herds are moved on transhumance to assure sufficient
forage and water. But in some semi-arid areas, nomads have acquired use rights to dry
season grazing areas such as in the Niger Delta in Mali (Gallais, 1975).
Sedentary pastoralism and mixed farming are systems of integrated crop and
livestock production. Sedentary pastoralism refers to permanent settlers who grow a few
crops but rely on livestock production as their dominant enterprise. Mixed farming refers
to a system whereby crop farmers add a livestock enterprise either for fattening or for
the use of oxen in cultivation and the subsequent fattening of the cull work oxen. The
integration of crops and livestock has a long history dating from schemes introduced by
colonial governments in the 1920s. Mixed farming is expanding but its success is crucially
dependent upon the presence of a cash crop such as cotton and groundnuts.
Systems of stratified production and ranching are alternatives to the dominant
system of pastoral production (FAO, 1977). Stratification involves removing young male
stock from the arid and semi-arid pastoral areas for fattening in the intermediate rainfall
zones and eventually for sale in urban centers. Stratification and industrial feedlots are
not widespread in West Africa because of the risk involved in securing a steady supply of
animals for fattening and in the fluctuation of the price of the finished steers.
Ranching has been practiced for decades in Zaire, Angola, and Eastern and Southern
Africa, and to a lesser degree in some countries in West Africa. During the 1960s and
early 1970s, ranching was endorsed by many experts and donors because ranching was
thought to (1) prevent overgrazing through the control of the number of animal units and
the practice of on-ranch rotational grazing, (2) prevent spreading of contagious disease,
(3) guarantee a supply of cattle to markets, and (4) provide a demonstration function for
small herders. But in recent years, the push to establish ranches has slowed in light of
evidence that the number of jobs created and the returns on investment are low
(Von Kaufmann, 1976). Moreover, concerns for equity are encouraging governments to
help small herders form ranching associations and group ranches (Ayuko, 1980, 1981).
Dairy production is concentrated in the eastern African region in Kenya, Tanzania,
Ethiopia, and the Sudan.I/ Although colonial policy in Kenya reserved dairy production
for large-scale farmers (mostly white), smallholder milk output has grown rapidly over the
past 20 years and it now accounts for 40 percent of Kenya's milk output. Some nomadic
groups such as the Fulani in western Africa supply milk to local villages.

!/The small number of economic studies of dairying include Zalla's (1981) study in
Tanzania and Hopcroft and Ruigu (1976), Ruigu (1978), and Stoltz' (1979) studies in Kenya.


Since all the countries in this review, with the exception of Ethiopia and Liberia, are
former colonies of France, England, Germany, Belgium, or Portugal, we shall begin by
examining precolonial economic activity and the links between colonial policies and
agricultural policy during the postindependence period. We then present an overview of
the main theoretical perspectives influencing scholars and policy makers. There has been
a lag in efforts to develop theories of development based on African values, institutions,
resource endowments, and data. Currently, two streams of thought--western develop-
ment economics and dependency/political economy models--dominate political debates,
research, and policy analysis in Africa.

Historical Perspective
Precolonial Economic Activity: 1800 to 1880s
Until recently, African history was essentially colonial history and, more specifical-
ly, the political history of the colonial powers starting around the 1880s. Economic
historians devoted little attention to the precolonial period because they assumed that the
development of Africa began with the imposition of colonial rule. Many scholars such as
Peter Bauer (1975) of the London School of Economics have argued that modern social and
economic life in Africa began with colonial rule and the creation of trade linkages with
Europe. But over the past 20 years, there has been an explosion of research on the history
of the precolonial period from 1800 to 1880. There is now convincing evidence that
extensive economic activity was taking place throughout Africa prior to the beginning of
colonial rule in the 1880s. The standard reference to the economic history of the
precolonial and colonial periods is McPhee's The Economic Revolution of British West
Africa (1971). Although McPhee called the rapid expansion in export crops a "revolution"
which was facilitated by the imposition of British rule starting in the 1880s, he carefully
traced the origins of the revolution to the 1820s.
A Swedish economic historian, Sundstrom has made a major contribution to the
economic history of the precolonial period. Sundstrom's book was published in Swedish in
1964, going virtually unnoticed until it was translated and published in English (1974). He
drew on English, French, German, Dutch, Portuguese, and Swedish sources in his
penetrating analysis of the internal trade of salt, textiles, iron, copper, and brass in West,
Central, and East Africa from the eighteenth to the twentieth centuries. Evidence that
internal trade was widespread, and of great antiquity, is also supported by Dike's (1956)
study on oil palm trade in the Niger delta of southern Nigeria in the nineteenth century

and Lovejoy's (1980) research on long-distance kola nut- trade in the nineteenth century.
A. Cohen (1971) traces the origins of long-distance livestock trade in West Africa to the
eighth century and highlights the central role of Islam as the binding force of a network of
traders scattered over a wide area encompassing several countries. Baier (1977) points
out that for many centuries long-distance traders have played a major role in the
economies of West Africa by moving salt, dates, and livestock from the desert and
savannah zones into the more fertile land on the coast and grain, cloth, gold, and
manufactured articles from the coast to the savannah and desert zones. The net impact
of the above studies is that they have essentially demolished the view that economic
development in sub-Saharan Africa began with the arrival of colonial governments in the
late 1880s.
A. G. Hopkins, an economic historian at the University of Birmingham, established
himself as a controversial and influential scholar with the publication of An Economic
History of West Africa (1973). Hopkins used neoclassical economic theory and a modified
formalist paradigm of economic anthropology in his analysis of the interaction between
internal trade and external market forces over the 1800-1950 period. Hopkins' grand
theme was that market forces played a key role in integrating West Africa into the world
economy. Although colonial rule did not begin until the 1880s, Hopkins pinpoints the large
increase in agricultural exports in the early 1800s as the beginning of "modern West
African agricultural development." Hopkins' book has been acclaimed by scholars in
numerous disciplines but attacked by George Dalton, a Northwestern University economic
anthropologist and leader of the substantive paradigm of economic anthropology. In
Dalton's (1976) 50-page review of Hopkins' book, he criticized the emphasis on the market
as the major force in development in the nineteenth century. Dalton contends that
reciprocity and redistribution were also important forces in the precolonial and colonial
The slave trade is at the heart of West African history. The late Walter Rodney
argued in How Europe Underdeveloped Africa (1974) that a European imposed pattern of
exchange of slaves for manufactured goods from the 15th through the 19th centuries was
the cause of the underdevelopment of Africa. Rodney argued that the slave trade reduced
Africa's population growth rate to zero and caused a severe technological arrest (although
Rodney neglected to explain how population growth leads to technical advance).
Now that economic historians have documented the presence of extensive interna-
tional trade and the growth of cash crops before the advent of colonial rule, there is a
need for micro research on the agricultural history of particular crops, the origins and

/A nut with a caffeine base. Kola is chewed as a stimulant and used in ceremonial
functions such as welcoming guests.

dynamics of internal trade, and the linkages between internal and international trade.
Carolyn Barnes' study (1979) of the agricultural history of coffee production among the
Gusii in Kenya over the 1933-1948 period and Baier's analysis (1980) of the economic
history of the central region of Niger from 1850-1960 are examples of the type of in-
depth research which is needed.

The Colonial Period: 1880-1960
An analysis of agricultural development in the postindependence period since 1960
should be rooted in an understanding of colonial strategies. Colonial strategies varied
widely throughout Africa and it is difficult to generalize about the impact of these
strategies on the countries included in this review. British colonial policy in Kenya, for
example, promoted extensive European settlements. Sorrenson's (1967) assessment of
Kenyan agriculture during the colonial period reveals that much of the best land was
reserved for Europeans starting in the late 1890s. Head taxes were introduced to
encourage small farmers to produce cash crops and to sell their labor to European
plantations and mines. On the other hand, British colonial policy in Nigeria, Ghana, the
Gambia, and Sierra Leone sharply restricted plantation development and settlement by
white farmers. In fact, British colonial policy in Nigeria prevented private plantations
from gaining long-term control over land and large firms such as Unilever gave up on
trying to establish plantations in Nigeria and eventually opened plantations in the Belgian
Congo (now Zaire). For a reference on British colonial policy in Africa, see Hancock
In contrast to British policy which restricted settlers and plantations in Nigeria,
French policy encouraged Europeans to establish plantations to grow coffee and cocoa in
the Ivory Coast (B. Campbell, 1974). In addition, the decrees of 1925 regulated forced
labor and ensured a steady flow of Ivorian workers for the European plantations.
Gradually, Ivorians started to grow coffee and cocoa on small plots scattered throughout
the forest but since they used European techniques they were called planters. The Ivorian
planters led the drive to independence after World War II and the planters remain a
powerful political force in the Ivory Coast today. For a discussion of French colonial
policy in West Africa, see Newbury and Kanya-Forstner (1969).
In Tanganyika (now Tanzania), the thrust of German colonial and commercial policy
from 1905-1912 was to develop plantation agriculture. Hut and poll taxes were imposed
to force Africans to provide wage labor for the plantations. Iliffe (1979) has shown that
by the end of German rule in 1912 an export orientation was firmly established.
An important assessment of colonialism is Duignan and Gann's collection of papers,
The Economics of Colonialism (1975). Papers include an analysis of trade policy by Meier,

the emergence of cash crop exports by Hogendorn, industrialization by Kilby, French
colonial policy by Thompson, agricultural research by Yudelman, and British colonial
policy by P. T. Bauer and a synthesis by the editors, Duignan and Gann. The editors and
most of the authors generally present a procolonial assessment of the contribution of
colonialism to African development. For example, Gerald Meier (1975) argues that the
slow rates of economic growth during the colonial period should not be attributed to the
fact that exports were primary products, rather it might have been due to the "absence of
more active policy making" by colonial governments. Studies of the colonial period
include Hill (1963), Suret-Canale (1971), Wrigley (1965), de Wet (1977), Brett (1973),
Dorward (1975), Yansane (1976), Baier (1977), Dumett (1971), Green and Hymer (1966),
Brooks (1975), Howard (1978), and Kitching (1980).
Since agricultural development programs in the postindependence period have been
directly or indirectly influenced by colonial policies, approaches, and attitudes about
agriculture's role in development, it is important to examine the following4-
(1) The degree to which Africans were excluded from or "forced" to participate in
selected colonial development programs;
(2) The colonial record on the training of Africans;
(3) The colonial position on promoting research on export vs. food crops;
(4) Who benefited from colonial land grants and the surpluses generated from
export agriculture?
A growing number of scholars have revealed that Africans were systematically
excluded from participation in many colonial development schemes and in producing
certain export crops and improved cattle. R. E. Baldwin's (1966) study of export growth in
Zambia (then Northern Rhodesia) from 1920-1960 reports that "the agricultural policy for
most of the period covered was designed to benefit European settlers. African farmers
were either ignored or discriminated against when their interests conflicted with those of
the European population ... and those few agricultural measures specifically directed
toward helping African farmers were often poorly conceived and ineffective" (R. E. Bald-
win, 1966). Reviewing colonial policy in Kenya, Heyer and Waweru (1976) state that
"there was an effective prohibition on African coffee growing until 1933 when it was
allowed on a very limited and experimental scale in three districts relatively far removed

-/Several other issues also could be examined including the colonial infrastructure
strategy which developed railroads and roads to link favorable natural resource zones to
coastal trading centers. This explains why there are still no major rail and road links
along the coasts of East and West Africa.

from European coffee growing." For further evidence on how colonialism excluded
Africans from growing a number of crops and from producing improved cattle, see
Uchendu and Anthony (1975a,b).
A number of scholars have documented how the colonial governments compelled
village chiefs to force farmers to grow selected crops (e.g. cotton), or to "contribute"
labor to maintain roads. Magasa (1978) presents a damming account of the repressive
policies adopted by the French colonial service to recruit labor for government projects
such as the Office du Niger irrigation scheme in Mali. The colonial legacy of top-down
approaches to agricultural change is still present in agricultural ministries in many
independent African countries.
Colonial governments gave little attention to the training of Africans. McKelvey's
survey of "Agricultural Research in Africa" (1965) reports the almost total failure of
colonial governments to develop institutions for training African agricultural scientists
and managers. For example, by the time of independence in the early 1960s, there was
only one Faculty (College) of Agriculture in French-speaking tropical Africa. Between
1952 and 1963, only four university graduates in agriculture were trained in Francophone
Africa and 150 in English-speaking Africa. Johnston (1964) observed that there were only
three African scientists working in all the experimental stations in the East African
countries of Kenya, Uganda, and Tanzania in 1964. Lele (1981) reports that training has
been given such low priority during the 20-25 years of postindependence that Africa has
failed (unlike India) to develop a proagricultural lobby of professional agricultural
scientists, planners, and administrators.
One of the most important legacies of the colonial era is the bias against research
on food production and on small farmers and herders. In numerous countries, colonial
regimes focused their research and development programs on export crops, plantations,
and land settlement schemes. For example, Hunt (1975a) reports that millet yields in
1972-73 in Kenya were at best only 50 percent of the maize yields and that "No
agronomicc) work has been done on developing improved bulrush millet or finger millet
strains" in contrast to a large amount of research on maize which was initiated during the
colonial period to benefit the white settlers in the highlands areas with high rainfall.
Heyer (1975) reports that colonial policy in Kenya, which established export crops in areas
with the most favorable natural resource endowments, helps to explain some of the
regional inequalities and the fragmented road and railroad systems. Although the number
of European farms was never very large in Kenya, reaching a maximum of 3,600 holdings
at the end of the 1950s, Heyer reports that "the development of Kenya's agriculture was
profoundly affected by the presence of white farmers" because at the end of the 1950s

European farms were responsible for an estimated 80 percent of the agricultural output
that reached urban and export markets (1981, pp. 93-94).
Economic historians have only recently started to grapple with the roots of rural
poverty in Central and Southern Africa. Palmer and Parson's collection of essays, The
Roots of Rural Poverty in Central and Southern Africa (1977), has been described by
Ranger (1978) as a "landmark in African agricultural and peasant studies." The theme of
the book is the alleged "strangulation" of peasant smallholderr) farming by the capitalist
market forces which first stimulated peasants to produce for the market and sell their
labor to the mines but later resulted in the exploitation of farmers through the taxation of
export crops. Although Ranger lauds Roots, he urges historians to shift from archival
research to more field studies and oral history to document how small farmers in Central
and Southern Africa were affected by specific colonial policies. Another important study
is Bundy's The Rise and Fall of the South African Peasantry (1979).
Any definitive assessment of the colonial period is faced with the unanswerable
question of what would have happened in the absence of colonialism. Until recently,
numerous scholars accepted at face value the development programs initiated in the
colonial period, while ignoring the broader question of possible exploitation. For
example, two scholars recently lamented the narrow approach to their doctoral disserta-
tion research. Brett, a U.K. political scientist, reflected on his doctoral dissertation
research on colonialism in Kenya as follows:
In London I worked in the general ethos of the Institute of Commonwealth
Studies where the tendency in Colonial History was towards a thorough
empiricism in a framework which did not question the overall validity of
Britain's colonial contribution to African development except where extensive
resettlement (e.g., encouragement of white farmers to settle in Kenya) had
been allowed. This environment certainly encouraged me to take at face value
a great deal in the material which I was given and did not lead me to make any
serious attempt to relate it to the broader questions of colonialism and
underdevelopment which have now forced themselves into the center of all
serious work in this general field. (1973, p. ix)
An American anthropologist, Sudarkasa, wrote in the preface to a book based on her
dissertation on market women in Nigeria:
When I wrote this study a decade ago, I tacitly gave support to the social
science fiction that what happened in Nigeria (and in all of Africa) in the 20th
century could be sanguinely described in terms of 'modernization', and that the
processes of European entrenchment and exploitation in Africa could be
subsumed under the benign if not indeed benificent concept of 'development'.
The fact that I did not delve into the factors underlying the conditions
described in this study is perhaps an indication of the success of my training in
Western social science. In any case, suffice it to say that I would bring a
decidedly different perspective to such an undertaking if I were doing it today.

In summary, although the full impact of the colonial policies on contemporary
Africa will require further research, the debate has been joined. Research on the colonial
period by neo-classical economists is now being supplemented by a growing number of
political economy studies of trade, capital formation, and the uses of surpluses. There is
a rich agenda for research by economic historians of diverse ideological positions. For
example, why has export growth in some countries, e.g., the Ivory Coast, been cumulative
and why do other countries such as Ghana, Zambia, and Senegal remain dependent on one
export crop after 60 years of export trade? Future research on the colonial period is
likely to show that the net effect of the colonial policies and programs on the peoples of
sub-Saharan Africa will be somewhere between Peter Bauer's assertion that "virtually all
components of modern social and economic life in Africa date from the colonial period"
(1975, p. 653) and Kwame Nkrumah's assessment that "without exception they (the
colonial powers) left us nothing but our resentment" (1963, p. xiii).

Transition to Independence
As African nations became independent in the late 1950s and early 1960s, most of
them pursued mixed economies with a heavy emphasis on foreign aid,-' nation-building,
industrial development, education, and diversification of their economies. A small
number of countries such as Mali, Ghana, and Guinea shifted abruptly to revolutionary
socialism in the early 1960s. But whether political leaders were espousing capitalism or
socialism they generally all gave low priority to agriculture. African leaders generally
viewed agriculture as a "backward" sector which could provide agricultural surpluses--
taxes and labor--to finance structural change and industrial/urban development. Agricul-
tural policies in many capitalist and socialist countries supported plantations, state farms,
land settlement schemes, and the replacement of private traders and money lenders with
government trading corporations and credit agencies. Moreover, the empirical record
shows that all countries under all types of governments--civilian, military, capitalist, and
socialist--have exploited and controlled the agricultural sector via (1) harsh taxation
policies and the underpricing of agricultural commodities; (2) concentration of secondary
schools, hospitals, and other social services in the capital city, or several large cities;
(3) top-down pattern of government-imposed agricultural development schemes; and
(4) unwillingness to transfer the administration of marketing, storage, and credit programs
to groups of farmers. Among the reasons for the low priority given to agriculture in the

/For example, one of Africa's most respected economists, Tom Mboya (1967) of
Kenya, laid out a development strategy for Africa which called for "a massive inflow of
capital over perhaps 30 years and an equally massive inflow of technical assistance
personnel over 10 to 15 years." Kenya's President Kenyatta encouraged investors "to
bring prosperity" to Kenya.

1960's were the following: (a) industrialization was perceived to be the most expedient
way to bring about structural change, a high rate of economic growth, and economic
independence; (b) investment in food production was assumed to be unnecessary because
of low man-land ratios and surplus land; and (c) export crops were perceived to be
"colonial" crops which contributed to dependency and price and income instability.
During the last decade, almost all African governments have shifted from industrial-
ization, export promotion, and agricultural transformation to espousing the multiple goals
of food self-sufficiency, improved nutrition, diversification of their economy, increased
trade within Africa, economic growth, and increased access of the poor to employment,
income, and social services. While nearly all African governments espouse the primary
goal of self-sufficiency in food production, objectives such as self-sufficiency are free. It
takes resources to increase food production and it will involve some difficult produc-
tion/equity conflicts (e.g., conflicts between regions with good agricultural potential and
regions with poor potential). Thus, as African nations enter the third decade of the
postindependence period, the 1980s, there is generally little debate over the goals of
development. The contentious issues are priorities, financing, time frame, the role of
agriculture, and the state in national development.

Theoretical Perspectives
Scholars working within Africa on development problems over the past 30 years have
been caught in a cross-fire of imported models and theories of development. The lag in
developing models of development by Africans and based on African data and institutions
is linked to the token role which colonial governments gave to the training of Africans for
positions in universities and research organizations.-/ As a result, the production of
knowledge about African development still remains to a substantial degree in the hands of
expatriates except in a few countries such as Nigeria and Kenya. Moreover, expatriates
dominated planning agencies and social science research institutes in Africa throughout
the 1960s and most of the 1970s. For example, the Economic Commission for Africa
(ECA) in Addis Ababa relied heavily on expatriate advisors such as A. F. Ewing and Rene
Dumont to provide the vision for ECA in the 1960s. But these ECA advisors failed to
articulate feasible development strategies for an agrarian-dominated continent. For
example, A. F. Ewing reported that "industry is the sole means of raising the productivity
of an economy" (1968, p. 11). An influential UN/ECA/FAO report on agriculture called
for "an accelerated movement into market agriculture through government measures

-IThe lack of African staff is still a critical problem in African universities. For
example, in August 1981, a total of 14 of the 16 resident faculty members in the
Department of Economics at the University of Dar es Salaam were expatriates.

aimed at both individual producers and large-scale projects" (1964, p. 39). But as we point
out later, large-scale projects have generally been ineffective in achieving both efficiency
and equity objectives.
The Executive Secretary of the United Nations Economic Commission for Africa,
R. K. A. Gardiner, requested the noted French agronomist, Rene Dumont, to undertake a
survey in order to initiate debate on how to transform African agriculture. In his report
entitled African Agricultural Development (UN/ECA/FAO, 1966), Dumont recommended
cooperative farming and lifelong claims to land rather than private ownership of land and
he urged African nations to learn from the Socialists' experiences in Eastern Europe,
USSR, China, and Cuba. But throughout the report, Dumont moralized about the need for
Africans to work harder and the need for austerity, integrity, education, and exemplary
moral qualities.-
The contrast between Africa and Latin America in producing knowledge about
development problems and strategies is striking. In Latin America in the 1950s and 1960s,
Latin American scholars such as Prebisch, Dos Santos, Furtado, Sunkel, and Pinto
produced a number of seminal papers and reports for the United Nations Economic
Commission for Latin America (ECLA).2' These scholars, under the leadership of
Prebisch, came to the conclusion that models on development based on the experience of
high-income countries in Western Europe and North America, including "classical" Marxist
models, were not relevant to Latin American conditions and that attention should be
directed to developing theories and models based on Latin American conditions. More-
over, Prebisch's (1959) influential thesis that the world economy is rigged against the
Third World--declining terms of trade--was one of the central issues in the North-South
meeting of 22 Heads of State in Cancun, Mexico in October 1981. The Prebisch thesis
remains an article of faith in the Third World.
As African nations became independent in the early 1960s and Western economists
assumed important roles in helping to prepare development plans and serving as policy
advisors, naturally questions were raised about the usefulness of Western economic
theory. One of the recurring debates in the 1960s centered on the economic motivation of
farmers and traders in subsistence economies. Another major issue was the relevance of
Western development models--dual-sector, labor surplus, land surplus, and stages of
growth models.

-Dumont's moralizing about the shortcomings of African leaders and the lack of
hard work in African society is a common theme in Dumont (1966, 1969) and Dumont and
Mottin (1980).

For a synthesis on Latin American scholarship on economic development in the
fifties and sixties, see Alain d'Janvry (1982).

Economic Behavior of Farmers and Traders
Many Western economists have long contended that Africans were not "economic
men" in the Western sense. For example, one hundred years ago, Alfred Marshall, the
founder of neoclassical economics, wrote about the savages "living under the dominion of
impulse; scarcely ever striking out new lines for themselves; never forecasting the distant
future; fitful in spite of their servitude to custom, governed by the fancy of the moment;
ready at times for the most arduous exertions, but incapable of keeping themselves long
to steady work" (Marshall, 1956). A major implication of this view is that Western
economic theory may have little to say about the behavior of farmers and traders in
It is instructive to note that this theme has been hotly debated in the literature of
several disciplines, including sociology, anthropology, and political science. One of the
first challenges to the notion that Western economics can be applied in low-income
countries came from the Dutch sociologist Boeke. On the basis of many years of research
in Indonesia, Boeke (1953) advanced the concept of sociological dualism to describe the
modern and traditional sectors in which the peasants in the traditional sector had limited
needs, a value system based on prestige, and they exhibited fatalistic behavior and
backward-bending supply curves of labor. As a result, Boeke asserted that Western-
financed development interventions in the traditional sector would be constrained by
these social and cultural barriers to change.i
In the 1950s and 1960s, the debate over the relevance of Western economics in
developing countries preoccupied many economic anthropologists. During this period,
economic anthropologists were divided into two schools of thought--the "substantivists"
and the "formalists." The substantivists contended that exchange in many low-income
countries is carried out according to principles of reciprocity and redistribution. As a
result, the substantivists contended that Western economic theory which stresses profit
maximization has limited application in many parts of the developing world, including
Africa. The substantive paradigm is conveyed through the writings of Polanyi et al. (1957)
and Dalton (1962, 1978).'2 The formalist paradigm was articulated by economic
anthropologists such as Raymond Firth and Harold Schneider (1974). The formalists
contended that Western economic theory could be selectively applied to the Third World

-But as Benjamin Higgins (1959) and other scholars observed, Boeke did not present
solid evidence to defend his case.

!/Pryor (1977) used econometric techniques to test 60 hypotheses about "primitive
and peasant societies" in a number of case studies and found that most of the hypotheses
of the substantive school were not supported. But one has to question the use of
econometric techniques to test the validity of the substantive paradigm.

including Africa, because African farmers, traders, and migrants were believed to respond
in a general way to economic incentives just as producers and consumers respond in high-
income market economies.- Fortunately, the often sterile debate between the substan-
tivists and formalists was relegated to the "dust-bin of history" during the 1970s.
Anthropologists generally reached a consensus that many of the views of both schools
were not mutually exclusive and that each school could provide hypotheses which shed
insights on economic behavior in agrarian societies in the Third World.
The debate over the economic behavior of farmers has recently reemerged among
political scientists working in Asia. 3. C. Scott (1976), for example, has advanced the
view that peasants are concerned with village cohesion as a means for assuring economic
security and survival. In Scott's view, the transition to a market-oriented economy is
likely to increase inequality and increase the insecurity of the poorest families in a
village. Joel Migdal (1974) has also stressed the importance of a village orientation in
decision making and the possible destruction of village safety nets as commercialization
of agriculture proceeds. Scott's and Migdal's view of a "moral economy of subsistence
farming" was challenged by Popkin in The Rational Peasant (1979). Drawing extensively
on historical data from Vietnam from precolonial times to the emergence of the
Communist movement, Popkin contends that peasants make individual investment deci-
sions which may be at the short-run expense of the village. Popkin is of the opinion that
the commercialization of agriculture, rather than increasing inequality and insecurity,
presents opportunities for peasants to increase their welfare.
The debate over the extent to which the behavior of farmers and traders in low-
income agrarian countries are influenced by "non-economic" considerations will likely
never be fully resolved. But the proposition that farmers and traders in Africa act in a
manner which is inconsistent with the postulates of Western economic theory is an
empirical question. During the colonial period and the first decades of independence, the
view that Africa's farmers were not "economic men" was reflected in (1) the target
income concept and the hypothesis that there was a backward-bending supply curve for
labor, and (2) the hypothesis that social and cultural factors are overriding barriers to the
adoption of innovations and the achievement of development objectives. We shall review
the literature on these two propositions.

1/The debate between the formalists and the substantivists has been summarized by
Posner (1980, pp. 608-609) as follows: "the formalists spend their time looking for explicit
markets in primitive societies and the substantivists spend their time showing how
resources in primitive societies are mostly allocated by non-market means."

Target Income/Backward Bending
Supply Curve of Labor
The origins of the target income or backward-bending supply curve of labor
hypothesis can be traced to the colonial period when plantation and mine owners reported
that they could not fill their vacancies at going wage rates because Africans were "lazy"
and that mine workers would often return to their families after they had earned their
target income. This view was implicitly and, in some cases, explicitly supported by
colonial administrators. For example, head taxes were introduced in numerous colonies to
force Africans to increase their supply of labor to plantations and mines. In the
postindependence period of the 1960s, numerous scholars repeated the theme that
Africans had limited wants and that they would not respond to market forces--higher
wages, for example--after they had earned a specific money or target income to pay for
taxes, bride-price, or consumer goods. Under this line of reasoning, it was assumed that a
5 percent increase in wages, for example, might lead to a reduction rather than an
increase in the supply of labor. The policy implication was that higher wages and higher
prices to farmers could not be relied upon to draw workers into the labor force and
farmers into the market economy. But during the 1950s and early 1960s, economists and
agricultural economists began to marshall an impressive array of empirical findings that
Africans responded "normally" to economic incentives such as changes in seasonal wages
and relative prices.
Two U.K. economists--Bauer and Yamey (1959)--were among the first researchers
to present quantitative evidence on the response of small farmers to price incentives.
Bauer and Yamey examined two of the statutory marketing boards which were established
in Nigeria during and after World War II--the Nigerian Cocoa Marketing Board and the
Nigerian Oil Palm Produce Marketing Board. Starting in 1947-48, these Boards offered
large differentials in prices paid to farmers in order to encourage the production of higher
grades of cocoa and oil palm. Bauer and Yamey's study revealed that the proportion of
Grade I cocoa purchased by the Nigerian Cocoa Marketing Board increased from 47
percent in 1947-48 to 98 percent in 1953-54 following the introduction of premium prices
for high quality cocoa. Bauer and Yamey demonstrated that the small Nigerian cocoa and
oil palm producers were indeed "economic men" and that they responded to market
incentives just as farmers in high-income countries had been doing for decades.
William O. Jones' (1960) masterful survey of studies of migration, trade, production,
and marketing in the precolonial and colonial periods produced impressive evidence that
"the economic drive is present in a great many Africans." Peter Kilby (1961) studied
Nigerian factory workers and found that they surpassed their European counterparts in
sheer physical exertion by as much as 50 percent when the proper financial rewards were

held out to them. Elliott Berg (1961) presented added evidence to suggest that the target
income/backward-bending supply curve of labor hypothesis should be discarded. More-
over, Miracle and Fetter (1970) argued that even if a backward-sloping, labor-supply
schedule had been validated by empirical studies this behavioral pattern could be
consistent with orthodox economic behavior if all of the costs associated with migration
and working were taken into account--costs such as uncertainty, disease, and long period
of separation from their families. Miracle studied the emergence of the wage labor force
in Kenya in the early 1900s and found that migrants who left the highlands of Kenya for
work 300 miles away on the coast encountered substantial risks of contacting disease, not
receiving wages when they became ill, being subject to brutality, and even death. Miracle
concluded (1976) that the costs associated with migrants leaving their families could have
produced a backward-bending supply curve of labor in this particular case. Helleiner
(1975) and Miracle (1976) summarized the debate by noting that it is doubtful that the
target-income hypothesis was ever valid for a large number of Africans, but, even if it
was, it is of limited validity today.

The Cultural Barrier Hypothesis
In the early 1960s, it was common to identify social and cultural factors as
overriding barriers to the adoption of innovations and the achievement of development
objectives. But the failure of the Community Development movement in Asia in the
1950s and 1960s- (which stressed social factors--such as felt needs, participation, and
self-help) and numerous examples of change in spite of perceived social barriers have led
to a consensus that the cultural barrier hypothesis--like any single barrier theory--should
be rejected.-
Turning to Africa, in a major study of agricultural change in six countries in sub-
Saharan Africa, Uchendu, an anthropologist, and Anthony, an agronomist, concluded that
there is no empirical support for the widely held belief that traditional values constitute a
general barrier to change (Uchendu, 1968; Anthony and Uchendu, 1974; Uchendu and
Anthony, 1975a,b). This same point is reiterated by Hutton and Cohen in their
reassessment of sociological approaches to the study of change among peasants in Latin
America and Africa. To quote:
Local suspicions, jealousies, ignorance, fatalism, passivity, and fears can play
their part, just as they can in any human situation, but we no longer use them
as general explanations; they can be relegated to their proper place, enabling
us to understand why sometimes they impede change and sometimes they do
not. (1975, p. 28)

-See Holdcroft (1978).

/See Streeten (1972) for an insightful note on the vacuity of theories of single
barriers to change.

Although social and cultural factors may not be a general barrier to change in rural
areas in the Third World, it is obvious that these factors still play significant roles in
shaping the overall pattern of development and who gains and who loses in the process.
For example, Parkin (1972) points out the enduring role of custom in village life of the
Giryama society on the coast of Kenya over a 25-year period. Parkin, a social
anthropologist, showed how the Giryama shifted from primarily subsistence farming and
herding to dependence on the cash economy through the production of copra which was
prepared from coconuts. Parkin found that successful farmers deliberately kept them-
selves from appearing to rise above their neighbors by participating in such customs as
bridewealth transactions and reciprocal funerary obligations while at the same time they
were purchasing land from small farmers who were forced to sell land to pay for the
escalating costs of funerals and bridewealth.-1/ Philip Mbithi (1977) similarly has shown
that in East Africa the social environment, including observance of rituals and taboos, is a
major influence on the timing of agricultural practices such as planting and harvesting.

In summary, the target income/backward-bending supply curve of labor hypothesis
was discredited by empirical studies in the 1960s and 1970s.-/ As one looks back upon the
debate, it is surprising to find the proponents of the hypothesis did not present empirical
evidence to support their position. The debate was carried on for years through an
exchange of hearsay and assertions. Nor is there any support for the view that social and
cultural values are an immutable barrier to change. But we must be cautious about
generalizing about the relative importance of social and cultural factors in some 1,000
ethnic groups in sub-Saharan Africa. There are many isolated villages where poor
transportation is still a constraint on the operation of market forces. Moreover, there are
many nonmarket institutions such as the extended family, clans, and age groups which still
play a significant role in shaping economic decisions. But there is unambiguous evidence
that African farmers, traders, and migrants will respond to economic incentives when
they are offered appropriate incentive structures.

Western Development Models
Although Western development economics is now 25 years old, one thing is clear--it
is still unencumbered by evidence from Africa. For example, although aggregate data

I/We are indebted to Sara Berry for calling Parkin's study to our attention.

/For further evidence supporting the concept of economic man in Africa, see the
overview of "Supply Response (Function)" studies in Part IV and the discussion of research
on livestock in Part VI.

from 101 countries are included in Patterns of Development: 1950-70 (Chenery and
Syrquin, 1975), the poor quality of data enabled them to use only 8 African countries in
their detailed analysis of the patterns of development. Moreover, dual-sector and stage-
of-growth models which dominated Western development economics in the 1960s and
1970s were primarily based on patterns of development, resource endowments, institution-
al structures, and empirical findings from Asia and Latin America. For example, the
well-known dual economy models of Lewis and Fei and Ranis depend on surplus labor,
institutionally determined agricultural wage rates, and the assumption of a closed
economy. While these models may have provided insight into the interaction of
agricultural and industrial development in Asia and Latin America, they shed little light
on patterns evolving in the rural economies of Africa.

Myint's vent-for-surplus model which focuses on the role of international trade as an
"engine of growth" in subsistence economies- has been used to explain the rapid growth
of agricultural exports during the colonial and postindependence periods in Africa. Myint,
an economist at the London School of Economics, abandoned the traditional classical
assumptions such as specialization and comparative advantage in certain crops in trying to
explain the sudden surge in exports in some countries such as Burma, Nigeria, and Ghana
in the late 1800s and early 1900s. Myint developed a vent-for-surplus model which
directly attributes the export boom to improved local transport, access to overseas
markets, and incentive goods from overseas. These factors provided a "vent" to tap the
surplus-productive capacity inherent in surplus land and family labor after subsistence
food needs of farm families had been met.
The vent-for-surplus model is appealing because it is not a global theory of develop-
ment like stages of growth and dual sector models. The model stresses the key role of
effective demand from European markets in mobilizing surplus labor and land in
underpopulated areas with a peasant smallholderr) type of production system.n' It is
presumed to be a costless type of growth which could be largely self-financed by small
farmers and local traders by reducing their leisure time. Governments or private
international firms only have to provide improved transport, communication, and access
to overseas markets.

/Starting in the late 1800s and early 1900s, there was a rapid growth in the
production and international trade of crops such as cocoa, coffee, oil palm, and rubber
throughout Africa. Many scholars contend that cash crop expansion through international
trade was the "engine of growth" of African economies during the colonial period.

2/The economic historian's emphasis on the role of international demand for export
crops in the study of the African history is a healthy contrast to the modest attention to
demand parameters in Western growth models in the past 25 years.

Szereszewski's (1961) and Polly Hill's (1963) research on the rapid growth of
smallholder cocoa production in Ghana in the 1880s and 1890s are empirical tests of the
vent-for-surplus model. Hill's meticulous field work in southern Ghana revealed that the
emergence of cocoa production involved more than a response to growing European
demand for cocoa and improved transport in Ghana. Hill found that migrant farmers--not
local small farmers--were the source of innovation in cocoa farming through their
leadership in organizing, financing, producing, and marketing cocoa. The spread of cocoa
farming by migrants can be viewed as a process of indigenous capital formation in a land
surplus economy. Migrants were also instrumental in the diffusion of cocoa production in
the Ivory Coast (Dupire, 1960) and in Western Nigeria (Berry, 1975). Although Berry
(1975) and Hogendorn (1975 and 1978) agree with Myint that increased effective demand
provided incentives for smallholders, they contend that Myint's vent-for-surplus model
needs to be refined to devote more attention to the role of local institutions in
facilitating capital formation and the spread of innovations. Also the vent-for-surplus
model of development does not explain why short spurts of export-led growth have led to
cumulative growth and diversification of economies such as Australia, New Zealand, and
Canada while the cocoa export boom in Ghana and copper exports in Zambia have not led
to a reduction of dependence on one key export.
The vent-for-surplus model has been used directly or indirectly in the following
studies: cocoa in the Ghana (Szereszewski, 1961; Hill, 1963); cocoa in Nigeria (Berry,
1975); groundnuts (peanuts) in Nigeria (Hogendorn, 1978); rubber in Ghana (Dumett, 1971);
peanuts in Senegal and the Gambia (Brooks, 1975); and kola nuts in West Africa (Lovejoy,
1980; Agiri, 1977). For a summary of studies using the vent-for-surplus model to analyze
the growth of agricultural exports before 1914, see Hogendorn (1975, 1978). For a radical
critique of Hogendorn's use of the vent-for-surplus model, see Freund and Shenton (1977)
and Hogendorn's reply (1977).
What is the value of Myint's model in Africa in the 1980s? Whereas Myint's model
highlights the potential of trade as an engine or handmaiden of growth, there are two
major shortcomings in using the model for policy guidance in Africa in the 1980s. First,
Myint's model fails to stress the investment in research, rehabilitation, and replanting
which is needed in order to maintain a country's share of world trade in a particular
commodity. Ghana and Nigeria are examples of countries which have lost world market
shares of cocoa and oil palm to the Ivory Coast, Malaysia, and Brazil over the past 20
years. Second, the model assumes away the food problem and the need to invest in
increasing the productivity of the food subsector. This problem is now at the crisis stage
in several African countries as the frontier is exhausted and investments in irrigation,

land reclamation, and tsetse fly control are needed to intensify agriculture and increase
food production.-

Labor Surplus
The concept of surplus labor as outlined by W. Arthur Lewis' (1954) model of
development with unlimited supplies of labor has never been seriously applied by scholars
working in sub-Saharan Africa because--as Lewis acknowledged--sub-Saharan Africa is
known to be a land abundant region relative to countries such as Egypt, India, Java, and
Bangladesh.-' This abundance of land historically has been reflected in "land extensive"
farming systems 3/in which long periods of fallow (often 10 to 20 years) are interspersed
with short periods of cultivation. Moreover, it was empirically shown in the 1960s that
most African countries were faced with both seasonal labor surpluses during the dry
season and labor shortages or bottlenecks during some periods of the farming season
(e.g., weeding). As a result, Byerlee and Eicher (1974) urged model builders to
concentrate on developing models of African development which were able to deal with
seasonal labor shortages and surpluses, farm/rural nonfarm labor interactions, determi-
nants of rural-urban migration, and capital and labor interactions between the agricultur-
al, industrial sectors and international trade.

-iW. Arthur Lewis (1978) contends that while international trade did serve as an
"engine of growth" in the 19th century this is not its proper role. He contends that
technological change (especially in food production) is the engine of growth in the Third
World today but acknowledges that trade can serve as a handmaiden of growth.

-In an article on the proletarianization of the peasantry in Rhodesia, Arrighi (1970)
criticized Lewis for viewing unlimited supplies of labor as a given rather than being
produced by the colonizers or capitalists. But Lewis specifically addressed these issues by
noting that in Africa the imperial powers impoverished the peasantry "by taking away the
people's land or by demanding forced labor in the capitalist sector, or by imposing taxes to
drive people to work for capitalist employers" (Lewis, 1954, p. 410). Still, as Hirschman
points out, these practices were not central to Lewis' model of unlimited supplies of labor
because "a decline in infant mortality could have the same effect in augmenting labor
supply as a head tax" (1981, p. 16).

-/We prefer using the term "land extensive" farming system rather than land surplus
models of development because even though farmers in Sierra Leone (see Spencer and
Byerlee, 1976, 1977) or Tanzania (see Shapiro, 1978) have 20 to 40 hectares of land under
their control they actually cultivate only a small portion of that land--3 to 5 hectares--in
any one year. The small area under cultivation is a function of the shortage of family
labor at critical periods in the production process and the need to keep the bulk of the
land in fallow in order for the bush to regenerate and restore soil fertility.

-For a discussion of seasonal labor bottlenecks, see De Wilde et al. (1970); Eicher
et al. (1970); Cleave (1974); and R. A. Swanson (1981).

Land Surplus
The concept of "surplus land" has received little attention in Western development
economics. While Myint's vent-for-surplus model hypothesizes that increased effective
demand from international trade can stimulate the use of both surplus labor and land,
there have been few attempts in sub-Saharan Africa to analyze the effects of surplus land
on labor allocation and development outside the vent-for-surplus framework. Gerry
Helleiner and Bent Hanson are among the few who have advanced models of development
with unlimited supplies of land. Helleiner (1966b) concluded that because of the diversity
of ecological zones and population densities in Nigeria it was impossible to classify
Nigeria as either a land or labor surplus economy. Once it is recognized that both
situations exist in a country such as Nigeria, the issue of labor mobility reemerges as an
important factor in development. Bent Hanson (1979) developed a model of development
with umlimited supplies of land but did not present empirical information to test his
The notion of surplus land has been criticized by numerous African scholars. Sunday
Essang (1973) contended that aggregate statistics can be misleading because even though
there are high land/man ratios (1) land in bush fallow cannot be considered surplus since
under existing technology long fallow periods are needed to regenerate soil fertility;
(2) aggregate statistics conceal the fact that much of the uncultivated land not in fallow
is of low quality and often is not suitable for agricultural production because of unreliable
rainfall; (3) the idea of land surplus ignores the point that uncultivated land often has a
positive opportunity cost since it provides firewood, timber, oil palm, wild game, and
forage; and (4) tsetse flies, river blindness, other diseases and health problems, and poor
transport (e.g., southern Sudan) preclude the use of large blocks of land which appear as
idle land in aggregate statistics. Essang further argued that development interventions
based on a land surplus model can lead to (1) bias in favor of large-scale farming,-
(2) inadequate attention to land improvement measures, and (3) indifference to population
On the basis of our review of the literature, we conclude that the concept of a land
surplus economy has little heuristic value. Most economies are too large, complex, and
diversified to be described as land or labor surplus. Furthermore, the notion of surplus,
whether of labor or land, overlooks complex institutional and administrative questions
about who controls the access to land and whether supporting services and adequate
incentives are in place.

I/In a subsequent article, Essang (1977) points out that the oil boom in Nigeria has
provided the foreign exchange earnings and government revenues for a resurgence of
government-directed large-scale farming schemes--especially irrigated farming systems
in northern Nigeria. Likewise, surplus land in western Sudan is cited by the government's
Mechanized Farming Corporation as a justification for mechanized farming.

As African nations entered the second decade of the postindependence period, the
1970s, the abstract theories and models of neoclassical economists were on the defensive.
The assumptions of the models were recognized to be irrelevant to Africa and the
resulting policy prescriptions were not taken seriously. In retrospect, the major
shortcoming of Western development models was their excessive macro orientation and
the inability of these models to provide a convincing specification of the agricultural
sector--the sector which employs 50 to 95 percent of the total labor force in African
economies. Most models ignored structural problems. Moreover, most models focused on
the supply side and ignored the structure of demand and its relationship to income
distribution and employment.
During the late 1960s and early 1970s, several scholars developed models based on
African resource endowments and institutions in an attempt to address the weakness of
imported development models. While the land-surplus model has rightly been rejected as
too global, valuable theoretical frameworks have been proposed for migration (Todaro,
1969), rural small-scale industry (Liedholm and Chuta, 1976), and consumption (King and
Byerlee, 1977). Byerlee and Eicher (1974) proposed a multi-sector rural economy model to
examine the linkages between rural and urban firms, both large- and small-scale, and
small- and large-scale agricultural producers. It is notable that the authors were
proposing these models to serve as a framework for conducting empirical research rather
than for deducing policy recommendations. This reflects the fact that a consensus had
emerged among Western development economists by the early 1970s that because of the
failure of Western development models to deal with the key problems of employment,
equity, and food supply, it was necessary to go back to the basics, building an
understanding of development in African rural economies based on meticulous micro-
economic research.-

Political Economy and
Radical Perspectives
Western development economics or orthodox economics was challenged in Africa in
the 1960s by the emergence and rapid growth of political economy/dependency/radical

IFor example, Hayami and Ruttan (1971) noted that there was a need to step up
micro research in the 1970s in order to provide the data necessary for a convincing
specification of the agricultural sector. This has left Western economists open to the
challenge from radical scholars that their micro studies are ahistorical, overstress
technical and infrastructural constraints, and give too little attention to the influence of
the world economy. For a critique of "conventional development research" and the role of
Western social scientists in Africa, see Amin et al.,(1978).

models of development and underdevelopment.- The political economy models of
development have their roots in the writings of Lenin on Imperialism and in the post-
World War II writings of the late Paul Baran. Baran, a Marxist economist at Stanford
University, wrote a seminal article, "On the Political Economy of Backwardness" (1952).
In Baran's political economy model, he did not rule out the possibility of broad-based
development in the LDCs via capitalism but he argued that in most underdeveloped
countries it would be impossible to bring about broad-based development without violent
changes in social and political institutions and without a dynamic industrial sector.
Although Baran was clearly ahead of his time in putting his finger on institutional and
structural barriers to development and the need to put the effective demand of the
masses at the center of development programs, his views on agriculture were naive and
misleading. For example, Baran wrote that since the marginal product of labor tends to
be zero in agriculture, "there is no way of employing it usefully in agriculture." Farmers
"could only be provided with opportunities for productive work only by transfer to
industry." Moreover, Baran advanced an anti-small farm view when he wrote "very few
improvements that would be necessary in order to increase productivity can be carried out
within the narrow confines of small-peasant holdings."

Dependency Theory
One of the most influential radical views of the development process in Africa is
dependency theory. The dependency interpretation of underdevelopment was first
proposed in the 1950s by the Economic Commission for Latin America (ECLA), under the
leadership of Raul Prebisch. The basic hypothesis of this perspective is that underdevel-
opment is not a stage of development but is the result of the development of the world
capitalist system. Although a number of different views of dependency have been put
forward by scholars such as Sunkel, Furtado, A.G. Frank (1966), Galtung (1971), and
others, the following definition of dependency by Dos Santos has been widely cited:
By dependency we mean a situation in which the economy of certain countries
is conditioned by the development and expansion of another economy to which
the former is subjected. (1970, p. 231)2/

In the 1960s, dependency theory was imported into Africa from Latin America.
Over the past 15 years, Samir Amin has provided leadership in developing a Marxist

/The emergence of political economy models in the mid-1960s occurred at the
same time that a number of countries (Ghana, Mali, Guinea, Tanzania) shifted from an
ideology of economy policy of capitalism to an ideology of socialism.

2/For critiques of the dependency school of thought in Latin America, see Cardoso
and Faletto (1979) and De Janvry (1982).

version of dependency theory. Amin, an Egyptian by birth, and a national accounts
specialist in economics, turned his early attention to an analysis of development in Mali,
Guinea, and Ghana (1965) and the Ivory Coast (1967). Amin subsequently elaborated on his
dependency views through analyses of the precolonial, colonial, and postindependence
periods in Africa. In Accumulation on a World Scale (1974c) and Unequal Development
(1976), Amin presents an analytical framework of underdevelopment in Africa based on
surplus extraction and the domination of the world capitalist system. Amin contends that
social structures in the periphery are "truncated" and can only be understood in relation to
the "world social structure" (1976, p. 294). One of the cornerstones of Amin's analysis is
the concept of the "social formation of peripheral capitalism."/ Amin argues that
peripheral formations are fundamentally different from those of the center because of
their "extraversion."2/ It is notable that Amin contends that "despite their different
origins the peripheral formations tend to converge toward a pattern that is essentially the
same" (1974c, p. 378).
Amin has provided valuable insights into the development process (1970, 1972,
1974a, 1974c, 1976), but his prescriptions for agriculture have been naive and have
changed over time. For example, he attributes Africa's agrarian crisis to the predomi-
nance of agrarian capitalism which he argues takes two major forms: kulakization (class
of planters who employ wage labor) and organization of export production subject to a
theocratic-political authority (such as the Mourides in Senegal). While both "formations"
clearly are prevalent, they by no means dominate the structure of agricultural production
in sub-Saharan Africa. During the 1960s, Amin favored animal traction, promoted
industrial crops, and argued that traditional social values were a serious constraint on
development at the village level. He also argued that the transition to privately owned
small farms was a precondition for socialism. By the mid-1970s, Amin reversed himself
and recommended the collectivization of agricultural production and he abandoned his
support for animal traction and industrial crops. These shifts reflect, in our view, the
weakness of deriving prescriptions for agricultural policy on the basis of global and
abstract analyses of the world economy. For critical surveys of Samir Amin's work, see
Sheila Smith (1980) and Schiffer (1981).

/Social formations are defined as "concrete, organized structures that are marked
by a dominant mode of production and the articulation around this of a complex group of
modes of production that are subordinated to it" (1976, p. 16). The characteristics of
peripheral capitalism are concisely described in Amin (1976), pp. 333-364.

2/Extraversion is defined as the dominance of the exporting sector over the
economic structure as a whole, which is subjected to and shaped by the requirements of
the external market (1976, p. 203).

During the past decade, there have been several attempts to evaluate the contribu-
tion of dependency models in understanding the causes of poverty and underdevelopment
in Africa. McGowan (1976) and Vengroff (1977) attempted to test dependency theory in
Africa but came up with inconclusive results. Kleemeier (1978) criticized McGowan and
Vengroff for attempting to use correlation analysis to infer causality. Palma (1978) raised
several important questions about whether the dependency hypothesis can be empirically
tested. Recent political economy research in Africa which focuses on agriculture include
Wilcock (1978); Ntangsi (1979); Henn (1978); and some of the selections in the book edited
by Heyer, Roberts, and Williams (1981). Collections of essays presenting political
economy/dependency perspectives on African development include Arrighi and Saul
(1968); Gutkind and Wallerstein (1976); Shaw and Heard (1979); Saul (1979); and M. A.
Klein (1980). For other influential writings, see Leys (1974). For an assessment of the
dependency literature in the Third World, see Tony Smith (1979).

A small group of French Marxist anthropologists--including 3. Suret-Canal, M.
Godelier (1972), C. Meillassoux, G. Dupre and P. Rey (1978), and C. Coquery-Vidrovitch--
and a few political scientists such as Goran Hyden have rejected the notion that a global
Marxist-Leninist ideology based on a generalization of the historical experience of Europe
and America is sufficient to explain patterns of development in Africa. The French
Marxist anthropologists also reject the views of anthropologists such as Firth, Bohannan,
Dalton, and Polanyi who stress forms of exchange rather than modes of production and the
theory of reproduction (Clammer, 1975). Regarding Western economics, the Marxist
anthropologists question the applicability of economic concepts derived from capitalism
to economic systems with different characteristics. For example, Meillassoux (1981)
argues that Western economists looking for simple economic explanations are often
confused by the fact that the system of circulation of goods in pre-capitalist societies is
conditioned by non-material phenomena.
The micro-Marxists attribute the roots of underdevelopment to the failure of
capitalism to produce a dynamic transformation of precapitalist economies while main-
stream Marxist scholars and dependency theorists stress the extraction of the surplus as
the root cause of underdevelopment. Micro-Marxists argue that there is a need to

-/Hirschman (1977) coined the term micro-Marxist to describe radical scholars who
concern themselves with "specific events and country experiences.

2See also M. Sahlin's Stone Age Economics (1974).

to understand the interaction between the precapitalist modes of production and the
capitalist mode in each particular setting.- As Hyden has said:
That modes of production differ in their articulation in the Third World
countries has only recently become a subject of research. How these forms of
articulation affect the development potential has not yet been fully explored.
(1980, p. 4)
Thus, the focus of research by the micro-Marxist has been on identifying the characteris-
tics of precapitalist modes of production in Africa. 3. Suret-Canale (1964) made a
seminal contribution when he attempted to apply the famous Asian mode of production to
precolonial Africa.
Meillassoux's (1964) study of the village of Gouro in the Ivory Coast is regarded by
some scholars as the definitive benchmark exposition of a micro-Marxist analysis of
precapitalist development and the transition from subsistence to commercial farming.-
Meillassoux's basic theme is that agriculture is based on communities (roughly equivalent
to households) which have the goal of self-sufficiency so they are not dependent on other
social classes. Circulation of foodstuffs is controlled by a social hierarchy based on
seniority. He argues that these communities, "agricultural self-sustaining formations,"
contain within themselves all the means necessary for providing the basic social and
material needs of their members but as production for external markets is grafted onto
these self-sustaining formations, it is inevitable that a class society results. A collection
of papers edited by Meillassoux which is based on this theme, The Development of
Indigenous Trade and Markets in West Africa (1971), provides an important counter-
interpretation to A. G. Hopkins' (1973) stress on the positive role of market forces in West
Africa's economic history.
Catherine Coquery-Vidrovitch (1978) has also made an influential contribution in her
attempt to specify an African mode of production. Unlike many other micro-Marxists,
Coquery-Vidrovitch stresses the historical importance of long-distance trade in African
development. A key element of her model is the exclusive ascendency of one group over
external trade.
Hyden (1980) is unique among scholars working in Africa in stressing the strength of
precapitalist societies in the face of the expansion of the market. Based on his extensive
experience in Tanzania, Hyden argues that smallholders have rejected both colonial and
postindependence interventions by the government of Tanzania because they recognize
that any improvement in material conditions would come at the loss of other values.

!/Some of the dependency scholars share this view even though their stress is on the
manner in which dependent development is conditioned by the world economy rather than
on the specific articulation of precapitalist modes of production.

2/Terray (1972) wrote a 100-page critique of Meillassoux's Gouro village study.

The political economy literature attempts to link rural poverty and underdevelop-
ment to historical forces, world capitalism, and surplus extraction. Political economy
scholars also emphasize the linkages between colonial policies and contemporary underde-
velopment and encourage agricultural scientists to move beyond the simple view that
African agriculture is unproductive primarily because of the lack of new technology. But
Amin and his followers who stress global interpretations of underdevelopment in Africa
have underplayed the large number of internal policies and factors which also contribute
to poverty and agricultural stagnation in Africa. The Achilles heel of dependency and
political economy theorists in Africa is likely to be the same one which discredited
Western dual sector models in the 1960s--abstract theorizing and the neglect of empirical
research at the micro level. The question remains: Can political economy and dependen-
cy scholars move beyond their abstract models to develop models based on studies of the
behavior of African farmers and herders, on African institutions, and on micro/macro
linkages in order to provide policy guidance in a region in which the majority of the people
are farmers?
A small group of micro-Marxists, primarily French anthropologists, have rejected
the view that Marxist ideology can be applied in Africa without modification. These
scholars are carrying out village studies with emphasis on francophone countries in West
Africa. The micro-Marxists have made important contributions to the study of the role of
precapitalist modes of production in shaping the development process, the analysis of the
transition to commercial farming, and the study of inequality. The ability of French
Marxist anthropologists to ask what some scholars call the "key questions" about
development undoubtedly explains the growing number of translations of Marxist works
into English (from French and German) over the past seven or eight years.- But, over the
next decade, the micro-Marxists also must face the challenge of translating their insights
into recommendations which can provide guidance to policy makers and donor agencies.

/A collection of papers by nine French Marxist anthropologists (including Copans,
Godelier, Roy, Coquery-Vidrovitch, and Meillassoux) is available in English in a paperback
volume edited by David Seddon (1978). Also see Meillassoux, Maidens, Meal and Money:
Capitalism and the Domestic Community (1981). See Raymond Firth (1975) for an
analysis of "Social Anthropology and Marxist Views on Society."


As African governments tried to establish their legitimacy and chart their own
courses in the postindependence period, they experimented with alternative development
strategies, programs, and policies. There is now a large and growing body of literature on
experiences with alternative policy instruments and development programs.- While a
review of the experiences cannot resolve the many policy debates facing African
governments, knowledge of empirical findings on alternative programs and policies may
help focus debates and identify directions for future research.
In this section, we review research on the major policy issues of the postindepen-
dence period, including (1) agrarian capitalism versus socialism, (2) agricultural planning
and agricultural sector modelling, (3) large- versus small-scale farming, (4) marketing
boards and food grain boards, (5) agricultural prices, (6) rural development programs, and
(7) accelerated food production campaigns.

Agrarian Capitalism and Socialism
One of the most important policy issues during the postindependence period of
1960-81 has been the ideology of economic policy--capitalism or socialism. In the
postindependence period beginning in the late 1950s, numerous countries shifted from
capitalism to socialism after a few years of independence. For example, soon after Ghana
(formerly the Gold Coast) became independent in 1957, its dynamic leader, Kwame
Nkrumah, shifted from capitalism to an ideology of "African socialism." Although
difficult to define, African socialism in Ghana and many other countries included the
establishment of state farms, government tractor hire stations, promotion of cooperative
farming and farmer associations, and a number of moves to reduce the influence of
private traders.
Socialism is now an important ideology in sub-Saharan Africa even though in many
countries it is socialism in name far more than practice.-2 About one-fourth of the

!For overviews of agricultural policy issues for Nigeria, see Byerlee (1973), Wells
(1974), Essang (1977), Idachaba (1980a,b, 1981), Idachaba et al. (1981), and Nigeria
(1980a,b); for Zambia, see Dodge (1977) and Turok (1979); for Ken a, see Heyer et al.
(1976); for Tanzania, see Green (1974), Kim, Mabele, and Schultheis (979), and Coulson
(1979); and for Sierra Leone, see Byerlee, Eicher, et al. (1982). Studies of agricultural
policy in West Africa are Club du Sahel (1977), CILSS/Club du Sahel (1978a,b; 1979;
1980a,b), and Pearson, Stryker, Humphreys et al. (1981).

2For example, in 1965, the Government of Kenya clearly stated its commitment to
socialism (Kenya, 1965) but it has been decidedly capitalist since the 1965 pronouncement,
even though fragments of state control are common. For example, in 1981, the
government set the prices of cereals and the National Cereals Produce Board handled the
marketing of maize (thereby excluding private traders). Moreover, parastatals such as the
Kenya Tea Development Authority are common throughout the agricultural sector. But
these fragments of state control do not add up to a socialist-controlled agriculture.

countries in our review are now partially or firmly committed to socialist economic
ideology.-' The reasons for countries moving from capitalism to socialist economic
ideology and back to capitalism (e.g., Ghana and Mali) and from capitalism to socialism
(e.g., Tanzania, Guinea-Bissau, Mozambique, Ethiopia, and Zimbabwe) should be carefully
analyzed by students of African development. We shall focus on three countries--Ghana,
Mali, and Tanzania--because the experience with socialism in these countries is well
documented. For the remaining countries, the literature is too fragmentary to attempt to
generalize about the performance of agrarian socialism. A widely cited skeptical view of
socialism is Elliot Berg's (1964) analysis of socialism in Guinea. The case for socialism is
put forward by Dumont (1966, 1969); Arrighi and Saul (1968); and Seidman (1972, 1977).
When Ghana became independent in 1957, it is reported that President Nkrumah had
at his disposal L150 million sterling in reserves (from its cocoa marketing board) in
London banks. In 1961, Nkrumah abruptly shifted from capitalism to a radical socialist
strategy which equated modernization with industrialization and the mechanization of
agriculture. Ghana established highly mechanized state farms because it was thought that
"small-scale private farming is an obstacle to the spread of socialist ideas" (Killick, 1978,
p. 48). Ghana imported several thousand tractors, took control of cooperatives, and even
established government-operated food retail shops for a brief period of time. The failure
of rural socialism in Ghana in the 1960s has imposed a severe toll on its population which
continues today. For an assessment of agrarian socialism in Ghana, see Miracle and
Seidman (1968, 1968a); Killick (1978); and Nweke (1978b, 1979b). Also, see Nkrumah's
analysis of class struggle in Africa (1970).
Mali was the second country to move to socialism soon after its dramatic break in
diplomacy with France. But socialism was shortlived; the rise and sudden demise of
socialism in Mali is recorded in Zolberg (1968); W. I. Jones (1972, 1976); and Martin (1976).
For a pro-socialist view, see Ernst (1977).
Tanzania's abrupt shift to socialism in 1967 has induced the most voluminous
literature on socialism in Africa. The vision of agrarian socialism in Tanzania is set forth
in Nyerere's essay "Socialism and Rural Development" (Nyerere, 1967) and in Nyerere
(1968). For a remarkably candid assessment of some of the problems in achieving rural
socialism in Tanzania, see Nyerere (1977). Tanzania is now in deep financial difficulties
because of drought in the mid-1970s, quantum jump in oil prices, war with Uganda, and
the stagnant performance of its agricultural sector under socialism.

!/Benin, Guinea, Guinea-Bissau, Congo (Brazzaville), Ethiopia, Tanzania, Zimbabwe,
Mozambique, Angola, Mauritius, Equatorial Guinea.

Although a number of observers dismiss Tanzania's experiment with its Ujamaa and
villagization programs as failures, there have been important gains in literacy and social
services. For example, Tanzania's life expectancy at birth increased from 42 to 52, an
increase of 25 percent over the 1960-79 period. The adult literacy rate increased from 10
to 66 percent over the 1960-76 period (World Bank, 1981b, p. 181). These are dramatic
achievements which are often overlooked when facile terms such as success or failure are
used in place of detailed evaluations. Valuable insights on socialism in Tanzania are found
in Hyden's recent book (1980), a collection of essays by Mwansasu and Pratt (1979), Barker
(1979), and Samhoff (1981). For a recent bibliography on Ujamaa villages, see McHenry
(1981). For a comparative study of four small socialist states, including Tanzania, see
Morawetz (1980).
There are many unanswered questions about Tanzania's experiment with agrarian
socialism. A central question is why was coercion used by President Nyerere to round up
farmers and move them into villages? Many pro-Tanzania scholars avoid this topic.
Second, how serious were exogenous factors such as the drought and war with Uganda in
undermining socialist programs at critical junctures over the past 5-7 years? Third, were
faulty economic policies the Achilles heel of socialism in Tanzania? Nyerere (1967)
pointed out long ago that the worst enemy of socialism is faulty economic policies.
Clearly, agrarian socialism is on the defensive in Tanzania and the Revolutionary Party
has taken a number of steps in 1981 to move to increase incentives to farmers, including
more emphasis on private plots. Even early and strong admirers of President Nyerere,
such as Rene Dumont, recently wrote "Nyerere, through all his writings has made all
Europe dream but the stark reality dispels all illusion" (Dumont and Mottin, 1980).
It is too early to pass judgment on the performance of agrarian socialism in Africa,
particularly because of the widely different definitions of socialism and the absence of
data on countries such as Benin and Mozambique. Moreover, as Gerry Helleiner pointed
out in his perceptive article on "Socialism and Economic Development" (1972), all
countries--capitalist or socialist--must break common economic constraints, including
capital formation, foreign exchange, human resources, and institutional and technical
bottlenecks. We would add that governments following either economic ideology must
develop agricultural institutions and incentive structures to solve the most basic prerequi-
site of development--achieving a reliable food surplus. Capitalism or socialism by itself
cannot solve food and poverty problems. There are many capitalist and socialist countries
in Africa that are in an economic morass and policy assessments should be based on the
effectiveness of specific programs in dealing with key constraints rather than facile
statements about socialism as an ideology of development. For an uneven but valuable

assessment of socialism in the Third World, see the volume edited by Desfosses and
Levesque (1975). For a recent reassessment of socialism in sub-Saharan Africa, see the
volume edited by Rosberg and Callaghy (1979).
References on agrarian socialism are as follows:
General: Desfosses and Levesque (1975); W. A. Lewis (1978a); Morawetz (1980).
Africa: Berg (1964); Friedland and Rosberg (1964); Arrighi and Saul (1968, 1973);
Rosberg and Callaghy (1970).
Ghana: Amin (1965); Miracle and Seidman (1968a, 1968b); Killick (1978).
Guinea: Amin (1965).
Guinea-Bissau: Goulet (1978); Urdang (1980).
Mali: Amin (1965); Zolberg (1967); Ernst (1977); W. I. Jones (1972, 1976); Martin
Mozambique: Isaacman (1979); Saul (1979).
Tanzania: Nyerere (1967, 1968); Dumont (1969); Feldman (1969); Helleiner (1972);
Van Hekken and Van Velzen (1972); Mwansasu and Pratt (1973); Lofchie (1976); Saul
(1977); Barker (1979); Coulson (1979); McHenry (1979, 1981); Von Freyhold (1979);
Due (1980); Dumont and Mottin (1980); Hyden (1980); F. Ellis (1980); Samoff (1981);
Zalla (1981).

Planning and Agricultural Sector Modelling
Planning was launched in the colonial period immediately following World War II.
For example, in 1946, Nigeria prepared a ten-year plan in response to colonial office
requirements. As countries became independent, in the late fifties and early sixties,
almost all countries launched medium-term plans which focused on high rates of growth of
GNP as the target and indicator of development and relied heavily on foreign aid.- For
example, the government of Mali pointed with pride that the 11 percent growth target in
its first Plan was the highest in any African development plan in the sixties (Zolberg,
1967). But Mali scaled down its target growth rate soon after the implementation of its
Plan was underway and a few years later dropped its plan following a coup. Gusten (1967)
expressed the obsession of economists over growth rates and macro planning in the 1960s
in his "Can the Nigerian Economy Grow at 6 Percent per Annum in the Near Future--A
Pre-Planning Exercise?" at a time when the civil war was underway in Nigeria.
Most assessments of planning over the 1960-80 period conclude that failures have
far outweighed achievements. Professor Aboyade, a distinguished Nigerian economist and

-/The volume edited by Helleiner (1968b) is a standard reference on agricultural
planning in East Africa during the sixties. For an analysis of Kenya's agricultural planning
since independence, see Heyer, Maitha, and Senga (1976); Leys (1974); and Holtham and
Hazlewood (1976). For appraisals of planning in East Africa, see Widstrand (1976) and
Apthorpe (1976).

architect of Nigeria's Second Five-Year Plan, concluded, "For most of tropical Africa
planning over the past two decades has been little more than false hope" (1973). Rimmer,
an economist at the University of Birmingham with considerable experience in Ghana,
noted, "any resemblance between development plans and the actual course of economic
change in African and other poor nations is purely coincidental" (1969). Shen (1974)
reviewed development plans and national income data for 22 tropical African countries in
the late 1960s and pointed out there were major problems in implementing plans due to
the unstable political climates and weak institutions for plan implementation. In a survey
of planning in developing countries, Killick (1976b) concludes that planning has not lived
up to its expectations. But Helleiner (1972a) wisely points out that one should not equate
the preparation of plans with economic planning. Although many elaborate national plans
were prepared by foreign experts in the 1960s, most were dropped or ignored soon after
they were published. Helleiner correctly points out that most African countries have
made substantial improvements over the past 20 years in the quality of decision making,
the evaluation of economic alternatives, and in the implementation of projects. These
achievements are easy to overlook in critiques which focus on whether countries achieved
the growth rates spelled out in the plans rather than examining the slow process of
improving the data base, training people and strengthening institutions.
Turning to agriculture, major agricultural sector assessments were carried out in
Nigeria, the Ivory Coast, Sierra Leone, Ghana, Zambia, and the Sahel over the past twenty
years. For a model of Senegal's agricultural sector, see Labonne and Legagneux (1977).
World Bank-sponsored modelling efforts have been completed by Goreux (1977) in the
Ivory Coast and by Blitzer (1979) in Zambia. Nigeria has been heavily studied.-' A four-
year Nigerian agricultural sector analysis was carried out in the late 1960s by the
Consortium for the Study of Nigerian Rural Development under the leadership of
Glenn L. Johnson. Johnson and his team analyzed three alternative policy packages over
the 1969-85 period and concluded that Nigeria should give urgent attention to food crop
research, eliminate all export duties and taxes on agricultural products, and shift the
fiscal burden from agriculture to petroleum in order to provide incentives to harness the
energy of its six million small farmers. Thirty-two working papers and a final report by
G. L. Johnson et al. (1969) were published by the Consortium. Building on these findings,
Johnson, Manetsch, and colleagues developed a generalized simulation approach to

-/Nigeria's planning experience has been well documented. Stolper's influential book
Planning Without Facts (1969) stressed the lack of data in preparing Nigeria's first Five-
Year Plan in the early 1960s. Other analyses of Nigeria's planning processes during the
1960s can be found in Aboyade (1973), Dean (1972), Kilby (1969), Gusten (1967), Wells
(1974), Eicher and Liedholm (1970), and Eicher and Johnson (1970).

agricultural sector analysis in Nigeria (see Manetsch et al., 1971). Byerlee (1973) later
developed a simulation approach to trace the indirect employment and income distribution
effects of alternative agricultural development strategies in Nigeria. Byerlee's ten-sector
dynamic macro model was linked to an employment-incomes model and to the agricultural
sector model developed by Manetsch et al. (1971).
A number of major agricultural sector modelling efforts have been aborted and
never published. For example, a foreign consulting firm completed a one million dollar
agricultural sector assessment in Ghana in 1969 but the work was regarded by Ghanians as
superficial and merely a reassembly of secondary data. The draft papers were never
formally published. Also, the Massachusetts Institute of Technology's simulation study of
the Sahelian region in West Africa offers some insights into the pitfalls of "crash" studies.
The drought in the Sahelian region over the 1968-72 period brought forth massive food
grain shipments and a concern by donors for generating information on how to aid in the
recovery and long-term development of the region. But the micro data base in the Sahel
was weak and spotty when the crash one-year research program was launched by MIT's
Center of Policy Alternatives. The MIT researchers relied on secondary data and used a
systems dynamic simulation approach to trace long-term development strategies for the
Sahelian region over the 1975-90 period. The participation of African governments and
African researchers in the MIT study was modest. The MIT project was discontinued in
1974 and preliminary results were published in a ten-volume study (see Seifert and
Kamrany, 1974).
Over the past 10-15 years, the modest performance of these major modelling efforts
can be traced to the static nature of the models, and the lack of micro data and African
participation in the conduct of the studies. As a result, many of the modelling efforts
were dropped after the departure of the foreign researchers. The shortcomings of these
major modelling efforts have led many donors to shift their support to more limited (3-4
months) agricultural sectoral assessments. Although it is easy to criticize these quick and
dirty assessments, they represent a "half-way house" between major 2-4 year modelling
efforts and the ad hocism of the project-by-project approach to planning. Nevertheless,
there is an obvious need to increase funding for macro studies of river basins, livestock
and cropping interrelationships, and food and nutrition policies. For example, in the
Sudan, Somalia, and the northwestern region of Kenya, there is a need to develop models
of the interrelationships between livestock, food crops, human resources, and rural
institutions over an entire region, country, or group of countries. But in most countries
until more Africans are trained, agricultural sector models which cost millions of dollars
and depend on foreign researchers will likely be perceived by Africans as "academic toys"
rather than productive tools which can help Africans improve planning and decision

Large Versus Small Farms
A continuing debate in agricultural policy has been the economics of assisting
smallholders versus promoting large farms, including plantations, state farms, land settle-
ments, and river basin settlements. In the 1960s, the debate over large- vs. small-scale
agriculture became known (especially in Eastern Africa) as the transformation versus the
improvement approaches. The transformation approach featured a wide variety of large-
scale farming (plantations, settlements, and state farms) and processing plants; it was
designed to bypass the lengthy process of improving small farms within the existing
village structure. The major ingredients of the transformation strategy were infusion of
capital-intensive technologies, such as tractor mechanization, central management (often
European), and mobilization and training of an unskilled labor force by removing people
from their villages.

Smallholder Farming: A Descriptive Overview
Smallholder crop cultivation is the predominant farming system in sub-Saharan
Africa. Smallholder farming is primarily characterized by reliance on family labor, a
small stock of physical capital, and abundant land relative to Asian countries.-/ Family
labor is the most important factor of production, with family labor inputs ranging from 80
to 90 percent of total labor inputs (Byerlee, 1980). Farming households generally have 6
to 10 family members and it is common for households to include more than one nuclear
family. Adult male farmers work an average of 5 hours per day or 1,000 to 1,500 hours
per year in farming activities but the number of hours of labor devoted to off-farm
activities, such as rural small-scale industries, is substantial. This is in stark contrast to
Egypt and many Asian countries where total hours worked by adults in farming range from
2,500 to 3,000 hours per year (Cleave, 1974). Women play an important role in farming,
processing, and marketing but the extent of their participation varies greatly by activity,
ethnic group, and religion. Children are an important source of labor for tasks such as
weeding, the collection of firewood, bird scaring, carrying water, and taking care of
sheep, goats, and cattle.
Most small farmers till their land with human labor and hand tools, including metal
hoes, cutlass or machete, digging sticks, and knives. Although the shift from hand
cultivation to animal traction cultivation (oxen and donkeys) has been promoted for more
than 50 years, animal traction is still a minor source of farm power in almost all countries
in sub-Saharan Africa. Capital investments in housing, storage, and perennial crops are

-/For example, small farmers in the Semi-Arid Tropics (SAT) zone of West Africa
have three times as much land at their disposal as farmers in the SAT zone in southern

mainly created by family labor using local materials. Although the separation of farming
and livestock production is common throughout sub-Saharan Africa, there is a slow but
discernible adoption of livestock enterprises by sedentary farmers.
Cash expenses generally represent a small proportion of the value of production.
Purchased inputs--seed, fertilizer, and chemical pesticides--are not widely used by
farmers. Most fertilizer is applied in the form of organic manure.
The land area controlled1/ by a typical smallholder varies considerably but it is
generally far larger than in Asia. For example, in a study of smallholder cotton
production in northwestern Tanzania, K. Shapiro (1978) reports that the typical survey
farm controlled about 25 acres of land of which 4 were in cotton, 5.5 in food crops, and
the remainder in fallow and grazing land. Likewise, Spencer and Byerlee (1976) report
that the typical smallholder in Sierra Leone controlled about 40 acres of land but actually
farmed only a small portion of the 40 acres. The idle land was regaining its fertility in
the bush fallow system.
The area cultivated per family has consistently been found to range from 2 to 10
acres throughout sub-Saharan Africa (de Wilde et al., 1967; Upton, 1973; Cleave, 1974;
Winch, 1976; Spencer and Byerlee, 1976; and Heyer and Waweru, 1976). Because of the
widespread reliance on hand tools and the lack of a landless labor class, the area
cultivated per farm family critically depends on the size and composition of the family
labor force.
While crop production is the major activity of smallholders, off-farm activities such
as trading, small-scale industry, livestock, and fishing are important activities throughout
sub-Saharan Africa. Micro-economic research has shown that farmers devote a signifi-
cant amount of their time to off-farm activities (Luning, 1967; Liedholm and Chuta, 1976)
and that there is an inverse seasonality with a large percentage of the total hours being
worked off-farm during the dry seasonss. Although small farmers strive to meet their
own food needs, 20-40 percent of staple food production is marketed in most countries.-

/Even though a farmer does not "own land" in the sense of having freehold title to
land, farmers in most countries have "control" or use rights to land for their lifetime;
these use rights can be passed on to heirs.

/For example, from 30 to 40 percent of small farm production was sold in Kenya
(Heyer and Waweru, 1976), 48 percent in Sierra Leone (Byerlee et al., 1977), and 24
percent of the total value of farm production in northern Nigeria (Norman, Pryor, and
Gibbs, 1979).

Large-Scale Farming: An Overview
Large-scale farming in sub-Saharan Africa dates to the colonial period with the
introduction of plantations and large European farms which produced for export markets.
Today, large-scale farming accounts for a substantial portion of export crop production in
only a few countries in sub-Saharan Africa.
In the 1960s, Western advisors generally endorsed large farms and plantations
because it was assumed that they would benefit from economies of scale, that they would
be convenient vehicles for newly independent governments to "bring rapid development"
to selected rural areas, and that they would provide rural employment for the growing
number of school leavers./ The rationale of the transformation approach is reflected in
the recommendation by the International Bank Mission to Tanganyika (later Tanzania) that
the Government of Tanzania should support land settlement schemes because "quicker
progress towards these ends is likely to be made, within the limitations of the resources
available for government action, by planned settlement of empty areas than through
exclusive concentration on improvement of methods (small farms) in settled areas" (IBRD,
1961, p. 131).2/ Success stories of large-scale agriculture from the colonial period such as
the world famous Gezira scheme in the Sudan (Gaitskell, 1959),3/ tea plantations in East
Africa, Firestone rubber estates in Liberia, and Unilever estates in the Belgian Gongo
(now Zaire) were often cited in the 1960s as examples of the superiority of large-scale
agriculture. But proponents of large schemes often overlooked or glossed over the
horrendous failures of large-scale schemes such as the East African groundnut scheme
introduced by the British colonial service in Tanganyika after World War II, the failure of
Mokwa settlement scheme in northern Nigeria in the 1950s (K. D. S. Baldwin, 1957), and
the mixed results with land settlement schemes and state farms in Africa

IA term widely used to describe recent graduates of primary and secondary schools
who are trying to enter the labor force.
2/The Government of Tanganyika followed the advice of World Bank experts and
started 23 settlement schemes over the 1963-67 period which featured heavy capital
investment, government management, and little participation by the settlers (Ingle, 1972).
But the 23 schemes met with little success and in 1967 Tanzania dropped the transforma-
tion approach and shifted to a rural socialist strategy which focused on helping small
farmers through the Ujamaa program and later the village development scheme (Nyerere,

2/For a critical appraisal of the Gezira scheme from a radical perspective, see
Barnett (1977).

(Chambers, 1969) and throughout the world (W. A. Lewis, 1964; FAO, 1976; Higgs, 1978).!/
State farms were another type of farm organization included in the transformation
strategy of agricultural development in the 1960s. State farms were adopted by the
Governments of Sierra Leone and Ghana in the late fifties and early sixties. Sierra Leone
established state farms to produce export crops in each of 12 regional provinces but the
farms were abandoned within a few years. In Ghana, state farms and tractor mechaniza-
tion were the centerpieces of Nkrumah's socialist strategy of development (Miracle and
Seidman, 1968a). Ghana's dismal record of tractor mechanization and state farms was
documented by Kline et al. (1969), Nweke (1978a,b), and Killick (1978).
Plantations were another type of large-scale farming pursued by African leaders in
the 1960s. Since it is almost impossible to gain access to data on private and government
plantations, it is difficult to pass judgment on the economic, social, and political costs,
and returns on plantations. Studies by Saylor and Eicher (1970) and Essang and
Ogunfowora (1975) of plantations in Nigeria are examples of the limited research on the
economics of plantations. Since British colonial land policy prevented plantation
development in Nigeria, plantations were insignificant until the constitutional change of
1951 permitted the establishment of plantations by both private and foreign capital.
Saylor and Eicher (1970) found that government plantations in Nigeria were generally
unprofitable because lack of technical data, poor management, and high turnover in
unskilled labor (frequently 100 percent per year), etc. The number of private plantations
increased during the 1951-65 period in Nigeria but marketing board taxes on oil palm and
rubber reduced the rate of return on plantations to almost zero by the early 1960s
(G. L. 3ohnson, 1968).

Small-Scale and Large-Scale Farming in Nigeria
Since both the transformation (large-scale) and improvement (small-scale) strategies
were pursued in different regions of Nigeria in the 1960s, Nigeria provides a unique case
study of the results of these two alternative strategies. During Nigeria's first Develop-
ment Plan (1962-68), the three regions in the southern part of the country (Western, mid-
Western, and Eastern) devoted some 70 percent of their capital and recurrent budgets in

!We should not give the impression that all settlement schemes in Africa are large-
scale and all settlement schemes are failures. Numerous examples of settlement schemes
for smallholders are found in Eastern Africa, including the famous one million acre
settlement scheme in Kenya which was conceived by R. Swynnerton and launched in 1953
to transfer land from large white farmers to small Kenyan farmers (Kenya, 1954). For an
appraisal of Kenya's settlement schemes, see MacArthur (1975) and Clayton (1978). For a
discussion of spontaneous settlements in Kenya, see Mbithi and Barnes (1975); in Senegal,
Rocheteau (1975); and in the Sahelian countries of West Africa, see CILSS/Club du Sahel

agriculture to the transformation approach (farm settlements, school leaver farms, and
plantations). On the other hand, the Northern region pursued an improvement strategy
during the 1962-68 Plan to help small farms through subsidized fertilizer, credit, and
farmer training centers. The political urgency of "getting on with development" and
"bringing development to the people" explain why the three southern regions in Nigeria
opted for crash schemes. As Nigeria approached independence in the late fifties, political
leaders in the three regions in the southern part of the country were not in a mood to wait
for research results on whether to pursue large-scale or small-scale agricultural develop-
ment strategies. For example, in 1959--one year before independence--a policy paper
issued by the government of the Western Region noted that "while scholars conduct
unbiased research ... agricultural development must go on" (Nigeria, 1959, p. 9). The
policy paper noted that political leaders from western Nigeria had visited the Gezira
scheme in the Sudan and Moshav settlements in Israel and concluded that experience in
other countries has shown that "... a system of co-operative farm settlements would be a
major step in the agricultural development of the Region" (Ibid).
In a detailed analysis of the western Nigeria's settlement scheme, Roider (1971)
found that after six years of operation, the government had spent $11,200 per settler, or
double the amount originally projected, while yields ranged from 25 percent (cotton) to 65
percent (rice) of the yields estimated in the feasibility study. Similar settlement schemes
were tried in the other two regions (Eastern and mid-Western) but by the end of the 1960s
it was obvious that the settlement schemes had failed in all three regions in southern
Nigeria (Andreou, 1981). The reasons for the failure of schemes in southern Nigeria were
almost identical to the findings of W. Arthur Lewis (1964) in his review of settlement
schemes and Nelson's (1973) study of 23 schemes in Latin America: lack of technical and
microeconomic data, superficial planning, overinvestment in housing and social services,
inappropriate mechanical technology, and lack of participation by settlers.
Eicher and Johnson (1970) evaluated the consequences of pursuing transformation
versus the improvement strategies and concluded that smallholder improvement programs
rather than land settlements or plantations should form the backbone of Nigeria's
agricultural strategy over the 1969-85 period. These findings were reinforced by Wells'
authoritative book (1974) on agricultural policy and plan implementation during Nigeria's
First Plan (1962-68).

Small-scale farming has many advantages relative to large-scale farming in Africa.
Although few donors support large-scale farms and ranches today, there is still substantial
support among African politicians and policy makers for large-scale agriculture. For

example, large-scale projects are still of central importance in Nigeria and in Ghana.
Moreover, foreign private investors have launched large-scale food production schemes in
several countries. But these large-scale schemes are experiencing unforeseen technical
and economic problems. For example, Uni-Royal's large-scale food production complex in
Liberia was terminated in the late 1960s. A $2.1 million foreign-financed maize farm and
grain storage complex in central Ghana also experienced numerous difficulties and it is
now operated as a state farm. Several observers have reported that large-scale wheat
schemes have failed in Tanzania.l In the Cameroon in 1979, we have unofficial reports
that a 4,000-hectare government wheat mechanized scheme with 35 tractors did not
produce enough--22 kilos of wheat per hectare in 1979--to recover its seeding rate of 100
kilos per hectare.
Although it is understandable why governments do not publicize the failures of many
large-scale farming schemes, we have gleaned enough information to conclude that in
most countries, large-scale/capital-intensive food production complexes cannot compete
with African smallholders for meeting staple food needs in the 1980s.- To be sure, there
is a limited scope for a few multi-national firms to produce fresh fruit and cut flowers for
European markets (FAO, 1976d). For example, string beans are flown from Upper Volta to
Europe during the winter season. A multi-national firm producing strawberries and other
fruit in Senegal for European markets went bankrupt in 1980. Delmonte has a 10,000
hectare pineapple plantation in Kenya. But there is very little empirical research on
these multi-national firms. The controversy surrounding multi-national firms is discussed
by Widstrand (1975) and Sklar (1976).
Most large-scale farming and land settlement schemes in Africa have been failures
over the past 50 years. But settlement is still an important topic being debated because
according to FAO estimates, two-thirds of the remaining arable land in the world is in
sub-Saharan Africa. The challenge is to learn from past settlement schemes. Research
has shown that the role of the government should shift from planner, finance, and
manager to providing agronomic research, disease control, and a minimum of infrastruc-
ture. Government policy should help African families build their own houses and clear
their own land in order to drive down the capital cost per settler.

1/But several reviewers pointed out that it is premature to write off wheat in
Tanzania. The economics of wheat production is an empirical question which should be
addressed by researchers in East Africa.

-/Research and empirical findings on the economics of smallholder farming are
reviewed in Part IV.

Marketing Boards and Food Grain Boards
The control of the agricultural surplus is a common denominator of both the colonial
and postindependence periods. Various taxes (head taxes, hut taxes) and compulsory
planting of selected export crops were imposed by colonial authorities to stimulate the
production of export crops and to capture the agricultural surplus. Shortly after
World War II, the British colonial governments introduced marketing boards in their East
and West African colonies following the relatively successful record of marketing boards
in Australia and New Zealand since the 1930s. The objective of the marketing boards was
to stabilize producer prices and foreign exchange earnings and to reduce interseasonal
price movements.
In the 1960s and 1970s, numerous African governments introduced grain boards to
control producer prices of food grains and to channel food to the urban centers.-' Boards
usually accumulate and carry stocks to mitigate both intra-annual and inter-annual
fluctuations in price and supply and develop distribution systems to facilitate the transfer
of grain from surplus to deficit regions.

Marketing Boards
The introduction of marketing boards was followed by a wave of studies by
economists. Leading the charge against the marketing boards was Peter Bauer who
contended in his West African Trade (1954) that marketing boards failed to stabilize
producer prices and reduce seasonal price variations, and that the boards dampened
producer incentives by paying producers one-half to two-thirds of world prices of exports
such as cocoa, oil palm, coffee, and rubber. Later, A. G. Hopkins (1973) showed that
export producers in the Gold Coast (Ghana) lost 41 percent and Nigerian producers lost 27
percent of their potential gross income through marketing board taxes over the 1947-61
Helleiner's meticulous study "The Fiscal Role of the Marketing Boards in Nigeria"
(1964) introduced new criteria for examining the performance of the boards. Instead of
being preoccupied with the issues of whether marketing boards stabilized producer prices,
producer incomes, and foreign exchange earnings, Helleiner contended that the more
important issue is whether the boards are effective in performing the fiscal role of
capturing the agricultural surplus for the development of both urban and rural areas.
Helleiner was of the opinion that no single taxing scheme could simultaneously achieve
multiple goals of stabilizing producer prices, foreign exchange earnings, and interseasonal

/While it is common to refer to these boards as government boards, most food grain
boards have a quasi-commercial character and some degree of autonomy from the

price variations. Helleiner concluded that, on balance, marketing boards were the best
mechanism for mobilizing the agricultural surplus in subsistence economies like Nigeria in
the 1950s and 1960s because of the lack of administrative capacity to impose other taxes
such as land and income taxes and the lack of other sources of revenue such as petroleum,
gold, phosphate, and timber. But Helleiner's qualified endorsement of marketing boards
was based on a shaky micro data base; he was unable to show that the rates of return on
government investments (financed by marketing board surpluses) in plantations, hotels,
airlines, and industrial estates were unambiguously higher than if Nigerian farmers had
received higher prices for their export crops and had reinvested their expanded earnings in
farming (e.g., such as the new hybrid oil palms) or if the boards had indirectly returned
the surplus to farmers through fertilizer subsidies and agricultural research. Idachaba
(1973) concluded that marketing boards in Nigeria substantially dampened producer
incentives and restricted output and employment generation in agriculture. Olayide,
Ogunfowora, and Essang (1974) found widespread inefficiencies in Nigerian marketing
boards and recommended alternative structures such as producer and marketing coopera-
tives. For other research on marketing boards, see Storm (1976) for Senegal; for
Tanzania, see Kriesel et al. (1970); in Ghana, see Kotey, Okali, and Rourke (1974); and in
West Africa, see Blandford (1979).
During the past 30 years, taxes on export crops via marketing boards have provided
a convenient way in many countries to capture and transfer the agricultural surplus to
finance government airlines, hotels, factories, plantations, and in a few cases subsidized
inputs such as fertilizer for farmers. Whether export taxes and marketing boards should
have been scrapped or continued over the past 30 years needs to be analyzed on a country-
by-country basis. The central questions have been the level of taxes imposed by the
boards, the use of the agricultural surpluses, and whether there were fiscal alternatives to
the boards. Clearly, countries such as Nigeria, Ghana, and Tanzania had few fiscal
alternatives to marketing boards in the 1950s and 1960s because they lacked the
administrative capacity to register land and collect land taxes and they lacked mineral
and petroleum exports. Marketing boards--like import substitution in Latin America in
the 1950s--served a useful role at a particular stage of economic history of some African
countries. A few countries, such as Nigeria, have been able to find alternative sources of
revenue in the seventies and have been able to reduce the tax burden on export crops. But
the problem of raising government revenues in agrarian economies without mineral or
petroleum exports remains a central issue in the 1980s.

Food Grain Boards
Several African governments have given food grain boards monopoly power over
domestic distribution of food grains, making private trade in grains illegal. An important
argument in favor of governmental monopolies in food grain trade is that it allows
governments to subsidize trade in remote, less productive areas. It is often stated that
boards need to be able to control the distribution of 20 to 25 percent of total grain
production to affect prices (Becker, 1974; Sorenson et al., 1975; Groleau and Kohler, 1979)
but it is common for private, and often illegal, trade in food grains to handle 85 percent
or more of the marketed surplus in countries with grain boards. As a result, even where
grain boards have been made legal monopolists, monopoly power is often more legal
fiction than fact (Lele and Candler, 1981). The inability of grain boards to dominate the
flow of grain makes it almost impossible for them to stabilize prices and to transfer grain
from surplus to deficit regions.
Many researchers have argued that the same issues plague food grain boards as
export marketing boards faced in the '50s and '60s: inefficiency, dampened producer
incentives, corruption, and a cadre of thousands of employees who are idling away their
time.- There is substantial evidence that administrative and operating costs of most
grain boards are quite high, reflecting inefficient operations (Sorenson et al.; Temu, 1975;
Wilcock, 1978; CILSS/Club du Sahel, 1977; Grolleaud and Kohler, 1979; Blandford, 1979;
Lele and Candler, 1981). While some boards have managed to reduce per unit costs over
the years (e.g., The Gambia Produce Marketing Board in Gambia; Blandford), the
operating costs of most have escalated mainly because of high overhead, thousands of
employees, and fleets of trucks. The marketing margins of grain boards have tended to be
high, clearly above those of private traders for comparable services (Berg, 1975). Large
margins effectively mean that a large percentage of the total sales value accrues to
government boards rather than producers.- This factor is particularly important since
high margins necessitate payment of low farm gate prices if retail prices are to remain at
a politically acceptable level.
Peter Temu showed that not only have mark-ups by grain boards in Tanzania been
high relative to private traders, they have been extremely variable, largely because of the

!When Senegal abolished its grain board--ONCAD--in late 1980, it eliminated
about 4,500 employees. Likewise, Mali's grain board--OPAM--is reported to have
between 6,000 and 8,000 employees on its payroll.

!/The World Bank (1981b, p. 59) reports, for example, that charges for marketing,
storage, and transportation in Kenya accounted for 34 percent of the f.o.b. border price
for maize, 23 percent for wheat, and 48 percent for rice during 1972-79. These figures
are typical of government grain board margins in other countries.

inability of boards to control their costs. Many governments have adopted a dual policy of
setting guaranteed producer prices in order to reduce the uncertainty faced by farmers
and controlling consumer prices of key staple foods. The margins available to boards are
a residual of the two controlled prices and frequently do not cover costs in any given year
(CILSS/Club du Sahel, 1977a, 1979).
Several researchers have found that improvements in operational efficiency should
enable most grain boards to carry out their duties with smaller margins. Sorenson et al.
and Temu show, for example, that administrative costs are a large component of the
margins received by grain boards. Several researchers have argued that extensive
duplication of functions among grain boards underlies some of these costs (Kriesel et al.,
1970). There are widely acknowledged problems of losses of grain during storage and
transportation. Supplementary urban grain storage with a network of rural storage
centers may reduce transport costs by eliminating the transfer of grain from rural areas
to capital cities at harvest and the redeployment of the grain both to rural areas in times
of crisis. Some grain boards are not taking advantage of opportunities to offset their
costs by generating revenues through spatial price arbitrage. Thodey (1969) shows, for
example, that price differences in Ethiopia exceeded transportation costs over half of the
time on routes between markets connected with Addis Ababa.
There is now growing evidence that grain boards are often a constraint on the
production and marketing of food. Wilcock (1978), for example, argued that the activities
of grain boards in the Sahel have contributed to the misallocation of resources because of
overcentralization, high operating costs, duplication of effort, and dysfunctional incentive
systems. Heyer and Waweru (1976) reported that the high degree of regulation and
control of food marketing has resulted in low producer prices and high consumer costs in
Kenya. The government of Senegal's decision to abolish its grain board--ONCAD--in late
1980 is a rare example of a country admitting that its board was ineffective. Berg (1975,
1979) has argued that the absence of a suitable substitute for the private trader in
primary markets is a major constraint on food self-sufficiency in the Sahel. While most
researchers have argued for minimizing the role of governments in food grain trade, Berg
contends that even mixed (government/private) marketing arrangements are unworkable.
Berg rejects what he calls the "imprudent peasant/monopolized market" model which he
feels dominates the views of African planners, offering as an alternative a "prudent peas-
ant/competitive market" model. Of three potential alternatives governments might
consider, including continuation of the status quo, increased public control, and competi-
tive liberalization, Berg firmly believes only competitive liberalization will work to the
benefit of both producers and consumers.

One of the primary results of creating legal monopolies for food grain boards has
been widespread illegal marketing which involves a volume often several-fold that of legal
marketing (Collins, 1976). Temu argued that the failure of grain boards to control
domestic food grain distribution systems even in countries where they are given legal
monopoly power can largely be traced to the disincentive effect of low and uncertain
producer prices offered by grain boards. Temu examined three basic approaches to
minimizing illegal marketing in Tanzania: (1) change the legal price so it better reflects
supply and demand conditions, (2) withhold and release stocks to counteract extreme price
fluctuations, and (3) improve enforcement of statutory prohibitions. After examining the
economic costs of the alternative approaches, Temu rejected the statutory approach as
unfeasible. He concluded that grain boards would be more effective if they reduce the
scope of their activities to a few specific functions, allowing a greater play of market
forces in Tanzania's economy.

Agricultural Prices
The manipulation of agricultural prices is a standard government technique to
influence the level and composition of agricultural production and the transfer of the
surplus to urban areas. The primary mechanism is to set prices of major agricultural
commodities and agricultural inputs administratively and to enforce the administered
prices through the buying and selling operations of parastatals and licensed buying agents.
African governments also influence relative agricultural prices through indirect pricing
policies, including export taxes, subsidized credit and fertilizer,- and overvalued
exchange rates. The major reasons for intervening in agricultural pricing have been to
(1) stabilize prices and production, (2) foster self-sufficiency, (3) generate tax revenue,
(4) curb the profit of middlemen, and (5) control the cost of living for urban consumers.
There have been few studies of the aggregate impact of pricing policies. Helleiner's
(1964, 1966a) studies of the fiscal role of marketing boards in Nigeria and his (1968a)
review of pricing strategies in Tanzania are among the exceptions. Helleiner argues that
the key in analyzing pricing policies is to ensure that the structure of price and tax
incentives--not the level--is working in the right direction. Helleiner concluded that
"clearly, the income distribution policy implicit in Tanzania's wage and agricultural
pricing policies has worked to the increasing disadvantage of the smallholder agricultural
sector" (1968a). Olayide, Ogunfowora, and Essang (1974) used systems simulation to test
the effects of marketing board pricing policies on the Nigerian economy; the authors

/For example, the government subsidizes 80 percent of the fertilizer prices paid by
farmers in northern Nigeria in 1981 (World Bank, 1981b).

concluded that depressed producer prices had reduced the growth of the economy and they
recommended the elimination of licensed buying agents and centralized fixing of prices.
On the question of input subsidies in Nigeria, see Idachaba's pioneering articles (1973,
During the 1970s, numerous researchers documented how agriculture has been
adversely affected by the impact of pricing policies designed to promote domestic
industry. For example, the ILO report on the Sudan found that the net incidence of taxes
and subsidies discriminated against agriculture in favor of industry and commerce (ILO,
1976). Reviewing policies in Kenya, Heyer, Maitha, and Senga (1976) report that despite
policy statements to the contrary, government programs are still heavily in favor of urban
and nonagricultural activities and that there was a large and growing net resource flow
out of the agricultural sector throughout the 1960s and 1970s. Evidence that pricing and
marketing policies followed by the Zambian government have had an adverse effect on the
rural-urban income gap and on self-sufficiency in food is presented in Doris Dodge's (1977)
study of agricultural policy in Zambia and in Maimbo and Fry's (1971) study of the terms
of trade between agriculture and the nonagricultural sector. For further references on
the impact of pricing policies in Zambia, see Kinsey (1978), Daniel (1979), and Turok
Byerlee, et al.'s (1982) analysis of employment-output conflicts and factor price
distortions in Sierra Leone reveals that (1) there is a wide choice of technology available
not only between large-scale and small-scale sectors, but within each sector; (2) the
choice of technology was found to be sensitive to relative factor prices; (3) administra-
tively established prices, wages, taxes, and subsidies have favored the adoption of larger-
scale, capital-intensive techniques in agricultural production, processing, fishing, and
industry; and (4) that a continuation of present government policies will have a serious
adverse effect on rural employment and national income.
The World Bank report (1981b) presents substantial support for the view that pricing
policies are a root cause of the food and agricultural crisis in Africa. In one example, the
authors argue that severe cutbacks in production in Tanzania and Ghana resulting from
low official prices have dissipated their export positions. In 1980, Tanzania's total exports
of major agricultural commodities (which account for nearly two-thirds of the total value
of all exports) were 28 percent lower than in 1966 and export earnings fell from 25
percent of the GDP in 1966 to 11 percent in 1979. Ghana has long been the world's
leading cocoa exporter but Ghana's cocoa production declined from a peak of 566,000
metric tons in 1965 to 249,000 tons in 1979 and Ghana's position in world trade of cocoa

fell to third in 1979, behind the Ivory Coast and Brazil.-' For the political economy
scholar who criticizes neoclassical economists for their fetish concern over getting prices
right (removing subsidies and taxes), one only has to provide empirical evidence from
Ghana to illustrate what happens when prices are distorted. Ghana's drive to promote
large-scale rice farming in northern Ghana in the mid-1970s through taxes, subsidies, and
an overvalued foreign exchange rate brought forth imported rice combines in a low wage
economy with substantial unemployment and rural underemployment (Winch, 1976). This
same mistake is being repeated in 1981 in Ghana's drive to promote state farms and
mechanized cotton harvesting. In mid-1981, the official exchange rate of 2.6 Cedis for
$ US 1 was in stark contrast to across the border in Togo where it took 33 Cedis to
purchase one dollar. So state farms use the 2.6 to I ratio to import $100,000 cotton
pickers. State farm managers are reporting a financial profit while small private farms
cannot even buy fertilizer on the market. Getting prices right is not the answer to
Ghana's economic problems but a move in this direction is a prerequisite for economic
reform in Ghana.
In summary, there is now substantial evidence that agricultural pricing policies have
tended to have an adverse effect on: (1) the gap between rural and urban income, (2) the
incentive to produce food and export crops, (3) the ability of governments to establish and
maintain food reserves, and (4) employment opportunities in farming, processing, and rural
industries. Surveys of agricultural pricing policy include De Wilde's (1980) case studies of
Kenya, Tanzania, and Ghana; Bates (1981); Mukui (1979); Lele and Candler (1981); Bovet
and Unnevehr (1981); and World Bank (1981b). For a devastating critique of pricing
policies in Tanzania, see F. Ellis (1980).

Rural Development Programs
Rural development programs have received a great deal of attention by policy
makers, scholars, and donor agencies over the past two decades./ The historical

i/Although these figures display an alarming loss of export markets for Tanzania
and Ghana, the decline is overstated because of commodities smuggled out of the country.
For example, smuggling of cocoa out of Ghana has long been the farmer's response to
harsh taxes. In Tanzania, in 1973/74, it was widely known that farmers smuggled maize
across the border in response to lower government producer price of maize. But even if
official export figures are understated by 10-15 percent for some commodities such as
cocoa in Ghana, the above figures do illustrate how Tanzania's and Ghana's export
positions have eroded.

-2The distinguishing characteristics of rural development programs are that devel-
opment is broadly defined to include increased rural welfare as well as increased
agricultural productivity and, in many cases, includes the notions of increased participa-
tion and broadly shared benefits. See 3. Cohen (1980) for a discussion of the definition of
rural development.

experience with designing, implementing, and evaluating rural development programs in
Africa is vast and would require a separate review. In this section, we shall mention only
a few of the important references, starting with the community development literature of
the 1950s.

Community Development and Animation Rurale
During the 1950s and early 1960s, rural development was primarily promoted
through community development (CD) and animation rurale (AR) programmes. CD
emerged from experiences in the United States in the 1930s and 1940s and from England
in the form of Fabian socialism after World War II. CD programs were introduced in the
British Colonies in the late 1940s with emphasis on building bridges, schools, and health
clinics. CD was viewed as a peaceful way to mobilize people to help themselves in
meeting their felt needs. India and Pakistan, the Philippines, and Korea were the primary
CD laboratories of the 1950s. The rise and fall of the CD experience of the 1950s in Asia
is chronicled in Holdcroft's valuable survey (1978). Holdcroft summarizes the reasons for
the rapid rise and fall of CD as follows:
1. CD architects overlooked class conflict in rural areas and assumed that CD
agents working at the village level could mobilize people to help themselves.
But, in practice, a CD worker was an all purpose worker with few technical
skills in agriculture and did not have the means to help farmers gain access to
credit and the landless to acquire land.
2. Since increasing agricultural production was not the main objective of CD
programs, when food crises emerged, governments shifted resources to minis-
tries of agriculture to step up food production.
3. CD was zealously promoted as a separate strategy from agricultural develop-
ment and, as a result, there was rivalry between the old line ministries--agri-
culture, health, education--and newly established ministries and departments of
community development.
CD was transferred to the Caribbean (e.g., Jamaica) and to several English-speaking
countries in Africa by the British colonial service in the 1950s. For an overview and
assessment of CD in Nigeria, see Jackson (1956). But CD never received the prominence
in Africa in the 1950s that it received in Asia.
Since 1960, a number of francophone countries in Africa have adopted animation
rurale as a means of mobilizing and educating their rural populations (Charlick, 1980). AR
essentially has played the same role in Francophone Africa as did CD in the former British
colonies. For a review of animation rurale programs, see Elliott (1974), Lele (1975), and
Charlick (1980). Gellar, Charlick, and Jones (1980) point out that except for a brief
period in Senegal AR has been used as a technique for fostering local organization and for
non-formal education rather than as a comprehensive rural change strategy. Two of the

key elements of AR are dialogues with villagers and the development of a network of
local animators to increase local organizational capability. In his review of AR in
Cameroon, Upper Volta, Senegal, and Niger, Charlick argues that in most AR programs
animation agents served as promoters for the programs of other technical services instead
of helping facilitate development from the bottom up and that nearly all important
decisions concerning development programs were made by government agents and then
imposed on villagers. By the end of the 1960s, the broadly defined populist mission of AR
programs had, in most countries, given way to a more limited role in which AR agencies
focused on a few tasks such as informal education (Gellar, Charlick, and Jones).
Although CD or AR agencies are still present in most African countries, they play a
limited role in the major policies affecting rural people in Africa. They often serve as
scapegoats allowing technical services to ignore the human side of rural production
programs. For example, a rural engineer may consider where a dam should be placed to
maximize its holding capacity while the CD or AR agent is called upon to generate
support for the project among the local population and to try to resolve conflicts in land
use rights. In retrospect, both AR and CD movements had two basic flaws: they
underplayed the technical constraints and they were too optimistic about the degree of
national political support for decentralized development (Gellar, Charlick, and Jones,

Integrated Rural Development
The 1970s can be labelled the decade of integrated rural development (IRD). IRD
projects were introduced in the late sixties and early seventies throughout the Third World
in the wake of the acknowledged failure of the Green Revolution to have much impact in
Latin America, the Caribbean, and Africa. IRD in Africa received a boost at the famous
Kericho Conference on rural development in Kericho, Kenya in 1966, when political
leaders emphasized the need to give attention to employment generation and rural
development projects (see Sheffield, 1967). Well-known IRD projects in Africa include the
Chilalo Agricultural Development Unit (CADU), Wolamo Agricultural Development Unit
(WADU), and various minimum package programs (MPP) in Ethiopia; the action priority
zones program (ZAPI) in Cameroon, the Lilongwe land development program (LLDP) in
Malawi, and the Special Rural Development Program in Kenya. These IRD projects were
based on the assumption that a critical minimum effort was necessary to have a
noticeable impact on target populations in a short time. The projects were therefore
concentrated in a limited area and were administered through semiautonomous agencies
which paralleled other government agencies.

Uma Lele's (1975) review of 17 IRD projects in Eastern, Southern, and West Africa
in the 1960s and early 1970s is a standard reference. Using a classification system based
on degree of integration, breadth of objectives, size of target population, and funding
agencies, she identified four major categories of programs, of which the most important
are the commodity programs (e.g., the Kenya Tea Development Association) and the
regional and national rural development programs. Lele found that most of the projects
evaluated were based on inadequate knowledge of technical possibilities and small farm
conditions and exhibited little understanding of the local institutional environment. She
argued that the key to success of IRD projects is the systematic acquisition of local
knowledge and flexibility in the course of implementation. Because of the severe lack of
trained manpower, she recommends beginning programs with only the few simple
interventions to remove critical constraints and then phasing in other programs. Lele
contends that it seems unlikely that the establishment of autonomous project authorities
will lead to the achievement of IRD objectives in the long run.
H. Dupriez (1978) reviewed several IRD projects funded by the European Economic
Community. Dupriez found that the administrators of these projects tended to regard
rural communities as undifferentiated masses and therefore frequently ignored social
structures and economic and political hierarchies. Dupriez was critical of the lack of
attention given to participation in many projects. For example, he was unable to find a
single instance in which decentralization actually allowed farmers an opportunity to
participate in decision making affecting projects.
Cohen (1980) analyzed IRD in a historical perspective and found that the common
denominators of IRD projects are a focus on small farmers and an attempt to promote
improvements in the quality of rural life and increase off-farm opportunities. According
to Cohen, the two major gaps in the IRD literature are the weak theoretical foundation
and the lack of a strategy to provide guidelines for practitioners. Cohen contends that
the increased scope of rural development activities during the 1970s exceeded the design,
implementation, and evaluation capacities of national governments and donor agencies.
Starting in the late 1970s, interest in IRD projects rapidly began to decline. In 1980
and 1981, the pendulum shifted from IRD to strategies which emphasize food production
by small commercial farmers. The decline of IRD does not reflect a retreat on equity
goals as much as growing recognition that pilot IRD programs rarely if ever were
implemented on a broader scale, that governments cannot afford to finance a wide range
of social services during the early stages of development, and that IRD (like CD in the
1950s) was not solving the most fundamental rural problem--achieving a reliable food
surplus. Although many IRD projects will continue to carry the IRD label in the 1980s, we
are of the conviction that agricultural production concerns will form the core of these

projects.- Moreover, as Lele (1981) pointed out, the stark reality is that there always
was a large gap between the donors' equity-oriented objectives and national governments'
goals of long-run modernization, growth, institution-building, and economic independence.
For an assessment of rural development projects in Kenya in the late sixties and the
seventies, see Heyer, Ireri, and Moris (1971); University of Nairobi (1975); and Heyer,
Maitha, and Senga (1976). For a recent book on IRD in Nigeria, see the volume edited by
Olayide, Eweka, and Bello-Osagie (1980). F. S. Idachaba (1981) presents an overview of
the lessons from IRD programs in Nigeria. For discussion of rural development projects in
Ethiopia, see 3. M. Cohen (1975); Tecle (1975); Sisaye and Stommes (1980); and B. Akilu
(1980). Hyden (1980) presents an insightful analysis of the Ujamaa program in Tanzania,
which stressed rural mobilization and collectivization as an approach to rural develop-
ment. Tanzania'a rural development experience is well documented in the bibliographies
by Kocher and Fleisher (1979) and McHenry (1981). Atayi and Knipscheer (1980) evaluate
the ZAPI program in Cameroon. For an overview of rural development policies in
Botswana, see Picard (1979). For a skeptical view of IRD, see Ruttan (1975).

Accelerated Food Production Campaigns
During the 1970s, several countries launched accelerated food production programs
to reverse the long decline in food production per capital and to reduce dependence on
food imports. Paralleling the attempts of Latin American countries to pursue import-
substitution industrialization in the 1-950s and 1960s, many African nations are finding
that "food import-substitution" programs cannot compensate for the long-term neglect of
the agricultural sector and overcome the many infrastructural and technical constraints
on food and agricultural production. For example, when Ghana launched its "Operation
Feed Yourself" in 1972, the government acknowledged that the historical priority on
cocoa and oil palm production was a critical factor in the decline of food production.
During the early phases of Operation Feed Yourself, the government invested heavily in
large-scale farms which turned out to be expensive and ineffective (Nweke, 1978b).
Moreover, the program was placed under the Ministry of Agriculture while research on
small farmers was in another ministry so there was little coordination of efforts. Finally,
Girdner et al. (1980) point out that if self-sufficiency in food could be achieved, Ghana
would remain heavily dependent on cocoa for export earnings.
In the early 1960s, Nigeria was a net exporter of food--mainly oil palm and
groundnuts--but by the early 1970s, Nigeria was importing food. Nigeria imported 1.4

/See the lack of attention given to regions of poor resource endowments and to
equity objectives in the World Bank's (1981) strategy for African development. The Bank's
strategy for African development in the 1980s could be labelled as a growth strategy
reminiscent of the 1960s.

million tons of basic staples in 1977 and this figure is projected to increase by 1985. The
value of Nigeria's food imports in 1981 was $1.3 billion. The prognosis is bleak. There is
no easy way for Nigeria to solve its food crisis in less than 10 to 15 years. In evaluating
Nigeria's National Accelerated Food Production Program, Abalu and D'Silva (1980b) link
Nigeria's present food crisis to repressive marketing board policies and they express little
hope that the accelerated food production program will be successful. Nigeria has
established a Green Revolution Committee to explore how to speed up food production
(Nigeria, 1980a,b). Also a number of important food policy studies have been recently
carried out under the leadership of Francis Idachaba (1980b, 1981; and Idachaba et al.,
1981). Idachaba points out in Food Policy in Nigeria (1980b) that there is a need to move
from ad hoc responses to a policy framework which examines the interrelationships among
sectors, subsectors, policy instruments, programs, and projects.
Sudan is one of the most glaring examples of the failure of a country to mobilize its
agricultural sector as an engine of growth and to feed its people. In the mid-1970s, it was
frequently asserted that the Sudan could become the "bread basket of the Middle East" by
drawing on OPEC loans and gifts to develop its vast reserve of idle land (Kiss, 1977). But
the issue today is not one of exporting food to the Middle East but one of Sudan's inability
to feed its 18 million people. For example, Sudan imported $30 million in US PL 480 food
in 1980 and it is now facing severe balance of payment problems and inflation. Sudan's
role as the bread basket of the Middle East remains a dream.
West Africa is now importing around I to 1.6 million tons of rice each year.
Economists from the Food Research Institute at Stanford University and the West African
Rice Development Association (WARDA) have recently completed a study of the rice
industry in five countries in West Africa.-/ The Stanford/WARDA study addresses three
main issues: (1) the private and social profitability of producing rice, (2) the effect of
governmental policies on the production, consumption, and trade of rice, and (3) the
potential for increased interregional trade. The findings reported in Pearson et al. (1981)
reveal that most techniques of rice production in West Africa are privately profitable but
the incentive to grow rice varies greatly among the countries, being the lowest in Mali
and Liberia. In terms of deliveries to urban centers, rice was found to be socially
profitable only in Mali and Sierra Leone. Even though the social profitability of locally
produced rice is greater in remote regions (because of high transportation costs), a
substantial proportion of local production in the Ivory Coast, Senegal, and Liberia could
not succeed without restrictive trade policies. An interesting finding is that, contrary to
general belief, West African governments did not subsidize rice consumers from 1965 to

!/The study centers on the Ivory Coast, Liberia, Senegal, Mali, and Sierra Leone.

1976. During this period, domestic rice prices were from one-quarter to one-third higher
than the cif import prices in four of the countries. Mali is the sole exception. The major
conclusions of the Stanford/WARDA study are: (1) Mali and Sierra Leone can achieve rice
self-sufficiency while promoting efficient use of resources, and with some production and
processing techniques could profitably export rice;/ (2) the Ivory Coast and Liberia
should emphasize other more profitable crops and import rice during periods of deficit,
and (3) Senegal needs to concentrate rice production in particular regions (such as the
Casamance region in southern Senegal) if it wants to increase food security without a high
cost in efficiency.2.
Among the bright spots on the food production front are food production programs in
Malawi and Zimbabwe (Blackie, 1981). In 1980, Zimbabwe exported 500,000 tons of
maize. Zimbabwe's 1981 maize crop is a record 2.5 million tons and approximately 1.0
million tons are available for export. On the other hand, neighboring Zambia is importing
a large percentage of maize, its staple food. But the question of food security has an
important political dimension. Although Zimbabwe is an exporter and Zambia an importer
of maize, they both face a common political problem--the lack of proven maize packages
and supporting institutions for small farmers. In Zambia, about 400 commercial farmers
produce an estimated 40 percent of the marketable surplus of maize. The Ministry of
Agriculture does not have proven maize packages for its 645,000 small farmers (Marter,
1978; CIMMYT, 1978). Should Zambia press on with achieving self-sufficiency in maize
via commercial farmers or small farmers? Unlike the Sahelian Zone of West Africa,
where the long-term prospects of rainfed farming are not favorable, the long-term
prospects for helping small farmers are favorable in Zambia. But can Zambia survive
politically in the short-and intermediate-run while it relies on commercial farms? The
same question confronts Zimbabwe (see Bratton, 1981). How much does one sacrifice in
terms of efficiency in order to transfer resources to small farmers?-'

-The Stanford study showed that the social profitability of rice-production in Sierra
Leone was positive because of extremely low rural wage rates. But over the 1974-1981
period, rural wage rates have doubled from I Leone to 2 Leones per day and Sierra Leone's
rice imports increased to 41,000 tons in 1980. The question of Sierra Leone becoming a
rice exporter now appears to be mute.

2'But a word of caution is in order. The five countries studied only have a total
population of 26 million--or one-third of the total population of Nigeria. It is risky for
the authors to discuss West African rice trade without including Nigeria in their study.

-/For a discussion of the SADCC strategy to achieve food security in the southern
African region, see Kgarebe (1981).

The overriding policy issue of the postindependence period has been the ideology of
economic development--capitalism, state capitalism, or agrarian socialism. Regardless
of whether economies are organized along the lines of capitalism, state capitalism, or
socialism, in most countries the state will continue to intervene actively in the pricing
and marketing of agricultural commodities.
During the late 1960s, a consensus emerged that macro-planning and the transfor-
mation approach to agricultural development were not working. Therefore, most
governments turned to small farmer strategies of agricultural development and many
governments initiated integrated rural development projects. Because of lagging food
production in most countries during the 1970s, support for IRD began to wane in the late
1970s and numerous countries launched crash food production campaigns. The empirical
record thus far is not encouraging. The prospects of dramatically increasing food
production in the short run are dismal in almost all African countries in the face of long-
term neglect of the agricultural sector.
It is more evident than ever that long-run food and agricultural development
strategies must be built on a solid foundation of micro research on small farmers, traders,
and herders. We now turn to a survey of micro research on smallholder farming, technical
change, and livestock.


The purpose of this chapter is to identify some of the major institutions carrying out
micro-economic studies on smallholder farming, discuss methodological issues in conduct-
ing rural surveys, evaluate analytical techniques used by agricultural economists in
analyzing survey data, and present empirical findings for the 1960-81 period. A word of
caution is in order. We obviously cannot go into depth on research in all of the countries
included in this review. The publications cited are illustrative of research completed over
the past 20 years. These studies will help identify research gaps and in defining research
agendas for the 1980s.

Overview of Institutions Carrying Out Farm Level Studies
Prior to the 1960s, research by agricultural scientists in sub-Saharan Africa focused
on export crops and commercial farming. Little was known about the socioeconomic
aspects of subsistence farms except for studies by anthropologists, geographers, and a
handful of studies by agricultural economists.I/ In the 1960s, a number of research
institutes were established and there was a large increase in the research on the
economics of smallholder farming. This overview identifies some of the major institutions
carrying out farm level studies over the 1960-81 period.
Several African governments and universities initiated farm management and socio-
economic surveys in the 1960s. The Farm Economic Survey Unit (FESU) in Kenya began
conducting research on large estates in 1958 and undertook its first survey of smallholders
in the 1961-62 cropping season. For a review of the FESU farm survey program, see
MacArthur (1968). In Kenya, the Institute of Development Studies and the Department of
Agricultural Economics of the University of Nairobi have made a major contribution to
knowledge about Kenya's agricultural economy (Heyer, Maitha, and Senga, 1976); small
farmers (Heyer, 1971); extension and diffusion of innovations (Ascroft et al., 1972); rural
development (Heyer, Ireri and Moris, 1971); dairy industry (Hopcroft and Ruigu, 1976); and
fertilizer (Mwangi, 1978).
In Tanzania, M. P. Collinson (1962-64) carried out surveys of small farms in Sukumu-
land District in the 1960s. The Economic Research Bureau, the Department of
Agricultural Economics, and the Bureau of Resource Assessment and Land Use Planning
(BRALUP) of the University of Dar es Salaam have carried out a wide range of farm level
studies. For bibliographies on Tanzania, see Kocher and Fleisher (1979) and McHenry

A few important studies did provide information on the economics of agricultural
practices. See, for example, Haswell (1953); Galletti, Baldwin and Dina (1956); Leurquin
(1960); and Boutillier et al. (1962).

In Malawi, three sets of surveys were conducted from 1962 to 1965. One focused on
a few progressive farmers in each of 12 districts. Later a random survey of cotton
growers was carried out in three villages. The third approach relied on weekly visits to
farmers. Each enumerator was responsible for only eight farms (Catt, 1966). In the late
1960s, the Agricultural Economic Survey (AES) program was given primary responsibility
for farm level research in Malawi. See Farrington (1975b) for a review of the survey
approach used in the AES.
In Uganda, D. Belshaw, D. Pudsey, M. Hall, and 3. Cleave carried out a number of
surveys in the 1960s. For reviews and assessments of farm level research in East Africa
through the early 1970s, see Hall (1970), Collinson (1972), and Cleave (1974).
There has been a long history of research in Zambia (formerly northern Rhodesia)
dating from surveys initiated by the Department of Agriculture in the 1930s in
collaboration with the Rhodes-Livingstone Institute of Social Studies (see W. Allan, 1965;
A. Richards, 1932, 1939). The University of Zambia sponsors research through its
Institute of African Studies (see Colson, 1971; Quick, 1978) and the Rural Development
Studies Bureau (Honeybone and Marter, 1975; Marter, 1978). In Zimbabwe (formerly
southern Rhodesia), farm level studies were carried out by A. W. Johnson (Massell and
Johnson, 1968). The University of Zimbabwe's Department of Land Management is
currently carrying out a wide range of studies of irrigation, tobacco, and marketing (see
M. Blackie, 1981).
Turning to West Africa, the Department of Agricultural Economics at the University
of Ibadan in Nigeria carried out a large number of farm management and marketing
studies in the 1960s under the leadership of H. A. Oluwasanmi, Martin Upton, and
Q. B. O. Anthonio. Rufus Adegboye (1969, 1977) was one of the first African agricultural
economists to point out the strategic importance of research on land tenure issues. An
innovative socioeconomic survey was carried out in Uboma village in eastern Nigeria in
1963-64 and reported in Oluwasanmi et al. (1966). Ibadan researchers also carried out
numerous studies of marketing boards and supply response during the 1960s. Later, under
the leadership of S. O. Olayide, O. Ogunfowora, and F. Idachaba, the Ibadan group carried
out major studies of employment, income distribution, and agricultural policy in the
A village studies research program was initiated by David Norman and his colleagues
at Ahmadu Bello University (ABU) in northern Nigeria in 1964. The results of these
studies have made a major contribution to our understanding of intercropping, constraints
on small farm production and marketing systems, and consumption patterns (Norman,
1972; Goddard, 1972; Abalu, 1976; Simmons, 1976a,c). For a description of the survey
approach used, see Norman (1973) and Abalu and D'Silva (1980a).

Since 1962, the Institute of Statistical, Social, and Economic Research (ISSER) of
the University of Ghana, Legon, has carried out important studies on food production
(C. K. Brown, 1972), consumption (Dutta-Roy, 1969), and the cocoa industry (Kotey,
Okali, and Rourke, 1974). Also, some of the most valuable studies of demography in
Africa have been carried out by demographers at the University of Ghana (see Caldwell,
1969; Caldwell et al., 1975).
In Francophone West Africa, the Senegalese Institute of Agricultural Research
(ISRA) in Senegal and the Institute of Rural Economy (IER) in Mali have been active in
village studies. The emphasis of both programs has been on forging a link between
farmers and research stations. In the 1970s, ISRA established "experimental units" or
research zones in the groundnut basin in an effort to demonstrate the potential of
introducing new technology that had been developed on the main research station at
Bambey. For an overview of the experimental unit program in Senegal, see ISRA (1977);
Faye and Niang (1977); Benoit-Cattin (1980); and Fall (1980a). IER publications in Mali
include Institut d'Economie Rurale (1977) and Traore (1980).
Since the mid-1960s, researchers from numerous American and European universi-
ties have carried out farm level studies in collaboration with local researchers in
universities and in government research institutes. The Food Research Institute of
Stanford University, for example, carried out an inter-disciplinary study of agricultural
change in six English-speaking countries in sub-Saharan Africa in 1965/66 (Kenya, Uganda,
Tanzania, Zambia, Ghana, and Nigeria). Survey results were reported in Anthony and
Uchendu (1970, 1974) and Uchendu and Anthony (1975a, 1975b). A summary is reported in
Anthony et al. (1979).
In Southern Africa, the University of Nottingham collaborated with the University
of Zambia in conducting farm surveys in two areas in Zambia in the late 1960s. The focus
of the UNZALPI1 project was to identify approaches for increasing the labor productivity
of small farmers. Survey results were reported in Elliott et al. (1970). Tench (1975) also
used UNZALPI survey data.
The University of Reading sponsored farm level research in Botswana in the mid-
1960s, in Malawi in collaboration with the AES, and in Ghana with the University of
Legon. Thornton (1973) summarized the findings of the village development project in
southeast Ghana. Palmer-Jones (1974) reported the results of a study of tea production
and marketing in Malawi.
In Sierra Leone, a national farm survey was carried out by Njala University College
and Michigan State University in 1973-75. The results were presented in Spencer and

I/Universities of Nottingham and Zambia Agricultural Labor Productivity Investiga-

Byerlee (1976, 1977) and Byerlee et al. (1977, 1982). A survey of 480 rural households in
the Eastern Region of Upper Volta was carried out by a Michigan State University
research team over the 1978-80 period. The results are reported in Barrett et al. (1982),
Lassiter (1981), and Wilcock (1981).
Since 1976, Purdue University has been conducting surveys in West Africa to
evaluate: (1) the economics of sorghum and millet production in the Sahelian countries,
and (2) the costs and benefits of small- and medium-sized irrigation perimeters. Major
findings were presented in Purdue (1980).
The University of Michigan has carried out studies in West Africa, focusing on
livestock (Shapiro, 1979; Ariza-Nino and Steedman, 1979, 1980), grain marketing (Berg,
1980; Sherman, 1981), and mixed farming in Upper Volta (Delgado, 1979a, 1980) and Niger
(Eddy, 1979).
The University of Bordeaux began multi-disciplinary survey work in the Maradi
Region of Niger in 1977. De Miranda and Billaz (1980) and Raynaut (1980) discuss their
survey methodology and preliminary findings.
Farm level research in sub-Saharan Africa has also been assisted by a network of
international and regional research institutes. The West African Rice Development
Association (WARDA) with its headquarters in Monrovia, Liberia was established in 1960
to promote cooperation in the development of rice improvement programs in 15 French
and English-speaking countries. Currently farm level studies of rice production systems
are underway in Mali, the Gambia, and the Ivory Coast under the supervision of WARDA
agricultural economist--Dunstan Spencer.
One of the most important international organizations to support farm level
research in Africa since the early 1960s has been the Institute for Economic Research
(IFO) in Munich. In the 1960s, IFO focused on East Africa but by 1980 more than 100
monographs had been published in their "Afrika-Studien" series covering many of the
English-speaking African countries. Most of the studies deal with agriculture and nearly
half are in English. See, for example, Ruthenberg (1968), Gusten (1968), Kraut and
Cremer (1969), Roider (1971), Chambers and Moris (1973), and Lagemann (1977).
The Scandinavian Institute of African Studies at Uppsala has published a wide range
of studies on cooperatives (Widstrand, 1972); tobacco production (Boesen and Mokele,
1979); women (Bukh, 1979); multinational firms (Widstrand, 1975); and rural development
(Chambers, 1974).
The Office de la Recherche Scientifique et Technique Outre-Mer (ORSTOM) in Paris
has a network of around 20 research centers covering most of the francophone countries
in sub-Saharan Africa. Publications by ORSTOM researchers include Kohler (1971, 1972)

in Upper Volta, Ancey (1974) and Ancey, Michotte, and Chevassu (1974) in the Ivory
Coast, and Copans et al. (1972), Delpechi and Gastellu (1974), and Rocheteau (1975) in
Senegal. See Couty and Hallaire (1980) for an overview of ORSTOM studies during the
1960-80 period.
The International Institute of Tropical Africa (IITA), established in 1968 at Ibadan,
Nigeria, has carried out technical and socioeconomic research on the major food crops in
the humid tropics. Studies by IITA researchers include Robinson (1974); Flinn, Jellema,
and Robinson (1975); Lagemann, Flinn, and Ruthenberg (1976); Lagemann (1977); Okigbo
and Greenland (1977); Bachmann and Winch (1979); Fotzo and Winch (1978); Diehl and
Winch (1979); Flinn and Lagemann (1980); Menz (1980); and Zuckerman (1977, 1979a,b,c).
In 1976, CIMMYT launched a program in Eastern and Southern Africa under the
leadership of Michael Collinson to help improve national research systems through
training and farming systems research on the problems of small farms. The farming
systems studies emphasize the rapid collection and processing of data. CIMMYT's goal
has been to publish results in 6 to 12 months in order to channel information on the
problems of farmers into the design of on-station research programs and to develop on-
farm experiments in order to test promising technologies under farm conditions. Although
the research focus is on the economics of wheat and maize, the crops under CIMMYT's
mandate, the farm level studies provide valuable information on other crops such as
sorghum, millet, cassava, cotton, and cowpeas. CIMMYT's cooperative studies in Kenya,
Tanzania, and Zambia are reported in CIMMYT (1977a, 1977b, 1978) and Collinson (1982).
ICRISAT started a cooperative research program in Upper Volta in 1975 with a
sorghum breeder followed by a plant pathologist, millet breeder, two agronomists, a striga
specialist, and an entomologist. A production economist was added in 1979 and farm level
research was started in 1980.-/
In addition to the above research programs and projects, research findings based on
farm level studies during the 1960-81 period are reported by the following researchers:
East Africa: Heyer (1966, 1971, 1972a); Catt (1970); Hutton (1973); Richards, Stur-
rock, and Fortt (1973); Shapiro (1973); Saylor (1974); Vail (1975); Humphrey (1975);
Hunt (1975a); Gerhart (1975).
Central and Southern Africa: Massell and 3ohnson (1968); Atayi and Knipscheer
(1980); Tollens (1975); Weinrich (1975); Perrault (1978); Marter (1978); Kinsey (1978).

1/ICRISAT has sponsored valuable literature reviews on production systems
(Norman, Newman, and Ouedraogo, 1981) and marketing (Harriss, 1979a,b) in the semi-
arid tropics of West Africa and a conference on socioeconomic constraints on develop-
ment of semi-arid tropical agriculture with reference to Africa and India (ICRISAT,

West Africa: Meillassoux (1964); Welsch (1965); Pelissier (1966); Luning (1967);
Upton (1967); Hill (1968); Smock and Smock (1972); Capron (1973); Monnier et al.
(1974); Maymard (1974); Berry (1975); Kleene (1976); Winch (1976); Marchal (1977);
Matlon (1979); Faulkingham (1977); Reyna (1977); Sawadogo (1977); Zuckerman
(1977); Venema (1978).

Methodological Issues in Rural Surveys
Issues in Survey Design
Since most small farmers are illiterate and do not keep farm account books,- three
methods have been used to generate information: (1) case studies, (2) infrequent surveys,
and (3) cost route or multiple visit surveys (Spencer, 1972). The case study or model farm
approach provides descriptive information on a single farm or a number of farms
purposively selected to be representative or to reflect the practices of progressive
farmers. Infrequent visit surveys- entail visiting a farm once or a few times to collect a
range of stock (inventory) data and information about current practices. In the cost route
(or multiple visit) approach,-3 farmers are visited regularly by an enumerator over an
entire cropping season or full year, generally one to three times weekly and from 50 to
150 times a year. The rationale for using the cost route approach is that it is an effective
way to capture flow (input/output) data on the magnitude and variability of labor--the
most important input on small farms.
During the 1960s, researchers in East Africa used all three approaches--case
studies, infrequent surveys, and cost route surveys--to collect farm-level information
(Hall, 1970). Clayton (1963), for example, used a model farm approach, drawing on data
from government farms. Researchers in the Farm Economic Survey Unit in Kenya used a
model farm approach in their whole-farm studies, based on data drawn from interviews
with progressive farms (MacArthur, 1968). Heyer (1966) used a case study approach,
relying on intensive observations on a small sample of farmers in Kenya. Researchers in
Tanzania used farm business type surveys on a large number of randomly selected farmers
(Collinson, 1962-64). In Uganda, researchers from Makerere University and the Ministry

'There have been occasional attempts to use literate children to keep rudimentary
records (MacArthur, 1968) but this approach has largely been abandoned in Africa.

!/There are numerous terms such as reconnaissance, exploratory, informal, and farm
business surveys for what are essentially infrequent visit type of surveys. The farm
business survey terminology is a western concept which was used in some African
countries in the 1960s but the term was subsequently dropped.

-Cost route derives its name from the repeated nature of the survey over the
course of a year in order to derive data to compute costs and returns of production.

of Agriculture used a cost route approach, interviewing randomly selected farmers three
times weekly for the entire crop year (Pudsey, 1967).
During the late 1960s and early 1970s, the case study approach was largely
abandoned by agricultural economists in English-speaking countries and researchers
shifted to surveys and random sampling to ensure that input/output data reflected typical
farm-level conditions. Although some researchers such as Collinson (1972) continued to
advocate infrequent visit surveys, the prevailing opinion was that the cost route approach
should be used if there were time, money, and administrative capacity (Kearl, 1976). The
rationale for cost route surveys can be traced to the 1960s when labor was identified as a
major constraint on smallholder production in Africa. Advocates of the cost route
approach such as Dunstan Spencer (1972) were aware of difficulties that one-shot surveys
had in coming to grips with the seasonality of labor use and argued that frequent visits by
enumerators were necessary to capture the seasonality of labor use and to control
measurement errors on the labor variable. The cost route method was widely used in farm
management surveys during the 1970s (Norman, 1972; Matlon, 1979; Winch, 1976; Fotzo
and Winch, 1978; Shapiro, 1973; Elliott et al., 1970; Spencer and Byerlee, 1976; Zucker-
man, 1979a; Purdue University, 1980). In order to reduce the cost of multiple visit
surveys, multi-stage sampling procedures were employed by numerous researchers. In this
approach, stock data are collected in a reconnaissance survey of a large population, often
several hundred households, and in the second stage, a smaller sample is randomly
selected for repeated surveying in order to collect flow data over a period which often
covers 12 months (Norman, Newman, and Ouedraogo, 1981).
Cost route or multiple visit surveys have thus far provided the most reliable data on
input flows, particularly labor inputs, but this type of survey is substantially more costly
per farm interview than one-shot surveys. As a result, there is a trade-off between
sample size and visiting frequency. While it has been recognized that one-shot surveys
will likely have high measurement errors for variables such as labor, the cost route or
frequent interview methodology has a number of inherent problems. For example, the
cost of interviewing the same farmer 50 to 150 times a year is extremely high, and there
is a proble-n of sustaining the interest of the farmer during repeated interviews.
Moreover, it often requires 6-12 months to plan a cost route study, a year to carry it out,
and sometimes 2-3 years to analyze and publish the results. Concern with the cost of cost
route surveys and the need to generate rapid results has led to a search for survey
methodologies which can produce results in a few months rather than 2-3 years.
Starting in the late seventies, there was a discernible shift from cost route to
infrequent visit surveys. These infrequent visit surveys, now more commonly called

exploratory or informal surveys, are popular among some advocates of farming systems
research such as Byerlee, Collinson et al. (1980) and Collinspn (1981, 1982). Although
there is clearly a need to expand the use of infrequent visit or informal surveys for more
rapid identification of major constraints on production, no one has yet identified a method
to collect reliable information on seasonal labor use through one-shot surveys. Because
labor is the most important input on small farms, designers of improved technical
packages need data on labor use by crop, by activity (e.g., weeding, ridging), and by week
during a few critical periods in the year.
Two approaches have emerged to deal with the conflicting goals of quick turn-
around time and reliable labor data. One approach is to incorporate the best features of
infrequent visits and cost route surveys into a new approach such as the activity approach
which was tried by Pascal Fotzo in his 1980-81 study of rice production in eastern Upper
Volta. Fotzo recorded data once on each activity (e.g., land preparation, planting, and
weeding) involved in the production of bas fond rice over a six-month period.- A second
approach is to prepare case studies for a small sample of farmers--say 30--from only one
or two villages (Matlon, 1980). The recent emphasis, by some farming systems research
teams, on combining survey research with anthropological type case studies has been
stimulated by increasing acceptance of the view that an understanding of farming systems
requires more than the input of economists and more than the computation of averages
from a farm survey. But whatever method is used to collect data a basic problem is one
of collecting far too much information. Farrington (1975b) and Abalu (1980) point out
that the frequent lag between data collection and publication of results is often caused by
the collection of far too much information.
The approach to farm-level studies has been significantly different in francophone
than in English-speaking countries. In Francophone Africa, farm-level researchers have
generally carried out case studies on a small number of farms rather than large-scale
rural surveys. Although we classify the approach often used by French researchers and
French-trained researchers as a case study approach, we do not mean that statistical
surveys are never used. For example, Boutillier et al. (1962) used a one-shot questionnaire
in which farmers were asked to recall information for the preceding year; they also
collected daily information for a seven-day period twice during the year on such variables
as expenditures and food consumption. But where formal attempts have been made to

-While the activity approach is promising, it may not reduce survey costs unless
there is detailed information available on the cropping calendar since farmers have to be
interviewed to see if an activity is completed. But data processing costs are reduced and
the researcher can benefit from insights gained through informal interviewing even though
the informal interviews are not formally recorded (Fotzo, personal communication,
Department of Agricultural Economics, Michigan State University).

quantify key variables, sample sizes are generally extremely small and households are
often selected on a non-random basis with the limited goal of deriving numbers for
descriptive purposes, as opposed to sampling for statistical inference.- Copans et al.
(1972) and Delpechi and Gastellu (1974), for example, report surveys in which the samples
were 10 to 12 adults representing 2 to 4 compounds.
Methodological debates in francophone countries have focused on alternative
conceptual frameworks for understanding social structure and social relationships and how
they affect the organization of production. One of the major issues debated during the
seventies was whether studying the production unit (exploitation agricole) is sufficient to
explain the dynamic forces affecting the decisions of small farmers or whether studies at
different levels--clans, villages, age groups, etc.--can provide more reliable information.
The term exploitation agricole refers to a production unit where people work together on
the main food grain field and eat as a group from the output of the main field (Ancey,
1975). While English-speaking researchers undertaking survey research have generally
relied on the household as the unit of analysis (often defined as "people who eat from the
same pot") researchers in francophone countries consider the question of the proper
decision unit to investigate to be an important unresolved issue (Monnier et al., 1974;
Maymard, 1974; Ancey, 1975; Kleene, 1976; Couty, 1979; Gastellu, 1980). For example,
Ancey (1975) argued that a single sampling unit such as exploitation agricole was not
broad enough to capture the multi-dimensional relationships affecting decision making on
African farms. He contended that different groups of decision makers operate at
different levels--individual fields, consumption groups, main field production, residence,
lineage, village, and supra-village--and that decisions of these various units often overlap
and compete with others. Based on survey experience in Senegal, Gastellu (1980) argued
that the starting place in survey research is to identify the units of production,
consumption, and accumulation, rather than relying on a single concept such as exploita-
tion agricole. He proposes that in order to identify these groups, researchers should
(1) evaluate who makes the main decisions in each activity, (2) evaluate the syntax of the
local language in order to discern obligations and expectations between people, and
(3) focus on identifying patterns of privileged exchanges-- food, gift, and labor, as well as

!/French researchers have often expressed the view that sample surveys are a tool
of statisticians which may be selectively used to supplement the qualitative understanding
of farmers provided by social science researchers using techniques such as participant
observation, recording life histories, and constructing geneologies (Couty, 1979; Benoit-
Cattin, 1980).

The value of using the individual rather than the entire production unit as a unit of
analysis has been stressed by Winter (1975), Rocheteau (1975), Kleene (1976), Couty
(1979), and Raynaut (1980). Raynaut argued, for example, that by taking into account
differences between individuals, researchers can better distinguish the roles that physical
and technical factors play in socioeconomic strategies of different household members.

References on Survey Design
In presenting results, researchers generally have devoted little space to justifying
the approaches they followed in collecting and analyzing survey data. But the choice of
data collection and analysis procedures may importantly influence survey results. For
example, the decision to use open-ended questionnaires as opposed to structured instru-
ments can exert a major influence on the results obtained. Open-ended questions allow
farmers to identify problems in their own words but tend to introduce intractable
problems in the analysis and interpretation of data. On the other hand, more structured
schedules may reduce ambiguity in interpretation of data but the choice of wording will
frequently bias the replies.-t- A large body of wisdom about designing rural surveys is now
available. A review of these reports should be the starting place for researchers
contemplating farm-level research.
Collinson (1972) reviewed methodological problems in collecting and analyzing farm
management data for planning purposes, drawing primarily on his early experience in
Tanzania. Spencer (1972) drew on his experience in conducting farm management and
marketing studies in Sierra Leone to discuss methodological problems in collecting flow
data. Norman (1973) reviewed his experience in directing farm-level surveys in northern
Nigeria. A volume edited by Kearl (1976) contains a valuable discussion of field data
collection experiences by numerous African researchers. Farrington (1975b) reviewed the
data collection methods used by the Agro-Economic Survey Unit in Malawi with emphasis
on labor flow information. More recently, Zuckerman (1979a, 1979b) has discussed the
approach used by IITA researchers in western Nigeria in the early 1970s. Abalu (1980)
critiqued cost route surveys in West Africa. For a discussion of infrequent (informal)
surveys, see Byerlee, Collinson et al. (1980) and Collinson (1982).

-Additional survey design issues which may influence survey results include:
(1) selection of the sampling frame, (2) procedures used for gaining knowledge of local
farming practices in order to design questionnaires, (3) approaches for securing support
and cooperation of interviewees, (4) choice of direct measurement techniques--primarily
for field size, yields, and intensity of labor use--to supplement recall information,
(5) alternative methods for gathering information about sensitive issues such as the size of
land holdings or livestock, buildings, and credit, and (6) methods for making field data
checks to reduce inconsistency and to verify recorded responses.

A discussion of methodological issues in rural surveys in several francophone
countries can be found in AMIRA's "notes de travail."-1 Ancey (1975) presented a
valuable picture of the complex organization of small farms. Winter (1975) and Thenevin
(1978) identified different survey techniques in gaining information for economic planning.
Couty (1979) presented a general overview of methodological issues encountered using
socio-anthropological approaches to farm-level research. Raynaut (1980) and De Miranda
and Billaz (1980) discussed multi-disciplinary surveys of farms in Maradi, Niger. Building
on his research experience in Senegal and the Ivory Coast, Benoit-Cattin (1980) outlined a
non-statistical survey approach to study how economic, technical, social, and ecological
variables interact to affect the organization of work on farms.

Data Processing
Processing of survey data has posed a major problem for researchers throughout
Africa. There has been a tendency to collect a wide range of data, paying little attention
to how the data is to be analyzed until after data collection is finished (Abalu, 1980). As
a result, portions of the collected data are frequently never even keypunched, let alone
analyzed. Moreover, the effort required to validate and aggregate data into files for
analysis has often led to delays of one to three years before preliminary results are
published. Researchers are slowly starting to realize that data processing must be
considered as an integral part of the entire survey design, data collection, and data
analysis chain of events. Data processing can take as many resources as data collection
but most manuals on survey methodology pay little attention to processing problems. In
addition, even though many African countries now have computer facilities, most
computer centers have limited capacity and access to software packages.
Several major decisions have to be made at or before the beginning of data cleaning
and validation. Often, little attention is given to the following two critical issues:
(1) how to stratify sample households into appropriate groups for subsequent analysis, and
(2) how to convert labor into a homogenous unit in order to make labor files more
manageable (Norman, 1972). Several approaches have been used for stratification. In
many cases, particular regions or villages have been purposively selected at the data
collection phase, reducing the need for ex post stratification. Clayton (1964) and

/AMIRA is an informal working group of researchers from ORSTOM, INSEE, and
the French Ministry of Cooperation who have had extensive farm-level research experi-
ence in francophone Africa. The group was formed in 1975 with the specific purpose of
debating and refining data collection methods. Four main issues are addressed in the
AMIRA papers: (1) what information to collect, for what uses, and for what objectives;
(2) how to collect the relevant information; (3) how to process and analyze the data; and
(4) how to use the information to improve decision making. (See Winter, 1978.)

Collinson (1972) advocated stratification by land per resident because capital use by
traditional farmers is limited. Researchers at Ahmadu Bello University stratified their
surveys in northern Nigeria: (1) by village or section, and (2) by land per resident. The
IITA 1970-71 survey in western Nigeria used age of the farmer to stratify their surveys
(Zuckerman, 1979a). Researchers have often relied on land tenure patterns, ethnic
groups, or agro-ecological zones to stratify samples of farms. For example, agro-climatic
zones were the main means of stratifying survey data by Barrett et al. (1982) in Upper
Volta and in Sierra Leone by Spencer and Byerlee (1976). Ancey (1977) classified 25 farm
models in the Sahelian zone of West Africa mainly on the basis of farm size, as well as
ecological characteristics. Poulain et al. (1979) stressed the use of ecologically
homogenous zones, and within each zone distinguishing among households using land
availability and the presence or absence of livestock.
There has been a clear dichotomy between francophone and anglophone countries in
the treatment of labor (Norman, Newman, and Ouedraogo, 1981). Studies in francophone
countries generally rely on a stock measure of labor called the actif, primarily because of
their heavy emphasis on case studies. An actif is often defined as a person between ages
15 and 54 or 59. The actif concept tends to understate the stock of labor available for
agricultural activities, particularly in peak seasons when most persons over 10 years work
in fields. If the proportion of labor coming from individuals under 15 and over 54 is
significantly different among sub-samples of households, the net return per actif will be
overestimated in the households where non-actifs make a major contribution to farming
activities. Also the actif concept does not make any allowance for differences in
productivity and labor use by seasons (Ancey, 1974). Examples of studies which have used
the actif approach include Kohler (1971) and Monnier et al. (1974) in Senegal; and Barrett
et al. (1982) in Upper Volta.
In studies in Anglophone Africa, researchers have generally aggregated labor units
using the concept of "man-equivalents." The concept of man-equivalents was applied to
stocks of labor in early research in Tanzania in order to discount the inputs of women and
children since it was assumed their labor was less productive than that of men (Collinson,
1962-65; Ruthenberg, 1968). More commonly, indices reflecting productivity difference
have been used to convert flow inputs by women and children into man-hours or man-days.
Most studies have used fairly arbitrary approaches for deciding on the weights to be used
in aggregating labor inputs and, as Collinson (1972) noted, there has been controversy over
how to derive appropriate weights. Norman (1972) contended that a work study approach
is an objective way to assign weights. Spencer and Byerlee (1976) used wages in rural
labor markets to derive relative weights. In a 1978-79 survey in eastern Upper Volta,

weights were obtained by asking a sample of farmers to estimate the relative productivity
of different classes of labor defined by age and sex for each major cropping activity
(Barrett et al., 1981).
The need to speed up the processing of survey data has emerged as one of the
dominant factors affecting survey design and data processing strategies. Quicker turn-
around times necessitate better coordinated data collection and processing. Norman and
Palmer-3ones (1977), for example, call for a more standardized economic methodology for
research on cropping systems. The FAO has developed a terminology for farm manage-
ment concepts in French, English, and Spanish and has also developed pre-coded
questionnaires and a standardized computer program (Friedrich, 1977). This program can
be used in different ecological zones and countries to generate partial and whole farm
budgets, crop and livestock enterprise tables on a farm-by-farm basis, as well as the usual
sample averages.-
Micro-computers, costing $4,000 to $15,000, have great promise for data cleaning,
validation, and preliminary analysis of survey data. Purdue University used micros in
their West African research in Mali, Senegal, and Upper Volta. Morris reports (in Purdue,
1980) that local programming skills are a bigger constraint than hardware problems. Mini-
computers (larger than micros and costing around $100,000) have been used in the World
Bank financed evaluation studies in northern Nigeria. While small computers will
undoubtedly play a major role in future research in Africa, much more experience with
both hardware and development of software packages will be needed before they will
match the claims of the manufacturers. In any event, the use of small computers cannot
solve the turn-around problem unless researchers stop collecting excess data (Candler and
Slade, 1981).

Applications of Analytical Techniques
Farm budgeting is a standard farm management technique which was introduced in a
large number of countries in English-speaking Africa in the 1960s.'2 Whole (total) farm
budgets were used to derive performance standards for extension programs (MacArthur,
1968). During the 1970s, researchers shifted to enterprise budgeting except in areas
where there was extensive intercropping. Enterprise budgeting was used to (1) compare

I/The initial FAO farm management program (FMDCAS) has been updated with a
second version (FARMAP). The new version is based on a modular design in order to be
more flexible.

2Ken a (MacArthur, 1968); Tanzania (Collinson, 1962-64; Ruthenberg, 1968);
Uganda Pudsey, 1967); Malawi (Catt, 1966); Nigeria (Upton and Petu, 1964; Upton, 1967);
and Senegal (Boutillier et al., 1962).

the costs and returns of different crop and non-farm enterprises (Lagemann, 1977;
Ruthenberg, 1980; Spencer, Byerlee, and Franzel, 1979; Lassiter, 1981); and (2) to
compare costs and returns in producing the same crop with different techniques (Winch,
1976; Spencer and Byerlee, 1976; Lang, 1979). A few researchers have used partial
budgeting to evaluate changes in enterprise mix or production techniques (Fotzo and
Winch, 1978). We shall concentrate on whole farm and enterprise budgeting.
Numerous methodological issues need to be addressed in constructing budgets from
cross-sectional data (Dillon and Hardaker, 1980). Four major problems have repeatedly
surfaced in deriving budgets, making it difficult to interpret and compare the results of
different studies. First, no standard approach has been used in deciding what to include in
farm budgets. For example, many researchers have only included yields of the main
crops, arbitrarily excluding secondary crops, livestock, and non-farm activities. Few
researchers have taken into account changes in stocks or have attempted to estimate the
opportunity costs of farm supplied inputs such as manure. Furthermore, there has been
little consistency in the treatment of land; some researchers included all land under the
control of the farm, other studies used only land owned or land cultivated in the year of
the survey.
Another problem in budgeting studies has been the valuation of inputs and outputs.
Because small farmers generally purchase few inputs and retain a major proportion of
farm output for family consumption, returns are highly dependent on the value research-
ers assign to inputs and outputs. Both measuring and valuing family labor pose major
problems in constructing farm and enterprise budgets. A few researchers have presented
results using a range of assumptions such as not including the cost of any family labor
during different times of the year, including the cost of only hired labor and inputing a
cost for all labor (Norman, Pryor, and Gibbs, 1979). More commonly, researchers have
treated hired labor as a variable cost, included a charge for land, and then computed a net
return to labor and management as a residual (Ruthenberg, 1980; Spencer, Byerlee, and
Franzel, 1979). Even this approach is fraught with difficulties. How does one value labor
in situations where farmers rely on reciprocal labor exchanges, providing meals for guest
workers as partial payment? Another issue is whether to include time walking to and
from fields. Most studies have ignored walking time even though in some cases it can be
substantial (Cleave, 1974).
The problem of valuing capital services is equally difficult. The most common
practice has been to exclude costs of capital for hand implements. Straight line
depreciation is often used for major capital items. Spencer and Byerlee (1976) and
Spencer et al. (1979) used a capital recovery formula to convert capital stocks to annual
service flows taking into account the average life of the equipment and a discount rate
assumed to approximate the social opportunity costs of capital.

The value of outputs has usually been derived by multiplying average yields over the
sample of farms by the average annual price in local markets. Use of a simple average of
observed prices throughout the year can substantially misrepresent the actual average
prices received by farmers because prices can fluctuate by more than 100 percent during
the year. Moreover, sales are often concentrated in the post-harvest period when prices
are lowest. Another common problem in valuing outputs is whether to attribute the value
of intercropped fields to individual crops or to treat the crop combination as a single
A third problem is interpreting budgets that have been constructed from average
input/output relationships on surveyed farms. Averages, as Upton and Casey (1974) and
Dillon and Hardaker (1980) point out, can obscure wide variations in soil characteristics,
managerial ability, class differences, access to inputs, and other variables which can
greatly influence the returns to any given farmer. Moreover, as Ruthenberg (1980) has
warned, budgets built from averages of survey data generally show a more diversified
production pattern than one is likely to observe on any individual farm. Ruthenberg
contended that data from a model or typical farm are often more useful than survey data
from farm budgets. G. E. Dalton (1973) points out that while net income figures derived
from farm budgets provide useful descriptive information, they do not provide insights
into the factors responsible for poor performance of a single enterprise or the reasons for
low productivity on an individual farm.
Finally, budgets based on cross-sectional data do not take account of changes over
time and space. Because of inflation, overvalued exchange rates and changes in relative
prices of crops over time, figures such as the cost of plowing with tractors or the returns
to labor in rice farming are generally not comparable in different countries and in
different years. Also, net income figures do not provide adequate information on the
welfare of rural households unless data are also available on the cost of living. Because of
periodic droughts and wide variations in the timing of rains, weather can greatly influence
the returns to land and labor in any given year. Finally, land clearing activities and time
spent planting perennial crops add value to production in future years but trying to
identify the returns to that labor poses an accounting nightmare.
Despite our stress on data limitations and problems of valuation, farm budgeting is a
valuable tool of analysis. Budgeting and comparative farm indices were found to be
invaluable in farm survey research in the 1960s in large part because they could be
constructed relatively quickly and the training requirements for building and interpreting
budgets is limited relative to programming and regression methods. Ruthenberg's (1980)
invaluable farming systems book attests to the descriptive value of comparative farm

budgets. Enterprise budgets have contributed to our understanding of relative costs and
returns in fallow and permanent cultivation systems, from improved versus traditional
production practices, from perennial and annual crops, and for farming and non-farm

Regression Analysis
Econometric studies of smallholder production relationships can broadly be divided
into three categories: (1) farm production functions; (2) analyses of the determinants of
levels of resource use; and (3) supply response studies.

Farm Production Functions
Production function studies in the 1960s were few and far between and of uneven
quality. Interest in this topic was stimulated by the desire to evaluate smallholder
productivity and efficiency in order to test Schultz's "poor but efficient" hypothesis
(Schultz, 1964). Production functions estimated from controlled survey data and data
from experiment stations have also been used to analyze factors influencing yields of
recommended technical packages and to identify optimal levels of resource use for new
To our knowledge, the earliest reported attempts to estimate production functions
for smallholders from survey data appears to be in the case studies of Tanzanian
agriculture reported in Ruthenberg (1968), Massell and 3ohnson's (1968) analysis of African
farmers in Zimbabwe, Upton's (1967) research in southwestern Nigeria, Welsch's (1965)
study of rice farmers in eastern Nigeria, and Luning's (1967) study of sorghum-millet-
groundnut farmers in northern Nigeria. During the 1970s, efforts to evaluate determi-
nants of farm household income using regression techniques have been carried out in
numerous countries, including Zambia (Elliott et al., 1970; Tench, 1975), Nigeria (Upton,
1970; Norman, 1972; Norman, Fine et al., 1976c; Matlon, 1979; Osuntogun, 1978; Norman,
Pryor, Gibbs, 1979; Mijindadi, 1980), Kenya (Saylor, 1974; Wolgin, 1975; Rukandema, 1978;
Moock, 1981), Tanzania (Shapiro, 1973; Shapiro and Muller, 1977), Sierra Leone (Spencer
and Byerlee, 1976), the Ivory Coast (Lang, 1979), Malawi (Farrington, 1975b, 1977a), and
Ghana (Prakah-Asante, 1976).
After almost two decades of experience in estimating smallholder farm production
functions using survey data, there are still many unresolved problems. Since production
function studies have relied on survey data of uneven quality, overcoming data limitations
has been one of the major determinants of the approaches followed in specifying
production functions. Major specification problems have included: (1) choice of a
functional form and (2) measurement and aggregation of inputs and outputs. The choice
of functional form has received surprisingly little attention in discussions about the

specification of production functions. Most researchers have used Cobb-Douglas func-
tions, generally with little or no discussion of why alternative functional forms were not
considered. A few researchers have presented comparative results using linear, quadratic,
and square root functional forms (Ruthenberg, 1968; Luning, 1967; Tench, 1975; Shapiro,
1973). Spencer and Byerlee (1976) used a CES production function to evaluate returns to
scale and the elasticity of substitution between labor and capital services in rice
production in Sierra Leone. In a series of studies conducted by researchers at the
Institute for Agricultural Research, Zaria, Nigeria, several functional forms were
used: linear for yield functions, quadratic or linear for estimating the relationship
between profits and yields, and Cobb-Douglas to evaluate the marginal value products of
inputs (e.g., Norman, Beeden et al., 1976a, 1976b).
Almost without exception, farm production functions have been estimated using
ordinary least squares. Three exceptions include Massell's use of analysis of covariance
(Massell and 3ohnson, 1968; Massell, 1967b), Spencer and Byerlee's use of non-linear
maximum likelihood estimation of their CES function, and Strauss' (1981) household firm
model using survey data from Sierra Leone. The production component of Strauss'
simultaneous equation model is Cobb-Douglas in inputs and constant elasticity of transfor-
mation in outputs. The model was estimated using a tobit model in order to account for
some products not being produced by all households.
In general, the dependent variable in farm production functions has been total farm
income from all cropping activities.- Limitations of survey data have largely prevented
estimation of crop-specific production functions. Yields from separate fields or from
individual crops in crop mixtures have typically been converted to value terms using an
average of annual prices in local markets. In a few cases, returns per acre instead of
total farm income has served as the dependent variable. In the few cases where
production functions have been estimated for specific crops, one or a few crops either
accounted for most of the area cultivated or were important sources of cash income
(Massell and Johnson, 1968; Spencer and Byerlee, 1977; Saylor, 1974). Yields of specific
crops have also served as the dependent variable in studies using experiment station data
or data from controlled surveys (Norman, Beeden et al., 1976a, 1976b; Norman, Hayward,
Hallam, 1974, 1975; Flinn and Lagemann, 1980).

/We refer to the research cited in this section as production function studies even
though many studies are not based on physical production functions of particular crops
since this terminology has been consistently used in the literature in Africa.

Data problems have forced researchers to rely on inadequate representation of
aggregate input categories--land, labor, capital---in estimating inter-farm production
functions. In some early studies, stock variables were used to represent flow inputs in
cases where flow data were not available. Ruthenberg (1968), for example, used family
man-equivalents as a proxy for labor. Massell and Johnson used the value of farm
implements at undepreciated replacement cost to represent capital. Upton (1967)
represented capital investments by acres of tree crops and number of livestock. Even in
cases where flow variables have been included, it has been common to use undiscriminat-
ing units such as man-equivalent days (Upton, 1967; Lang, 1979).2
Because of specification and measurement problems, the coefficient of multiple
determination of most farm income studies based on aggregate physical farm inputs has
been low and the standard deviations of individual coefficients have been large.
Researchers have followed two major approaches in trying to increase the amount of
income which is explained by their models. First, several researchers have abandoned
efforts to rely exclusively on variables representing underlying physical production
functions and have included indices of social and personal characteristics as well as
physical input variables (e.g., Upton, 1967, 1970; Elliott et al, 1970; Shapiro, 1973; Saylor,
1974; Tench, 1975; Rukandema, 1978). Upton (1967) was one of the first researchers to
analyze the relationships between production, consumption, and social characteristics.
Upton used Guttman scaling to rank farmers by factors such as progressiveness and
correlation analysis was then used to evaluate the strength of relationships between social
characteristics and production.
One of the first studies to formally incorporate social variables into functions of the
value of farm output was based on farm survey data collected in Zambia under the
UNZALPI Project. Elliott et al. (1970) incorporated indices to represent factors such as
education, general awareness, and knowledge of farming and attempted to attribute
proportions of the explained production per acre to social variables instead of the usual
farm inputs. Tench (1975) presents a valuable discussion of conceptual and statistical
issues involved in evaluating smallholder productivity in Zambia. In addition to estimating
a basic Cobb-Douglas model relying on aggregate input categories for land, labor, and

-/In several cases, capital has been left out because (1) there was little variation in
the use of capital across farmers, (2) little capital was used by farmers except a hand hoe,
or (3) farmers who used fertilizers or insecticides used them improperly (Ruthenberg,
1968; Shapiro, 1973; Lang, 1979).

-'There is substantial evidence that both the length and intensity of the working day
vary significantly during the year (Norman, 1972; Cleave, 1974; Farrington, 1975a;
Shapiro, 1978).

capital, Tench explored alternative formulations, including (a) inclusion of several indices
of social and personal factors and (b) regressing the value of output on a set of variables
representing only social and personal factors. Surprisingly, Tench found that the multiple
coefficient of determination based on social and personal indices was nearly as large as
the coefficient based on the usual physical inputs. Tench (1975) also used the technique of
principal components analysis to identify a set of social, personal, and technical variables
which were most closely associated with variation in the value of farm output. Using
survey data from Tanzania, Saylor (1974) used stepwise regression to analyze the
relationship between output per hectare and per unit of labor and several technical
variables such as the time of planting, plant densities, and ridging as well as social
variables such as modernity and contact with extension workers.
Researchers have also attempted to improve the specification of their models by
disaggregating land, labor, and capital into more specific components. Upton (1967) and
Norman, Pryor, Gibbs (1979) decomposed labor into family labor and hired labor. More
commonly, labor has been decomposed by either age-sex category (Shapiro, 1973, 1978;
Tench, 1975) or by cropping operation (Shapiro, 1973; Saylor, 1974; Rukandema, 1978).
Specification of land has been improved by incorporating dummy variables for soil quality
and fertilizer use. Capital has been divided into cash expenditures and depreciation on
capital equipment (Norman, 1972).
One of the major methodological issues in specifying farm production functions has
been the decision as to whether to include a variable for management. Attempts have
been made by Massell and Johnson (1968), Upton (1970), and Shapiro (1973) to derive
proxies for management in order to reduce the specification bias that arises when
management is correlated with included variables. Massell and Johnson used survey data
from two areas in Zimbabwe, the Cheweshe Reserve and Mt. Darwin District, to fit Cobb-
Douglas production functions for each of the three major crops--groundnuts, millet, and
corn. Farmers were interviewed weekly for the entire crop year. Massell and Johnson
attempted to reduce management bias by use of analysis of covariance and by incorporat-
ing dummy variables to distinguish three groups of small farms by management level. A
discussion of the procedures used in the analysis of farms in the Cheweshe Reserve and
major findings was also presented in Massell (1967a). Massell (1967a) represented
management differences using dummy variables; he later drew on data from the Mt.
Darwin District to illustrate his method for reducing management bias through the use of
analysis of covariance (Massell, 1967b). Massell treated each firm-product combination as
a separate observation, making the assumption that the relative efficiency of farmers is
constant across all the crops they grow.

Upton (1970) used the technique of principal components analysis to derive a single
index as a proxy for management. Shapiro (1973) derived several indices for social and
personal characteristics using Guttman scaling and then combined the indices into a single
index whereby farmers were ranked according to their degree of "modernization."
Because there was a high correlation between the modernization ranking and a farmer's
ranking in terms of technical efficiency, Shapiro argued that the modernization index
could be used as a proxy for management efficiency. In estimating production functions
for his sample of farmers in Geita District, Tanzania, Shapiro experimented with several
specifications, both including and excluding the management variable. Shapiro concluded
that the effect of including management was to raise the productivity of labor and to
decrease the productivity of land. Shapiro and Muller (1977) expanded on Shapiro's
analysis by incorporating indices of information in their analysis of farmer efficiency.
Although the above research by Massell, Upton, and Shapiro represented important
methodological contributions, researchers have largely abandoned the use of a manage-
ment variable and have tried to incorporate management indirectly through variables such
as the timing of key activities (e.g., date of planting, date of weeding) and by specifying
variables reflecting the quality of inputs to account for differences in "technical
efficiency" among farmers. In production function studies for maize, cotton, and
sorghum, sowing date, plant density, amount of fertilizer, and time from sowing to
weeding have been found to have significant relationships with yields (Norman, Beeden
et al., 1976a, 1976b; Norman, Hayward, Hallam, 1974, 1975). In their recent research on
maize production in Nigeria, Flinn and Lagemann (1980) found that coefficients of
variables representing the timing of field activities--especially the date of planting--
proved to be more significant and more stable than variables representing the level of
labor inputs.
In summary, due to data limitations, and consequent estimation procedures and
specification problems, the results of production function studies should be considered as
rough approximations to be treated with caution, as several researchers have acknowl-
edged (Ruthenberg, 1968; Massell and Johnson, 1968; Shapiro, 1973; Norman, Beeden
et al., 1976b; Upton, 1979).

Determinants of Resource Use
The second major use of regression analysis has been to identify factors affecting
the level of resource use by farmers. Applications have focused on determinants of
(1) the amount of land cultivated or labor used in farming activities, or (2) the rate of
diffusion or probabilities of adoption of a recommended input such as fertilizer or an
improved variety.

In general, regression analyses of factors affecting labor or land use have depended
on somewhat ad hoc selection of variables to be included as regressors. A limited amount
of the variation in resource use is generally explained in these applications. The major
purpose of these regressions has been to test hypotheses in order to improve an
understanding of farming system relationships. Using simple linear regression, Upton
(1967) and Luning (1967) demonstrated a positive and significant relationship between the
number of people available for farming activities and the area cultivated. Norman, Pryor,
and Gibbs (1979) estimated several regressions relating to levels of resource use, including
hours worked per hectare, total hours worked on the family farm, and hours worked per
hectare in northern Nigeria. Man-hours per acre were found to be inversely related to the
number of acres cultivated but positively related to the proportion of higher quality land.
Also, as might be expected, the authors were able to show that total family hours devoted
to work on the farm was directly related to the size of family and area cultivated and
inversely related to use of hired labor. On the other hand, days worked per male adult on
the farm was inversely related to the number of male adults available to work.
Farrington (1975a) used farm survey data in Malawi collected in the 1970-71 and
1971-72 cropping seasons to test the hypotheses that the length of the working day varies
according to the type of worker and the type of crop operation. The main regressors were
proxies for the energy requirements of a cropping operation and the urgency with which
task had to be performed. Farrington found that energy requirements had a stronger
influence on the length of the working day than did urgency, but the model explained only
a small amount of the observed variation in the length of the working day.
Researchers have included a wide range of independent variables in studies of
adoption patterns. Gerhart's (1975) study of maize diffusion in Kenya took into account
such factors as population density, proximity to a research station, average annual
rainfall, education, knowledge of credit, number of extension visits, and farm size and
found that agroclimatic zone was the most important variable in explaining adoption.
Binary dependent variable models have been used to evaluate factors affecting the
decision to adopt improved technologies. Falusi (1974/75), for example, used a multi-
variate probit model to analyze factors affecting the decision to use fertilizer in Nigeria.
Aklilu (1980) used a logit model in his study of fertilizer adoption in Ethiopia.-

Supply Functions
Since the mid-1960s, numerous studies have been carried out on the magnitude and
direction of smallholders' supply response for a wide variety of cash crops. Evidence on

!/See also "Agricultural Extension in Part IV.

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