lHESlS A

 

; LlfifiARY
:ichigan Qtate 5
University {

l‘ /

 

 

This is to certify that the

dissertation entitled

The Economics of Bas-Fond Rice Production in the
Eastern Region of Upper Volta: A Whole Farm
Approach.

presented by

Pascal Tagne Fotzo

has been accepted towards fulfillment
of the requirements for

Ph.D. , Agricultural Economics
degree in

 

 

Major professor

 

April 4, 1983
Date

 

MSU i: an Affirmative Action/Equal Opportunity Institution 0-12771

 

lllll Hlllll \lllllllllllllll

31 1293\ 01070 7853

 

 

1V1531_J RETURNING MATERIALS:
Place in book drop to

LJBRARJES remove this checkout from

”as. your record. FI.“__‘__ES_ will

 

 

 

be charged if book is
returned after the date
stamped below.

 

{98 new ~13

 

 

 

THE ECONOMICS OF BAS-FOND RICE PRODUCTION IN THE EASTERN REGION
OF UPPER VOLTA: A WHOLE FARM APPROACH

by

Pascal Tagne Fotzo

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY
Department of Agricultural Economics

1983

ABSTRACT

THE ECONOMICS OF BAS-FOND RICE PRODUCTION
IN THE EASTERN REGION OF UPPER VOLTA:
A WHOLE FARM APPROACH

By

Pascal Tagne Fotzo

Little is known about the costs and returns of current bas-fond
(saucer swamp) rice production techniques in Eastern Upper Volta, and
the possibilities for expanded production. This study gathered detailed
input/output data on four major bas-fond rice production systems, dif-
fering in degree of water control, in the Eastern Region of Upper Volta.
The multiple-Visit activity approach was used; 116 farmers were inter-
viewed from June, 1980, through February, 1981, at the end of each major
field activity (land preparation, weeding, harvesting, etc.). .

Financial enterprise budgets for all crops were prepared for each
production system. Gross margins and returns to land, family labor and
management were computed. Economic costs and returns to the rice enter-
prises were also analyzed for each production system. In addition, a
linear programming model was developed for one representative farm in
each production system, to investigate whether and how rice cultivation
could be expanded or intensified.

The findings showed that the least cost (and economically profit-
able) technique fbr producing rice is traditional cultivation in unim-

‘ proved swamps. Given current technologies and yield levels, production
under improved water control results in negative economic returns.

Using financial prices, rice entered the optimal solution in all LP

Pascal Tagne Fotzo

models, at acreage levels which held in each system under a wide range
of gross margins per hectare. Seasonal family labor supply was found to
be critical in determining the maximum level of total gross margins
attainable by the farmer from cropping activities.' Although rice is the
only feasible rainy-season crop on bas-fond land, this result implies
that price policy alone may not stimulate expanded or intensified rice
production. However, higher producer prices and a more productive tech-
nological package for rice would increase yields and help justify further
investment in water control.

The policy recommendations of this study stress the need (1) to
de-emphasize major investments in dam irrigation and to give priority to
partial water control and rainfed agriculture, (2) to develop a package
of improved rice production practices using the Farming Systems Research

approach, and (3) to revise the producer price of paddy as an incentive

to domestic rice farmers.

In memory of my late Father,

"SOUOP" TAGNE

who taught us basic survival skills

and

to my Mom,
NJIKE

for her support and understanding.

ii

ACKNOWLEDGMENTS

My sincere thanks to the many individuals and institutions who have
facilitated my graduate program at Michigan State University.

I especially wish to express my deepest appreciation to Professor
Carl K. Eicher, Chairman of my Guidance Comittee, for his inspiration
and guidance throughout my graduate program and.for the positive role he
played as my mentor. The guidance and helpful criticism of Dr. Eric w.
Crawford, my thesis director, are gratefully acknowledged. In addition,
I wish to thank the members of my guidance and thesis comnittees':
Professors Lester V. Manderscheid, Carl Liedholm and Gerald Schwab.

Financial support for my graduate program and this research project
came from various sources. I am grateful to the Ford Foundation for
Providing the financial assistance which enabled me to complete my Ph.D.
coursework and comprehensive examinations and also to cover part of my
dissertation typing and reproduction. Drs. Werner Kiene and Steven
31995, Ford Foundation Lagos, will be remembered for the understanding '
and flexibility they showed throughout my program. I am equally grate-
ful to the Department of Agricultural Economics at Michigan State Uni-
versity for their generous support for my data collection and processing
work and the many travel opportunities they offered me throughout my
graduate program. Professors Eicher and Manderscheid will be remembered
for the role they played in this respect. I am also thankful for the

One-year leave of absence that I was granted by the Ecole Nationale

iii

Superieure Agronomique (ENSA), University Centre of Dschang, even though
one year was not enough to fulfill either part or all requirements for
the degree of Ph.D. Dr. Jean Ongla, former head, Department of Rural
Economy at ENSA, will always be remembered for his stimulation and sup-
port during this painful undertaking.

Turning to the field phase of this research project, I wish to
express my inlnense gratitude to the farmers who willingly sacrificed
their time to supply the data for this study. My thanks go also to all
the four enumerators and two office assistants who worked to make this
research project a successful one. Among my colleagues in the MSU
research team in the EORD, a conmendation is due to Drs. David Wilcock _
and Gregory Lassiter and Kifle Negash for their encouragement and help
in planning and carrying out this research. Thanks also to all the
Upper Volta government officials in the study area and at the central
level whose support and assistance was invaluable.

Data processing and secretarial help came from a variety of sources.
A conmendation is due to Paul Holberg and Chris Nolf for their extra
efforts in moving my problem through the computer. My thanks go to
Lucy Wells and Cindy Spiegel for typing the drafts of this dissertation,
and to Lois Pierson for typing the final draft.

Finally, loving thanks to my family and family friends for the

many ways in which they facilitated the completion of my graduate work.

iv

TABLE OF CONTENTS

Page
LIST OF TABLES ......................... ix
LIST OF FIGURES ........................ xiv
LIST OF MAPS .......................... xiv
LIST OF ABBREVIATIONS AND ACRONYMS ............... xv
CHAPTER ONE: INTRODUCTION ................... l
1. THE COUNTRY ....................... l
1.1. General Description ................ l
1.2. The Eastern Region ................ 3
1.3. The Economy .................... 4
1.4. The Agricultural Sector .............. 5
2. PROBLEM SETTING AND NEED FOR THE STUDY ......... 9
3. OBJECTIVES OF THE STUDY ................. 14
CHAPTER THO: RESEARCH APPROACH AND DATA COLLECTION METHODS . . 17
1. SCOPE OF THE STUDY ................... l7
2. RESEARCH APPROACH .................... 18
3. THE FIELD RESEARCH SITES ................ 19
4. SELECTION OF SAMPLE FARMERS .' .............. 22
5. INTERVIEW FREQUENCY ................... 24
6. ENUMERATOR TRAINING AND RESEARCH CALENDAR ........ 27
7. DATA PREPARATION AND ANALYSIS .............. 3O
8. DATA LIMITATIONS .................... 32
8.1. Cross-Sectional Analysis Problem ......... 33

8.2. Representativeness of the Rice Production
Systems Studied ........... - ..... 33
8.3. Reliability of Sample Farmers' Responses ..... 34
8.4. Problem of Estimating Labor Time ......... 34

8.5. Problem of Estimating Quantities of Other
Inputs and Outputs ............... 35
9. SUMMARY ......................... 36

CHAPTER THREE: DESCRIPTIVE PROFILE OF THE AGRICULTURAL

4.

PRODUCTION SYSTEMS IN THE EASTERN REGION
OF UPPER VOLTA .................

CROPS GROWN AND DEGREE OF MIXED CROPPING ........

SOME AGRICULTURAL CONSTRAINTS IN THE EASTERN REGION . . .
2.1. Land Tenure ....................
Marketing and Processing .............
Water Control ...................
Research and Technological Problems . . . ... . . .
. . Extension Problems ................
PRODUCTION SYSTEMS STUDIED ...............
3.1. Definition of the Systems .............
3.2. Descriptive Profile of Sample Farmers .......
3.2.1. Land ...................
3.2.2. Labor ...................
3.2.2.1. Size and Composition of
the Household ..........
3.2.2.2. Average Age of Heads of
Households ...........
3.2.3. Importance of Non-Farm Activities .....

SUMMARY .........................

NNNN
(31-th
O . 0

CHAPTER FOUR: A FINANCIAL AND ECONOMIC ANALYSIS OF THE FOUR

1.
2.

RICE-BASED PRODUCTION SYSTEMS IN THE EORD . . . .
DISTINCTION BETWEEN FINANCIAL AND ECONOMIC ANALYSIS . . .

DEFINITION OF CROP ENTERPRISES AND PROCEDURES USED
TO SELECT CROP MIXTURES ...............

CROP ENTERPRISE BUDGETS .................
3.1. System 1: Production System Based on Farmers
Growing Rice in Traditional Bas-Fonds .....
3.1.1. Overview of the Enterprise Budgets
in System 1 ..............
3.1.2. Comparison and Appraisal of the Six
Major Enterprises in System 1 .....
3.1.2.1. Gross Margin (GM) ........
3.1.2.2. Net Margin (NM) ........

3.1.2.3. Net Returns to Land, Family
Labor and Management (NRLFLM)
(NRLFLM) ...........

3.1.2.4. Returns to Land and Management
(RLM) ..............
3.1.2.5. Costs of Production .......
3.2. System 2: Production System Based on Farmers

Growing Rice on Semi-Traditional Bas-Fonds . . .

3.2.1. Overview of the Enterprise Budgets in
System 2 .................

vi

58
58

59
61
63
63
71
71

73

74
76

76
76

4.

3.2.2. Comparison and Appraisal of the Seven

Major Enterprises in System 2 .....
3.2.2.1. Gross Margin (GM) ..... - . . .
3.2.2.2. Net Margin (NM) .........
3.2.2.3. Net Returns to Land, Family
Labor and Management
(NRLFLM) ...........
“3.2.2.4. Returns to Land and Management
(RLM) ............
3.2.2.5. Cost of Production .......
3.3. System 3: Production System Based on Farmers
Growing Rice in Improved Bas-Fonds .......
3.3.1. Overview of the Enterprise Budgets in
System 3 ................
3.3.2. Comparison and Appraisal of the Five
Major Enterprises in System 3 .....
3.3.2.1. Gross Margin (GM) ........
3.3.2.2. Net Margin (NM) .........
3.3.2.3. Net Returns to Land, Family
Labor and Management
(NRLFLM) ...........
3.3.2.4. Returns to Land and Management
RLM .............
3.3.2.5. Costs of Production .......

3.4. System 4: Production System Based on Farmers
Growing Irrigated Paddies with Fertilizer
3.4.1. Overview of the Enterprise Budgets

in System 4 ..............
3.4.2. Comparison and Appraisal of the Six
‘ Major Enterprises in System 4 .....
3.4.2.1. Gross Margin (GM) ........
3.4.2.2. Net Margin (NM) .........
3.4.2.3. Net Returns to Land, Family
Labor and Management
(NRLFLM) ...........
3.4.2.4. Returns to Land and Management
(RLM) ............
3.4.2.5. Costs of Production .......

3.5. Comparison and Appraisal of the Rice Enterprises
Across the Four Production Systems Studied . . .
3.6. Social Profitability of Rice Cultivation Under

the Four Systems Studied ............
3.6.1. Shadow Prices of Domestic Factors and

Output .................

3.6.1.1. Fertilizer ...........

3.6.1.2. Water Control Costs .......

3.6.1.3. Import Parity Price of Rice . . .

SUMMARY .........................

vii

85

87
88

88
88
95
97

97

98
98

99

99

108
108

110
112

112
112
113
113

117

CHAPTER FIVE: RICE PRODUCTION VERSUS PRODUCTION OF OTHER
MAJOR COMPETING CROPS: AN OVERVIEW OF THE
ANALYTICAL MODEL ................

1. THE ANALYTICAL MODEL ..................
1.1. Building the "Representative" Farm Model for

Each Production System Studied .........
1.2. Model Structure ..................
1.2.1. Resource Constraints Used .........
1.2.2. Activities and Objective Function .....

l 2 3 Derivation of the Numerical Coefficients
of the Model ..............

2. TABLEAUX FOR THE FOUR REPRESENTATIVE FARMS OF THE
SYSTEMS STUDIED ...................

CHAPTER SIX: EVALUATION OF THE BASIC MODEL, PRESENTATION OF
RESULTS, AND SENSITIVITY ANALYSIS ........

1. EVALUATION OF THE MODEL .................
2. PRESENTATION OF THE RESULTS FROM THE AVERAGE MODEL

3. SENSITIVITY OF THE BASIC MODEL TO CHANGES IN
SELECTED PARAMETER VALUES ..............
3.1. Comparison of Optimal Production Strategies Under
Different Models (A, B, c, cl, 0, and E) . . . .
3.2. The Opportunity Cost of Scarce Resources and
Constraints to Increased TGM ..........

4. SUMMARY .........................

CHAPTER SEVEN: SUMMARY, POLICY RECOMENDATIONS, AND FURTHER
RESEARCH ....................

1. SUMMARY ..........................

2. POLICY IMPLICATIONS AND STRATEGY TO IMPROVE THE
PERFORMANCE OF RICE PRODUCTION IN THE EORD ......
2.1. Major Policy Issues and Reorientation .......
2.2. Identification of an Appropriate Bottomland
Rice Production Strategy in the EORD ......
2.2.1. Small Scale Irrigation Using Dikes .
2.2.2. Improved Production Practices .......

3. SUGGESTIONS FOR FUTHER STUDIES .............
APPENDIX A: SURVEY FORMS USED FOR THE STUDY .........
APPENDIX B: FIELD SPECIFIC FAMILY LABOR PROFILE FOR CROPS

OTHER THAN RICE IN ALL THE FOUR PRODUCTION
SYSTEMS STUDIED .................

BIBLIOGRAPHY ..........................

viii

122

124
125
127
134

136

180
181

183
183

195
197

200
200
200

201

Table
1.1
2.1

3.1

3.2
3.3

3.4

3.5
3.6

4.1.1

4.1.2

4.1.3

4.1.4

4.1.5

4.1.6

4.5.1,

LIST OF TABLES

Historical Evolution of the Rice Sector in Upper Volta . .

Details of the Schedules Used for Colleting Agro-

Economic Data in the EORD, 1980-81 .........

Hectares of Crops Grown and Percentage Grown as Sole
Crops and in Different Crop Mixture Classes, 1980-81

Rainfall Data in mm in Fada-N'Gourma, by Period, 1975-80 .

Average Number of Fields in Total and in the Bas-Fonds,

per Household, 1980-81 ...............

Average Area of Farms in Total and in the Bas-Fonds,

1980-81 .......................

Average Age of Household Heads by System, 1980-81

Relative Percentage of Household Heads with at Least

One Non-Farm Activity per System, 1980-81 ......

Average Costs and Returns per Hectare for Rice,

System 1, in the EORD, 1980 .............

Average Costs and Returns per Hectare for Sorghum/

Millet/Cowpeas, System 1, in the EORD, 1980 .....

Average Costs and Returns per Hectare for Maize,

System 1, in the EORD, 1980 .............

Average Costs and Returns per Hectare for Groundnuts,

System 1, in the EORD, 1980 .............

Average Costs and Returns per Hectare for Bambera Nuts,

System 1, in the EORD, 1980 .............

Average Costs and Returns per Hectare for Soybeans,

System 1, in the EORD, 1980 .............

Comparative Analysis of the Major Enterprises in
.System 1, Based on Survey Data from 26 Households,

1980 ........................

ix

Page
10

28

41
44

51

51
55

56

64

65

66

67

68

69

Table
4.2.1

4.2.2

4.4.1

4.4.2

4.4.3

Average Costs and Returns per Hectare for Rice,

System 2, in the EORD, 1980 .............

Average Costs and Returns per Hectare for Sorghum/

Millet/Cowpeas, System 2, in the EORD, 1980 .....

Average Costs and Returns per Hectare for Maize,

System 2, in the EORD, 1980 .............

Average Costs and Returns per Hectare for Groundnuts,

System 2, in the EORD, 1980 .............

Average Costs and Returns per Hectare for Soybeans,

System 2, in the EORD, 1980 .............

Average Costs and Returns per Hectare for Okra,

System 2, in the EORD, 1980 .............

Average Costs and Returns per Hectare for Cotton,

System 2, in the EORD, 1980 .............

A Comparative Analysis of the Major Enterprises in
System 2, Based on Survey Data from 30 Households,

1980 ........................

Average Costs and Returns per Hectare for Rice,

System 3, in the EORD, 1980‘ .............

Average Costs and Returns per Hectare for Sorghum/

Millet/Cowpeas, System 3, in the EORD, 1980 .....

Average Costs and Returns per Hectare for Maize,

System 3, in the EORD, 1980 .............

Average-Costs and Returns per Hectare for Groundnuts/

Bambera Nuts, System 3, in the EORD, 1980 ......

Average Costs and Returns per Hectare for Soybeans,

System 3, in the EORD, 1980 .............

A Comparative Analysis of the Major Enterprises in
System 3, Based on Survey Data from 30 Households,

1980 ........................

Average Costs and Returns per Hectare far Rice,

System 4, in the EORD, 1980 .............

Average Costs and Returns per Hectare for Sorghum/

Millet/Cowpeas, System 4, in the EORD, 1980 .....

Average Costs and Returns per Hectare for Maize,

System 4, in the EORD, 1980 .............

Page

77

78

79

80

81

82

83

89

9O

91

92

93

100

101

102

Table
4.4.4

4.4.5

4.4.6

4.5.4

4.5.5

4.5.6

4.5.7

5.1

5.10

Page
Average Costs and Returns per Hectare for Groundnuts,
System 4, in the EORD, 1980 .............. 103
Average Costs and Returns per Hectare for Bambera
Nuts, System 4, in the EORD, 1980 ........... 104
Average Costs and Returns per Hectare for Okra,
System 4, in the EORD, 1980 .............. 105
A Comparative Analysis of the Major Enterprises in
System 4, Based on Survey Data from 30 Households,
‘1980 ......................... 107
A Comparative Financial Analysis of the Four Major
Rice Production Techniques in the EORD, Based on
Survey Data from 116 Households, 1980-81 ....... 111
Import Parity Price of a Ton of Paddy Produced in
the EORD, 1980 .................... 114
A Comparative Economic Analysis of the Four Major
Rice Production Techniques in the EORD, 1980 ..... 116
Main Labor Periods and Activities Covered, System 1,
1980 ......................... 129
Main Labor Periods and Activities Covered, System 2,
1980 ......................... 130
Main Labor Periods and Activities Covered, System 3,
1980 ......................... 131
Main Labor Periods and Activities Covered, System 4,
1980 ......................... 132
1 Mean Percentage of Household Land Allocated to Each
Crop Category, 1980 .................. 135
Average Number of Workers per Household in All the
Four Production Systems Under Study in the EORD,
1980 . . . .‘ ..................... 138
Summary of Area and Proportion of Each Type of Land
Farmed by Farm System, 1980 Survey .......... 141
Minimum Area Under Sorghum/Millet/Cowpeas for the
Four Production Systems, 1980 ............. 142
Maximum Area that Can be Grown in Maize and Soybean
for Four Production Systems, 1980 ........... 143 ‘
Basic EORD Farm Linear Program, System 1, 1980 ...... 145

xi

Table

.12
.13

010101

6.13

6.14

Page

Basic EORD Farm Linear Program, System 2, 1980 ...... 146
Basic EORD Farm Linear Program, System 3, 1980 ...... 147
Basic EORD Farm Linear Program, System 4, 1980 ...... 148
Actual Versus Optimal Allocation of Land to Each Crop

Category, by System, 1980 (percentages) ..... '. . . -150
Total Labor Use Under Actual Allocation Versus Total

Labor Use Implied by the Model, by Crop Category

and by System, 1980 .................. 152
Family Labor Profile for Selected Fields Belonging to

the Rice Enterprise, System 1, 1980 .......... 153
Family Labor Profile for Selected Fields Belonging to

the Rice Enterprise, System 2, 1980 .......... 154
Family Labor Profile for Selected Fields Belonging to

the Rice Enterprise, System 3, 1980 .......... 155
Family Labor Profile for Selected Fields Belonging to

the Rice Enterprise, System 4, 1980 .......... 156
Results from the Basic Model: Enterprises in the

Optimal Solution for System 1 ............. 158
Results from the Basic Model: Enterprises in the

Optimal Solution fur System 2 ............. 159
Results from the Basic Model: Enterprises in the

Optimal Solution for System 3 ............. 160
Results from the Basic Model: Enterprises in the

Optimal Solution for System 4 ............. 161
Shadow Prices of Resources and Constraints Used in

the Model, by System, 1980 .............. 165
Summary of the Production Strategies which Maximize

Total Gross Margins (TGM) Under Different Versions

of the Basic Model, System 1, 1980 .......... 170
Summary of the Production Strategies which Maximize

Total Gross Margins (TGM) Under Different Versions

of the Basic Model, System 2, 1980 .......... 171

Summary of the Production Strategies which Maximize
Total Gross Margins (TGM) Under Different Versions
of the Basic Model, System 3, 1980 .......... 172

xii

Table

6.15

6.16

6.18

6.19

Page

Summary of the Production Strategies which Maximize

Total Gross Margins (TGM) Under Different Versions

of the Basic Model, System 4, 1980 .......... 173
Shadow Prices of Resources and Constraints Used in

the Various Models, System 1, 1980 .......... 176
Shadow Prices of Resources and Constraints Used in

the Various Models, System 2, 1980 .......... 177
Shadow Prices of Resources and Constraints Used in

the Various Models, System 3, 1980 .......... 178

Shadow Prices of Resources and Constraints Used in
the Various Models, System 4, 1980 .......... 179

xiii

LIST OF FIGURES

Figure
3.1 Incorporation of the Time Dimension into Cropping
Patterns Under Indigenous Conditions ........
3.2.a Sketch of the Bas-Fond Improvement Structure, Type 1a:
Opened Structure ..................
3.2.b Sketch of the Bas-Fond Improvement Structure, Type 1b:
Semi-Opened Structure ................
3.3 Sketch of the Bas-Fond Improvement Structure, Case of
Dam Irrigation ...................
3;4 Different Categories of Farm Labor in the EORD,
1980-81 .......................
LIST OF MAPS
Plan
1.1 Upper Volta International Boundaries ..........
1.2 Upper Volta Administrative Divisions ..........
2.1 Location of Research Sites ...............

xiv

.0
DJ
a> no lg

ARGMVl

AVV

BAEP
CENATRIN

CERCI

COREMMA

CSPPA
DER

EORD
FAO
FSR
GDP
GNP
GOUV
GM

IITA
IMF
IRAT

LIST OF ABBREVIATIONS AND ACRONYMS

Atelier Regional de Construction des Materiels Agricoles
(Regignal plant for the manufacture of agricultural
tools .

Autorite des Amenagements des Vallées des Voltas (Volta
bottom land Development Authority)

Bureau Analyse Economique et Planification

Centre National pour 1e Traitement de 1'Information
(National Center for Data Processing)

Centre d' Experimentation du Riz et des Cultures Irriguees
(Research Center on Rice and Irrigated Crops)

Cooperative Regionale de Montage de Materiels Agricoles

Caisse de Stabilisation des Prix des Produits Agricoles

Developpement des Entreprises Rurales

Eastern ORD

Food and Agriculture Organization of the United Nations
Farming Systems Research

Gross Domestic Product

Gross National Product

Government of Upper Volta

Gross Margin (gross income minus the variable expenses
attributable to that enterprise)

International Institute of Tropical Agriculture
International Monetary Fund
Institut de Recherches Agronomiques Tropicales et de

Cultures Vivrieres (Research Institute on Tropical
and Food Crops)

XV

TWNACER = Office National des Céréales (National Cereals Agency)

01131 = Office National des Barrages et de l'Irrigation (National
Agency for Dams and Irrigation)

0RD = Office Regional de Development (Regional Development Office)

SAED = Société Africaine d'Etudes et du Developpement (Private
Research Consulting Firm)

SATEC = Société d'Aide Technique et de Cooperation (French Rural
Development Agency)

S/M/C = Sorghum/millet/cowpeas

SMIG = Salaire Minimum Interprofessional Guaranti

SOVOLCOM = Sociéte Voltaique de Commercialisation (Voltaic Marketing
Organization)

TAC = Technical Advisory Comittee .

TGM = Total Gross Margin

USAID = United States Agency for International Development

UV = Upper Volta

.Note: Currency unit = CFA (Communauté Financiére Africaine)

US $1 = 220 CFA (average exchange rate in 1980)

Weights and Measures: Metric system was used in this study

mm = millimeter
ha = hectare
kg = kilogram
T = ton
km = kilometer
0C = degree Celsius
km2 =r square kilometer

xvi

CHAPTER ONE

INTRODUCTION

The objective of this chapter is three—fold. First, to describe
the geographic and socio-economic characteristics of Upper Volta, parti-
cularly the Eastern Region, for those who are not familiar with them.
Second, to outline the salient features of agriculture in Upper Volta
which are pertinent for understanding agricultural problems in the
Eastern Region of the country. Third, to present the objectives of the

study and the organization of the dissertation.

1. THE COUNTRY
1.1. General Description

Upper Volta is one of the four landlocked Sahelian Countries in
West Africa, the other three being Chad, Niger and Mali. Upper Volta
is bounded along its northern and western borders by Mali. Niger forms
the eastern boundary and the southern boundary is shared by Benin, Togo,
Ghana and Ivory Coast(see Map 1.1).

Official estimates indicate a total population of about 6.15 mil-
lion in 1980, which makes Upper Volta the most densely populated country
in the Sahel. The population is unevenly distributed, varying from a

density of 10 persons/km2 in the Volta valleys to 40 persons/km2 in the

AFRICA

UPPER VOLTA

 

 

A moo

IVORY COAST

MAP 1.1 UPPER VOLTA INTERNATIONAL BOUNDARIES

Source: Adapted from Martin Greenwold Associates,
Inc., "Maps on File", 1982.

 

|\i

all

Mossi plateau in Central Upper Volta (IFAD, 1981). An important
implication of this variation in population density is the variation in

land use intensity throughout the country.

1.2. The Eastern Region

Viewed as a whole, the Eastern Region of Upper Volta is a vast
peneplain with the difference between the highest elevation and the low-‘
est elevation being only 101 m. The Eastern Region covers an area of
49,992 kmz which represents about 18 percent of the total area of the
nation (Mehretu and Wilcock, 1979). Because of the flat topography,
the drainage system in the Eastern Region is very poor resulting in a
large number of saucer swamps,genera11y called bgsrjpggg, found along
many of the permanent to semi-permanent rivers.

TWO main categories of has-fonds can be distinguished: (1)
improved bas-fonds and (ii) traditional has-fonds. According to
Weldring (1979), 224 hectares of the first category and 379 hectares of
the second are already under cultivation, but about 2,187 hectares could
possibly be put under cultivation. Many organizations are involved in
has-fond land development in the Eastern Region--they include the FAO,
the C.T.S. (a Swedish project) and the D.E.R. project (Partnership for
Productivity project).

The dominant characteristics of the climate are sustained heat
(average temperatures around 33°C can be expected in the hot season) and
seasonal rainfall. The alternate north-south movement of the continental
air masses, as they follow the annual migration of the sun, bring about
Sharp periodic differences in rainfall. Two seasons can be distinguished:

a short rainy season running from June to-September (rains are heaviest

 

0“
l\
to.

o'-
I
V I

o.-

...

01:

'c

5‘

h:

 

 

in August) and a long dry season running from October to May. However,
December, January and February are relatively cool because of the
passage of the harmattan (a seasonal wind from the north). Rainfall
varies between 1,000 mm in the southern part of the region to 600 mm in
,the northern part (Mehretu and Wilcock, 1979). Together with the vari-
ation in population denSity, these climatic characteristics are import-
ant determinants of the land use pattern in the Eastern Region.

The population of the Eastern Region of Upper Volta is estimated at
about 443,000, which represents only 7 percent of the national popula-
tion. It is one of the least densely populated areas in the country.
The population of the Eastern Region comprises the following ethnic
groups: Gourmantché (64 percent), Mossi (28 percent), and Peuhls and
Others (8 percent). These groups are differentiated by their traditions,
customs, and languages. However, common to all these groups is the fact
that they live in small villages made up of several compounds which con-

tain members of a family or close relations (Swanson, 1975).

1.3. The Economy

Upper Volta is recognized as one of the poorest countries in the
world. Living standards are very low, as illustrated by an average life
expectancy at birth of 39 years and an adult literacy rate of only 5
percent. The 1980 GNP per capita was 210 dollars (World Bank, 1982).
Although heavily rural and agricultural, Upper Volta's agricultural pro-
duction has been declining in real terms over the past decade, -1.2 per-
cent between 1970 and 1979 (World Bank, 1982). According to the World
Bank Development Report, 1981, the distribution of GDP in 1980 was as

’9

n1

1;)

11..

“

follows: 40 percent agriculture, 18 percent industry, 13 percent
manufacture and 29 percent services. In 1980 the exports of agricul-
tural products accounted for 87 percent of total export earnings. The
same World Bank report indicates that capital equipment constituted 29
percent of total import bill while food made up 22 percent of the
imports.

The strong trading links between Upper Volta and France have con-
tinued since independence. Together with Ivory Cost, France was still
the main trading partner of Upper Volta in 1980. The economic bonds
between Upper Volta and the other former French colonies in West Africa
are very weak, despite a common heritage and common interests. The main
exports of Upper Volta are livestock, oil seeds and cotton lint; and on
the import side, major trade items include capital equipment, industrial
raw materials, foodstuffs and chemicals.

No national accounts have been published since 1975. However,
according to IMF estimates, the share of commerce increased from 1975 to
1978 from 15 to 32 percent, while agriculture declined from 45 to 38
percent. Growth of GDP at constant prices has been estimated at 4 per-
cent in 1978 after a 7 percent high in 1977 (World Bank, 1981). The
rate of inflation measured on the basis of the price index for an African
family in Ouagadougou was estimated at 32.9 percent in 1977 and 30.5
percent in 1980 (IFAD, 1981).

1.4. The Agricultural Sector

Agriculture (including livestock) plays a commanding role in the

daily life of the people of Upper Volta and is by far the main element

in the country's economy. Although somewhat less than 33 percent of

the entire land area is classified as arable, nearly 90 percent of the
population is dependent on agriculture as a livelihood (IFAD, 1981).

Upper Volta has a complex geographical structure and a wide variety

of climate. Following an IFAD report (1981), Upper Volta can be
divided into four regions in terms of its agricultural potential: (1)
the southwest with relatively fertile soils, moderate population density
and an average rainfall of 900nm; (ii) the Mossi plateau with poor soils
and a highpopulation density (nearly 40'persons per kmz); (iii) the
southeastern savannah zone with characteristics similar to those of the
southwest zone, but poor road infrastructure; and (iv) the dry northern
zone where livestock is the major activity.

Upper Volta is currently in its third five-year development plan
(1978-1982).1 The first priority cannon to all three development plans
is rural development. The main objectives of the government's policy.
for the rural sector include: (i) to develop rainfed agriculture by pro-
"Pting improved farm practices, while integrating cropping and livestock
“ti V1 ti es; (ii) to step-up migration from the densely populated and
relatively infertile north-central plateau .to areas in the West and
Southwest which have low population densities and a good agricultural
potent1&1; (iii) gradually to intensify the development of swampland and
WMgated agriculture, thus helping to protect the nation against the

catastrophe caused by drought; and (iv) to ensure national self-sufficiency
\

1 o
1980 The third five-year plan was not yet officially released as of mid-
' HOwever. the plan's Avant Projet is generally used as the guide-

line for- development programs.

 

7

in food crops, particularly by replacing rice imports (Third five-year

Development Plan, .1978). These priorities are pursued by a variety of

projects financed by external donors.
The country is divided into eleven Regional Development Offices

(ORDs), one for each pre'fecture except for the pre'fecture of the Hauts-

Bassins in the Southwest which contains two ORDs (See Map 1.2). The

ORDs, which were created in 1967, enjoy some autonomy with respect to

project planning and implementation and use of resources within the

broad policy guidance of the Plan. The objectives of the 0RD structures

were defined as promotion of agricultural production, development of-

rural infrastructure and equipment, and social development.

Agricultural production practices in Upper Volta are characterized

by extremely low yields, particularly for the food crops. About 90 per-

cent of the cultivated land in Upper Volta is devoted to cereal crops.
Yields achieved in Upper Volta are among the lowest in the world; grain
c"PP .Yi e1 ds range from 300 to 800 kg/ha, and yields for cotton are about
500 kg! ha. However, research station cereal yields range upward from

"000 kglha (IRAT, IRCT, 1979). For most food crops, field activities

are al 1 done by hand. Chemicals and fertilizers are rarely used on food
CY‘Ops.

The fact that hi gh-yieldi 119 varieties of food crops have not been
developed to any important degree is a reflection of the neglect of
research in this area. The government still ddes not have an organized
r"’i'i‘eal‘ch structure. Research responsibilities are still assumed by the

F"ench‘style vertically organized crop-specific research institutions

“RAT: IRCT, ORSTOM, etc). This applies to virtually all agronomic as

well 35 1 ivestock research. More recently, international research

Pecowumz macecamemowm paupumcm "ougsom

monmH>Ha m>~h<mhmuzazo< <h4¢> «mad: N.— a<z
_ hr _ opuum

..e: A pmeeu see,—

 

Auoeu :ogoomoav
acoauconoo accumuu

 

 

institutes--ICRISAT, IITA and WARDA--are undertaking research on farm

production systems, looking both at the cropping and socio-economic

aspects of the systems. For more details, see ICRISAT, 1980; IITA/

SAFGRAD, 1980; and WARDA, 1979.
Animal traction for land preparation and cultivation is being pro-

moted as a means to increase agricultural production. However, data on

the impact of animal traction on the productivity of land suggest that

only modest results have been achieved thus far. For more details on

animal traction in West Africa, see Kline, et a1. (1969); Sargent, et a1.,
1981; and Barret, et a1., 1982.

As a result of food deficits and nutritional problems in Upper
Volta. the government initiated a multiplicity of high priority programs
9'19 Projects aimed towards the recuperation of lowlands. The Fonds de
De'vel oppement Rural (FDR) and L' Office National pour les Barrages et
Pe.V‘lmétres Irrigués (ONBI) are the two main institutions in charge of
the above programs, working in close collaboration with the ORDs. One
1""3071iant implication of these programs is that rice, which is the main
"OP grown in these lowlands or bas-fonds, has seen its domestic produc-
tion Thereased in recent years. Although this is an attempt to match
the increased local demand estimated at 8 kg/inhabitant/year, rice

OUT-put has fallen short of required levels (see Table 1.1). Domestic

ri -
Ce p‘l‘Oduction only covers about 65 percent of domestic demand.

2. PROBLEM SETTING AND NEED FOR THE STUDY

As one of its first drought recovery projects in the Sahel, USAID

Pmflded substantial material and technical assistance to the Eastern ORD.

10
TABLE 1.1

HISTORICAL EVOLUTION OF THE RICE SECTOR
IN UPPER VOLTA]

 

 

Rice Imports

 

 

 

Padd Pr d ction -

Yea r ()‘IOOOOmTfl ("1(11010e0d mrti)ce)
1951 31 2.341
1953 45 2.183
1954 . 25 4.755
1955' 34 3.222
1955 35 4'084
1957 ’34 7.459‘
1968 44 1.327
1959 40 1.475
1970 34 2' 546
1971 34 1.127

1972 37 1.598
1973 30 1 .000
1974 . 32 18.700
1975 39 15.200
1975 40 15. 400
1977 4] 18. 382
1978 28 10. 237
1979 47 25. 455
L 29 29.000
\

1

Source: FAO,1965-81a&b.

11

Michigan State University's contract2

provided the technical assistance

component of the USAID Integrated Rural Development Project in the

Eastern 0RD. This technical assistance was a combination Of program
implementation and applied research for improving program design and
execution. The project team included a production economist whose task
at the beginning of the project (1977-1979) was to organize and coor-
dinate a survey of 480 farmers throughout the EORD, focusing primarily
on dry] and crop production and use of animal traction. In the course of
the implementation Of the project, the following concerns emerged as
priority activities of the EORD:

1. Increased water control in order to satisfy human consumption
and agricultural production needs;

2. As most foodstuffs consist Of grains produced under dryland
conditions, vagaries of weather Often cause shortages. The
situation has been particularly serious in the the 7-8 years
as a result of the drought that struck the whole Sahelian
zone. Hence, it was decided that priority should be given
to areas with more reliable water availability, namely the
has-fond areas;

3- Promotion of rice production in the seasonally flooded bas-

‘ fond lands. 1
1" Upper Volta, rice still has a marginal share in the average diet

Of the Depulation. Average per capita consumption is about 8 kg 5 year.
\

21““ chigan State Uni versity's four-year contract to provide the
techni Ca] assistance component Of the USAID Integrated Rural DevelOpment
PY'OJec-t in the Eastern 0RD of Upper Volta was signed and went into
effect On May 1, 1977 under the reference number AID/Afr-C-l3l4.

1’ ‘ :4
'4:':._
I .‘F
1.”:t '-
.

O I.
b a)

an":
Fl. .I
.

on '1
:3.

0|

«Yr
.1:
ll»

Dob

‘D!

12

However, although the staple food is sorghum and millet, rice is Of
growing concern to the government of Upper Volta (GOUV) because in
towns, rice is rapidly replacing sorghum and millet in the diet. This
change in food habits of the urban population caused rice imports to
soar from 2,341 tons in 1961 to about 29,000 tons in 1980. Although
these quantities may seem small, Upper Volta has a large trade deficit,
with the value of exports averaging only a third of the value of imports.
This provides an incentive for the country to cut imports of foods which
it can produce satisfactorily itself (WARDA, 1981). Rice is therefore
a high priority investment for the GOUV, the EORD and many donor agen-
ci es (USAID, World Bank, SATEC, etc.).

Despite the need to expand rice production in Upper Volta in
general and in the Eastern Region of the country in particular, very
little is known about the productivity and/or the efficiency of resources
comi tted to bas-fond rice production under alternative cropping
systenis. Until very recently little was known about the agricultural
systems of small farmers in West Africa. In the past decade various
organizations and research groups have begun to study these systems,
mainly in the coastal countries. Few studies have included the Sahelian
countries. But the drought in the early 1970's focused attention on
these countries. Before programs aimed at increasing rice production
can be PPOposed or instituted, it is first necessary to understand the
9x15131119 farming system and at least attempt to predict the reaction Of
that System to various policies. However, research reports on the econ-

omy 0f has-fond rice production in Upper Volta are very few.

AA 0.
v. ‘
u

a
O '0
r

‘f

D
'0'.

c-IUU

‘I-
-'
Uhv.

‘ .
1:"

H‘,

O.
(\p
.1

l \

13

TWO studies by WARDA (1975 and 1979) show that intensive irrigated
rice growing was introduced in Upper Volta through technical assistance
from Taiwan. Under an agreement signed in 1964 by the governments Of
Upper Volta and Taiwan, development work was carried out at Boulbi and
Louda from 1965 to 1968, downstream Of already existing small dams.
Yields of 4 tO 6 tons Of paddy per hectare were Obtained in both areas.
Unfortunately, the water supply in those areas was inadequate to pro-

duce two crops a year. In any case, rice production in the bas-fond
areas per se was completely overlooked.

Other rice research reports3 on different ecological areas in
Upper Volta are available, but they emphasize purely agronomic issues
(varietal improvement, fertilizer trials, pest control, etc.). More-
over, little empirical ~study has been done Of the bas-fonds areas. How
farmers operate in terms Of resourse use and production techniques,
their cost Of production both in absolute and relative terms, and their
Performance in terms Of output and incomes, are questions which have
been overlooked. However, it is clear that the production of a staple
cTOP in West Africa will respond not only tO the availability Of fer-
tilizers and pesticides, but also to price incentives. It is therefore
l'mportant to examine the relative profitability Of different crops,

Particularly of sorghum and millet and maize relative to rice, in

Eastern Upper Volta.

\

3Angladette, 1966; IRAT, 1980; Mayer and Bonnefond, 1973; and

 

otPW

of“

‘1

( D
-.-

14
3. OBJECTIVES OF THE STUDY

The general goal Of this study then was to provide baseline
microeconomic data necessary for regional planning and for suggested
improvement in farm practices, with a special focus on the relative
position of rice vis-a-vis other major crops in the cropping system Of
farmers in the has-fonds areas Of the Eastern Region of Upper Volta.
This study will serve at least two needs:

1. TO provide data for planning purposes and ex-ante agricultural

policy evaluation.

2. TO provide technical and economic knowledge to researchers,
extension services and farmers; It is generally agreed that
small farmers use their limited resources and knowledge effi-
ciently in their traditional farming systems. TO improve
the small farmer's welfare or income, according to this
"efficient but poor hypothesis," it is necessary to provide
improved technology and better information on market prices
and prospects.

This study analyzes data collected from a sample Of 116 farmers in
1’Our different has-fonds production systems. Four representative farms,
9'18 in each production system, are developed for more detailed analysis
Of production, labor use and income.

More Precisely, the Objectives Of this study are:

1- TO diagnose the problems and constraints on bas-fonds rice
PI‘Oduction in the Eastern 0RD. Information from this diag-
nostic stage can be used to guide experiment station research

such as the development Of new technological components,

15

particularly varieties and water control methods. The
diagnostic effort can also uncover'constraints at the farm
level which are the result Of policy decisions or problems
in policy implementation (e.g., problems of input avail-
ability, product marketing, management Of water at the dam
sites, etc.).

TO determine the profitability Of rice growing relative to
other crops in the farming system. .In assessing the rela-
tive profitability Of different crops, two elements must be
considered: relative input and output prices and relative
yields. Simple enterprise budgets are developed for this
purpose.

TO determine the relative labor requirements of different
enterprises both in terms Of total labor use and labor profile
throughout the cropping season.

TO evaluate existing cropping enterprises within a whole-
farm framework through the development Of linear programming
models for the four representative farms. Optimal farm
plans for each Of the four representative farms are Obtained
under several alternative combinations Of resource levels,
resource requirements, and Objective function coefficients.
The shadow prices Of different resources and constraints are
also analyzed for the four production systems.

'TO discuss the policy implications Of the analysis for future
programs in the bas-fonds areas Of the Eastern Region in

particular and of Upper Volta in general.

“ion.
a

'0'.

I I
F

-

.1_1

~:d

I
lent
0

Un-

l6

4. ORGANIZATION OF THE DISSERTATION

The research approach and data collection methods are presented in

Chapter Two. A descriptive profile Of agricultural production systems
Chapter Four con;

in the Eastern Region is presented in Chapter Three.

tains the analysis Of crop enterprise budgets for the four production
An

systems, both from the financial and the economic point Of view.
overview Of the whole-farm analytical model is presented in Chapter Five

Ir1 Chapter Six the evaluation Of the basic model, the presentation Of
results, and the sensitivity analysis are given. And finally, in

Chapter Seven, policy implications of the findings are discussed.

CHAPTER TWO

RESEARCH APPROACH AND DATA COLLECTION METHODS

As noted in Section 2 in Chapter One, the farm level survey Of
agricultural production in representative bas-fonds was requested by
the EORD in order to complement the 1978-79 farm survey, which focused
primarily on dryland crop production. The growing interest Of the EORD
in promoting bas-fond rice production was motivated by the large amount
Of potentially highly productive land Of this type, only a small portion

Of which was currently being cultivated. The 1980-81 survey carried
out i n this study was intended to fill important gaps in the present

knowl edge Of the economics Of bas-fond production in the Eastern Region.

1. SCOPE OF THE STUDY

Primary emphasis was given to cropping activities on fields located
in the has-fonds areas, rice being the chief crop grown. In order to
understand the potential for increasing bas-fond production, it was also
necessary to devote attention to household fields outside the bas-fonds,
and to other important household activities such as those involving
Hvestock and Off-farm employment. However, detailed i nput/output data

Were not collected for these activities.

I"Put-output and sales data were collected for all fields in the

SamPle. Information was also obtained on the full-range Of income-earning

17

18

activities other than crop production. The role of non-cropping
activities in terms of labor time and earnings was ascertained in order

to assess the potential for expansion or/and intensification Of bas-

fond production .

2 . RESEARCH APPROACH

Three sources Of information on farmer circumstances were used:

1. Background information on the farmer's environment from
secondary sources (research institutes, government reports,
meetings, etc.).

2. Field surveys involving structured interviews with farmers
by enumerators.

3. Personal interviewing and Observations in farmers' fields
and their surroundings tO Obtain information on the agri-
cultural production processes.

The author arrived in Upper Volta on May 26, 1980. In-country ori-
entat‘i on consisted Of meetings in Ouagadougou with Officials from USAID,
Agri cultural Services in the Ministry Of Rural Development, SOVOLCOM,
the Department Of External Trade Of the Ministry Of Conmerce and the
UPPer Volta Rural Development Fund. The Objective Of these contacts
was to gather information on quantifiable parameters likely to affect
farmer- decision making such as price trends, production trends, and
import trends Of rice since 1961, and also to gather information on
"3121 onal rice marketing and price policy. This secondary information
“as considered useful for drawing policy implications from the analysis

Of Tam level data. Orientation in the EORD itself took place in the
fohn Of:

19

1. Meetings with the EORD personnel including the director and
the heads of the Bureaus Of Economic Analysis/Planning Agri—
cultural Production, and Rural Works.

2. Attendance at monthly 0RD sector chief meetings.

3. Extensive travel to major bas-fond production areas to
develop familiarity with their salient agronomic features
and to initiate contact with farmers with significant
bottom-land production in these zones.

The main Objective Of data collection was to provide infOrmation

on the inputs and outputs of the agricultural production processes at

work in the research area. The inputs include all labor, land, capital

and technology. Technology was understood here in the broad sense of

the timing and nature of Operations, as well as the implements used.
Data on the outputs include total yield of the major crops grown as

well as the disposition Of these crops. An effort was also made to

ascertain the role Of livestock and non-farm activities in terms Of

labor time, other costs and earnings in order to assess the potential
for expansion or intensification of the cropping enterprises studied.
In brief, the study was seeking to obtain all the data necessary for
cal cul ation Of the factor requirements for each output under a given

technology, the .returns to factors in different enterprises, and the

Opportunity cost of resources used.

3. THE FIELD RESEARCH SITES

This section deals with the Choice Of research sites, their char—

acteristics and population. From an administrative standpoint, the EORD

 

20

covers the whole Eastern Department with its headquarters being Fada

N'gourma. The EORD itself is divided into eight sectors, and the sectors

are in turn divided into 24 sub-sectors.
The major bas-fond areas in the EORD were visited by the author

accompanied by Dr. Wilcock, MSU field team leader and Mr. Lompo, Chief

Of Section, Ame'nagements Hydro-Agricoles, EORD, during the first two

weeks Of June, 1980. After interviewing senior 0RD personnel in Fada

N'gourma, the author consulted production maps and tables as well as
available literature, e.g., the reports by J. R. Weldring (1979) on
watershed management projects for agricultural purposes in the EORD and

the report by D. Wilcock and A. Mehretu (1979) on Planning Regional

Development for the EORD Of Upper Volta. The following areas were then

visited:
Zanre, Dianga, Tamosgo and Lantaogo in the DiabO sector;

Komboari, Panpangou I and Panpangou II in the Fada sector;

Sampieri, Sambalgou, Botou, Gangana, Koyenga and Boudieri
in the Kantchari sector; '
Tapoa, Mordeni, Tanla, LogObou, Kouakouli, Kalbouli,
Bomoandi and Dangou in the Diapaga sector.
These areas were visited in order to develop familiarity with the
Salient features of the has-fonds and the farmers involved in these

has-fonds. During the visits, the following information was collected:

- estimated number Of farmers involved in each has-fond;
- degree Of mastery Of water in each has-fond;
- method Of land preparation; and

- use Of manure and/or fertilizer.

 

21

At the end Of the reconnaissance survey, five basic criteria were
used to select the villages or areas to be studied. First the location
chosen had to present great potential in terms Of bas-fond development
and to be representative Of either traditional has-fonds or improved
bas-fonds. The basic objective was to select an area With Character-
istics such that rice activity reflects a choice between viable enter— Eh
prises (millet, sorghum and possibly some grain legumes or cotton) as

Opposed to an area where rice represents the only way to earn cash.

 

Second, the area chosen needed tO be in a region that was Of interest '-
for future policy intervention since the purpose of the exercise was to
be Of use in such planning. For this reason, areas for possible con-
sideration were confined to the areas where there were already extension
agents involved in has-fonds development. Third, in the location chosen,
there had to be at least 30 farmers engaged in rice growing activity in
the bas-fonds during the cropping season 1980/1981. Fourth, the research
site had to be accessible tO a fairly good road leading to a major con-
sumption center. This restriction was made to ensure that the farming
system studied had an outlet for food grains and cash crops. This re-
striction was also imposed in order to facilitate the supervision Of
the enumerators by the principal investigator. Fifth, farmers in the
research site chosen had to be willing to cooperate in order to aid in
the successful completion of the study.

On the basis of these criteria, the following areas were selected

as the research sites:

- Zanre in the DiabO sector;

22

- Komboari, Panpangou I and II in the Fada sector;
- Kalbouli and Kouakouli in the Diapaga sector; and
- Tapoa in the Diapaga sector.

(See Map 2.1 for the location Of these research sites.)

4. SELECTION OF SAMPLE FARMERS

The unit Of data collection adopted for this study was the "house-
hold," defined here as the aggregate of persons who normally reside
together (not necessarily under the same roof), eat together and have
common granaries. The members of a household are generally bound by
ties of kinship, although a household member may be non-relative as
well.

After the selection Of research sites, the first task was the
selection Of sample households. The major concern in household selec-
tion was to include households which would provide truthful and complete
information, as well as representative data. While random sampling
helps ensure obtaining the latter, the first two elements can never be
Ptaken for granted. SO a two-step approach was adapted for the selection
Of the households retained for the study.

First, a random sample Of 35 to 40 households was selected for each
Of the production systems to be studied. In the second stage, house-
holds were dropped from the final sample for any Of the following three
reasons:

- lack of cooperation;

- cases where the household was Obviously unrepresentative of

the population, e.g., village chiefs;

 

23

 
    

(Kal bouli - Kouakouli

Tapoa - Diapaga

 

Konboari - Panpangou

klanre - DiabO

 

Scalelep.49.¢9.0°.top-.

MAP 2.1 LOCATION OF RESEARCH SITES
(all in circles)

Source: BAEP, 0RD de L'est, Fada N'gourma

24

- cases where the information Obtained at the inventory stage
(household and field inventory) was Obviously unreliable.
SO, with this procedure, the final sample retained for the survey
consisted of:
26 households in Zanre - Diabo - System 1
30 households in Kalbouli - Kouakouli - System 2 c

30 households in Komboari - Panpangou - System 3

ISL-

30 households in Tapoa - System 4 i

116 total households

 

5. INTERVIEW FREQUENCY

Since it was anticipated that the most important input would be
labor, detailed data on labor utilization was collected. And since the
amount Of labor furnished by each worker may be highly variable over
the agricultural season and since the timing Of agricultural labor input
may be as important as the quantity, it was important to have some form
of data collection procedure Operating at or around the time the labor
is being performed.

The data collection approach usually followed in these cases is
the cost route survey which is that Of Spencer (1972), Norman (1973)
and others who used multiple visits in their production economics
research.

1

Other methods used for collecting micro-level data from farmers

include: (1) model or case farm study, (ii) farm account books, and

 

( ;For a comprehensive discussion of these methods, see Spencer
1972 .

25

(iii) the farm business survey. The model farm study and the farm
account book methods were ruled out in this study because in the first
method, farms studied are atypical and therefore cannot be used to re-
present a given system Of production; and the latter method was inap-
propriate due to the illiteracy Of farmers, Which means that they
cannot and do not keep records. The Options remaining for collecting
micro-level data were then the farm business and the cost route methods.
In the farm business approach, the enumerators visit the farmers once
or twice to complete the survey questionnaire. This approach usually
covers a large sample Of statistically selected farmers, hence minimiz-
ing sampling errors. But it was this author's view that this approach
could result in high Observational errors, mainly with respect to the
labor input profile. SO, because of the objectives Of this study and
the type of analyses planned, the cost route method was chosen.

Critics Of the cost route survey have raised the issue of its cost
effectiveness, and the issue of the sustained interest of the farmer
during repeated interviews. The cost route method has been widely used
in farm management and production economic studies carried out in Africa
and has been modified over the years. More precisely, the frequency
with which the farmers are formally interviewed depends on the
researcher's confidence in the ability Of farmers to remember the
required details Of their past Operations.

In an attempt to incorporate the best features of farm business

surveys and cost route surveys,our approach to data collection, later

 

26

on referred to as activity approach,z

employed less intensive methods Of
family labor data collection. We shifted from weekly or bi-weekly inter-
views Of sample farmers to interviewing them once for each of the major
field activities (land preparation, planting, weeding, etc.) on a plot

by plot basis. Since it was felt that sample farmers would be able to
remember the details, particularly Of labor input for periods longer

than one week, it was decided that they would be visited at weekly inter-
vals but that actual recording of labor input would only take place at
the end Of each major field activity. However, weekly visits without

any recording involved did help in terms of establishing a strong rapport
between the enumerator and the sample farmers while increasing the chance
for accurate information to be collected. These visits were also very
helpful in deepening the author's understanding Of the production pro-
cesses and Of farmers' goals and priorities.

For hired labor and for expenditures on and use Of purchased non-
labor inputs, monthly interviews were considered adequate. In contrast
to flow data, stock variables are susceptible to measurement by "one
shot" interview, at least in theory. Examples pertinent to this study
are land holdings, farm implements, and livestock ownership. Although
some Of this information may be sensitive, these data can be gathered
with accuracy in one or two interviews. In addition to flow and stock
questionnaires, an additional questionnaire was administered by the

author himself through personal interviewing. This questionnaire tried

 

2While the activity approach is promising, it may not reduce survey
costs unless there is detailed information available on the cropping
calendar Of sample farmers. But data processing costs are reduced and
the researchers can benefit from insights gained through informal
interviewing.

27

to get at attitudes, Objectives, the relative importance Of various
enterprises, and reasons underlying particular production practices.
One or two interviews were conducted per household on a sub-sample Of
farmers (five farmers per research zone). It was anticipated that this
questionnaire would provide an essential context for interpreting the
quantitative survey data and fill in gaps which the stock and flow
questionnaires dO not or cannot cover. Table 2.1 gives the details
about farm recording schemes used in this study.

During the reconnaissance survey, it was noted that if there were
two or three wives or an adult son in a household, each responsible for
a plot, there were then as many as five decision makers within the same
production unit: each wife as a maternal head Of her family and as
independent manager Of her plot, and the household head or the sample
farmer making overall "policy" decisions over the land in general, the
main sorghum/millet field, and some or all Of the individual plots.

The practical significance Of this for data collection was that reliable
data on labor use or cropping practices could only be Obtained from the
household member who actually did the work or who was in charge Of the
plot. SO in this study, efforts were made to interview directly the

household member responsible for the plot under consideration.

6. ENUMERATOR TRAINING AND RESEARCH CALENDAR

Enumerators were recruited from the set who previously served in
the 1978-79 farm survey carried out by the MSU team. The four enumera-
tors recruited for the bas-fond study participated in a four-day train-

ing course during July, 1980, consisting Of instruction in the purpose

28

 

 

Amoco m>_umuwpe=ov

 

323.5. 93 “B 2...... 85 8.533 mismmunzm ... :o 8.5 5.93::er 33:3 .2
_ . magnum; one .
ecospmm>cw .ucmeoewzamg sonny .xuoumm>s_ so on)» pace xuoumm>_4 .Np
common pmezupauwcmm mewcau movuw>ppoe peapoe
avocado; .mupocomso; opasem mo moonsoe Ppm com mono mo_uv>wpoa Eeeetcoz .FP
so>e=m to new age p< ooze newsmesmeoe eoc< cop
mmmFFP> swag» :_ msomm.~_>
s5 umpuemcogu mam»? coo» comes so mpwepou our»; mxeu mptmp sgo>u cowpeELomcw mu_ca .m
:o_uuem:eeu sum>m so ospm> use wasp emugoumm apnoea: museums mmpmm .w
meme; uo—a an pope a co couscous oxeuueeo Pouch mzac «pun zgo>m mveoooe uma>cmz .5
so>g=m so use ecu mcpccwmmm muss» scouco>=w acmeopaew some .m
pope an pope mpgpcoz mm: page? Lonmpucoz .m
be: .3 so: :58: .252 323-52 .e
memes uo—a an papa a co emeeouom auv>_uum upmpe sgo>o so; gonmp s__5em .m
e—ogomso; oPQEam some as spasms comm usage mppmumo mono mamcmu upogmmsoz .N
so>gam so new new mcpccvmmn as» a< oupeh agou=m>cs uo—a ._
mxcasom sucoacoee mama so masw

 

 

—mnoma— .omOm mxh 2H <h<o umzozoumtomo<

wzmhouoooo mom cum: muoaomxum m1» no m4~<hmo
—.N 39?

29

Of the study and its objectives, definition Of survey terms, interviewing
techniques, how to approach the farmer, delineation Of enumerator's re-
sponsibility, and practice using the different survey forms. At the end
of the training, one enumerator was assigned to each research zone. The
number of sample farmers per research zone varied between 26 and 30.

Actual data collection began in July, 1980, with the collection Of
plot and household inventory data. Resource utilization was obtained
according to the schedule indicated in Table 2.1. The enumerators inter-
viewed farmers from July, 1980, through January, 1981. Towards the
middle of the harvesting period, a mobile team Of two BAEP staff was
sent to the research sites to measure sample farmers' fields in collabo-
ration with the enumerator already established in the area.

Because of the importance Of close supervision throughout the crop-
ping season, during the months Of July-November, a major activity of the
author was to visit the enumerators as Often as possible. Because Of
the dispersion Of the research sites, great distances and poor roads,
this proved to be a time-consuming and Often arduous task. 0n the
average, each enumerator was visited on site every 12 days for the first
three months (July-September), and every 18 days for the last four months
(Octobeeranuary). The main activities during the author's supervision
tours were to:

- ensure that enumerators were on schedule; ,

- check over completed forms;

- deliver instructions and explanations, and provide guidance;

- ensure a prompt and coordinated delivery Of materials;

- help the enumerators in organizing a work schedule; and

- conduct personal interviewing.

30

In general, individual performance for all the four enumeratOrs during
the survey was very satisfactory. Perhaps the single most important
factor in successful data collection is staff morale. The author spared
no effort in this respect and feels strongly that the investment was

justified in terms Of the results.

7. DATA PREPARATION AND ANALYSIS

Organizing and editing completed survey fOrms started as early as
in August, 1980, and ran through the month of January, 1981. For each
Of the 116 sample farmers surveyed, approximately 2,200 completed survey
fOrms had to be sorted, organized and accounted for at the end Of the
cropping season. Completed survey forms were checked, noting data errors,
inconsistencies, and illegible or incomplete responses.

Since the questionnaires used in this survey were not precoded, this
necessitated setting up an adequate coding system as each type Of survey
form was being completed in the field. Two options with respect to data
processing were available to the author:

- use of local computer facilities (Centre National de Traitement

de l'Information [CENATRIN]); or

- use Of Michigan State University (MSU) computer facilities.

The lack of a pool of clerks combined with the fact that CENATRIN was
over-committed and therefore not dependable ruled out the first option.
FurthermOre, the fact that during the processing of the 1978-79 MSU
Survey data, it was necessary to switch from CENATRIN tO the MSU computer
facilities with further "conversion" problems and attendant delays was

not a precedent to be followed.

31

With the second Option being the only feasible one, all the survey
forms were transferred to MSU where all the coding, keypunching and file
creation took place. Files were "Cleaned" by a combination Of computer
validation routines, rejection Of sample Observations and judgments by
the author to remove Obvious contradictions.

Data analysis began with calculation Of means, variance, frequency
distributions and modes for the major parameters. This provided a quick
impression of variability and extreme values.

The next step was to prepare a set Of crop enterprise budgets,
which involved computing the following input, output, and productivity
measures:

- average yields Of major crops produced per hectare (kg/ha);

- per hectare labor requirement per activity for major crops;

- per hectare farm budgets for each major enterprise or group

Of enterprises containing the following items:
a) technical data
(1) .yield in kilograms
(2) kilogram seed rate
(3) fertilizer used
(a) percent farmers using
(b) application rate
(4) total labor use
(5) average hourly wage for hired labor
(6) price per kilogram of major crops
b) costs and returns
(1) value Of output

(2) variable costs

32

(a) seed
(b) fertilizer
(c) non-family labor
(3) gross margin
(4) depreciation on farm implements
(5) net margin to family labor and management
(6) net margin per hour Of family labor

A comparison was made Of the performances Of the crop enterprises,
both across the fOur systems Of production being studied, and also with
respect to data from IRAT and AVV. This analysis Of the relative pro-
fitability Of the crop enterprises in the different systems was a nec-
essary first step in identifying the major production constraints and
possibilities fOr improved productivity.

The final step in the data analysis consisted of looking at the
implications of re-allocating farmers' resources using the linear pro-
gramming (LP) technique. LP allows all the constraints and revenue con-
siderations affecting agricultural production in the EORD to be
incorporated in a simultaneous framework. .The underlying hypothesis
is that the optimizing framework Of LP can help explain why farmers
typically engage in some activities more than others, given the resource
constraints they face, the particular desires they may have concerning
production (such as on-farm self-sufficiency in sorghum/millet), and

the desire to make the most Of what they have.

8. DATA LIMITATIONS

In a study of the economics of rice production in the bas-fonds Of

the EORD, many factors are likely to affect the quality Of data collected

33

and thus the inferences that can be drawn from such data. Among these
factors are:

the cross-sectional analysis problem;

the representativeness Of the systems studied;

the reliability of farmers' responses;

the problem Of estimating labor time; and

the problem Of estimating quantities Of other outputs and inputs.
8.1. Cross-Sectional Analysis Problem

Because only one season's data was used in the analysis, the find-
ings must be interpreted accordingly, taking into consideration specific
conditions (e.g., rainfall) that prevailed during the survey period.
Coupled with this problem is the fact that the study looked at the use
Of has-fonds only during the rainy season. A longitudinal analysis where
data are collected from the same Observational unit at successive points
in time would have been desirable to enable us to take into consideration
changes over time in the farmers' environment. However, this was not
possible in this study.

8.2. Representativeness Of the Rice
Production Systems Studied

NO statistical procedures were employed to choose the systems
studied. Since our main Objective was not to estimate response func-
tions, but rather to illustrate the costs and returns Of alternative
techniques of producing rice currently existing in the EORD, a purpo-

sive choice of the systems to study was appropriate.

34

8.3. Reliability Of Sample Farmers' Responses

Despite the fact that the Objectives Of the study were clearly
explained to the farmers and that at the onset Of the study they all
agreed to cooperate throughout the study, from time to time certain
farmers were reluctant to give out the information requested. Their
main allegations were that "prior investigations have yielded nO visible
fruit." When this problem occurred in an area, the enumerator in that
area would report the case to the author. In all the four cases re—
ported, the author, with some persuasion, was able to re-convince the
farmers Of the value Of their continued cooperation. Frequent informal
visits tO sample farmers involving no recording coupled with personal
interviewing by the author were believed to be very helpful in this

process .

8.4. Problem of Estimating Labor Time

TO measure field labor use, data were collected separately for
family, hired, and social labor on the basis Of age-sex category (men,
women, children) and activity. Field hours represent the time actually
spent by the farmer on the field, not including travel time to and from
the field. Because it was not possible to estimate the time the farmer
spent in eating or resting when the sun gets very hot or the time gone
from a farm operation briefly to set a trap for animals or for other
reasons, the time stated as spent on farm work tends to be overestimated.
For non-family labor, information on total expenditures, non-wage pay-
ments and the wage rates (case Of hired labor only) were Obtained in

addition to field hours.

35

Work rates may be affected by the age and sex Of the worker as well
as by the task being performed. Some researchers have used complicated
weighting procedures for aggregating the different categories Of labor
(Spencer, 1972). It is doubtful whether such elaborate weighting pro-
.cedures are necessary for the following reasons:

.- dividing the sample into different age classes may involve
substantial errors; the ages Often have to be estimated
because farmers cannot tell their ages in years with pre-
cision;

- from survey results and as Spencer (1972) pointed out, in

many parts Of Africa, women and children rarely participate
in activities in which they are less efficient than adult
males. Where they are commonly employed (seeding, weeding,
harvesting), women and children are as efficient as men.

Because Of the above reasons, coupled with my interactions with.
sample farmers regarding this issue of relative productivity Of differ-
ent categories Of labor, actual field-hours were used in the analysis,
i.e., the work productivity indices for all age-sex categories were
assumed to all be equal to one regardless of task performed.

8.5. Problem Of Estimating Quantities of Other
Inputs and Outputs

TO estimate farm size, 965 plots were measured towards the end Of

3

the harvesting period using tapes and field compasses. The quantity

of output was estimated by recording (generally on a weekly basis) the

 

3For description Of various techniques of field measurement, see
Collinson (1972).

36

number Of units Of product leaving the farm and their intended disposal.
Most Of the containers used to carry the output to the storage were
local measures, i.e., baskets, tins, bags, etc. Local measures were
also used to collect quantities Of inputs such as seeds and fertilizers.
For each research site, the amount represented by each local measure
was weighed for various products. For each type of container and for
each crop, six to eight measurements were taken and the average weight

in kg was used to convert all the quantities into kilograms.

9. SUMMARY

The purpose Of the field survey was to Obtain farm level data in
order to estimate the financial costs and returns of the major rice pro-
duction systems currently existing in the basefond areas Of the EORD.
The multiple-visit activity approach was used to collect input-output
data on the major crops from 116 farmers during the 1980-81 crop season.
Ihformation was also Obtained on the full range Of income-earning acti-
vities other than crop production, using a single-visit survey technique.

Enumerators were recruited from those who previously served in the
1978-79 farm survey carried out by the MSU research team in the EORD.
The four enumerators recruited participated in a four-day training
course consisting Of instruction in the purpose Of the study and its
Objectives, definition Of survey terms, interviewing techniques, how to
aPProach farmers, and practice of using the different survey forms.
Enumerators were required to visit farmers weekly but tO collect field
records only at the end of each major field activity. .Field data were

periodically cross—checked for completeness and consistency.

37

Labor utilization data were collected on an activity by activity
basis, for both family and non-family labor. Data were recorded in
terms Of field hours. For non-family labor (hired and social labor),
infOrmation regarding wage rates, total labor expenditures and the
estimated value Of payments in kind was also collected. Total produc-
tion estimates were determined by counting on a weekly basis the number
of units Of product leaving the field and their intended disposal.
Quantities in local measures were converted to kilograms. TO estimate
farm size, 965 plots were measured toward the end Of the harvesting

period using tapes and field compasses.

CHAPTER THREE

DESCRIPTIVE PROFILE OF THE AGRICULTURAL PRODUCTION SYSTEMS
IN THE EASTERN REGION OF UPPER VOLTA

The purpose of this chapter is to present the organization Of
agricultural production in the EORD and the problems that it faces, and
to lOOk at the profile Of sample farmers and rice production systems

studied.

1. CROPS GROWN AND DEGREE 0F MIXED CROPPING

A total Of 22 crops were grown in the EORD, sorghum and millet
accounting for about 31 percent of the total number Of fields under cul-
tivation in 1980-81, and rice for another 27 percent. Maize and peanuts
make up 27 percent (13 and 14 percent, respectively), while soybeans,
tubers, bambera nuts, and vegetables account for the remainder. In
terms Of acreage, sorghum and millet fields account for about 45 per-
cent Of the total area under cultivation and rice only accounts for
about 19 percent.

Rice was generally grown in pure stands (less than 1 percent Of
the total area under rice was mixed either with sorghum or millet around
the edges Of the bas-fonds). Sorghum and millet themselves were either
grown in pure stands (about 13 percent of total area under cultivation)

or mixed with cowpeas and Okra (32 percent of total area under

38

39

cultivation). Maize accounts for 7 percent of total area under
cultivation and is either grown in pure stands or mixed with some banana
trees and some vegetables.

Peanuts account for 6 percent Of total area under cultivation.
Peanuts are either grown in pure stands or mixed with tubers or sesame.
Soybeans, tubers and cowpeas grown in pure stands account for the re-
mainder Of total area under cultivation.

On the average, 57 percent Of the cultivated land was sole cropped.
The remaining land cultivated was primarily devoted to crop mixtures,
i.e., two or more crops grown on a given piece Of land at the same time.
It should be noted here that the different crops may be together for a
short or long time. For the purpose Of this thesis, any degree Of over-
lapping in terms Of time was considered to be mixed cropping. As
Norman (1974) pointed out, confusion exists in the literature concerning
the use of the terms mixed cropping, interplanting, intercropping, etc.
(For more clarification Of the terminology, see Figure 3.1.) The major
crop mixtures found in the EORD were: millet/sorghum, sorghum/cowpeas
or millet/cowpeas, millet/sorghum/cowpeas, millet/sorghum/cowpeas/Okra,
maize/sorghum, and maize/cowpeas. The mixtures consisting Of two crops
were the most common although occasionally as many as four crops were

fOund. See Table 3.1 for all details about crop mixtures.

2. SOME AGRICULTURAL CONSTRAINTS IN THE EASTERN REGION

Five major categories of agricultural constraints can be distin-
9Uished in the Eastern Region of Upper Volta. They are: (i) land tenure,
(ii) marketing and processing, (iii) water control, (iv) research and

technological problems and (v) extension problems.

4O

.eump . :MELOZ umULzom

.mowooom ococomevo mo Ppoz mo mopooom psom one so mopoowco> uoogoemwo oo spoon opsoo moowowc_eoo

moon» .czosm coco zoom on oo_goo eomco_ o oxoo see poem as use .Fpoecsoe ooo=o_oompo xppooomoom mo moose

oo oo:_ecoo apweommooo: no: moo mooso opom .omeooo eo .omp< .ocomoeo on see mcwooooocooow oco mopocopo

ueoocw noon to mucoEopo ocooxwe o o? czoeo moo moose woos so does» woos: open: mooco ago so m:_ooo_go>o

osowpm on see woos» mosoooco opowo_:e co opoooo c? .oposoxo com .opopmmoo moo meeooooo ago to moopooweo>
oco mooeoocwoeoo one» oooosooeooo omeooo so we om .cowooocomooo osoososom ovomw_oepm a we meghe

p.593 5
3 no.5 :92!
9:26 totem -

 

 

 

 

 

 

 

 

0 menu a 4
m echo - O «u .553 o _
< n20 - X .53. 1 . . . 4 . u . q q q . 2.5
23:08 30k
9:39.... 28 XI] LX
u: mono... c Xvi
9.2 .33....qu a s _ .1 id x
.5380 v.53 cos-5:38
.3 .8358 QED :35:—
ucssiuoucu 0 XI ilo 3
one»
mean—93 £9.50 9 0 Xi Lea 30559.5.
quou .uiosvcdw
sci—no.3 no.3. mm;— .oEEhuafin—N
93.395 233:2 Till-4 OIIIIIIIIOvT X

 

 

omzomhmozou maozwouozH mmcza mzmMPh<m ozumaomo eth zommzmzuo uzuh mzh mo zo~p<¢ommooz~
F.m mmooau

41

.oo_o on» go czocm ooeo cones mg» on ooooomeo no; omoeouoo; one Ppo .mooeo mo .o; Pouoo on» mowpopoopou :H

ouooooo oco oooooo mooopoco

.mcopco oco.moouoeou .eooooo .ogxo mooopocu

.ommp .oooo >o>e=m "ooeoom

m
N
p

 

 

 

 

 

co. m._ mooomucoz
ca co o.~ .Nmopoooomo>
cop m. mam»
~.m m.¢m m.m o>ommeo memosh one mooom
m.p ~.wm ¢.~_ mcooozom
cop e.~ mooozou masomoo
o. m.om m.m¢ ~.¢m moooocooeo mcpocw
o.PF ~.om 5.5m n.mm oNVoz
o. o.me o.me ~.e s.mm oo—~_z
m.~ o.mp ~.m¢ ~.¢m o.mmp Aways: oeo ommv Eogmcom
m. _.op P.mm o.~m more pomeou
moo: so once moss» o3» moose
opom ozogo pope on»
mogouxpz oogu m ouooz co ooeu mmopo coco
Foooh gone: so 982

CBOLG .o: Pouchm—O chugmm

 

 

Pwuowmp .mmmm<4o umzthz memo hzmmmmuuo
2H oz< moomo meow m< zzomo mo<hzmummm az< zzomw macaw mo mmm<homz

~.m m4m<h

42
2.1. Land Tenure

Shifting cultivation has probably been the dominant agricultural
pattern of Gourtmantche farmers for centuries. In this system of culti-
vation, a farmer will cultivate a piece Of land, and, after some time,
abandon it to fallow and move on tO clear a new piece Of land. Nonethe-
less, as Swanson (1975) pointed out, the low population density in the
EORD combined with this practice Of shifting cultivation should not
obscure the fact that personal and family land rights represent a sen-
sitive issue which is going to become a major problem area in the
development Of this region. For instance, some farmers today, even
though they have never cultivated bas-fonds land, are attempting to lay
claim tO newly improved bas-fonds by demonstrating that their cultivated
fields or fallow land lie adjacent to such bas-fonds and that this land
thus represents part of their property, thus raising disputes over
ownership Of bas-fond lands. However, it should be pointed out that
outside currently established villages, there are great portions Of yet

unclaimed arable land.

2.2. Marketing and Processing

Cereals are marketed by two parastatals and private traders. The
two parastatals are OFNACER (National Cereal Agency) and SOVOLCOM (Voltaic
Marketing Agency). OFNACER is the governmental agency responsible for
marketing all cereals (sorghum, maize and rice). It operates by direct

purchases and purchases through licensed buyers, to ensure that farmers

43

1 for cereals, and also through its retail

receive the Official prices
outlets which act as a buffer stock to make the Official consumer prices
effective at the retail level. SOVOLCOM, which is 97 percent government
owned, Operates under the umbrella of the Ministry Of Commerce and‘
Industry. It buys locally, imports, and sells a variety Of commodities
including agricultural goods throughout the country.

Even though OFNACER is responsible for the marketing Of all cereals,
it was not giving enough attention to rice, which is the main crop
grown in the bas-fonds in the EORD. In fact, in 1980, the regional
director of OFNACER for the Eastern Region suspended all the paddy buy-
ing activities from his program because Of inadequate storage and prO-
cessing facilities. Throughout the whole EORD, only two small processing
units (130-160 kg of paddy per hour) were in operation during the survey
period.' This scarcity Of processing units coupled with the fact that
batches of paddy were not homogenous resulted in a poor quality Of

milled rice (60 percent Of broken rice), making the imported rice pre-

ferred to domestic rice.

2.3. Water Control

Table 3.2 showed that rainfall data in the Eastern Region has a

skewed distribution mainly towards the end Of the growing season (see

 

1Prices Of agricultural products are set by the government. They
are determined in part by the government's policy to promote production.
For 1978/80, they were set as follows:

 

Product Price (CFA/kg)
White Sorghum ' 40
Red Sorghum 3O
Millet 4O
Paddy Rice 63
Groundnuts (unshelled) 37

Cowpeas 45

44

.owm— .mcoo .m.>o=auoo om\¢m .oz usoooom .o<u "muszom

 

 

 

o.o~ N.emo .eoee Peeeee
P.o o._ .o.o o.o o.o o.o one o._ eeeeeeeo
m.o m., o.o o.o e.o o.o m.o o.o Leasesez
o.m m.o~ m.. o.~ o.~ e.m ..m N.“ eeeeoeo
N.e_ m.eo_ m.o_ e.o~ e.- o._m e.om m.~e Leaseoeem
o.ep N.mo~ «.me m.oe e.Po o.om o.om m.am omeeee
o.~P «.Nep o.PN p.o~ e.~m o.m~ P.o~ ~.- ssee
P.o_ m.mm_ m.~_ m.e~ N.- e.m~ ~.m_ m.e~ eeee
~.~ o.oo 0.0m m.e_ e.op a... ~.o m.a se:
m.~ ~.om m.o e.m e.N o.~ ~.N e.o .seee
0., N.o m.~ 8.. m._ ~.o m.o o.o eeeez
~.o m.o _.o o.o o.o o.o o.o 5.0 seeeeeee
..o _.o o.o o.o o.o ..o o.o o.o seeeeee

 

mama xcpom coev

to Lease: eeeeese _M\om eo mN mm 5» FN ON eo o. mp 5» .P as ea o m as P meeeez
emeee>< s_eoeez

 

Amxoo goocopoo :PV moowcoa

 

 

 

 

omumnop .oommwm >m .<z¢=oozu<o<m 2H z: zH <h<o 44<mz~<m
~.m m4m<h .

45

monthly average column). This will have serious implications both for
the screening Of short cycle varieties and for the development Of
improved cultural practices. It also appears from Table 3.2 that, at
least as far as the cereals are concerned, the main constraint is not
insufficient level of rainfall, but it is the management Of available
water that is lacking. This was reinforced by the author's personal
Observations which revealed that some sample farmers suffered from both
drought and overflooding during the same cropping season.

Speaking about management Of available water, there are two main
types of improvement Of bas-fonds in the EORD, based on the topography
Of the area and the.availability Of a permanent river. The first type
of improvement consists Of building dikes to retain water longer on the
plots than otherwise. Figures 3.2 (a and b) illustrate this system Of
improvement.

The main problem Of this system Of improvement is that after heavy
rains, the wings Of the dikes are wiped out with the consequence that -
water does not stay on the plot long enough for the investment in
building dikes to be Worthwhile.

The second system of improvement consists of building a dam in an
attempt to fully master the control of water and irrigate the plots
downstream at will. Figure 3.3 below illustrates this second system.
The main problem with this system is the water losses along the main and
secondary canals coupled with the fact that plots are not protected from

overflooding by a protection dike.

46

 

\

s f 7 '
93/ VEO '0g .‘_3
/ \ V‘“‘\g 6'\ 7/\ \
4%)! -..“
/

FIGURE 3.2.a

SKETCH OF THE BAS-FOND IMPROVEMENT STRUCTURE,
TYPE 1a: OPENED STRUCTURE

Adapted from Ministry Of Rural Development, HER, "Annual report,"

1980.
i/ i.

... ...d__ ___‘\~ .

\

M \
/\

‘_‘ ‘\

Source:

  

M

 

 

protection dike 33

'3?
/</
' /
/’

 

 

FIGURE 3.2.b

SKETCH OF THE BAS-FOND IMPROVEMENT STRUCTURE,
TYPE 1b: SEMI-OPENED STRUCTURE

Adapted from Ministry Of Rural Development, HER, I'Annual report,“
1980.

Source:

47

I . . r " .
eserv
P ~— I .0] r
’ ~ “ \
""""‘ ---~---- \
...- _ .... ...-...
T" \

"“' overflooding,,

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

  

h" main canal

/'

protection dike‘
1’

 

 

 

FIGURE 3.3

SKETCH OF THE BAS-FOND IMPROVEMENT STRUCTURE,
CASE OF DAM IRRIGATION

Source: Adapted from Ministry of Rural Development, HER, "Annual report,"
1980.

48

2.4. Research and TechnologiCal Problems

The main agricultural research centers in Upper Volta are located
around Ouagadougou and BObO-Dioulasso. NO experiments are mounted in
the EORD. Because agricultural research is non-existent in the EORD,
there is currently no proven technological package available to the
extension services. The adaptation of different rice varieties to dif-
ferent types Of bas-fonds as well as the relevance Of mineral fertili-
zation is yet to be defined. Furthermore, in order to have a better
quality of rice, batches Of paddy should be as homogenous as possible.
This is only possible with stabilized varieties maintained through the
provision Of new seeds every year. This ideal situation does not exist
in the EORD. Most farmers grow on the average two or three varieties
of rice (C74, Gambiaka, Dourado, Sintane Diofor, 1R8) Of different sizes.
Average yields Of these varieties are low, around 0.8-1.5t/ha. The
existing seed multiplication units do not nearly satisfy the needs of
farmers, forcing them to rely on their Old stocks for the next cropping

season.

2.5. Extension Problems

In addition to the lack of technological improvements to be trans-
ferred to farmers as mentioned above (in section 2.4), two other prob-
lems cripple the extension service. They are: a) the level of training
of agents involved in bas-fonds improvement, and b) the turn-over Of
personnel. .

With respect to the level Of training of personnel, most bas-fond

extension agents have had no previous exposure to water management in

49

the has-fonds during their training in Matourkou. Moreover, they are
totally unfamiliar with most Of the rice growing activities.

With respect to the turn-over Of the personnel, nearly 50 percent
of the extension agents dealing with bas-fonds were new (i.e., less than
one year) at their post during the 1980-81 survey. Consequently, their
knowledge Of their area Of jurisdiction and of their farmers was very

poor.

3. PRODUCTION SYSTEMS STUDIED

3.1. Definition Of the Systems

On the basis Of the reconnaissance survey, four systems Of produc-
tion were delineated and representative areas were selected in consulta-
tion with the senior ORD personnel. Areas selected were:

1. Zanre in the Diabo sector which represents the system based

on traditional has-fonds. NO attempt is made in this system
to control water in the has-fonds (System 1).

2. Kalbouli - Kouakouli in the Diapaga sector. These areas are
representative Of the transitional stage between the system
based on traditional bas-fonds and the system based on
improved bas-fonds (System 2). Improvements here in an
attempt to control water are rudimentary and are all done by
farmers without any government intervention.

3. Komboari - Panpangou I and II, in the Fada sector which
represents the system based on improved bas-fonds.. The
improvement here consists Of the building Of dikes to re-

tain water longer on the plots than otherwise (System 3).

50

4. Tapoa in the Diapaga sector, which represents the system
based on bas-fonds located downstream Of a dam. The degree
of'water control is higher here than in any Of the previous
systems (System 4). Irrigation is done by gravity in this

system.

3.2. Descriptive Profile of Sample Farmers

In the survey area, two main categories of resources available to

sample farmers could be distinguished: land and labor.

3.2.1. Land

All cultivated land can be subdivided into two major categories:
dryland and low-lands (or bas-fonds). In the dryland category, we can‘
further distinguish plots around the house (46 percent Of total land
under cultivation) and bush fields (20 percent Of total cultivated land)
generally located beyond 2 km from the village. In the bas-fond
category, We can distinguish rice fields from fields under other crops
such as sorghum, vegetables and cassava. Tables 3.3 and 3.4 illustrate
the spatial organization Of fields on each research site or zone. 1

Tables 3.3 and 3.4 show that on the average, farms are made up Of
eight fields with an average Of .45 ha per field. One thing that can be
seen from these tables is the high degree Of fragmentation of fields;
the bigger the farm, the higher the number Of fields. For instance,
in System 3 where the average area Of farm is the highest, 6.32 ha, that
is where we find also the highest average number Of fields, 10.7.
Another striking thing to note is the relative importance Of rice (both

in terms Of number Of fields and in terms Of hectarage) in the cropping

51

TABLE 3.3

AVERAGE NUMBER OF FIELDS IN TOTAL AND IN THE
BAS-FONDS, PER HOUSEHOLD, 1980-81

 

 

 

 

 

 

 

 

 

 

 

Average Total Fields in the Bas-Fonds
S t Number Of

ys em Fields Per Average Percent of
Household Total Number Total Fields

1 8.6 2.7 32

2 8.1 4.2 52

3 10.7 1.8 17

4 6.3 2.1 34

Total 8.5 2.8 32

TABLE 3.4
AVERAGE AREA OF FARMS IN TOTAL AND IN
THE BAS-FONDS, 1980-81
Average Average Area Rice Fields
Total Of Fields 3

System Area (ha.) in the Bas- Average Total Percent of
Per Farm Fonds (ha.) Area (ha.) Total Area

1 3.55 .40 .40 ll

2 3.49 1.89 .71 20

3 6.32 .30 .22 3

4 1.86 .32 .32 17

 

Total 3.81 .74 .41 ll

 

52

system. In terms Of hectarage, rice represents on the average only 11
percent Of the total farm area, varying from 3 percent in System 3 to
31 percent in System 1. This relative importance of rice in the crop-
ping system will surely have to be reckoned with in any effort to expand
rice production in the EORD, i.e., rice is still a relatively minor crop

in the cropping system Of EORD farmers.

3.2.2. Labor

Fifty-three percent Of the plots surveyed were prepared using human
energy while on 32 percent of the fields, zero-tillage was the main
method of land preparation. Only 15 percent Of the fields were prepared
using animal traction. SO the crucial role Of human energy (i.e.,
labor) in the production system cannot be over-emphasized.

Three main categories Of labor were identified in the research
sites: family labor, hired labor, and social labor. Figure 3.4 depicts
the various categories Of labor found in the surveyed villages. Hired
'labor is generally labor which is engaged for cash payments, although
there may or may not also be a payment in kind such as meals, tobacco
or drinks. Social labor includes communal and exchange labor. Communal
labor is labor provided by the community (friends and/or relatives and/or
in-laws) without cash payment or without the expectation of receiving
labor in return. Exchange labor occurs when one or more farmers or
members of their household agree to work on someone else's farm with
the understanding that members Of the other farmer's household will
undertake an equivalent amount Of work on their farm without payment.

This category Of labor is Often referred to in the literature as rota-

tional group labor (Norman, 1975).

53

Total Labor

 

 

 

 

 

 

 

 

 

Non-Family Labor Family Labor
Hired Social Labor
Labor
Exchange Communal
Labor Labor
FIGURE 3.4

DIFFERENT CATEGORIES OF FARM LABOR
IN THE EORD, 1980-81

54

3.2.2.1. Size and Composition Of the Household

In peasant agriculture, the size of the household largely determines
the total area cultivated. Thus, any policy measure which aims at
expanding output through land expansion must give serious consideration
to the size Of the household, the basic unit of agricultural production
in traditiOnal agriculture.

The average size of the household in the research sites was about
eight members with the following composition: 24 percent adult male, ‘
26 percent adult female, 23 percent children and 27 percent infant.2
This relative proportion has implications for the family labor force
available since it depends upon the age at which children are expected
to help on the farm and/or in other productive activities and whether
small children and Old men are included or not. Similarly, a distinc-
tion should be made between the size of the household and the size Of
the family farm labor force because the number Of hours available per
person per period for farm work depends upon the number Of hours family
members do farm work and the extent Of Off-farm commitments such as
trading, weaving, agricultural processing, and school attendance. Thus,
the number Of hours available for farm work depends upon the availability

of Off-farm jobs and the attitudes towards education, leisure and income.

3.2.2.2. Average Age of Heads Of Households

The average age Of the heads of household varied from 36 years in
System 2 to 51 years in System 3 (see Table 3.5). The weighted average
age Of household heads across all the four production systems was 44

years.

 

2All children less than seven years Of age are considered infants
and do not belong to the family labor work force. Adults are defined as

those ranging from 15 years tO 65 years of age.
/

55

TABLE 3.5

AVERAGE AGE OF HOUSEHOLD HEADS
BY SYSTEM, 1980-81

 

 

 

 

System 1 Average Age (Years)
1 46
2 36
3 51
4 42
Total , 44

 

 

3.2.3 Importance Of Non-Farm Activities

Recently, many development agencies and national governments have
shown an increased interest in the role Of non-farm rural enterprises
in the overall rural development effort because of their employment and
income generation potential.' SO, even though rural non-farm enterprises
were not the main thrust Of this study, some information was Obtained in
the questionnaire concerning the full range Of income-earning activities
other than farm activities.

Survey data showed that on the average, 67 percent Of household
heads are involved in at least one non-farm activity, this percentage
varying from 58 percent in System 2 to 80 percent in System 4 as shown
in Table 3.6. In comparing and contrasting these percentages in the
different systems, it is interesting to note that farmers who are closer
to a major center (e.g., Fada, Diapaga, Namounou, Diapangou) were more
involved in non-farm activities than those located further away with

poor road infrastructure. Wherever non-farm activities exist, they

56

TABLE 3.6

RELATIVE PERCENTAGE OF HOUSEHOLD HEADS WITH AT LEAST
ONE NON-FARM ACTIVITY PER SYSTEM, 1980-81

 

 

 

 

System "3353.2: 33230323215322} 313th Tot:?r§§2t5§:tem
1 15 58
2 18 A 50
3 21 7O
4 24 30
Total 73 67

 

 

range from agricultural processing (dOlO making, shea butter extraction,
rice hulling, etc.), to weaving, pottery, blacksmithing, tailoring,

masonry, and repair shops.

4. SUMMARY

An overview of agricultural production in eastern Upper Volta shows
that a wide variety Of crops are grown, both for domestic consumption
(e.g., sorghum, millet, maize, groundnuts, etc.) and for export (e.g.,
cotton). Farmers generally are growing more than one crop, but on dif-
ferent plots. Fifty-seven percent Of total cultivated area was sole
cropped. There are numerous factors that affect agricultural develop-
ment in the EORD among which are: land tenure, marketing and processing
infrastructure, lack Of improved technology, poor water management, and
poor extension services. With regard to water, the main problem is not
insufficient level of rainfall, but rather the poor management Of avail-

able water.

57

Four production systems were defined based on the degree Of water
control. System 1 represents the traditional bas-fonds or unimproved
swamps. NO attempt is made in this system to control water in the bas-
fonds. System 2 is based on semi-traditional has-fonds; in this
system water control improvements are rudimentary (based on the use Of
dikes) and are all done by farmers without any government intervention.
In System 3, the improvement consists Of the building Of dikes to
retain water longer on the plots than otherwise. The topographic map-
ping and dike building are done with heavy government assistance.

System 4 represents the system based on irrigated bas-fonds; here the
degree Of water control is higher than in any Of the previous systems.
The dam structure here is built and managed by the government, and irri-
gation is done by gravity.

On the average, farms were made up of eight fields with an average
area Of .45 ha per field. Rice represented on the average only 11 per-
cent Of the total cultivated area, varying from 3 percent in System 3
to 31 percent in System 1. Three main categories Of labor were identi-
fied in the research sites: family labor, hired labor and social labor.
Finally, it was found that on the average, 67 percent of household heads

were involved in at least one non-farm activity.

CHAPTER FOUR

A FINANCIAL AND ECONOMIC ANALYSIS OF THE FOUR
RICE-BASED PRODUCTION SYSTEMS IN THE EORD

The purpose Of this chapter is two-fold: first, to estimate crop
enterprise budgets by system Of production or zone to provide the data
base fOr establishing the relative profitability Of rice versus sorghum
and millet, groundnuts, etc., and second, to appraise the alternative
rice production techniques existing in the EORD, both from the private

and economic point of view.

1. DISTINCTION BETWEEN FINANCIAL AND ECONOMIC ANALYSIS

In the financial analysis, the main Objective was to answer the
question whether a particular enterprise under a given system Of prO-
duction will pay its way in strict monetary terms (are returns greater
than monetary costs?). Towards this end, inputs were valued at the
average market prices that farmers paid for each type of resource,
e.g., seeds, fertilizers, labor, etc.

Output was valued at the average unit price received at harvest
period by farmers in each research zone. For each enterprise budget,
financial returns to land, family labor, operating capital and manage-
ment were computed. Other performance measures computed from the budget
data included: net returns per field hour Of family labor, costs Of

production, etc.

58

59

1 the Objective was to find out the social

In the economic analysis,
opportunity costs Of different enterprises under each system Of produc-
tion, i.e., to determine the relative enterprise and system profitabil-
ities from the national point of view. Among the fOur systems under
study, all resource prices except seeds and expenditures on labor,
contain subsidies. Subsidies were estimated and resource costs were
increased by the amount Of subsidy to Obtain adjusted prices to reflect
true social values. Output was valued at its estimated import parity
price. These prices essentially reflect the true economic costs Of
factor inputs involved in the production process and the true social
value Of output. Furthermore, the costs Of investments to control water

borne by the government were taken into consideration in the economic

analysis.

2. DEFINITION OF CROP ENTERPRISES AND PROCEDURES
USED TO SELECT CROP MIXTURES

Enterprises were defined on the basis Of distinct products or groups
Of jointly produced products. Thus crops like rice, maize, soybeans,
etc., were considered as single enterprises. Also, several systems Of
rice production (swamp rice, irrigated rice) were considered distinct
enterprises because Of the unique conditions under which each system
operates and because Of the importance Of rice in the population studied.

A budget for an enterprise in this study was calculated from house- I

holds in which that enterprise was considered important, i.e., households

in which at least 10 percent of all area under cultivation was absorbed

 

1Economic analysis will be done only for rice enterprises across
the four systems under study.

60

by the enterprise under consideration, or households where that
enterprise was an important part of family consumption (e.g., Okra,
groundnuts and maize), or households where that enterprise was an
important source Of cash income. This chapter presents crop enter-
prise budgets by system for rice, sorghum/millet/cowpeas (S/M/C),
maize, groundnuts, bambera nuts, soybeans, cotton and Okra. Few Of
these enterprises are commonly based upon inter-cropping and inter-

planting2

though a field or plot can Often consist Of a mosaic Of
sole-cropped sub-suplots. True inter-cropping mixtures in the survey
zones do consist Of combinations Of sorghum plus millet plus cowpeas
and groundnuts plus bambera nuts. Although systematic calculations Of
field size were made, we did not measure the area planted to each crop
'type or the plant density in a given crop mixture mainly because Of
the prevalence Of complex patterns Of crop inter-cropping or inter-
planting. In summary, it was ascertained that S/M/C and groundnuts/
bambera nut mixtures typically require a given bundle of production
inputs and yield an output Of Xi kilograms Of sorghum, millet, cow-
peas, or groundnuts and bambera nuts.

The identification Of the most important crop mixtures for enter-
prise budget purposes proved tO be a complex process. Following Matlon
(1977) and Cranord (1980), data for each field were first checked for
consistency by comparing the mixtures implied by the planting and har-
vest data with mixtures reported by farmers in a separate interview

(plot use inventory'survey form). We also checked whether crops reported

as planted had been harvested and vice versa. Second, for each research

 

2For more details on inter-cropping and inter-planting, see
Igbozurike, 1971.

61

site, fields belonging to the same mixture type were then aggregated
despite the superficial differences and field variability. However, it
should be noted that very simple data manipulations were performed to
aid the process Of narrowing down the number Of mixtures. Two mixtures
were eventually selected on the basis of their importance (in terms Of
total number Of fields and total cultivated area) in the local farming
system. Mixtures containing rice and sorghum, maize and sorghum, maize
and Okra and groundnuts and cowpeas were eliminated. Such mixtures
cover only a small number Of fields and very little cultivated area.
Vegetable sub-plots such as Okra were considered as a separate enter-
prise rather than as a part Of other mixtures because household demand
for Okra was assumed here to be essentially a fixed function Of house-

hold composition rather than a function Of profitability considerations.

3. CROP ENTERPRISE BUDGETS

The aim Of input-output budget analysis here is to derive farm
recommendations which are consistent with the farmer's desires to
increase expected income and to make the best possible use Of the
resources available to him. Furthermore, enterprise budgets are
important in farm income analysis because they help to explain the
internal structure Of the farm as a whole and to show the relative
contribution of each enterprise to the whole organization. So, these
enterprise studies are very instrumental in an attempt to (i) assess
the profitability Of each enterprise relative to the resources used,

(ii) compare relative efficiency of various enterprises on the farm and

(iii) provide a basis for making rational decisions about the kind and

size Of enterprise to be expanded.

62

The budget data are presented here by rice-based production system
for all the major enterprises. For each enterprise budget, three main
parts can be distinguished:

l. Input use: This section includes non-labor and labor input
use as well as costs attached to their use and some agronomic data.
Costs were classified into two categories: variable costs (seeds,
fertilizer, hired labor, non-wage paymentsB) and fixed costs (fixed
costs here only refers to the depreciation of tools and equipment; no
animal depreciation or appreciation was included). Furthermore, no
attempt was made in this study to value land, the reason being that land
for most parts is communally owned in the EORD. Because of this com-
munal system of land tenure, there is no market price for agricultural
land. Token fees rather than real economic rents are sometimes paid by
farmers for use of land but.these are minor in most cases and paid by
a small proportion of farmers. Return to land and management was
therefore considered as a residual. Depreciation on tools and equip-
ment was cOmputed using the straight line method and assuming zero
salvage value and it was allocated to each crop enterprise proportionally
to the area covered by that enterprise.

2. Output: This section comprises the yields of crops included
in the enterprise as well as prices used to value production. Prices
used to value production represent average prices realized by all
farmers in the system at harvest period. However, it should be noted

that the value of total output as well as the expenditures on seeds are

 

3Non-wage payments referred to here are food and drink provided by
the household head to non-family laborers when they are performing work
on his fields.

63

seldom realized in cash since very few farmers sell their total output
or purchase their seeds.

3. Performance measures: This section looks at various perfor-
mance measures available to compare the relative efficiency of various
enterprises within a given production system and between production

systems.

3.l. System l: Production System Based on Farmers
Growing Rice in Traditional Bas-Fonds

In this production system the most important enterprises in terms
of objective of study, area cultivated, labor used, and income generated
were: rice, sorghum/millet/cowpeas (S/M/C), maize, groundnuts, bambera

nuts and soybeans.

3.1.1. Overview of the Enterprise Budgets in System 1

Costs and returns to the six major enterprises in System 1, derived
from the 1980/Bl survey are shown in Tables 4.l.l-4.l.6. The average
areas planted in rice, S/M/C, maize, groundnuts, bambera nuts and soy-
beans were in hectares, .411, .429, .079, .319, .476 and .252, respec-
tively. The mean labor utilization per hectare in all field activities
ranged from l06.6 hours for the bambera nut enterprise to 1,627.5 hours
for the S/M/C enterprise. In this system, Bl-lOO percent of the total
labor input was family labor and 0-19 percent was social labor. No
hired labor was used in this production system (Tables 4.l.l-4.l.6).
This is particularly important as we know that the use of non-family
labor is resorted to when family labor has become a constraint on produc-
tion of the subsistence farmer. As regards this system, the low depend-

ence on non-family labor gives us an idea of the poor employment

64

TABLE 4.1.1

AVERAGE COSTS AND RETURNS PER HECTARE FOR RICE
SYSTEM 1, IN THE EORD. 1980

 

 

 

CFA CFA
I. INPUT USE
A. Basic data
1. l of cases 64
2. Average size (ha.) .411
B. Non-labor input use
1. Seed rate (kg/ha) 23.6 @(83./CFA/kg) 1.961
Total 1.961
C. Agronomic data
1. Percentage of area ploughed using:
1.1 Animal traction 19.7
1.2 Hand tools 80.2
1.3 Zero tillage .1
2. Percentage of area fertilized:
2.1 Chemically 0
2.2 Organically 0
0. Labor input use
1. Pre-harvest activities
1.1 Family labor 194.1 hrs.
1.2 Social labor 14.0 hrs 0
1.3 Hired labor 0 hrs 0
1.4 Sub-total 208.1 hrs
2. Harvest activities
2.1 Family labor 51.1 hrs
2.2 Social labor 42.3 hrs 897
2.3 Hired labor 0 hrs 0
2.4 Sub-total 93.4 hrs 897
3. Total 301.5 hrs 897
E. Total variable costs 2.858
F. Tools and equipment (depreciation on) 850
B. Total costs 3,708
II. OUTPUT
A. Crop yields (kg/ha)
Rice paddy 458.3
8. Unit price (CFA/kg)
Rice paddy 60.6
C. Total value of output 27.773
III. PERFORMANCE MEASURES
A. Gross income 27.773
Less: Total variable costs 2.858
8. Gross margin 24,915
Less: Total fixed costs 850
C. Net margin 24.065
Less: Opportunity cost of equity capital 0 lzlmonth (1.961 x .01 x 8) 157
0. Net returns to land, FL (family labor) and management 23,908
Less: Opportunity costs of FL: (245 hrs 0 47 CFA/hr) 11,515
E. Net returns to land and management 12.393
F. Net returns per field-hour of family labor 23,908 + 245 97.6
6. Output - seed ratio 19.4
H. Costs of production (CFA/kg)(3,708 + 458.3) 8.1
I. Total costs of production (CFA/kg)(15,3ao + 458.3) 33.6

 

 

65

TABLE 4.1.2

AVERAGE COSTS AND RETURNS PER HECTARE FOR SORGHUM/MILLET/COHPEAS
SYSTEM 1. IN THE EORD, 1980

 

 

CFA CFA

 

II.

III.

INPUT USE

A. Basic data
1. I of cases , 88
2. Average size (ha) .429,

8. Non-labor input use
1. Seed rate (kg/ha)

l.l Sorghum 14.8 6 (84 CFA/kg) 1,243

1.2 Millet 20.4 8 (70.6 CFA/kg) 1,440

1.3 Cowpeas 12 5 8 (66.9 CFA/kg) 836

Total 3.519

C. Agronomic data ‘

1. Percentage of area ploughed using:
1.1 Animal traction 4
1.2 Hand tools 5
1.3 Zero tillage

2. Percentage of area fertilized:
2.1 Chemically 11
2.2 Organically 2.

0. Labor input use
1. Pre-harvest activities
1.1 Family labor 1,247.2 hrs.
1.2 Social labor 22.3 hrs 190
1.3 Hired labor 2.5 hrs 6 105 CFA/hr) 263
1.4 Sub-total l.272.0 hrs 453
2. Harvest activities
2.1 Family labor 266.3 hrs
2.2 Social labor 89.2 hrs 365
2.3 Hired labor 0 hrs 0
2.4 Sub-total 355.5 hrs 365
3. Total 1.627.5 hrs 820

E. Total variable costs 4.339
F. Tools and equipment (depreciation on) 1.065
G. Total costs 5.404
OUTPUT
A. Crop yields (kg/ha)

l. Sorghum 311

2. Millet 270

. 3. Cowpeas a]

8. Unit price (CFA/kg)

l. Sorghum/millet 48.5

2: Cowpeas 63.8
C. Total value of output 30,794
PERFORMANCE MEASURES

A. Cross income 30,794
Less: Total variable costs 4.339

8. Gross margins 26.455
Less: Total fixed costs 1.065

C. Net margin 25.390
Less: Opportunity cost of equity capital 0 lxlmonth [(3.519 x 3) t (453 X 3)] (~01) 295

0. Net returns to land, FL (family labor) and management 25,095
Less: Opportunity costs of FL: (1,513 hrs 8 47 CFA/hr) 71,111

Net returns to land and management -46.016
Net returns per field-hour of family labor 25.095 + 1.513 16.6
Output - seed ratios - 21.0. 13.2. 3.3

Im‘nm

Costs of production (CFA/ hg of grain) (5,404 o 622) 3,7
I. Total costs of production (CFA/kg of grain) (76.810 i 622) 123.5

 

 

6M5

TABLE 4.1.3

AVERAGE COSTS AND RETURNS PER HECTARE FOR HAIZE
SYSTEM 1, IN THE EORD, 1980

 

 

 

CFA CFA
I. INPUT USE
A. Basic data
1. i of cases 15
2. Average size (ha) .079
8. Non-labor input use
1. Seed rate (kg/ha) 23.7 8 (57 CFA/kg) 1,351
Total 1.3si
C. Agronomic data
1. Percentage of area ploughed using:
1.1 Animal traction 24.6
1.2 Hand tools 75.0
1.3 Zero tillage .4
2. Percentage of area fertilized:
2.1 Chemically O
2.2 Organically 72.0
0. Labor input use
1. Pre-harvest activities
1.1 Family labor 389 hrs
1.2 Social labor 13 hrs
1.3 Hired labor 0 hrs
1.4 Sub-total 402 hrs
2. Harvest activities
2.1 Family labor 149 hrs
2.2 Social labor 0 hrs
2.3 Hired labor 0 hrs
2.4 Sub-total 149 hrs
3. Total 551 hrs
E. Total variable costs 1,351
F. Tools and equipment (depreciation on) 382
G. Total costs 1,733
II. OUTPUT
A. Crop yields (kg/ha)
1. Shelled corn 1,194
8; Unit price (CFA/kg)
1. Shelled corn 35.2
C. Total value of output (CFA) 42,029
III. PERFORMANCE MEASURES
A. Cross income 42.029
Less: Total variable costs 1,351
8. Gross margin 40,678
Less: Total fixed costs 382
C. Net margin 40,296
Less: Opportunity cost of equity capital 8 lzlmonth (1,351 x 8 x .01) 108
0. Net returns to land, FL (family labor) and management 40,188
Less: Opportunity costs of FL: (540 hrs 8 47 CFA/hr) 25,380
E. Net returns to land and management 14,808
Net returns per field-hour of family labor (40,188 : 540) 74,4
6. Output - seed ratio 50.4
H. Costs of production (CFA/kg) (1,733 a 1,194) 1.4
I. Total costs of production (CFA/kg) (27,221 a 1.194) 22.8

 

 

67

TABLE 4.1.4

AVERAGE COSTS AND RETURNS PER HECTARE FOR GROUNDNUTS
SYSTEM 1. IN THE EORD, 1980

 

 

 

CFA CFA
I. INPUT USE
A. Basic data
1. 4 of cases 21
2. Average size (ha) .319
B. Non-labor input use
1. Seed rate (kg/ha) 16.4 8 (91.1 CFA/kg) 1,494
Total 1.494
C. Agronomic data
1. Percentage of area ploughed using:
1.1 Animal traction 43.3
1.2 Hand tools 56.1
1.3 Zero tillage .6
2. Percentage of area fertilized:
2.1 Chemically o
2.2 Organically O
0. Labor input use
1. Pre-harvest activities
1.1 Family labor 137.3 hrs
1.2 Social labor 0
1.3 Hired labor 0 hrs
1.4 Sub-total 137.3 hrs
2. Harvest activities
2.1 Family labor 56.1 hrs
2.2 Social labor 10.3 hrs
2.3 Hired labor 0 hrs
2.4 Sub-total 66.4 hrs
3. Total 203.7 hrs 0
E. Total variable costs 1.494
F. Tools and equipment (depreciation on) 217
G. Total costs 1,711
11. OUTPUT
A. Crop yields (kg/ha)
1. Shelled groundnuts 168
8. Unit price (CFA/kg)
1. Shelled groundnuts 54.2
C. Total value of output 9,106
III. PERFORMANCE MEASURES
A. Gross income 9,106
Less: Total variable costs 1,494
8. Gross margin 7,612
Less: Total fixed costs 217
C. Net margin 7,395
Less: Opportunity cost of equity capital 0 lt/month (1,494 x 8 x.Ol) 119
0. Net returns to L. FL (family labor) and management 7,276
Less: Opportunity costs of FL: (193 hrs 8 47 CFA/hr) 9.071
E. Net returns to land and management -1.795
F. Net returns per field-hour of family labor (7,276 + 193) 37.7
6. Output - seed ratio 10.2
H. Costs of production (CFA/kg) (1,711 a 168) 10.2
I. Total costs of production (CFA/kg) (10,901 a 168) 64.9

 

 

68

TABLE 4.1.5

AVERAGE COSTS AND RETURNS PER HECTARE FOR BAMBERA NUTS
SYSTEM 1, IN THE EORD, 1980

 

 

CFA CFA

 

I. INPUT USE
A. Basic data .
1. I of cases 9
2. Average size (ha .476
8. Non-labor input use
1. Seed rate (kg/ha) 5.4 O (125 CFA/kg) 675
Total 675
C. Agronomic data
1. Percentage of area ploughed using:
1.1 Animal traction 11.6
1.2 Hand tools 87.9
1.3 Zero tillage .5
2. Percentage of area fertilized:
2.1 Chemically o
2.2 Organically o
0. Labor input use
1. Pre-harvest activities
1.1 Family labor 79.8 hrs
1.2 Social labor 0 hrs
1.3 Hired labor 0 hrs
1.4 Sub-total 79.8 hrs
2. Harvest activities
2.1 Family labor 26.3 hrs
2.2 Social labor .5 hrs
2.3 Hired labor 0 hrs
2.4 Sub-total 26.8 hrs
3. Total 106.6 hrs .
E. Total variable costs 675
F. Tools and equipment(depreciation on) 139
G. Total costs 814
11. 01min
A. Crop yields (kg/ha)
1. Shelled pois de terre 32
8. Unit price (CFA/kg)
1. Shelled pois de terre 73.3
C. Total value of output 2,346
III. PERFORMANCE MEASURES
A. Cross income 2.346
Less: Total variable costs 675
8. Gross margin 1,671
Less: Total fixed costs 139
C. Net margin 1,532
Less: Opportunity cost of equity capital 0 lZ/month (675 x .01 x 8) 54
i
0. Net returns to land. FL (family labor) and management 1,478
Less: Opportunity costs of FL: (106 hrs 0 <47 CFA/hr) 4.982
E. Net returns to land and management -3,504
F. Net returns per field-hour of family labor (1,473 + 106) ‘3-9
6. Output - seed ratio 5.9
H. Costs of production (CFA/kg) (814 a 32) 25.4
182.8

I. Total costs of production (CFA/kg) ( 5,850 + 32)

 

 

69

TABLE 4.1.6

AVERAGE COSTS AND RETURNS PER HECTARE FOR SOYBEANS
SYSTEM 1. IN THE EORD. 1980

 

 

 

CFA CFA
1. INPUT USE
A. Basic data
1. l of cases 7
2. Average size (ha) .252
8. Non-labor input use
1. Seed rate (kg/ha) 37.5 0 (117 CFA/kg) 4.391
2. Fertilizer (18-35-0) rate (kg/ha) l6 2 8 (100 CFA/kg) 1.620
Total 6.011
C. Agronomic data
1. Percentage of area ploughed using:
1.1 Animal traction 100
1.2 Hand tools 0
1.3 Zero tillage O
2. Percentage of area fertilized:
2.1 Chemically 15.1
2.2 Organically O
0. Labor input use
1. Pre-harvest activities
1.1 Family labor 988.8 hrs
1.2 Social labor ' 0 hrs
1.3 Hired labor 0 hrs
1.4 Sub-total 988.8 hrs
2. Harvest activities
2.1 Family labor 133.5 hrs
2.2 Social labor 0 hrs
2.3 Hired labor 0 hrs
2.4 Sub-total 133.5 hrs
3. Total 1,122.3 hrs 0
E. Total variable costs I 6.011
F. Tools and equipment (depreciation on) 550
G. Total costs 6.561
11. OUTPUT
A. Crop yields (kg/ha)
Threshed soybeans 828
8. Unit price (CFA/kg)
Threshed soybeans ' 103.3
C. Total value of output 85,532
III. PERFORMANCE MEASURES
A. Cross income‘ 85,532
Less: Total variable costs 6,011
8. Gross margin 79,521
Less: Total fixed costs 550
C. Net margin 78.971
Less: Opportunity cost of equity capital 0 l%/month (6,011 x .01 x 8) 481
0. Net returns to L. FL (family labor) and management 78,490
Less: Opportunity costs of FL: (1.122 hrs 0 47 CFA/hr) 52,734
E. Net returns to land and management 25,756
F. Net returns per field-hour of family labor (78,490 + 1,122) 70.0
G. Output - seed ratio 22.1
H. Costs of production (CFA/kg) (6,561 + 828) 7.9
I. Total costs of production (CFA/kg) (59,776 + 828) 72.2

 

 

7O

Opportunities for the non-farm rural population which exist within this
system of production. This could have important implications for poli-
cies aimed at discouraging or encouraging rural-urban migrations or the
creation or promotion of non-farm rural enterprises.

The mean-expenditures per hectare for social labor in rice and
SIM/C enterprises where it is used were 897 CFA and 820 CFA, respectively.
For the rice enterprise, all expenditures were on harvest activities
whereas under the S/M/C enterprise, 55 percent of these expenditures
were spent in pre-harvest activities (Tables 4.1.1 and 4.1.2). Tables
4.1.1 and 4.1.2 further indicate that expenditures on social labor in
the form of drink and food account for 24 percent and 15 percent of total
farm expenditures. So in this system, the expenditures on social labor .
which is generally claimed to be unpaid labor, is quite appreciable.

The seed rates observed in this production system.were quite often
far from the recommended rates. The mean quantity of paddy rice seeds
used by farmers of this system was 23.6 kg/ha which is only about 30
percent of the recommended rate of 80 kg/ha. In the case of the ground-
nut enterprise, on the average, only 21 percent of the recommended seed
rate was used. Now, in the case of maize and soybeans, the average
seed rates observed (23.7 kg/ha and 37.5 kg/ha, respectively) were very
close to the recommended rates of 25 kg/ha and 40 kg/ha, respectively.

The yields in Tables 4.1.l-4.l.6 clearly reflect certain general
characteristics of the 1980/81 season, such as average rainfall and its
distribution. Crop yields were generally low, which is indicative on the
one hand of the low level of soil fertility in the EORD and on the other
hand of the poor crop varieties at the disposal of farmers coupled with

low seeding rate. Cereal yields in EORD in general are among the lowest

71

in the world (IRAT, 1979). This observation amply demonstrates that
one of the major problems facing this system of production is how to
increase cereal yields from the present levels of 300-500 kgs of grains
per hectare. It should be pointed here that low farm productivity in
this system is not the result of any single factor. It reflects a com-
bined effect of physical, technological, human and institutional factors.
3.1.2. Comparison and Appraisal of the Six Major Enterprises
in System 1

This section puts together the findings of the analysis of the six
major enterprises comprising this system of production. A summary of
general characteristics, costs and returns as well as performance
measures for all six enterprises and provided in Table 4.5.1. The-
discussion will mainly concern the analysis of the performance measures
so as to identify the enterprises with the highest financial return and

lowest cost of production.

3.1.2.1. Gross Margin (GM)

Among the six major enterprises of System 1, the variation in
gross margin ranged from 1,671 CFA/ha to 79,521 CFA/ha (Table 4.5.1).
The soybean enterprise had the highest gross margin and the groundnut
and bambera nut enterprises had the lowest gross margin. One thing
interesting to note though is that all enterprises were able to cover
their variable costs and therefore are all valid candidates to stay in

the farm business organization.

3.1.2.2. Net Margin (NM)
To compute the net margin, depreciation on tools and equipment was

deducted from the gross margin. Among the six major enterprises, the

72

TABLE 4.5.1

COMPARATIVE ANALYSIS OF THE MAJOR ENTERPRISES IN SYSTEM 1,
BASED ON SURVEY DATA FROM 26 HOUSEHOLDS. 1980

 

 

 

 

Enterprises
Criteria
Ground- Bambera
Rice S/M/C Maize nuts nuts Soybeans
I. General Characteristics
1. 4 of cases (fields) 64 15 21 9 7
2. Average size (ha) .411 .429 .079 .319 .476 .252
3. Average yield (kg/ha) 458.3 311/270/41 1,194 168 32 828
II. Financial Situation (CFA/ha)
1. Gross income 27.773 30,794 42,029 9,106 2.346 85,532
2. Variable costs 2.858 4,339 1,351 1.494 675 6,011
3. Total expenditures
(including depreciation
on tools 5 equipment) 3,708 5,404 1,733 1,711 814 6.561
4. Opportunity costs of
4.1 Family labor 11,515 71,111 25,380 9,071 4,982 52.734
4.2 Equity capital 157 295 108 119 54 481
5. Total costs 15,380 76,810 27,221 10,901 5,850 59,776
III. Performance Measures
1. Gross margin
(11.1 - 11.2) 24.915 26,455 40,678 7,612 1.671 79.521
2. Net margin
(11.1 - 11.3) 24.065 25.390 40.296 7.395 1.532 78.971
3. Net returns to land,
family labor a management
(CFA/ha) (111.2 - 11.4.2) 23,908 25,095 40,188 7,276 1.478 78.490
4. Net returns to land a
management (CFA/ha)
(111.3 - 11.4.1) 12.393 «46,016 14,808 ~1,795 -3,504 25,756
5. Net returns per hour of
family labor (CFA/phr)
(111.3 + total family
labor) 97.6 16.6 74.4 37.7 13.9 70.0
6. Total cost of production
(CFA/kg) 33.6 123.5 22.8 64.9 182.8 72.2
7. Output - seed ratio 19.4 21/13/3 50.4 10.2 5.9 22.1

 

 

73

variation in NM ranged from 1,532 CFA/ha to 78,971 CFA/ha. The ranking

of enterprises was the same as for the first performance criterion, GM.

3.1.2.3. Net Returns to Land, Family Labor and Management (NRLFLM)

To compute the NRLFLM, an opportunity cost was assigned to equity
capital. Normally, operating capital is treated as an input without any
opportunity costs in the accounting period. But since more operating I
inputs (e.g., seeds) are tied up for more than eight to ten months, they
effectively become capital expenditure items, which have an opportunity
cost.4

' The NRLFLM for the six enterprises ranged from 78,490 CFA/ha for

soybean enterprises to 1,478 CFA/ha for the bambera nut enterprise
(Table 4.5.1). In all enterprises the return per field hour of family
labor was less than the average wage rate paid to the hired labor,
105 CFA/hr. This result suggests that there may be financial gain in

5 or in urban areas where the SMIG

seeking employment on other farms
(minimum guaranteed wage in urban areas) is 90 CFA/hr. However, in the
case of rice enterprise, the returns of 97.6 CFA/hr is slightly above
the minimum wage rate paid to unskilled labor in urban areas. Thus,
there is no financial advantage of family members seeking wage employ-

ment in urban areas when they are needed on their rice fields.

 

4The private opportunity cost of equity capital is assumed to be
12 percent per year, which is the rate used by Tapsoba (1981). Equity
capital is made up of seeds and cash used to pay labor or buy food and
drink for workers during pre-harvest activities. Eight months and three
months were considered as relevant periods for the computation of oppor-
tunity cost of seeds and labor cash cost, respectively.

5However rural wage labor opportunities are so rare that few
farmers could consider wage labor as a viable alternative to farming.
It should also be kept in mind that farmers usually use family labor
beyond the point where hired labor is used.

74

The returns to family labor per hour give a rough indication of the
shifts farmers are likely to make if the present cost and return struc-
ture continues. The following shifts could be expected:

a) farmers will put more of their land and labor under rice,
maize and soybeans if the minimum sorghum needed for family consump-
tion is attained; and

b) low returns per person-hour of family labor in bambera nut
enterprise may force farmers to abandon this crop despite the critical
role it plays in the diet before the harvesting of other crops. Further-
more, no research is being currently done on this crop which is produc-

ing a very low yield (32 kg/ha).

3.1.2.4. Returns to Land and Management (RLM)
To compute the RLM, opportunity costs were assigned to family

labor;6

It was assumed that the internal opportunity cost of the family
labor was equal to the weighted average net returns per field hour of

family labor across all the enterprises in each production system. The
weighted average net returns per field hour of family labor was computed

using the following general formula:

 

6Some planners may argue that a farm family's labor has no social
"opportunity cost," but the members of that family probably have a dif-
ferent perspective. It is important that in the financial analysis, we
try to determine whether the farmer will want to participate in any rice
expansion scheme that we want to launch. This, of course, will depend
on how the farmer values his own time. ‘If we call the farmer's price
that he puts on his own labor his reservation price, this price will be
a bit different for each person depending on how rich or poor he is, on
what alternative work or other uses of time there are, and on how ener-
getic or lazy he is. Because farmers very rarely work for wages off the
farm, it was felt that the hired labor wage was not an appropriate indi-
cator of the opportunity cost of family labor. Instead, the best alter-
native to the use of family labor in any given enterprise is the returns
available from other on-farm production activities.

NR1 =§NR1jHij, i=1,2,3,4
' j = k, 2 ... 7
Where:
NRZ = weighted average net returns per field hour of

family labor in production system i

NR1.j = net returns per field hour of family labor for
enterprise j in system i

family labor used for enterprise j
weight = in system i
total family labor used across all
enterprises in system i

5:
II

For System 1,

N‘R‘ g (97.5)(245) + (16.6)(14513) + + (7o.9)(1.122)
1 245 + 1,513 + ... + 1,122

47.4 CFA/hr

Returns per hour of family labor found by Lassiter (1981,p. 31) average
only 39.4 CFA for hoe farmers and only 36 CFA for animal traction
farmers. But as he acknowledged, these returns vary greatly across
zones, primarily due to the high variability of yields. It is not
rare.to find in Lassiter'ssample, returns per hour as high as 156 CFA
(Lassiter, 1931337). Other reasons for this high variability in returns
per hour of family labor include labor market segmentation and labor
scarcity in some specific areas.

Using 47 CFA/hr as the opportunity cost of family labor, in System
1 three enterprises realized a negative return to land and management
(Table 4.5.1); they are: S/M/C, bambera nuts and groundnuts. However,
it should be stressed here that negative RLM observed does not mean that
farmers are losing money on these crops but rather, it means that net
margin is not enough to yield a positive return to the fixed land and

management factors.

76

3.1.2.5. Costs of Production

Two types of costs of production were computed. The first cost of
production took into consideration only variable and fixed costs. Using
this type of cost of production, maize showed the lowest cost of pro-
duction (1.4 CFA/kg of shelled corn) and bambera nuts had the highest
cost of production (25.4 CFA/kg) (Tables 4.1.1. -4.1.6.). The second type
of cost of production computed was obtained by adding the opporunity
costs of equity capital and family labor to total expenditures and the
result was divided by the yield. Among the six enterprises, maize still
showed the lowest opportunity cost of production (22.8 CFA/kg) and
bambera nuts had the highest cost of production (182.8 CFA/kg). The
second highest total cost of production was fbund in S/M/C (123.5 CFA/kg),
probably due to the large quantity of family labor input per hectare. ’
Rice showed the second lowest total costs of production (33.6 CFA/kg of

paddy).

3.2. System 2: Production System Based on Farmers
Growing Rice on Semi-Traditional Bas-Fonds

In this production system the most important enterprises in terms
of area cultivated, labor used and income generated were: rice,
sorghum/millet/cowpeas (S/M/C), maize, groundnuts, soybeans, okra and

CO‘C‘COD .

3.2.1. Overview of the Enterprise Budgets in System 2

Costs and returns to the seven major enterprises in System 2,
derived from the 1980/81 survey are shown in Tables 4.2.1-- 4.2.7. The
average areas planted in rice, S/M/C, maize, groundnuts, soybeans, okra

and cotton were in hectares, .270, 1.574, .049, .070, .131, .011, and .277,

AVERAGE COSTS AND RETURNS PER HECTARE FOR RICE

77

TABLE 4.2.1

SYSTEM 2. IN THE EORD. 1980

 

 

CFA CFA

 

II.

III.

INPUT USE

A.

'E.

F.
G.

Basic data
1. i of cases
2. Average size (ha)

Non-labor input use
1. Seed rate (kg/ha)
2. Pesticides (kg/ha)

Total

Agronomic data
1. Percentage of area ploughed using:
1.1 Animal traction
1.2 Hand tools
1.3 Zero tillage
2. Percentage of area fertilized:
2.1 Chemically
2.2 Organically

Labor input use

1. Pre-harvest activities
1.1 Family labor
1.2 Social labor
1.3 Hired labor
1.4 Sub-total

2. Harvest activities
2.1 Family labor
2.2 Social labor
2.3 Hired labor
2.4 Sub-total

3. Total

Total variable costs
Tools and equipment (depreciation on)
Total costs

OUTPUT

A.

C.

Crop yields (kg/ha)
1. Rice paddy

Unit price (CFA/kg)
1. Rice paddy

Total value of output

PERFORMANCE MEASURES

A.

I‘D'flm

Gross income
Less: Total variable costs

Gross margin
Less: Total fixed costs

Net margin

Less: Opportunity cost of equity capital 0 lzlmonth [(5,044 x 8) + (68 x 3)] (.01)

76

.270

50.8
2.2

1,172

41.7

Net returns to Land, FL (family labor) and management
Less: Opportunity costs of FL (2,210 hrs 9 20 CFA/hr)

Net returns to land and management

90.8 CFA/kg)
196.1 CFA/kg)

hrs
hrs

hrs

hrs
hrs

hrs
hrs

Net returns per field-hour of family labor (41,381 a 2.210)

Output - seed ratio

Costs of production (CFA/kg) (7.085 t 1.172)

Total costs of production (CFA/kg) (

51,691 4 1,172)

4,613
431

5,044

1.049

1,049
1,117

6.161
924
7,085

48,872

48.872
6.161

42.711
924

41,787
406
41,381
44,200
-2,819
18.7
23.1
6.0
44.1

 

 

78

TABLE 4.2.2

AVERAGE COSTS AND RETURNS PER HECTARE FOR SORGHUM/MILLET/COHPEAS
SYSTEM 2. IN THE EORD, 1980

 

 

 

CFA -CFA
INPUT USE
A. Basic data
1. I of cases 40
2. Average size (ha) 1.574
8. Non-labor input use
1. Seed rate (kg/ha)
1.1 Sorghum/millet 15.1 8 (88.1 CFA/kg) 1,330
1.2 Cowpeas 1.4 8 (93 CFA/k ) 130
2. Pesticides (kg/ha) 1.0 8 (196.1 kg? 196
3. Total 1.656
C. Agronomic data
1. Percentage of area ploughed using:
1.1 Animal traction 4.1
1.2 Hand tools 2.5
1.3 Zero tillage 93.4
2. Percentage of area fertilized:
2.1 Chemically O
2.2 Organically 0
0. Labor input use
1. Pre-harvest activities
1.1 Family labor 452.3 hrs
1.2 Social labor 7.7 hrs 75
1.3 Hired labor 0 hrs 0
1.4 Sub-total 460 hrs 75
2. Harvest activities
2.1 Family labor 99 hrs
2.2 Social labor 23.5 hrs 262
2.3 Hired labor 0 hrs 0
2.4 Sub-total 122.5 hrs 262
3. Total 582.5 hrs 337
E. Total variable costs 1.993
F. Tools and equipment (depreciation on) 3,315
G. Total costs 5.309
OUTPUT
A. Crop yields (kg/ha)
l. Sorghum 579
2. Millet 13
3. Cowpeas 1
8. Unit price (CFA/kg)
1. Sorghum/millet 32.6
2. Cowpeas 61.7
C. Total value of output 19,361
PERFORMANCE MEASURES
A. Cross income 19,361
Less: Total variable costs 1,993
8. Gross margin 17.368
Less: Total fixed costs 3.316
C. Net margin . 14.052
Less: Opportunity cost of equity capital 9 lz/month [(1,656 x 8) + (75 x 3)] (.01) 135
0. Net returns to Land, FL (family labor) and management 13.917
Less: Opportunity costs of FL: (551 hrs 9 20 CFA/hr) 11,020
E. Net returns to land and management 2.897
F. Net returns per field-hour of family labor (13.917 4 551) 25.3
6. Output - seed ratio 38. 9. 1
H. Costs of production (CFA/kg of grain) (5.309 a 593) 8.9

1. Total costs of production (CFA/kg of grain) (16,464 4 593) 27.8

 

 

79

TABLE 4.2.3

AVERAGE COSTS AND RETURNS PER HECTARE FOR MAIZE
SYSTEM 2. IN THE EORD, 1980

 

 

 

CFA
1. INPUT USE
A. Basic data
1. a of cases 23
2. Average size (ha) .049
8. Non-labor input use
1. Seed rate (kg/ha) 12.8 9 (84.3 CFA/kg) 1,079
2. Pesticides (kg/ha) 6 9 (196.1 CFA/kg) 118
Total 1.197
C. Agronomic data
1. Percentage of area ploughed using:
1.1 Animal traction 11.4
1.2 Hand tools 87.7
1.3 Zero tillage .9
2. Percentage of area fertilized:
2.1 Chemically O
2.2 Organically 20.8
0. Labor input use
1. Preeharvest activities
1.1 Family labor 1,058 hrs
1.2 Social labor 22.6 hrs 228
1.3 Hired labor 0 hrs 0
1.4 Sub-total 1,080.6 hrs 228
2. Harvest activities
2.1 Family labor 82.9 hrs
2.2 Social labor 30.2 hrs 171
2.3 Hired labor 0 hrs 0
2.4 Sub-total 113.1 hrs 171
3. Total 1.193.7 hrs 399
E. Total variable costs 1.596
F. Tools and equipment (depreciation on) 158
G. Total costs 1,754
11. OUTPUT
A. Crop yields (kg/ha) '
1. Shelled corn 514
8. Unit price (CFA/kg) 29.1
1. Shelled corn
C. Total value of output 14,957
111. PERFORMANCE MEASURES
A. Cross income 14.957
Less: Total variable costs 1,596
8. Gross margin 13,361
Less: Total fixed costs 158
C. Net nmrgin 13.203
Less: Opportunity cost of equity capital 8 1%/month [(1,197 x 8) + (228 x 3)] (.01) 103
0. Net returns to Land, FL (family labor) and management 13.100
Less: Opportunity costs of FL: (1,141 hrs 8 2O CFA/hr) 22,820
E. Net returns to land and management «9.720
F. Net returns per field-hour of family labor (13,100 e 1,141) 11.5
6. Output - seed ratio 40.1
H. Costs of production (CFA/kg) (1.754 + 514) 3.4
Total costs of production (CFA/k9) (24,677 + 514) 48.0

 

 

80

TABLE 4.2.4

AVERAGE COSTS AND RETURNS PER HECTARE FOR GROUNDNUTS
SYSTEM 2. IN THE EORD. 1980

 

 

CFA CFA

 

II.

III.

INPUT USE

A.

Basic data
1. 4 of cases 33
2. Average size (ha) .070

Non-labor input use

1. Seed rate (kg/ha) 15.7 9 (74.1 CFA/kg)

Total

Agronomic
1. Percentage of area ploughed using:
1.1 Animal traction
1.2 Hand tools
1.3 Zero tillage
2. Percentage of area fertilized:
2.1 Chemically
2.2 Organically

‘V

DO mum
010301

Labor input use
1. Pro-harvest activities
1.1 Family labor 504.3 hrs
1.2 Social labor 0 hrs
1.3 Hired labor 0 hrs
1.4 Sub-total 504.3 hrs
2. Harvest activities
2.1 Family labor - 276.5 hrs
2.2 Social labor 31.7 hrs
2.3 Hired 1aobr 0 hrs
2.4 Sub-total 308.2 hrs
3. Total 812.5 hrs

Total variable costs

Tools and equipment (depreciation on)

Total costs

OUTPUT

C.

Crop yields (kg/ha)
1. Shelled groundnuts 215

Unit price (CFA/kg)
l. Shelled groundnuts - 40.4

Total value of output

PERFORMANCE MEASURES

A.

Gross income
Less: Total variable costs

Gross margin
Less: Total fixed costs

Net "141910
Less: Opportunity cost of equity capital O 1%.month (1,163 x 8 x .01)

Net returns to Land, FL (family labor) and management
Less: Opportunity costs of FL: (781 hrs 8 20 CFA/hr)

Net returns to land and management

Net returns per field-hour of family labor (4.283 t 781)
Output - seed ratio

Costs of production (CFA/kg) (4,310 i 215)

Total costs of production (CFA/kg) (20,023 a 215)

1.163
1,163

COO

3,043
0

3.043
3,043

4,206
104
4.310

8,686

8,686
4.206

4,480
104

4.376
93

4.283
15.620

-11,337
5.5
13.7
20.0
93.1

81

TABLE 4.2.5

AVERAGE COSTS AND RETURNS PER HECT
SYSTEM 2. IN THE EORD.

ARE FOR SOYBEANS
1980

 

 

CFA

CFA

 

II.

III.

INPUT USE

A.

Basic data
1. 4 of cases
2. Average size (ha)

Non-labor input use
1. Seed rate (kg/ha)

Total

Agronomic data
1. Percentage of area ploughed using:
1.1 Animal traction
1.2 Hand tools
1.3 Zero tillage
2. Percentage of area fertilized:
2.1 Chemically
2.2 Organically

Labor input use

1. Pre-harvest activities
1.1 Family labor
1.2 Social labor
1.3 Hired labor
1.4 Sub-total

2. Harvest activities
2.1 Family labor
2.2 Social 1ab0r
2.3 Hired labor
2.4 Sub-total

3. Total

Total variable costs
Tools and equipment (depreciation on)
Total costs

OUTPUT

C.

Crop yields (kg/ha)
l. Soybeans

Unit price (CFA/kg)
l. Soybeans

Total value of output

PERFORMANCE MEASURES

A.

Gross income
Less: Total variable costs

Gross margin
Less: Total fixed costs

Net margin
Less: Opportunity cost of equity capital 0 lzlmonth

Net returns to land. FL (family labor) and managemen
Less: Opportunity costs of FL: (821 hrs 9 20 CFA/

Net returns to land and management

Net returns per field-hour of family labor (37,348 a
Output - seed ratio

Costs of production (CFA/kg) (5,728 a 362)

Total costs of production (CFA/kg) (22,512 a 362)

13
.131

58.8 9 (72.4 CFA/kg) 4.

O
10.4

627.1 hrs
29.4 hrs

0
656.5 hrs
194.1 hrs
90

284.1 hrs
940.6 hrs

362

120

43,440

[(4,257 x 8) + (765 x 3)] (.01)

t
hr)

821)

257

765

0
765
629
629

4.257

1,394
5,651

77
5.728

43.440
5.651

37.789
77
37.712
364

37.348
16.420

20.928
45.5
6.2
15.8
62.2

 

 

82

TABLE 4.2.6

AVERAGE COSTS AND RETURNS PER HECTARE FOR OKRA
SYSTEM 2, IN THE EORD. 1980

 

 

 

CFA CFA
1.. INPUT USE
A. Basic data
1. 4 of cases 23
2. Average size (ha) .011
B. Non-labor input use -
1. Seed rate (kg/ha) 69 8 (144 CFA/kg) 9.936
C. Agronomic data
1. Percentage of area ploughed using:
1.1 Animal traction O
1.2 Hand tools 74.5
1.3 Zero tillage 25.5
2. Percentage of area fertilized:
2.1 Chemically O
2.2 Organically 18.2
0. Labor input use
1. Pre-harvest activities
1.1 Family labor 666.7 hrs
1.2 Social labor 0 hrs
1.3 Hired labor 0 hrs
1.4 Sub-total 666.7 hrs 0
2. Harvest activities
2.1 Family labor 216.7 hrs
2.2 Social labor 0 hrs
2.3 Hired labor 0 hrs
2.4 Sub-total 216.7 hrs 0
3. Total 883.4 hrs 0
E. Total variable costs 9.936
F. Tools and equipment (depreciation on) 130
G. Total costs ' 10.066
11. OUTPUT
A. Crop yields (kg/ha)
1. Fresh okra 643
8. Unit price (CFA/kg)
1. Fresh okra 74.2
C. Total value of output 47,711
111. PERFORMANCE MEASURES
A. Gross income 47,711
Less: Total variable costs 9,936
8. Gross margin 37,775
Less: Total fixed costs 130
C. Net margin 37,645
Less: Opportunity cost of equity capital 9 lx/month (9,935 x 3 x .01) 795
0. Net returns to land FL (family labor) and management 36,850
Less: Opportunity costs of FL: (883 hrs 0 20 CFA/hr) 17,660
E. Net returns to land and management 19,190
Net returns per field-hour of family labor (36.850 4 883) 41.7
6. Output - seed ratio 9.3
H. Costs of production (CFA/kg) (10.066 1 643) 15.6
1. Total costs of production (CFA/kg) (23,521 a 643) 44.3

AVERAGE COSTS AND RETURNS PER HECTARE FOR COTTON

83

TABLE 4.2.7

SYSTEM 2. IN THE EORD. 1980

 

 

CFA CFA

 

II.

III.

INPUT USE

A. Basic data
1. 4 of cases
2. Average size (ha)

8. Non-labor input use
1. Seed rate (kg/ha)
2. Fertilizer (18-35-0) rate (kg/ha)

Total

C. Agronomic data

1. Percentage of area ploughed using:
1.1 Animal traction
1.2 Hand tools
1.3 Zero tillage

2. Percentage of area fertilized:
2.1 Chemically
2.2 Organically

0. Labor input use

1. Pre-harvest activities
1.1 Family labor
1.2 Social labor
1.3 Hired labor
1.4 Sub-total

2. Harvest activities
2.1 Family labor
2.2 Social labor
2.3 Hired labor
2.4 Sub-total

3. Total

E. Total variable costs

F. Tools and equipment (depreciation on)
C.‘ Total costs

OUTPUT

A. Crop yields (kg/ha)
1. Cotton

8. Unit price (CFA/kg)
1. Cotton

C. Total value of output
PERFORMANCE MEASURES

A. Gross income
Less: Total variable costs

8. Gross margin
Less: Total fixed costs

C. Net margin ,

00

746.4
13.1

759.5

-57.3
144.9

0
202.2
961.8

212

hrs
hrs

hrs

hrs
hrs

hrs
hrs

4,110
1.630

5,740

455
455

5.000

O
5.000
5.455

11.195

11.695

13.314

13.314
11.195

2,119
500

1.619

Less: Opportunity cost of equity capital 0 11/month [(5,740 x 8) + (455 x 3)] (.01) 473

0. Net returns to land FL(fami1y labor)
Less: Opportunity costs of FL: (804

Net returns to land and management
Net returns per field-hour of family

Output - seed ratio

zoo-rim

Costs of production (CFA/kg) (11,695

and management
hrs 0 20 CFA/hr)

labor (1.145 i 804)

4 212)

1. Total costs of production (CFA/kg) (28,248 a 212)

1,146
16.080

-143934

1.4
3.5
55.1
133.2

 

 

84

respectively. Except for the S/M/C enterprise, plots were much smaller
in System 2 than in System 1. The mean labor utilization per hectare in
all field activities ranged from 582.5 hours for the S/M/C enterprise to
2,278.0 hours for the rice enterprise. Compared to System 1, the labor
use in S/M/C enterprise is low because of the difference in land pre-
paration technique. Zero tillage was used on 93.4 percent of the S/M/C
fields. In this system of production, family labor input as a percent-
age of total labor input varied between 83 percent for the cotton enter-
prise and 100 percent for the okra enterprise. Social labor varied
between 0 percent and 17 percent. No hired labor was used in this pro-
duction system (Tables 4.2.1-4.2.7). The mean expenditure per hectare
for social labor where it was used in this system ranged from 337 CFA
under the S/M/C enterprise to 5,455 CFA for the cotton enterprise,
representing 6 to 47 percent of total expenditures.

The seed rates observed in this production system were quite far
from the recommended rates.7- The mean quantity of paddy rice seeds used
' in this system was 50.8 kg/ha which is only about 64 percent of the
recommended rate of 80 kg/ha. In the case of groundnut enterprise, on
the average, only about 20 percent of the recommended seed rate was
applied. Maize, millet and cowpeas suffer the same problem and seed
rates observed were only 51, 11 and 4 percent of the recommended rates,
respectively. However, in the cases of cotton and soybeans, seed rates

observed were 50 percent and 47 percent above the recommended rates.

 

7Rates referred to here are rates recommended by the research unit
of the AVV settlement schemes in the center-east region of Upper Volta.

85

Crop yields were generally low in this sytem, particularly for
groundnuts, soybeans, cotton and millet where they were only 215, 362,
212, and 13 kg/ha, respectively.

3.2.2. Comparison and Appraisal of the Seven Major
Enterprises in System 2 '

Following the scheme used in analyzing System 1, the discussion
here will mainly fOCUS on the performance measures so as to identify the
enterprises with the highest financial return and the lowest cost of pro-

duction.

3.2.2.1. Gross Margin (GM)

Among the seven major enterprises of System 2, the variation in
gross margin ranged from 2,119 CFA/ha to 42,711 CFA/ha (Table 4.5.2).
The rice enterprise had the highest gross margin and the groundnut and
cotton enterprises had the lowest gross margin; however, all enterprises

in this system were able to cover their variable costs.

3.2.2.2. Net Margin (NM)

Among the seven major enterprises the variation in NM ranged from
1,619 CFA/ha under the cotton enterprise to 41,787 CFA/ha under the rice
enterprise. 50, according to the first two performance measures con-
sidered, rice appears to provide the highest returns per hectare of all

crops in this system.

3.2.2.3. Net Returns to Land, Family Labor and Management (NRLFLM)

The NRLFLM for the seven enterprises ranged from 1,146 CFA/ha for
the cotton enterprise to 41,381 CFA/ha for the rice enterprise (Table
4.5.2).

£36

 

 

 

 

m.m m.m N.» ~.m_ ..o. _\a\om ..mN o_oae coon - psaoeo .A
~.mm_ m... ~.~o ..no o.o¢ e.AN ..e. on\<cov
=o.»u=cocq mo mumou .auo» .o
e.— A._e m.me m.m m.__ m.m~ 2.”, Asa _aooo . m.___v
.Lea\<au. Lona. »_.eec
uo sac: can mccsamc uoz .m
emm.e_- om..m_ m~m.o~ amn._p- o~h.a- Aao.~ o_a.~- A..e.__ - m.___v
A~=\<muv acosomacms
. a tea. on mcczumc um: .o
we... cmm.em mem.~m nm~.e oo_.m_ e.g.n_ .mn.F¢ A~.¢.__ - ~.___.
»o3\<muv «coaomueus a Loan—
p—Em» .tcu— o» mes—50.. aw: .n
a_e.. meo.an ~_~.Am can.e no~.n_ Nee... “as... .n.__ - _.__v
. 2:58 E?! 8.. .~
o._.~ maa.am omA.Am one.¢ ..on.m_ mo~.~. ..A.~. Am.__ - _.~_.
.ae\<aov e.oe.e “meta ..
mosamao: oucuELomcoo .__~
me~.m~ _~m.m~ ~pm.- nmo.e~ use.¢~ eo¢.o_ .ae._m named .eoo» .m
me. mos «on no no. mn— oce ..p.aeo so.=eu ~.o
oac.e_ coo.hp o~e.e_ ewe.m. cua.- cue... oo~.e. Leea— »_.eec ..e
ac mumou Auvcaucoaac .e
mam... oo°.o_ «Na.m o_m.e ems._ acn.m mac.“ .oeoea,=eo a m_ooo
:o copua—uucauc mcpuapue.v
mocaupucogxo _uuoh .n
ma.... omm.o .me.m oo~.. oom._ naa._ .e_.o nomoo o_o._co> .~
e_m.m_ ..A.~e o...me one.m smm.e_ pon.o_ Nem.m. ueoue. omega .—
..e\<au. eo.oeeo.m ...oeee.a .3.
~.~ nee Non m.~ ..m _\n_\o~m NA... Aoe\ex. e_u.s uneco>< .n
New. ..e. _m_. cue. ace. eem._ OAN. Rage on.“ oaetos< .~
4 MN m. mm m~ oe oh numeo co . .—
mueumvcouungogu _cco:ow .~
couaou ocxo mcuuazom mascvcaoca o~.~z u\z\m gov:
..eo».cu
mom—sncuucu

 

 

comp .moaozmmac: cm 2oz; (kc: >u>¢am 2o aum<a .N :uhm>m z— mum—xmzuhzu acacx wzh no m_m>a<z< u>~h<¢<mxou <

~.m.v u4m<p

87

In all the enterprises considered in this system, the return per
field hour of family labor was less than the minimum average wage rate
paid to the hired labor in the other systems studied (i.e., 105 CFA/hr)
and also less than the minimum wage paid to unskilled labor in urban
areas (i.e., 90 CFA/hr). This result suggests that there may be some
financial gain for family members in seeking employment on other farms
outside this system or in urban areas. Returns per field hour of family
labor in this system ranged from 1.4 CFA/hr in the case of cotton enter-
prise to 45.5 CFA/hr in the case of soybeans. As regards the rice
enterprise, the return per field hour of family labor was only 18.7
CFA/hr. If the present costs and returns structure continues, the fol-
lowing shifts could be expected in this system:

a) farmers will put more of their land and labor into soybeans,
okra, S/M/C and rice in that order; and

b) low returns per field hour of family labor in cotton enterprise
may force farmers to abandon this crop despite the government support

of this export crop.

3.2.2.4. Returns to Land and Management (RLM)

Four enterprises realized a negative return to land and management.
They are cotton, groundnuts, maize and rice. The RLM for the seven
enterprises ranged from -14,934 CFA/ha for the cotton enterprise to
20,928 CFA/ha for the soybean enterprise (Table 4.5.2). The high nega-
tive returns to land and management for cotton and groundnuts was
probably due to the low yields of these two enterprises. And because
of this low yield, the NM was not enough to yield a positive return to

land and management.

88

3.2.2.5. Cost of Production

Taking into consideration only variable and fixed costs, (direct
costs) maize showed the lowest cost of production (3.4 CFA/kg of shelled
corn) and cotton had the highest cost of production (55.1 CFA/kg)--
Tables 4.2.1-4.2.7. The second type of cost of production computed
-was obtained by adding the opportunity costs of equity capital and
family labor to the total expenditures and the result was divided by
the yield. Among the seven enterprises, S/M/C showed the lowest total
cost of production (27.8 CFA/kg) and cotton had the highest cost of
production (133.2 CFA/kg). The second highest total cost of production
was found in groundnuts (93.1 CFA/kg), probably due to the low yield of
groundnut in this system. Rice showed the second lowest total costs of

production (44.1 CFA/kg of paddy).

3.3. System 3: Production System Based on Farmers
Growing Rice in Improved Bas-Fonds

The most important enterprises in terms of the objective of study,
area cultivated, labor used and income generated were: rice, sorghum/
millet/cowpeas (S/M/C), maize, groundnuts/bambera nuts (GN/BN), and

soybeans.

3.3.1. Overview of the Enterprise Budgets in System 3

Costs and returns to the five major enterprises in System 3, derived
from the 1980/81 survey are shown in Tables 4.3.1-4.3.5. The average
areas planted in rice, SIM/C, maize, GN/BN, and soybeans were in hectares,
.488, .707, .431, .582 and .494, respectively. The mean labor utflization
per hectare in all field activities ranged from 175.6 hours for the

soybean enterprise to 1,435.8 hours for the S/M/C enterprise. In this

89

TABLE 4.3.1

AVERAGE COSTS AND RETURNS PER HECTARE FOR RICE
SYSTEM 3. IN THE EORD, 1980

 

 

CFA

CFA

 

II.

III.

INPUT USE

A.

E.
F.
G.

Basic data
1. 4 of cases 45
2. Average size (ha) .488

Non-labor input use
1. Seed rate (kg/ha) 38.0 9 (100.4 CFA/kg) 3,815

Total

Agronomic data
1. Percentage of area ploughed using:
1.1 Animal traction
1.2 Hand tools
1.3 Zero tillage
2. Percentage of area fertilized:
2.1 Chemically
2.2 Organically

UhN

0° DNA-d
N400

Labor input use

1. Pre-harvest activities
1.1 Family labor 267.6 hrs
1.2 Social labor 69.7
1.3 Hired labor 9
1.4 Sub-total 346.

2. Harvest activities
2.1 Family labor
2.2 Social labor
2.3 Hired labor
2.4 Sub-total

3. Total

hrs 450
hrs 9 125 CFA/hr) 1.200
hrs 1,650

N

6

9

.5 hrs

.8 hrs 231
hrs 0

3 hrs 231

2 hrs

0|“)

Total variable costs
Tools and equipment (depreciation on)

Total costs

OUTPUT ' '

A.

Crop yields (kg/ha) -
l. Paddy rice 501

Unit price (CFA/kg)' 65.7
Total value of output 32,916

PERFORMANCE MEASURES

A.

161-vim

Gross income
Less: Total variable costs

Gross margin
Less: Total fixed costs

Net margin
Less: Opportunity cost of equity capital 8 lZImonth [(3.815 x 8) + (1,650 x 3)] (.01)

Net returns to land, FL (family labor) and management
Less: Opportunity costs of FL: (518 hrs 9 55 CFA/hr)

Net returns to land and management

Net returns per field-hour of family labor (26,631 4 518)
Output - seed ratio

Costs of production (CFA/kg) (5,930 + 501)

Total costs of production (CFA/kg) (34,775 1 501)

3.815

1 .881
5,696

234’
5.930

32.916
5.696

27.220
234

26.986
355

26.631
28.490

-1,859
51.4
13.2
11.8

 

 

94)

TABLE 4.3.2

AVERAGE COSTS AND RETURNS PER HECTARE FOR SORGHUM/MILLET/COHPEAS
SYSTEM 3. IN THE EORD, 1980

 

 

CFA CFA

 

INPUT USE

A. Basic data
1. 4 of cases 56
2. Average size (ha) .707

8. Non-labor input use
1. Seed rate (kg/ha)
1.1 Sorghum 19.
1.2 Millet 22.
1.3 Cowpeas ' 7. 3 CFA/kg) 491
2. Pesticides (kg/ha) 96.1 CFA/kg) 176
3. Total 2.989

O (63.1 CFA/kg) 1,218
O (48.2 CFA/kg) 1,104
8 (6
O (1

C. Agronomic data

1. Percentage of area ploughed using:
1.1 Animal traction 8.1
1.2 Hand tools .8
1.3 Zero tillage 91.1

2. Percentage of area fertilized:
2.1 Chemically O
2.2 Organically O

0. Labor input use

1. Pre-harvest activities
1.1 Family labor
1.2 Social labor
1.3 Hired labor
1.4 Sub-total

2. Harvest activities
2.1 Family labor
2.2 Social labor
2.3 Hired labor
2.4 Sub-total

3. Total

...
O
-‘

hrs

hrs 1,383
hrs 8 145 (CFA/hr) 565
hrs 1.948

d

U
N —l
-‘ N 8“
U as» u
e e e I e
b .4009

hrs

a
5

415
hrs 0
hrs 415

hrs _ 2,363

-‘
fiN
U4.”
GIVO
0‘

E. Total variable costs 5.352
F. Tools and equipment (depreciation on) 661
G. Total costs 6.013
OUTPUT
A. Crop yields (kg/ha)

1. Sorghum 125

2. Millet 315

3. Cowpeas 15
8. Unit price (CFA/kg)

1. Sorghum/millet 42.6

2. Cowpeas 76.8
C. Total value of output 19,896
PERFORMANCE MEASURES

A. Cross income 19.896
Less: Total variable costs - 5,352

8. Gross margin 14.544
Less: Total fixed costs 661

C. Net margin 13.883
Less: Opportunity cost of equity capital 0 11/month [(2,989 x 8) + (1,948 x 3)] (.01) 298

0. Net returns to land, FL (family labor) and management 13,585
Less: Opportunity costs of FL: (1.347 hrs 9 55 CFA/hr) 74,085

E. Net returns to land and management -60.500
F. Net returns per field-hour Of family labor (13,585 . 1,347) 10.1
G. Output - seed ratio 5714/17'
H. Costs of production (CFA/kg of grain) (6,013 . 455) 13.2
1. Total costs of production (CFA/kg of grain) (80,396 . 455) 176.7

 

 

91

TABLE 4.3.3

AVERAGE COSTS AND RETURNS PER HECTARE FOR MAIZE
SYSTEM 3. IN THE EORD, 1980

 

 

 

CFA CFA
1. INPUT USE
A. Basic data
1. I of cases 39
2. Average size (ha) .431
8. Non-labor input use
1. Seed rate (kg/ha) 17.8 9 (55.4 CFA/k ) 986
2. Fertilizer (18-35-0) (kg/ha) 1.3 O (100 CFA/kg? 130
Total 1.116
C. Agronomic data
1. Percentage of area ploughed using:
1.1 Animal traction 15.5
1.2 Hand tools 19.6
1.3 Zero tillage 65.1
2. Percentage of area fertilized:
2.1 Chemically 69.0
2.2 Organically 83.1
0. Labor input use
1. Pro-harvest activities
1.1 Family labor 231.1 hrs
1.2 Social labor 0 hrs
1.3 Hired labor .1 hrs 0 106 (CFA/hr) 11
1.4 Sub-total 231.2 hrs 11
2. Harvest activities
2.1 Family labor 46.1 hrs
2.2 Social labor 8.3 hrs - 280
2.3 Hired labor .1 hrs 9 106 (CFA/hr) 11
2.4 Sub-total 53.5 hrs 291
3. Total 284.7 hrs 302
E. Total variable costs , 1,418
F. Tools and equipment (depreciation on) 179
G. Total costs 1.597
11. OUTPUT
A. Crop yields (kg/ha)
1. Shelled corn 885
8. Unit price (CFA/kg)
l. Shelled corn 38.2
C. Total value of output 33.807
111. PERFORMANCE MEASURES
A. Gross income 33,807
Less: Total variable costs 1,413
8. Gross margin 32.389
Less: Total fixed costs 179
C. Net margin 32.210
Less: Opportunity cost of equity capital 9 lt/month [(1,116 x 8) + (11 x 3)] (.01) 90
0. Net returns to land, FL (family labor) and management . 32,120
Less: Opportunity costs of FL: 277 hrs 8 55 CFA/hr) 15,235
E. Net returns to land and management 16,885
F. Net returns per field-hour of family labor (32,120 g 277) 116.0
6. Output - seed ratio 49.7
H. Costs of production (CFA/kg) (1.597 4 885) 1.8

Total costs of production (CFA/kg) (16,922 i 885)

19.1

 

 

92

TABLE 4.3.4

AVERAGE COSTS AND RETURNS PER HECTARE FOR GROUNDNUTS/BAMBERA NUTS
SYSTEM 3, IN THE EORD, 1980

 

 

CFA CFA

 

, 11.

III.

INPUT USE

A. Basic data
1. v Of cases 47
2. Average size (ha) .582

8. Non-labor input use
1. Seed rate (Kg/ha)
l.l Groundnuts 7.78 (142.8 CFA/k ) 1,100
1.2 Bambera nuts 4.30 (84.4 CFA/kg? 363
2. Total 1.463

C. Agronomic data

1. Percentage of area ploughed using:
1.1 Animal traction
1.2 Hand tools
1.3 Zero tillage

2. Percentage of area fertilized:
2.1 Chemically
2.2 Organically

NN
-»c> uo-uab

N

DNAO

0. Labor input use

1. Pre-harvest activities
1. 1 Family labor
1.2 Social labor
1.3 Hired labor
Lawmmui

2. Harvest activities
2.1 Family labor
2.2 Social labor
2.3 Hired labor
2.4 Sub-total

3. Total

N
N

hrs

rs
hrs 0 106 (CFA/hr) 21
hrs 65

m 41‘"
O

n: no
86 ~a
coco loin (3904040

hrs

hrs 205

hrs 9 (CFA/hr) 0

hrs 205

hrs 270

#N
o—J
UU‘ION

E. .Total variable costs . 1.733
F. Tools and equipment (depreciation on) 291
G. Total costs 2,024
OUTPUT
A. Crop yields (kg/ha)

1. Shelled groundnuts 119

2. Shelled bambera nuts 6
8. Unit price (CFA/kg)

1. Shelled groundnuts 67.4

2. Shelled bambera nuts 35.8
C. Total value of output 8,235
PERFORMANCE MEASURES

A. Gross income 8,235
Less: Total variable costs 1,733

8. Gross margin 6,502
Less: Total fixed costs 291

C. Net margin 6,211
Less: Opportunity cost of equity capital 8 lt/month [(1. 463 x 8) + (65 x 3)] (. 01) 119

0. Net returns to land. FL (family labor) and management 6.092
Less: Opportunity costs of FL: (480 hrs 8 55 CFA/hr) 26,400

E. Net returns to land and management -20,3OO
F. Net returns per field-hour of family labor (6,092 4 480) 12.7
G. Output - Seed ratio 15/1
H. Costs of production (CFA/kg) (2.024 a 125) 16.2
1. Total costs of production (CFA/kg) (28,543 a 125) 228.3

 

 

93

TABLE 4.3.5

AVERAGE COSTS AND RETURNS PER HECTARE FOR SOYBEANS

 

 

 

SYSTEM 3. IN THE EORD, 1980
CFA CFA
1. INPUT USE
A. Basic data
1. d of cases 18
2. Average size (ha) .494
B. Non-labor input use
1. Seed rate (kg/ha) 25.0 9 (157.3 CFA/kg) 3,932
Total 3.932
C. Agronomic data
1. Percentage of area ploughed using:
1.1 Animal traction 58.4
1.2 Hand tools 2.6
1.3 Zero tillage 39.1
2. Percentage of area fertilized:
2.1 Chemically o
2.2 Organically o
0. Labor input use
1. Pre-harvest activities
1.1 Family labor 147.2 hrs
1.2 Social labor .9 hrs 56
1.3 Hired labor 0 hrs 0
1.4 Sub-total 148.1 hrs - 56
2. Harvest activities
2.1 Family labor 23.5 hrs
2.2 Social labor 1.5 hrs 0
2.3 Hired labor 2.5 hrs 0 106 (CFA/hr) 265
2.4 Sub-total 27.5 hrs 265
3. Total 175.6 hrs 321
_E. Total variable costs ' 4.253
F. Tools and equipment (depreciation on) 950
G. Total costs 4,348
11. Output
A. Crop yields (kg/ha)
1. Soybeans 707
8. Unit price (CFA/kg)
l. Soybeans 116.7
C. Total value Of output 82.507
111. PERFORMANCE MEASURES
A. Gross income 82.507
Less: Total variable costs 4,253
8. Gross margin 78,254
Less: Total fixed costs 950
C. Net margin 77.304
Less: Opportunity cost of equity capital 9 lz/month [(3,932 x 8) + (56 x 3)] (.01) 316
0. Net returns to land, FL (family labor) and management 76.988
Less: Opportunity costs of FL: (171 hrs 6 55 CFA/hr) 9.405
E. Net returns to land and management 67,583
F. Net returns per field-hour of family labor (76,988 4 171) 450.2
G. Output - seed ratio 28.2
H. Costs of production (CFA/kg) (4.348 t 707) 6.1
' 1. Total costs Of production (CFA/kg) (14.059 . 707) 19.9

 

 

94

system, 76 to 97 percent of the total labor input was family labor. FOr
the rice enterprise, family labor accounted for 76 percent of total
labor input while for maize, GN/BN or soybean enterprises, it accounted
up to 97 percent Of total labor input per hectare. In this system of
production, hired labor was used in all enterprises, its contribution

to total labor input under all enterprises being less than 2 percent.
But total non-family labor contribution to total labor input in this
system varied between 3 percent and 24 percent. The relatively high
dependence on non-family labor in this system gives us an idea of

some employment Opportunities for the rural population existing

within this production system. And at the same time, it lets us suspect
that family labor may rapidly be becoming a constraint on production in
this system, or that farmers in this system are beginning to substitute
hired labor for family labor in order to free family labor for other
purposes (e.g., leisure).'

The mean expenditure per hectare for non-family labor in this
system varied from 270 CFA for GN/BN enterprise to 2.353 CFA for S/M/C
enterprise. Total'labor expenditures on non-family labor in the form of
wages, food and drink varied from 7 percent of total farm expenditures
(TFE) under the soybean enterprise to 39 percent of TFE under the S/M/C
enterprise.

The seed rates observed in this production system were quite Often
far from the recommended rates. The mean quantity Of paddy seeds used
by farmers here was 38 kg/ha which is only 47 percent Of the recommended
rate Of 80 kg/ha. In the case of the maize enterprise, the mean quan-
tity of seeds used was only 71 percent Of the recommended rate. Also,

in the case of soybeans, the average seed rate Observed, 25 kg/ha was

95

only 62 percent of the recommended rate. In this production system,
fertilizer was used only in maize but at a very low rate (1.3 kg/ha)
which is only 1.3 percent Of the recOmmended rate. The average appli—
cation rate of fertilizer is so low that it suggests only farmer experi-
mentation on a small proportion of "fertilized" area.

Crop yields were generally low, particularly for sorghum, ground-
nuts and bambera nuts. However, it should be noted that yields Obtained
under crop mixtures may understate the potential yields of those crops
when grown in pure stands. A cowpea yield of 15 kg/ha represents only
the average contribution of cowpeas to the grain production enterprise,
taking into consideration that sometimes no cowpeas were harvested even
though they were planted. While the yield so computed may understate
the potential yield of cowpeas as an enterprise itself, it correctly
measures its average importance or cOntribution to the grain production
enterprise. However, it remains true that one of the major problems
facing this production system is how to increase cereal yields from their '
present levels of ZOO-500 kg/ha.

3.3.2. Comparison and Appraisal of the Five Major Enterprises
in System 3

A summary of general characteristics, costs and returns as well as
performance measures for all five enterprises are provided in Table
14.5.3. The discussion will mainly concern the analysis of the perfor-
rnance measures so as to identify the enterprises with the highest

'Financial return and lowest cost Of production.

I

96

TABLE 4.5.3

A COMPARATIVE ANALYSIS OF THE MAJOR ENTERPRISES IN SYSTEM 3.
BASED ON SURVEY DATA FROM 30 HOUSEHOLDS. 1980

 

 

 

 

Enterprises
Criteria
Rice SIM/C Maize GN/BN Soybeans
I. General Characteristics
1. I of cases 45 56 39 47 18
2. Average size (ha) .488 .707 .431 .582 .494
3. Average yield (kg/ha) 501 125/315/15 885 119/6 707
11. Financial Situation (CFA/ha)
1. Gross income 32.916 19,896 33,807 8,235 82.507
2. Variable costs 5,696 5,352 1.418 1,733 4.253
3. Total expenditures
(including depreciation on
tools and equipment) 5,930 6.013 1.597 2,024 4,348
4. Opportunity costs Of
4.1 Family labor 28,490 74.085 15.235 26.400 9.405
4.2 Equity capital 355 298 90 119 316
5. Total costs 34,775 80,396 16,922 28,543 14,069
111. Performance Measures
1. Gross margin (CFA/ha)
(11.1 - 11.2) 27.220 14.544 32,389 6,502 78,254
2. Net margin (CFA/ha)
(11.3) 26,986 13.883 32.210 6,211 77,304
3. Net returns to land.
family labor a management
(CFA/ha) (111.2 - 11.4.2) 26,631 13,585 32,120 6,092 76,988
4. Net returns to land a
management (CFA/ha):
(111.3 - 11.4.1) -l.859 -60,500 16.885 -20,300 67,583
5. Net returns per hour of -
family labor (CFA/phr)
(111.3 + Total FL) 51.4 10.1 116.0 12.7 450.2
6. Total costs of production
(CFA/kg) 69.4 176.7 19.1 228.3 19.9
7. Output - seed ratio 13.2 6/14/2 49.7 15/1 28.2

 

 

97

3.3.2.1. Gross Margin (GM)

Among the five major enterprises of System 3, the variation in
gross margin ranged from 6,502 CFA/ha to 78,254 CFA/ha (Table 4.5.3).
The soybean enterprise had the highest gross margin and the GN/BN enter-
prise had the lowest gross margin. One thing interesting to note though
is that the gross margins for all enterprises studied was positive and
hence, all the enterprises are valid candidates to stay in the farm

business organization according to the neg-classical economic theory.

3.3.2.2. Net Margin (NM)

Among the five major enterprises, the variation in the NM ranged
from 6,211 CFA/ha to 77,304 CFA/ha. So far, the soybean enterprise
appears to provide the highest returns per hectare of all crops under
consideration in this system followed by maize (32,210 CFA/ha) and rice
(26,986 CFA/ha) (Table 4.5.3).

3.2.2.3. Net Returns to Land, Family Labor and Management (NRLFLM)

The NRLFLM for the five enterprises ranged from 76,988 CFA/ha for
soybean enterprise to 6,092 CFA/ha for the GN/BN enterprise (Table
4.5.3). In all enterprises, except for soybeans and maize, the return
per field hour of family labor was less than the minimum wage rate paid
to unskilled urban workers (i.e., 90 CFA/hr). This result suggests that
there may be some financial gain in seeking employment in urban areas or
other farms where the minimum Observed wage to hired labor is 45 CFA/hr
(e.g., rice farms). In the case of soybean and maize enterprises where
the returns are 450.2 CFA/hr and 116.0 CFA/hr, respectively, returns
here are far above the minimum average agricultural wage rate and also

have the SMIG. Thus there is no financial advantage of family members

98

seeking wage employment in other enterprises or in urban areas when they
are needed on their soybean and maize fields. In the case of rice, there
is some financial advantage of family members seeking wage employment in
other enterprises or in urban areas since the returns are only 51.4
CFA/hr. But compared to returns under S/M/C enterprise, farmers are
better Off working on their rice fields. The following shifts could be
expected to take place:

a) farmers will put more of their land and labor into soybeans
and maize if the minimum sorghum needed for family consumption is
attained; and

b) low returns per hour of family labor in rice enterprise
compared to the SMIG, may force farmers to abandon this crop since as
a grain, it is not yet an important part Of the diet. Some incentive
structure must be urgently found if rice growing is to survive in this
system where some important investments in water control have already

been made.

3.3.2.4. Returns to Land and Management (RLM)

All enterprises in this system, except soybeans and maize, realized
a negative return to land and management (Table 4.5.3). The RLM for
the five enterprises ranged from 67,583 CFA/ha for soybeans to —60,500
CFA/ha for the S/M/C enterprise.

3.3.2.5. Costs of Production

Taking into consideration only variable and fixed costs, maize
showed the lowest cost of production (1.8 CFA/kg) and rice had the
highest cost of production (11.8 CFA/kg of paddy) (Tables 4.3.l-4.3.5).

The second type Of cost of production computed was Obtained by adding

99

the Opportunity cost of equity capital and family labor to total
expenditures and the result was divided by the yield. Among the five
enterprises, maize showed the lowest total costs of production (19.1
CFA/kg) and GN/BN had the highest total cost of production (228.3
CFA/kg). The second lowest total cost of production was found in soy-
beans (l9.9 CFA/kg). The third lowest total cost Of production was
found in rice (69.4 CFA/kg).

3.4. System 4: Production System Based on Farmers Growing
Irrigated Paddies with Fertilizer

In this production system, the most important enterprises in terms
Of the objectives Of study, area cultivated, labor used, and income
generated were: rice, sorghum/millet/cowpeas (SIM/C), maize, ground-

nuts, bambera nuts and Okra.

3.4.1. Overview Of the Enterprise Budgets in System 4

Costs and returns to the six major enterprises in System 4, derived
from the 1980 survey are shown in Tables 4.4.l-4.4.6. The average areas
planted in rice, S/M/C, maize, groundnuts, bambera nuts and Okra were
in hectares, .151, .766, .093, .118, .042 and .065, respectively. The
mean labor utilization per hectare in all field activities ranged from
145.4 hours for the Okra enterprise to 3,054 hours for the rice enter-
prise. In this system, 98 to 100 percent Of the total labor input per
hectare was family labor. No hired labor was used in this system. The
low dependence on non-family labor gives us an idea of the poor employ-
ment Opportunities which exist in this production system.

The mean expenditure per hectare for social labor in rice and

S/M/C enterprises where it is used were 1,220 CFA and 41 CFA, respectively.

100

TABLE 4.4.1

AVERAGE COSTS AND RETURNS PER HECTARE FOR RICE
SYSTEM 4, IN THE EORD, 1980

 

 

 

CFA CFA
1. INPUT USE
A. Bésic data
1. a of cases 62
2. Average size (ha) .151
8. Non—labor input use
1. Seed rate (kg/ha) 57. 7 O (125 CFA/k ) 7,212
2. Fertilizer 18-35—0) (kg/ha) 12 20. 3 O (56 CFA/kg? 6,737
3. Pesticides kg/ha) 3. 3 O 196.1 CFA/kg) 647
Total 14,596
C. Agronomic data
1. Percentage of area ploughed using:
1.1 Animal traction 14.7
1.2 Hand tools 70.9
1.3 Zero tillage 14.4
2. Percentage of area fertilized:
2.1 Chemically 76.4
2.2 Organically O
0. Labor input use
1. Pre-harvest activities
1.1 Family labor 1.925 hrs
1.2 Social labor 22 hrs 301
1.3 Hired labor ' 0 hrs 0
1.4 Sub-total 1.947 hrs 301
2. Harvest activities
2.1 Family labor 1,051 hrs
2.2 Social labor 56.3 hrs 919
2.3 Hired labor , 0 hrs 0
2.4 Sub-total 1.107 hrs 919
3. Total 3.054 hrs 1.220
E. Total variable costs 15,816
F. Tools and equipment (depreciation on) 1.226
G. Total costs 17,042
11. OUTPUT
A. Crop yields (kg/ha)
1. Paddy rice 1,736
8. Unit price (CFA/kg)
l. Paddy rice 51.8
C. Total value of output 89,925
111. PERFORMANCE MEASURES
A. Gross income 89,925
Less: Total variable costs 15,816
8. Gross margin 74,109
Less: Total fixed costs 1,226
C. Net margin 72,883
Less: Opportunity cost of equity capital 0 l%/month [(14,596 x 8) + (301 x 3)] (.01) 1,177
0. Net returns to land, FL (family labor) and management 71,706
Less: Opportunity costs of FL: 2,976 hrs 8 27 CFA/hr) 80,352
E. Net returns to land and management -8.646
F. Net returns per field-hour of family labor (71,706 4 2,976) 24.1
G. Output - seed ratio 30.2
H. Costs of production (CFA/kg) (17,042 + 1.736) 9.8
1. Total costs of production (CFA/kg) (98,571 a 1,736) 56.8

 

 

1(11

TABLE 4.4.2

AVERAGE COSTS AND RETURNS PER HECTARE FOR SORGHUM/MILLET/COHPEAS
SYSTEM 4, IN THE EORD, 1980

 

 

CFA _ CFA

 

III.

INPUT USE

A.

Basic data
1. I Of cases 48
2. Average size (ha) .766

Non-labor input use

1. Seed rate (kg/ha)
1.1 Sorghum 7.9 O (93.2 CFA/kg) 736
1.2 Millet ’ 2.7 O (83.4 CFA/kg) 225
1.3 Cowpeas 3 2 O (93 0 CFA/kg) 298

Total 1.259

Agronomic data

1. Percentage of area ploughed using:
1.1 Animal traction O
1.2 Hand tools 98.
1.3 Zero tillage

2. Percentage of area fertilized:
2.1 Chemically
2.2 Organically

(”N

00

Labor input use
1. Pre-harvest activities
1.1 Family labor - 1.233 hrs
1.2 Social labor .9 hrs 29
1.3 Hired labor 0 hrs 0
1.4 Sub-total 1.233 hrs 29
2. Harvest activities
2.1 Family labor 205.3 hrs
2.2 Social labor 2.2 hrs 12
2.3 Hired labor 0 hrs 0
2.4 Sub-total 207.5 hrs 12
3. Total 1.440 hrs 41

Total variable costs 1,300
Tools and equipment (depreciation on) » 4,919
Total costs 6.219

OUTPUT

Crop yields (kg/ha)

1. Sorghum 401
2. Millet 20
3. Cowpeas 12

Unit price (CFA/kg)
l. Sorghum/millet 33.9
2. Cowpeas 56. 9

Total value of output 14,955

PERFORMANCE MEASURES

A.

10

Cross income ’ 14,955
Less: Total variable costs 1,300

Gross margin 13,655
Less: Total fixed costs 4,919

Net margin 8,736
Less: Opportunity cost of equity capital 0 lz/month [(1,259 x 8) + (29 x 3)] (.01) 102

Net returns to land, FL (family labor) and management 8,534
Less: Opportunity costs of FL: (1,417 hrs 8 27 CFA/hr) 38,259

Net returns to land and management -29.625
Net returns per field-hour of family labor (8,634 4 1,417) 6.
Output - seed ratio . 51/7/4
Costs Of production (CFA/kg of grain) (6.219 t 433) ' 14.4
Total costs of production (CFA/kg of grain) (44,580 a 433) 103.0

 

102

TABLE 4.4.3

AVERAGE COSTS AND RETURNS PER HECTARE FOR MAIZE
SYSTEM 4, IN THE EORD, 1980

 

 

 

CFA CFA
1. INPUT USE
A. Basic data
1. I of cases 18
2. Average size (ha) .093
8. Non-labor input use
1. Seed rate (kg/ha) 24.6 8 (81.7 CFA/kg) 2,010
Total 2.010
C. Agronomic data
1. Percentage Of area ploughed using:
1.1 Animal traction 18.7
1.2 Hand tools 75.0
1.3 Zero tillage 7.3
2. Percentage Of area fertilized:
2.1 Chemically O
2.2 Organically 31.8
0. Labor input use
1. Pre-harvest activities '
1.1 Family labor 601.2 hrs
1.2 Social labor 0 hrs
1.3 Hired labor 0 hrs
1.4 Sub-total 601.2 hrs
2. Harvest activities
2.1 Family labor 238.1 hrs
2.2 Social labor 0 hrs
2.3 Hired labor 0 hrs
2.4 Sub-total 238.1 hrs
3. Total 839.3 hrs 0
E. Total variable costs 2.010
F. Tools and equipment (depreciation on) . 211
G. Total costs I . 2.221
11. OUTPUT
A. Crop yields (kg/ha)
l. Shelled corn 2,197
8. Unit price (CFA/kg)
l. Shelled corn 31.8
C. Total value of output 69,895
111. PERFORMANCE MEASURES
A. Gross income ' 69.895
Less: Total variable costs 2,010
8. Gross margin 67,855
Less: Total fixed costs 211
C. Net margin 67.644
Less: Opportunity cost of equity capital 0 lzlmonth (2,010 x 8 x .01) 161
0. Net returns to land, FL (family labor) and management 67,483
Less: Opportunity costs of FL: (839 hrs 9 27 CFA/hr) 22,653
E. Net returns to land and management 44.830
F. Net returns per field-hour Of family labor (67,483 a 839) 80.4
G. Output - seed ratio 89.3
H. Costs Of production (CFA/kg) (2.221 a 2,197) 1.0

Total costs Of production (CFA/kg) (25.035 4 2,197) 11.4

 

 

103

TABLE 4.4.4

AVERAGE COSTS AND RETURNS PER HECTARE FOR GROUNDNUTS
SYSTEM 4, IN THE EORD, 1980 ~

 

 

 

CFA CFA
1. INPUT USE
A. Basic data
1. 4 of cases 20
2. Average size (ha) .118
B. Non-labor input use
1. Seed rate (kg/ha) 44.0 9 (80.8 CFA/kg) 3,555
Total 3.555
C. Agronomic data
1. Percentage of area ploughed using:
1.1 Animal traction 24.3
1.2 Hand tools 72.9
1.3 Zero tillage 2.8
2. Percentage Of area fertilized:
2.1 Chemically O
2.2 Organically 18.2
0. Labor input use
1. Pre-harvest activities
1.1 Family labor 355.6 hrs
1.2 Social labor 0 hrs
1.3 Hired labor 0 hrs
1.4 Sub-total 355.6 hrs
2. Harvest activities
2.1 Family labor 323.9 hrs
2.2 Social labor 0 hrs
2.3 Hired labor 0 hrs
2.4 Sub-total 323.9 hrs
3. Total 679.5 hrs 0
E. Total variable costs 3.555
F. Tools and equipment (depreciation on) 303
G. Total costs 3.858
11. OUTPUT
A. Crop yields (kg/ha)
1. Shelled groundnuts 422
8. Unit price (CFA/kg)
1. Shelled groundnuts 41.2
C. Total value of output 17,386
111. PERFORMANCE MEASURES
A. Gross income 17,386
Less: Total variable costs 3,555
8. Gross margin 13,831
Less: Total fixed costs 303
C. Net mar in 13.528
Less: pportunity cost Of equity capital 9 lz/month (3,555 x 3 x .01) 234
0. Net returns to land. FL (family labor) and management 13,244
Less: Opportunity costs of FL: (679 hrs 0 27 CFA/hr) 13,333
E. Net returns to land and management -5,039
Net returns per field-hour Of family labor (13,244 4 579) 19,5
G. Output - seed ratio 9.6
H. Costs of production (CFA/kg) (3,858 a 422) 9.1
1. Total costs Of production (CFA/kg) (22,475 a 422) 53.3

 

 

104

TABLE 4.4.5

AVERAGE COSTS AND RETURNS PER HECTARE FOR BAMBERA NUTS

SYSTEM 4, IN THE EORD, 1980

 

 

 

CFA CFA
1. INPUT USE
A. Basic data
1. a of cases 3
2. Average size (ha) .042
B. Non-labor input use
1. Seed rate (kg/ha) 79 9 (109.1 CFA/kg) 8,619
Total 8.619
C. Agronomic data
1. Percentage of area ploughed using:
1.1 Animal traction O
1.2 Hand tools 100
1.3 Zero tillage O
2. Percentage of area fertilized:
2.1 Chemically O
2.2 Organically O
0. Labor input use
1. Pre-harvest activities
1.1 Family labor 770 hrs
1.2 Social labor 0 hrs
1.3 Hired labor 0 hrs
1.4 Sub-total 770 hrs
2. Harvest activities
2.1 Family labor 750 hrs
2.2 Social labor 0 hrs
2.3 Hired labor 0 hrs
2.4 Sub-total 750 hrs
3. Total 1.520 hrs 0
E. Total variable costs ' 8,619
F. Tools and equipment (depreciation on) 130
G. Total costs 8,749
11. OUTPUT
A. Crop yields (kg/ha)
1. Shelled bambera nuts 540
8. Unit price (CFA/kg)
1. Shelled bambera nuts 48.2
C. Total value of output 26,028
111. PERFORMANCE MEASURES
A. Gross income 26.028
Less: Total variable costs 8.619
8. Gross margin 17,409
Less: Total fixed costs 130
C. Net nmrgin 17.279
Less: Opportunity cost of equity capital 9 lZ/month (8,619 x 8 x .01) 690
0. Net returns to land, FL (family labor) and mana ement 16.589
Less: Opportunity costs of FL: (1,530 hrs 0 7 CFA/hr) 41,310
E. Net returns to land and management -24.721
F. Net returns per field-hour of family labor (16.589 1 1.530) 10.8
G. Output - seed ratio 6.8
H. Costs of production (CFA/kg) (8.749 t 540) 16.2
1. Total costs of production (CFA/k9) (50,749 4 540) 94.0

 

 

105

TABLE 4.4.6

AVERAGE COSTS AND RETURNS PER HECTARE FOR OKRA
SYSTEM 4, IN THE EORD, 1980

 

 

 

CFA CFA
1. INPUT USE
A. Basic data
1. 4 of cases 20
2. Average size (ha) .065
8. Non-labor input use
1. Seed rate (kg/ha) 12.3 9 (144 CFA/kg) 1,771
Total ° 1.771
C. Agronomic data
1. Percentage of area ploughed using:
1.1 Animal traction 2.3
1.2 Hand tools 92.3
1.3 Zero tillage 4.4
2. Percentage of area fertilized:
2.1 Chemically O
2.2 Organically 6.8
0. Labor input use
1. Pre-harvest activities
1.1 Family labor 120.8 hrs
1.2 Social labor 0 hrs
1.3 Hired labor 0 hrs
1.4 Sub-total 120.8 hrs
2. Harvest activities
2.1 Family labor 24.6 hrs
2.2 Social labor 0 hrs
2.3 Hired labor 0 hrs
2.4 Sub-total 24.6 hrs
3. Total 145.4 hrs 0
E. Total variable costs 1.771
F. Tools and equipment (depreciation on) 171
G. Total costs 1.942
11. OUTPUT
A. Crop yields (kg/ha)
1. Fresh okra 370
8. Unit price (CFA/kg) .
1. Fresh okra 83.3
C. Total value of output 30.821
111. PERFORMANCE MEASURES
A. Cross income 30,821
Less: Total variable costs 1,771
8. Gross margin 29.050
Less: Total fixed costs 171
C. Net margin 28.879
Less: Opportunity cost of equity capital 9 ltlmonth (1,771 x 8 x .01) 142
0. Net returns to land, FL (family labor) and management 28.737
Less: Opportunity costs of FL: (145 hrs 9 27 CFA/hr) 3.915
E. Net returns to land and management 24.822
F. Net returns per field-hour Of family labor (28,737 a 145) 198.2
G. Output - seed ratio 30.1
H. Costs of production (CFA/kg) (1.942 a 370) 5.2
1. Total costs of production-(CFA/kg) (5,999 . 370) 16.2

 

 

106

For the rice enterprise, 75 percent Of all expenditures were on harvest
activities whereas under the SIM/C enterprise, 71 percent Of these
expenditures were spent on pre-harvest activities (Tables 4.4.l-4.4.2).

The seed rates observed in this production system were quite far
from the recommended rates. The mean quantity of paddy rice seeds used
by farmers of this system was 57.7 kg/ha which is about 72 percent of
the recommended rate Of 80 kg/ha. For the groundnut enterprise, only 55
percent Of the recommended seed rate was applied, and for maize, the
average seed rate observed (24.6 kg/ha) was very close to the recommended
rate Of 25 kg/ha.

Crop yields were generally low, but relative to other systems
studied, yields were quite high. This certainly reflects certain general
ecological conditions such as rainfall, humidity, soil type, etc., of
the area where System 4 was located. In general, yields are higher
with a higher rainfall and a better rain distribution. And as Baker
and Lassiter (1980, pp. 48—49) pointed out, the higher rainfall and
better distribution of rain in the Diapaga area goes with higher plant
density, except for millet, which explains why, in general, we have
higher yields in this system for all crops. It can also be observed
that a large quantity of family labor input per hectare was used in
this system of production. These Observations amply demonstrate that
.yields in any system Of production is not the result of any single
factor.

3.4.2. Comparison and Appraisal of the Six Major Enterprises
in System 4
A summary Of general characteristics, costs and returns, as well as

performance measures for all six enterprises are provided in Table 4.5.4.

107

TABLE 4.5.4

A COMPARATIVE ANALYSIS OF THE MAJOR ENTERPRISES IN SYSTEM 4,
BASED ON SURVEY DATA FROM 30 HOUSEHOLDS, 1980

 

 

 

 

Enterprises
Criteria
Ground- Bambera
Rice S/M/C Maize nuts nuts Okra
I. General Characteristics
1. 4 of cases 62 48 18 20 3 20
2. Average size (ha) .151 .766 .093 .118 .042 .065
3. Average yield (kg/ha) 1.736 401/20/12 2,197 422 540 370
11. Financial Situation (CFA/ha)

1. Gross income 89.925 14,955 69,895 17.386 26.028 30.821
2. Variable costs 15,816 1,300 2.010 3.555 8,619 1,771
3. Total expenditures

(including depreciation on

tools and equipment) 17,042 6,219 2,221 3,858 8,749 1,942
4. Opportunity costs of

4-1 F4811! 1490' 80.352 38.259 22.653 18.333 41.310 3.915

4.2 Equity capital 1,177 102 161 284 690 142
5. Total costs 98,571 44,580 25,035 22,475 50.749 5.999

111. Performance Measures

1. Gross margin (CFA/ha)

(11.1 - 11.2) 74,109 13,655 67,855 13,831 17,409 29,050
2. Net margin (CFA/ha)

(11.1 - 11.3) 72.883 8,736 67,644 13,528 17,279 28,879
3. Net returns to land, family

labor and management (CFA/ha)

(111.2 - 11.4.2) 71,706 8,634 67,483 13,244 16,589 28,737
4. Net returns to land 6

management (CFA/ha)

(111.3- 11.4.1) -8.646 -29,625 44,830 -5,089 -24,721 24,822
5. Net returns per hour of

family labor (CFA/phr)

(III.3+ Total FL) 24.1 6.1 80.4 19.5 10.8 198.2
6. Total costs of

production (CFA/kg) 56.8 103.0 11.4 53.3 94.0 _ 16.2
7. Output-seed ratio 30.2 51/7/4 89.3 9.6 6.8 30.1

 

 

108

The discussion will mainly concern the analysis of the performance
measures so as to identify the enterprise with the highest financial

return and lowest cost of production.

3.4.2.1. Gross Margin (GM)

Among the six major enterprises of System 4, the variation in gross
margin ranged from 13,655 CFA/ha to 74,109 CFA/ha (Table 4.5.4). The
rice enterprise had the highest gross margin and the S/M/C enterprise

had the lowest gross margin, but all the GMs were positive.

3.4.2.2. Net Margin (NM)
Among the six major enterprises, the variation in NM ranged from
8,736 CFA/ha to 72,883 CFA/ha. Rice so far, appears to provide the

highest return per hectare of all crops in this system.

3.4.2.3. Net Returns to Land, Family Labor and Management (NRLFLM)

The NRLFLM for the six enterprises ranged from 8,634 CFA/ha for the
S/M/C enterprise to 71,706 CFA/ha for the rice enterprise (Table 4.5.4).
In all enterprises, except for Okra the return per field hour of family
labor was less than the minimum wage rate paid to unskilled urban
workers, i.e., 90 CFA/hr. This result suggests that there may be some
‘financial_gain in seeking employment in urban areas. However, in the
case of the Okra enterprise, the returns of 198 CFA/hr is far above
the minimum wage rates. Thus, there is no financial advantage Of family
members seeking wage employment when they are needed on their Okra
fields. In the case of rice, the returns Of 24.1 CFA/hr of family labor
is far smaller than the minimum wage rates. This may be enough to ini-

tiate an exit process in this fragile industry. The following shifts

109

could be expected:

a) farmers will put more of their land and labor under Okra, maize
and groundnuts if the minimum sorghum needed for home consumption is
attained; and

b) low returns per hour of family labor in rice may force farmers

to abandon this crop despite the government support of this crop.

3.4.2.4. Returns to Land and Management (RLM)

All enterprises, except okra and maize, realized a negative return
to land and management (Table 4.5.4). The high negative RLM under
S/M/C, bambera nuts and rice, was probably due to the large quantity of

family labor input per hectare.

3.4.2.5. Costs of Production

Taking into consideration only variable and fixed costs, maize
showed the lowest cost Of production (1.0 CFA/kg) and bambera nuts had
the highest cost of production (16.2 CFA/kg) (Tables 4.4.l-4.4.6). The
second type of cost of production computed was Obtained by adding the
Opportunity costs of equity capital and family labor to total farm ex-
penditures and the result was divided by the yield. Among the six
lenterprises, maize still showed the lowest total cost of production
(11.4 CFA/kg) and S/M/C had the highest total cost of production (103.0
CFA/kg). The second highest total cost Of production was found in
bambera nuts (94.0 CFA/kg). Okra showed the second lowest total costs
of production (16.2 CFA/kt), and rice showed the third lowest total
costs of production (56.8 CFA/kg).

110

3.5. Comparison and Appraisal Of the Rice Enterprises
Across the Four Production Systems Studied

Four basic systems Of water control in rice farming were identified
in the EORD. One relies on uncertain surface flooding (System 1) and
attains a yield of less than 500 kg/ha (with no modern inputs). The
other three provide partial or complete water control, with yields of
.5 to 1.2 ton per hectare for improved swamps (Systems 2 and 3) and 1.7
tons per hectare for the dam syStem with fertilizer use. Table 4.5.5
summarizes the general characteristics, costs and returns as well as
measures Of efficiency for all four rice production techniques in the
EORD. The discussion will focus mainly on the private profitability
measures. These measures are based on average costs and returns fOr
existing methods of rice farming in the EORD.

The NRLFLM at the farm level ranged from a high of about 71,706 CFA/
ha on irrigated paddies in System 4 to a low of 23,908 CFA/ha on tradi-
tional bas-fonds in System 1. The greatest part of this difference is
caused by variations in the method of water control which has a clear
impact on yields. Rice cultivation is cheapest by a wide margin on
traditional bas-fonds (System 1). Most expensive is the production on
irrigated bas-fOnds (Table 4.5.5). Cost variations among the three
systems with partial or complete water control are quite appreciable
(cf. family labor use between Systems 2 and 3 or Systems 3 and 4). Part
of the difference in labor requirements as we mentioned earlier is prob-
ably due to the difference in the method Of land preparation, yield
differences and/or intensity Of weeding. For instance, while in System
2, 100 percent Of rice fields were prepared using hand-tools, in System

3, zero tillage during the 1980 survey was used on almost 78 percent of

111

TABLE 4.5.5

A COMPARATIVE FINANCIAL ANALYSIS OF THE FOUR MAJOR RICE PRODUCTION TECHNIQUES
IN THE EORD. BASED ON SURVEY DATA FROM 116 HOUSEHOLDS, 1980-81

 

 

Production Techniques

 

 

Traditional Semi- Improved Irrigated
Criteria Bas-Fonds Traditional Bas-Fonds Bas-Fonds
Bas-Fonds
I 11 111 IV
I. General Characteristics
1. i of cases 64 76 45 62
2. Average size (ha) .411 .270 .488 .151
3. Average yield (k /ha) 458.3 1,172 501 1.736
4. Seed rate (kg/ha? 23.6 50.8 38.0 57.7
11. Financial Situation (CFA/ha)
1. Gross income 27,773 48,872 32,916 89,925
2. Variable costs 2.858 6.161 5,696 15,816
3. Total expenditures
(including depreciation on
tools 3 equipment) 3,708 7,085 5,930 17,042
4. Opportunity costs of
4.1 Family labor 11,515 44,200 28.490 80.352
4.2 Equity capital 157 , 406 355 1,177
5. Total costs 15,380 51,691 34,775 98,571
111. Performance Measures
1. Gross margin
(11.1 - 11.2) 24.915 42.711 27.220 74.109
2. Net margin
(11.1 - 11.3) 24,065 41,787 26,986 72,883
3. Net returns to land,
family labor a management
(CFA/ha) (III-Z - 11-4-2) 23.908 41.381 26.631 71.706
4. Net returns to land 6
management (CFA/ha)
(111.3 - 11.4.1) 12,393 -2,819 -l.859 -8,646
5. Net returns per hour of
family labor (CFA/9hr)
(111.3 + Total FL) 97.6 18.7_ 51.4 24.1
6. Total costs of production
(CFA/k9) 33.6 44.1 59.4 55.3

 

 

112

rice fields; and in System 4, higher yields realized called for a higher
level of family labor use. Other factors that may explain differences
in labor requirements among the four systems include soil quality, field
size and length of growing season.

3.6. Social Profitability Of Rice Cultivation Under
the Four Systems Studied

The objective Of this section is to determine the economic returns
to the four alternative methods of rice cultivation in the absence of
distorting government policies and imperfection in factor and product
markets. Market imperfections due, for example, to economies Of scale
or the existence Of externalities or monopoly elements in the economy,
are difficult to measure and are probably not so quantitatively import-
ant as the distortions introduced by government. Hence, emphasis here

is placed on the effects Of government policies on the rice production.

3.6.1. Shadow Prices of Domestic Factors and Output

The shadow price of a scarce factor will be adequately approximated
by its market price if the imperfections and other distortions in the
market are minor. These conditions are largely fulfilled for labor and
seeds. SO the factor price distortions facing rice producers are mainly
budget subsidies on inputs such as fertilizers and land improvement

costs as well as on the average invested capital.

3.6.1.1. Fertilizer

Farmers in the EORD are in general paying low prices for urea and
fertilizer (18-35-0). The price paid by farmers per kilo Of urea and
18-35-0 were 60 CFA and 56 CFA, respectively. Their true cost-prices

113

are estimated at 90 CFA and 80 CFA per kilo, which translate into 50 and
43 percent subsidy for urea and fertilizer (18-35-0), respectively.8
Thus an average of 44 percent subsidy weighted by the quantities of
each fertilizer used was added to farmer's total expenditures on fertil-

izer to approximate its true economic cost.

3.6.1.2. Water Control Costs
Part of or total cost of land improvement work and/or water control
in the EORD is supported by government funds. These investment costs

vary from 18,000 CFA/ha in System 2 to 68,000 CFA/ha in System 4.9

There
is no fee charged to farmers using these newly developed areas. Thus,
the annual investment costs borne by the government were added to

farmers' total farm expenditure to approximate its true eoonomic cost.

3.6.1.3. Import Parity Price Of Rice

The world price is used to evaluate the profitability of domestic
production of rice in the EORD since rice imports are the major alter-
native tO increased rice output in the EORD.

To determine the gross economic benefits from each alternative
technique of rice cultivation, the import parity price of domestic pro-
duction was computed as shown in Table 4.5.6. The import parity price
of a kilogram Of paddy was determined to be 56.4 CFA (Table 4.5.6).

SO in System 2, farmers were receiving a lower price for their output

Of rice (41.7 CFA/kg Of paddy) comparatively to the world price

 

8The source of cost-price information is World Bank Report No.
3296-UV, September 1980.

9The source of information on investment costs is FDR, "Rapport
technique," Campagne, 1977-78. ,

114
TABLE 4.5.6

IMPORT PARITY PRICE OF A TON OF PADDY
PRODUCED IN THE EORD, 1980

 

 

 

 

 

Item Value (CFA)

1. Cost C.I.F. Abidjan (CFA/mt) 97,000

plus port handling 1,870

plus transhipment to rail 1,000

plus rail transport Abidjan-Ouaga 9,500

plus road transport Ouaga-Fada 6,500

plus unloading Fada 300

2. Wholesale price Fada area (CFA/mt) 116,170

less milling costs 12,000

Sub-total 104,170

3. Paddy equivalenta 67,711

less transport to mill (in Fada) 2,100

less bag costs 3,000

less collection costs of paddy 6,200

4. Economic price of paddy (CFA/mt) 56,411
1979 SOURCE: Adapted from FAO trade year book and OFNACER's Report,

aAt the average milling rate Of 65%.

equivalent (56.4 CFA/kg).

In Systems 1 and 3, farmers were receiving

60.6 CFA/kg and 65.7 CFA/kg, respectively, which were greater than the

world price equivalent.

In Systems 2 and 4, farmers were receiving a

lower price for their output, i.e., 41.7 CFA/kg and 51.8 CFA/kg,

respectively.

115

In comparison with the financial analysis (Table 4.5.5), the
economic costs of production for the four techniques Of rice cultiva-
tion were higher, varying from an increase Of 35 percent increase in
System 1 to 132 percent increase in System 3 (see Table 4.5.7). These
increases in costs are explained by the high rate of subsidy and the
costs Of water control which were not taken into consideration in the
financial analysis. Moreover, the variation in increases observed may
be due to the mix Of subsidized resources in the different systems
studied. The fact that System 3 showed the highest percentage increase
in cost is mainly due to low yield achieved in this system.

Table 4.5.7 shows that the least cost technique for producing rice
even from an economic point of view remains the traditional cultivation
on unimproved swamps. The introduction of water control to date does
not compete effectively with this basic system, and total costs per kilo-
gram of paddy rise in every case. The most efficient means of producing
rice under "secure" water control appears to be by partial water control
(System 2 where economic losses are only 4,798 CFA/ha). Production on
irrigated bas-fonds with complete water control and fertilizer use is
most expensive with economic losses averaging 70,966 CFA/ha. Neverthe-
less, when considering whether rice production should be increased by
promoting a particular technique of cultivation, it is possible to decide
in general terms that only System 1 will be economically feasible.
Despite the importance of water control in raising yields and expanding
physical output of rice, the high capital costs of increasing water
control are seldom offset by sufficiently higher yields, which means

that net returns to land and labor dO not necessarily rise.

116

TABLE 4.5.7

A COMPARATIVE ECONOMIC ANALYSIS OF THE FOUR MAJOR RICE
PRODUCTION TECHNIQUES IN THE EORD, 1980

 

 

Production Techniques

 

Criteria
Traditional Semi-Traditional Improved Irrigated
Bas-fonds (I) Bas-fonds (II) Bas-fonds (III) Bas-fonds (IV)

 

1. Total cost of labora

(CFA/ha) 12.412 45.317 30.371 81.572
2. Seeds (CFA/ha) 1,961 4,613 3.815 7.212
3. Fertilizer (CFA/ha) -- -- -- 9.701
4. Interest on investment

and depreciation on 1,970 2:969 653 2.391

farm tools (CFA/ha)”

5. Hater control costs
(CFA/ha) -- 18.000 32.000 68,000

6. Total economic costs per
hectare (CFA/ha) 16.343 70.899 66.839 168.876

7. Economic cost per kilo
of paddy - 35.7 60.5 133.4 97.3

8. Gross economic value of
production (CFA/ha) 25,848 66,101 28,256 97,910

9. Net economic returns
(8 ' 6) (CFA/M) 9.505 '49798 ‘389583 '70.”

 

 

‘Non-famnly labor is valued at the market wage rate. and famnly labor at its opportunity cost.

bInterest on investment - average invested value (AIV) x interest rate and AIV - £29!1§1§12!_E£E£
hssumnng a zero salvage value). A 35 percent money rate of interest was assumed, giving us
a real rate of interest of 5 percent since the 1980 inflation rate was estimated at about 30
percent (IFAD, 1981).

117

4. SUMMARY

It was found that in System 1, the gross margins ranged from 1,671
CFA/ha to 79,521 CFA/ha. The soybean enterprise had the highest gross
margin and the groundnut and bambera nut enterprises had the lowest
gross margin. For all the six enterprises, the returns per field hour
of family labor varied from 97.6 CFA for rice to 13.9 CFA for bambera
nuts. For the rice enterprise, the returns Of 97.6 CFA/hr of family
labor suggest that there is no financial advantage of family members
seeking wage employment in urban areas when they are needed on their
rice fields (minimum guaranteed wage in urban areas is 90 CFA/hr). Among
the six enterprises comprising System 1, maize showed the lowest total
cost of production, 22.8 CFA/kg and bambera nuts had the highest cost
Of production, 182.8 CFA/kg. Rice showed the second lowest total cost
Of production, 33.6 CFA/kg Of paddy. Using 47 CFA/hr as the opportunity
cost of family labpr. three enterprises in System 1 realized a negative
RLM; they are S/M/C, bambera nuts and groundnuts.

In System 2, the gross margin ranged from 2,119 CFA/ha to 42,711
CFA/ha. The rice enterprise had the highest gross margins and the.
groundnut and cotton enterprises had the lowest gross margins; however,
all enterprises in this system were able to cover their variable costs.
Returns per field hour of family labor in this system ranged from 1.4 CFA
for cotton to 45.5 CFA fOr soybeans. For the rice enterprise, the
return per field hour of family labor was only 18.7 CFA. If the present
costs and returns structure continues, the following shifts could be

expected:

118

a) farmers will put more of their land and labor into soybeans,
okra, S/M/C and rice in that order; and

b) low returns per field hour of family labor for the cotton
enterprise may force farmers to abandon this crop despite the heavy
government support Of this export crop.

The return per hour of family labor alone cannot determine the change in
enterprise mix. Other factors include: (1) fixity of some inputs to
some enterprises (e.g., lowland fields during the wet season can only be
used to grow rice); (2) the family size and the division Of labor within
the family; (3) the labor requirements of different crops as well as
their market potentials(e.g., currently, the market potentials of soy—
beans and Okra are very limited and their home consumption levels are
very low); and (4) the relocating costs and the job search costs. Four
enterprises (cotton, groundnuts, maize and rice) realized a negative
return to land and management. Among the seven enterprises, S/M/C
showed the lowest total cost of production, 27.8 CFA/kg, and cotton had
the highest cost of production, 133.2 CFA/kg. The second highest total
cost Of production was found for the groundnut enterprise, 93.1 CFA/kg,
probably due to the low yield of groundnut (215 kg/ha) in this system.
Rice showed the second lowest total costs of production, 44.1 CFA/kg of
paddy.

It was found that in System 3, gross margin ranged from 6,502 CFA/ha
to 78,254 CFA/ha. The soybean enterprise had the highest GM and the
groundnut/bambera nut mixture had the lowest GM. For all the five enter—
prises comprising this system, except for soybeans and maize, the return
per field hour of family labor was less than the minimum wage rate paid

to unskilled urban workers, i.e., 90 CFA/hr. For the maize and soybean

119

enterprises, returns per field hour of family labor were 116.0 CFA and
450.2 CFA, respectively. For the rice enterprise, returns per field
hour of family labor were 51.4 CFA. As'a result, if the present costs
and returns structure persist, farmers will likely put more of their
land and labor into soybeans and maize if the minimum sorghum needed for
home consumption is attained; and low returns per hour of familylabor
under rice may force farmers to abandon this crop since as a grain, it
is not yet an important part Of the diet. Some incentive structure must
be urgently found if rice growing is to survive in this system where
some important investments in water control have already been made.
Except for soybeans and maize, all enterprises in this system realized
a negative return to land and management. Among the five enterprises,
maize showed the lowest total costs of production (19.1 CFA/kg) and
GN/BN had the highest total cost Of production (228.3 CFA/kg).

In System 4, the variation in GM ranged from 13,655 CFA/ha to
74,109 CFA/ha. The rice enterprise had the highest gross margin
the S/M/C enterprise had the lowest gross margin ; but all the GMs were
positive. In all the six enterprises comprising System 4, except for
Okra, the returns per field hour of family labor was less than the
minimum wage rate paid to unskilled urban workers, i.e., 90 CFA/hr. For
Okra, returns per field hour were 198.2 CFA, while for rice, it was only
24.1 CFA. These low returns per hour for rice may be enough to initiate
the exit process from this fragile industry. All enterprises, except
maize and Okra, realized a negative return to land and management. Among
the six enterprises comprising System 4, maize showed the lowest total
cost of production, 11.4 CFA/kg, and S/M/C had the highest total cost Of
production, 103.0 CFA/kg. The second highest cost of production was

120

found in bambera nuts, 94.0 CFA/kg. Okra showed the second lowest total
cost of production followed by rice (56.8 CFA/kg).

The financial analysis of the different rice production techniques
showed that System 4 yielded the highest gross margins per hectare
(74,109 CFA) but the second lowest returns per field hour of family labor
(24.1 CFA) due to the high labor requirement of this system (irrigated
bas-fonds). The traditional bas-fond (System 1) yielded the lowest gross
margins per hectare (24,915) but the highest returns per field hour Of .
family labor (97.6 CFA). The most expensive way to grow rice was found
in System 3 (69.4 CFA/kg of paddy), probably due to water costs coupled
with low rice yields. The least expensive way Of producing rice was
found in System 1 (traditional has-fonds), i.e., 33.6 CFA per kg of paddy.

The economic analysis of the different rice enterprises from
society's perspective showed that the least cost technique for producing
rice remains traditional cultivation in unimproved swamps. The intro-
duction of water control to date does not compete effectively with this
basic system, and total costs per kilogram Of paddy rise in every case.
The most efficient means Of producing rice under "secure" water control
appears to be by partial water control (System 2). Production on irri-
gated bas-fonds with complete water control and fertilizer use is the
second most expensive with cost per kilogram Of paddy some 173 percent
above the least expensive, traditional bas-fonds, and 96 percent above
the more attractive improved alternative (System 2). When considering
whether rice production in the EORD should be increased by promoting a
particular technique of cultivation, only production in traditional bas-
fonds will be economically justifiable under current technologies and

yield levels.

CHAPTER FIVE

RICE PRODUCTION VERSUS PRODUCTION OF OTHER MAJOR COMPETING CROPS:
‘ AN OVERVIEW OF THE ANALYTICAL MODEL

In Chapter Four, net margins per hectare or per hour of family labor
for the different enterprises in each system were compared, with a pos-
sible view to substituting enterprises with high net margins for those
with low ones. Four important points have to be remembered, however.
First, the different enterprises may be utilizing very different types
Of land: there may be only a limited area of the farm suitable for
growing rice, for example, and dryland crops may be using land totally
unsuitable for rice production because of the rainfall. Secondly, dif-
ferent enterprises Obviously have different requirements for the various
"fixed" resources-~family labor and land. Thirdly, expansion Of an
existing enterprise may necessitate a large increase in some resources,
such as labor, land or even new machinery. And fourth, when based only
on one cropping season data, net margin analysis, though a useful and
widely-applicable tool in farm management, is known to be weak in iden-
tifying optimum solutions under the economic and environment realities
in operation.

For all these reasons, a simple comparison Of net margins per
hectare or per field hour of family labor has only a very limited use
for farm planning and policy orientation, except perhaps as a rough

initial guide. Hence, in order to make realistic evaluations of

121

122

existing cropping enterprises in the different systems under study, an
approach is needed which will simultaneously take into account inter-
relationships among all production processes through their dependence

on common resources. The whole-farm modeling approach was considered
adequate in this context. Such an approach could also be used to assess
short— and long-run consequences of technologies in terms Of social

costs and benefits and macro-level planning Objectives.

1. THE ANALYTICAL MODEL

The approach Of whole-farm modeling is primarily based on the fact
that different production processes depend on common resources. It will
therefore be appropriate to look at the farm as a system. Following the
TAC (1978) report, I

A farm system or whole-farm system is not simply a collection
of crops and animals to which one can apply this input or that
and expect immediate results. Rather it is a complicated
interwoven mesh of soils, plants, animals, implements, workers,
other inputs and environmental influence with the strands held
and manipulated by a person called the farmer who, given his
preferences and aspirations, attempts to produce output from
the inputs and technology available to him. It is the farmer's
awareness of his immediate environment both natural and socio-
economic that results in his farm system.

There exist a wide range of programming models that can be used to

evaluate cropping systems.1

Following Ghodake and Hardaker (1981,
pp. 8-9), methods of whole-farm modeling include, in approximate order

of increasing complexity:

 

1For more details on these models and their applications, see Dillon
and Hardaker, 1980; Barnard and Mix, 1973; Anderson, Dillon and Hardaker,
1977; Anderson, 1974; Hardaker, 1974; and Ghodake, 1981.

123

- whole-farm budgeting

- simplified programming

- linear programing

- linear risk programming

- quadratic risk programming

- linear stochastic programming

- non-linear stochastic programming
4 goal programming

— Monte-Carlo programming

- systems simulation

The different models listed above will not be discussed here;
however, it is useful to consider some criteria that are relevant to
the choice of a particular method for use in cropping system evaluation.
As Ghodake and Hardaker (1981, pp. 9—11) pointed out, a number Of
criteria that are judged to be relevant include:

a) the capacity to handle many constraints and variables, the
need for which arises from the complexity Of agricultural‘
production; b) the capacity to incorporate risk in a realis-
tic way; c) the capacity to incorporate the real goals and
Objectives Of farmers; and d) the need to introduce a cri-
terion of degree Of objectivity, for if the system evalua-
tion performed is to be accepted by scientists, extension
workers and policy makers, they should depend no more than
is absolutely necessary on subjective judgements by the
analyst using the method. Of course, complete Objectivity
is not attainable, but methods do vary in the extent to
which they depend on judgements by the analyst.

Linear programming (LP) was considered adequate as a tool of analysis
in this study; however, no contention is made here that farmers are
always profit maximizers.

Although linear programming is by no means a new technique in

agricultural studies, Heyer (1971) has pointed out that "the large body

124

of literature that now exists on the use of LP in farm production
analysis includes remarkably little about small-scale farmers in devel-
oping countries." According to Hopkins (1975),

This method of analysis seemed particularly appropriate to

a changing situation where new crops and techniques would

not only affect farmers’ incomes, but would also imply re-

percussions in the pattern Of farming activities and

resource allocations too complex to be analyzed by con-

ventional budgeting or other farm planning methods.
LP is therefore used in this study in order to help answer questions con-
cerning: a) the Optimal combination of enterprises that will maximize
total gross margins in light of existing constraints; b) the marginal
value product of each resource and/or constraint; c) the cost of forcing
in non-Optimal activities (or enterprises); and d) the ranges of the
gross margins of different enterprises in the basis for which the Opti-
mal plan remains constant, ceteris paribus.

1.1. Building the "Representative" Farm Model
for Each Production System Studied

Following Collinson (1972), five elements are important to analysis
and planning using the representative farm technique in traditional
agriculture: a) cropping pattern; b) labor profile and supply levels;
c) scale of operation and output; and d) ethnic characteristics and
technology used.

In constructing the initial representative farm for each production
system, labor requirements and supply levels per period, farm sizes, and

2

gross margins per enterprise were averaged across the sampled households

and all activities in the model were defined on a per hectare basis.

 

2Later on in Section 3 of Chapter Six, average labor coefficients
for each production system were replaced with field specific labor
coefficients.

125

The problem with this procedure is that the mean computed for most
characteriStics may not necessarily typify the whole sample Of farms
due to considerable inherent variation among farms. However, it is
not usually possible to obtain a close match between the circumstances
assumed for the representative farm and the circumstances Of any large
proportion of actual farms for each single characteristic. Rather,
following Dillon and Hardaker (1980, pp. 50-51), the representative
farm approach was used here to derive technical coefficients in an
attempt to identify general guidelines about the economical use of

farm resources for farms of a particular type in a given area.

1.2. Model Structure

This section is devoted to a description of the main structural
elements Of the model and the derivation of the numerical coefficients
which it comprises.

According to Heady (1971), there is a homogeneity in the agricul-
tural planning environment among regions and countries in that all farms
have:

(1) plans or goals, (2) limited physical resources such as
land, labor, capital and water which restrain the range of
plans or programs which are feasible, (3) institutional or
subjective restraints which restrict the range of feasible
plans considered or put into actual operation, (4) an Ob-
jective function of some type to be maximized or goal to
be approached, (5) weights which must exist to evaluate or
express the contribution of alternative feasible plans
toward objective function maximization or goal attainment,
and (6) enterprises, technologies or activities which are
competitive in the use of resources.

The model, therefore, is comprised Of an Objective function and a set
of inequalities in which the right-hand side represents a vector of

resource supplies or other constraints. The left-hand side Of the

126

inequalities contains technical coefficients of requirements for these
resources multiplied by variables representing the levels Of enterprises
to be produced. These inequalities define the technologies to be used.
In matrix notation, the model for each production system has.the fOllow-

ing familiar form:

n .
Maximize 2 C.X., j=1,2...7
i=lJ
Subject to

n < ._

.E Aijxj 3 bi’ 1-l,2...23
j-l

.>
XJ __0

where Cj is the return per unit of quantity j allocated,
Xj is the number of units of quantity j allocated,

Aij is the use Of resource i per unit Of quantity j allocated,

bi is the endowment Of resource i.

The resource constraints considered in the model are land and
family labor. Borrowing from Delgado (1979), other resources such as
capital are not dealt with explicitly for three main reasons. First,
the capital cost associated with the cultivation techniques and the
animal traction is minimal since most farmers use hand-tools. Further-
more, sample farmers used virtually no purchased inputs and cash is only
needed for marketing activities of some crops (e.g., soybeans). Second,
the maximum production constraints serve the same purpose as a capital
constraint in cases relating to specific activities (e.g., soybean
enterprise). The MAXSOY constraint effectively limits production Of

soybeans to the area that can be sustained by actual techniques and cash

127

available. If marketing soybeans in distant markets becomes an Option
fOr the farmer, both in terms of financial affordability and physical
availability, then this assumption may have to be revised. Third, the
production factors (land and labor) considered in the model are those
common to all the farmers in the EORD, although the magnitude of use
differs within and between the production systems studied. In any case,
the assumptions made are consistent with the technology and resources

currently available to farmers.

1.2.1. Resource Constraints Used

a) Land Constraints

Following Norman (1973, pp. 5-6), a basic distinction can be made
between lowland and upland. Lowland is usually centered around rainy
season watercourses or swampy areas with poor drainage between mid-July
and mid-September. It supports relatively labor intensive crops such
. as rice, fruits, vegetables and tubers. Since the survey only covered
the rainy season, only rice will be considered as a possible crop to be
planted on lowland fields. Upland can be further divided into two cate-
gories depending on its proximity to human habitation: housefields (HF)
and bushfields (BF).

This division of land into three categories allows us to include
constraints on areas of particular crops or crop mixtures, reflecting,
for eXample, water, distance or fertility considerations. Although all
farmers in the survey area indicated that more land can be Obtained just
by clearing, this statement can only apply to the bushfield category.
However, even in this category, distance and labor requirements mean
that bushfields are not freely available. So, in our models, the bush-

field category Of land was considered as a constraint.

128

b) Labor Constraints

An understanding of the demand and supply of labor is an important
pre-condition in the design of improvements Of small-holder agriculture
in the EORD and other areas in developing countries. Generally, sea-
sonal labor profiles are based on division of the year or cropping
season into planning periods that may be chosen either conventionally,
such as calendar weeks or months, or to correspond with the biological
timetalbe of field activities (e.g., land preparation, planting, weed-
ing, harvesting). Although the latter approach was used in collecting
the survey data, conventional planning periods were used in the linear
programming mOdel.

Seventeen labor periods for the model were defined (Tables 5.1-
5.4). The length of each labor period was established by analyzing the
crops' cycles as reported by Lassiter (1981, p. 20), and labor profiles
for the main crops and crop mixtures found in each production system
during the survey. Because the LP solution alogrithm treats the entire
time period as a single point in time, making no distinction between
the beginning and the end of the period, as Crawford (1980) pointed out,
it was necessary not to allow labor needed in one period to be drawn
from a different period. For example, if labor is needed in the first
two weeks of June for planting, the program should not be allowed to
draw from labor available in the second part of June. This was accom-
plished by narrowing down the labor period to two-week periods around
the most critical field activities for the major crops considered under

each system of production.

129

TABLE 5.1

MAIN LABOR PERIODS AND ACTIVITIES COVERED,
SYSTEM 1. 1980

 

 

 

Labor Heeks Oates ’ a
Period Covered Covered. 1980 Principal Field Activities

1 1-18 Jan. l-Mey 4 Land preparation/grains (sorghumlmillet and maize)

2 19-20 May 5-May 18 Land preparation] rain (sorghumeillet and rice)
planting/grain sorghumlmilleta and rice)

3 21-22 May 19-June 1 Land preparation and planting/groundnuts +
activities in period 2

4 23-24 June 2-June 15 Activities in period 3 oont'd

5 25-26 June 16—June 29 Activities in period 4 cont‘d + planting/bambera nuts

6 27-28 June JO-July 13 Handing/grain (sorghumwmillet and rice). planting]
maize and seybeens

7 29-30 July l4-July 27 Needing/grain. weeding/soybeans

8 31-32 July ZC-Aug. 10‘ Planting/cowpeas. weeding/grains

9 33-34 Aug. 11-Aug. 24 Heading/grain, weeding/groundnuts, fertilization/
soybeans

10 35-35 Aug. 25-Sept. 7 Relative slack, weeding/grains

11 37-38 Sept. O-Sept. 21 Activities in period 10 cont‘d

12 39.40 Sept. 22-Oct. 5 Harvest/grains (rice + maize). further weeding]
wmeu.Mwuuwmeu

13 41-42 Oct. 6-Oct. 19 Harvest/grains + cowpeas, harvest/Mara nuts.
harvest/soybeans

14 43-44 Oct. ZO-Nov. 2 Harvest/grains a cowpeas

15 46-46 Nov. 3-Nov. 16 Harvest/grains + nuts

16 47-48 How. 17-Nov. 30 Harvest/grains

17 49-53 Dec. l-Oec. 31 Slack. harvest/grains

 

 

 

t

 

'Note that most field activities run across more than one period even though they are not always

repeated in the table.

130

 

 

 

TABLE 5.2
MAIN LABOR PERIODS AND ACTIVITIES COVERED,
SYSTEM 2. 1980
m; ' m ' «"23?an Principal Field Activities.
1 1-18 Jan. Hey 4 51:2)1and preparation/grains (sorghum/millet and
2 19-20 my 5-May 18 Planting/grains
3 21-22 nay is.»- 1 Planting/grains. lane preparation/maize
4 23-24 June 2-June 15 Heading/grains. planting/size and cowpeas
5 25-25 Jame 15-Jime 29 Planting/groundnuts and okra, land preparation]
. m
'6 27-25 June n-July 13 Heading/grains. planting/groundnuts. planting/cotton
7 ‘25-! July 14-July 27 ‘ meg/grains. planting/(soybeans. cotton. bauera
a 31.3: July za-Aua. 10 Activities in period 7 oont'd ‘
9 33034 M. ll-Aug. 24 WHO/Mt:
10 35.36 Aug. ZS-Sept. 7 Needing/soybeans and cotton
11 ai-sa sapt. O—Sept. 21 Harvest maize. weeding oont'd. harvest/cowpeas
12 39-40 Sept. 22-Oct. 5 Further weeding a relative slack
13 41-42 Oct. 5-Oct. 19 Harvest/Ilse. harvest/sorgh-
14 43-44 ' . Oct. ZO-Nov. 2 Harvest/grains. harvest/okra
15 45-46 Nov. 3-Nov. 16 Harvest/grains and cotton
16 47-48 lbv. l7-Nov. 30 Harvest/grains and nuts
17 49-53 Dec. l-Oec. 31 Slack. harvest/grains and cotton

 

 

 

 

‘llote that most field activities run across more than one period even though they are not always
repeated in the table.

 

 

131

 

 

 

TABLE 5.3
MAIN LABOR PERIODS AND ACTIVITIES COVERED.
SYSTEM 3, 1980
35°; .23.“.1. mgr,” principal Field Activities'
1 1-18 Jan. 1-May 4 Slack, land preparation/(SINK) and rice
2 19.20 my Sony 18 Planting/grains
3 21-22 May 19-June 1 Planting/grains and cowpeas
4 23-24 June 2-June 15 Planting/groundnuts, maize, weeding/grains
5 25-26 June l6-leie 29 Needing/grains. planting/soybeans
6 27-28 June 30-July 13 Needing/grains and groundnuts. planting/We nuts
7 29-fl July l4-July 27 Needing/grains
a ' 31-32 July za-m. lo ' Activities in period 7-oont'd
9 33-34 Aug. ll-Aug: 24 ' Relative slack a further weeding
10 35-36 Aug. 2.5-Sept. 7 Needing/grains. hanesting/groiwldnuts
11 37-38 Sept. O-Sept. 21 Harvesting/min. 4' further weeding
12 39—40 Sept. 22-Oct. 5 Harvest/maize 4- groundnuts
13 41-42 Oct. 6-Oct. 19 Harvest/(Mere nuts + cowpeas r grains)
14 43-44 Oct. 2041». 2 Harvest/(grains + soybeans)
15 45-46 Nov. 3-Nov. 16 Harvest/grains (rice. sorghu. millet), harvest/nuts
16 47-48 Nov. 17-Nov. 30 Harvest/grains
17 49-53 Dec. 1-Oec. 31 Slack. harvest/grains and nuts

________——————=—-————————————___—————————————
'llote that most field activities mm across more than one period even though they are not always
repeated in the table.

132

. TABLE 5.4

MAIN LABOR PERIODS AND ACTIVITIES COVERED.
SYSTEM 4, 1980

 

 

 

3:; cm“; “wafflm Principal Field Activities.
1 1-18 Jan. 1-May 4 Slack. land preparation/grains
2 19-20 May 5-May 18 Planting/grains
3 21-22 May l9-June 1 Planting/cowpeas. planting/grains and okra
4 23-24 June 2-June 15 Planting/(maize r groundnuts). weeding/sorghu-millet
5 25-26 June 164m 29 Planting/9W 4 activities in period 4 cont'd
6 27-28 June SO-July l3 Planting/bambera nuts. weeding/grains and groundnuts
7 29-30 July 14-July 27 Heading/soybeans; weeding/grains and okra,
- fertilization/rice
. 6 31-32 July ZB-Aug. 10 Activities in period 7 oomt'd
9 33-34 Aug. ll-Aug. 24 Needing/grains. fertilization/rice
lO as-as Aug. 254.». 1 Further weeding/grains
11 37-38 Sept. 8-Sept. 21 Harvest/maize. weeding/grains
12 39-40 Sept. 22-Oct. 5 Harvest/groundnuts + maize
13 41—42 Oct. 6-Oct. 19 Harvest/okra + groundnuts, harvest/grains
14 43-44 Oct. 20-Nov. 2 Harvest/grains. harvest/Mara nuts
15 45-46 Nov. 3-Nov. 16 Harvest/grains r nuts
16 47—48 Nov. 17-Nov. 30 Harvest/grains 9 relative slack
17 49-53 Oec. 1-Oec. 31 Harvest/grains. slack

 

 

‘lbte that most field activities run across more than one period even though they are not always

repeated in the table.

133

c) Production Constraints

In addition to the limits imposed on output by resource supplies,
the model incorporates some direct constraints on the level of certain
enterprises like sorghum/millet, maize, okra and soybeans. The minimum
and maximum output levels serve to express limitations other than those
Of the basic land and labor constraints; indirectly, they represent
scarce resources which are relevant to the sorghum/millet, maize, soy-
bean, and Okra enterprises.

Farms in the EORD are generally characterized by a strong subsis-
tence orientation. Commonly, a significant proportion of family food
needs is produced on the farm. Thus, the general level Of health and
welfare Of the members Of the household is strongly dependent on the
degree Of success achieved in farm food production. SO, despite the fact
that the main purpose Of an LP model Of farm behavior is to identify pro-
duction strategies which maximize net farm revenue,or total gross margins,
the model also needs to be realistic by incorporating other important
household Objectives. This is why the minimum food grain constraint was
introduced here; sample farmers typically will not rely upon the market
for their supply Of the food staple, sorghum/millet. This objective was
specified in terms of the minimum foodcrop area farmers are comfortable
with, as opposed to the area necessary to feed the family in an average
season. The problem is to specify this level correctly, in order that
the model may give realistic results.

Maximum production constraints ensure that the Optimal program only
includes levels of activities that are plausible in the real world.
However, we want to keep these production constraints to a minimum in

order to allow some flexibility to the model to choose freely from

134

existing alternatives, given the real resource constraints. The limit
on maize takes into consideration special soil characteristics required
by this crop which are not found on all housefields. Maize is typically
planted immediately outside the house. This is the most fertile soil
Of the farmer's land holdings which has been receiving the manure since
the establishment of the household compound. Without the production
constraint, MAXMAI, the program would be free to allocate the entire
supply of housefields to maize, ceteris paribus, even though in practice
farmers will not do so. Here again, the principal problem is to know
the correct level to specify the maximum.

The remaining ceiling applies to the soybean enterprise. In the
real world, soybean production is limited by processing facilities and

marketing outlets not included directly in the model.

1.2.2. Activities and Objective Function

The objective function Of the model used for each production system
involves maximization of total gross margins subject to resource use 1
constraints and production constraints.

The choice of possible crop activities is limited to mixtures which
are typically grown in the survey area as was discussed in Chapters
Three and Four. There are nine major crop categories. These are rice,
sorghum/millet/cowpeas, maize, groundnuts, bambera nuts, soybeans, cotton,
okra, and groundnuts with bambera nuts. Survey data on the allocation
of land to different crop activities is shown in Table 5.5. The high
proportion of land put under S/M/C appears to reflect a concern with
assuring an on-farm supply of staple foods. This allocation of land to

different crop activities, besides its intrinsic interest for agricultural

135

TABLE 5.5

MEAN PERCENTAGE OF HOUSEHOLD LAND ALLOCATED
TO EACH CROP CATEGORY, 1980

 

 

Mean Household Percentage

 

 

 

Cro
Categgry System 1 System 2 System 3 System 4
Rice (LF only) 31.4 54.3 16.0 32.2
S/M/C (HF+BF) 39.3 37.1 45.2 57.5
Maize (HF only) 14.1 3.4 12.3 3.1
Groundnuts (HF+BF) 8.0 2.2 - 4.1
Bambera Nuts (HF+BF) 5.1 - - 5
Soybeans (8F only) 2 O 1.6 6 5 -
Cotton (BF only) - l.l - -
Okra (HF only) - 3 - 2.5
GN/BN (HF+BF) - - 19.9 -
100.0 100.0 100.0 100.0

 

 

LF stands for lowland fields

HF stands for housefields

BF stands for bush-fields

S/M/C stands for sorghum/millet/cowpeas

GN/BN stands for mixture groundnut/bambera nuts

136

planning, will also be useful for comparing the actual allocation
portrayed in Table 5.5 with the Optimal allocations suggested by the

LP model later on.

1.2.3. Derivation of the Numerical Coefficients Of the Model

a) Labor Coefficients and Restriction Levels

The calculation of the numerical coefficients for the different
enterprises is a crucial process in the design Of any linear programming
model. "As with the formulation of representative farm types, a major
issue here is whether to calculate average coefficients, or to use actual
coefficients from representative enterprise types, or to devise synthetic
coefficients based on subjective evaluation Of the data." (Crawford,
1980).. Another approach suggested by Balcet and Chandler (1981) would
be to estimate crop yields and labor requirements for the different
enterprises by using multiple regression techniques. This latter method
was ruled out because it was not possible to measure the area planted to
different crops within a mixture type Of each field. As a result, labor
input per hectare could not be associated with specific crops within the
mixture. Hence, the initial approach used in this study was to compute

averages3 per hectare for each enterprise type. The total number of

 

3The problem with this procedure, as Collinson (1972, p. 134) poinuxi
out, is that "interfarm differences in timing create different peak re-
quirements on particular farms, which are damaged when averaged--and peaks
on one farm are Offset by relatively slack periods on another, so the
whole labor profile is flattened." The effect Of smoothing labor peaks
is to reduce the incidence and size of seasonal labor bottlenecks. The
implication of using figures for a mean household, then, is to lower the
opportunity cost of rice in terms of foregone production of other crops.
This is because this cost is incurred only as a result of the realloca-
tion of labor during peak periods from other crops to rice. As peak labor
requirements for crops are reduced, so is the opportunity cost Of grow-
ing rice.

137

hours allocated by each household to each major crop category was
calculated by period for the 17 periods defined above.) The figures for
each household were divided by the total household land area in hectares
devoted to the crop category in question. The ratio Obtained for each
period, household and crop category was averaged over households to give
the mean total household hours allocated to each major crop category by
period. Accordingly, the basic model employed enterprises defined in
terms Of these averages. Comparisons of crop yields and total labor
requirements for these enterprises with the figures reported from
similar geographical areas by Lassiter (1981) indicated general con-
sistency. A comparison of the results from the basic model with results
from an alternative model incorporating labor coefficients based on
individual fields will be discussed in Chapter Six.

Now, turning to the supply of family labor, in principle, estimating
this supply is quite straighthrward. The total labor time available in
any period is found by adding for each available worker the time he or
she can allocate to cropping activities in that period. In practice,
however, while it is usually easy to determine the number of workers-
available, estimating the available labor time of each can present some
difficulties.4 The average size of the family labor force in the areas
surveyed was fairly variable, ranging from 5.3 to 7.6 workers per house-

hold for all the systems studied (Table 5.6).

4It should be kept in mind that total household labor availability
can be broken into five sectors: crop, livestock, domestic, non-
agricultural work and social activity. This study focuses just on crop
labor use. The labor available for each period for cropping activities
is constrained in the model by the number Of work days and the number Of
people in the household available to help on the farms. Five working
days per week is assumed to be the available number of work days for
cropping activities. A seven-hour day of 7 a.m. to 11 a.m. and l p.m.
to 4 p.m. (Systems 1, 2, 3) or of 7 a.m. to 2 p.m. (System 4) as reported
by the farmers defines the available hours of labor per day.

 

138

.apo>wuumammg .e was m .N .p msmumaw c? m~wm upozomso; any we ucmucwg om ucm acmugma
mu .ucmugma mm .ucmocma mm usonm umuzupumcou mgmasms upozmmzo; mcwxgozucoc ugh .mugow Loam, »__smm
as» :_ umnapocp ac: ago; Adam mo «com» me gm>ov cos upo new Ammo mo memo» cm>om cusp mmwpv coca—Pgo

Ppmsm .mogom Lona. xpwsam us» do mNPm on» ecu upogmmso; on“ co mem use cmwzumn mums mm: cowuocppmrc
a .cm>mzo= .apo>ppumqmmc .e ucm m .N ._ msoumaw :_ m.~ new N.“ .m.~ .m.op mm: m~_m upogmmzo;_mmmcm>m
on» .vFozmmao; as» we cowupmoasou mg» so new m~pm ms» co mucmgmu mgmxgoz mo conga: wmmcm>m wshp

 

 

. . . . mcmxcoz we
3 . mm 2 a 3 2 _ em 2 m ..Z is; $23

 

xm: .cpz com: .xmz .cpz can: .xc: .=.z cum: .xmz .cwz cam:

 

>~ HHH _H H

 

copaozuosa mo msmumam

 

 

ommp .am0m mzp z~ >o=hm macz: m2uhm>m zonhusooma
«so; mzh 44< zH ogozmmzo: mum mmmxmoz mo ammzaz mw<mm><

o.m m4m<h

139

The procedure used to derive the exact number of hours of family
labor available for crops over each period Pj is based upon the house-
hold averages contained in Table 5.6. The labor availability for crops
in period Pj is set equal to the maximum amount of hours available per
worker per week for crops during the jth period of the year times the

average number of workers times the number of weeks in period Pj.

f3 = AWE x 35 x Lj

Where fj = total family labor available in period Pj
W3 = average number of family workers available in period P j
Lj = length of the period Pj expressed in weeks,

and where 35 represents the maximum amount of hours available per family
worker per week for cropping activities.

In Systems 2 and 3, there were no children in the sample going to
school and no migration of adult workers was observed during the dry
season. So, Nj was constant throughout the year in these two systems.

So, for each labor period covering two weeks (LABOZ through LABl6),
fj-= 5.6 x 35 x 2 = 392 hours (for System 2) and fj = 5.7 x 35 x 2 =
399 hours (for System 3). In System 2, since LABOl covers l8 weeks, for
that labor period, f5 = 392 x %§-= 3,528 hours; and in System 3, for the
labor period LABOl, fj = 399 x 129- = 3,591 hours. LABl7 covers four weeks
in both systems. So fj is equal to 784 hours (392 x 2) and 798 hours
(399 x 2) in Systems 2 and 3, respectively.

In System l where children represented about 39.5 percent of the
household size and where 5 percent of them attended school in the village,

it was estimated that 2 percent [(.395) (.05) = .02] of available family

140

labor for crops was lost during the school year. Similarly, it was
estimated that in System 4, 4 percent of available family labor for crops
was lost during the school year. In System 1, during two out of the l8
weeks comprising the labor period LABOl, there is no loss of available
family labor due to schooling. So total labor available during this
period (LABOl) is (521 x %§) + 532 = 4,700 hours. During the period
running from mid-June through the beginning of October (LABOS through
LABll), fj = 7.6 x 35 x 2 = 532 hours for each labor period. During the
periods running from May to mid-June (LABOZ to LABO4) and from the
beginning of October to the end of November (LABlB to LABl6), children
are in school--so available labor every two weeks is: fj = 532 -

(.02) (532) = 521 hours. During the month of December, children are
still in school the first two weeks, so available labor is 52l; but in
the last two weeks of December, children are on Christmas vacation, so
available family labor is 532 hours, which makes total labor available
for the period LABl7 equal to: fj = 521 + 532 = l,053 hours. Using

the same procedure as in System 1 above, the amount of family labor
available per period in System 4 was found to be as follows: f5 =

5.3 x 35 x 2 = 371 hours for j'= LABOS through LABll; fj = 371 - (.04)
(371) = 356 hours, for j = LA802 through LABO4 and LABl3 through LABlG;
fi = (355 x 1;) + 371 = 3,219 hours for j = LABOl and f3. = 356 + 371 --
727 hours for j = LABl7. Tables 5.10 - 5.l3 contain all the information

on labor coefficients and their right—hand side (RHS) values.

b) Land Restriction Levels
Housefields, bushfields and lowland fields are the three land types
specified in the model. The average housefield area per farm was used

as the total land available in this category (Table 5.7). Lowlands and

141

TABLE 5.7

SUMMARY OF AREA AND PROPORTION OF EACH TYPE OF LAND
FARMED BY FARM SYSTEM, 1980 SURVEY

 

 

 

Percentage of Total Average
Systems Land Category Household Area Field Area
Cultivated in 1980 (ha.)
1 Housefield (HF) 47 1.6
Lowlandfield (LF) 30.7 1.1
Bushfield (BF) 22.3 .85
2 HF 25.7 .9
LF 54.3 1.9
BF 20.0 .7
3 HF 42.7 2.7
LF 15.8 1.0
BF 41.4 2.62
4 HF 64.5 1.2
LF 32.2 .6
BF 3.2 .06

 

 

bushfields that may be cropped were also limited under current practices.
So, average field areas in these categories were used as the total land

available in these categories.

c) Production Restriction Levels
The minimum food grain area in this study was derived using actual
farm plantings of sorghum/millet/cowpea (S/H/C) in 1980 in terms of both
_ absolute acreage and proportion of total land under cultivation suitable
for S/M/C growing. The minimum food grain area for each production
system was computed as the smallest percentage of land under S/M/C in
the sample times the area of the representative farm (Table 5.8). This

procedure ensures that the results are commensurate with the scale of

142

TABLE 5.8

MINIMUM AREA UNDER SORGHUM/MILLET/CONPEAS FOR
THE FOUR PRODUCTION SYSTEMS, 1980

 

 

 

Systems T?::I+Lgpg Minimum Percentage Minimum Area Under
nder S/M/C S/M/C (in ha)
1 2.45 21 .51
2 1.6 13 .21
3 5.32 27 1.44
4 1.26 5 .08

 

 

the representative farm. Additional care was also taken to ensure that
the minimum percentage selected was not an outlier case--for example,

in System 4, the smallest percentage of land under S/M/C happened to

be zero. However, the household in which this percentage occurred
contained only one person, which was not typical of the system under '
study. Hence, the next smallest percentage, i.e., 6 percent, was
selected. The same concerns were kept in mind in the selection of maxi-
mum production levels. '

The wide variation in the minimum area under sorghum/millet between
systems shown in Table 5.8 may indicate that in some systems of produc-
tion farmers are beginning to rely more and more on the market for their
food-grain supply and/or that sorghum/millet is becoming less important
in the diet of some farmers.

The maximum production constraint for maize was specified as either
the maximum percentage of housefields in the sample attributed to maize ‘
times the area of housefields on the representative farm, or the maximum

household area across the sample in maize, whichever is smallest

143

(Table 5.9). This procedure ensures that the chosen output ceiling is
both a maximum based on the sample data and that it reflects the scale

of the representative farm.

TABLE 5.9

MAXIMUM AREA THAT CAN BE GROWN IN MAIZE AND SOYBEAN
FOR FOUR PRODUCTION SYSTEMS, 1980

 

 

 

 

Crops
Systems Maize Soybeans
(max. area in ha) (max. area in ha)
1 .396 .242
2 .655 .373
3 .523 .490
4 .482 -

 

 

Finally, given the current market conditions and the assumed labor
coefficients, the maximum soybean level was computed as either the maxi-
mum percentage of bushfields in the sample attributed to soybeans times
the total area of bushfields on the representative farm or the maximum
household area across the sample in the soybean enterprise, whichever

was smallest (Table 5.9).

d) Objective Function Coefficients

The gross margins from one hectare of each crop enterprise were
used in the basic model (TablesS.lO-5.l3) as the objective function
coefficients. These figures were taken from Tables 4.5.1, 4.5.2, 4.5.3

and 4.5.4, and they are all expressed in CFA per hectare.

144

2. TABLEAUX FOR THE FOUR REPRESENTATIVE
FARMS OF THE SYSTEMS STUDIED

The tableaux of the basic models for the fbur production systems
are displayed in Tables 5.lO-5.l3. For each table, activities (or
enterprises) run across the top of the table. Right below the enter-
prise names, objective function values expressed in CFA/ha are found.
The very first column gives the names of restrictions (resource con-
straints or production levels) imposed on the model. The column labeled
81's contains the levels of resource supplies or production levels which
should be met or cannot be exceeded. The figures in the body of the
matrix or table are the a... input-output coefficients expressed on a

1J
per hectare basis.

145

 

an on . p o o o p o e

 

2 .... . e a o o o o

2 ... n o c . a o o

2 3.... . o . 2 a a a

2 .2 . u . I e ~ .

2 .2 . 2 2 2 a 2 2

2 .2 . .. 2 .2. 8 .. 2

2 .2 . . . 2 8 . . 2 2 . 22.
2 a... . o o 2 2. a u a 8 . 22.
2 «a . 2 2 .2 2. 2 2 8. .2 o :2.
2 «3 . o 2 2. .2 o 2 o 2. . 2.2
2 «a . 2 2 2. 2. 2 2 2 2. 2 82..
.. 5 . a 2 .2 2 a 2 8 .2 8 82.
2 5 . .n 2 .2 .2 ... 2 8.. .2 2 .85 .
. 5 . u 2 .2 8 u 2 2 .2 a 82.
a 2.. . 2 2 .2 a 2 2 .2 2. 2 8.2
. .2 . o o 2 .. . o a o . 2 2 82..
o .2 . a n 8 a a n e 2 . 83
a .2 . o o 2 e e o . 2 . 82..
. 8... . a . ... o a o a = o .3:

mafia.

. 8. . . . . . o o o a a :2 8.22....-
~ ... . o a o o . o e o . .2. 0.285....
. e . . o o o . o . . . . e .2. 8.2.8....

..1. 3

.8333!

 

    

p3.— ~3£ out! .3... 3o.- ~3£ 5.3 38.3 30.3
...... .... .22.»... .... .... .... .2 2.

3‘ .n a». .0 oz» 3 3‘ .0 8.! an" 3.-

 

 

a. .. gun .i :3 in 8n 2‘
3.. iv

146

VII AI

mm M
E .
ma —

was» so..u..umo¢ canon
.33: 9.33.. .2. a ..o 0... 2.3.3.... .2.

 

 

 

   

.. 2. . a o . c o o o c c 23....
.. ... . o o o e . o a o c S2...
.. 2. . a o e a o o . o o =5!
.. ... . o . a o o e a . a 9......
... ... . a .. .. .. o o o .. 8 ......
2 .2 . 2 8 a c a 2 a .. .2 2.5
2 .2 . 8 .. . S .... .... S . .. 8. 29:
.. .2 . .... . ... o 8 .... m . a 22..
2 .2 . a . .. o a o 2 . a 22:
2 .2 . . ... n. .. n. . 8 ... 2 25:
2 .2 . .. .. .2 ... a .. 2 .. .. ....S
2 .2 . a. n. n. 8 .. a. .. n. 2. 2.5.
2 .2 . .. .. .... .. 2 .. 8. .. ... 82..
.. .2 . 8.. .. ... 8. ... ... ... ... .2 8...:
2 .2 . ... .. 8 .2 ... ... .2 .. 8. .85
a .2 . u. o. .2 e a a. 2.. ... 2. 82..
. .2 . o 2 .. a .. . .. 2 2. 82..
. .2 . o .o a o a o .. .. 2. .85
o .2 . a 2 o o e .. .... 2 ... 89:
... .2 . o .... o a a o .. .... .2 .8...
. ...... . o .. o o o o o .. a .8...
. _msamz

35932.8 .3...
. .... . . . o . . o o a c :2 22......-
. 8.. . o o a o o o o a . ...: «22.2233
. 8. . o a . c o . . . a C... ...-:88:

...... 252:8... ...:

. 8.. .... 8.. 8.... 2.... 2... 8 ... 8.. .2... .2... ...... 2.2.2.2...

.8— 35»... to ...—3 9.8 .838 on A! 0...: At: 3:.
3.... .w ”5... "52m ......»33

33. .w

 

 

 

82 .w 55»... £5.83; :3 52.. gm “:3 :6 03-.—

147

 

M u. n

u an ~

v ... .
2.2 83233.. +022.

.uvnca Obauuog LOG I :0 0L. mflwu—>-uu~ ——<l

 

 

 

 

2 .... . o a . .. ... ... o .62....
.. .... . a o .. o . o o . .15...
.. .... 2 o . o o .. . o 9......

F 22....

mac—ahumcou covuuatOLa
... 2. . .. 2 o .. o 2 . ......
2 2. . ... ... . ... .. ... 2 2....
2 2. . .. 2 . .. .. 2 2. . 2....
.. ..2 . 2. 2 a 8. . 2 ... 29..
2 .2 . 2 m . 2 .. m 2 2......
2 .2 . 2 2 .. 2 2 2 ... ......
2 .2 . 2 .. .. 2 . .. .. ......
2 2. . 2 2. 2 2 2 2. 2 2...:
.. .2 . .. .2 2 .. 2 .2 2 2....
.. 2. . .. 2. .2 .. .. 2. .. 8...:
2 2. . 2 ... 2 2 2 ... 2 .2...
a 2. . 2 .2 . 2 .. .... 2 89..
a .2 . 2 .2 . 2 2 .2 2 3....
. ..2 . n 2 .. n 2 2 .. .8...
o 2. . . ... o . . ... o 8...:
m 2. . o 2 .. o o 2 .. .8...
. .22 . o 2 .. o o 2 .. .8...

mg:

mac—...mcou so...
. .... . . . . .. .. a .. .... 3.2.5....
. .... . o .. a o a a . .... 2.2.228.
. ... . a a o . . . o C... 3.2.3.5..

...... 3552.8 ......

. ...... .... .2... ...... .2... .2... ...... 2.... .2... 2.... 23:23....

8.3.5.... .... .... .8 t... 2.... .2. 8...
......a ...... 2.... .....m .3232...

  
  

 

 

coo. .n :uhm>w .t<¢ua¢s ¢<uz~a 23¢. neon u~m<¢ ~—.m o—aah

.

148

.8
new
.3

n
u
—

g u...
...-3 :30, so. 0 ‘ 0... «0.3.39- :1-

 

 

 

 

 

  

 

2 2. . o o o . .. .. a o o
2 a. . . a o c o o . .. a
2 8 n o c . o e o o . ..
... ... . e .. . o . o o e . ...
2 .2 . o a 8. o . o .. 8. .2
2 .2 . 3. 2 2 o 8. 2 a 2 2. 2.2
.. .2 . a 2. 2 . e . 2. 2 2 2 2....
2 .2 . o 8 o o. .. ... .. o 2. 2!.
2 ... . o . 2 n o . 2. 2 .2 ...: .
2 ... . o .. ... a a .. .. ... 2. ...:
2 ... . ... 8 .2 . ... .- ... .2 .2 2.2
.. ... . 8. .. ... .. 8. .2 .2 ... ... 8....
.. ... . I. a .2 2 I. .- .. 2. .2 8!.
2 ... . 8 2 ... 2 I 8 I. ... .2 .82
. ... . .2 2 .. 2 .2 2 .. .. ... 8...
. ... . 8 2 8 2 2 2 a. a. ... 2!.
. .2 . o a 2 2 o o . 2 ... 82.
. .2 . c a 2 . o n .. 2 2 . 82.
. 2. . a a 2 o o o e 2 2. ...-2
. 2... . o o . a o o e . a. .82
if...
n 8 . . . . . e o o o .. .2. .223...-
. . . e a c o e o o o . .2. 8.2.13.8.
. . . . . o o . . . . . .. ...... 8.2.8....
...... ......
8.... .8... ...... .8... 8...” .8... .8... 8.... .2... 5.2:...-
. ...... .... .... ...»... 3..."... an... .3” :8 8...“... aka... 8.... an.» 8..

 

 

 

‘— .' E2 is a: in g 93'

2.. a!»

CHAPTER SIX

EVALUATION OF THE BASIC MODEL, PRESENTATION OF
RESULTS, AND SENSITIVITY ANALYSIS

Before discussing the results from the basic model, it is appropriate
at this point to examine very briefly to what extent this model is an
accurate representation of the true system from the standpoint of crop-
ping pattern and labor use. So, the evaluation of the basic model will
be discussed first, then in section 2, results from the basic model will
be presented; in section 3, various experiments with the basic model
will be discussed; and finally, in section 4, a summary of the results

emerging from the modeling exercise will be presented.

1. EVALUATION OF THE MODEL

As Crawford (1980) pointed out, aconmon problem with linear pro-
gramming models is their tendency to produce solutions in which crOpping
patterns are highly specialized. These are very uncommon in the EORD.
So the issue here is how closely the solutions to the basic model re-
flect actual cropping patterns? Table 6.] compares the optimal alloca-
tion suggested by the model with the actual allocation. In System l
where the actual allocation of land to different enterprises comprised
six enterprises, four are included in the optimal solution mix. Ground-

nuts and bambera nuts which actually made up l3 percent of land allocated

T49

150

TABLE 6.]

ACTUAL VERSUS OPTIMAL ALLOCATION OF LAND TO EACH CROP
CATEGORY, BY SYSTEM, 1980 (percentages)1

 

 

Representative Farm

 

Crop Category
System l System 2 System 3 System 4

 

 

 

 

Rice 31 (31)1 44 (54)' 19 (15) 32 (32)
Sorghum/millet/cowpeas Sl (39) 40 (37) 30 (45) 4 (57)
Maize 11 (l4) 0 (3) TO (12) 26 (3)
Groundnuts 0 (8) 0 (2) - - 2 (4)
Bambera nuts 0 (5) - - - - 14 (l)
Soybeans 7 (2) l3 (2) 9 (6) - -
Cotton - - 0 (l) - - - -
Okra - - 3 (I) - - 22 (2)
Groundnuts/bambera nuts - - L - - 32 (20) - -
Total lOO 100 l00 100
I

The number in parentheses show the actual allocation of land in
percentages to each crop category. ,

to crops were excluded from the optimal solution. S/M/C and soybeans
saw their share of land in the optimal allocatiOn increased by 30 and
250 percent, respectively. The maize shareof land was reduced in the .
Optimal solution by Zl percent, while the area under rice in the opti-
mal solution remained about the same as in the actual cropping pattern.
In System 2 where the actual allocation of land to different enterprises
comprised seven enterprises, four were included in the optimal alloca-

tion. Maize, groundnuts and cotton, which actually made up about

151

7 percent of the total land under cultivation were dropped from the
optimal solution. The percentage of land under soybeans, S/M/C and

okra increased by 550, 8 and 200 percent, respectively, while the per-
centage amount of land under rice in the optimal allocation decreased

by 18 percent. In System 3, all five enterprises included in the actual
cropping pattern were included in the optimal solution. SIM/C and

maize saw their percentage decreased by 33 and 17 percent, respectively,
while soybeans, rice and GN/BN saw their percentage increased by 50, 18
and 60 percent, respectively. In System 4, while all the six enterprises
originally present in the current farm plan were also included in the
optimal farm organization, there was, however, a 93 percent decrease in
the percentage amount of land allocated to S/M/C in favor of maize,‘
bambera nuts and okra whose percentage in the optimal solution were
increased by 766, 1,300 and 1,000 percent, respectively. So, it can be
concluded that results of the optimal plan are quite comparable with
those of the actual plan in terms of the number of enterprises included
in the solution but the percentage of land area allocated to each crop .
category varies quite a bit.

Another important issue in deciding whether a model is believable
and an appropriate one is the question of how closely does the labor use
reflect the true system under study. Table 6.2 compares the labor use
suggested by the model with the actual labor use. In System 1, total
labor uSe suggested by the model is 25 percent higher than the actual
labor use, whereas in Systems 2, 3 and 4, total labor use implied by the
model is 28, 29 and 19 percent lower than the actual labor use, respec-
tively. Tables 6.3 to 6.5 contain information on average family labor

use as well as labor use on some specific rice fields for all the

TABLE 6.2

152

TOTAL LABOR USE UNDER ACTUAL ALLOCATION VERSUS
TOTAL LABOR USE IMPLIED BY THE MODEL BY CROP
CATEGORY AND BY SYSTEM, 19801

 

 

Labor Use by Representative Farm (hrs)

 

 

 

Caggggry System 1 System 2 System 3 System 4
Rice 271 (273) 2,746 (4,201) 517 (517) 1,786 (1,786)
Sorghum/

millet/

cowpeas 2,738 (2,103) 628 (716) 2,169 (3,852) 121 (1,539)
Maize 215 (269) 0 (137) 143 (216) 403 (50)
Groundnuts 0 (54) 0 (62) -. - 27 (54)
Bambera nuts 0 (19) - - - - 395 (15)
Soybeans 270 (79) 296 (41) 83 (68) - -
Cotton - - 0 (32) - - - -
Okra - - 88 (9) - - 58 (7)
Groundnuts/

bambera nuts - - - - 302 (505) - -
Total 3,494 (2,797) 3,758 (5,198) 3,714 (5,258) 2,790 (3,451)

 

 

1
category, by system.

The numbers in parentheses show the actual labor use, by crop

153

TABLE 6.3

FAMILY LABOR PROFILE FOR SELECTED FIELDS BELONGING
TO THE RICE ENTERPRISE, SYSTEM I, 1980

 

 

Hours per Hectare

Labor Field Reference

 

 

 

Period* Average
' 18/7** 6/7

1 (18)

2 (2) 5

3 (2) 4

4 (2) 29 81

5 (2) 28

6 (2) .33 909
7 (2) 14

8 (2) 28 150

9 (2) ‘ 55 222
10 (2) 6

11 (2) 6 263
12 (2) l

13 (2) 4

14 (2) 15 122 182
15 (2) 15

16 (2) 3

l7 (4) 0
Total (hrs/ha) 246 353 1,576
Area (ha) .411 .36 .456

 

 

*The numbers in parentheses represent the length of the period
expressed in weeks.

**Fie1d labor profile used in Version A of the model.

154

TABLE 6.4

FAMILY LABOR PROFILE FOR SELECTED FIELDS BELONGING
TO THE RICE ENTERPRISE, SYSTEM 2, 1980

 

 

Hours per Hectare

 

Field Reference

 

 

 

Labor

Per1°d* Average 27/2 32/2 37/4 52/16 36/6**
1 (18) 5

2 (2) 188 312 450 252
3 (2) 174 150 145 450 252
4 (2) 135 ° 225

5 (2) 173 67 857

5 (2) 210 75 88 686 350 590
7 -(2) 130 50

8 (2) 193 150 207 214 225

9 (2) 281 87 198 371 225 402
10 (2) 215 62 257 357 574
11 (2) 82 8 249 343

12 (2) 13

13 (2)
14 (2) 8

15 (2) 130 37 200 159
15 (2) 193 33 239 143 200 271
17 (4) 80 418
Total (hrs/ha) 2,211 917 1,501 2,971 2,325 2,940
Area (ha) .270 .24 1.43 .07 .04 .24

 

 

*The numbers in parentheses represent the length of the period
expressed in weeks.

**Field labor profile used in Version A of the model

\

155

TABLE 6.5

FAMILY LABOR PROFILE FOR SELECTED FIELDS BELONGING TO
THE RICE ENTERPRISE, SYSTEM 3, 1980

 

 

Labor

Hours per HeCtare

 

Field Reference

 

 

 

Peri°d* Average 57/2** 61/3 77/3
1 (18)

2 (2) 12

3 (2) 6

4 (2) 67 723 1,191

5 (2) 26 687
6 (2) 14

7 (2) 35 96 58 190
8 (2) 12

9 (2) 23 18 332
10 (2) 28 298 97

11 (2) . 48 1,021 282

12 (2) 20 436

13 (2) 19 500 194

14 (2) 48 160

15 (2) 119 926 288

16 (2) 35 130

17 (4) 5 190
Total (hrs/ha) 517 3,723 2,694 1,398
Area (ha) .488 .468 .33 .518

 

 

*The numbers in parentheses represent the length of the period
expressed in weeks.

**Field labor profile used in Version A of the model.

FAMILY LABOR PROFILE FOR SELECTED FIELDS BELONGING

156

TABLE 6.6

TO THE RICE ENTERPRISE, SYSTEM 4, 1980

 

 

Hours per Hectare

 

 

 

 

F23$Od* Average Field Reference
87/4 102/3 ll3/4** 116/5

1 (18) 20

2 (2) 115 30 40
3 (2) 81 45 80

4 (2) 172 120

5 (2) 121 160 165 118
6 (2) 273 47 20
7 (2) 256 702 533 329 773
8 (2) 354 15 1,973 910
9 (2) 174 387 82 268
10 (2) 187 185 390 367
11 (2) 175 353 18 340
12 (2) 154 15 13 94 550
13 (2) 115 62 67 35 432
14 (2) 51 , 120 10 59 79
15 (2) 123 430 13 3,120 282
16 (2) 182 3 741 131
17 (4) 423 185 1,767 850
Total (hrs/ha) 2,976 1,923 5,707 4,724 5,160
Area (ha) .151 .065 .30 .17 .15

 

 

*The numbers in parentheses represent the length of the period
expressed in weeks.

**Fie1d labor profile used in Version A of the model.

157

different types of rice cultivation studied. The same information on
the other enterprises can be found in Appendix B. The major points that
can be made from examining Tables 6.3-6.6 are twofold: (i) the average
total amount of family labor use per hectare is comparable to the total
family labor use on some specific fields of similar size in some cases
(e.g., fields 18/7, 36/6, 113/4 in Systems 1, 2 and 4, respectively),
but quite variable in other cases (e.g., fields 6/7, 27/2, 57/2, 116/5

in Systems 1, 2, 3 and 4, respectively); (ii) in all cases, the labor
profile from specific fields was quite different from the average labor
profile. So, it was not possible to conclude whether or not total family
labor use and profile demonstrated the closest resemblance between the
model and reality. However, the hypothesized weakness of the average
model in terms of its tendency to underestimate peak labor demand will
be tested in section 3 of this chapter by using a field-specific labor

profile in lieu of an average labor profile in the model.

2. PRESENTATION OF THE RESULTS FROM THE AVERAGE MODEL

The models displayed in Table 5.10 to 5.13 in Chapter Five yielded
the results shown in Tables 6.7 to 6.10. For Systems 1 and 4, respec-
tively, the optimal value of crop production was 20 and 45 percent
higher than the value of crops produced by the average farm in 1980.
For Systems 2 and 3, respectively, the optimal value of crops produced
was 16 and 13 percent lower than the value of crop package produced by
the average farm in 1980. This difference in results can be attributed
to two main reasons. First, land is fully used in Systems 1 and 4

whereas in Systems 2 and 3, only 81 and 84 percent of land are used.

158

. .co.u:_om use gauze v.39: Auscveacgm
...... a .o c mm..a.uucu .<.u c.~pmm 36.89 p... uxzxm .o vacuum: so: xc use a. .o—aEaxu so. .ucaca
.838. mg. ..g y: con: :o..:—cm 8:. gauge v.56: gaps: mom.ga.uaco as. 28:. muuxuoga c. a.mae:= as»;

..=\<.u a.m.w.m— an o==o>ms a... u—aocpoaac asepxae ammuEOov

.....mmmumc ac: epac: ....ucm span: mpg. can. muugacmo. .o :o—uuuc—puo. mum—2566 as. g..z .Amuacucaogm

a... om.ea.uu:o cu...uuam 6:» mo oguuuu; oco cog. use: seem a. .852.» on. a:.u.om .=o..:.om us.

82:. mm..a.oucu =o>—a as» mo mgouuog ago m=.u.o. so umou as. 38;. has» .cc..:pom pue.uao newscau on»
cases. vague —pusm . so. apco u.—o> we. .uaauao a. ugoecuum as» .8 ago: as. :u.;: .mu.=m.. umogp.

.<.u 8...... . as... e6..ue=. 8......86 ....e...: "8.62

 

 

 

o...... .... a... .... m.=z ....e.. ..
a.m.... .... a... .... m.=e.e=a.u ..
o...m.. ... ...... mm... ... ......gmamv .\=\m ..

o...m..ea a.m.... ..ma. ... meaeaaem ..

o...... . .... a... ..z. ”.52 ..aee.. .m
8...... .... a... ..z. m.=eue=a.e ..
o...oo.o. 8...... ...o. o.. ....x ..

a.m.... ... ...... mm... a... ......mmeazv 8.2.. ..

a.m....e. 9.. 8.... o... a... ..

A<uuv on..a.mu:m wcamm1sugaa. scamMTmmzo. Au: go. as; up. mum..a.uucu

e. mhmumwmomuaumau ...... ”...: ea..=.em we. 2.: .u>au
. .238 6:... .65 32.. .2: o

m=.m..z mmogw mo moacox

 

p subm>m «cm zo—~:.om 4<z—Hm° uzh
z— mum—zmzwhzu "duce: um—(m uzh tocu.mha:mm¢ m.o m—nop

159

.m;\<uu mNNeN an m:cm>m. a... «Famcwauao Eaewxme mmmmgomu xpwcmmmmum: no: u.=o3 .....cm
upaoz mpg» was. mmugaomm. .o copumucppom. mumpaeou mg. gap: .Amuacucaogm a.m. wmwgagmucm um_m_umam
.w.o mpnmh mam .co..=.om Pmswuao .cmgcau ms.
8:38.. mace. —mem a so. xpco nwpm> we. .uanuao a. ugmocmum us. we a.ma mg. cope: .mmgsaww awash.

on. .o a.m.uo; «no can. egos saga o. .mEme on. mcwugom

.<.o .....o. u as... ea..ae=. 5......86 .8..e.xaz "6.62

 

 

 

 

m.... o... oo.c .... m.=eee=aew .m

m.o-oe a.m.oop moms. moan. a.m. u\z\m .m

a.m...mmm m.am¢o~ msmnm op. m.xo ..

mummm m——~ oo.o :oaaou .o

~.mee_m came. mmunm moon. mcmmnzom .m

m.~e~ owe. oo.o Aux. muzcccaocw .e

oepmm .mmm— oo.o o...: .m

..mmow. ....m. an... em. ..=. o\z\m ..

...m.o_. m.~mmm~ ...Ne ope~._ 86.x ..

A<muwmawwmwmmuuucm venom Lena: ccaom Logo. Amufimwwn Amm>mwv mumpgaempcm
a. m:_u.o. .o .mou A<muv nope: :owuspom mmogw .mswuno

poe_uao ;o_;2 .o>o
m=2mgoz mmogw .o mmmcmm

 

 

N Zuhm>m mom zomhzaom 4<zuhao map

zH mummmaamhzu "Juno: u~m<m uzh tom. mh4=mwm m.m «Foch

160

.<mu ~.mmom__ u m=_~> covuucs$ m>_»umnno umNPspxmz "muoz

 

 

 

m._ooo_ a.momo memo momm. Aumv muzz «Luggam
\muscucaogw .n
n.9momm m.¢¢c_P «ewe. _moe._ Aumv Q\z\m .o
o.¢m~m~om ¢.a_m~ emNmu come. mcmmaaom .m

a.momo ~.m~__ Nomm -e_.F Amxv muzz mgmnsmm
\mazcuczogu .q
a.mmmumom «.Nooe ammum comm. mN_mz .m
¢¢m¢_ oc.o Amzv U\z\m .N
c.o--om o.pm_m ONNNN coco.P mow; .P
A<uov mmmwmagoucu canon swag: venom ngoa Am;\<uuv Am; :wv mmm_gqgmu=m

Fus_u o :oz mcpmgmz _w>m4
cw mcwugou mo “mag A<uuv mayo: =o_u=_om mwoga _mswpgo

Pmevuao =o_;3 gm>o
mcpmgmz mmogu fie mmmcam

 

 

m zmhm>m mom zouha40m 4<z~hqo mzh

zH mummmammhzw "Juno: umm<m mzh tom; mhgzmmm m.o mpnmh

161

.<uu-..¢~m¢¢ u u=_m> govuu==$ «>Paummao om~.e_xmz "mpoz

 

 

 

 

co. moo“. co. Ammv musz mgmnsmm .m

¢O¢NP Aev .mwm_ _mmmp ammo. Anny mpacuczoga .m

Aev mmom_ “Koo. mmom_ «Na. Aumv Q\z\m .N

omoamoa ¢.mo~mp omomm cc. mgxo .m

m.moeom Pawn. moe~_ eoou. Ally muzz mgmnsmm .m

oco. .mmmP co. Amxv mpzcucsogu .e

a.mmmsmom Pumm. mmmso we. m~Pm= .m

«mus. flew mmom. mmmmp ammo. Am=v u\z\m .N

mo.¢~c¢ mo.mmmm . ao_¢~ on. mu?“ ._

A<uwwswwmwwmmuucu ucaom gmaa: ucaom gmzod Amummwwp AMM>MWV mmmvgagmucu
a? mcwugou $0 “mom A<muv mu_o= =o_u=_om mwogw Pms_uao

pmepuao ;u_;3 L26
m:_mgmz mmogu we mmmcoa

 

 

e zmhm>m mom zomhaaom 4<z~hao mrh

z“ mmmEgam—hzm Juno: 3mg ...:t. tom“. m._.._:mmm 2.6 ~32.

162

respectively. Land left idle in System 2 is mostly lowland fields.
whereas in System 3, land left idle was partly housefield and partly
bushfield. One implication of this excess supply of land is that in
System 2, rice intensification strategy rather than lowland expansion
strategy should be pursued and in System 3, for dryland crops. inten-
sification rather than land expansion strategy will be a much more
appropriate strategy for agricultural growth in these systems. Second,
l980 rainfall was lower than average and poorly distributed. Systems

2 and 3 were the most affected areas, and hence realized relatively
poor yields and lower gross margins per hectare.

It is significant to note that rice is grown by the net revenue
maximizing representative farmer in all the production systems under
study. Besides identifying the mix of enterprises which maximizes total
gross margins (TGM), LP also generates additional useful economic infOr-
mation about the Optimal solution, such as the ranges of the gross
margins (GM) of the individual enterprises for which the optimal solu-
tion holds, ceteris paribus. This information is useful in assessing
how stable the optimal solution is in the face of possible changes in
costs and/or prices. The wider the range for an individual enterprise,
the more likely the enterprise is to remain in solution.

Thus, in System l, under ceteris paribus assumption, the gross
margin for rice (24,915 CFA/ha) could vary between 0.0 CFA/ha and

9,024,915.0 CFA/ha without its optimal level moving from l.l hectare.1

 

1This is partly because rice is the only crop which uses lowland in
the model, hence it faces no competition from other crops. This speci-
fication reflects conditions during the survey period (the rainy season).
Other crops are grown on lowland during the dry season in System 4 only.

163

Similarly, the GM for S/M/C (HF) (26,455 CFA/ha) could vary between
7,612.0 CFA/ha and 40,678 CFA/ha without its optimal level moving from
1.2 ha. In System 2, the GM for rice (42,7ll CFA/ha) could vary from
25,592.5 CFA/ha to 110,157.7 CFA/ha without its optimal level moving
from l.24 ha. In System 3, the GM for rice (27,220 CFA/ha) could vary
between 3,16l.0 CFA/ha to 9,027,220 CFA/ha without its optimal level
moving from 1.0 ha. Finally in System 4, the GM for rice (74,l09 CFA/ha)
could vary from 3,983 CFA/ha to 9,074,109 CFA/ha without its optimal
level moving from .60 ha. From the above descriptive information, it
can be concluded that in all production systems under investigation,
apart from System 2, rice is entering the optimal plan under a relatively
wide range of gross margins per hectare. So, given the model structure,
it will take a big change in costs or prices for the level of rice acti-
vity in the optimal plan to change from their current levels, assuming
other things remaining unchanged. In section 3 below, we will relax
the lowland field constraint and look at its impact on the level of rice
cultivation in the optimal plan for various production systems studied. .
The last column of Tables 6.7 to 6.10 gives us the cost of forcing
in non-optimal enterprises. This measure indicates the extent of improve-
ment needed in the GM of each excluded enterprise before it could compete
for a place in the optimal solution. This measure may serve as a basis
for suggestions for adaptive trials in the different systems studied.
For example, in System 1, other things remaining unchanged, groundnuts
would need a GM of l8,843 CFA/ha compared to its current level of 7,6l2
CFA/ha before it could enter the optimal solution. Bambera nuts would
need a GM of 24,784 CFA/ha compared to its current level of l,67l CFA/ha

before it could enter the optimal solution. In System 2, maize would

164

need a GM of 33,146 CFA/ha before it could enter the optimal solution.
And cotton would need a GM of 33,273 CFA/ha compared to its current level
of 2,119CFA/ha. These target gross margins could be achieved either
through yield increases, or reduction in costs of production, or price
increases, or a combination of all these three alternatives.

Another category of useful economic infOrmation about the optimal
solution that the LP generates is the marginal value product of each
resource used in the production process. This information is often use-
ful in indicating where effort should be directed to relax operationall
constraints. Such information can be valuable in evaluating the feasi-
bility and profitability of relaxing particular constraints. The
results in Table 6.11 show the gain in total gross margins to be obtained
from a marginal unit addition to the level of each individual constraint,
ceteris paribus.

In System 1 for example, an additional hectare of housefield or
bushfield would add 26,455 CFA to TGM while an additional hectare of
lowland field would add 24,915 CFA. Labor periods are non-binding con-
straints in this system. In System 2, an additional hectare of lowland
field will add nothing to TGM whereas an additional hectare of house-
field or bushfield will add respectively l0,527 CFA and 7,824 CFA to
TGM. Two labor periods are binding constraints in this system; they are
labor periods 8 and 9, corresponding to the month of AugUst. In period
8 which corresponds mainly to the weeding of grains, an additional hour
or labor will add 139 CFA to TGM whereas in period 9 which corresponds
to weeding of grains and groundnuts, an additional hour of labor will

add 57 CFA to TGM.

165

TABLE 6.11

SHADOW PRICES OF RESOURCES AND CONSTRAINTS USED
IN THE MODEL, BY SYSTEM, 1980

 

 

Resources or Shadow Prices (CFA)

 

 

Constraints System 1 System 2 System 3 System 4
Housefields (HF) 26455 7824 0 13133
Lowlandfields (LF) 24915 0 24059 70126
Bushfields (BF) 26455 7824 0 13133
LABOl 0 0 0 0
LABOZ 0 0 0 0
LA803 0 0 0 0
LABO4 0 0 0 0
LABOS 0 0 0 0
LABO6 0 0 22 0
LABO7 0 0 82 0
LABOB 0 139 0 11
LABO9 0 57 0 0
LABlO 0 0 0 0
LABll 0 0 0 0
LABlZ 0 0 0 0
LA813 0 0 0 0
LABl4 0 0 0 0
LABlS 0 0 0 0
LABlG 0 0 0 0
LABl7 0 0 0 0
MINFG 0 0 0 -2123
MAXMAI 14223 0 27722 53934
MAXSOY 53066 0 75435 N.A.

MAXOKRA N.A. 11286 N.A. 15782

 

 

166

In System 3, an additional hectare of housefield or bushfield will
add nothing to TGM while an additional hectare of lowland field will add
24,059 CFA to TGM. Labor periods 6 and 7, corresponding to the month of
July,-are binding constraints in this system. In period 6 which cor-
responds mainly to the weeding of grains and groundnuts, an additional
hour of labor will add 22 CFA to TGM and in period 7 which corresponds
to the weeding of grains only, an extra hour of labor will add 82 CFA
to TGM.

In System 4, all the three categories of land are binding constraints
and only labor period 8 corresponding to the first part of August is a
binding constraint. In period 8 which corresponds to weeding of grain
and fertilization of rice an extra hour of labor will add 11 CFA to TGM.

In all the production systems, labor is a binding constraint only
in July-August. This is a rather surprising result given that through
the discussions with farmers, labor was also a problem during the months
of October-November (periods 13 and 14) when the bulk of harvesting
takes place. One possible explanation for this unrealistic result could
be that by using the maximum farm labor available each period on an
average farm, we are making total labor from the model very different
from the total labor use in reality. In an attempt to address this
issue, in section 3 below, instead of using the total maximum available

farm labor per period (i.e., fj = Nj x 35 x Lj where all the symbols
are the same as defined in Chapter Five, section 1.2.3.) we will use
the average labor use per period in the labor supply vector (i.e.,

f5 _3 NJ x Hj x Lj where fj, Nj and Lj are the same as before and

R3 is the average number of hours per worker per week).

167

The last four rows of Table 6.11 give us an idea on how the TGM
will be affected if the production constraints are modified. In Systems
1, 2 and 3, increasing the MINFG by one hectare will have no impact on
TGM whereas in System 4, TGM will be reduced by 2,123 CFA. In Systems
1, 3 and 4, increasing the MAXMAI conStraint by one hectare will
increase TGM by 14,223 CFA, 27,722 CFA and 52,934 CFA, respectively;
increasing the MAXMAI constraint in System 2 by one hectare will have
no impact on TGM. In Systems 1 and 3, increasing the MAXSOY constraint
will add to TGM by 53,066 CFA and 75,435 CFA, respectively; increasing
the MAXSOY constraint in System 2 by one hectare will have no impact
on TGM. Soybean is not grown in System 4. In Systems 2 and 4, where
okra is grown, increasing the MAXOKRA constraint will add 11,286 CFA
and 15,782 CFA to TGM, respectively. However, the feasibility and pro-
fitability of relaxing production constraints should not be based only
on gross margins considerations. The current levels of consumption of
the enterprises concerned, the storage and other marketing problems
faced by these enterprises should also be taken into accoUnt. I

The major results of this section show that rice is grown by the
revenue-maximizing average farmer in all the production systems studied.
Furthermore, rice is entering the solution in each system under a rela-2
tively wide range of gross margins per hectare, given the model struc-
ture. One implication of this result is that it will take a big change
in costs or prices for the level of rice activity in the optimal plan
to change from their current levels, assuming other things remaining
unchanged. So, increased rice production will necessitate improvement
in the production methods for other major competing crops, which would

free up labor which would allow more intensive rice production. It

168

should be remembered here that rice was the only major crop grown in
the lowland fields during thewrainy season, and thus did not face any
competition by other crops for this category of land.

In Systems 1 and 4, all the three categories of land were binding
constraints. In System 2, however, lowland field was not a binding
constraint. But in System 3, lowland field was the only binding land
constraint. It should be noted that some land is left idle in Systems
2 and 3 as a result of labor conflict between crops during the months of
July-August; relatively high labor demand on sorghum/millet fields
during this period being the major contributing factor.

3. SENSITIVITY OF THE BASIC MODEL TO CHANGES IN
SELECTED PARAMETER VALUES

Experiments with the basic model were designed to assess the impact
of possible changes in the economic and technical environment of the
farmers surveyed. The hypotheses elaborated in sections 1 and 2 of this
chapter were tested by running five separate versions of the basic model.
Version A is the case where the average labor coefficients for each
production system were replaced with field specific labor coefficients.
This version allowed us to test whether the opportunity cost of growing
rice was lower in the average model, hence leading the farmer to produce
more rice because peak labor demand was underestimated. Version B cor-
responds to the case where the availability of lowland was increased by
50 percent. This version allowed us to test whether the use of scarce
development funds to support water control (hence more lowland avail-
able for cultivation) in the has-fonds would lead to increased hectarage

under rice in the different production systems. Version C represents.

169

the case where there is a 10 percent increase in gross margins per hectare
for the.S/M/C enterprise. This version also includes the case where the
minimum food grain constraint (MINFG) is lowered by half, implying that
farmers are more willing to rely on the market for their grain supply.

1 which is a variant of Version C corresponds to the complete

Version C
elimination of the MINFG constraint from the model. These two versions
(C and C1) allowed us to test whether increased productivity in S/M/C
coupled with an easing of the MINFG constraint will lead to more bas-
fonds rice production by the representative farmer. Version E corres-
ponds to the case where the initial labor supply vector made up of maxi-
mum family labor available per period is replaced by a labor supply 1
Vector made up of average family labor use per period. This model allows
us to test whether the fact that labor is only constraining in July-
August was due to the fact that total labor availability in the model

was unrealistically high by comparison to actual farm conditions. Note

that in Versions B, C, C1

(and E, average labor coefficients are used.

The production strategies which maximize total gross margins (TGM)
under assumptions A through E above are given in Tables 6.12 to 6.15,
alongside the actual production strategy derived from the basic model-
(Version D).

3.1. Comparison of Optimal Producti n Strategies
.Under Different Models (A, B, C, C , D, and E)

The main result here is that the basic model is very sensitive to
the choice of labor coefficients and the labor supply vector as defined
under Versions A and E,'and much less sensitive to the other parameter

values as defined under Versions B, C and C]. In System 1 under Version

17W)

TABLE 6.12

SWRY OF THE PRODUCTION STRATEGIESa HHICH MAXIMIZE
TOTAL GROSS MARGINS (TGM) UNDER DIFFERENT VERSIONS
OF THE BASIC NOEL. SYSTEM 1, 1980

 

 

 

 

 

 

Hodelb
Enterprises A B C _ C1 O E
Rice 1.100 1.65 1.100 1.65 1.100 1.10
SIM/C (HF) .51 1.20 1.200 1.20 1.200 .51
SIN/C (BF) .61 .61 .61 .61 .6100
Maize .40 .40 .400 .40 .40 .40
Groundnuts (HF) .69 .1725
Groundnuts (BF)
Balbera Nuts (HF) .0609
Bambera Nuts (BF)
Soybeans ‘ .2400 .2400 .2400 .24 . .24 .2307
Cotton
Okra
GN/BN (HF)
eu/au (as)
Maximized objective
function value (CFA) 97645 124350 115434 129137 110646 76933

 

aSolutions in the table are expressed in hectares.

bin Version A, field specific labor coefficients are used in the model.
In Version 8, availability of lowland was increased by 50 percent.
In Version C, gross margin per ha for SIM/C enterprise was increased by 10 percent and the MINFG
7 constraint was lowered by half.
In Version C . gross margin per ha for SIM/C enterprise was increased by 10 percent and the MINFG
constraint was dropped from the model.
Version 0 is the basic mbdel.
In Version E. the family labor supply vector was made up of average family labor use per period.

171

TABLE 6.13

SUMMARY OF THE PRODUCTION STRATEGIESa HHICH MAXIMIZE
TOTAL GROSS MARGINS (TGM) UNDER DIFFERENT VERSIONS
OF THE BASIC MODEL, SYSTEM 2. 1980

 

 

Model

Enterprises A a c c1

 

 

Rice .5291 1.2419 1.2419 1.2419 1.2419 .792
S/M/C (HF) .800 .800 .80 .80 .80 .900
SIM/C (BF) .3445 .3395 .3395 .3395 .3395 .3392
Maize

Groundnuts (HF)

Groundnuts (BF)

Bwera Nuts (HF)

Baebera Huts (BF)

Soybeans .3555 . .3605 .3605 .3605 .3605 .3008
Cotton '

Okra - .100 .100 .10 .10 .10

GN/BM (HF)

GN/BN (BF)

 

Maximized objective
function value (CFA) 59689 90233 92213 92213 90233 67760

...-n ———————————————-—g—___ .

 

 

 

-. —-

4'Solutions in the table are expessed in hectares.

 

bIn Version A, field specific labor coefficients are used in the model.

In Version B, availability of lowland was increased by 50 percent.

In Version C, gross margin' per ha for SIM/C enterprise was increased by 10 percent and the MINFG
constraint was lowered by half.

In Version C . gross margin per ha for S/H/C enterprise was increased by 10 percent and the MINFG
constraint was dropped from the model.

Version 0 is the basic model.

In Version E, the funny labor supply vector was made up of average family labor use per period.

1272

TABLE 6.14

SUMMARY OF THE PRODUCTION STRATEGIESa HHICH MAXIMIZE
TOTAL GROSS MARGINS (TGM) UNDER DIFFERENT VERSIONS
OF THE BASIC MODEL. SYSTEM 3. 1980

 

 

 

 

 

Model
Enterprises A B C C1 O E
Rice .428 1.500 1.00 1.5 1.0 .9743
S/M/C (HF) 1.000 .1478
S/H/C (BF) 1.603 1.6091 1.6029 1.6091 1.8849
Maize .400 .52 .520 .52 .52 .52
Groundnuts (HF)
Groundnuts (BF)
Bambera Nuts (HF)
Bambera Nuts (BF)
Soybeans . .3835 .490 .49 .49 .49 .49
Cotton '

- Okra .

GN/BN (HF) .9123 1.1477 .9123 1.1477
GN/BN (BF) .5271 .5209 .5271 .5209 .2451
Maximized objective
function value (CFA) 69159 128689 118999 131019 116659 112865

 

1'Solutions in the table are expressed in hectares.

bIn Version A. field specific labor coefficients are used in the model.

In Version 8, availability of lowland was increased by 50 percent.

In Version C, gross margin per ha for S/H/C enterprise was increased by 10 percent and the MINFG

1 constraint was lowered by half.
In Versionlz, gross margin per be for S/H/C enterprise was increased by 10 percent and the MINFG
constraint was dropped from the model.
Version 0 is the basic model.
In Version E, the family labor supply vector was made up of average family labor use per period.

1173

TABLE 6.15

Similar OF THE Fnooucnow STRATEGIES“ HHICH NAXIMIZE
TOTAL GROSS MARGINS (TGM) UNDER DIFFERENT VERSIONS
OF THE BASIC nooEL. SYSTEM 4. 1980

 

 

 

 

ModeTF
Enterprises A a c c1 o E
Rice .366 .6948 .60 .7311 .60 .3071
S/M/C (HF) .050 .08 .04 .0596 .3180
S/M/C (BF) .024 .0207
Maize .48 .48 .48 .48 .48 .2539
Groundnuts (HF) .0579 .1786
Groundnuts (BF) .0396
Bowen Nuts (BF) .1785 .2221 .1942 .2604 .0495
Balbera Nuts (BF) .060 .0393
Soybeans
Cotton
.Okra .01 .40 .40 .40 .40 .40
GN/BN (HF)
GN/BN (BF)

 

Maximized objective ' -
function value (CFA) 60664 99884 94969 101751 94829 60249

W
'Solutions in the table are expressed in hectares.

hIn Version A, field 'specific labor coefficients are used in the model.
In Version 8, availability of lowland was increased by 50 percent.
In Version C, gross margin per ha for S/H/C enterprise was increased by 10 percent and the MINFG
constraint was lowered by half.
In Version C . gross margin per ha for S/H/C enterprise was increased by 10 percent and the MINFG
constraint was dropped from the model.
Version 0 is the basic model. -
In Version E. the family labor supply vector was made up of average family labor use per period.

174

A, less S/M/C is grown and groundnuts which is less labor demanding

(204 hrs/ha as compared to 1,627 hrs/ha under S/M/C) enters the solution
at the level of .69 ha. In System 2, labor demand is so constraining,
mainly in periods 9 and 10 which correspond to the month of August that
only .529 ha of rice is grown comparatively to 1.24 ha grown under the
basic model (Version 0). In System 3, not only is a smaller hectarage
of rice and S/M/C grown, but also the mixture GN/BN does not enter the
solution anymore under Version A. Finally, in System 4, less rice and
okra are grown under Version A, and two enterprises (groundnuts and
bambera nuts) are no longer part of the optimal solution.

One over-riding observation is that across all the production
systems (except System 1) less rice is grown under Version A, confirming
the hypothesis that under the basic model, the opportunity cost of
growing rice is low and the use of true specific field data in the .
matrix results in lower hectarage of rice. It is also interesting to
note that the results for Version A, which represents fairly closely

current conditions, Show that rice growing is compatible with a revenue

maXimizing strategy.
Under Versions B, C and CT, the optimal solution to the basic model
showed very little sensitivity to the parameter values under considera-
tion. In Systems 1 and 3 under Version B, all additional lowland avail-
able was put under rice cultivation whereas in System 4, only part of
the lowland made available is grown in rice. In System 2, additional
lowland made available was left idle.

In Systems 1, 2 and 3 under Version C, the optimal solution to the
basic model remained unchanged. But in System 4, there was a slight

modification in the optimal plan: fewer hectares of S/M/C and bambera

175

nuts were produced and groundnuts did not enter the new plan. This was
probably due to increased intensity in rice cultivation under this
system. Under Version C], the same results as in Version C occurred,
except that in System 4, no S/M/C enterprise was produced and more land
was put under rice cultivation (.731 ha instead of .60 ha). So the main

1, the parameter values

conclusion here is that under Versions B, C and C
would have to change substantially in order to affect the optimality of
the solutions in Tables 6.7 to 6.10, given the model structure.

Under Version E, the optimal solution to the basic model was very
sensitive to the parameter values under consideration except in System
3.. In System 1, labor is so constraining during the months of September-
October-November that less sorghum/millet is grown and groundnuts and
bambera nuts, which are less labor demanding, enters the solution at
.17 ha and .06 ha, respectively. In System 2, a smaller hectarage of
rice and soybean is grown because of labor bottlenecks in June and
September. Finally in System 4, less rice and maize are grown and the
groundnut enterprise, which is less labor demanding than rice and maize,
enters the solution at the level of .1786 ha. One over-riding observa-
tion under Version E is that labor has become constraining in period 6
(land preparation) in System 2, periods 7-9 (weeding) in Systems 1 and
2, and in periods 11-16 (weeding-harvesting) in Systems 3 and 4. This
confirms the hypothesis that under the basic model (Version 0), too much
labor was made available with the attending result that labor turned out
to be a constraining factor only in periods 7-9. However, it is inter-
esting to note that even under Version E, both rice and S/M/C enterprises

are compatible with a revenue maximizing strategy.

176

TABLE 6.16

SHADOH PRICES OF RESOURCES AND CONSTRAINTS
USED IN THE VARIOUS MODELS.
SYSTEM 1. 1980

 

 

Resources or Shadow PFICES (CFA)

 

 

Constraints ' a A B C CT’ 0 E
Housefields (HF) 7612 26455 29100 29100 26455 O
Lowlandfields (LF) 24915 24915 24915 24915 24915 22121
Bushfield: (BF) 7612 26455 29100 29100 26455 O
LABO1 0 O 0 0 0 O
LABOZ O O O O O O
LA803 o' o o o o o
LABO4 O O O 0 O O
LABOS O O O 0 O 0
LABOG O 0 O O O O
LABO7 O O 0 0 O 1
LABOB o o o o o ' o
LABO9 40 O O 0 0 O
LABlO 0 O O O 0 O
LABll 0 O O 0 O 0
LABl2 O 0 O O O 543
LA813 O O 0 0 O 0
LABI4 O O O O O 149
LABl 5 O 0 O O 0 0
LA816 O O 0 O O O
LABl7 0 O 0 0 O 0
MINFG O 0 O N.A. 0 -9158
MAXMAI 33066 14223 11573 11573 14223 23730
MAXSOY 71909 53066 50421 . 50421 53066 O
MAXOKRA N.A. N.A. N.A. N.A. N.A. N.A.

 

 

aIn Version A, field specific labor coefficients are used in the model.
In Version 8, availability of lowland was increased by 50 percent.
In Version c. gross margin per ha for S/H/C enterprise was increased by 10 percent and the MINFG
constraint was lowered by half.
In Version C ,gross margin per ha for S/M/C enterprise was increased by 10 percent and the MINFG
constraint was dropped from the model.
Version 0 is the basic model.
In Version E, the family labor supply vector was made up of average family labor use per period.

1177

TABLE 6.17-

SHADOW PRICES OF RESOURCES AND CONSTRAINTS
USED IN THE VARIOUS MODELS,
SYSTEM 2. 1980

 

 

Resources or Shadow Prices (CFA)

 

 

Constraints a A B C c1 0 E
Housefields (HF) 13558 7824 9951 9951 7824 13520
Lowlandfields (LF) 0 0 0 0 0 0
Bushfields (BF) 13558 7824 9951 9951 7324 13520
LABOl o 0 o 0 o o
LABOZ 0 0 o o o o
LABO3 0 0 o 0 0 232
LABO4 0 o 0 o 0 0
LABOS o o 0 o 0 o
LA505 o o o o o 0
LABO7 0 0 0 o o 0
LAaoa 0 139 128 125 139 253
LABO9 54 57 54 54 57 114
LABlO . 30 o 0 0 o o
LABll o 0 o o 0 0
LABlZ o 0 0 0 0 0
LABl3 0 o o 0 o o
LABl4 o 0 0 0 o 304
LABlS o o 0 o o o
LA816 o o o 0 o o
LAB17 o o o o 0 0
MINFG 0 o o N.A. o 0
MAXMAI o o o o o o
MAXSOY .0 0 o 0 0 o
HAXOKRA 13303 11255 9955 9955 11286 0

aIn Version A, field Specific labor coefficients are used in the model.
In Version 8, availability of lowland was increased by 50 percent.

In Version C, gross margin per ha for S/H/C enterprise was increased by 10 percent and the MINFG

constraint was lowered by half.

In Version C ,gross margin per ha for SIM/C enterprise was increased by 10 percent and the MINFG

constraint was dropped from the model.
Version 0 is the basic model.

In Version E, the family labor supply vector was made up of average family labor use per period.

1773

TABLE 6.18

SHADOH PRICES OF RESOURCES AND CONSTRAINTS
USED IN THE VARIOUS MODELS.
SYSTEM 3, 1983

 

 

Resources or Shadow PI‘ICOS (CFA)

 

 

Constraints a A B C c1 0 E
Housefields (HF) o 0 o 0 o o
Lowlandfields (LF) 0 24059 24041 24041 24059 0
Bushfields (BF) 0 o 0 0 0 0
LABOl 0 0 0 0 o 0
LABOZ o o o 0 0 0
LA803 o 0 0 0 0 o
LABO4 o o 0 o o o
LABO5 0 o o 0 o 0
LA806 545 22 31 31 22 434
LABO7 .0 82 79 79 82 154
LA808 ‘ o 0 0 0 0 0
LABO9 . 143 0 0 0 0 o
LABlO o o 0 0 0 o
LABll 27 o o o o 111
LABlZ o o 0 0 o 0
LABl3 o 0 0 0 o 0
LABl4 o 0 0 0 o o
LABlS o 'o o o o 178
LA816 o 0 o o 0 25
LABl7 o o o 0 o o
MINFG -452359 0 o . N.A- 0 0
MAXMAI 0 27722 27533 27533 27722 31388
MAXSOY 0 75435 75521 75521 75435 75533
MAXOKRA N.A. N.A. N.A. N.A. N.A. N.A.

 

 

 

aIn Version A, field specific labor coefficients are used in the model.
In Version 8, availability of lowland was increased by 50 percent.
In Version C, gross margin per ha for S/M/C enterprise was increased by 10 percent and the MINFG
1 constraint was lowered by half.
,gross margin per ha for S/M/C enterprise was increased by 10 percent and the MINFG
constraint was dropped from the model.
Version 0 is the basic model.
In Version E, the family labor supply vector was made up of average family labor use per period.

In Version C

1179

TABLE 6.19

SHADOH PRICES OF RESOURCES AND CONSTRAINTS
USED IN THE VARIOUS MODELS.
SYSTEM 4. 1983

 

 

Resources or Shadow PPI C85 (CFA)

 

 

Constraints a A B C c1 0 E
Housefields (HF) o 0 13133 13133 13133 895
Lowlandfields (LF) 0 0 70125 70126 70125 0
Bushfields (BF) 22243 0 13133 13133 13133 954
LABOl o o 0 o o o
LABOZ o o o 0 o o
LA803 o o 0 o o o
LABO4 o o o 0 o o
LABOS 0 o o o o o
LA806 o o o o o o
LABO7 95 205 o 0 o 159
LABOB o 3 ' 11 11 - 11 0
LA809 o o o 0 o o
LABlO o o 0 0 o o
LABll 0 0 o 0 0 o
LABlZ o o o 0 o 430
LABl3 o o o o o 58
LABl4 o o . o I 0 go 22
LABlS o o o o o 22
LA816 o o o 0 0 0
LA817 o o o 0 o 75
MINFG .22490 -23340 -758 M- -2123 o
MAXMAI 51844 30658 53934 30558 53934 0
MAXSOY N.A. N.A. N.A. N.A. N.A. N.A.
MAXOKRA 0 24495 15782 24495 15782 25468

 

 

a'In Version A, field specific labor coefficients are used in the model.

In Version 8, availability of lowland was increased by 50 percent.

In Version C, gross margin per ha for S/M/C enterprise was increased by 10 percent and the MINFG
constraint was lowered by half.

In Version C , gross margin per ha for S/M/C enterprise was increased by 10 percent and the MINFG
constraint was dropped from the model.

Version 0 is the basic model.

In Version E. the family labor supply vector was made up of average family labor use per period.

180

3.2. The Opportunity Cost of Scarce Resources
and Constraints t0 Increased TGM

The opportunity cost of a scarce resource represents the change in
.TGM that one extra unit of the factor could produce when employed in the
most profitable fashion, given that resources are already optimally allo-
cated and that available supplies of the factor in question have already
been used. The results for Versions A through C1 for the four produc-
tion systems were given in Tables 6.16 through 6.19, alongside the
opportunity costs of resources and constraints used in the basic model.

One interesting result that emerges from this analysis is that the
family labor profile and the labor supply levels are critical in deter-
mining the maximum level of TGM attainable by the farmer from his crop-1
ping activities. This result is not surprising since family labor is
the major input in farming in the EORD. The use of chemical fertilizer
or insecticides or herbicides is extremely uncommon. Even animal trac-
tion was uncommon in 1980 in the survey area. Hence the intensity and
timing of total family labor use are important factors in determining
attainable gross farm incomes. The labor coefficients in periods 9 for
System 1, 9 and 10 for System 2, 6 and 9 for System 3, and 7 for System
4 are critical in determining the maximum level of attainable TGM under
Version A. Under Version E, in addition to the labor periods that are
constraining under Version A, labor is also a binding constraint for all
systems during at least one period between periods 11-15. Consequently,
greater availability of labor in these periods would permit increases
in the maximum level of attainable TGM if families were willing to devote
more time to farm work. Presumably the fact that it is not the case at

the present time is because of incentive problems, the desire for

181

increased leisure, and the preference for pursuing other activities.
Under Versions 8, C and C], the opportunity costs of resources and con-
straints used are only slightly different from the results under the

basic model in Table 6.11.

4. SUMMARY

Several interesting results emerge from the modeling exercise.
First, except for System 4, the cultivation of low-yielding sorghum/
millet is compatible with a revenue-maximizing strategy. Even when we
do not impose a minimum food grain production constraint, the S/M/C
enterprise is still competitive enough to appear in the optimal plan,
given the current costs and returns structure and the production con-
straints imposed to reflect either the market limits or the soil con-
ditions. Second, investment in water control to expand the rice land
available to farmers will only induce farmers to grow more rice in.
Systems 1, 3 and 4. In System 2 where investment comes entirely from
local initiative without any government assistance, an additional
hectare of lowland made available will add nothing to TGM. 50 land
expansion as a strategy to expand rice production in this system will
have a very limited potential. In Systems 1 and 3, a land expansion
strategy would have some potential to the extent that the increase in
lowland availability is compatible with labor bottlenecks in periods
6, 11, 14 and 15. In System 4, increasing rice production through land
expansion should be undertaken with caution since lowland appears to
be less of a constraint here. When the lowland available was increased
to .9 ha in the model, only .7 ha is used in the optimal plan (Table

6.15). Third, except in System 2, there is evidence that increased

182

productivity in the S/M/C enterprise coupled with the elimination of
the minimum food grain constraint will lead to increased hectarage under
rice; the logic here being that with less acreage under S/M/C, labor

is freed up and can now be used in rice production, which is relatively
more attractive in terms of gross margin per hectare. Fourth, current
levels of rice growing under the different systems are pretty stable
within a wide range of variations of gross margins per hectare given
the model structure, implying that in the model, price policy alone may
not be enough to bring about expanded rice production. But it is quite
conceivable that the improvement in the technological package may result
in higher yields and justify further investment in water control, for
it should be kept in mind that improved water control may have an
impact either on the land under cultivation or on the intensity of
farming. Fifth, for groundnuts and bambera nuts to be competitive in
the various production systems, tremendous efforts in terms of yield
increases and marketing outlets are still needed. Gross margins per
hectare for these two enterprises will have to be increased at least by
ten-fold. Sixth, from the modeling perspective, the labor profile
using field specific labor coefficients and the labor supply vector
using average labor use‘per period yielded results which were fairly
close to reality, i.e., labor being a binding factor during land pre-

paration, weeding and harvesting.

CHAPTER SEVEN

SUMMARY, POLICY RECOMMENDATIONS,
AND FURTHER RESEARCH

The purpose of this chapter is to present the summary of findings
and the policy implications of the study, to draw conclusions and to

make suggestions for further studies.

1. SUMMARY

The current vagaries of weather in Upper Volta, increasing food
imports in general and rice imports in particular, and the need for agri-
culture to play a greater role in the socio-economic development of the
country have intensified interest in increasing food production in the
country. The country is divided into eleven 0RDs (Regional Development
Office), one for each of the ten departmental subdivisions of the
country, with one department subdivided into two ORDS. The Eastern
Region of Upper Volta has a large surplus agricultural production po-
tential and it has a natural terrain which is suited to rice production
in bas-fonds (saucer swamps). But relatively little information iS
available on the costs and returns of producing rice in these bas-fonds.

This study focuses on the Eastern 0RD (EORD) and is designed to
generate farm level data to estimate the relative profitability of dif-

ferent crops for the four major rice production systems in the Eastern

183

184

Region of Upper Volta in 1980. This study further investigates whether
and how rice cultivation can be increased in the EORD.

The research approach employed for this study was described in
Chapter Two. Three major sources of information on farmers' circum-
stances were used:

(1) background information on the farmer's environment from
secondary sources (research institutes, government reports, meetings,
etc.);

(2) field survey involving structured interviews with farmers by
enumerators; and

(3) personal observations in farmer's fields and environment to
obtain information on the agricultural production processes. It was‘
anticipated that the most important input would be labor, so detailed
data on family, hired, and communal labor utilization was collected,
using an "activity" or modified cost route survey method. The cost
route method of interviewing farmers once or twice a week over the whole
length of the agricultural calendar was not used in this study because
of the doubt about its cost effectiveness and the difficulty of main-
taining interest of farmers during repeated interviews. Instead an
activity approach was used to collect input-output data on each of the
major activities--clearing-burning-ploughing-weeding-mulching-spraying-
fertilization-cutting-threshing-winnowing-bagging-transport--involved
in the production processes during the period running from June 1980
through February 1981. The activity approach consisted of interviewing
farmers once for each of the major field activities (altogether 12 to 14
activities depending on the type of production system), on a plot by .

plot basis soon after its completion.

185

The sample size of 116 farmers was determined by the number of the
farmers producing bas-fond rice in the EORD and by the number of farmers
that an enumerator could effectively handle. Typically each enumerator
covered 30 farmers. Later inputs were recorded only at the end of each
major field activity, but enumerators visited farmers at weekly inter-
vals to check on their progress since detailed information on the crop-
ping calendar of sample farmers was not available to the researcher at
the beginning of the survey.

The activity approach which was used in this study as a method to
generate information on small farmers' conditions looks promising in
terms of (l) maintaining the farmers' interest in rural surveys, and
(2) avoiding collection of too much information. However, this approach
may not reduce survey costs unless there is detailed information avail-
able on the cropping calendar Since farmers have to be interviewed to
see if an activity is completed. But data processing costs are reduced,
in the sense that there are fewer survey forms to be coded and key .
punched with the attending result that data files created are much
smaller than otherwise, and the researcher can benefit from insights
gained through informal and personal interviewing. While the data col-
lected using the activity approach can be used for budgeting, some
difficulties are encountered when programming models are the intended
analytical tool; more specifically, conversion of labor coefficients
derived by using data collected on an activity by activity basis into
calendar week coefficients proved to be a painful process. Furthermore,
it will be very difficult to define the family labor supply on an
activity by activity basis.

186

Chapter Three describes the agricultural production systems in the
Eastern Upper Volta. An overview of agricultural production in the
Eastern Upper Volta shows that the agricultural sector is based primarily
on subsistence agriculture: over 90 percent of farmers meet their food

requirement needs by growing staple food crops under shifting cultivation.

A wide variety of crops are grown, both for domestic consumption (e.g.,
sorghum, millet, maize, peanuts, etc.) and for export (e.g., cotton).
Farmers generally grow more than one crop but on different plots.
Fifty-seven percent of the cultivated land was sole cropped, and all
but 1 percent of the rice fields was sole cropped.

There are numerous factors that constrain the expansion of agricul-
ture in the EORD; these factors include: capricious weather, poor soils,
water management deficiencies and/or costs, lack of improved seeds,
poor extension services, poor marketing infrastructure, etc. Rainfall
has a skewed distribution toward the end of the growing season. The
main constraint is not the lack of rainfall, but poor management of
available water. Agricultural research is nonexistent in the EORD and
the extension system is crippled by the low level of training of exten-
sion agents coupled with the high turn-over of the 0R0 personnel.

Four production systems were defined based on the degree of water
control. System 1 represents the traditional bas-fonds or unimproved
swamps. No attempt is made in this system to control water in the bas-
fonds. System 2 is based on semi-traditional bas-fonds; in this system
water control improvements are rudimentary (based on the use of dikes)
and are all done by farmers without any government intervention. In
System 3, the improvement consists of the building of dikes to retain
water longer on the plots than otherwise. The topographic mapping and

dike building are done with heavy government assistance. System 4

187

represents the system based on irrigated has-fonds; here the degree of
water control is higher than in any of the previous systems. The dam
structure here is built and managed by the government, and irrigation is

done by gravity.

0n the average, farms were made up of eight fields, with an average
area of .45 hectares per field. Rice represented on the average only
11 percent of the total cultivated area, varying from 3 percent in System
3 to 31 percent in System 1. Three main categories of labor were iden-
tified in the research Sites: family labor, hired labor and social
labor. Finally, it was found that on the average, 67 percent of house-
hold heads were involved in at least one non-farm activity.

In Chapter Four, the relative profitability of different crops was
estimated for each of the four major rice based production systems.
Financial enterprise budgets were constructed from survey data, and for
each enterprise budget, the net cash income (or gross margins) was de-
rived. Other performance measures that were derived using the budgets
included: net margin per hectare, net financial returns to land, family
labor and management, and costs of production.

It was found in System 1 that gross margin ranged from 1,671 CFA/ha
to 79,521 CFA/ha depending on the crop. The soybean enterprise had the
highest gross margin and the groundnut and bambera nut enterprises had
the lowest gross margin. For all the six enterprises (rice, S7M/C,
maize, groundnuts, bambera nuts and soybeans) in System 1, the returns
per field hour of family labor varied from 97.6 CFA for rice to 13.9 CFA
for bambera nuts. This result suggests that it may be profitable in
some instances for family members to seek employment in urban areas where
the minimum guaranteed wage is 90 CFA/hr. However, for the rice enter-

prise, the returns of 97.6 CFA/hr of family labor suggest that there is

188

no financial advantage of family members seeking wage employment in urban
areas when they are needed on their rice fields. It was found that the
weighted average net return per hour of family labor across all enters
prises in System 1 was 47.4 CFA, which is only about half of the minimum
urban guaranteed wage. However, it should be kept in mind that whenever
the return per hour of family labor is less than the minimum urban guar-
anteed wage, this will not necessarily lead to rural exodus because of
relocating the job search costs that may be so high that the farm worker
decides to remain in agriculture. Among the six enterprises comprising
System 1, maize showed the lowest total cost of production, 22.8 CFA/kg,
and bambera nuts had the highest cost of production, 182.8 CFA/kg. Rice
showed the second lowest total cost of production, 33.6 CFA/kg of paddy.
Using 47 CFA/hr as the opportunity cost of family labor, three enterprises
in System 1 realized a negative return to land and management; they are
groundnuts, bambera nuts and S/M/C.

In System 2, the gross margins ranged from 2,119 CFA/ha to 42,711
CFA/ha, depending on the crop. The rice enterprise had the highest gross
margin and groundnut and cotton enterprises had the lowest gross margin.
However, all enterprises in this system were able to cover their vari-
able costs as in System 1. Returns per field hour of family labor in
the system ranged from 1.4 CFA for cotton to 45.5 CFA for soybeans. For
the rice enterprise, the return per field hour of family labor was only
18.7 CFA. If the present costs and returns structure continues, the
f0110wing shifts could be expected: a) farmers will likely put more of
their land and labor into soybeans, okra, S/M/C and rice in that order;
and b) low returns per field hour of family labor for the cotton enter-
prise may force farmers to abandon this enterprise despite the heavy

government support of this export crop. However, the return per hour to

189

family labor algge_cannot determine the change in enterprise mix. Other
factors that can affect the farm re-structuration include: (1) the
fixity of some inputs to some enterprises, e.g., lowland fields during
the rainy season can only be used to grow rice; (2) the family size and
the division of labor within the family; and (3) the labor requirements
of different crops as well as their market potentials, e.g., currently
the market potentials of soybeans and okra are very limited and their
home consumption level is very low. Four out of seven enterprises (rice,
maize, groundnuts and cotton) realized a negative return to land and
management. Among the seven enterprises, S/M/C had the lowest total
cost of production, 27.8 CFA/kg and cotton had the highest cost of pro-
duction, 133.2 CFA/kg. The second highest total cost of production was
found for the groundnut enterprise, 93.1 CFA/kg, probably due to the low
yield of groundnut (215 kg/ha) in this system. Rice Showed the second
lowest total cost of production, 44.1 CFA/kg of paddy.

It was found that in System 3, gross margin ranged from 6,502
CFA/ha to 78,254 CFA/ha, depending on the crop. The soybean enterprise
had the highest GM and the groundnut/bambera nut mixture had the lowest
GM. For all the five enterprises (rice, S/M/C, maize, groundnuts/
bambera nuts, soybeans) comprising this system, except for soybean and
maize, the return per field hour of family labor was less than the mini-
mum wage rate paid to unskilled urban labor, i.e., 9O CFA/hr. For the
maize and soybean enterprises, returns per field hour of family labor
were 116.0 CFA and 450.2 CFA, respectively. For the rice enterprise,
returns per field hour of family labor were 51.4 CFA. As a result, if
the present costs and returns structure persist, farmers will likely put
more of their land and labor into soybeans and maize if the minimum

sorghum needed for home consumption is attained. Low returns per hour

190

of family labor under rice may force farmers to abandon this crop since
as a grain, it is not yet an important part of the diet. Some incentive
structure must be urgently found if rice growing is to survive in this
system where some important investments in water control have already
been made. Except for soybeans and maize, all enterprises in this
system realized a negative return to land and management. Among the
five enterprises, maize showed the lowest total cost (If production
(19.1 CFA/kg), and GN/BN showed the highest total cost of production
(228.3 CFA/kg). Rice showed the third lowest total cost of production
(69.4 CFA/kg).

In System 4, the variation in GM ranged-from 13,655 CFA/ha to
74,109 CFA/ha, depending on the crop. The rice enterprise had the
highest gross margin and the S/M/C enterprise had the lowest GM; but
all the GMs were positive. In all the Six enterprises (rice, S/M/C,
maize, groundnuts, bambera nuts, and okra) comprising System 4, except
for okra, the returns per field hour of family labor was less than the
minimum wage rate paid to unskilled urban workers, i.e., 9O CFA/hr.

For okra, returns per field hour were 198.2 CFA, while for rice it was
only 24.1 CFA. This low return per hour for rice may be enough to dis-
courage farmers from further rice cultivation. All enterprises except

' maize and okra realized a negative return to land and management. Among
the Six enterprises comprising System 4, maize showed the lowest total
cost of production, 11.4 CFA/kg. The second highest cost of production
was found in bambera nuts, 94.0 CFA/kg. Okra showed the second lowest
total cost of production (16.2 CFA/kg),followed by rice (56.8 CFA/kg).

Among the four different rice enterprises, the labor utilization

ranged from 302 hours per hectare in System 1 to 3,054 hours per hectare

191

in System 4. So, the highest the level of water control, the more labor
intensive is the rice cultivation. The contribution of family labor to
total labor on rice plots varied from 76 percent in System 3 to 97 per-
cent in Systems 2 and 4. In System 1, family labor contribution to total
labor was 81 percent.

The financial analysis of the different rice production techniques
showed that System 4 yielded the highest gross margin per hectare
(74,109 CFA) but the second lowest returns per field hour of family
labor (24.1 CFA) due to the high labor requirement of this system (irri-
gated bas-fonds). The traditional bas-fond (System 1) yielded the lowest
gross margin per hectare (24,915 CFA) but the highest returns per field
hour of family labor (97.6 CFA). The most expensive way to grow rice
was found in System 3 (69.4 CFA/kg of paddy), probably due to water‘
costs coupled with low rice yields and the least expensive way of pro- '
ducing rice was found in System 1 (traditional bas—fonds), i.e. 33.6 '
CFA/kg of paddy.

An economic analysis of the different rice enterprises from the
society's perspective showed that the least cost technique for producing
rice remains the traditional cultivation or unimproved swamps (System 1).
The introduction of water control to date does not compete effectively
with this basic system, and total costs per kilogram of paddy rise in
every case. However, the possibility of increasing rice production
within traditional cultivation is limited. The most efficient means of
producing rice under "secure" water control appears to be by partial
water control (System 2). Production on irrigated has-fonds with com-
plete water control and fertilizer use is the second most expensive with

cost per kilogram of paddy some 173 percent above the least expensive,

192

traditional bas-fonds, and 96 percent above the more attractive improved
alternative (System 2). When considering whether rice production in the
EORD should be increased by promoting a particular technique of cultiva-
tion, only production in traditional bas-fonds will be economically

juStifiable under current technologies and yield levels.

In Chapters Five and Six, a linear programming model for the four
rice—based production systems was presented and results from the basic
model discussed. In addition, several experiments with this basic
average-coefficient model were designed to assess the impact of possible
changes in the economic and technical environment of the sample farmers.
These changes involved varying land availability, labor coefficients,
and supply and some production constraints. The goal of this exercise
was to provide direction to policy makers rather than attempting to
recommend improved plans for individual farmers.

The major results of the models Show that rice is grown by the
revenue-maximizing average farmer in all the production systems studied.
Furthermore, rice is entering the solution in each system under a rela-
tively wide range of gross margins per hectare given the model structure,
except in System 2. It can therefore be concluded that it will take a
big change in costs or prices for the level of rice activity in the
optimal plan to change from their current levels, assuming other things
remaining unchanged. As a result, increased rice production will neces-
sitate improvement in the production methods for other major competing
crops, which would free up labor which would allow more intensive rice
production. It should be remembered here that rice was the only major
crop grown in the lowland fields during the rainy season, and thus did

not face any competition by other crops for this category of land.

193

In Systems 1 and 4 all the three categories of land were binding
constraints. In System 2, however, lowland field was not a binding con-
straint. But in System 3, lowland field was the only binding land con-
straint. It should be noted that some land is left idle in Systems 2
and 3 as a result of labor conflict between crops during the months of
July-August; relatively high labor demand on sorghum/millet fields during
this period being the major contributing factor.

In all the production systems studied, labor is a binding constraint
only in July-August. This is a rather surprising result given that
farmers identified labor as a problem during the months of October-
November (periods 13 and 14), when the bulk of harvesting takes place.
One possible explanation for this unrealistic result could be that by
using the maximum farm labor available each period on an average farm,
we are making total labor from the model very different from the total
labor use in reality.

Several interesting results emerge from the modeling exercise with
labor supply based on average labor use. First, the cultivation of low-
yielding sorghum/millet is compatible with a revenue-maximizing strategy
in Systems 1, 2 and 3. Even when we do not impose a minimum food grain
production constraint, the S/M/C enterprise is still competitive enough
to appear in the optimal plan, given the current costs and returns
structure and the production constraints imposed to reflect either the
market's limits or the soil conditions. Second, the investment in water
control to expand the rice land available to farmers will only induce
farmers to grow more rice in Systems 1, 3 and 4. In System 2, where
investment comes entirely from local initiative without any government

assistance, an additional hectare of lowland made available will add

194

nothing to TGM. So land expansion as a strategy to expand rice productflxi
'in this system will have a very limited potential. In Systems 1 and 3,

a land expansion strategy would have some potential to the extent that
the increase in lowland availability is compatible with labor bottlenecks
in periods 6, ll, 14 and 15. In System 4, increaSing rice production
through land expansion should be undertaken with caution Since lowland
appears to be less of a constraint here. When the lowland available was
increased to .9 ha in the model, only .7 ha was used in the optimal plan.
Third, except in System 2, there is an evidence that increased producti-
vity in the S/M/C enterprise coupled with the elimination of the minimum
food grain constraint will lead to increased hectarage under rice; the
logic here being that with less acreage under SIM/C, labor is freed up

and can now be used in rice production which is relatively more attractive

in terms of gross margin per hectare. Fourth, the family labor profile
and the labor supply levels are critical in determining the maximum
level of total gross margin (TGM) attainable by the farmer from his
cropping activities. The labor coefficients in periods 9 for System 1,
9 and 10 for System 2, 6 and 9 for System 3 and 7 for System 4 are
critical in these respective systems in determining the maximum level of
attainable TGM when field specific labor coefficients are used. This
result is not surprising since the family labor is the major input in
farming in the EORD. The use of chemical fertilizer or insecticides or
herbicides, is extremely uncommon. Even animal traction was uncommon
in 1980 in the survey area. Hence the intensity and timing of total
family labor use are important factors in determining attainable gross
farm incomes. Consequently, greater availability of labor in these

periods would permit increases in the maximum level of attainable TGM

195

if families were willing to devote more time to farm work. Presumably

the fact that it is not the case at the present time is because of

incentive problems, the desire for increased leisure and the preference
for pursuing other activities. Fifth, current levels of rice growing
under the different systems are stable within a wide range of variations
of gross margins per hectare, given the model structure, implying that

in the model price policy alone may not be enough to bring about expanded
rice production. But it is quite conceivable that the improvement in

the technological package may result in higher yields and justify further
investment in water control, f0; it should be kept in mind that improved
water control may have an impact either on the land under cultivation

or on the intensity of farming. SiXth, for groundnuts and bambera nuts
to be competitive in the various production systems, tremendous efforts
in terms of yield increases and marketing outlets are still needed.

Gross margin per hectare for these two enterprises will have to be in-
creased at least by ten-fold. Seventh, from the modeling perspective,
the labor profile using field specific labor coefficients and the labor
supply vector using average labor use per period yielded results which
were fairly close to reality, i.e., labor being a binding factor during

land preparation, weeding and harvesting.

2. POLICY IMPLICATIONS AND STRATEGY TO IMPROVE THE
PERFORMANCE OF RICE PRODUCTION IN THE EORD
Several policy implications can be derived from the findings of
this study as they relate to (l) the improvement of the four major rice
based production systems in the EORD; and (2) the identification of an

appropriate bas-fond rice production strategy which could assist with

196

the increase of rice production in the EORD. The improvement of economic

profitability of the different types of rice cultivation can come either

through the improvement of the efficiency of resources already committed
to rice production or the introduction of improved inputs or through
appropriate pricing policies, or a combination of all the above alter-
natives. This study provides policy makers in Upper Volta someof the
data required to identify the trade-offs among investment levels in
water control for alternative rice production systems. Four basic
systems of water control are employed in rice farming in the EORD.
System 1 represents the unimproved swamps with yield of .458 ton of
paddy per hectare on the average. Systems 2 and 3 rely on uncertain
surface flooding, and water is controlled in these two systems by using
dikes. Yields in these two systems averaged .5 and 1.2 ton of paddy per
hectare in Systems 3 and 2, respectively. The main difference between_
System 2 and System 3 is that the investments in water control in

System 2 come entirely from local initiative whereas in System 3 water
control is achieved with government assistante in terms of topographic
mapping and building of dikes. System 4 represents the system with
complete water control, based on dam irrigation and with yield of 1.7
tons of paddy per hectare on the average. The survey data have permituxi
a detailed analysis of the farm—level relative position of rice in the
four production systems studied. However, we do not have marketing and
other macro—economic data to rigorously trace the direct and indirect
implications of these alternative production systems for the Eastern
Region of Upper Volta and the national economy. Nevertheless, this

study will pose major policy issues facing GOUV policy makers and then

197

conclude with some recommendations as to how to improve the perfbrmance

of rice production in the EORD.

2.1. Major Policy Issues and Reorientation

Key policy issues which GOUV policy makers should consider include:

Capital investment. Capital investment in the agricultural produc-
tion systems was found to be low. With regard to durable capital, the
present study shows clearly that the predominant capital item among
survey farmers is hand-tools, mainly cutlasses and hoes. However, it
should be kept in mind that historically, the development of mechanical
technology suitable for small farmers has proved to be more difficult
than advances along biological lines (improved seeds and fertilizers).
It is important that adaptative research trials be directed toward
divisible and low cost technology given the low resource levels of
farmers in the EORD. Policy measures should also aim at developing
much required skills in small scale rural industries both as an addi-
tional source of income and as a support to the farm sector technology.
It is suggested here that the operations of ARCOMA-COREMMA should be
reviewed in order to help this institution play a vital role in the
rural equipment which will help lessen the human energy requirements
associated with hand-tools.

Research priorities. Even though the results of the modeling exer-
cise suggest that current levels of rice growing in the different produc-
tion systems are stable within a wide range of variations of gross margin
per hectare, implying that price policy alone may not be enough to bring
about expanded rice production, it is quite conceivable that improvement

in the technological package of rice growing may result in higher yields,

198

thus increasing rice production without any expansion in the area under
rice cultivation. With the S/M/C enterprise still the major crop mix-
ture in the cropping systems of EORD farmers, this enterprise which
absorbs most of the family labor, will have to be looked at concurrently.
Unfortunately, these two enterprises fall under the jurisdiction of two
different research institutes: IRAT for the S/H/C enterprise and CERCI
for the rice enterprise. So, it is recommended that trials on farmers'
fields be jointly planned and carried out by CERCI and IRAT. This recom-
mendation ensures that technological packages will be formulated and
refined under actual farmers' management practices, while also meeting
consumer's preferences.

Land expansion/Levels of investments in water control. High mar-
ginal value productivities found for the lowland fields in Systems 1, 3
and 4 seems to indicate insufficient lowland fields in these systems and
thus suggest a land expansion strategy. But there are a number of unfa—
vorable medium— and long-term consequences of this strategy. Increasing
rice production can be achieved either by improving yields (increased
intensity of farming) or by expanding the area under production, or both.
However, the higher level of investment in water control necessary for
expanding area or increasing yields can only be profitable if increased
costs are more than offset by increased revenues. The economic analysis
showed that the costs of improved water control in the EORD are very
high and cannot be covered by the returns under existing technology.
With current average yields of .5 ton of paddy per hectare in System 3,
the expansion of rice land available to farmers should be discouraged.
In System 4, lowland availability to farmers appears to be less of a

constraint, and a rice farming intensity strategy rather than a land

199

expansion strategy should be promoted. However, in order to draw
definitive conclusions regarding the returns to land improvement under
partial or complete water control, there is a need for studies on the
utilization of the irrigated areas during the dry season.

Paddygpricing. The present official farm gate price of 63 CFA per
kilogram of paddy is rarely followed by market participants. Actual
observed farm gate prices per kilo of paddy were 61, 42, 66 and 52 CFA
in Systems 1, 2, 3 and 4, respectively. The economic costs of production
per kilogram of paddy were 34, 44, 69 and 57 CFA in Systems 1, 2, 3 and
4, respectively. So, except in System 1, the economic cost of production
was greater than the farm gate price received by sample farmers. The
marketing of domestic rice is further complicated by its poor quality
because of its higher percentage of brokens, with the attending result
that domestic rice is less desired by consumers. If the policy objective
is agreed upon as being to satisfy consumer's demand through domestic
production and imports while providing some incentives to rice farmers,
possible policy instruments available to the GOUV policy makers to
achieve this objective include: (1) floor pricing in order to allow
farmers to cover their cost of production; and (2) a tax on imports of
rice in the short run and subSidization of rice research in-country with
the tariff revenues. Here it is suggested that measures should be taken
to make sure that the farm gate price of paddy in the EORD is at least
equal to the import parity price of paddy, i.e., 56.4 CFA per kilogram.
The "Caisse deiPeréquation" of the Ministry of Commerce could play a
'vital role in the implementation of such policies. However, further

investigation will be needed to determine the exact magnitude of these

200

policy instruments. The problem of the poor quality of domestically
produced rice should also be addressed at the same time that the pricing
of paddy is being discussed.

Extension services. The problem of high turn-over of the EORD

 

personnel was identified. The personnel policy of the EORD as well as
that of the other institutions involved in the development of the Eastern
Region (e.g., CERCI, FDR, ONBI) will have to be revised to ensure some
continuity in the implementation of various rice programs. In addition,
given the relatively high profitability of other crops such as soybeans
and okra and the importance of the S/M/C enterprise in the cropping
system of the EORD farmers, more attention Should be given to these
crops through a personnel redeployment.
2.2. Identification of an Appropriate Bottomland Rice
Production Strategy in the EORD
Two major components of an appropriate bas-fond rice production

strategy are identified below for the evaluation of GOUV policy makers.

2.2.1. ”Small Scale Irrigation Using Dikes

The EORD should de-emphasize major investments in dam irrigation and
give priority to partial water control and rainfed agriculture. Reasons
for this recommendation include:

(1) Dam irrigation appears to be too costly to be profitable under
current technologies and yield levels.

(2) Managerial and maintenance problems continue to persist, making
demands on scarce administrative talent and reducing the profitability
of these operations. In fact rice production with dam irrigation created

the highest net economic losses among the four types of rice cultivation.

201

(3) Even though from the strict economic sense only rice production
in traditional baS-fonds is justifiable, partial water control has poten-
tial for bas-fond rice production. Economic losses per hectare in
System 2 are only about 5,000 CFA. Furthermore, farmers in this system
can play a more active role in both the investment in and operation of
the system, thereby reducing the need for government control, supervision

and expenses, while creating more employment.

2.2.2. Improved Production Practices

Our analysis showed that not only yields achieved in various systems
under study are low, but also most farmers are not following recommended
production practices (seed rates, fertilizer use and/or rates, pesticide
use, use of animal traction, etc.). The main reason for this situation
is that most recommendations from CERCI and IRAT have been so far based
on experiment station trials, sometimes discounted to approximate
responses under farm—level conditions. This approach is a poor substi-
tute for FSR approach, with the attending result that yield-increasing
technology, especially for food crops, is still lacking. Therefore,
there is an urgent need to get Farming Systems Research (FSR) underway
in the EORD in order to test and to assess new technologies with farmers
and to evaluate with them what the tangible benefits of improved produc-
tion techniques can be. This recommendation will call for a close colla-
boration between CERCI, IRAT and ICRISAT which are the main research

institutions of importance to the EORD.

3. SUGGESTIONS FOR FURTHER STUDIES

This study has generated a data base on the economics of bas-fond

rice production and the relative profitability of the different levels

202

of water control in the EORD. However, there is a need for studies on
the utilization of the "irrigated" areas during the dry season which
will be helpful for an accurate assessment of the returns to land
improvement under partial (Systems 2 and 3) or complete water control
(System 4). In addition, the author recognizes the need to integrate
this farm level study with other important components of the sub—sector
such as marketing, trade policy, tax on consumers in some areas, etc.
These complementary studies will provide the missing link between farm-
level trade-offs and macro-level policies. Finally, there is now a
need to initiate on-farm trials as we mentioned above. This will call
for a much more close collaborative effort from IRAT, CERCI and ICRISAT.
Further investigations are also needed into alternative data col-
lection methods. The assessment of the activity approach used in this
study was presented in Chapter Two and in the Summary of Findings. In
order to address some of the shortcomings of this approach, it is still
necessary to identify various data gathering Shortcuts which could be
'useful in any future evaluation of existing cropping systems, particu-

larly if resources of the study are more limited.

APPENDIX A

SURVEY FORMS USED FOR THE STUDY

203
APPENDIX A

SURVEY FORMS USED FOR THE STUDY

BFl - Plot inventory

BF2 - Household inventory

BF3 - Family labor use

BF4 - Non-family labor use

BF5 - Purchase of non-labor inputs

BF6 - Use of non-labor inputs

BF7 - Harvest records

BFB - Sales records

BF9 - Inventory of tools and equipment

BF10 - Family labor spent on non-family fields
BFll - Price information records

BF12 - Off-farm employment records

BF13 - Plot area measurement

BFl4 - Yield plot records

BF15 - Livestock inventory

BF16 - Labor use in livestock activities

BF17 - Estimated costs and returns in livestock activities
BFlB - Qualitative data information

*Measurement units used to collect information on output

*Assessment of the quality of data gathered by the enumerators

204

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

zmmhmmhzm.n mh<a mzmhmzozm

 

“.28 ”use: 8 2oz ”a
omega: .23 owe—Em ~38 9533 Son 328:; not 35:: Epsom—.88.”. ”g
mmmcmz «:3 ago I
3 3 \j
pmuowmp .>:m .hmm.4 mo .o.x.o.4 mo mozouim<m mun unamzozoouiomw< mashm

 

205

 

 

 

 

 

 

 

 

 

2: , 8: A3 :3 E A8 A3 3 A8 2:

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

cm can mance» mgumm ounces
cumwmua 55 3258 .2an 2385.3 53.2.. 32:83: :28 3:332 .6 more 2.8.9:.
. . mauve ezoa mop—>Faom mup>puu< ice :8 an exam uo>m um
ewuoecfi _m oceemmmuoc moeu=< . opemvocwea amo>pz mwcc< anemone meoz
$32.8 :omwem: 853mm} 3 =85
zmHhmmhzm.o mh<o zzmhwzczu
mzo~ mw<zmz an :02 "zomh<uummhzmnm

 

monocws moo mP—Pmu a. mu pawppep acoeomcmoom "mom mzume

ommjwz ocoN meow;
\ xii» \1 x \ \ \ \

pmiowmp .>:m .hmm.4 ma .n.m.o.4 mo mnz¢mim<m mun maomzozoumiomw< mozhm

 

 

 

6
0
2

 

 

 

 

 

 

occomcma occomema accomema
can can can
mm_muop moccomeoa mmFauou moccomcma mo—ouou moccomeon
macaw: mu menz macaw: on «small mmeawz mu wcaz
on mez<ezu (Awe messes fame mezzo:

 

 

 

mHH>Hku< mhhmu <

mmHumkm<m Pz<>< mum

 

mzmz mmo mHmoomh<u

 

Amv
oeeseeoe
.muumo

mu cow»
izumxm.u
mama

 

Amy
m—Foogmq
e— muse»
Lam muw>wuom
muumu we mcowu
isowxm.u menz

 

mczupzu

 

A_v

moee>_eog

 

393:2. on 2:3 5 mustang v.33 2 9523 2 can $055.53

 

oppoogaa oeumu cam Amy mmucmpa Amy oczppsu

 

zmemmhzm.o uh<o

 

 

oppmucen up we oz pm soz

mamhmsozm

 

MZON

mw<zmz 2n zoz

 

”onH<uHmthuoH

panama at moeaeos map can vcezom —_e>meu mu weave ”mum mzuHe

amaze:

ocoN

«guru

 

\ \1 \ \ \ x \ \. \
—miommF .>:m .hmu.4 ma .o.x.o.4 mo mozomim<m mun mzouzozouuiomw< mozhm

 

207

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

moccom «econ moccom was... 323 2:8
-gma icon -cma -goa .goa icon
83.888 8: 5% >233 8 .3283 8 5838 8 8:55
mums: 1:833 h muwamu mocao: men: macaw: wen: macaw: ocnz r mczop muumo oppmu
OLE—m2 Amy mflcwu—hm Amv man—5.”— A¢V ago: “WMWMMH cowwao IMWWO Maw—H Imp—van Om:—
mme<hOb mmmzumuo uhH>Hhu< mhhmu < . noxo.u um Eoz
maeueeaee ez<>< memes“: mug “excome<u oooe
zmmhmmhzm.o mh<a mauhmaczm
szN mu<zmz an 2oz "onh<uHmthuo~

mowmemco mopeuou momconwu am ouc:m5.=u nee: mmgnsms map can eccaom —_e>ucu we masoh "cum uzuHe

mance: ocoN «gown
\JJJJJJ

_wiowm— .>:m .hmm.4 mo .o.m.o.4 we mozcmim<m mun unomzozoomiomm< mosh“

 

208

 

Aev Ame . A_v

 

 

 

 

 

 

 

 

 

 

 

 

 

wave: can pace unoeocowmw>oeqae.v mueaom o a»: a so aesom.u
oe_oee=c e o 2 open
zmmhmmhzm.o mh<a mauhmaczu
mzo~ mw<zmz so :92
"onh<oHeH»zwo_

.we>=oo.u =_ee up mac moguae covuoauoeq mu meaouume mop Lao; mo_euou momcmawa "mom uzoae

ounce: ocoN «new;
\\\\\\\Vu
Fw-omm— .>:z .hmm.4 mo .a.m.o.4 mo mozcm1m<m mun maomzozoumiomu< wank“

 

mmcu=< .moupupummo .mFmemcm .mmocosmm .

 

mopamczuico: mgzouoee

 

209

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

acme uses no?» omPpr: mppmocma
-moapasoc on ioccowme>oeene.u wave: oenz -emwp_u=.c “souuee ”MMWMMhM up
.zo acecaoo xpce ooeaom muwucoza mama so wcauez on oz
zmmhmmhzm.a mh<a msmhmaozm
szN . mw<zmz no 202 "zomh<u~mmhzunn

Ammpnegauicoc meaouuem mop csoa acmeozapcav
mg>=oo.uicems up one oeuse cowpuauoeq ow meaauoee on copuempppuz "mum mxume

mango: o=o~ «gown
\ w V \ N V \ V \
Fmiowmp .>:m .Hmm.4 ma .o.m.o.4 mo mozoeim<m mm: maofizozoomiomc< moahu

 

21D

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

”mmmmw omexooum mou=o> copaossomcoo mmwvca acneoz Aev «seam upstage oppooeea one:
mmo<m2>zm zoee<meeee= mmeeoomm wheezema. on easy a, one:
zuflhuzhzm.o mp<o . ‘ mamhmsczm
mzoN mm<zuz an 2oz

 

"onh<uHmH»zmo~
mongoooc mot oeum_mwm "mum mzumu

>11 .emm.m we .o.¢.o.e we omeewz oeoN oeoee
mezoo-m<m mu: macezozoum-oeo< ma=em \ \iiw \ \1 \ \iiwiiu

 

 

A

.Fmiowmp :omwem mp mu mupauoca mmp mac emepmwmmcmcm.ze

 

211

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

mucm> cummumugu mpuwe, were: F33 mwats 8.252 2:9. «.5ng 3.693. memo
on osoo ease seem \xeea xeea mmaazm> mmeeez<=o .oeeeeeo on ease
me<o mememeozm
mzoN mmezmz as so: "zoeeemeaeezmee

 

mmu=m> mmu meumwmmm ”men wrong

mmmcmz mcoN mguFu

3J\\o\\\
pmuommp .>:x .hmu.4 ma .o.m.c.4 mo mozou1m<m mun uzomzozoumiomw< moahm

 

 

212

 

A_v

 

 

 

 

 

 

 

 

 

 

 

 

memces an me.mu.== vacwuee :8 page» _mpe was
pmpcmume at mppu: m_> pmaaoe x'ce um:oe.u xwee mm< mensoz mu m )5
zwuhmmhzm.a mh<o mzmhmzozm
NZON ww<zmz 3: 2oz "zomh<ummmhzmoH

(mpoo_eme smegmuee pm ages quamzo an mgpaucm>cm ”mum uzume

mmecmz meoN meow;
\w\\\\\|.|\\
Fmiommp .>:m .hmu.4 mo .a.m.o.4 mo mazom-m<m mun maomzozoomioaw< moahm

 

213

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

mmceom macaw mmccom mceom mmccom mccom
-Lmq -cma -cma -Lmq .gmq icon
oesowee 2: 25 8233 8 .5238 8 >238 8 2:8
mu_== -cmao mumgmm mmezm: menz mmcam: mcaz mmgam: mcaz «camp «Hume mppmo
m : ._¢m>aeu mv wu_> amaze. Jena mp
2352 Q 22 m E 5.8.5.2 E 3.5: 8 me: 553 -38 .23 8 oe
mh~>mpu< mhhmu < -mxm.u um 2oz
82 552.. 285$. :22 $85: ma 5825 28
, zmmhmzpzm.a mh<a mzmhmzozu
uzoN mw<zuz :2 2oz "zonp<ommfihzua~
some acmemwma am wacuze.u musesu mmp memo mmocme av mmgnsme mmp emu vccaom pwm>ecp mu masmh "open mzuam
mmmcmz m:o~ mgupm

 

 

\1 \i.\ \ii\1 \ x, \ \
_miomm— .>:m .hmm.4 mo .o.m.o.4 ma monmim<m mmo maomzozoomiomu< moahu

 

214

 

 

E

mgsmu=m> mu «gasoz

E
og.au.== xvgg

 

 

95mm: on w... E:

ANV APV

wage; muwzuoga

 

 

 

 

 

 

"mamhmsozm zoz

 

mzum<z so uh<a

 

mzo¢<z so to: ”zo~h<o~mmhzmou

 

xzmuop Wmsugg xzm xpgn— u_._i._m mzumm

caucus acoN mguwm
\\\\\\\VV

 

pmuowmp .>:¢ .hmm.4 mo .o.m.o.4 mo mozcmnm<m mun mzouzozoouuomw< uozhm

215

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

wgsuaz

  

   

unsuuz

  

 

mhszm
mz~<w

 

 

mm4<h0h
mmmzuuua

mhz<uzm

 

zo~h<gzuuo whhmu < hz<a_u_h¢<m m¢<zuz an mmmewz mm: w—«oamh<u xm\\\V

   

w . warm“ . maoa:

 

 

 

zmHhmmhzm_o mh<o

mamhmaczm

 

 

 

szN

mw<zmz am so: ”zomh<o~u~hzma~
mmpoorgmoucoc :owuonsuuo.u u=~=m>ogn m==m>ws am mpaou "Npmm mzumm

mmocmz ocON «guru
\\\\\\\\\

 

anommp .>:m .hmm.4 mo .o.m.o.4 mo mnzo;nm<m mun m:¢~:¢zoumucxw< mnahm

216

 

mmgm \ \w \ \ \, \ mvowmgwazm Ex cowmmmucoo m— mu mucmumpo

 

a mgaumssmw mu Lzmggm

 

 

 

 

 

 

 

 

 

 

 

 

 

muou
mu oz #

«you

samamcog guasw~< mu oz

gamsmcog gaze N
m44uua<a <4 mo mHzoomu _ <

 

 

 

 

 

 

 

 

 

 

mmhzmmmmm mmmzhgzu

 

 

mw<¢=mmz mo mh<o «submaozm

 

 

 

mzo~ . mu<zmz no :02 "onh<u~mHquoH

comFMm maumu mmm>Pup=u mmppmusmn mmu mmwuwmgmnam mac mangamm: "mme wzomm

 

macaw: ocoN ago?“

\-1N. N \ \ \ \, N11\
pwuommp .>:¢ .hmm.4 ma .o.m.o.4 mo mazomnm<m mm: maouzozoumaomw< moshm

 

 

217

 

\ _ouou so mompo mo monsoz
\ Pmpou an mpozooo mo moosoz
~E\ mwggmo moo opmuou monogam

 

 

 

oPFoogmo m— Amocmogoos_.o
oggmo Lam mogoupzo ogogo gmov oFPougmo mp

moor; mo mgoz mpozooo mo ogoz we oz moo ucosmumposm moucomwgo mmgaupzu oo o: um Eoz

 

 

 

 

 

 

 

 

 

mmmm<u mum mu<4o zm mmHz mo mp<n . mzmhmzozm

 

 

 

mch mw<zmz so :02 “onh<uHmthuoH
mmuwmowo mo mongoo "epmm mzumm

 

mango: ocoN ozo_m

\JJ\\V\I\\
Pwnommp .>:z .hmm.4 mo .o.m.o.4 mo mozomum<m mm: mzomzozoumuozw< mozhm

 

218

 

 

Amy zeo

 

 

 

 

 

 

 

 

zmv ANV AFV
xzme_=m.o ogoz oo< oxom opgomoumu mooomu
zmmhumhzm.o uh<o zzmhuaozm

 

 

uzON

Fm-omaF .>=m .hmu.4.mo

mu<zmz so zoz

 

"zouh<ommmhzmom

 

op> .mpgmgu an mgzu¢=a>=o ”mpmm “zoom

mango: ozoN ozowm
\ \x \ \¢:N \ x» \ \

 

 

.n.z.o.4 ma Mozomnm<m mun mzomzozoumuozw< mozhw

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

   
 

 

mhz<m2m
th>HHQ< mkhmo < maHuHHm<a Hz<>< mmamZME mmo mamoomh<u

 

 

zmmhmmhzm.o mh<a mzmhu302m
uzON mw<zmz no so: "onh<uHmthmoH

 

ago—w; mo ocwgoo 2 2:6 magma so muggy: on: LS 2.58 526.5 mo mosoh "gm
macaw: 23 ago E
W4 \ x N \ \ \ \
8.82 .22 .hmm... we did... we mozouumg mmo moomzozoumuomg ESE

 

220

 

 

9»: mm:—

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Fmpou xwgo oucomgwwahfluo osgou moosoz Fmfimmm.o moowomm mo wove: ouwpcmao oozh
mm=F<z
mmpzm> moo mhzmm mzo<o , mmo<POP mmmzmmmo
zmopmahzm.o mh<a maupm=ozu
mzo~ mo<zmz =9 :02 "zooh<oHoszmaH

 

omo>m_w.P mo ocwasoo op mono mason mcwmm um mopmuou momcoowo "spun mzofim

mango: mcoN «cop;
\ N V \¢m\ \ x \ \

—muomm~ .>:m .hmm.4 ma .o.x.o.4 mo mozomum<m mun mzomzozoumuomo< moshm

 

 

221

ETUDE AGRD-ECONOMIQUE DES BAS-FDNDS
DE L'0.R.D. DE L'EST, RHV, 1980-81

/ I! o/ / / / / Z /

Fiche Zone Menage

 

FICHE BFIB: Informations généraies sur ies différents systemes de

 

 

 

production.

IDENTIFICATION:
Nom du menage _‘
Zone

 

Date d'entretien

 

A. PROBLEMES FONCIERS ET UTILISATION DES DIFFERENTES PARCELLES.
1. Quei nombre de categories de terres exploitables avez-vous 5
votre disposition?
Terres libres (Tinjaii)
Terres en Jachére (Kuawaogu)
Terres de Sorgho ou de mii (Kuanu)
Petits lopins de terre (Liioii)

 

 

 

 

2. Si vous voulez accroitre vos superficies cultivees, pourriez-
vous trouver du terrain?

oui ‘ non

 

2.1. Si non, pourquoi?

Pas de terres disponibies dans Ia region
Main d'oeuvre familiale Iimitee

Main d'oeuvre saiariée Iimitée

Manque d'argent pour acquerir terrain

Manque d'argent pour les intrants agricoIes
(semences engrais, ... etc)

o | .
Autres raisons (prec1ser)

 

 

 

 

 

 

 

 

 

2.2 Si oui,
Quelies cuitures vous aimeriez voir les superficies augmentees?

 

 

222

- A queIIe distance de la maison compterez-vous trouver de terrain?
Moins d' an km

Moins de 2 kms

Plus de 2 kms

 

 

 

3. Avez-vous i'intention d'augmenter vos superficies en riz?
oui non
3.1. Si oui,
Quand?
Pourquoi vouiez-vous augmenter vos superficies en riz?
besoins alimentaires
besoins en argent
Prestige
Recommende par 0RD
Autre raisons (précicer)

 

 

 

 

 

 

 

 

 

3.2. $1 non,
pourquoi?

 

 

 

4. Dans votre menage, pour queis produits 1a superficie a- -t-e11e
augmentee et poor quels produits a-t-eIIe diminuee durant les
5 dernieres annees

 

 

 

 

Mil-sorgho ‘ I = augmentee

Riz 2 = diminuee

Arachide 3 = pas de changement
Niébé/haricot

Soja

 

 

 

Raisons pour chaque cas:

 

 

 

223

5. Est-ce que la rotation des cultures augmente vos rendements?
oui non

-6. Quel type de rotation avez-vous trouvé trés bénéfique?

 

 

 

7. Pourquoi faites-vous des associations de cultures?

Manque de terrain

Pour diminuer le travail

Pour diminuer le nombre de champs

Pour réduire les problémes avec
les insectes

Pour conserver le sol
Par tradition
Autres raisons(preciser)

 

 

 

 

 

 

 

 

 

B. DISPONIBILITE DE LA MAIN D'CEUVRE ET PROBLEMES Y AFFERANT

l. Pendant quelle (s) période (5) de l'annéc avez-vous des problémes
de main d'oeuvre?

 

 

2. Pendant ces périodes de pointe, quelles cultures sont considérées
comne les plus importantes et recoivent en priorit'e la main d'oeuvre
disponiple? (Preciser cultures et operations dans l'ordre des
priorites.)

 

 

3. Est-ce qu'un champ de riz demande plus de travail qu'un champ de
sorgho/mil?

oui non
mais?
'oui non
Commenter

 

224

C. INVITATION DE CULTURES ET AUTRES ACTIVITES COMMUNAEES

1.8zelles sent les activites communales auxquelles vous participez
gulié rement?
Invitation de culture

Construction de route
Construction de case
Autre (preciser)

 

 

 

 

 

 

2. Pour chaque activite preciser les periodes de l'année et les
membres de votre famille y participant '

 

 

 

3. Commenter sur le role et l'importance des activites communales

 

 

 

4. Est-ce que vous etes oblige de fournir du travail ou de la
nourriture aux membres de la communauté qui ont eu un faible
rendement a cause de maladie ou d' autres raisons (a specifier)?

 

 

 

D. SECURITE ALIMENTAIRE ET PROBLEMES GENERAUX

l. Quelles cultures sont plantées pour faire face a la famine en cas
de mauvaise saison?

 

/

2. En cas de deficit cerealier, quelles mesures sent prises pour
reduire la consommation ou obtenir de la nourriture d' ailleurs?

 

 

 

225

3. Dans la famille, qui est responsable du stock alimentaire et de
sa distribution?

 

4. Avez-vous eu des problémes particuliers avec certaines de vos
parcelles cette saison?

oui non

 

Si oui
Nom_parcelle Problémes

 

 

 

 

Problemes:

Invasion d'insecte
maladie

sécheresse

inondation .

damages causes par animaux
autre raison (a préciser)

mm-fide

E. COMMERCIALISATION DES PRODUITS

l. Quelles circuits commerciaux sont disponsibles a vous pour
a) vos cultures de rente?

 

 

b) Pour le surplus alimentaire

 

 

c) Pour le riz

 

 

2. A quelle distance de votre maison se trouvent les marches
a) Pour les cultures de rente

 

 

b) Pour le surplus alimentaire

 

 

226

c) Pour le riz

 

 

F. INNOVATIONS ET CONTACTS AVEC LES SERVICES DE VULGARISATION

I.

Pendant les cinq derniéres annees, mentionne un changement ou une
innovation importante que vous avez introduit dans votre exploita-
tion dans les domaines suivants:

. .'
cultures Vivrieres

 

- cultures perennes ou de rente

 

 

- elevage

 

 

- transformation des produitsagricoles

 

 

- commercialisation

 

 

- moyens de stockage

 

 

- equipements agricoles

 

 

. I . I
. Quand avez-vous pour la derniere fo1s consulte un encadreur

agricole ?
un agent d'elevage ?

227

3. Quelles utilisations faites-vous de vos parcelles de bas-fonds
pendant la saison seche?

 

 

 

4. Par rapport aux trois annees précédentes la récolte 1980 a été:
l = tres bonne

 

 

 

 

 

 

 

 

2 = bonne
3 = moyenne
4 = mauvaise
5 = trés mauvaise
Sorgho
Petit mil Pois de terre
Coton Manioc
Riz Haricot
Mais Soja

 

 

Patate Sesame

228

POIDS DES UNITES DE MESURE UTILISEES

IDENTIFICATION: Nom du menage
Zone

Date

 

CULTURE PESEE

POIDS EN KGS

 

Unite de
mesure Cult. Var. Forme

Poids de Poids
la tare

du recipient
rempli

 

 

 

 

 

 

 

 

229

EVALUATION SUBJECTIVE DE LA QUALITE DES DONNEES
(A REMPLIR PAR LES ENQUETEURS)
CHEF DE MENAGE:

 

ZONE:
DATE:

 

 

Comment évaluez-vous les responses du Chef du menage aux types de données
suivantes:

 

TYPE DE QUALITE RAISONS'DE LA
DONNEES ET ESTIMATIONS ISOUS OU SUR-
PRECISION ESTIMATION

 

Composition de son menage

 

ge des membres du menage

 

Temps de travail des membres
du menage

 

Nombre de parcelles sous
culture

 

Temps de travail des membres
en dehors du menage

 

Quantités recoltées de
chaque parcelle

 

lComposition du cheptel mort

 

 

 

 

 

 

Composition du cheptel vif

 

AUTRE COMMENTAIRES:

 

 

 

 

 

APPENDIX B

FIELD SPECIFIC FAMILY LABOR PROFILE FOR CROPS OTHER THAN
RICE IN ALL THE FOUR PRODUCTION SYSTEMS STUDIED

230

TABLE 8.]

FAMILY LABOR PROFILE FOR FIELDS BELONGING TO
S/M/C ENTERPRISE, SYSTEM I. 1980

 

 

Hours per Hectare

 

Labor Field Reference

 

Periods
Average 2/2 6/2 24/2

 

 

1 (13) 47
2 (2) 76 173 88
3 (2) 48 288
4 (2) 73 so
5 (2) 124 256 242
6 (2) 127
7 (2) 127 72 135 231
8 (2) 144 290 385
9 (2) 176 475 155
10 (2) 139 33
11 (2) 131 333 503 692
12 (2) 43 198 53
13 (2) 46 17
14 (2) 33 33 58
15 (2) 76 76 l85
16 (2) 86 103 115
17 (4) 17 88
Total (hrs/hai 1513 1372 l835 2409

Area Ind) .429 .36 .62 .23

 

 

231

TABLE B.2

FAMILY LABOR PROFILE FOR FIELDS BELONGING TO
MAIZE ENTERPRISE, SYSTEM 1, 1980

 

 

Hours per Hectare

 

 

 

 

$233335 Average Field Reference
5/9 5/9
1 (TB) 3
2 (2)
3 (2)
4 (2)
5 (2) 185
6 (2) 34 56 I40
7 (2) 104 22 56
8 (2) 4O
9 (2) 23
IO (2)
ll (2) ms
12 (2) 31 I67 421
T3 (2) 13
14 (’2)
15 (2)
16 (2)
l7 (4)
Total (hrs/ha) 538 244 6l8
Area (ha) .079 .09 .036

 

 

232

TABLE B.3

FAMILY LABOR PROFILE FOR FIELDS BELONGING T0
GROUNDNUT ENTERPRISE, SYSTEM 1, 1980

 

 

Hours per Hectare

 

 

 

 

Piaggds Average Field Reference
2/7 23/4
1 (l8)
2 (2)
3 (2) 3
4 (2)
5 (2) l5
6 (2) l8 123
7 (2) 27
8 (2) 24 123 248
9 (2) 15
lO (2) 17
ll (2) 16 177
12 (2) 6
l3 (2) 4
l4 (2) 29 46 106
15 (2) lO
l6 (2) 4
l7 (4) 5
Total (hrs/ha) 193 308 532
Area (ha) .319 .132 .336

 

 

233

TABLE 8.4

FAMILY LABOR PROFILE FOR FIELDS BELONING TO
BAMBERA NUTS ENTERPRISE, SYSTEM 1, 1980

 

 

Hours per Hectare

 

 

 

 

Pgtggds Average Field Reference
2l/5 20/7
l (18)
2 (2)
3 (2)
4 (2)
5 (2) 12
6 (2) 2
7 (2) 3l
8 (2) 9 138
9 (2) lO 962
10 (2)
ll (2) l7 704
l2 (2)
l3 (2) l
14 (2) ll l2
l5 (2) 12
16 (2) 2 235
17 (4)
Total (hrs/ha) 107 lSO 1901
Area (ha) .476 .26 l.4l

 

 

234

TABLE 3.5

FAMILY LABOR PROFILE FOR FIELDS BELONGING TO
SOYBEANS ENTERPRISE. SYSTEM 1, 1980

 

 

Hours per Hectare

 

 

 

 

 

Pgeggds Average Field Reference
4/6 4/6
1 (18)
2 (2)
3 (2)
4 (2) ll
5 (2)
6 (2) 34 T96 lO7
7 (2) 276 I73 95
3 (2)" 32 46 25
9 (2) 219
l0 (2) 157
ll (2) 135
l2 (2) 128 123 67
T3 (2) 68 108 59
14 (2) 65 15 8
15 (2)
16 (2)
l7 (4)
Total (hrs/ha). 1125 662 362

Area (ha) .252 .26 .476

 

 

235

TABLE B.6

FAMILY LABOR PROFILE FOR FIELDS BELONGING TO
S/M/C ENTERPRISE, SYSTEM 2. 1980

 

 

Hours per Hectare

 

 

 

 

Pgeggds Average Field Reference
28/1 46/2 55/2 44/1
1 (18) 11 150
2 (2) 51 93 112 78
3 (2) 14
4 (2) 62 230 144
5 (2) 34 192 156
6 (2) 40 44 217
7 (2) 72 121 51
8 (2) 57 35 94 148
9 (2) 29 20 71 42
10 (2) 23 19 6 38
11 (2) 41 102 54 9
12 (2) 20 48
13 (2) l
14 (2) 2 44
15 (2) 22 47 15 134
16 (2) 61 153
17 (4) 11 57
Total (hrs/ha) 551 922 293 823 689

Area (ha) 1.574 3.1 .59 .48 1.49

 

 

236

TABLE B.7

FAMILY LABOR PROFILE FOR FIELDS BELDNING TO
MAIZE ENTERPRISE, SYSTEM 2, 1980

 

 

Hours per Hectare

 

 

 

 

Pigggds Average Field Reference _
52/4 42/8 29/3 34/5
l (18)
2 (2)
3 (2) 45
4 (2) 22
5 (2) 49 736
6 (2) 155 264 333
7 (2) 357 956 222 490
8 (2) 237 185 222 273
9 (2) 103
10 (2) 71 89
11 (2) 15 9
12 (2) 66 179 111 64 67
13 (2) 12
14 (2) 5
15 (2) 3
l6 (2)
17 (4)
Total (hrs/ha) 1140 1585 889 1082 646
Area (ha) .049 .39 .009 .11 .045

 

 

237

TABLE B.8

FAMILY LABOR PROFILE FOR FIELDS BELONING TO
GROUNDNUTS ENTERPRISE, SYSTEM 2, I980

 

 

Hours per Hectare

 

 

 

 

5:5?365 Average Field Reference
55/7 52/9 41/8 32/4 46/8
1 (18)
2 (2)
3 (2)
4 (2)
5 (2) 6
6 (2) 26
7 (2) 228 538 150 2267
8 (2) 138 267 27
9 (2) 32 200 400 274
10 (2) 49 8 75 133 274
11 (2) 22 27
12 (2) 4
13 (2)
14 (2) 202 38 1050 1133 32 390
15 (2) 60 114
16 (2) 13
17 (4)
Total (hrs/ha) 780 785 1275 4200 200 939
Area (ha) .070 .13 .04 .015 .37 .095

 

 

238

TABLE 3.9

FAMILY LABOR PROFILE FOR FIELDS BELONGING TO
SOYBEANS ENTERPRISE, SYSTEM 2, 1980

 

 

Hours per Hectare.

 

 

 

 

5333665 Average Field Reference
48/11 40/6 46/7
1 (18)
2 (2)
3 (2)
4 (2)
5 (2) 47
6 (2)
7 (2) 248 1075 629
8 (2) 211 206
9 (2) '13 75 321
10 (2) 41 126
11 (2)
12 (2) 23 175
13 (2)
14 (2) 80 69 217
15 (2) 65
16 (2) 94 28 175
17 (4)
Total (hrslha) 822 478 1325 1293
Area (ha) .131 .32 .04 .14

 

 

239

TABLE 8.10

FAMILY LABOR PROFILE FOR FIELDS BELONGING TO
OKRA ENTERPRISE, SYSTEM 2, 1980

 

 

Hours per Hectare

 

 

 

 

Pigggds Average Field Reference
52/28 40/5 37/5 52/17
1 (18)
2 (2)
3 (2)
4 (2)
5 (2) 23
6 (2) 157 333 133 171
7 (2) 80 104
8 (2) 113
9 (2) 53 667 171
10 (2) 83 104 133
11 (2) 157 62 171
12 (2) 23 167
13 (2) 17 67
14 (2) 127 563 167 67 171
15 (2) 50 167 133 85
16 (2)
l7 (4)
Total (hrs/ha) 883 833 1500 533 769 ‘
Area (ha) .011 .048 .006 .015 .012

 

 

240

TABLE B.11

FAMILY LABOR PROFILE FOR FIELDS BELONGING T0
COTTON ENTERPRISE, SYSTEM 2, 1980

 

 

Hours per Hectare

 

 

 

 

Pigggds Average Field Reference

45/8 52/5

1 (18)

2 (2)

3 (2)

4 (2)

5 (2)

6 (2)
7 (2) 184 300 150
8 (2) 190 229

9 (2) 22
10 (2) 56 56
11 (2) V 273 1443

12 (2) 22 48

13 (2)

14 (2)

15 (2) 25 133

16' (2)
17 (4) 32 28
Total (hrs/ha) 804 1776 511

Area (ha) .277 .21 .50

 

 

241

TABLE 3.12

FAMILY LABOR PROFILE FOR FIELDS BELONGING T0
S/M/C ENTERPRISE, SYSTEM 3, 1980

 

 

Hours per Hectare

 

 

 

 

Pgtggds Average Field Reference
65/2 82/1 69/5
1 (18) 39 177 28
2 (2) 38 274 35
3 (2) 70 366 55
4 (2) 24
5 (2) 147
6 (2) 204 679 763 168
7 (2) 124 48 321 120
8 (2) , 136 119 468 37
9 (2) ‘ 152 190 361 46
10 (2) 118 395 37
11 (2) 61 60 229 32
12 (2) 23 71 34
13 (2) 5 86
14 (2) 28 60 14
15 (2) 29 23
16 (2) 121 71 214 111
17 (4) 28 366
Total (hrs/ha) 1347 1571 3803 683

Area (ha) .707 .419 .524 .926

 

 

242

TABLE 8.13

FAMILY LABOR PROFILE FOR FIELDS BELONGING TO
- MAIZE ENTERPRISE, SYSTEM 3, 1980

 

 

Hours per Hectare

 

 

 

 

5231335 Average Field Reference
61/14 83/3 59/4
1 (18)
2 (2)
3 (2) 1
4 (2) 16
5 (2) 31
6 (2) 27 50
7 (2) 50 562 56 55
8 (2) 49 812 89
9 (2) 16 71
10 (2) 34
11 (2) 9 37 5
12 (2) 20 12 50 25
13 (2) 21 250 268
14 (2) 1
15 (2)
16 (2)
17 (4)
Total (hrs/ha) 275 1675 175 492

Area (ha) .431 .08 .28 .23

 

 

243

TABLE 8.14

FAMILY LABOR PROFILE FOR FIELDS BELONGING T0
GROUNDNUTS/BAMBERA NUTS ENTERPRISE, SYSTEM 3, 1980

 

 

Hours per Hectare

 

 

 

 

5231365 Field Reference
Average
66/12 68/18 69/9
1 (18)
2 (2)
3 (2) 2
4 (2) 3 115
5 (2) 15
6 (2) 25 229
7 (2) 73 429 77 179
8 (2) 57 175 40
9 (2) 52 159 19 50
1O (2) 34 349 9 25
11 (2) 23 159
12 (2) 16 15
13 (2) 29 60
14 (2) 109 32
15 (2) 21 71 523
16 (2) 10 63 284
17 (4) 11
Total (hrs/ha) 480 1365 177 1520

Area (ha) .582 .308 .31 .578

 

 

244

TABLE B.15

FAMILY LABOR PROFILE FDR FIELDS BELONGING TO
SOYBEANS ENTERPRISE, SYSTEM 3, 1980 .

 

 

Hours per Hectare

 

 

 

 

P2318ds Average Field Reference
64/6 72/12 68/12
1 (18)
2 (2)
3 (2)
4 (2)
5 (2) 2 272
6 (2) 2
7 (2) 34 258 994 99
8 (2) 39 110 62
9 (2) - 29 545 74
10 (2) 31 136 2017 173
11 (2) 8
12 (2) 5
13 (2) 7
14 (2) 8 15 99
15 (2) 4 138 49
16 (2) 1
17 (4)
Total (hrs/ha) 170 955 3260 827

Area (ha) .494. .466 .149 1.96

 

 

245

TABLE 8.16

FAMILY LABOR PROFILE FDR FIELDS BELONGING TO
S/M/C ENTERPRISE, SYSTEM 4, 1980

 

 

Hours per Hectare

 

 

 

 

11231665 Average Field Reference
91/1 102/13 112/3
1 (18) 2 86
2 (2) 13
3 (2) ' 20
4 (2) 21 18 7
5 (2) 48 29
6 (2) 44 18 - 129 6
7 (2) 198 170 471 145
8 (2) 235 134 329 131
9 (2) 171 177 257 55
10 (2) 243 114 286 124
11 (2) 212 329 57
12 (2) 27 57
13 (2)
14 (2) 14 429
15 (2) 28 16 44
16 (2) 160 108 129 14
17 (4) 2
Total (hrs/ha) 1,438 737 2529 583

Area .(ha) .766 2.05 .14 .87

 

 

246

' TABLE 8.17

FAMILY LABOR PROFILE FOR FIELDS BELONGING TO
MAIZE ENTERPRISE, SYSTEM 4, 1980

 

 

Hours per Hectare

 

 

 

 

 

tggggds Average [Field Reference.
105/2 115/2 107/2
1 (18)
2 (2) _
3 (2)
4 (2) 3
5 (2) 40 153
6 (2) 87 30 162
7 (2) 180 15 167
8 (2) 7D 7 63
9 (2) 107 59
10 (2) 115 59 126
11 (2) 71 505
12 (2) 155 59 175
13 (2)
14 (2) 12
15 (2)
16 (2) 222
17 (4)
Total (hrs/ha) 840 . 333 456 1022

Area (ha): .093 .27 .068 .095

 

247

TABLE 8.18

FAMILY LABOR PROFILE FDR FIELDS BELONGING TO
GROUNDNUTS ENTERPRISE, SYSTEM 4, 1980

 

 

HoUrs per Hectare

 

 

 

 

Fggggds Average Fie1d Reference
94/7 113/2 101/4
1 (18)
2 (2)
3 (2) 3
4 (2)
5 (2) 10
6 (2) 36 73 89
7 (2) 94 55 250
8 (2) 62 55
9 (2) 37 83
1o (2) 80 250 35
11 (2) 17 291
12 (2) 4
13 (2) 90 873
14 (2) 224 53
15 (2) 22
16 (2)
17 (4)
Total (hrs/ha) 679 1345 583 177
Area (ha) .118 .055 .48 .11

 

 

248

TABLE 8.19

FAMILY LABOR PROFILE FOR FIELDS BELONGING TO
BAMBERA NUTS ENTERPRISE, SYSTEM 4, 1980

 

 

 

Hours per Hectare

 

 

 

$231365 Average Fie1d Reference

110/4 102/7 105/4
1 (18)
2 (2)
3 (2)
4 (2)

5 (2) 30 326
6 (2) 140 714
7 (2) 80 217 103
8 (2) 380 2071 326 68
9 (2) 100 429 217
10 (2) 50

11 (2)
12 (2)
13 (2)

14 (2) 1143 2609 547

15 (2) 740

16 (2)
17 (4)

Total (hrs/ha). 1520 4357 3696 718

Area (ha) .042 .014 .0092 .058

 

 

 

249'

TABLE 8.20

FAMILY LABOR PROFILE FOR FIELDS BELONGING TO
OKRA ENTERPRISE, SYSTEM 4, 1980

 

 

Hours per Hectare

 

 

 

 

thggds Average Fie1d Reference
114/4 110/10 107/8 103/5
1 (18)
2 (2)
3 (2) 5
4 (2) 10
5 (2) 15 167 107 370
6 (2) 33 83 273 123
7 (2) 22 167 182 71 123
8 (2) 12 83 36 123
9 (2) 17 167 91 123
10 (2) 3 83
11 (2) 3 167
12 (2) 3 167 123
13 (2) 20 167 273 36
14 (2) 1
15 (2)
16 (2)
17 (4)
Total (hrs/ha) 144 1250 818 250 988

Area (ha) .065 .012 .011 .028 .016

 

 

BIBLIOGRAPHY

BIBLIOGRAPHY

A. References Cited

Abalu, 6.0.1., and Silva, B.D. (1979). "Socioeconomic Analysis of
Existing Farming Systems and Practices in Northern Nigeria. "
Paper presented at the Workshop on Socioeconomic Constraints to
Development of Semi-Arid Tropical Agriculture, ICRISAT.

 

Abalu, 6.0.1. (1976). "A note on crop mixtures under indigenous con-
ditions in Northern Nigeria." Journal of Development Studies,
12:214.

Anderson, J. R. (1974). "Simu1ation: Methodology and Application in
Agricultural Economics. " Review of Marketing,andAgricultural
Economics, 42(1): 3-55.

Anderson, J. R. , et a1.(1977).ricu1tura1 Decision Analysis. Ames,
Iowa: Iowa State University Aress.

Andrews, D.J. (1970). "Relay and Intercropping with Sorghum at
Samaru: Proc. Seminar on Intercropping." Ford Foundation/IITA]
IRAT, Ibadan, pp. 1-12.

Andrews, D.J. (1972). "Intercropping with Sorghum in Nigeria." Exper-
imental Agriculture, VIII:2.

Arrivets, J. (1973). "Résu1tats de Rizicu1ture Sans Amenagement en
Zone Soundanienne - L'exemple de Bas-Fonds du Centre Haute-Volta."
L'Agronomie Tropicale, 28(1):34-53.

Bar10w, C. , et a1. (1980)."On Measuring the Economic Benefits of New
Technologies to Small Rice Farmers. " IRRI Paper No. 78-10. Las
Banos, Philippines.

Barnard, C. S. and Nix, J. S. (1973). Farm Planningkand Control. Cam-
bridge, United Kingdom: Cambridge University Press.

CERCI. (1975-78). "Rapport annuel." Mimeo, Bobo - Dioulasso.

Collinson, M. P. (1972). Farm Management in Peasant Agriculture, A
Handbook for Rural Development Planning in Africa. New York:
Praeger.

Crawford, E. H. (1980). "A Programming Simulation Study of Constraints
Affecting the Long-Run Income-Earning Ability of Traditional Dry-
land Farming Systems in Northern Nigeria. " Ph. D. Dissertation,
Cornell University, Ithaca, New York.

Delgado, C.L. (1978). "Livestock versus Food Grain Production in South-
eastern Upper Volta: A Resource Allocation Analysis.“ Ph.D. Disser-
tation, Cornell University, Ithaca, New York.

 

 

250

251

FAO/FIDA. (1980). "Rapport de la Mission de Projet de Developpement
Rural dans 1'ORD de 1'Est, Haute-Volta." Rapport No. 24/80 DDC
UPV S..

FAO. (1965-803). "Production Yearbook" Vols. 19-34, Rome.
FAD. (1965-80b). "Trade Yearbook." Vols. 19-34, Rome.

FDR. II. (1978). "Comite Technique International de Coordination du
Fonds de Developpement Rural." Technical Report.

Fotzo, P.T. (1977). "Resource Productivity and Returns in Rice Produc-
tion Under Alternative Farming Systems: A Comparative Study in the
Northwest Province of Cameroon." Masters Thesis, University of
Ibadan, Ibadan, Nigeria.

 

Flinn, J.C., gt_gl, (1980). "Incorporating Multiple Objectives into
Plans for Low-Resource Farmers." Australian Journal ongricul-
tural Economics, 24(1):35-45.

 

 

Ghodake, R.D. (1981). "The Potential of Mathematical Programming for
the Analysis of Yield Gaps in Semi-Arid Tropical Agriculture."
ICSISAT Economics Program Progress Report 24, Patancheru, A.P.,

n 1a.

Ghodake, R.D. and Hardaker, 5.8. (1981). "Whole Farm Modeling for
Assessment of Dryland Technology." ICRISAT Economics Program,
Progress Report 29, Patancheru, A.P., India.

Hazell, P.B. (1971). "A Linear Alternative to Quadratic and Semi-
variance Programming for Farm Planning Under Uncertainty." Ameri-
can Journal of Agricultural Economics, 53(1):53-62.

Heyer, J. (1971). "A Linear Programming Analysis of Constraints on
Peasant Farms in Kenya." Food Research Institute Studies, 10:
55-67. “

 

IFAD-ADB. (1981). "Rural Development Project in the Eastern 0RD."
Appraisal Report.

IRAT. (1979-80). "Campagne, 1979 - Resume des Travaux, Conclusions."
Ouagadougou, Haute-Volta. . .

Kamuanga, M. (1982). "Farm Level Study of the Rice Production System
at the Office du Niger in Mali: An Economic Analysis." Ph.D.
Dissertation, Michigan State University, East Lansing, Michigan.

Kelean, Y. (1972). "Contribution a 1:étude des Aptitudes des sols a
Rizi-culture de Bas-Fonds sans Amenagement dans 1e Nord-Dahomey

(gas-Fonds de Serarou et Bagou).: L'agronomie Trgpicale, 27(3):
,3 -357.

252

Kelean, Y. and Fussor, Y. (1973). "Méthodes d'Investigation pour
l' Analyse et 1e Classement des Bas-Fonds dans quelques Regions de
1' Afrique de 1' Guest. Proposition de Classification d' Aptitudes
des Terres a la Rizicu1ture. " L' Agronomie Tropicale, 28(2): 156-
172.

Lassiter, 6.6. (1981). "Cropping Enterprises in Eastern Upper Volta."
African Rural Economy Working Paper No. 35, Department of Agri- -
cuTtural Economfics, Michigan State University, East Lansing,
Michigan.

Mehretu, A. and Nilcock,D. (1979)."P1anification Regionale: Aspects
Geographiques de la Region de 1' Est. " BAEP ORD de 1' Est.

Norman, B.N. (1973). "Economic Analysis of Agricultural Production
and Labour Utilization Among the Hausa in the North of Nigeria."
African Rural Employment Paper No. 4, Department of Agricultural
Economics, Michigan State UniverSity, East Lansing, Michigan.

. (1974). “Rationalizing Mixed Cropping Under Indigenous
Conditions: The Example of Northern Nigeria." Journal of Develop-
ment Studies, 11:3-21.

 

 

Okigbo, B.N. (1978). CroppipgSystems and Related Research in Africa.
Association for the AdVancement ongricfiltural Scienceiin Africa.

Robinson, K.L. (1974). "The Economics of Increasing Staple Food
Produgtion in West Africa." Ibadan, Nigeria, IITA. (Unpublished
paper .

Spencer, D.S.C. (1972). "Micro-Level Farm Management and Production
Economics Research Among Traditional African Farmers: Lessons from
Sierra Leone. " African Rural Emplgyment Paper No. 3, Department
of Agricultural Economics, Michigan State University, East Lansing,
Michigan.

Swanson, R. A. (1975). Gourmantche Agriculture; Part I. "Land Tenure
and Field Cultivation. Eastern 0RD, BAEP Upper Volta, Document
No.7. .

Tapsoba, E. (1981). "An Economic and Institutional Analysis of Formal
and Informal Credit in Eastern Upper Volta: Empirical Evidence and
Policy Implications." Ph.D. Dissertation, Michigan State Univer-
sity.

WARDA. (1975). "Varietes de riz recommendees dans la region de
1'ADRAO. " M_i__meo, Monrovia.

WARDA. (1979). "Le Riz en Afrique de 1'Ouest." Mimeo, Monrovia.
HARDA. (1981). "Rice Statistics Yearbook." Mimeo, Monrovia.

253
Weldring, J. (1979). "Synthése sur les Amenagements Hydro-Agricoles
dans l'ORD de 1'Est, Fada N'Fourna." Direction du Fonds de
Developpement Rural, Ouagadougou, Upper Volta, A Report.
World Bank. (1981). World Development Report, 1980. Washington, D.C.

(1982). World Atlas. Washington, D.C.

 

 

B. Other References Consulted

 

Angladette, A. (1966). "Le Riz.” in Techniques Agricoles et Produc-
tions Tropicales, Nos. 5, 6, 7. Ed} Maison Neuve et Larasse, Paris.

Barrett, V., at al, (1982). "Animal Traction in Eastern Upper Volta:
A Technical, Economic and Institutional Analysis." MSU Inter-
national Development Paper No. 4, Department of Agricultural Econ-
omics, MiEhigan State University, East Lansing, Michigan.

Beneke, R.R. and Winterboer, R. (1980). Linear Programming Applica-
tions to Agriculture. The Iowa State University Press, Ames,
Iowa. .

Brown, M. (1979). "From Farm Income Analysis to Agricultural Project

Analysis." Worldpgank Staff Occasignal Papers, up, 29, The John

Hopkins University Press, Baltimore and London.

Candler, W. and Slade, R. (1981). "Collection of Reliable Farm Level
Data in LDCs." Journal of Agricultural Economics, 30(1):65-70.

.Collinson, M.P.' (1982). "Farming Systems Research in Eastern Africa:
The Experience of CIMMYT and Some National Agricultural Research
Services, 1976-81.” MSU International Development Paper No. 3,
Department of Agriculturai Economics, Michigan State University,
East Lansing, Michigan.

Delgado, C. (1976). "The Southern Fulani Farming System in Upper
Volta: A Model for Integration of Crop and Livestock Production in
the West African Savannah." African Rural Economy Paper No. 20,
Department of Agricultural Economics, Michigan State UniverSity,
East Lansing, Michigan.

Dillon, J.L. and Hardaker, J.B. (1980). "Farm Management Research for
Small Farmer Development." Rome, FAO, Agricultural Services Bul-
letin No. 41.

 

Eicher, C.K. and Baker, D.C. (1982). "Research on Agricultural Devel-
opment in Sub-Saharan Africa: A Critical Survey." MSU International
Development Paper No. 1, Department of Agricultural Economics,
Michigan State University, East Lansing, Michigan.

Erhabor, P.O. (1982). "Efficiency of Resource Use Under Small-Scale
Irrigation Technology in Nigeria: A Case Study of Shadoof and Pump
in Ringim LGA, Kano, Nigeria." Ph.D. Dissertation, Purdue Univer-
sity.

254

Friedrich, K.H. (1977). "Farm Management: Data Collection and Analysis."
Rome, Food and Agriculture Organization, Agricultural Service Bul-
letin No. 34.

 

Fotzo, P.T. (1981). "Analyse Preliminaire des Enquetes Agro-Economiques
des Bas-Fonds de l'ORD de 1'Est, RHV." Section Recherche Appliquee,
BAEP, 0RD de 1'Est, Fada N'Gourma, Upper Volta, A Report.

Galt, D., et a1. (1982). "Farming Systems Research (FSR) in Honduras,
1977-8131 Case Study." MSU International Devemeent Workin Paper
No.1, Department of AgricUltural Economics, Michigan State Un ver-
s1ty, East Lansing, Michigan.

Gilbert, E. H. , et a1. (1980)."Farming Systems Research: A Critical
Survey. " MSU—Rural Development Paper No. 6, Department of Agricul-
tural Economics, Michigan State University, East Lansing, Michigan.

Humphreys, C.P. and Pearson, S.R. (1979-80). "Choice of Technique in
Sahelian Rice Production." Food Research Institute Studies, 17(3),
Stanford University.

Kline, C.K., Lt L1. (1969). "Agricultural Mechanization in Equatorial
Africa." _Institute of International Agriculture, Research Report
No. 9, Michigan State UniVersity, East Lansing, MiEhigan.

Lagemann, J. (1981). FarmingASystems Research as a Tool for Identify-
ing an%Conducting ResearCh and"Deve10pment’Projects. TTurrialba,
Costa ica

Lang, H. (1979). The Economics of Rainfed Rice Cultivation in West
Africa: The Case oiilvor§:Coast. Soarbrucken, West Gérmeny, Verlog
Breinterbach. .

 

Lassiter, G. C, Lt L1. (1982). "Applied Research in Eastern Upper Volta:
Major Findings and Recommendations for Future Programs. " Mimeo,
Department of Agricultural Economics, Michigan State University,
East Lansing, Michigan.

Mann, R.D. (1974). A Survey Technique for Identif in the Needs of
Small Farmers, and an Example of Its Use in Gambia. London, Inter-
mediate Technology Publications, Ltd.

Meyer, J. and Bonnefond, R. (1973). "Les Rizicu1tures Paysannales -
Ameliorations Ppssibles." Republique Francaise, Secretariat aux
Affaires Etrangeres, Paris, Presses Muray.

Pearson, S.R., Lt L1. (1981). Rice in West Africa: Poligy_and Econ-
omics. Stanford, Stanford University Press.

 

Rigoulot, J.P. (1980). "An Analysis of Constraints on Expanding Rice
Output in the Casamance Region of Senegal." African Rural Economy,
Working Paper No. 31, Department of Agricultural Economics, Michi-
gan State University, East Lansing, Michigan.

255

Sargent, M., g§_a1. (1981). "An Assessment of Animal Traction in
Francophone-west Africa." African Rural Economy Working_Paper No.
34, Department of AgriculturaTEconomics, Michigan StateiUniVer-
sity. East Lansing, Michigan.

Spencer, D.S.C., §3_91, (1979). "Annual Costs, Returns and Seasonal
Labor Requirements for Selected Farm and Non-Farm Enterprises in
Rural Sierra Leone." African Rural Economy Working Paper No. 27,
Department of Agricultural Economics, Michigan State University,
East Lansing, Michigan.

Valdes, A., g; a1. (1979). Economics and the Design of Small Farms'

Technology. Ames, Iowa:‘Iowa State UniverSity Press.

Winch, F. (1976). "Costs and Returns of Alternative Rice Production
Systems in Northern Ghana: Implications for Output, Employment and
Income Distribution." Ph.D. Dissertation, Michigan State Univer-
sity.

\

 

. H ——r

STRT

N V.

111

930 0707 53

L

111

RQRIES

 

1.1"."

n‘