v5.

Lflma
.2.

, 3.253.:
L *- .

1.: 1 L...
.5431. r.

(a.
.J- .

e .
. .7 .1. x) 931-
....rnéofx it. 4. 2:113...»
I l 10.... 5...: v . l.
7...... .

: .l
a S 1
291.33;

.. xii}.
w ,.!.1 .
.31

u) I. .
:3. .4 5...!tsn

.7.rsuwt «it. :.
.1: . )\1txl.: Q:
. ..

 

.. x: :
E 1:...

 

 

Humane
3
9001

This is to certify that the

dissertation entitled

Essays on the Economics of Cotton Production in Zimbabwe:
Policy Implications for Technology Adoption, Farmer
Health and Market Liberalization

presented by
Blessing Mukabeta Maumbe

has been accepted towards fulfillment
of the requirements for

Ph.D. degree in Agricultural Economics

 

 

     

gag/W

__Sco.t.t_M- Swinton

Major professor

 

Date May 31, 2001

MS U is an Affirmative Action/Equal Opportunity Institution 0- 12771

 

 

LIBRARY

Michigan State

University

 

 

PLACE IN RETURN BOX to remove this checkout from your record.

TO AVOID FINES return on or before date due.

MAY BE RECALLED with earlier due date if requested.

 

DATE DUE

DATE DUE

DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 c:/CIRCIDateDue.p65-p.15

ESSAYS ON THE ECONOMICS OF COTTON PRODUCTION IN ZIMBABWE:
POLICY IMPLICATIONS FOR TECHNOLOGY ADOPTION, FARMER
HEALTH AND MARKET LIBERALIZATION

Blessing Mukabeta Maumbe

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements

for the degree of

DOCTOR OF PHILOSOPHY

Department of Agricultural Economics

2001

ABSTRACT

ESSAYS ON THE ECONOMICS OF COTTON PRODUCTION IN ZIMBABWE:
POLICY IMPLICATIONS FOR TECHNOLOGY ADOPTION, F ARMER

HEALTH AND MARKET LIBERALIZATION

By

Blessing Mukabeta Maumbe

Cotton is grown in more than 30 Sub-Saharan African (SSA) countries and is a
vital source of employment, foreign exchange and raw material for textile industries. But
cotton industries in Africa face several strategic threats. In Africa, and in particular in
Zimbabwe, long-term viability of cotton production will depend on both sustainable
technology and policy. A key part of technological sustainability is mitigation of
pesticide-related farmer health risks. Integrated Pest Management (IPM) provides some
tools to reduce pesticide risks, but why farmers do not adopt IPM in African cotton
production is less understood. In the policy arena, both economic and military policy can
affect agricultural production. Zimbabwe’s cotton sector has faced important policy
shocks from Structural Adjustment Programs (SAPS) and the end of civil conflict in

neighboring Mozambique.

The first essay analyzes acute health effects associated with pesticide use in two
cotton-producing zones of Zimbabwe. The initial cost of illness regression model shows
that farmer’s reported acute symptoms are key determinants of farmer health costs.
Poisson regression model results reveal that pesticide-induced acute symptoms are linked
to the most toxic pesticides, use of leaking Sprayers, label illiteracy, alcohol intake, and
taking meals in the fields after spraying. Exposure averting and mitigating strategies that
significantly reduce the incidence of acute symptoms include protective clothing,
knowledge of first aid and predisposition toward reform calendar-based spray strategies.

The second essay examines determinants of cotton-IPM adoption in the same two
zones. Results from a Poisson model shows that farmer’s knowledge is the most
important factor influencing the uptake of IPM technology. Pesticide-related health risks
played no significant role in the adoption of IPM technology.

The third essay analyzes the determinants of Zimbabwe’s cotton supply since
1980. It finds that SAPS have a negative impact on cotton acreage for both the large-scale
commercial and small-scale communal farmers. Results show that large-scale cotton
growers respond strongly to economic incentives while institutional factors matter for
smallholders. Cessation in 1992 of the conflict in Mozambique is associated with positive
cotton supply response among Zimbabwe’s smallholders. Opportunities for cotton
expansion lie with widespread diffusion of technical innovations and refinement of the

on-going SAPS in order to generate positive supply response in future.

Copyright by
BLESSING MUKABETA MAUMBE

2001

DEDICATION

To my wonderful parents, Enoch and Violet who have given me so much love, and I give
that love in return to you. To the struggling smallholder cotton growers in the semi-arid

regions of Africa, hoping that you reap a harvest of hope not sorrow.

ACKNOWLEDGMENTS

I owe a profound debt of gratitude to many people who provided moral and
intellectual support as l was working on this dissertation. 1 am especially grateful to my
thesis advisor and chairman of my graduate committee, Professor Scott Swinton for his
valuable support and guidance throughout the completion of this work. His brilliance and
advice had and continue to have, a profound influence on my academic and professional
development. I was blessed with energetic and enthusiastic graduate committee members;
Professors Carl Eicher, Eric Crawford, Chris Petersen and Carl Liedholm. 1 would like to
offer my sincere gratitude for their sterling assistance in the formulation and execution of
this dissertation. My fellow graduate students in the Department of Agricultural
Economics played a significant role in many different ways during the five years of my
doctoral program at Michigan State University and deserve my salute.

This work could not have been completed without mentioning the many people
whose advice and wisdom helped shape this study. I am grateful to Dr. Derek Byerlee,
Gerd Fleischer, Jaimie Cuzon, and Gershon Feder all of the World Bank. 1 am equally
indebted to Professor George Norton of Virginia Tech University, Blacksburg for his
help and insights in the conceptualization of my study. I wish to acknowledge the support
of Takashi Yamano, and Margaret Beaver of the MSU Food Security II Project for
assisting me with data file management and analysis.

I am also grateful to Professor Ananthius Mphuru, the Dean of the College of
Agriculture at Africa University in Zimbabwe, for his valuable support over the years and
giving me the opportunity to go on study leave to accomplish this task. I also appreciate

the help rendered by Dr. Robert Armstrong and Mrs. Florence Chokuda both members of

vi

Faculty at Africa University. Special thanks to Professor Mandivamba Rukuni and
Godfrey Mudimu of the Department of Agricultural Economics at the University of
Zimbabwe for their assistance and advise. This study benefited from the enthusiasm and
energy of Richard Mupfiga, Monica Chiunda, Lemy Mutaviri, and Zvanakireni Chitiza
who tirelessly worked with me as part of my research team from Africa University.
Together we traveled hundreds of miles day and night, and we spent many hours in
remote cotton fields and learned a lot from each other. Of course, my best teachers were
the cotton growers in Sanyati and Chipinge, who volunteered their time and generously
shared their cotton production experiences with us. Data entry benefited from the diligent
work of Tracy Mapenzeuswa to whom I am equally grateful.

My greatest debt of gratitude, as usual, is to my wife Kudzayi, and daughters
Amanda, Wayne and Samantha. I really appreciate your patience, love and great support.
To my parents Enock and Violet, and my eldest brother Michael, I thank you for laying
the foundation to my education, and above all, for believing in me. Special thanks goes to
the University United Methodist Church community for their moral and spiritual support
during my stay in East Lansing. Finally, I am eternally grateful to the WK. Kellogg
Foundation for funding my studies at Michigan State University, and to Rockefeller

Foundation for providing the resources to complete the fieldwork in Zimbabwe.

vii

TABLE OF CONTENTS

List of Tables .......................................................................................... x
List of Figures .................................................................................................................. xiii
INTRODUCTION .............................................................................................................. l
1.1 The Importance of Cotton Production in Africa ........................................................... 1
1.2 Study Objectives and Thesis Organization .................................................................. 4
1.2.1 Farmer Health Risks in Smallholder Cotton Production in Africa ..................... 5
1.2.2 The Adoption of Integrated Pest Management (IPM) in Zimbabwe .................. 7
1.2.3 The Impact of Structural Adjustment Programs (SAPS) in African
Agriculture ........................................................................................................... 9
References ......................................................................................................................... 11
CHAPTER 22ECONOMIC AND HEALTH IMPACTS OF PESTICIDE USE IN
ZIMBABWE’S SMALLHOLDER COTTON PRODUCTION ...................................... 15
2.0 INTRODUCTION ...................................................................................................... 15
2.1 Objectives of the Study ............................................................................................... 19
2.2 Methodology and Data ................................................................................................ 20
2.2.1 Theoretical Model ............................................................................................. 20
2.2.2 Data Sources and Sampling Strategies ............................................................. 23
2.2.3 Empirical Model: Estimation of the Health Cost of Pesticide Exposure .......... 25
2.3 RESULTS AND DISCUSSION ................................................................................. 28
2.3.1 Farmer Health Cost of Pesticide Use: Cost of Illness Model Results ................ 29
2.3.2 Evidence of Health Impairments Among the Cotton Growers .......................... 30
2.3.2.1 Incidence of Skin Effects ....................................................................... 30
2.3.2.2 Incidence of Eye Effects ........................................................................ 33
2.3.3 Multiple Health Impairments: Total Symptom Incidence Model Results ......... 35
2.3.4 Poisson Protective Clothing Model Regression Results .................................... 37
2.4 Conclusions ................................................................................................................. 39
2.4.1 Policy Options and Knowledge Gaps ................................................................ 42
References ......................................................................................................................... 54
Appendices ........................................................................................................................ 61
Appendix 2A1: Langragian Function ............................................................................... 61
CHAPTER 3: ADOPTION OF COTTON I.P.M. IN ZIMBABWE: THE ROLE OF
TECHNOLOGY AWARENESS AND PESTICIDE-RELATED HEALTH RISKS ...... 69
3.0 Introduction ................................................................................................................. 69
3.1 Problem Overview ...................................................................................................... 72
3.2 Study Objectives ......................................................................................................... 74
3.3 The Diffusion of IPM Technologies in Less Developed Countries (LDCs) .............. 75
3.4 Current Status of IPM Adoption in Less Developed Countries (LDCs) .................... 77
3.5 Defining Smallholder Cotton-1PM Adoption ............................................................. 79
3.6 Methodology and Data ................................................................................................ 80
3.6.1 Economic Behavioral Model ............................................................................. 80

viii

3.6.2 Data and Estimation ........................................................................................... 84

3.6.3 Empirical Model ................................................................................................ 86
3.7 Results and Discussion ............................................................................................... 87
3.7.1 Farmers’ Adoption Patterns and Pest Management Perspectives ...................... 87
3.7.2 Factors Affecting Cotton-IPM Adoption Among Smallholder Farmers ........... 88
3.8 Conclusion .................................................................................................................. 90
3.8.1 Policy Implications and Suggestion for Future Research .................................. 90
References ......................................................................................................................... 94
Appendices ...................................................................................................................... 101
CHAPTER 4: COTTON SUPPLY RESPONSE TO MARKET LIBERALIZATION IN
ZIMBABWE ................................................................................................................... 104
4.1 Introduction ............................................................................................................... 104
4.1.1 Overview of Cotton Production Challenges in Africa ..................................... 107
4.1.2 Contemporary Issues in Cotton Production in Zimbabwe ............................... 109
4.1.3 Adoption of Structural Adjustment Programs (SAPS) In Zimbabwe .............. 111
4.2 Methodology and Data .............................................................................................. 113
4.2.1 Theoretical Model for Cotton Supply Response .............................................. 113
4.2.2 Empirical Econometric Model ......................................................................... 120
4.2.3. Data Considerations ........................................................................................ 121
4.3.1 Large Scale Commercial Cotton Supply Response Results ............................ 122
4.3.2 Smallholder Cotton Supply Response Results ................................................. 125
4.4 Conclusion ................................................................................................................ 127
References ........................................................................................... 1 34
Appendices ...................................................................................................................... 1 50
CHAPTER 5: SUMMARY AND CONCLUSION ........................................................ 157
APENDICES: QUESTIONNAIRES .............................................................................. 164
APPENDIX A1: Household Questionnaire .................................................... 164
APPENDIX A2: Household Member Health Questionnaire ................................. 185
APPENDIX A3: Field Level Questionnaire ................................................... 194

ix

LIST OF TABLES

Table 2.1: Descriptive statistics for Sanyati and Chipinge districts, 1998/99 .................. 44
Table 2.2: Cost of Illness Model Results for Sanyati and Chipinge Districts, 1998/99. .. 45
Table 2. 3: Poisson Model Results for Self-reported Acute Skin Symptom Incidences. .46
Table 2. 4: Poisson Model Results for Self-reported Acute Eye Symptom Incidences. .. 47

Table 2. 5: Poisson Results for Self-reported Total Acute Symptom Incidences, 1998/99
........................................................................................................................................... 48

Table 2.6: Poisson Protective Clothing Adoption Model Results, 1998/99 ..................... 49

Table 2.7: Probit Results for Eating in Cotton Fields in Survey Districts, 1998/99 ......... 50

Table 2.8: Probit Results for Leaking Sprayers in Survey Districts, 1998/99 .................. 51
Table 2. 9: Probit Results for First Aid Knowledge in Survey Districts, 1998/99 ........... 52
Table A1.1: Pesticide-related Health Symptoms for Cotton Growers, 1998/99 .............. 62

Table A1.2: Pesticide-related Farmer Health Characteristics in Study Regions, 1998/99 63

Table A1.3: Distribution of Pesticide Storage Strategies in Study Regions, 1998/99 ...... 64

Table A1.4: Label Literacy Among Cotton Growers in Survey Areas, 1998/99 ............. 65

Table A1 .5: Knowledge of Triangles in Correct Order of Toxicity, 1998/99 .................. 65

Table A1.6: Pesticide Use and Toxicity Classes by Farmers in Survey Areas, 1998/99 . 65
Table A1 .7: Cotton Pesticides Used in Sanyati and Chipinge Districts, 1998/99 ............ 66
Table A1.8: Spaying Equipment Distribution in Cotton Growing Regions, 1998/99 ...... 67
Table A1 .9: Average Cotton Pesticide Applications By Region, 1998/99 ....................... 67
Table A1.10: Pesticide-related Treatment Patterns in Survey Areas, 1998/99 ................. 67
Table A1.11: Types of Protective Clothing Worn By Pesticide Applicators, 1998/99. 68
Table A. 12: Mean Health Costs AS a Proportion of Household Income and Costs ......... 68

Table 3.1:Descriptive Statistics on Variables Used in Poisson Regression Model .......... 92

Table 3.2: Determinants of Cotton IPM Practice Adoption in Sanyati District, 1998/99 93
Table A3.1: Cotton IPM Practice Probit Model Results for Sanyati District, 1998/99.. 101
Table A3.2: Cotton IPM Practice Probit Model Results for Chipinge District, 1998/99102
Table A3.3: List of Cotton IPM-related Practices Used in Survey Areas, 1989/99 ....... 103

Table 4.1 Description of Variables Used in Zimbabwe’s Cotton Supple Response
Models, 1980-1997 ......................................................................................................... 131

Table 4.2: Zimbabwe Cotton Acreage Response Model Results, 1980-1997 ................ 132

Table A4.1: Results of Diagnostic Tests on the Cotton Acreage Response Models ...... 145

xi

Table A4.2: Zimbabwe Cotton Yield Response Model Results, 1980-1997 .................. 146

Table A4.3 Agricultural Policy Changes in Zimbabwe, 1980- 2000 ............................. 147

Table A4.4: Zimbabwe Data for Cotton Supply Response Analysis, 1980-1997 .......... 148

Table A4.5: Zimbabwe Data for Cotton Supply Response Analysis Continued, 1980-1997
......................................................................................................................................... 148

Table A4.6 Dickey Fuller (DF) Unit Root Test (constant without trend) ...................... 149

Table A4.7: Zimbabwe Average Annual Growth Rates of Official Prices, 1980-1995 . 149

xii

LIST OF FIGURES

Figure 2.1: Structure of Econometric Analysis ................................................................. 53
Figure 4.1: Zimbabwe Trends in Cotton Area Cultivated, 1980-2000 ........................... 133
Figure A4.1: Structure of the Cotton Industry in Zimbabwe .......................................... 150
Figure A 4.2: Zimbabwe Cotton Yield Trends, 1980-1999 ............................................ 151
Figure A4.3: Zimbabwe Seed Cotton Production Trends, 1980-199 ............................. 151

Figure A4.4: Zimbabwe Trends in Relative Real Agricultural Producer Prices, 1980-1996

......................................................................................................................................... 152
Figure A4.5: Zimbabwe Nominal Crop Producer Prices, 1980-1998 ............................ 152
Figure A4.6: Zimbabwe Real Crop Producer Prices, 1980-1998 ................................... 153
Figure A4.7: Comparison of Local and World Cotton Prices, 1980-1997 ..................... 153
Figure A4.8: Zimbabwe Macroeconomic Trends, 1980-1998 ........................................ 154
Figure A4.9: Zimbabwe Cotton Producer Price Trends, 1980-1997 .............................. 154
Figure A4.10: Zimbabwe Maize Producer Price Trends, 1980-1998 ............................. 155
Figure A4.1 1: Zimbabwe Minimum Wage Trends, 1980-1999 ..................................... 155
Figure A4.12: Zimbabwe Inflation Rate Trends, 1980-2000 ......................................... 156

xiii

Figure A4.13: Zimbabwe Real Interest Rate (nominal-inflation) 1980-1997 156

xiv

CHAPTER 1

INTRODUCTION
1.1 The Importance of Cotton Production in Africa

Cotton iS grown in more than 30 Sub-Saharan African (SSA) countries (Lele,
1989). Although SSA countries differ greatly in their geographic and physical conditions,
political disposition, weather patterns and cultural heritage, they do Share some common
patterns in their economic structures. In most if not all the African countries, agriculture
is the dominant sector crucial to the well being of the economy. Over the past century,
African agriculture has not grown fast enough to realize its potential as the engine for
economic growth and poverty reduction (Binswanger, and Townsend, 2000).
Nonetheless, cotton the “white gold” is one of Africa’s most successful agricultural
export stories (Bingen et a1. 1998). The annual market value of African lint exports is
about USS] billion (Lele, 1989). Given that about 50% of Sub-Saharan Africans live
below the poverty line, efforts to accelerate cotton production will offer tremendous
opportunities for reducing the problem of rural poverty (World Bank, 1996).

In Zimbabwe, cotton was introduced as a commercial crop in the early 19th
century and Since 1980, its cultivation has expanded rapidly among the smallholders
(Blackie, 1983). Zimbabwe ranks among the top five cotton producing countries in Africa
(J owa, 1996). Cotton is the most important cash crop to Zimbabwe’s smallholdersl and it
ranks second after tobacco in terms of national foreign exchange earnings. More than half

a million Zimbabweans depend on cotton for their livelihood. Cotton is also a major

 

' In Zimbabwe, three primary sub-sectors fall under the smallholder sector; these are Communal Area
(CA), Small Scale Commercial Sector (SSCS) and the Resettlement Areas (RA).

source of cooking oil and animal feed (Mariga, 1994). Nonetheless, cotton production in
Zimbabwe has reached a turning point.

Prior to independence in 1980, smallholders contributed an insignificant amount
of cotton output. But at independence, the GOZ adopted a deliberate policy to stimulate
smallholder agriculture and these efforts paid off. Between 1980 and 1985, Zimbabwe
experienced a major cotton revolution as smallholder cotton area increased from 15,000
ha to 130,000ha. Smallholders’ Share of national output rose from 8% to 40% over the
same period. Favorable agricultural policy that led to a major increase in support services
such as extension, credit and research being directed to serve smallholders helped
stimulate smallholder cotton production (Rukuni, 1993). That cotton is a drought tolerant
crop might have aided its widespread expansion in semi-arid regions of Zimbabwe as
well (Mariga, 1994).

However, the smallholder cotton revolution is under serious threat from risky
pesticide use (Gillham, 1993; .Iowa, 1996). Part of the reason is that the most toxic
chemicals are among the cheapest. Also, pesticide use in Zimbabwe has been Shown to be
price inelastic (Sukume, 1999). Cotton is exposed to the depredations of various insect
Species and to achieve acceptable yields Zimbabwe’s smallholders have placed heavy
reliance on chemical control of pests. Calendar-based Spray strategies traditionally used
for controlling cotton pests have been blamed for the emerging problems of pest
resistance and the outbreak of secondary pests like the red Spider mite (Meerman, 1991).
More importantly, pesticide-related farmer health risks aggravate the problems of

chemical dependent pest management strategies being used in Zimbabwe’s agriculture

(Loewenson and Nhachi, 1996). The challenge is to provide alternative effective pest
control with less use of toxic pesticides.

In addition to farm level constraints, economic reforms present new challenges for
cotton growth in Zimbabwe. Like many other African countries, Zimbabwe adopted
International Monetary Fund (IMF )/World Bank inspired Structural Adjustment
Programs2 (SAPS) in the 19903 following a major slow down in economic growth, lack of
investment and operating losses of marketing boards. The common elements of the
adjustment packages are reduction of trade barriers, elimination of subsidies and price
controls, privatization of state owned firms, and realignment of foreign exchange regimes
(Ajayi, 1994, UNRISD, 1994). A critical question is whether these reforms have
delivered the expected agricultural growth pay-off. Several studies have observed that
reform effects in Africa are mixed, and non-adjusting countries face sluggish markets,
weak institutions and continued economic stagnation and decline (World Bank,1994 and
2001). Nonetheless, initial evidence from Zimbabwe’s experience with SAPS provides
important lessons for other countries going through the same cycle of reforms. There is
relatively limited research on the evaluation of SAPS effect on agricultural supply
response of large versus smallholder farmer groups in Africa.

A significant threat to post-independence agricultural production growth in
Zimbabwe has been the civil war in Mozambique. Zimbabwe’s economy suffered during
the civil war in Mozambique that spilled over into southeast border areas. Transport costs

from land locked Zimbabwe to seaports escalated during this period, while the local

 

2Stabilization (or macroeconomic adjustment) refers to a standard set of policies designed to manage
balance of payment, inflation and the government budget deficit through elimination of subsidies, wage
freeze, and restricted growth in government expenditure. Adjustment (or structural transformation) policies

people were displaced from their farms. Lost economic opportunities from armed
conflicts and civil strife have not been seriously studied yet they fall disproportionately
on the rural poor. Also, the recurrent drought at a rate of one in every five years has
reversed gains in predominantly rain-fed smallholder agriculture. Understanding farm-
level constraints, the impact of the policy setting especially SAPS, and the effects of the
end of the war in neighboring Mozambique in engendering cotton supply response in

Zimbabwe is a key focus area of our study.

1.2 Study Objectives and Thesis Organization

In view .of the importance of agricultural sector for economic growth in SSA,
governments need to determine policies best suited to stimulate agricultural production
(Bond 1983). Crafting such policies requires a good understanding of the major barriers
to cotton production in Africa that include pest infestation, farmer health risks and
misguided policies that marginalize agriculture (Gilham, 1993, Kiss and Meerman, 1991;
World Bank, 1995). These key threats to sustainable cotton production in Africa are
discussed in three interrelated essays whose main objectives are to:

0 Evaluate the pesticide-induced health risks associated with conventional
pesticide application in cotton production in Zimbabwe.

0 Determine the factors affecting IPM use among Zimbabwe’s smallholder
cotton farmers.

0 Estimate supply response models for both large-scale commercial (LSC)

and smallholder cotton growers.

 

focus on removing restriction on and interference with the market. Adjustment policies have stabilization
consequences and vice versa (Davies et al., 1988).

0 Suggest policy guidelines for the future growth in Zimbabwe’s cotton
production.

Data to address the first two objectives of this study was collected from smallholders in
Sanyati and Chipinge, two leading cotton regions in Zimbabwe. Sanyati is situated in the
Middleveld and is a traditional cotton region being among the first in Zimbabwe where
Farmer Field School (FF S) —based cotton Integrated Pest and Production Management
(IPPM) extension efforts were introduced in 1996. Most of the district is located in
natural region III suitable for semi-intensive cultivation and the mean annual rainfall is
650-800 mm. The Cotton Research Institute (CRI) conducts research trials in the rural
areas and is located in Kadoma less than an hour’s drive from Sanyati. Chipinge district
is a relatively new frontier cotton zone. It is situated in Manicaland Province (Lowveld)
and it falls under natural region IV suitable for semi-extensive to extensive farming with
a mean annual rainfall range of 450-650mm (Central Statistical Office, 1989). Time
series data used to address the third objective were obtained from the Ministry of

Agriculture and Reserve Bank of Zimbabwe (RBZ) publications.

1.2.1 Farmer Health Risks in Smallholder Cotton Production in Africa

Farmers in Afiica face a severe health threat from toxic chemicals currently used
in cotton production (Loewenson, R and C.F.B. Nhachi, 1996). Of concern is the fact that
local constituencies for the protection of public health are much less influential than in
developed countries (Kiss and Meerman, 1991). An estimated 11 million cases of
pesticide intoxications are reported to occur annually in Africa although the true extent of

the problem might be underestimated due to poor data compilation and misdiagnosis of

some of the cases (Jeyaratnam, 1990). Consensus is rapidly growing that health issues in
Africa constitute a massive threat to development and have the potential to reverse gains
made in agricultural growth (Binswanger and Townsend, 2000;World Bank, 2001). In the
past, the health of poor farmers has been trivialized and rarely addressed seriously in
many developing countries (Rengam, 1999). Previous effort to put economic gain ahead
of farmer health does not make economic sense in the long run. Nonetheless, the health
status of the cotton growers is a key ingredient to ensure the attainment of sustainable
agriculture. Thus an analysis that identifies the sources of pesticide-related acute health
risks and their extent and severity is beneficial to both farmers and policy makers.

The first essay, in Chapter 2 analyzes the health risks of pesticide use in
Zimbabwe’s smallholder cotton production. The essay uses cost of illness models and
Poisson regression analysis to identify factors causing acute health costs and symptom
incidences among smallholder cotton growers. Zimbabwe’s smallholder cotton growers
are ofien exposed to highly toxic chemicals when they mix, load and Spray pesticides.
Seminal work on farmer health and agricultural production was conducted in the
Philippines and Ecuador (Antle and Pingali, 1994; Pingali et al., 1994, Crissman et al.,
1994). In Africa, the empirical evidence in support of the link between pesticide use and
farmer health is still patchy (Loewenson, R and C.F.B. Nhachi, 1996). The relatively low
level of education and literacy rates in SSA makes the overall risks of pesticides exposure
worrisome (Kiss and Meerman, 1991). The experience of Zimbabwe cotton growers in
transition from calendar-based chemical control of pests to IPM use provides important
lessons for SSA countries undergoing Similar transformation. The study uses primary

data from Chipinge and Sanyati districts in Zimbabwe. The data covered pesticide use

patterns and farmer behavioral practices that contribute to health risks. Since pesticides
pose human health risks, to do nothing and let the problem deteriorate into an epidemic
that might induce a major exodus from cotton production would be naive for policy

makers and catastrophic for the cotton industry in Africa.

1.2.2 The Adoption of Integrated Pest Management (IPM) in Zimbabwe

Although IPM adoption is assumed to result in improved crop productivity, the
circumstances surrounding its adoption in Africa are still unclear and require careful
analysis. Elsewhere, the financial and health benefits of IPM use including its role in
improving the long run sustainability of agricultural systems has been demonstrated
(Norton and Mullen, 1994). Despite the fact that IPM could raise cotton yields
Significantly, very few farmers in Zimbabwe have adopted cotton IPM as most still rely
on calendar-based spraying techniques. Inefficient pesticide use based on prophylactic
Spray strategies without regard to need increases pest resistance and thus limit the
efficacy of most pesticides (Eisa, et al., 1993). Consequently, a wider range of pesticides
in increasing amounts and strengths must be used to achieve the same level of plant
protection. Further, some cotton pesticides destroy natural pest predators creating a
“pesticide treadmill” which results in a steady increase in the cost of pest control and
health hazards (Lipton and de Kadt, 1988). To resolve this crisis, current pest control
strategies that rely on repeated applications of conventional broad-spectrum pesticides
need to be modified. Therefore, Zimbabwe’s cotton growers should move away from
reliance on pesticides and encouraged to use IPM technology. A fundamental decision
confronting cotton growers is the need to adopt effective and affordable pest management

strategies yet the motivational factors are not well understood.

To articulate these issues, Chapter 3 analyzes the factors affecting the adoption of
IPM in smallholder cotton production in Zimbabwe. Specifically, the study examines
IPM-related practices used to control both cotton pests and disease under smallholder
conditions. IPM use is operationally defined as a count of IPM related practices used in
cotton pest management. This study is based on primary survey data collected in fall
1999 in Sanyati and Chipinge district. As in most cotton growing areas, insect pests are a
major problem. The present study on Zimbabwe is among the first in Africa to use
Poisson regression as previous studies have either used bivariate statistical analysis and
ordinary least squares (OLS) thus leading to statistically undesirable results (Ramirezi
and Shultz, 2000). Although IPM is a relatively new phenomenon in much of Africa, it
represents an alternative way to stabilize yields, and reduce both pesticide use and readily
avoidable health impacts. The magnitude of insect3 damage under prophylactic spray
techniques has been estimated to reduce cotton yields by up to 60% (Jowa, 1996). AS a
result, insect pest infestation has induced Zimbabwe’s smallholder cotton yields to
decline from 800kg/ha to 600kg/ha over the past two decades (Jowa, 1996). Knowledge
generated by this study provides a basis for the diffusion of IPM among smallholders in
Zimbabwe. The question about IPM use is not whether, but when and how the technology
could be extended to all cotton growers at risk from exposure to toxic pesticides every
year. A key issue though is that successfiil adoption of IPM requires a policy

environment committed to its implementation (World Bank, 1997).

 

3 The most damaging insect pests implicated in the cotton yield losses in Zimbabwe are the aphids (aphis),
red spider mite (tetranychus), red bollwonn (diparopsis), heliothis bollwonn (helicorvepa), spiny bollwonn
(earias), whitefly flJemisia and trialeurodes) and the stainers (dysdercus), while verticillium wilt and

1.2.3 The Impact of Structural Adjustment Programs (SAPS) in African Agriculture

More than any other factor, SAPS in Africa have generally been undertaken as a
response to fiscal crisis and declining agricultural growth. Since 1980, more than 30
African countries have implemented agricultural policy reforms as part of SAPS (Ajayi,
1994). Many advocates of these reforms have argued that SAPS will result in agriculture-
led economic growth in Africa. The central question is have SAPS enhanced or
undermined cotton growth in Africa? However, reform packages implemented in Africa,
although they vary in content and sequence, provide a degree of considerably uncertainty
and results have been mixed. The majority of the adjustment experiences considered
“successful” is small in number relative to the total group of countries undergoing
reforms (United Nations Research Institute for Social Development, 1994). There is an
emerging consensus that future success requires closer attention to the impact of SAPS on
agricultural production especially given that more than 70% of the 650 million Africans
derive their livelihoods from agriculture (Economic Commission for Africa, 1991; World
Bank, 1994).

In Zimbabwe, it is not well established whether SAPS impedes or enhance cotton
production, the second major export crop. The contribution of macroeconomic factors has
not been clearly studied either. Understanding the effect of SAPS in agricultural supply
response would make an important difference to the design of structural adjustment
packages (Rodrik, 1990). Further, objective analysis of SAP impacts on agricultural
supply response will help us decide whether the pessimism (or optimism in some

instances) associated with early generation of the policy reforms is legitimate (Bond,

 

bacterial blight are among the common diseases (Gillham,1993; Kiss and Meerman, 1991; Zimbabwe
Cotton Handbook, 1998).

1983). So far, there is no firm knowledge about the social impact of SAPS despite the fact
that the issue has attracted a lot of attention since the mid-19803 (Azam, 1994).

In the third essay in Chapter 4, we contribute to this debate and analyze cotton
supply response to SAPS, macroeconomic policies and exogenous shocks such as
drought, and the return to peace in Mozambique. To articulate these issues, the analysis
uses aggregate time series data from Zimbabwe focusing on the post-independence
period, 1980 to 1996. The generally held view is that SAPS result in a positive impact on
cotton production but the empirical evidence to refute or support such claims is scant.
Also uninvestigated is the role of regional peace in engendering agricultural supply
response. In this respect, we hypothesize that both peace in Mozambique and the reforms
enhance cotton acreage expansion among the smallholders and LSC cotton growers. The
essay uses autoregressive models to estimate acreage and yield response to
macroeconomic policy changes. The main advantage of analyzing large and smallholder
farmers separately is to capture equity implications emanating from SAPS and the fact
that the growers respond differently to price changes.

The conclusions from the three essays, including lessons for other African

countries and opportunities for further research are summarized in Chapter 5.

10

References

Antle, J .M., and PL. Pingali. “Pesticides, Productivity, and Farmer Health: A Philippine
Case Study.”American Journal of A gricultural Economics. 762(418-430), 1994.

Azam, J. “The Uncertain Distributional Impact of Structural Adjustment in Sub-Saharan
Africa.”Structural Adjustment and Beyond in Sub-Saharan A flica: Research and
Policy Issues. pplOO-l 13. Rolph Van Der Hoeven and Fred Van Der Kraaij,
James Currey, eds. London, United Kingdom, 1994.

Bingen, J. “Cotton Democracy and Development in Mali.” The Journal of Modern
African Studies. 36: (265-85), 1998.

Binswanger, HP. and RF. Townsend. “The Growth Performance of Agriculture in Sub-
Saharan Africa.” American Journal of Agricultural Economics. 82:(1075-1086),
2000.

Blackie, M.J. Case Study, The Zimbabwe Cotton Marketing Board, Department of Land
Management, Working Paper, 2/ 83, University of Zimbabwe, Harare, Zimbabwe,
1983.

Bond, M.E. “Agricultural Responses to Prices in Sub-Saharan African Countries.” IMF
Staff Papers. pp703-726, Washington DC. 1983.

Central Statistical Office. Statistical Year Book. Government Printers, Harare,
Zimbabwe, 1989.

Crissman, C.C., D.C.Cole, and F. Carpio. “Pesticide Use and Farm Worker Health in
Ecuadorian Potato Production.” American Journal of A gricultural Economics.
762(593-597), 1994.

Economic Commission for Africa, A fi‘ican Alternative Framework to Structural
Adjustment Programs for Socio-Economic Recovery and Transformation: A
Popular Version. UNECA Addis Ababa, 1991.

Eisa, H.M., S. Barghouti, F.Gillham, and MT. Al-Saffy. Cotton Production Prospects for
the Decade to 2005: A Global Overview. World Bank Technical Paper Number
231, Washington DC. 1994.

11

Gilham, F.E.M. Case Study Report on Cotton Research and Development Work in South
Africa, Tanzania, Uganda and Zimbabwe, International Cotton Advisory
Committee, 1993.

Jeyaratnam, J. “Acute Pesticide Poisoning: A Major Global Health Problem.” World
Health Statistics Quarterly. 43:3(139-144), 1990.

J owa, P. “IPM on Cotton in Zimbabwe.” IPM Implementation Workshop
of the East, Central and Southern Africa. Harare, Zimbabwe, 1996.

Kiss, A and F. Meerrnan. Integrated Pest Management and African Agriculture. World
Bank Technical Paper Number 142. African Technical Paper Department Series.
Washington D.C.1993.

Lipton, M., and E. de Kadt. Agriculture-Health Linkages.WHO Offset Publication,
No.104, Geneva, 1988.

Mariga, I. K. “Cotton Research and Development, 1920-1990.” Zimbabwe ’s Agricultural
Revolution. pp 219-233. Mandivamba Rukuni and Carl K. Eicher, eds. University
of Zimbabwe Publications, Harare, 1994. '

Meerman, F. “Pesticide Management in Cotton in Zimbabwe”, Integrated Pest
Management and African Agriculture, World Bank Technical Paper Number 142,
Washington D.C.1991.

Loewenson, R and C.F.B. Nhachi. “Epidemiology of the Health Impact of Pesticide Use
in Zimbabwe.”Pesticides in Zimbabwe: Toxicity and Health Implications. pp25-
35. Charles F.B. Nhachi and Ossy M.J.Kasilo, eds. University of Zimbabwe
Publications, Harare, Zimbabwe, 1996.

Lele, U. Cotton in Africa: An Analysis of Diflerences in Performance, Managing
Agricultural Development in Africa. Madia Discussion Paper No.7, World Bank,
Washington DC. 1989.

Pingali, P.L., C.B.Marquez, and F.G.Palis. “Pesticides and Philippine Rice Farmer

12

Health: A Medical and Economic Analysis.” American Journal of A gricultural
Economics. 76:(587-592), 1994.

Ramirezi, O.A., and SD. Shultz. “Poisson Count Data Models to Explain the Adoption of
Agricultural and Natural Resource Management Technologies by Small Farmers

in Central American Countries.” Journal of A gricultural and Applied Economics.
32:1(21-33), 2000.

Rengam, S.V. “The Struggle Against Pesticides.” Women and IPM: Crop Protection
Practices and Strategies. Royal Tropical Institute/Intermediate Technology
Publications, Amsterdam, The Netherlands, 1999.

Rodrik, D. “How Should Structural Adjustment Programs be Designed?” World
Development. 18:7(933-947), 1990.

Rukuni M and CK Eicher. “Zimbabwe’s Agricultural Revolution: Lessons for Southern
Africa.” Department of Agricultural Economics, Michigan State University, East
Lansing, Stafl Paper No. 93-1, 1993.

Sukume, C. “Economics of Pesticide Use in Zimbabwe Agriculture.” Pesticide Policies
in Zimbabwe: Status and Implications for Change. Publication Series Special
Issue No.1. Godfrey D. Mudimu, Hermann Waibel and Gerd Flesicher, editors.
Pesticide Policy Project, Hannover University, 1999.

United Nations Research Institute for Social Development. Structural Adjustment in a
Changing World. Geneva, Switzerland, 1994.

World Bank. Adjustment in Afiica: Reforms, Results, and the Road Ahead. A World
Bank Policy Research Report, Washington DC, USA, 1994.

World Bank. Cotton Production Prospects for the Next Decade.World Bank Technical
Paper Number 287, Washington DC. 1995.

World Bank. Taking Action For Poverty Reduction in Sub-Saharan Africa: Report of an
Africa Region Task F orce. Report Number 15575-AFR, Washington DC. 1996.

13

World Bank. Integrated Pest Management: Strategies and Policies for Effective
Implementation. Environmentally Sustainable Development Studies and
Monographs Series No.13. Washington D.C.1997.

World Bank. World Development Report: Attacking Poverty. Oxford University Press,
Washington DC. 2001.

Zimbabwe Commercial Cotton Growers Association. Zimbabwe Cotton
Handbook. Harare, 1998.

I4

CHAPTER 2

ECONOMIC AND HEALTH IMPACTS OF PESTICIDE USE IN ZIMBABWE’S
SMALLHOLDER COTTON PRODUCTION

2.0 INTRODUCTION

The health hazards of using pesticides4 are now globally acknowledged (Burrows,
1983; Fernandez-Comejo, 1994; Jowa, 1995; van Emden and Peakall, 1996; World
Wildlife Fund (WWF), 1998). This recognition has led to social decisions to restrict the
use of certain pesticides and forced renewed efforts to promote alternative crop protection
methods (Pincus et al., 1999; Sheets and Pimentel, 1979; WHO, 1990). Recently in
Southern Africa, attention has been focused on the uptake of alternative approaches like
Farmer Field School (FF S)-based Integrated Pest Management (IPM)5 although its
implications for farmer’s health remains to be seen.

There are few studies that focus on health risks in agricultural production and yet in
Africa, the human health problem is growing (World Bank, 2000; Hurley et al., 2000;
Sunding and Zivin, 2000). Farmer health issues can no longer be downplayed while
Simply worrying about them will not solve the problem, much less understand it. Both
economic research and studies conducted in the medical field corroborate that
occupational health problems of the agricultural industry as whole have received scant

attention (Watterson, 1988; Smith et al., 2000). However, there are good reasons to

 

4Pesticides is any substance or mixture of substances intended for preventing, destroying, repelling, or
mitigating any pest and it includes insecticides, herbicides, fungicides, rodenticides etc. (EPA, 1999).

5 IPM emphasizes natural pest control, host plant resistance, and using chemicals as a last resort. Unlike
conventional control, pesticides recommended under IPM approaches target specific pests, are applied at
lower rates and are less toxic to beneficial organisms. Therefore, 1PM is considered a sustainable pest
management practices that confers both health benefits and environmental safety.

expect improved health to result in improved functionality and productivity (Strauss et
al., 1998). Using Philippines farm-level data Pingali and his collaborators conclude that
pesticide use has a negative effect on farmer health, while farmer health has a positive
effect on productivity (Antle and Pingali,l994; Rola and Pingali, 1993). Recent attempts
to measure health costs of pesticide use have been made in Ecuador and the United States
(Antle et al., 1998; Crissman et al., 1994; Sunding and Zivin, 2000). But the evidence
from Africa is thin.

Some key questions remain unanswered: What are the main sources of health risks in
smallholder agriculture in Africa? Do these risks threaten farmer livelihoods or the
sustainability of cotton production in Africa? Are farmers fully aware of the health risks?
What intervention strategies are required to minimize the risks? Many now believe that
reduction of pesticides use should be the primary goal of IPM (Burrows, 1983; Pesticide
Policy Project, 1999; Czapar et al., 1995; Jorge Fernandez-Comejo, 1996; Ajayi, 1999).
Nonetheless, the question whether cotton IPM confers any health benefits to smallholder
pesticide users has not been systematically explored.

Nhachi and Loewenson (1996) looked narrowly at occupational health problems
among commercial farm workers in Zimbabwe. No study has looked Specifically at the
health effects under smallholder farming context. In Africa, less work has been done to
understand the important sources of occupational health risks for farmers. In West Africa,
a survey on pesticide-related occupational health effects found that the social cost of
acute poisoning in cotton is substantial (Ajayi, 1999; Fleischer, et al., 1998). However
comprehensive analysis of the individual-Specific acute effects is lacking in several

countries in Sub-Saharan Africa.

16

The human health threat appears particularly evident in LDCS where environmental
laws tend to be lax and farmers lack sufficient information about the products they apply
(WHO, 1990; Chitemerere, 1996; Mbanga, 1996; Tjomhom et al., 1997; The Pesticide
Trust (TPT), 1993). The risk of exposure iS worsened by the farmer’s inability to afford
protective equipment and also the general lack of health insurance markets in most LDCS
(Antle et al., 1994;World Bank, 2000). Although the problem is acknowledged, the extent
of the health problems among farm workers in Africa remains unclear. Only a few
African countries monitor health effects or keep statistics and information about pesticide
poisonings (World Bank 1996; Rother and London, 1998). As a result, pesticide
poisonings are grossly under-reported and under-notification is still a serious problem on
the continent.

Cotton is vulnerable to a wide range of diverse insect pests6 the control of which
makes it one of the leading users of chemical pesticides. Although in general Africa uses
relatively less pesticides than Asia and Latin America, pesticides are widely used in
African cotton production. Most smallholders rely on pesticides as one of the efficient
and effective ways to manage the diverse number of cotton insect pests. Such chemical
crop protection strategies have been sustained by vertically integrated production systems
and contract farming (Fleischer, 1999; World Bank, 1996).

Prolonged exposure to pesticides has been associated with several chronic and acute
health illnesses like non-Hodgkin’s lymphoma, leukemia, cardio-pulmonary disorders,

neurological and hematological symptoms, and adverse Skin effects. Some older

 

6 The cotton pests of economic significance in Zimbabwe are cotton bollwonn, aphids, jassids, stainers,
whitefly, and diseases such as wilt and bacterial blight. Pest-related yield losses in Africa have been
estimated to range between 40 and 65 percent (Iowa, 1996; Zethner, 1995).

pesticides have been reported as causing parkinsonism, asthma, cancer, tumors,
miscarriages, birth defects, still births, male sterility, genetic mutations and behavioral
changes (Watterson, 1988; Moses, I992; Femandez-Cornejo, 1997; Yudelman, 1998;
Cuyno, 1999). Impaired vigilance, reduced concentration, memory deficit, linguistic
disturbances, allergies, hypertension, depression and repeated irritability are some of the
frequently reported clusters of symptoms linked to pesticides exposure (TFECHLD,
1988). Chronic effects are particularly alarming in the light of new studies linking
immune system suppression to some pesticides (Rengam, 1999; EPA, 2000).

When farmers take the health risks into consideration, the risk-reducing role of
pesticides becomes less important than previously assumed. An increasing body of
evidence suggests that the benefits of pesticides are obtained at a substantial cost to the
society (Antle and Pingali, 1994; Antle et al., 1998; Cole et al., 1998; Pingali et al., 1995;
Crissman and Cole, 1994; Pincus et al., 1999; Nhachi and Loewenson 1996; Watts, 1993;
WWF, 1998; Czapar et al., 1998; WHO, 1990). The principal insight from these studies
is the negative impact of pesticides on the health of farmers and farm workers,
particularly in LDCS.

Global acute poisoning figures are quite alarming as the negative trends are fueled by
the widespread use of more relatively toxic organophosphates and carbarnates (Antle and
Capalbo, 1994;WHO, 1990). Nonetheless, IPM was developed in the 1950s as a response
to the negative side effects of using pesticides, but has not been in use in many Afiican
countries until only recently (Maumbe and Swinton, 2000). IPM use in LDCS can be
enhanced by the development of institutions that Simplify choices by making it more

attractive and accessible to farmers (Zilberrnan and Castilo, 1994; van Emden and

18

Peakall, 1996). Little empirical evidence is actually available about the health benefits of
IPM use (Antle and Capalbo, 1994).

Africa cannot afford to be left behind as the world shifts from broad-spectrum
pesticides to IPM and highly selective, rapidly degradable, and environmentally safe
products (Fomey, 1999; Cartwright, et al., 1993; TPT, 1993; Jowa, 1995). Distorted
policies that subsidize pesticide use worsen health hazards experienced in most African
countries (Zethner, 1995; Fleischer, 1999). At the same time, poor access to health
services and a medical profession who lacks the ability to recognize pesticide-related
morbidity raises further concerns (TPT, 1993). Such policy shortcomings, information
gaps and institutional rigidities create a bias towards pesticide-dependent paths of
technological development (Pincus et al., 1999).

The potential health risks from pesticide exposure vary with the pesticide
products, quantities, and application methods used. Since pesticides are not widely
applied in Africa, their health effects have eluded attention. But pesticides are heavily
applied to cotton, and their effects on humans deserve attention. This study examines the
degree and determinants of acute pesticide health symptoms among Zimbabwe’s
smallholder cotton growers. The results are specific to Zimbabwe, but the analysis

provides useful lessons for cotton growers in other African countries.

2.1 Objectives of the Study

The central focus of this paper is to analyze the pesticide-related health risks
among smallholder cotton farmers in Zimbabwe. The study examines the extent of
pesticide—induced acute health risks among cotton growers in two cotton-growing regions

of Zimbabwe. The specific objectives the study addresses are to:

D Describe pesticide-induced farmer health impairments and pest
management practices,

[3 Estimate a health cost function for smallholder cotton growers,

D Identify the determinants of incidences of acute farmer health risks,

D Determine pesticide-related farmer behavior that mitigates or averts the
observed illnesses and,

Cl Propose policies that help reduce pesticide-related health risks.

We hypothesize that due to the problem of thermal discomfort in tropical agriculture,
ownership of protective clothing does not lower farmer’s health risks as available
protective clothing are not worn. In terms of the impact of pesticide-related behavior, we
hypothesize that contact with extension workers is negatively correlated with smallholder
cotton farmer’s health risks. Finally, we hypothesize that FF S-IPM graduates experience
lower pesticide-induced health risks than non-IPM farmers, assuming that growers who

are aware of IPM apply less pesticides which in turn lowers health risks.

2.2 Methodology and Data
2.2.1 Theoretical Model

In order to understand the benefits and costs pesticides impose on society, we
need to assess social costs and benefits as well as the private ones. The social costs of
pesticides include pesticide residue monitoring, exposure to pesticide drift, and contact
with freshly sprayed foliage by non-applicators, health care facilities and clean up
programs. Due to limited information and methodological flaws, the costs of extemalities

caused by pesticides and the implied mitigation costs have not been adequately accounted

20

for, resulting in a tendency to over-estimate net benefits of pesticides use (Pincus et al.,
1999). Over-estimation of pesticide net benefits has led to benefit assessment of
pesticides based on social welfare theory (Waibel, 1999). This requires placing a value on
the adverse health effects associated with pesticide use (Cropper, 1994). Including health
in farmer’s utility function tends to reduce demand for harmful divisible production
inputs (Swinton, 1998). Consider first a single farm household, with one adult member,
the farmer who aims to maximize profit with respect to pesticide use. For simplicity, we
assume perfect information regarding positive and negative extemalities of pesticide use.
The farmer derives utility from his or her level of health H, and consumption goods G
obtained from marketed cash crop output Y, such that:

(1) U=U(H,Y, G)
We assume that the utility function is concave and increasing in consumption goods G,
marketed output Y and health status H. The level of health of the pesticide applicator is
represented by a health function (Cole et al., 1998, Strauss et al., 1998, Pitt et al.,1986
and Hurley et al., 2000):

(2) H=f((X" . E(X”,A). M p, h, 71)
Where H represents a measure of health status or outcomes that depends on personal
characteristics that impact health such as age, gender, use of alcohol and smoking;
exposure to pesticides X ” and pesticide exposure averting behavior EtX”, A). Exposure
mitigating strategies M P, that includes human capital variables such as IPM knowledge,
are assumed to improve health via judicious use of pesticides. Similarly, a farmer’s

knowledge of basic first aid7 is assumed to mitigate pesticide—induced health risks. We

 

7 First aid is the initial effort designed to save the life of a worker exposed to pesticide poisoning while
medical help is on the way or in worst cases is unavailable.

21

further assume that farmer’s health status is influenced by institutional factors such as
access to local public health infrastructure or health services h, whose price is W, and ,u is
an individual specific health (genetic or hereditary) endowment. Assume that (all/6X ’)<0,
(an/65a", A)) >0, (oil/6M”) >0, and (6H/6h)>0. The farm output production function is
unconventional as it also captures how farmer’s health and risk perception may affect
production; that is,

(3) Y=r(X”, x”, L(H,- L" , 19.2)
where Y is cotton output, X P represents variable pesticide inputs, X ” is non-pesticide
inputs, L refers to household head’s effective field labor, P is a vector of output price and
Z is conditioning variables such as age, gender, educational levels and risk perception.
Farmer’s health status affects the number of healthy man-days and the total time
available for leisure and work. Therefore, effective labor input or management is a
function of available work hours or man-days L" and also worker’s functionality which
depends on individual health status (11,) so that L = f (11,, L") (Antle et al., 1994; Strauss et
al., 1998). Assuming that the smallholder farmer operates in perfect competitive
conditions, the household’s utility is maximized subject to constraints of budget, safe
minimum health and environmental standard. Letting P denote product Y’s price, W, and
W, are prices of household consumption goods G and health services 11. W,, and W, is the
price vector of inputs X P and X ” respectively. In addition, W. and W... refer to the price of
exposure averting and mitigating inputs A and M ” respectively. H,” and EQ," is the
minimum acceptable health level of the farmer and environmental conditions

respectively. The resulting extended profit function is given by:

22

«P, X, G, A, M’) = PY-W,X’-W.X°-W.A- W... M' -W,G-W,. h. The utility maximizing first
order conditions are obtained from the optimization of a constrained fimction represented
below;

(4) Maxrykxngua) W130, H)
s.t. a) pY(L,x”,-XP,- A; M”) —W,,X"-W0X”- WaA-Wm MP- "1,0.th 2 0
fit) HMS=f(X", E(X". A), M”, Wu!)-
aivEcquorx”, X”, X”, p)
In constraint (iii) minimum environmental quality EQM x P”, X ”T and X PS refer to pesticide
disposal, toxicity and storage hazards. We did not collect environmental data of sufficient
variability to warrant the analysis of this constraint as our focus is on farmer health.
When health costs are considered, optimal levels of pesticide use will solve:
(5) P(a” Y/ dX")= Wp- Wk [(0” U/é’Ifl(a”H/dX)//[(dU/a"H)(dH/dh)]

The second term on the RHS of equation (5) is a money-metric of the marginal utility
loss from pesticide-related health damage. Since it is greater than zefo, the result implies
reduced optimal pesticide use (Appendix 1). In practice, the exact determination of the
socially optimal level of pesticide use may not always be feasible due to uncertainty
regarding the magnitude of pesticide effects and lack of reliable data (Oskam, 1994 cited

in Fleischer, 1999).

2.2.2 Data Sources and Sampling Strategies
Farm level data were obtained from a primary survey conducted from June to
December, 1999 in two leading cotton-producing regions of Zimbabwe. Due to

differences in dialects we pre—tested the questionnaire in each region resulting in

23

modifications of wording, and sequence of the questions. In Sanyati district, located in
the Middleveld (altitude 600-1200m), clusters of villages with exposure to Farmer Field
School Integrated Pest and Production Management [FFS-IPPM] training were first
identified. Next, we stratified on the basis of farmer participation or non-participation in
FFS-IPPM. The sampled IPM and non-IPM farmers were randomly drawn from within
the same villages. In the Chipinge district, located in the South-eastem Lowveld of
Zimbabwe (altitude 300-600m) no FFS-IPPM program was available for rain-fed cotton
production. Clusters were determined on the basis of relative distance from markets and
farm Sizes. Chipinge district lies in the Lowveld of Zimbabwe (altitude 300-600m).

Primary data were gathered on field pest management practices and farmer health
status. Health variables included incidences, treatments and degree of severity of
pesticide-related acute illnesses. Pest management data were collected on type of
pesticide used by target insect, number of applications made in each cotton field, as well
as cotton pesticide storage and disposal practices. Usable responses were obtained from a
total of 280 growers. The main incentive for participating in the survey was the certificate
of participation awarded to farmers who completed the interview. All farmers gave
informed consent prior to the interview.

Data used to estimate the empirical cost of illness models are grouped into the
following categories, (1) farmer characteristics, (2) health-related factors and (3)
institutional variables. Farmer characteristics include age, education and gender. Health
variables included the incidence and severity of pesticide Skin, eye, and stomach

poisoning acute symptoms, as well as personal habits like smoking and alcohol intake.

24

Institutional variables include access to a borehole for drinking water and health center

for treatment.

2.2.3 Empirical Model: Estimation of the Health Cost of Pesticide Exposure

The empirical analysis is in three stages [Figure 2.1]. The first stage involves the
estimation of a cost of illness model for smallholder cotton growers in the two districts.
In the second stage, we estimate Poisson illness incidence models in order to help us
understand the determinants of the Specific acute illness episodes experienced by the
pesticide applicators. The final stage uses probit analysis of the Significant pest
management and farmer behavioral practices driving the observed incidences.

Empirical variables that influence health outcomes come in two different
specifications, continuous and dichotomous variables. For continuous variables like
number of specific acute symptom type ACUTESYM, severity of acute symptom episodes
ACUTINDX, alcohol use and smoking duration, that also take zero variables we added 0.]
factor in all cases to account for the problem of taking logarithm of zeros and to
distinguish responses with the value zero from positive non-zero responses. We specify
the health cost function simply as HC=f(SYM, DSS, C, W,.) where previously undefined
SYM is the sum of pesticide-induced acute symptom types experienced by the farmer
which takes the range 0 to 3, DSS sums the degree of severity8 of acute symptoms and C
is a vector of farmer characteristics. The general form of the health cost model estimated
iS given by:

Health cost = f (A cute symptoms, farmer characteristics, institutional factors) + e

 

8 Acute symptom severity is defined on a monotonic scale of O to 3 with O=absence of symptom, l=mild,
2=severe and 3= very severe.

25

An individual farmer’s health cost is calculated as a sum of both cash and non-cash costs.
The health cost variable is a composite of (1) farmer treatment costs per visit either to the
clinic or local private physician, (2) annual levy of Z$100 contributed to the local rural
health facility and (3) opportunity cost of lost time due to days farmer was ill and
recuperating. We used the 1998/99 agricultural minimum wage of 281,400 per month or
2838.00 per day for costing lost days due to illness. Missing in the composite value of
health cost variable are individual patient’s travel cost to the nearest health facility,
waiting time costs prior to treatment, cost of leisure forgone due to illness and the cost of
home-based health care. We did not collect data on the cost of traditional healing
strategies because farmers are generally reluctant to volunteer such information. We
assume health costs of pesticide exposure to hired labor are borne by the hired workers
themselves (Antle and Capalbo, 1994). The rationale for using the logarithmic functional
form is that it is parsimonious in parameters, can be interpreted as first order
approximation to the true cost fimction, and is globally well behaved (Antle and Pingali,
1994)

(5) Ln(HC) =6, + 6, Ln (AGE) + 621m (EDUCATION) + 6; Ln (ACUTE SYIIIPTOMS) +
64 Ln(ACUTE SYMPTOM SE VERI TY ) + 65 Ln(ALCOHOL ) + 5,Ln(SMOKE) +
67 Ln(HEAL TH CENTER DISTANCE) + 6, (GENDER) + 59 (FIRST AID ) +
5n (BOREHOLE) + e

The dependent variable is the logarithm of an individual farmer’s aggregate health costs

(HC) measured in Zimbabwe dollars. The exogenous variables are defined in Table 2.1.
We decided to assess the relative health impact of pesticides on the basis of

toxicity ranking as defined by the Plant Protection Research Institute in collaboration

with the Zimbabwe Hazardous Substance and Articles Control Board. Four pesticide

26

hazard classes are distinguished by their color codes: green, amber, red, and purple, in
rising order of toxicity. Survey farmers did not use any green label pesticides, so our
analysis uses only three pesticide classes. Color codes are assigned based on three
criteria, (1) acute oral lethal dose (LD50) 9 of the pesticide, (2) the concentration of the
formulation and (3) the persistence of the pesticide in the ecosystem (Nhachi, 1999). We
focused on acute effects since these are health problems that occur very close to the time
when one is exposed to the pesticides (Moses, 1992). We measure pesticide exposure as a
product of the active ingredients per application and the number of chemical applications
made (Honsby et al., 1996; EPA, 1999).

Acute symptom incidences are estimated as a Poisson regression model Specified
as E(Y,-) = (fl1;)+v (i =I,...n) where E(Y,) is the expected value of the dependent
variable of the it” observation, v is stochastic error term, fl is a 1* k vector of parameters,
X, is a k*l vector with the values of the k independent variables in the i"I observation,
and n is the sample Size. The elasticity estimate at X,, a continuous independent variable
is given by (0170,) /d¥,,)(X,-,-/E(Y,-)) =,B ,X 1,. Now, assuming thath is dichotomous, the
percentage change on E(Y) when X j changes from O to 1 is given by 100(exp (’3 “YA-I) a
different relative impact on the dependent variable (Ramirez and Shultz, 2000). The
estimated empirical Poisson regression model for the specific and multiple symptom
incidences is stated below:

(6) SINCID= M C, M, X ”, A, M ”, l) + ,u where SINCID is a positive integer count of

the number of self-reported pesticide caused acute symptom episodes or incidences

 

9 Oral LDso refers to the orally ingested dose (mg of toxicant/kg of body weight) of a pesticide which kills
50 percent of the test population animals. Similarly, the dermal LDso is the dose of a pesticide applied to
the skin which kills 50 percent of the test population animals (The Pesticide Manual, 1997).

27

experienced by the farmer. C is a vector of farmer characteristics as already mentioned
above, M refers to farm management variables, exposure to pesticides X" dosage used on
the farm, and A is a vector of pesticide averting behavior among farmers such as use of
protective clothing. M P refers to pesticide mitigating strategies, and I is institutional
variables affecting farmer health such as relative location to a health care facility.

A full description of the variables used to estimate the model is presented in Table 2.1.

2.3 RESULTS AND DISCUSSION

Results Show that pesticide use in smallholder cotton production is associated
with adverse health consequences for the poor farmers. The illness Spectrum associated
with pesticides exposure varies from acute, systemic to chronic effects (Appendix table
A1.1 ). In 1998/99 season, the estimated average cost of pesticide-related health risks is
23180 and Z$316 for Sanyati and Chipinge districts respectively. The costs represent
45% and 83% of the household seasonal total chemical outlay in Sanyati and Chipinge
respectively. The health costs are attributed to the pesticide applicators and we infer that
costs are much higher per household when we take into consideration other members of
the household who are potentially exposed. Factoring in these unrecognized costs reduces
the net benefits of pesticides among growers.

Farmer illness imposes a Significant constraint to achieving higher net returns in
agriculture (Ruttan, 2001). During the season, farmers lose an average of about 4 and 10
days recuperating from pesticide-induced illnesses in Sanyati and Chipinge respectively.
Although the average distance to the nearest health facility is 5km in Sanyati and 9km in
Chipinge district, the proportion of farmers who visited the clinic to seek medical

attention after acute pesticide poisoning or irritation was very low, about 3% in Sanyati

28

and 7% in Chipinge. Use of home-based mitigating strategies and religious beliefs that
encourage the use of prayer to end health ailments partly explain why farmers do not
often seek medical assistance from health facilities in the study zones. The study focused
mainly on acute effects of pesticides such as Skin and eye irritations, and stomach
poisoning. However, between 40% and 50% of households in Sanyati and Chipinge
respectively have at least one family member suffering from chronic illness. Our
assessment was limited to cancer, blindness, back problems, and asthma or lung
problems.

The predominant route of exposure of pesticide use in the survey areas was
through the skin and eyes. More than half of the applicators in Sanyati (67%) reported
skin irritations compared to 55% from Chipinge. The proportion of farmers reporting eye
irritations was also higher in Sanyati (38%) than Chipinge (26%). Only 7% and 12% of
the farmers in Sanyati and Chipinge reported having experienced stomach poisoning after
Sprayinglo respectively. We did not ask the farmers to indicate the Specific chemicals
responsible for the reported acute symptoms. However, in both districts, male farmers are
the major risk group as they are responsible for most of the Spraying. The descriptive

statistics for the main variables used in our analysis are summarized in Table 2.1.

2.3.1 Farmer Health Cost of Pesticide Use: Cost of Illness Model Results

The number and severity of acute pesticide symptoms are the main determinants of the
cost of farm household illnesses. The elasticity of health costs with respect to positive

symptoms is 0.16 and 0.29 in Sanyati and Chipinge respectively. The results clearly

 

'0 The dominant chemicals used by farmers are monocroptophos, larvin, carbaryl, dimethoate and
fenvalerate. Among these, the most commonly implicated chemicals in stomach poisonings and skin
irritations are the organophosphates (e.g. monocrotophos and dimethoate) and carbamates (e.g. carbaryl

29

suggest that Chipinge cotton growers experience greater health costs than their Sanyati
counterparts. In addition, the elasticity of health cost with respect to symptom severity is
0.11 in Chipinge and 0.09 in Sanyati, reinforcing the fact that Chipinge farmers could be
facing higher risk of pesticide exposure costs compared to Sanyati. Health conditioning
variables such as alcohol intake and smoking habits were also assessed to see their effect
on health costs in both regions. In Sanyati, the coefficient for smoking is statistically
significant at 10% and its positive Sign indicates that farmers who smoke incur relatively
higher pesticide-related health costs. None of the health conditioning variables are

Significant in the Chipinge cost-of-illness model.

2.3.2 Evidence of Health Impairments Among the Cotton Growers

Given the critical contribution of pesticide—related acute symptoms to health costs,
the second stage analysis investigated determinants of these symptoms. Poisson
regression models were used to relate the incidences of self-reported acute symptoms to
farmer characteristics, farm management variables, health, pesticide exposure, exposure

averting and mitigating, institutional and perception variables.

2.3.2.1 Incidence of Skin Effects

In Sanyati, skin irritation incidences are positively associated with the use of a
knapsack sprayer, and calendar-based prophylactic spraying (Table 2.3). The use of a
knapsack Sprayer is positive and significant at all conventional levels while the

prophylactic spray approach iS also positive but significant at 5% level. This means that

 

and 1arvin)(TFECHLD, 1988; WHO, 1990; Cole et al., 1998).

30

greater skin incidences are associated with calendar-based Spraying. Also, the positive
significant coefficient for the knapsack Sprayer implies that farmers using a knapsack
Sprayer obtain higher incidences of skin irritations. In Chipinge, the coefficient for
knapsack is Significant but negative, perhaps because most Chipinge farmers use ultra-
low volume (ULV) Sprayers.

The coefficient on the use of leaking sprayers is statistically Significant at 1%
level and 10% level in Chipinge and Sanyati district. Results suggest that use of defective
Sprayers explain the occurrence of positive skin incidences in the cotton regions. A
significant proportion of farmers in Sanyati (39%) and Chipinge (34%) reported that
sprayers leak on their back when they apply chemicals. The high risk of pesticide
exposure attributed to leaking sprayers implies the prevalence of poor maintanance
procedures or lack of local expertise for sprayer maintanance. In addition, hazardous
storage of pesticides in Chipinge is positive and Significant at 1% level implying the
critical role of unsafe storage practices in determining the presence of positive skin
incidences. Farmers who keep pesticides in the same huts used for cooking and or
Sleeping are more predisposed to have Skin irritations than ones who keep pesticides
locked in a separate store-room elsewhere.

In Chipinge, the incidence of skin problems is positively associated with the use
of both “purple” and “red” pesticides. The “red” and “purple” pesticides comprise mostly
the organophosphates and organochlorines that are known to cause skin irritation
problems when spray men do not use protective clothing. None of the pesticide dosage

variables were significant determinants of skin incidence in Sanyati.

31

The coefficient for attendance at extension meetings was positively Significant at
1% level in both Sanyati and Chipinge. This implies that farmers attending extension
meetings reported more incidences of pesticide-related Skin irritations. This raises the
question of the health content of the extension messages directed at the farmers. It is
possible that farmers’ participation in extension meetings makes them aware that
pesticides can cause skin irritation. But it also appears that extension contact does not
Significantly improve their knowledge of how to prevent Skin irritations.

Among the farmer characteristics, formal education and farmer’s age have a
positive and significant effect on self-reported Skin incidences in Chipinge while in
Sanyati, age was highly significant with a negative coefficient. The evidence from
Chipinge suggests that older and more educated farmers tended to report more positive
incidences of skin incidences. However, in Sanyati older farmers are less likely to report
skin irritation problems. This result probably reflects the fact that older farmers are more
experienced with chemical handling and application than their younger counterparts.

Alcohol consmnption and cotton area cultivated are both positively related with
skin irritations in Sanyati at the 10% level of Significance. Alcohol affects judgment
about use of safety precautions when farmer is exposed to pesticides.

The coefficient on farmer’s use of protective clothing was statistically Significant
at 1% with a negative Sign in both districts. The negative sign suggests that there is a
Significant reduction in dermal contamination when farmers spray with the necessary
protective clothing. This was however contrary to our hypothesis that because of the

weather-related discomforts, smallholders do not effectively use their protective clothing

32

resulting in greater health risks .The evidence clearly shows that those who wear
protective clothing are able to significantly reduce their pesticide exposure.

The perception variable on reforming calendar-based Spray was negative and
Significant at 5% and 1% in Chipinge and Sanyati respectively; that result is consistent
with our expectations. Although label illiteracy coefficient for Sanyati was insignficant,
in Chipinge, it was negative and highly significant at 1%, contrary to our expectations.
We expect poor literacy to be associated with inability to follow instructions for safe use
of pesticides leading to greater exposure risks in terms of higher positive symptom
incidences. A possible explanation for the reason why illiterate Chipinge farmers reported
less skin incidences could be attributed to the problem of lack of hazard awareness. It is
possible some illiterate farmers may notice Skin sensations but fail to make the link that
their symptoms are a function of their exposure to pesticides thus resulting in more
illiterate farmers reporting less pesticide caused Skin incidences. Similarly, negative
correlation between skin effects and formal employment in Chipinge can be explained by
the fact that farmers with full-time jobs are farming part-time and may not Spray
chemicals as frequently as their full-time counterparts. Also, the part-time farmers may
suffer from the same hazard awareness problem explained above, leading to less reported

cases of skin irritation incidences. The coefficient for formal employment in Sanyati was

insignificant.

2.3.2.2 Incidence of Eye Effects

Results from Sanyati show that incidence of eye symptoms increases with the use
of a knapsack sprayer, calendar-based spraying and alcohol intake (Table 2.4). In

contrast, eye incidences in Chipinge were positively associated with farmer’s age, cotton

33

area cultivated, formal employment, smoking and amber pesticides. In Sanyati, farmer’s
age is statistically significant at 5% level with a negative Sign. The result suggests that
older farmers are less likely to experience eye irritation incidences than their younger
counterparts; a fact we attribute to greater experience with pesticide use or a higher risk
averse behavior. In Chipinge, the habit of smoking is positively associated with eye
incidences at 10% level while alcohol consumption had a negative but highly significant
effect on reported eye incidences. The effect of alcohol is contrary to theoretical
expectations. The result may indicate that alcohol consumption interferes with farmer’s
capacity to recall or recognize health hazard of pesticides thus allowing drinkers to report
less eye incidences, when the opposite was expected. In Sanyati, alcohol consumption
had a Significant positive influence on reported eye symptoms while the effect of
smoking was insignificant.

Exposure to cotton IPM training among Sanyati growers has an unexpected
positive significant effect on eye symptoms. This could be attributed to greater awareness
of pesticide-caused eye symptoms among IPM graduates. Farmers’ use of protective
clothing and first aid knowledge are negatively associated with reported eye incidences in
Sanyati and Chipinge respectively. Use of protective clothing in Sanyati is significant at
the 1% level suggesting that these farmers effectively use eyeglasses, and or facemasks
when spraying pesticides. Protective clothing use in Chipinge has the expected negative
Sign but is insignificant.

The coefficient on the need to review weekly spray approach is negative and
Si gnificant at all conventional levels in Sanyati. We interpret this result to mean that

cotton growers who hold the perception that there is a need to reform the calendar-based

34

preventative spraying technique are less likely to report eye incidences because they are
more likely to rely less on chemical pest management strategies. The coefficient for the
perception variable was insignificant in Chipinge district. The foregoing section analyzed
the relative impact of different factors on specific pesticide-induced illness. However,
since farmers may be exposed to multiple illnesses at the same time, we address this issue

in the next stage of our analysis.

2.3.3 Multiple Health Impairments: Total Symptom Incidence Model Results

The Poisson total symptom incidence model results presented in Table 2.5
Shows that pesticide-related acute symptoms in Sanyati are significantly associated with
dosage of the most toxic pesticides, male farmers, larger farm sizes, and extension
meetings attended. The practice of taking meals in cotton fields where there are no
washing facilities is positively and significantly correlated with positive acute symptoms
at 5% level. In addition, after combining the acute symptoms, label illiteracy contributes
significantly to positive acute symptom incidences among the cotton growers in Sanyati.
This result suggests that a higher risk of pesticide exposure confronts cotton growers who
are either partially or completely label illiterate. Likewise, recent evidence from
Cambodia indicates negative correlation between adult education levels and probability
of illness (Deolalikar and Laxminarayan, 2000). The aggregate incidence model also
Shows that farmer’s age, protective clothing, radio ownership, and first aid knowledge are
all Significantly associated with reduced occurrence of acute pesticide symptoms in
Sanyati. Clearly, first aid knowledge and protective clothing have theoretically consistent
results as they are predicted to play a mitigating and averting role in pesticide-induced

ailments respectively. Further, farmers who hold the view that calendar-based Spraying

35

ought to be reformed are less likely to experience acute symptoms in Sanyati and that is
consistent with our expectations. Exposure to IPM training had the hypothesized negative
Sign, but was not significant. The evidence is therefore insufficient to prove that IPM
training can Significantly lower the incidence of pesticide-induced multiple health
impairments in Sanyati. Empirical evidence on the health benefits of IPM on pesticide
use is still mixed. Studies on IPM in Vietnam and West Africa showed that farmers
practicing IPM have substantially lowered occupational health risks (Kenmore, 1997).
Some evidence of reduced average human toxicity with insect IPM adoption has been
cited in the literature (F emandez-Comejo, 1997). Nonetheless, evidence emerging from
Zimbabwe reveals that pesticide-related health risks do not Significantly determine farmer
adoption of IPM practices (Maumbe and Swinton, 2000).

In Chipinge, the total incidence model results show that the coefficients for
leaking sprayer, formal education and extension meetings are positive and statistically
significant at 1% level. The results suggest that Sprayer malfunction contributes to greater
incidences of acute dermal symptoms among cotton growers in the Lowveld area. The
positive effect of extension and formal education on symptom incidences iS unexpected.
Such a perplexing outcome could be attributed to the ‘chemical solutions’ that are
generally proposed by traditional extension workers. The results suggest that the specific

nature of health education needed to limit pesticide exposure is potentially missing in
extension messages. That traditional extension services lack a health focus and need
revitalization has been mooted in the literature (Sasakawa-Global 2000, 1999; Fleischer,
1 999)- It is also possible that farmers who attend extension meetings are more aware of

t 11 a health risks of pesticides and were more likely to report such incidences during the

36

survey. Male pesticide applicators and cotton area treated with pesticides positively
influence the number of reported acute symptom incidences, all are significant at 5%
level. The total incidences of acute symptoms are positively related to “red” pesticides.
The need to reform the calendar based-preventative spray technique, credit use, formal
employment, knapsack sprayers, and first aid knowledge all contribute to less acute
symptom incidences in Chipinge. In addition, the incidence of acute symptoms is
significant and negatively related to “amber” pesticide classes in Chipinge suggesting the
possibility that this class comprises mostly safer chemicals. We did not expect a negative
effect as any exposure to pesticides is likely to cause acute poisoning among unprotected
cotton growers. Since pesticide use is a function of farmer behavior and choice, in the
next section we provide an analysis of the factors driving some of the pesticide

management behavior that is likely to endanger farmer’s health.

2.3.4 Poisson Protective Clothing Model Regression Results

In order to understand why farmers engaged in practices that mitigated and
averted pesticide symptoms, the third stage of the analysis looked at determinants of
these behavioral practices. Pesticide averting behavior involved wearing protective
clothing while spraying. The number of protective clothing items owned by the farmer
consistently reduced pesticide-related health symptoms in both Sanyati and Chipinge. To
understand the individual decision maker’s choice of using protective clothing, we
conduct a Poisson regression analysis of the count of individual protective clothing items
adopted by the farmers in the two districts.

Evidence from Sanyati reveals that farmers who attend a greater number of

extension meetings are likely to use protective clothing when making pesticide treatments

37

( Table 2.6). Attendance at extension meetings was significant at the 10% level.
Ownership of knapsack sprayers is positively associated with the use of protective
clothing at the 5% level of significance. The coefficient for IPM awareness was positive
and Significant at 10% implying that farmers who were exposed to F FS-IPM training are
likely to have more protective clothing equipment. The predicted value for skin illness
episodes was negative and significantly associated with the use of protective clothing,
and this is theoretically consistent as the likelihood of skin incidence is bound to decrease
with increased use of protective clothing. The highly Significant but negative effect of age
was unexpected. The coefficient for label illiteracy had the correct Sign but was not
significantly different from zero.

In Chipinge, protective clothing ownership was positive and significantly
associated with attendance at extension meetings, educational attainment, and use of the
relatively toxic red pesticide category. These results suggests that better informed
individuals are more likely to wear protective clothing when handling and applying
pesticides and the use of “red class” pesticides appears to instill the same discipline. The
predicted skin illness incidences are significant and negatively associated with the use of
protective clothing at all conventional levels. This suggests that Chipinge farmers
generally perceive that pesticide-induced skin-related health risks decrease with greater
use of protective clothing equipment. Alcohol intake and the ownership of knapsack
Sprayers are negatively associated with use of protective clothing at 10% and 5% level of
Significance respectively. Meanwhile formal employment and label illiteracy are both
highly Significant and negatively associated with protective clothing use in the Lowveld.

This suggests that farmers who engage in off-farm employment and thus grow cotton on

38

a part-time basis are less likely to invest in protective clothing equipment. The evidence
also Shows that those farmers who exhibit higher levels of pesticide label illiteracy are
more likely to make pesticide treatments with inadequate protective clothing equipment.

Apart from risk averting behavior, pesticide exposure is also affected by explicitly
risky behavior such as taking meals in cotton fields and using leaky sprayers. Risky
behavior such as taking meals in cotton fields is associated with smoking and label
illiteracy in Sanyati and Chipinge respectively (Table 2.7). Formal education, extension
contact, qualified master farmers, and protective clothing use were negatively associated
with eating in cotton fields. Use of leaky sprayers is explained by among others larger
farm sizes and prophylactic spray strategies, while exposure to IPM training appears to
reduce the use of leaky sprayers in Sanyati (Table 2.8).

Finally, pesticide symptoms are also affected by the ability to treat symptoms.
Knowledge of first aid was enhanced by farmer’s education and cotton area cultivated
(Chipinge) and formal employment (Sanyati) (Table 2.9). AS expected, knowledge of
first aid was negatively associated with exposure variables such as taking meals in cotton

fields and label illiteracy.

2.4 Conclusions

Balancing the numerous benefits that may accrue from pesticide use in cotton
production, farmers face adverse health risks. Looking at the cost of illness model results
for both regions, the conclusion one comes to is that pesticide-induced acute symptoms
Significantly influence smallholder farmer health risk costs. Cotton growers lose a mean
of 28180 in Sanyati and Z$316 per year in Chipinge on pesticide-related direct and

indirect acute health effects. Pesticide-related health costs average 45% and 83% of the

39

household annual pesticide budget in Sanayati and Chipinge respectively. The average
number of days spent recuperating from illnesses attributed to pesticides varies from 4
days in Sanyati to 10 days in Chipinge in 1998/99 growing season.

Exposure variables that are significant in predicting acute symptom incidences
among cotton growers in the study zones are the use of toxic pesticides, leaking Sprayers,
unsafe pesticide storage, pesticide label illiteracy and taking meals in cotton fields. Also,
personal habits like smoking and alcohol consumption confound farmer health risks in
the cotton growing regions of Zimbabwe. In contrast, practices that effectively reduces
health risks in the study zones are the use of protective clothing, knowledge of first aid
and a disposition towards reforming calendar-based Spraying strategies.

The need for farmer education in exposure averting strategies is evident
particularly in the new cotton region of Chipinge. Since farmers in Chipinge face
relatively greater exposure to pesticide risks, targeting farmers in the new regions will
contribute most to minimizing the health risks. Evidence from the traditional cotton
producing zone of Sanyati suggests that farmer’s participation in FPS-based IPM
training does not significantly reduce the incidence of acute symptoms, contrary to our
expectation. However, the IPM awareness variable positively influences the farmer’s
decision to adopt protective clothing for making pesticide treatments. The mixed results
on IPM impact could be attributed to the fact that IPM use is still in its infancy in Sanyati
district where it was introduced in 1997, two years prior to our survey.

Although the pesticide label provides information on pesticide hazard categories, it
iS ineffective for the many farmers who are illiterate. Nor has the use of color codes

helped much as revealed by the inability of 28% in Sanyati and 58% in Chipinge to

40

associate colored triangles to pesticide toxicity. Ignorance about pesticide toxicity
prevalent among survey farmers ought to be seriously addressed by policy makers.
Perhaps the use of local languages on labels for pesticides targeted to small farmers and
educational campaigns about the dangers of pesticides could alleviate the situation.

A very small proportion of cotton growers in both regions reported that pesticide-
related health problems resulted in a visit to seek medical attention to a local health
facility. The evidence seems to suggest that some smallholders treat acute pesticide
effects as minor Side effects that do not warrant medical attention (Flesicher, 1999). The
minimal access to formal health care services further suggests reliance on informal health
care system and or adherence to religious values that discourage seeking treatment for
any ailments. In addition, our study seems to corroborate the fact that due to poor
reporting systems, formal health sector statistics under-reports pesticide-induced acute
symptoms, as most cases do not seek medical care (Chitemerere, 1996; Rother and
London, 1998; WHO, 1990).

An important result of our empirical analysis is that access to extension is
positively significantly correlated with incidences of acute symptoms. This raises
questions about the health content of the extension message. Alternatively, and perhaps
more fundamental, is the question of who attends the meetings and whether safety
information from extension agents filters directly to the pesticide applicators.
Nonetheless, the evidence implies the need to effectively utilize traditional extension
services for the delivery of pesticide-related farmer health and safety information. This
is important given that lack of contact with the formal health system may perpetuate

farmer’s ignorance about pesticide-related health risks. Therefore, a firm commitment to

41

utilize existing extension networks and rural pesticide vendors for the delivery of health
and safety information could heighten farmer awareness and decrease pesticide-related

health risks.

2.4.1 Policy Options and Knowledge Gaps

In Zimbabwe, much effort is currently devoted to promoting new strategies like
F F S-based IPM techniques. While IPM allows for judicious use, what is lacking is an in-
depth study to provide essential information on rational use of pesticides. A clear policy
implication of these findings is that farmers would be healthier if less toxic pesticides are
used in cotton production because they cause Significant health problems for the farmers.
However, a policy to phase out or reduce the use of the risky “purple” and “red”
pesticides without identifying safer substitutes could be short sighted for Zimbabwe. It is
also possible that toxic pesticides may not be the only problem but also the way they are
handled or both.

The health risks of pesticide use can be reduced by effective regulatory systems
that could aid in increasing on-farm human safety. Some important areas for intervention
include expanding farmer first aid education, eliminating the risk of taking meals in
cotton fields, improvements in Sprayer maintanance, and promoting the safe use of
protective clothing. Cotton growers face serious knowledge constraints and on-farm
safety policy should enhance farmer awareness of the link between pesticide hazards,
human health and on-farm pesticide management practices.

Educational policy that targets labeling policy reform is needed in order to
improve farmer’s hazard awareness so they can in turn make more informed choices

about cotton pesticides and required protective clothing. Without label reforms, pesticide

42

choices will continue to be distorted and thus endanger farmers’ health. Widespread label
illiteracy particularly in Chipinge is associated with difficulty in interpreting pictorial
safety labels. Given the existing pesticide dependency in smallholder cotton production,
policy failure to improve farmer health and safety will result in ruined agriculture and
human livelihoods.

Future research work Should attempt to measure health costs to all the individuals
exposed to pesticides including children and hired workers. Self-reported health
conditions due to pesticide exposure leads to problems of bias and endogeneity. Future
attempts to increase the accuracy of measurement of pesticide-related health symptoms
Should consider an independent assessment of farmer’s health condition by a health
expert. Also, the true health risk budget can be obtained by incorporating the costs of
pesticide-induced chronic illnesses including deaths. Longitudinal farmer health study
designs could provide more insights about causality of chronic health effects. More
importantly, growing evidence about the linkage between smallholder cotton production
and farmer health risks implies a greater need to transform the current delivery of rural
health services and provide extension a more prominent role in the diffusion of health
information. Finally, a more permanent solution requires pest management methods that

are effective, economic, long lasting and not damaging to farmer health.

43

Table 2.1: Descriptive statistics for Sanyati and Chipinge districts, 1998/99

 

 

 

 

Variable Sanyati Chipinge—---
Mean Standard Dev. Mean Standard Dev.

Farmer characteristics

Age (years) 46.40 14.20 42.70 12.58
Education (years) 6.54 3.72 6.54 3.75
Male farmers (0,1) 0.83 0.38 0.91 0.29
Health-related and pesticide exposure variables

Acute symptom cases 1.12 0.84 0.95 0.88
Acute symptom severity 0.60 1.00 1.01 1.42
Health cost (ZS) 180.15 157.16 315.63 504.59
Purple pesticides“ (mg/kg/farm) 415.69 2,982.97 1,218.90 6,568.76
Red pesticides (mg/kg/farm) 2,428.98 5,758.26 4,600.08 9,499.08
Amber pesticides (mg/kg/farm) 3,423.04 14,334.94 5,496.06 16,536.52
Eat in cotton fields (1,0) 010 0.30 0.28 ' 0.45
Label illiteracy (1,0) 0.32 0.47 0.54 0.50
Sprayer leak (1,0) 0.39 0.49 0.34 0.48
Storage hazard (1,0) 0.36 0.48 0.21 0.41
Smoking duration (years) 2.14 5.1 1 2.78 7.03
Alcohol intake duration (years) 3.65 6.63 9.66 13.70
Farm management variables

Cotton area (ha) 4.57 3.98 8.74 l 1.56
Cotton bales (bales) 8.12 7.63 19.30 16.82
Extension meetings 4.67 6.37 13.04 1 1.24
Knapsack (1,0) 0.69 0.47 0.42 0.49
Ultra-Low Volume (1,0) 0.05 0.22 0.26 0.44
Formal employment (1,0) 0.46 0.50 0.43 0.50
Prophylactic spray (1,0) 0.30 0.46 0.26 0.44
Pesticide exposure averting and mitigating variables

IPM Train (0.1) 0.48 0.50 - -
Number of protective clothing 3.76 1.54 1.76 1.77
First aid knowledge (0.1) 0.61 0.49 0.19 0.40
Institutional variables

Access to borehole (1,0) 0.37 0.48 0.67 0.47
Distance to health center (km) 4.93 2.82 9.30 5.63
Radio ownership (1,0) 0.68 0.47 0.73 0.45

 

Source: Field survey data, 1999

 

" Pesticide dosage/concentration is expressed as active ingredients that are measured in mg/kg. Farmer’s
exposure is measured as product of pesticide concentration and rate of pesticide application per farm.

44

Table 2.2: Cost of Illness Model Results for Sanyati and Chipinge Districts, 1998/99.

Dependent variable: Natural logarithm of farmer health costs (Z S)

 

 

 

 

Independent variables Sanyati District Chipinge District
coefficient t-statistic coefficient t-Statistic

Farmer characteristics

FARMER’S AGE 0.0060 0.04 0.1800 0.83

MALE FARMER -0.1700 -l.51 0.0049 0.02

FORMAL EDUCATION 0.0330 1.34 0.0310 0.66

Health-related variables

ACUTE SYMPTOMS‘ ***0.1600 4.54 "*02900 4.35

SYMPTOM SEVERITY2 ***0.0890 2.63 "0.1200 2.01

ALCOHOL CONSUMPTION 0.0067 0.30 -0.0210 -0.73

SMOKING *0.0470 1.92 -0.0032 -0.09

Institutional variables

BOREHOLE ACCESS -0.0074 -0.09 -0.0820 -0.63

HEALTH CENTER DISTANCE -0.0088 -0.63 0.0340 0.42

FIRST AID KNOWLEDGE -0.0830 -1.04 -0.0730 -0.46

Adjusted R1 31 35

N 137 131

p-value 0.000 0.000

 

"*=significance at 1% level, "=significance at 5% level, * =significance at 10% level

Note:
1. Three types of pesticide-induced acute symptoms were assessed in detail, eye irritations, skin
irritations and stomach(gastro-intestinal effects) irritations.
2. Symptom severity was assessed on a scale of l to 3 with l= mild, 2=severe and 3=very severe.
The severity variable is a product of positive acute symptom and its severity aggregated across all
the three acute symptoms under investigation. Its value ranges from 0 to 9.

Source: Field survey data, 1999

45

Table 2. 3: Poisson Model Results for Self-reported Acute Skin Symptom Incidences.

 

 

 

 

Sanyati District Chipinge District

Skin Symptom Incidences Skin Symptom Incidences
Independent variables coefficient z-value Coefficient z-value
Farmer characteristics
FARMER’S AGE ***-0.0350 -4.95 ”0.0190 1.95
FORMAL EDUCATION -0.0100 -0.41 ***0.0680 2.50
MALE FARMER 0.3300 1.18 —0.2200 —0.80
Farm management variables
COTTON AREA *0.0330 1.65 ***-0.0700 -2.40
EXTENSION MEETINGS "*00670 4.15 "*00480 6.87
FORMAL EMPLOYMENT 0.1700 0.86 * "-0.9100 -5.31
KNAPSACK * "0.6700 3.17 * "-0.7700 -3.69
PROPHYLACTIC SPRAY “0.4300 2.25 0.0770 0.37
Health-related variables
ALCOHOL CONSUMPTION *0.0270 1.89 0.0081 0.98
SMOKING -0.0290 --1 .51 0.0070 0.98
Exposure variables
PURPLE PESTICIDE DOSAGE -0.0230 -0.30 ***0.1 100 2.43
RED PESTICIDE DOSAGE 0.0230 0.33 ***0.0440 2.56
AMBER PESTICIDE DOSAGE 0.0088 -0.46 "*-0.0640 -3.61
SPRAYER LEAK *0.3000 1.62 ***0.5300 3.15
MEALS 1N COTTON FIELDS 01600 -0.52 0.3100 0.29
LABEL ILLITERACY -0. 1700 -0.97 "*-0.5600 -3.22
STORAGE HAZARD -0.2900 -1.38 ***0.6700 3.66
Exposure averting and
mitigating variables
IPM GRADUATE 0.0260 0.1 1 - -
FIRST AID KNOWLEDGE -0.2800 - l .56 -0.4000 -1.51
PROTECTIVE CLOTHING **“‘-0.3 100 -5.58 ***-0.3700 -6.34
Institutional variables
BOREHOLE ACCESS 0.2200 1.21 0.1100 0.56
RADIO OWNERSHIP * "—0.5800 -3. 13 -0.0360 -0.18
Pest management perception
variable(s)
REVIEW CALENDAR SPRAY ***-0.6200 -2.74 "-0.4200 -2.23
N 133 1 17
Log likelihood chi-square 125.27 307.73
x2 -p value 0.0000 0.0000

 

 

* " *=significance at 1% level,

Source: Field survey data, 1999

46

"=significance at 5% level, * =significance at 10% level

 

Table 2. 4: Poisson Model Results for Self-reported Acute Eye Symptom Incidences.

 

Sanyati District
Eye Symptom Incidences

Chipinge District

Eye Symptom Incidences

 

 

 

Independent variables coefficient z-value Coefficient z-value
Farmer characteristics

FARMER’S AGE "-0.0210 -2.05 "0.0360 1.99
FORMAL EDUCATION -0.0120 -0.04 -0.0900 -1.57
MALE FARMER 0.5000 1.30 -0.2500 -0.55
Farm management variables

COTTON AREA 0.0160 -0.54 * *0.0260 2.29
EXTENSION MEETINGS 0.0190 0.82 -0.0100 -0.53
FORMAL EMPLOYMENT -0.3800 - l .41 "0.8000 2.11
KNAPSACK ** *0.8400 2.72 * *-0.7800 -2.00
PROPHYLACTIC SPRAY ***0.7500 2.74 ***-1.5900 -2.98
Health-related variables

ALCOHOL CONSUMPTION “0.0420 2.25 * "-0.0530 -3.35
SMOKING 0.0004 0.02 *0.0520 1.87
Exposure variables

PURPLE PESTICIDE DOSAGE 0.0710 0.51 -0.0024 -0.26
RED PESTICIDE DOSAGE -0. 1000 -1.13 -0. 1800 -2.31
AMBER PESTICIDE DOSAGE 0.0220 1.03 "0.0950 2.65
SPRAYER LEAK 0.4800 0.19 0.0078 0.02
MEALS IN COTTON FIELDS 0.4300 1.1 1 0.5900 1.53
LABEL ILLITERACY 0.3300 1.39 -0.3200 -0.82
STORAGE HAZARD 0.0490 0.16 -0.3100 -0.77
Exposure averting and

mitigating variables

IPM GRADUATE "0.8800 2.36 - -
FIRST AID KNOWLEDGE 0.0220 0.09 *-0.8200 -1.82
PROTECTIVE CLOTHING * **-0.2100 -2.67 -0.0400 0.72
Institutional variables

BOREHOLE ACCESS 0.1300 0.53 -0.3200 -0.91
RADIO OWNERSHIP 0.0420 0.15 0.6400 1.56
Pest management perception

variable

REVIEW CALENDAR SPRAY ***-1.1000 -2.77 -0.5500 -1.50
N 133 117

Log likelihood chi-square 111.54 96.10

xz-p value 0.0000 0.0000

 

”*=significance at 1% level, "=significance at 5% level, * =significance at 10% level

Source: Field survey data, 1999

47

Table 2. 5: Poisson Results for Self-reported Total Acute Symptom Incidences, 1998/99
Sanyati District Chipinge District
Total Incidences'2 Total Incidences

 

 

 

Independent variables coefficient z-value coefficient z—value
Farmer characteristics

FARMER’S AGE ***-0.0790 11.01 *0.0150 1.71
FORMAL EDUCATION 0.0210 0.95 * * *0.1000 3 .69
MALE FARMER **"1.4100 4.37 “0.8900 2.32
Farm management variables

COTTON AREA * M0.0630 3.40 "0.0120 2.08
EXTENSION MEETINGS * “ ‘0.0930 6.44 "“00170 2.06
FORMAL EMPLOYMENT 0.0530 0.30 * * *-0.6800 -3.72
KNAPSACK 0.1900 1.19 * “-0.8800 -4.27
Health-related variables

ALCOHOL CONSUMPTION *0.0280 1.79 0.0130 1.30
SMOKING -0.0060 -0.34 0.0140 1.01
Exposure variables

PURPLE PESTICIDE DOSAGE * * *0.3100 2.77 0.0380 0.57
RED PESTICIDE DOSAGE *0.1300 1.86 ** *0.0970 2.86
AMBER PESTICIDE DOSAGE -0.0210 -0.58 "-0.0620 -2.26
SPRAYER LEAK 0.1700 1.09 "*08200 4.91
MEALS IN COTTON FIELDS ”0.6000 2.29 0.2900 1.52
LABEL ILLITERACY *0.2600 1.75 0.4200 0.25E-1
Exposure averting and

mitigating variables

IPM GRADUATE -0.2800 -1.35 - -
FIRST AID KNOWLEDGE * * ‘-0.5 100 -3.46 M05000 -2. l4
PROTECTIVE CLOTHING * "-0.1500 -3.56 *-0. 1000 -1.76
Institutional variables

BOREHOLE ACCESS *"06900 4.23 0.2000 1.13
CREDIT USE -0.3000 -1.51 * "-0.6200 -2.98
RADIO OWNERSHIP * * *-1.2100 -7.66 0.1600 0.78
Pest management perception

variable

REVIEW CALENDAR SPRAY ***-1.1500 -4.75 *-0.3100 -1.71
N 133 119

Log likelihood chi-square 495.54 165.81

x2 —p value 0.0000 0.0000

 

"*=significance at 1% level, **=significance at 5% level, * =significance at 10% level

Source: Field survey data, 1999

 

I 2 Acute symptom incidences refer to short-term illness episodes experienced by the farmers and these
in c lude both the dermal (eye and skin irritation) and oral (ingestion) symptoms. Therefore, the total
incidence model aggregates skin, eye and stomach (gastro-intestinal) poisoning episodes incurred by the
farrn er during and or soon after spraying pesticides.

48

Table 2.6: Poisson Protective Clothing Adoption Model Results, 1998/99

Dependent variable: Count of protective clothing ownership

 

 

 

 

Sanyati District Chipinge District
Independent Coefficient z-value Coefficient z—value
Variables Estimate Estimate
Farmer characteristics
Farmer’s age ***-0.0140 -2.97 0.0120 1.32
Education 01300 -1.06 ***0.5700 2.48
Male farmer 0.0790 0.56 0.2200 0.64
Farm management
variables
Total area cultivated 0.0046 0.34 -0.01 10 -1.29
Extension meetings ‘0.0190 1.88 “*00460 5.58
Formal employment 0.0840 0.73 **-0.5200 -2.93
Certified master farmer -0.0440 -0.36 -0.1200 -0.63
Knapsack “0.2300 2.15 "-0.4100 -2.06
Prophylactic spray 0.1800 1.50 0.2600 1.38
Health-related variables
Predicted skin incidences ***-0.3000 -3.91 ***-0.4300 -5.80
Predicted eye incidences -0.0880 -0.99 -0.0850 -0.49
Alcohol consumption 0.0860 0.66 ‘-0.2800 -1 .63
Smoking -0.0410 -0.29 0.0790 0.35
Exposure variables
Purple pesticide class -0.3100 -0.62 -0.3900 -1.20
Red pesticide class 0.2400 0.58 *"0.6000 3.23
Amber pesticide class 0.0082 -0.86 -0.2100 -1.50
Label illiteracy -0.0410 -0.36 * * *-0.8000 -4.76
Exposure averting and
mitigating variables
First aid knowledge 0.0610 0.60 0.2500 -1.37
IPM awareness *0.1200 1.72 - -
Institutional variables
Distance to health center 0.0085 0.44 -0.0036 -0.25
Radio ownership -0.1700 - l .43 -0.2500 -1.36
N 133 117
Log Likelihood x’ 44.69 101.43
x2 —-p value 0.0019 0.0000

 

 

***=significance at 1% level, ”=significance at 5% level, "‘ =significance at 10% level

Source: Field survey data, 1999

49

Table 2.7: Probit Results for Eating in Cotton Fields in Survey Districts, 1998/99

Dependent variable: Eating in Cotton Fields

 

 

 

 

SANYATI DISTRICT CHIPINGE DISTRICT
Independent variables coefficient z-value coefficient z—value
Farmer characteristics
Farmer’s age -0.0003 -0.75 -0.0017 -0.45
Male farmer -0.0130 -1.45 -0.2700 -1.45
Formal education **-0.0190 -2.07 0.1 100 1.04
Farm management
variables
Extension meetings 0.0004 0.55 * "-0.0130 -2.89
Cotton area cultivated 0.0012 0.63 0.0047 0.44
Cotton bales harvested 0.0007 0.87 0.0027 0.90
Knapsack sprayer 0.0140 1.51 * "-0.3100 -3.68
Master farmer certified *-0.0130 -1.62 0.0570 0.50
Health-related variables
Alcohol intake 0.1300 0.83 0.0970 1.14
Smoking **0.0710 2.23 ***-0.2900 -3.16
Exposure variables
IPM graduate 0.0013 0.12 - -
Label illiteracy 0.0270 1.49 ** *0.2000 2.30
Exposure averting and
mitigating variables
First aid knowledge 0.0074 0.87 -O. 14 1.46
Protective clothing ***-0.1000 -2.58 *0.047 1.90
Institutional variables
Borehole **-0.0220 -1.95 -0.03 80 044
Health center distance -0.0030 -1.54 -0.0034 -0.46
Pest Management
Perception variable
Review calendar spray ***0.1 100 2.51 -0.1 100 - l .3400
N 137 134
Log likelihood x2 35.50 42.79
x2 —p value 0.0006 0.0003

 

***-significance at 1% level,**-significance at 5% level, "‘-significance at 10% level

Source: Field survey data, 1999

50

l3

Dependent variable: Leaking sprayers

Table 2.8: Probit Results for Leaking Sprayers in Survey Districts”, 1998/99

 

 

 

 

SANYATI DISTRICT CHIPINGE DISTRICT
Independent variables coefficient z-value coefficient z-value
Farmer Characteristics
Farmer’s age 0.3200 0.82 -0.0028 -0.68
Male farmer «0.0210 -0. 15 -0.1 103 -0.66
Formal education *0.2300 1.85 -0.0056 -0.39
Farm Management
variables
Extension meetings *0.1300 1.62 -0.0017 072
Cotton area cultivated ***0.0560 2.44 0.0061 1.60
Formal employment 0.03 10 0.28 -0.0200 -0.21
Knapsack sprayer 0.8400 0.80 -0.1000 -1.06
Knapsack Age -0.6400 -1.36 -0.0027 -0.54
Master farmer certified -0.0730 -0.59 -0.1373 -1.25
Prophylactic spray *0.1800 1.67 0.0988 0.61
Exposure variables
IPM graduate ***-0.3700 -3.06 - -
Label illiteracy -0.0500 -0.48 -0.0109 -0. 12
Exposure averting and
mitigating variables
First aid knowledge -0.0700 -0.71 -0.0475 -0.43
Protective clothing -0.0520 -1.60 -0.0197 -0.75
N 138 134
Log likelihood x’ 41.70 13.86
1’ -p value 0.0001 0.3838

 

Source: Field survey data, 1999

 

5]

**"‘-significance at 1% 1eve1,**-significance at 5% level, *-significance at 10% level

S prayer leakage probit model for Chipinge District is not significant.

Table 2. 9: Probit Results for First Aid Knowledge in Survey Districts, 1998/99

Dependent variable: First Aid Knowledge

 

SANYATI DISTRICT CHIPINGE DISTRICT

 

 

 

Independent variables coefficient z-value coefficient z-value

Farmer characteristics

Farmer’s age 0.0009 0.23 0.0051 1.47
Formal education -0.0620 -0.55 * * *0.3300 3 .01
Farm management variables

Cotton area cultivated 0.0040 0.33 "0.0054 1.97
Formal employment *0. 1600 1.69 0.5300 0.71
Knapsack sprayer -0.0570 -0.57 -0.0450 -0.64
Prophylactic spray 0.0210 0.21 0.0680 0.85
Health-related variables

Alcohol intake 0.1100 0.90 0.0520 0.69
Smoking -0.0640 -0.50 -0.0570 -0.71
Exposure variables

Meals in cotton fields 0.1 100 0.75 ‘-0. 1400 -1.80
Label illiteracy *-0.1700 -1.73 0.0760 1.09
Exposure averting variables

Protective clothing 0.0420 1.40 0.0300 1.60
Institutional variables

Received Treatment" -0.0270 -0.1 1 *-0.1500 -1.77
Pest management perception variable

Review calendar spray -0.1 100 -1.34 0.0520 0.78
N 134 123

Log likelihood x2 42.79 27.95

x2 —p value 0.0003 0.0072

 

I""""-significance at 1% 1evel,**-significance at 5% level, *-significance at 10% level

Source: Field survey data, 1999

 

Treatment received is a dichotomous response variable which assesses whether or not a farmer received
fortnal treatment from a health facility [i.e. clinic or hospital] or private physician.

52

 

Cost of Illness Model
(OLS Log-linear Model)

 

 

Acute Symptom Incidence Model
(Poisson Regression)

 

 

 

+Protective Clothing + Eating in fields
(Poisson Model) (Probit Model)

+ First aid knowledge + Leaking sprayers
(Probit Model) (Probit Model)

 

 

   

Figure 2.1: Structure of Econometric Analysis

53

 

REFERENCES

Antle, J .M., and P.L.Pingali. “Pesticides, Productivity and Farmer Health: A Philippine
Case Study.” American Journal of A gricultural Economics. 76(1994):418-30.

Antle, J .M., and SM. Capalbo. “Pesticides, Productivity, and Farmer Health:
Implications for Regulatory Policy and Agricultural Research.” American Journal
of A gricultural Economics. 76 (1994):598-602.

Antle, J.M., D.C. Cole, and CC. Crissman. “Further Evidence on Pesticides, Productivity
and Farmer Health: Potato Production in Ecuador.” Agricultural Economics
Journal. 18(1998): 199-207.

A j ayi, O. “Measuring the Indirect Health Benefits of IPM: Methodology for Estimating
Pesticide Health Costs in Africa.” Proceedings of the International Workshop an
Evaluation of IPM Programs: Towards a Framework for Economic Evaluation.

Garbsen, Hannover, Germany 25-27 May, 1999.

Chitemerere, R.C. “Pesticides and Occupational Safety and Health in Zimbabwe.”
Pesticides in Zimbabwe: Toxicity and Human Health Implications. Charles F .B.
Nhachi and Ossy M.J. Kasilo, eds. pp 17-24. Harare, University of Zimbabwe
Publications, 1996.

C artwright, B., J .K.Collins, and G.W. Cuperus. “Consumer Influence on Pest Control
Strategies for Fruits and Vegetables.” Successful Implementation of Integrated
Pest Management for Agricultural Crops. Anne R. Leslie and Gerrit, W. Cuperus,
editors. Lewis Publishers Ann Arbor, 1993.

Cole, C.D., F.Carpio, J.A. Julian, and N.Leona. “Health Impacts of Pesticide Use in
Carchi Farm Populations”. Economic, Environmental and Health Trade-ofls in
Agriculture: Pesticide and the Sustainability of Andean Potato Production.
Charles C. Crissman, John M. Antle and Susan M. Capalbo, editors. Kluwer
Academic Publishers, Boston, USA, 1998.

C uyrio, L.C.M. “An Economic Evaluation of the Health and Environmental Benefits of
the IPM Program (IPM CRSP) in the Philippines.” Ph.D dissertation, Virginia
Polytechnic Institute and State University, 1999.

54

Crissman, C.C., D.C.Cole, and F. Carpio. “Pesticide Use and Farm Worker Health in
Ecuadorian Potato Production.” American Journal of Agricultural Economics.

76:(1994):593-97.

Cropper, M.L. “Economic and Health Consequences of Pesticide Use in Developing
Country Agriculture: Discussion.” American Journal of A gricultural Economics.

76(1994):605-O7.

Czapar, G.F., M.P.Curry, and ME. Gray. “Survey of Integrated Pest Management
Practices in Central Illionois.” Journal of Production Agriculture. 8:(1995):483-

86.

Deolalikar, A.B., and R. Laximinarayan. “Socioeconomic Determinants of Disease
Transmission in Cambodia.”Discussion Paper 00-32. Resources for the Future,

Washington, DC. 2000.

Environmental Protection Agency (EPA). “What is a pesticides?” Office of Pesticide
Programs, (URL) www.cpagovflaesticides/whatis.htm. 1999.

Environmental Protection Agency. “Environmental Protection Agency Endocrine
Disruptor Screening Program, Report to Congress.” Washington DC. 2000.

F ernandez-Cornejo, J. “Environmental and Economic Consequences of Technology
Adoption: IPM in Viticulture.” Agricultural Economics Journal. 1 8(1997): 145-55.

F Omey, D.R. “Importance of Pesticides in Integrated Pest Management” Pesticides:
Managing Risks and Optimizing Benefits. ACS Symposium Series 734. Nancy N,
Ragsdale and James N, eds. Seiber. American Chemical Society, Washington

DC. 1999.

F le i Scher, G., V.And01i, M.Coulibaly, and T. Randolph. “Economic Analysis of the
Pesticide Sub-Sector in Cote d’Ivoire.” Pesticide Policy Project Publication Series

No. 06/F. University of Hannover, Germany, 1998.

Fleischer, G. “The Role of Economic Analysis of Pesticide Use and Policy- Experiences

55

From Country Case Studies.” Pesticide Policies in Zimbabwe: Status and
Implications for Change. pp208-220. Pesticide Policy Project, Publication Series
Special Issue No.1, Godfrey Mudimu, Herman Waibel and Gerd Fleischer, eds.
Hannover University, Germany, 1999.

Honsby, A.G., R.D.Wauchope, and AB. Hemer. Pesticide Properties in the
Environment. New York, NY: Springer-Verlag. 1996.

Hurley, T.M., J .B. Kliebenstein, and PF. Orazem. “An Analysis of Occupational Health
in Pork Production.” American Journal of A gricultural Economics. 82(2000):323-
33.

J owa, P. “Present Status of IPM on Cotton and Future Needs in Zimbabwe.” Risk Fund

Project, Ciba-Geigy Zimbabwe. [PM Planning and Implementation Workshop: 6-
8 June, 1995, Holiday Inn, Harare, Zimbabwe. 1995.

Kenmore, P. “A Perspective on IPM.”ILE1A Newsletter. Center for Research and

Information on Low-External-Input and Sustainable Agriculture. Wageningen,
Netherlands, 1 3(1997):8-9.

Lele, U., N. Van De Walle, and M, Gbetibouo. “Cotton in Africa: An Analysis of
Differences in Performance. Managing Agricultural Development in Africa.”
Discussion Paper 7. The World Bank, Washington DC. 1989.

Loewenson, R and C.F.B. Nhachi. “Epidemiology of the Health Impact of Pesticide Use
in Zimbabwe.” Pesticides in Zimbabwe: Toxicity and Health Implications. pp 25-
35. Charles F .B. Nhachi and Ossy M.J.Kasilo, eds. Harare, University of
Zimbabwe Publications, 1996.

Maumbe, B.M. and SM. Swinton. “Why Do Smallholder Cotton Growers in Zimbabwe
Adopt IPM? The Role of Pesticide-Related Health Risks and Technology
Awareness.” Selected Paper, American Agricultural Economists Association,
Tampa, Florida, July 30th -August 2nd, 2000.

Mbanga, T. “Pesticides and the Agricultural Chemicals Industry Association.”
Pesticides in Zimbabwe: Toxicity and Health Implications. pp 12-16. Charles F.B.
Nhachi and Ossy M.J.Kasilo, eds. Harare, University of Zimbabwe Publications,
Zimbabwe, 1996.

56

 

Moses, M. Harvest of Sorrow: Farm Workers and Pesticides. Pesticide Education Center,
San Francisco, 1992.

Nhachi, C.F.B. “Toxicology of Pesticides and the Occupational Hazards of Pesticide Use
and Handling in Zimbabwe.” Pesticide Policies in Zimbabwe: Status and
Implications for Change. pp 125-133. Godfrey D. Mudimu, Hermann Waibel and
Gerd Fleischer, eds. Pesticide Policy Project, Publication Series Special Issue
No.1, University of Hannover, Institute of Horticultural Economics, Hannover,
Germany.

Owens, N.N., S.M. Swinton, and E0. Van Ravenswaay. “Will Farmers Use Safer
Pesticides?” Staff Paper 97-1, Department of Agricultural Economics, Michigan
State University, East Lansing, USA, 1997.

Page, S. L.J. “Natural Pest Management in Zimbabwe.” ILEIA Newsletter. Center
for Research and Information on Low-Extemal-Input and Sustainable Agriculture.
Wageningen, Netherlands, 13(1997): 12-13.

Pesticide Policy Project. Pesticide Policies in Zimbabwe: Status and Implications for
Change, Publication Series Special Issue No.1, Godfrey Mudimu, Herman Waibel
and Gerd Fleischer, eds. Hannover University, Germany, 1999.

Pincus, J., H.Waibel, and F. Jungbluth. “Pesticide Policy: An International Perspective.”
Approaches to Pesticide Policy Reform- Building Consensus for Future Action. A
Policy Workshop. Hua Hin, Thailand , July 3-5, 1997. Nipon Poapongsakorn,
Lakchai Meenakanit, Hermann Waibel and Frauke J ungbluth, editors. Pesticide
Policy Project Publication Series No.7. University of Hannover, Germany, 1999.

Pingali, P.L., C.B. Marquez, F.G. Palis, and AC. Rola. “The Impact of Long Term
Pesticide Exposure on Farmer Health: A Medical and Economic Analysis in the
Philippines.” Impact of Pesticides on Farmer Health and the Rice Environment,
P.L. Pingali and PA. Roger, editors. Boston, MA: Kluwer Academic Publishers,
1995.

57

Ramirez, 0, A., and S. D. Shultz. “Poisson Count Models to Explain the Adoption of
Agricultural and Natural Resource Management Technologies by Small Farmers
in Central American Countries.” Journal of A gricultural and Applied Economics,
32 (2000):21-33.

Rengam, S.V. “The Struggle Against Pesticides.” Women and 1PM: Crop Protection
Practices and Strategies. Royal Tropical Institute, Amsterdam, Netherlands,
1999.

Rother, H, and L. London. Pesticide Health and Safety Policy Mechanism in South
Africa: The State of the Debate. Occupational and Environmental Health
Research Unit, Working Paper Number 1, Department of Community Health,
University of Cape Town, Cape Town, South Africa, 1998.

Ruttan, V.W. Technology, Growth and Development: An Induced Innovation
Perspective. New York, NY: Oxford University Press, 2001.

Sheets, T.J., and D. Pimentel, Pesticides: Contemporary Roles in Agriculture, Health,
and Environment. Humana Press, Clifton, New Jersey. USA, 1979.

Sasakawa-Global 2000. Innovative Extension Education in Africa. International
Workshop on Innovative Training Programs for Mid-Career Agricultural
Extension Professionals in Sub-Saharan Africa, Red Cross Training Institute,
Addis Ababa, Ethiopia, July, 6-8, 1999. Sasakawa Africa Association, Steven A.,
Breth, editor. Mexico City, Mexico, 1999.

Smith, M.E., J. K. Lewandrowski, and N.D.Uri. “Agricultural Chemical Residues as a
Source of Risk.” Review of Agricultural Economics. 22(2000):313-25.

Straus, J. and D.Thomas. “Health, Nutrition and Economic Development.” Journal of
Economic Literature. 36 (1998):766-817.

Sunding, D., and J. Zivin. “Insect Population Dynamics, Pesticide Use and F arm-worker
Health.” American Journal of A gricultural Economics. 82 (2000):527-40.

Swinton, S.M. “The Effect of Health Risk on Agricultural Input Demand.” Staff Paper

58

Number 93-2, Department of Agricultural Economics, Michigan State University,
East Lansing, Michigan, 1993.

Swinton, S.M., N.N. Owens and ED. van Ravenswaay. “Health Risk Information to
Reduce Water Pollution.” Flexible Incentives for the Adoption of Environmental
Technologies in Agriculture. Frank Casey, Andrew Schmitz, Scott Swinton and
David Zilberman, eds. Kluwer Academic Publishers, 1999.

Task Force of Environmental Cancer and Heart and Lung Disease. The Eflects of
Pesticides on Human Health. Proceedings of a Workshop May 9-11, 1988
Keystone Colorado, Scott, R. Baker, and Chris F. Wilkinson, editors. Princeton
Publishing Co., Inc. Princeton, 1988.

The Pesticide Manual. Eleventh Edition. British Crop Protection Council, Surrey, United
Kingdom, 1997.

The Pesticide Trust. The Pesticide Hazard: A Global Health and Environmental Audit.
Compiled by Dinharm, B. The Pesticide Trust. Biddles Limited, Guildford and
King’s Lynn, London, United Kingdom, 1993.

Tjomhom, J .D., G.W. Norton, K.L. Heong, N.S. Talekar and VP. Gapud. “Determinants
of Pesticide Misuse in Philippine Onion Production.” Philippine Entomologist,
11(1997):139-49.

Van Emden, HF, and D. B. Peakall. Beyond Silent Spring: Integrated Pest Management
and Chemical Safety. Chapman and Hall, London, United Kingdom. 1996.

Waibel, H. “Methodological Aspect in Benefit Assessment of Pesticides.” Pesticide
Policies in Zimbabwe: Status and Implications for Change. Godfrey. D Mudimu,
Hermann Waibel and Gerd Fleischer, editors. Pesticide Policy Project. Publication
Series Soecial issue No.1. Hannover, Gemany, 1999.

Watterson, A., Pesticide Users’ Health and Safety Handbook: An International Guide,
Van Nostrand Reinhold, New York, USA, 1988.

Watts, M. Poisons in Paradise: Pesticides in the Pacific. Pesticide Action Network Asia
and the Pacific, Penang, Malaysia, 1993.

59

World Health Organization. Public Health Impact of Pesticides Used in Agriculture.
Geneva, Switzerland, 1990.

World Bank, World Development Report 2000/2001: Attacking Poverty. Oxford
University Press, New York, 2000.

World Bank, Integrated Pest Management: Strategies and Policies for Effective
Implementation. Environmentally Sustainable Development Studies and
Monographs Series No.13. Washington DC, 1996.

World Wildlife Fund. “Chemicals that Compromise Life: A Call to Action.” Issue Brief.
Washington DC, 1998.

Yudelman, M., A.Ratta, and D. Nygaard. Pest Management and Food
Production: Looking to the Future. 2020 Vision, Food Agriculture and the
Environment, Discussion Paper 25, International Food Policy Research Institute,
Washington DC, USA, 1998.

Zilberman, D., and F. Castillo. “Economic and Health Consequences of Pesticide Use in
Developing Country Agriculture: Discussion.” American Journal of Agricultural
Economics. 76(1994):603-04.

Zimbabwe Ministry of Lands, Agriculture and Rural Resettlement. Guidelines for the
Safe and Eflective Use of Pesticides. Agricultural Technical and Extension
Services, Harare, Zimbabwe, 1986.

60

APPENDICES

Appendix Al: Langragian Function

The associated Langragian function is given by:
(I) L = U(Y,GJD +1(0’Y(X) -Wx)-(WaA+ Wg G+th))+fl(Hm-H(XP.M” 11, A))

First order conditions:

(2) dL/dx: 1(p6Y/6x—w) —u 6H/6x =0
(3) JL/dG: dU/dG-le =0
(4) dL/dh: -kWh—u6H/6h =0
(5)6L/61: pY(x)-wx- WaA- WgG-Whh =0
(6)6L/6u: Hm—H(G,x,h) =0
(7) 6L/6H: dU/6H+p =0
(8) 5L/5A: JeWa-y 6H/5A =0

Rearrange ( 7) to make u the subject of the formula and substituting the result into
equation (4) i.e. first order conditions with respect to health care services to obtain the
following: (9) u = - ( 6U/ 6H)
(10) k = ((6U/6ID(6H/6h))/ W1.
Substituting u in equation (9) into equation (2) and divide through by i. as follows:
(11)p 6Y/6x = w- (6U/6H*§H/6x) ll
Now substitute equation (10) into equation (11) to obtain the following:
(12) p 6Y/6x = w - (Wh/(JU/6H*6H/dh)(6U/6H*6H/dx)
Equating LHS of equations (11) and (12) we obtain the following expression:

(13) w - (dU/aH*oH/dx) m = w - ( W,. /(dU/6H*6H/6h)(6U/6H*6H/5x)

(14) (6U/6H*6H/6x)/.1 = ( Wh/(6U/6H*bH/6h)(6U/6H*6H/6x)

(15)(6U/6H*6H/6x) /.z= / (6U/(SH*6H/6h) /1] = Wh(6U/6H*6H/6x)
(6U/6H*6H/6x)/ W, (6U/6H*6H/6h )

 

 

61

Table A1.1: Pesticide-related health symptoms for cotton growers, 1998/99

 

 

 

PESTICIDE-RELATED SAN YATI (N=140) CHIPINGE (N=140)
SYMPTOMS % Number % Number
Acute Health Costs (Z5 Mean) 180 3 16
Acute Health Symptoms
Skin irritations 67.4 95 55.0 77
Eye irritations 37.6 53 26.4 37
Stomach poisoning 7.1 10 12.1 17
Systemic Health
Symptoms
Nausea 1.4 2 5.7 8
Vomiting 1.4 2 0.0 0
Abdominal pains 9.2 13 2.9 4
Blurred vision 5.0 7 6.4 9
Dizziness 19.9 28 10.0 14
Nasal bleeding 1.4 2 0.7 1
Severe headache 3.5 5 0.0 0
Coughing 1.4 2 1.4 2
Sneezing 9.2 13 0.0 0
Diarrhea 0.0 0 1.4 2
Multiple symptoms 7.8 1 1 23.6 33
None of the above 39.7 56 47.9 67
Chronic Health symptoms
Type of Individual Affected
Household head 20.6 29 1 1.5 16
Spouse 21.3 30 4.3 6
Child 2.1 3 13.7 19
Resident relative 5.0 7 7.9 1 1
None 49.6 69 61.4 85

 

Chronic diseases surveyed: 1=Cancer 2=Back pain 3=Lung problem 4=Blindness

Source: Field survey data, 1999.

62

 

Table Al.2: Pesticide—related farmer health characteristics in study regions, 1998/99

 

 

FARMER CHARACTERISTICS SAN YATI CHIPINGE
N=141 N=140
Pesticide Applicator Health -Related
Characteristics
Current smoker (%) 22.7 26.4
Current drinker (%) 47.5 58.6
Previous smoker (%) 15.6 7.1
Previous drinker (%) 16.3 12.9
Eat in cotton fields (%) 9.9 27.9
Smoke in cotton fields (%) 2.1 5.0
Knowledge of first aid (%) 61.0 19.3
Ownership first aid kit (%) 6.4 2.1
Prophylactic spray (%) 29.8 25.9
Leaking sprayer (%) 39.0 34.3
No protective clothing (%) 3.6 34.3
Ownership of sprayer (%) 73.8 69.3
Average scouting (Sessions) 17 16
Smoking duration (Years)
Mean 2.13 2.79
Standard Deviation 5.09 7.03
Alcohol Drinking Duration (Years)
Mean 3.62 9.66
Standard deviation 6.61 13.70
Sick Days/Recuperation Period (Days)
Mean 1.60 4.34
Standard Deviation 3.1 1 10.29
Clinical treatment costs (ZS)
Mean 5.29 16.32
Standard Deviation 37.31 147.46
Health Information Sources
Mean 2.06 1.29
Standard deviation 1.14 0.63

 

Source: Survey data, 1999.

63

Table A1.3: Distribution of pesticide storage strategies in study regions, 1998/99

 

 

PESTICIDE STORAGE SANYATI (%) N=141 CHIPINGE (%)
STRATEGY N=140

Pesticide store-room 22.0 41.4
Pesticide and food store-room 35.5 17.9
Bedroom 30.5 17.1
Locked suitcase/trunk 4.3 2.1
Bush 0.7 0.7
None of the above 6.4 17.9

 

Source: Field Survey, 1999.

64

Table A1.4: Label literacy among cotton growers in survey areas, 1998/99

 

 

 

 

 

 

 

 

COLOR CODE TOXICITY SANYATI (%) N=138 CHIPINGE (%) N=140
RANKING
Correct ranking all color codes 71.8 41.7
Most toxic color code identified 1.7 5.0
Least toxic color code identified 8.9 7.2
Complete ignorance 9.3 46.1
Source: Survey data, 1999
Table Al.5: Knowledge of triangles in correct order of toxicity, 1998/99
KNOWLEDGE OF COLOR CODE SANYATI SANYATI CHIPINGE CHIPINGE
RANKING NUMBER PERCENT NUMBER PERCENT
None 13 9.3 65 46.1
Partial 15 10.6 17 12.2
Complete 101 71.8 58 41.7
Total 129 91.7 140 100.0
Source: Survey data, 1999
Table A1.6: Pesticide use and toxicity classes by farmers in survey area, 1998/99
PESTICIDE TOXICITY PESTICIDE SANYATI CHIPINGE
COLOR CODES (CLASSES) CLASS DISTRICT DISTRICT
DESCRIPTION (%) N=140 (%) N=140
1. Purple Very Dangerous 5.1 5.1
11. Red Dangerous 54.3 19.9
111. Amber Poisonous 40.6 75.0
1V. Green Harmful if swallowed 0.0 0.0

 

Source: Field Survey data, 1999

65

Table A1.7: Cotton pesticides used in Sanyati and Chipinge districts, 1998/99

 

 

Chemical Chemical Common Color Chemical Dermal Oral
Trade Name Name Code Group LDso LDso
mg/kg mg/kg
AGRITHRIN F ENVALERATE AMBER Pyrethroid >5000.0 45 1
FERNVALERATE FENVALERATE AMBER Pyrethroid >5000.0 45 l
AGRITHRIN SUPER ESFENVALERATE RED Pyrethroid >5000.0. 75 to
88
PFUMO F ENVALERATE AMBER Pyrethroid >5000.0 45 1
MOTO F ENVALERATE AMBER Pyrerhroid >5000.0 451
CARBARYL CARBARYL AMBER Carbamate >2000.0 590.0
SEVIN CARBARYL AMBER Carbamate >2000.0 590.0
DIMETHOATE DIMETHOATE RED Organophosphate 1000.0 60.0
ROGOR DIMETHOATE RED Organophosphate 1000.0 230.0
KARATE LAMPDA- RED Pyrethroid >5000 1 1000
CYHALOTHRIN
LARVIN THIODICARB RED Carbamate >2000.0 166.0
MARSHAL/SHASHA CARBOSULFAN RED Carbamate >2000 250
ONCOL BENFURACARB AMBER >2000 222.6
THIODAN ENDOSULFAN PURPLE Organochlorine 256.0 44.9
THIONEX ENDOSULF AN PURPLE Organochlorine 256.0 44.9
MITAC AMITRAZ AMBER Acaricide >2000.0 1000.0
to
2000.0
MONOCROTOPHOS MONOCROTOPHOS PURPLE Organophosphate l 26 1 8
NUVACRON MONOCROTOPHOS PURPLE Organophosphate 126 18
AZODRIN MONOCROTOPHOS PURPLE Organophosphate 126 1 8
SECURE CHLORFENAPYR RED Organophosphate >2000 44 1
F ASTAC ALPHAMETHRIN AMBER Organochlorine >2000 79-400

 

Source: 1.Field Survey, 1999

2. www.cdms.net/ldt/mp27QOOO.pdf

 

3.The Pesticide Manual, 1997

66

Table A1.8: Spaying equipment distribution in cotton growing regions, 1998/99

 

 

SPRAYER TYPE SANYATI DISTRICT CHIPINGE DISTRICT
(%) N=141 (%) N=140

Knapsack 68.80 42.00

Ultra-Low —Volume (U LV) 5.00 26.40

No sprayer 26.20 31.40

Hired sprayer 10.00 1.00

 

Source: Field survey data, 1999

Table A1.9: Average cotton pesticide applications by region, 1998/99

 

 

CHEMICAL APPLICATIONS SANYATI (%) CHIPINGE (%)
PER SEASON N=129 N=114

Less than 3 64.00 20.00
Between 4 and 7 26.80 31.20
Between 8 and l l 0.00 14.20
Between 12 and 14 0.00 3.50

More than 15 0.00 11.30

 

Source: Field survey data, 1999

Table Al.10: Pesticide-related treatment patterns in survey areas, 1998/99

 

 

PESTICIDE-RELATED HEALTH SANYATI CHIPINGE
AILMENT °/o TREATED % TREATED
Stomach Poisoning 2.8 5.0

Skin Irritation 2.8 7.9

Eye Irritation 2.1 7.9
Mean 2.6 6.9

N 140 140

 

Source: Field survey, 1999

67

Table A1.11: Types of protective clothing worn by pesticide applicators, 1998/99.

 

 

PROTECTIVE CLOTHING SANYATI (%) N=140 CHIPINGE (%) N=140
Rubber boots 88.0 37.8
Protective eye glasses 29.6 10.0
Long sleeved overall 93.8 53.3
Face mask 56.4 20.6
Respirator 57.0 25.0
Rubber gloves 61.2 27.8

 

Source: Field survey data, 1999

Table A.12: Mean health costs as a proportion of household income and costs

 

 

HEALTH COSTS AS PERCENT OF COST/INCOME SANYATI (%) CHIPINGE(%)
ITEM N=140 N=140

Total chemical cost outlay 45 83
Cotton sales revenue 2 5

Off -fann income 14 6

 

Source: Field survey data, 1999

68

CHAPTER 3

ADOPTION OF COTTON I.P.M. IN ZIMBABWE: THE ROLE OF
TECHNOLOGY AWARENESS AND PESTICIDE-RELATED HEALTH RISKS

3.0 Introduction

In Africa, crop protection is centered on chemical control of pests with alternative
approaches still in minimal use (Adesina, 1994; Ajayi, 1999). Cash crops like cotton have
relied on chemical pesticides but the limitations of chemical pest control have become
increasingly clear to both farmers and policy makers. Pesticide use in Africa has been
tied to small farm credit programs (Fleischer, 1999). Although the application of
chemical pesticides has alleviated pest problems in the short term, pesticide use has led to
negative extemalities such as secondary pest outbreaks, development of pesticide
resistance and the destruction of natural enemies thereby putting farmers in a vicious
pesticide treadmill (Burrows, 1983;World Bank, 1996).

The calendar-based techniques are increasingly being questioned for a number of
reasons. First, these traditional chemical-based pest management tactics have failed to
provide essential ingredients for sustainable crop production, which includes the
attainment of multiple benefits such as effective pest control, raising agricultural
productivity and minimum damage to the environment. Second, chemical control of pests
has elevated occupational health hazards particularly in less developed countries (LDCS)
where farmers do not afford protective clothing (Cole et al., 1998; Loewenson and
Nhachi 1996; WHO, 1990). Rising concern for public health risks of pesticide use as well
as its burden on the environment has added momentum to the need to re-evaluate the
current chemical-based pest management practices (Rola and Pingali, 1993). Until the

past decade, the debate advocating the substitution of pest control based exclusively on

69

chemicals with new approaches such as integrated pest management (IPM)15 has not been
strong in Africa. Local constituents advocating the protection of the environment and
public health are still in their development stages on the continent. Yet the farmers’ low
level of literacy and education makes the overall risk of exposure to pesticide greater than
elsewhere in the world despite the fact that Africa uses about 2% of the world chemical
sales (Kiss and Meerman, 1993;Fleischer, 1999).

The benefits of knowledge-based technologies such as IPM in reducing over-
application of pesticides thus improving productivity, human health and the environment
have been demonstrated in a number of studies conducted mostly in developed countries
(Fernandez-Cornejo, 1998; Swinton, et al., 1999; Norton and Mullen, 1993;Thomas et al.,
1990) and also in Asia (Antle and Pingali, 1994) and South America (Antle et al., 1998).
The momentum for the development of IPM is relatively high in Asia but is still very
limited in Africa (Adesina, 1994). The general consensus on IPM recognizes that control
of pests exclusively with pesticides satisfies a short-term need.

An increasing number of development agencies including the Food and
Agriculture Organization (FAO), the International Labor Organization (ILO) and the
World Health Organization (WHO) observe that priority should be given to education of
pesticide users and promoting systems that restrict or eliminate pesticide use (Weber,
1996). Smallholder cotton growers can make the transition from the use of calendar-

based chemical pest management through exposure to Farmer Field School (FFS)'6. The

 

'5 1PM is a sustainable approach to managing pests by combining biological, cultural, physical, and
chemical tools in a way that minimizes economic, health, and environmental risks (Vandeman et a1 1994).

'6 FFS is a participatory training approach that uses discovery-based learning techniques in pest and crop

management. Its aim is to help farmer groups understand agro-ecosystems analysis in order to cope with
biotic (insect, pests and weeds) and abiotic (water soil and weather) stresses (Rola, undated). Farmer-to-

70

concept of F FS arose from the dual problem of development of pesticide resistance and
increasing health risks among farmers in rice—based monocultures in Asia. The FSS
philosophy revolves on four principles; (1) growing a healthy crop, (2) weekly field
observations, (3) conserving natural enemies and (4) understanding the field ecology
including water and nutrient management (Fleischer et al., 1999). In Zimbabwe, this
approach is being used to disseminate Integrated Production and Pest Management
(IPPM)l7 technology widely viewed as the means to ameliorate the pesticide menace.
IPPM, unlike single item innovations such as high-yielding varieties (HYVs), relies on
multiple pest management practices, soil and water conservation, and weather
assessments in making pest management interventions. It is essential to understand how
such an information-intensive technology is adopted in practice if its prospects for
widespread implementation are to be fulfilled.

Although several studies have examined the adoption of IPM in cotton,
(Thomas,1990; Ladewig and McIntosh, 1990; F emandez—Comejo, 1996; Napit et al.,
1988), none addresses the smallholder context and none focuses on Africa. Our study
differs from previous studies in that we focus on an emerging innovation still in its early
stage of the diffusion cycle in a region that has received no similar systematic studies in
the past. This study looks at the adoption of different cotton pest management practices
by smallholders in transition from conventional calendar-based chemical pest control to
FFS-IPPM strategy. In particular it examines the roles of 1) IPM technology awareness

and 2) health experience related to pesticide use.

 

farmer approaches are then used to spread IPM knowledge as the process involves selecting farmers who
excel in each FFS group and empowering them to subsequently train other farmers in their own villages.

7l

3.1 Problem Overview

The indiscriminate use of toxic pesticides is associated with farmer health and
environmental risks. The severe danger from pesticide use implies that a reduction of
pesticides has to take place. The development of risk-reducing technologies such as IPM
is now the preferred approach in pest management worldwide. Although farmer pesticide
use in Africa is relatively low compared to Asia, there are signs of misuse that require
urgent solutions. In Asia, heavy pesticide use in food crops especially rice has triggered
widespread farmer health problems (Antle and Pingali, 1994). However, pesticide use
among smallholders in Africa is associated with cash crop cultivation especially cotton
and tobacco (Sukume, 1999). A key question therefore is how can Africa’s export crop
production avoid the errors of Asia’s pesticide misuse. Developed countries have laws
and regulations to limit the negative effects of pesticides yet comparable systems of laws
and surveillance have been established only recently in LDCs (Frank, 1996).

Given that IPM is viewed as a more effective pest management option, the next
question is: How best can it be implemented under smallholder cotton production systems
in Africa? The IPM approach has been well received in Asia, Indonesia and Philippines
in particular, and it is in Africa’s best interest to draw useful lessons from Asia’s success
with F FS-IPM. The opportunity cost of not adopting IPM is relatively high in LDCs were
most farmers using toxic pesticides have the additional burden of being illiterate and lack
protective clothing (Kiss and Meerman, 1993). Despite the fact that IPM is widely
recommended, it is still less widely used in LDCs (CAB International et al.,1991). For

instance, pesticides remain the dominant pest management tactic in most African

 

3IPPM combines 1PM approaches to manage pests and improve crop production management under mixed
farming systems in rural areas of Zimbabwe; it aims to increase crop productivity through interventions in

72

countries even though majority farmers cannot afford pesticides (Ajayi, 1999; Kiss,
1995). Currently, there is little information about actual adoption of IPM in smallholder
agricultural production in Africa.

Empirical evidence from Asia where IPM has been well received shows that
pesticide use can have negative effects on farmer health causing reductions in farmer
productivity (Antle and Pingali, 1994). Assessment of the Indonesia National IPM
program and Philippine IPM for rice farmers reveals that IPM is a successful framework
for alleviating pest problems leading to higher crop returns and a reduction of both
environmental liabilities and human health risks associated with intensive use of agro-
chemicals (Rola and Pingali,l993; Cuyno, 1999; World Bank, 1997). In India, pest
suppression was found to be more efficient in bio-control-based IPM with consequent
increase in cotton yields of up to 33% compared to farmers’ practices of plant protection
(Rajendran and Bambawale, 1994). However, a slow down in IPM adoption in the
Philippines has been attributed to the fact that its benefits are not apparent in the short-
run (Rola, undated).

In contrast, empirical evidence on the adoption of IPM in Africa is scant (Jowa,
1993; F oti, 1999). In Kenya, F FS has empowered local farmers to make more efficient
crop management decisions that include assessing crop health and natural enemy activity
prior to applying pesticide treatment (Loevinsohn, et al., 1998). The strength of the
discovery-based, experimental group-learning model relative to the traditional ‘top-
down’ pest control recommendations is that it takes into account important crop

interactions and prevailing field conditions. The F FS approach is now considered the

 

both pests and production management.

73

standard procedure to implement IPM in Asia and is slowly spreading to Latin America
and Africa (Fleischer et al., 1999).

Zimbabwe cotton offers a useful test case for determinants of IPM adoption
among smallholders with and without exposure to comprehensive extension training.
Zimbabwe cotton growers make intensive use of pesticides to control major pests such as
aphids (aphis), heliothis bollworm (helicoverpa), termites, stainers (dysdercus) and red
spider mites (tetranychus). Cotton IPM-F F S was initiated among smallholders in the
Sanyati district of the Midlands Province in north central Zimbabwe during 1997 with
help from F AO’s IPM Global Facility. By 1999, two classes of farmers had graduated
from FFSs with IPPM training in cotton production. This early stage of IPPM awareness
offers a timely opportunity to analyze IPM adoption determinants among Sanyati cotton

farmers, including the technology awareness effect embodied in FFS training.

3.2 Study Objectives

The main purpose of this study is to determine the factors that influence the
adoption of IPM practices in smallholder cotton production in Zimbabwe, and to explore
the resulting policy implications. Knowledge of the key factors driving the adoption of
IPM will facilitate policy formulation, program planning and targeting, and diagnosing
constraints in existing methods of IPM dissemination. Therefore, the study addresses a
serious challenge facing researchers, extension workers and policy makers involved in
the development and implementation of an appropriate IPM strategy for smallholder
mixed-cropping systems in Africa. Results also provide insights into the prospects for
widespread implementation of IPM in Africa. We hypothesize that pesticide-related

health risks positively influence IPM adoption. The remainder of the paper will be

74

organized in the following way. First, the evolution and adoption of IPM in LDCS are
highlighted. Second, we develop a working definition of IPM for Zimbabwe cotton.
Third, we present an economic behavioral model for IPM adoption followed by
specification of the empirical model. Next, results of the econometric estimation are
presented and discussed. The final section summarizes the paper and discusses key policy

implications.

3.3 The Diffusion of IPM Technologies in Less Developed Countries (LDCs)

In a few countries where IPM has been introduced in Africa (e. g. cotton in Sudan
Uganda, and Zimbabwe is now in IPM mode even if only recently), implementation
weaknesses in some instances have been associated with farmer’s inability to recognize
both key pests and beneficial insects a key dimension of IPM use. Also in some cases,
farmers are unable to distinguish between stress caused by water deficiency and high
temperatures relative to that arising from disease and insect damage. However, the impact
of factors that constrain early phases of diffusion processes tends to differ and decline as
the technology reaches final stage of the diffusion process (F eder and Umali, 1993).

One of the essential aspects of IPM diffusion is the integration of technical and
social knowledge (World Bank, 1997). In particular, knowledge about specific pests as
well as location specific farm management systems is critical for the successful design
and dissemination of IPM approaches. Some major limiting factors to the successful
implementation of IPM-related technologies are lack of farmer-focused research and the
availability of effective and competitive alternative non-chemical techniques (World

Bank, 1997). In many countries, imbalances exist between IPM dissemination and

75

extension curriculum that emphasizes chemical control at the expense of non-chemical
options (Pincus et al., 1997).

Apart from Asia, there is also growing evidence of successful development and
use of IPM in South America (soybeans in Brazil) (Gallagher, 1988). However, evidence
from Ecuador highlights the fact that farmers are at risk of excessive exposure due to
widespread ignorance of pesticide poisoning symptoms and lack of personal protective
equipment (Crissman et al., 1994). Therefore, one of the leading concerns of pesticide
use in LDCs is that farmer’s health is seriously compromised by unsafe application
practices (Rola and Pingali, 1993; Tjomhom et al., 1997). The problems of pest
resistance, pest resurgence and emergence of secondary pests in Africa further justify the
need for IPM diffusion Kiss (1995).

In Africa, rice IPM pilot programs based on F F S-concept were launched in
Ghana, Mali, Cote D’Ivoire and Burkina Faso in1994. Over the past five years, IPM-
FFSs have expanded to Sudan, East and Central, and Southern Africa regions.
Increasingly, the IPM approach has become popular with both African governments and
development agencies interested in broader issues of integrated crop and pest
management, and various versions of IPM have been tried in the different countries
(Gallagher, 1998).

Despite successes in a few countries, widespread implementation of IPM is still
an elusive goal in most parts of the world. The momentum for the diffusion of improved
technology such as IPM is slowed by policies that discriminate against agriculture in
many countries (Birkhaeuser et al., 1991). Past experience shows that immediate and

uniform adoption of agricultural innovations is very rare. In addition, technology

76

adoption and diffusion differs across socio-economic groups and over time (F eder et al.,
1982). In Africa, the use of Economic Threshold Levels (ETL)18 is still underdeveloped
and requires refinement (Kiss and Meerman, 1993). Also, IPM technologies oriented
toward single pests pose serious weaknesses as the challenge lies with development of
ETL for several pests (Rola and Pingali, 1993). Outbreak of secondary pests in Africa

makes this approach imperative for successful cotton IPM adoption.
3.4 Current Status of IPM Adoption in Less Developed Countries (LDCS)

Experiences from LDCS suggest that successful adoption of IPM on a wide scale
requires the following key elements;1) establishing an enabling environment for IPM by
eradicating policies in support of environmentally unsustainable pest management and
strengthening regulatory institutions, and 2) targeted support for measures that promote
the uptake of IPM, such as, public awareness, research, extension and training with an
emphasis on decentralized farmer centered initiatives (World Bank, 1997). The desired
broad constituency in favor of IPM adoption can be achieved through a clear definition of
institutional roles and responsibilities of pest management stakeholders. Also, the World
Bank advocates the adoption of a national IPM strategy as crucial for enlisting the
necessary commitment to IPM adoption. Such a strategy can secure broad institutional
support by addressing both upstream policy elements and on-farm IPM uptake. The

introduction of a national IPM strategy has been adopted relatively easily in countries

 

'8 Economic Threshold Level is the breakeven point at which the dollar value for an increment of loss in
yield quantity or quality is equal to the cost of a control method that successfully eliminates pest damage
and yield loss (Kiss and Meerman, 1993).

77

where research evidence has proved that pesticides are not increasing yields significantly
(World Bank, 1997).

A critical constraint to IPM adoption in Sub-Saharan Africa (SSA) is the
shortage of low-cost IPM technologies that are relevant to the mixed farming systems
prevalent on the continent. Besides, encouraging a broad base of farmers to experiment
with new practices remains a challenge (World Bank, 1997). IPM adoption relies on
farmer-to-farmer diffusion, yet knowledge diffusion by graduates is gender biased as men
diffuse to men and women to women. Similarly, an age bias among graduates of FFS in
SSA has been reported as IPM adoption by older farmers dominates younger farmers
(Loevinsohn, et al., 1998). Inadequate interaction between researchers, extension workers
and farmers has inhibited local understanding and adoption of the IPM technologies that
are being introduced in Africa (Gallagher, 1998).

In Asia, evidence from Philippines shows that farmers have misconceptions about
pests and natural enemies; with leaf eaters generally considered as most important pests.
Mismatch between pest damage and responsible pests and confusion between rice and
vegetable pests seemed common among farmers. Further, additional IPM adoption hurdle
in Asia has been the widespread lack of knowledge about pest resurgence and action
thresholds among rice and vegetable farmers (Lazaro et al., 1995). According to Rola and
Pingali, (1993), biological control tended to receive less attention in most IPM activities
in Asia. Similar deficiencies were identified in cotton-IPM adoption studies in Uganda
where farmers failed to recognize some species of insects as beneficiaries (Kiss, 1995).

Empirical evidence on whether multi-component technology like IPM is adopted

individually or in package has been mixed and it still requires further research (Feder and

78

Umali, 1993). Evidence of stepwise adoption patterns of agro-chemical technological
components has been reported in the literature (Byerlee and Hesse de Polanco, 1986).
Conversely, the sequential adoption hypothesis was later challenged in a study of maize
production in Swaziland where farmers were reported to adopt technologies in clusters
(Rauniyar and Goode (1992) cited in Feder and Umali, 1993). A later study indicates that
the adoption decision is inherently multivariate and univariate modeling excludes
valuable economic information (Dorfman, 1996). Since, uncertainty about productive
performance of a technological package decreases with experience, while confidence
increases with positive experience, usually early adopters choose to adopt only parts of a
package rather than a complete package (Feder and Umali, 1993). Generalizing adoption
patterns is difficult due to differences in technology adOption arising from diverse agro-

climatic regions and farmer’s socio-economic conditions.

3.5 Defining Smallholder Cotton-IPM Adoption

The successful assessment of any IPM strategy begins with a clear definition of
what is being assessed. Typically, IPM involves a number of pest management practices
that are both location and crop specific. There is no consensus in the literature as to what
specific pest management practices constitute IPM. IPM definitions have been classified
as either “input-oriented” or “output-oriented” (vainton and Williams, 1998). The latter
focus on desired outcomes such as profitability, human health and environmental quality
while the former relate to specific IPM practices. Assuming an input-oriented approach,
pest management practices can be grouped together and IPM defined as low, medium and
high level (Vandeman, 1994; Mullen et al., 1997). Other studies have assigned points to

different practices and defined adoption along a scale (Hollings-worth et a1 cited in

79

Swinton and Williams, 1998). Yet others have considered both the proportion of practices
and the degree of economic importance of the pest. In our study we use the “input-
oriented approach” and focus on the actual number of IPM practices. We characterize the
cotton growers in terms of how many IPM practices have been adopted.

For the purpose of this study, the specific cotton IPM and production practices
examined include; (1) alternating pesticides to slow development of pest resistance, (2)
use of less toxic and safer chemicals, (3) adjusting pesticide application frequency and
timing, (4) pest scouting (5) adjusting planting dates, (6) use of beneficial insects in pest
management, and cultural practices such as (7) crop rotation, (8) legally enforced closed
season (or field sanitation) to stop pest carry-over, and (9) use of trap crops. The potential
range of adoption was from 0 to 9. However, we did not ask the farmers to rank the

relative importance of each IPM practice.

3.6 Methodology and Data
3.6.1 Economic Behavioral Model

Typically, individual households are the primary decision makers concerning
agricultural innovations, implying that a household behavioral model is key to
understanding the adoption-diffusion process (Feder et al., 1993). Assume the model of
an individual. household producing multiple crop outputs using multiple inputs that

include pesticides. The household maximizes a utility-function U(7r) that is increasing in

net returns (7:) subject to constraints from fixed factors.
Several assumptions are made in specifying the model. First, we assume that
farmers consider health costs as cost of production. This implies that farmers do care

about both economic and pesticide-related health problems associated with the use of

80

agro-chemicals. Also, agrochemical exposure is assumed to reduce health status of the
farmer. Second, cotton production and management decisions can be described as static
profit maximization or cost minimization. Third, farmers are sensitive to downside yield
risk. Fourth, family and hired labor are homogenous and are considered as perfect
substitutes when used in cotton production. The labor market is competitive and the
returns to farm work and off-farm work are equilibrated. Finally, we also assume that
agro-chemicals contribute to cotton productivity only indirectly via reduction in the
population of pests, which are considered in our case as the damage agents.

In that respect, smallholder cotton yields are an indirect function of pesticides
applied since production functions that treat pesticides as yield increasing inputs over-
estimate marginal productivity. Lichtenberg and Zilberrnan (1986) were among the first
to point out that pesticides should be modeled as damage control inputs just like
sprinklers for frost protection. Suppose that the actual cotton yield (19 is given by:

(I) Y = Y ”(Iowa-ma)» and Y”=f(p ,, p ., K.L,I.Z)
where the potential pest-free cotton yield Y0 is a function of cotton price p y, prices
p x for all variable inputs including pesticides, labor, fertilizer, seeds, and credit; K is
fixed physical capital such as land, L is labor, I is pest management information and Z
represents conditioning factors such as soil type, rainfall, farmer’s education, gender,
experience and managerial capacity. But the actual yield Y, depends on pest damage and
its abatement. Therefore, D(.) represents the pest damage function"), N is the pest
pressure and X p is the pesticide or damage control agent purchased at price p p. Pesticide

efficacy range is such that (0<k(X p)<1) where k(X”) describes the “kill function”. When

 

'9 Damage function expresses the relationship between pest pressure and yield loss; it varies with presence
of different pests and the abundance of natural enemy species that feed on pests (Rola and Pingali, 1993).

81

a chemical is completely effective, that is (k(X p)=1), then Y = Y 0. We also assume that
profits are affected by the level of non-pesticide inputs X” and health services H ‘
purchased at p0 and p}. respectively. Following from the work of Antle et al. (1994) and
Swinton (1998), we specify the relevant smallholder maximization problem as follows:

(2) Maxn(Y,X,H, I) =PyY(pypx,K, L, I,Z)-prp—p0X"-p;,HS

s4 (0 Y=f(X".X ‘9- D(N+s)(1- karma:
(ii) EXP=Pa(Q .- s .9 - w (L‘-Lf) + R
(iii) L‘_< Lf+L,, + L,
( iv)I,= Y’(Io, FFS, Age, Education, Experience, V)
(v)H=¢(A”H‘,I,X”,X”)

where (i) cotton output (10 is increasing in non-pesticide inputs X”, D(.) is a concave and
increasing in pest population (N), but N is reduced by concave “kill function” k(.) which
is increasing in pest management input X”. Both cotton yield and pest population are
stochastic; Y+,u where p ~N(0, 0,, 2) and N+e where 8~N(0 a, 2). Pesticides and non-
chemical production inputs are distinguished by variables X p and X a respectively. In
equation (ii) EXP is expenditure on non-agricultural products, Pa Q a is cash receipts
from agriculture, PaS c is value of household consumption of self-produced agricultural
staple, R is remittances from relatives and w is wage rate, (iii) L" is total effective labor
requirement, L1. is total hired labor input, LfIS family labor and L, is shared labor from
the community. (iv) 1, refers to farmer’s pest information knowledge, 1,, represents
farmer’s initial level of pest management information before exposure to IPM training,
FFS refers to participation in FFS-based IPM training. I, is also affected by among others

farmer’s personal and village level characteristics (IO. (v) H is a measure of farmer’s

82

health endowment and A p is pesticide-averting behavior. Beside human health, pesticide
use X p also influences environmental quality. However, the data in this study does not
provide sufficient farm-level variation to identify this effect, so it is not included in the
presentation of the economic behavioral model.

Solving the input and output choice problem, we can derive a factor demand
functions for X p , which is stated as follows:

(3)X”=g (Py Pg, Ph,H,L, I, K, Z)

In particular, the demand for IPM practices X p will depend on farmer characteristics,
available farm resources (L, K), biophysical characteristics of the farm setting (Z), pest
pressure N, institutional and relative prices (P), health effects of pesticides (H) and IPM
awareness (I). The specific empirical measures of these attributes are presented in Table
1 and discussed below. It is hypothesized that exposure to IPPM training through F FS
will do the following; (i) improve farmer’s cotton pest management knowledge, (ii) raise
cotton yields, (iii) lower pesticide use, (iv) improve farmer’s health status and (v) raise
farm profitability (Waibel, et a1. 1998). The expected outcomes can be summarized as
follows:(i)bI/ 6FFS 2 0, (ii)a”Y/ 0”] 2 0 ( iii)@("/ 0"] _< 0, (iv)a’H/ d! 2 0, (v)0”7t/ a 2 0.
Adoption analysis of cotton-IPM, a recently introduced technology in Zimbabwe, will
provide insights about the important factors driving the uptake of IPM by both the current
and future adopters who will ultimately accept the technology. The inclusion of farmers
with different years of experience with FFS-based IPPM can form the basis to explore the
adoption process itself. Such information will provide more timely strategic adjustments

in future IPPM implementation.

83

3.6.2 Data and Estimation

The empirical model of IPPM adoption among smallholder cotton growers is
estimated using cross-sectional data obtained from a survey conducted through personal
interviews in Sanyati, one of the leading cotton growing districts in Zimbabwe. Sanyati
was one of the first districts to offer FFS-IPPM training to local cotton farmers in 1997.
The district is located in the north central part of the country in the Midlands province. It
lies in natural regions III and IVZO. Farmers in Sanyati have a mean cotton growing
experience of 14 years. Survey farmers in Sanyati were identified using a stratified
random sampling approach on the basis of villages with FF S groups. The second level of
stratification was based on the cotton farmer participation in F FS-based IPPM training
groups. A total of 141 farmers were interviewed in Sanyati.

Data used in the analysis was collected at two different levels; household and
field. The unit of observation was the household. Variables used in the empirical model
are based on previous findings from economic theory of adoption decisions (Feder, Just
and Zilberman; D’Souza; Dorfrnan). In addition, univariate adoption models for
individual IPM practices provided a further basis for selecting the final variables used in
the model. Farmer characteristics that condition adoption behavior include farmer’s age
HHAGE, extension meetings attended COTEXTMTG, cotton growing experience
COTYEARS, level of formal education EDUYEARS, gender of the head of household
HGENDER and whether certified as a Master Farmer MASTERFM. We expect farmer
characteristics to positively influence IPM adoption although gender effect cannot be

established a priori. Also, the age effect can be ambiguous a priori. As age increases

 

20 A natural region is an agro-ecological zone demarcated by rainfall pattern, as well as crop and livestock
production potential in the region. Mean annual rainfall for NRs Ill and IV are 800mm and 400 mm.

84

hence experience, the time horizon to reap IPM benefits decreases yet experience could
lead to better knowledge of IPM and its benefits (Khanna, 2001).

Farm resource endowment variables include total cotton labor LABDAYS, cotton
land area cultivated COTAREA, value of productive assets PROASSESTS, ownership of
draft animals DRAFTOWN, use of credit CREDIT and access to formal off-farm
employment FOMEMPLT. Farm management practices that could influence pest
management include use of improved cotton variety ALBARFQ902, production of staple
maize crop ALTCROPM, whether the cotton field was fallowed the previous season
FALLOW, absence of specific three-year cotton rotation program NROTPROG and number
of tillage practices used TILPRACS in cotton production. Institutional and relative price
factors are access to information media MEDIA, average walking time from homestead to
cotton fields AVEWALKT and distance to cotton markets DISTMKT.

The health risk variables used to estimate the empirical model are number of
pesticide-related acute symptoms ACUTESYM, number of individual protective clothing
units used by the farmer in making pesticides treatments SAFINDEX, measure of farmers
ability to rank the toxicity level of pesticides based on color codes on pesticide container
labels LABELIT and whether or not the head of the household drinks alcohol HHDRINK.
The last category of the determinants of IPM adoption is the awareness and perception
variables. IPMAWARE is a measure of post FFS-IPM training years judging from the time
when the farmer enrolled. Farmer’s perceptions of downside yield risk YLDRISK and
current chemical-based pest management strategies PMGTVIEW were also used in the

regression model. All variables and their definitions are listed in Table 3.1.

85

3.6.3 Empirical Model

In this study, we use a Poisson maximum likelihood regression model to predict
the discrete but non-categorical pest management strategies used by cotton growers in
Zimbabwe. The shortcomings of using OLS, ordinary probit and or logit analysis in the
case of count data are highlighted in the econometric literature (Greene, 1997, Madalla,
1983). The number of additional pest management practices used on a given crop
indicates the farmer’s reliance on multiple biological and cultural pest management, a
key ingredient of IPM use (V andeman et a1 , 1994). Since an integrated package of
cotton-IPM as an off-the-shelf system does not yet exist, farmers have the flexibility to
combine different counts of practices that address specific pest complex in their fields.
The predicted values Y 1 Y 2,...,Y ,, are assumed to have independent Poisson distribution
with parameters A, , 42,-", 21 n respectively (Madalla, 1983). According to Greene, 1997,
the basic equation for the Poisson regression is represented as follows:

(2) Prob(Y; =y.-) =(e 4“ ,1,- ) / y.- where yi =0, 1, 2,.....

The parameter 115 is assumed to be log-linearly related to regressors xi. Therefore,

(3) Ln (41:): 0’1:

The log-likelihood function is given by :
(4) Ln L =Zi=1,...,n (4» i +y 1031’ i - In yr!)
The expected number of events per period is given by;

(5) E(y ,~/x ,-)=Var (yi/x ,-)= r,- = e 4’1”“

Based on the conceptual framework presented the empirical model is estimated
using the following groups of explanatory variables; (1) farmer characteristics (F C), (2)

farm resource endowment (FR), (3) farm management practice (F P), (4) pest damage

86

(PD), (5) institutional and relative prices (IP), (6) health risk (HR), and (7) awareness
and perception variables (AP). Therefore the general form of the empirical model

estimated is stated mathematically as follows;

(6) ZIPM Practice ,- = 6 ( F C, FR, FP, PD, IP,HR, AP) +e ,-

3.7 Results and Discussion

3.7.1 Farmers’ Adoption Patterns and Pest Management Perspectives

The majority of farmers (30%) use at least three IPM practices identified earlier.
The mean number of IPM practices used is 4.36. About 11% reported using as many as
seven IPM practices in 1998/99. All the farmers reported using at least two IPM—related
strategies. The survey contained about 48% of the farmers who had some exposure to
F FS-IPPM training. The F FS-IPPM graduates were more likely to be male farmers with
high school education cultivating more than one cotton field. A cluster analysis of the
different IPM practices did not reveal any discernible pattern in terms of adoption of IPM
practices in clusters. The leading IPM practices adopted were field sanitation (97%), crop
rotations (87%), scouting (89%), alternating pesticides (32%) and preservation of
beneficial insects (30%). Correlation analysis suggests practices were adopted
independently. Farmers in Sanyati applied an average of three pesticide treatments on
each plot in 1998/99. This suggests either a low level of pest infestation that season or a
shift towards more efficient non-chemical pest suppression methods.

An important finding is that cotton growers expressed diverse views about
motivations for pesticide interventions and use of IPM. Among those exposed to IPM

training, 85% believed that IPM knowledge is an effective pest management tool and 8%

87

felt the opposite. Farmers satisfied with IPM methods constitute a critical mass of the
early adopters. The rest were either not sure or had no opinion about IPM. Including the
non-FF S graduates, the beliefs about IPM were that 40% felt it was effective while 54%
had no idea. The rest felt IPM was ineffective, labor intensive and inferior to the
traditional calendar-based chemical control of pests.

Chemical interventions were made for different reasons with 30% of the
respondents stating that they spray on fixed calendar basis, while 66% said they applied
chemicals only after pest scouting. Other reasons cited for guiding chemical interventions
were specific growth stages for the cotton plant. Only 14% of the farmers felt that the
major problem in cotton production in Sanyati was pest management. A majority felt that
poor prices (53%) were most important, and some felt drought (6%) was a serious

problem needing urgent attention.

3.7. 2 Factors Affecting Cotton-IPM Adoption Among Smallholder Farmers

Poisson regression results on the determinants of adoption of aggregate IPM
practices in Sanyati District are summarized in table 3.2.The analysis was carried out
using STATA version 6.0 (STATA, 1997). The Poisson regression model was significant
at the 1% level. However, much variability in IPPM practices adopted was not explained
by the model, which had a McFadden R2 of only 10.2%.

The key result of our analysis is the technology awareness coefficient that is
significant at the 1% level. Farmers exposed to cotton-IPM techniques through FF S
training, are more likely to use several IPM related practices in cotton pest management.
The uniqueness of IPM as a knowledge intensive technology makes the human element

an overriding factor in its adoption decisions. Training augments skills and knowledge

88

useful in making innovative decisions. Further it enhances farmer’s ability to distinguish
technologies that generate opportunities for economic gain from those that do not
(Wozniak, 184). The coefficient for the total area cultivated to cotton is positive and
significant at the 10% level. This implies a scale-effect in the use of IPM technology in
Sanyati. The results suggest that farmers with smaller acreage under cotton are likely to
require special adoption incentives such as improved access to technical information to
encourage them to use more IPM practices in reducing cotton pest damage. This result
matches other several studies that identify farm size as the most prominent variable
explaining adoption decisions (Wozniak, 1984; Dorman, 1996)

The coefficient for the cotton variety ALBARFQ902 is positive and significant at
10%. This implies that farmers growing this HYV are more inclined to use several IPM
practices. The unique characteristics of this variety are that it is resistant to bacteria blight
and jassids (empoasca), and is more tolerant to aphids and drought. Although it is
susceptible to verticillium wilt (verticillium dahliae), ALBARFQ902 has the highest score
on pest and disease resistance among all the cotton varieties grown in the middleveld in
Zimbabwe.

Contrary to expectation, the pesticide-related health risk variables ACUTESYM
SAFINDEX, LABELIT and HHDRINK are insignificant. The lack of statistical significance
associated with health risk variables does not support the hypothesis that IPM adoption
decisions are based on pesticide-related health risks. Analysis of IPM adoption decisions
of label literate farmers only did not reveal significant relationship between health risks
and IPM adoption. Lack of significance could be attributed to the fact that more than 60%

of the IPM practices assessed did not use pesticides. Although farmer characteristics

89

estimated from adoption models are important in the policy arena our results also show
such variables as insignificant in the adoption of multiple IPM practices. What we may
have demonstrated is several variables that can be excluded from consideration when

targeting likely cotton IPM adopters under smallholder conditions.

3.8 Conclusion

The main conclusion is that technology awareness embodied in access to F F S-
based IPM training is really important in motivating the use of multiple components of
risk-reducing technologies such as IPM. Investment in IPM farmer education and literacy
programs targeted to non-adopters will have long-term beneficial impacts on cotton-IPM
use. IPM use requires an experimental cotton plot and farmers with more land are more
likely to adopt IPM practices. In addition, cotton growers who planted the pest resistant
variety ALBARFQ902 were likely to adopt more IPM practices. Experience with pesticide-
related health problems did not significantly affect IPM adoption, suggesting there is still
a greater need to sensitize farmers about the health risks of using pesticides. It may be
that using IPM does not significantly reduce these risks hence the link between pesticides

and human health should be further explored in the context of Zimbabwean smallholders.

3.8.1 Policy Implications and Suggestion for Future Research

Analysis of IPM practices being adopted, the characteristics of smallholders that
are adopting and the factors motivating adoption of FFS-IPM is still fragmentary in
Africa (Adesina, 1994). Our results indicate that diffusion factors such F F S-IPM training
and farmer-to-farmer extension delivery approaches will play a critical role in the

delivery and adoption of IPM. Success of IPM adoption will depend on farmer’s

9O

knowledge and awareness of the technology. Further, rapid adoption could occur if
farmers complement adoption of IPM practices with a conscious choice on HYV varieties
that confer pest and disease resistance qualities such as ALBARFQ902. Both economists
and policy makers involved in crafting incentives for widespread diffusion and adoption
of F FS-based cotton IPM need to target not only farmers with large cotton acreages but
also smaller farmers when planning future IPM programs in Africa. In addition,
information about type of farmers most likely to adopt IPM technology and the extent of
its adoption is expected to guide agro-chemical firms in their future new product
development and marketing strategies given that one of the dimensions of IPM is to
emphasizes the use of safer and less toxic products. The pesticide-related health effects
require fiirther research to determine their costliness and also their impact on the net
benefits of pesticides use under smallholder conditions. As farmers become aware of the
health risk costs, it is expected that they will pay more attention to the health factors in
IPM adoption decisions. Also, gaps exist in terms of understanding the sequence
followed by smallholders in adopting individual IPM practices during its diffusion cycle.
Future research must therefore determine the risk-reducing role of individual IPM
practices and the bundling strategies used by smallholder cotton growers in adopting

compatible combinations of different IPM techniques.

91

Table 3.1:Descriptive statistics on variables used in Poisson Regression Model

 

 

Variable Name Definition Units Mean Standard
Deviation
Dependent Variable
Number of IPM practices Count of IPM practices count 4.36 1.62
Farmer’s Characteristics
HHAGE Farmer’s age years 46.25 14.26
COTYEARS Cotton growing experience years 14.21 10.37
COTEXMTG Number of extension meetings number 4.64 6.36
EDUCYEARS Number of years in formal education years 6.57 3.72
HGENDER Head of household’s gender (0,1) 0.83 -
MASTERFM Certified Master Farmer (0,1) 0.26 -
Farm Resource Endowment
LABDAYS Total labor used in production man-day 80.93 43.06
COTAREA Land area cultivated to cotton (Ha) 4.55 3.97
PROASSET Value of productive assets (ZS) 9,506.41 8,298.98
CREDIT Farmer used credit (0.1) 0.18 -
DRAFTOWN Farmer owns draft power (0,1) 0.69 -
FOMEMPLT Farmer is in formal employment (0,1) 0.47 -
Farm Management Practices
ALTCROPM Maize is major alternative crop (0,1) 0.08 -
ALBFQ902 Cotton variety ALBARFQ902 grown (0,1) 0.81 -
FALLOW Field was fallowed previous year (0.1) 0.42 -
NROTPROG No specific crop rotation program (0,1) 0.13 -
TILPRACS Number oftillage practices number 1.43 0.87
Pest Damage Variable
PSTPRESS ' Pest pressure (scale 0-1) index 0.46 0.61
Institutional and Relative Prices
MEDIA Farmer has access to information media(0,1) 0.67 -
AVEWALKT Average walking time to cotton fields minutes 12.97 15.78
DISTMKT Distance to markets (km) 13.49 7.69
Pesticide-Related Health Risks
ACUTESYM Pesticide-related acute symptoms number 1.13 0.84
SAFINDEX Count of protective clothing count 3.75 1.53
LABELIT Count of correct label interpretation count 2.16 1.26
HHDRINK Farmer drinks alcohol (0.1) 0.48 -
Technology Awareness and Perception .
IPMAWARE 22 Farmer’s experience in FF S-IPM years 0.83 0.93
YLDRISK23 Downside yield risk perception index 2.61 2.82
PMGTVIEW Maintain calendar-based methods (0,1) 0.65 - -

 

 

2' Pest Pressure Index =2 (pest pressure)/39. The pest pressure indicators are 0=None, l= Light 2=
Average and 3=Severe. The pest pressure is assessed for 13 different cotton pests where a count of 39
represents severe cases for all cotton pests.

22 IPMAWARE is measured as post FFS-1PM training years; O=no 1PM training, 1: 1998/99 FFS graduate
and 2=1997/98 F F S graduate.
23 Downside yield risk=(YM-YL)2 where YM represents mean cotton yield and YL is the perceived lowest
cotton yield from the main cotton field during a poor season. The assumption is that farmers care more

about downside yield risk.

92

Table 3.2: Determinants of Cotton IPM Practice Adoption in Sanyati District, 1998/99

 

 

 

Variable Coefficient Standard Error Z-value
Farmer Characteristics

HHAGE -0.0002 0.0044 -0.0460
COTEXMTG 0.0072 0.0074 0.9700
COTYEARS 0.0008 0.0058 0. 1400
EDUYEARS 0.0100 0.0140 0.7500
HGENDER 0.1200 0.1400 0.8400
MASTERFM -0.1 100 0.1 100 -0.9700
Farm Resource Endowment

LABDAYS -0.0003 0.0013 -0.2300
COTAREA *0.0230 0.0120 0.0190
PROASSET 0.0032E—5 0.0064E-5 0.5000
CREDIT 0.0430 0.1300 0.3400
DRAFT OWN -0. 1700 0.1200 -0.1400
FORMEMPLT -0.03 80 0.1000 -0.3600
Farm Management Practices

ALTCROPM -0.1200 0.2000 -0.6100
ALBFQ902 *0.2600 0.1400 0.0190
FFALLOW 0.1100 0.0960 0.0100
NROTPROG -0.05 80 0.1500 -0.3 800
TILPRACS 0.0330 0.0710 0.4700
Pest Damage

PSTPRESS 0.0030 0.0730 0.0410
Institutional and Environmental

MEDIA -0.0540 0.1 100 -0.4900
AVEWALK 0.0006 0.0030 -0. 1900
DISTMKT 0.0066 0.0081 0.8200
Pesticide-Related Health Risks

ACUTESYM 0.0200 0.0650 0.3100
SAF INDEX 0.0340 0.0340 0.9900
LABELIT -0.0150 0.0390 --0.3800
HHDRINK -0.1000 0.1 100 -0.9500
Technology Awareness and Perception

IPMAWARE ***0.1900 0.0670 0.0290
YLDRISK -0.0080 0.0170 —0.4600
PMGTVIEW -0.0370 0.1000 -0.3500
N 136

McFadden R-squared 0.1022

Log Likelihood Ratio 53.32

x2 —p value 0.0027

 

*=significance at 10%, "=significance at 5%, "*=significance at 1%

Source: Field survey, 1999

93

References

Adesina, A., A. “Pesticide Use and Policy in West Africa: An Overview.” Proceedings
of the Gottingen Workshop on Pesticide Policies. Gottingen, Germany.S. Agne,
G. Fleischer and H. Waibel, eds. 28 February-4 March, 1994.

Ajayi, 0. “Measuring the Indirect Health Benefits of IPM: Methodology for Estimating
Pesticide Health Costs in Africa.” International workshop an ”Evaluation of IPM
Programs: Towards a Framework for Economic Evaluation. " Garbsen Hannover

Germany 25-27 May, 1999.

Antle J .M., D.C. Cole, and CC. Crissman. “The Role of Pesticides in Farm Productivity
and Farmer Health.”Ec0nomic, Environmental, and Health T rade-ofls in
Agriculture: Pesticides and the Sustainability of the Andean Potato Production.
Charles C. Crissman, John M. Antle and Susan M. Capalbo, eds., Kluwer
Academic Publishers. 1998.

Antle J .M., and PL. Pingali. “Pesticides, Productivity, and Farmer Health: A Philippine
Case Study.” American Journal of A gricultural Economics. 76(1994):418-29.

Besley, T., and Anne Case. “Modeling Technology Adoption in Developing Countries.”
Journal of American Economic Review. 83(1993):396-402.

Birkhaeuser, D., R.E. Evenson, and G. F eder. “The Economic Impact of Agricultural
Extension: A Review.” Economic Development and Cultural Change.

39(1991):607-50.

Burrows, M.T. “Pesticide Demand and Integrated Pest Management: A Limited
Dependent Variable Analysis.” American Journal of Agricultural Economics.
65(1983):806-10.

Byerlee, D., and EH. de Polanco. “Farmers’ Step-wise Adoption of Technological
Packages: Evidence from the Mexican Altiplano.” American Journal of
Agricultural Economics Association. 68(1986):5 19-27.

Cole, D.C., F. Carpio, J.A. Julian, and N Leon. “Health Impacts of Pesticide Use in
Carchi Farm Populations. Economic, and Environmental, Health Tradeofls in
Agriculture: Pesticides and the Sustainability of A ndean Potato Production.

94

Charles C. Crissman , John, M. Antle and Susan M. Capalbo, eds., Kluwer
Academic Publishers. 1998.

Cuyno, L.C.M. “An Economic Evaluation of the Health and Environmental Benefits of
the IPM Program (IPM CRSP) in the Philippines.” Ph.D Dissertation, Virginia
Polytechnic Institute and State University, Blacksburg, Virginia. 1999.

Crissman, C.C., D.C.Cole, and F. Carpio. “Pesticide Use and Farm Worker Health in
Ecuadorian Potato Production.” American Journal of Agricultural Economics.
76(1994):593-97.

Dorfman, J .H. “Modelling Multiple Adoption Decisions in a Joint Framework.”
American Journal of A gricultural Economics. 78( 1 996):547-5 7.

D’Souza, G., D. Cyphers, and T. Phipps. “Factors Affecting the Adoption of
Sustainable Agricultural Practices.”Agricultural and Resource Economics Review.
22 (1993):]59-65.

F eder G., and J.Quizon. Integrated Pest Management in Asia: Are There Real
Returns to IPM and its Diffusion? Evaluation of IPM Programs-Concepts and
Methodologies. Pesticide Policy Project in Cooperation with F AO. Hannover,
March 16-18, 1998.

F eder, G., R.E.Just, and D. Zilberman. “Adoption of Agricultural Innovations in

Developing Countries: A Survey.” Economic Development and Cultural Change.
33(1985):255-98.

Feder, G., and D. L. Umali. “The Adoption of Agricultural Innovations: A Review.”
Technological Forecasting and Social Change. 43(1993):2 1 5-3 9.

Fernandez-Cornejo, J. “Environmental and Economic Consequences of Technology
Adoption: IPM in Viticulture.” Journal of Agricultural Economics. 18(1998): 145-
55.

Fleischer, G., F. Jungbluth, H. Waibel, and J.C. Zadoks. A Field Practitioner ’s Guide to
Economic Evaluation of IPM. Pesticide Policy Project in Cooperation with Food
and Agriculture Organization. University of Hannover, Germany, Publication
Series No.9.1999.

95

Foti, R. “Comparative Economics of Integrated Pest Management and Pesticide Use for
Smallholder Cotton Farmers in Gokwe.” MSc Thesis, University of Zimbabwe,
Department of Agricultural Economics and Extension, Harare, 1999.

Frank, D. “World Cotton Production: Current Situation and Outlook.” Cotton Service
Buro. Proceedings of the International Conference on Cotton Connection:
Towards Sustainable Cotton Production. Hamburg, Germany, 1996.

Gallagher, K.D. “Farmer Field Schools for Integrated Pest Management in Africa with
Special Reference to East Africa.” Proceedings of the National Pre-Season
Planning Workshop on the Implementation of Field School Groups for Integrated
Production and Pest Management. 31 August-1 September, 1998. ZIPAM,
Darwendale. Government of Zimbabwe and FAQ Global IPM Facility, Rome,
Italy.

Greene, W.H. Econometric Analysis. Third Edition. New York University. Prentice Hall,
New Jersey. USA. 1997.

Harper, J.K., M.E. Rister, J .W. Mjelde, B.M. Drees, and MO. Way. “Factors Influencing
the Adoption of Insect Management Technology.” American Journal of
Agricultural Economics. 72(1990):997-1005.

Jowa, P. “Cotton IPM in Zimbabwe.” IPM implementation Workshop for East, Central
and Southern Africa. Harare, Zimbabwe. Natural Resource Institute, eds.,
Edingburg, United Kingdom, 1993.

Khanna, M., O.F. Epouhe, and R. Hombaker. “Site Specific Crop Management:
Adoption Patterns and Incentives.” Review of A gricultural Economics. 21(1999):
455-72.

Khanna, M. “Sequential Adoption of Site-Specific Technologies and Its Implications For
Nitrogen Productivity: A Double Selectivity Model.” American Journal of
Agricultural Economics. 83(2001):35-51 .

Kiss, A. “Pest Management Needs and Trends in Africa Today.” International Food

96

Policy Research Institute Workshop on “Pest Management, Food Security, and
the Environment; The Future to 2020. ”Nairobi, Kenya. May 10-11. 1995.

Kiss, A., and F. Meerman. Integrated Pest Management and African Agriculture. World
Bank Technical Paper Number 142. African Technical Paper Department Series.
Washington DC, USA. 1993.

Lazaro, A.A., K.L. Heong, B. Canapi, V. Gapud, and G.W.Norton. “Farmers’ Pest
Management Knowledge, Attitudes and Practices in San Jose, Philippines: A
Baseline Survey.” Integrated Pest Management Collaborative Research Support
Program (IPM CRSP) Working Paper 95-2. Virginia Tech. Blacksburg, VA,
USA. 1995.

Loevinsohn, M., G. Meijerink, and B. Salasya. “Integrated Pest Management in
Smallholder Farming Systems in Kenya.” Kenya Agricultural Research Institute.
Narobi, Kenya, 1998.

Loewenson, R and C.F.B. Nhachi. “Epidemiology of Health Impact of Pesticide Use in
Zimbabwe.” Pesticides in Zimbabwe: Toxicity and Health Implications. pp 25-35.
Charles, F. B. Nhachi and Ossy M. Kasilo, eds., University of Zimbabwe
Publications, Harare, Zimbabwe, 1996.

Lichtenberg, E., and D. Zilberman. “The Econometrics of Damage Control: Why

Specification Matters.” American Journal of Agricultural Economics.
68(1986):261-73.

Maddala, G.S. Limited-Dependent and Qualitative Variables in Econometrics.
Econometric Society Monographs No, 3. Cambridge University Press. USA.
1983.

Mulugetta, M. “An Economic Analysis of Smallholder Wheat Production and
Technology Adoption in the South-Eastem Highlands of Ethiopia.” PhD
Dissertation, Michigan State University, East Lansing, USA.1994.

Mullen, D.J., G.W. Norton, and Dixie, W., Reaves. “Economic Analysis of
Environmental Benefits of Integrated Pest Management.”Journal of A gricultural
and Applied Economics. 29(1997):243-53.

97

Napit, K. B., G.W. Norton, R.F. Kazmierczak, Jr., and E.G. Rajotte. “Economic
Impacts of Extension Integrated Pest Management Programs in Several States.”
Journal of Economic Entomology. 81(1988):251-56.

National Research Council. Ecologically Based Pest Management: New Solutions for a
New Century. National Academy Press, Washington DC. 1996.

Norton, G.W., and J.D.Mullen. Economic Evaluation of Integrated Pest Management
Programs. Publication No. 446140. Virginia Polytechnic Institute and State
University, Blacksburg. USA. 1993.

3

Pincus, J., H. Waibel and F. Jungbluth. “Pesticide Policy: An International Perspective.’
Approaches to Pesticide Reform-Building Consensus for Future Action.
Publication Series Number 7, Pesticide Policy Project, Hua Hin, Thailand, 1997.

Rajendran, T.P., and OM. Bambawale. “IPM Strategies for Rain-fed Cotton Cultivation
in Central India.” Pesticide Action Network Germany, Cotton Connection,
Towards Sustainable Cotton Production: Proceedings of the International
Conference “Cotton Connection- For Ecologically, Socially and Economically
Sustainable Cotton Production” 25th-26th November, 1994, Hamburg, Germany.

Rauniyar GP, and F .M. Goode. “Technology Adoption on Small Farms.” World
Development. 20(1992):275-82.

Rola, A.C. “Farmer Field School, IPM Knowledge and Farmer Efficiency: The Iloilo-
Philippines Case Study”. University of Philippines at Los Banos, College,
Laguna, Philippines, undated.

Rola, A.C., and PL. Pingali. “Pesticides, Rice Productivity and Farmers’ Health: An
Economic Assessment.” Washington DC: World Resources and International
Rice Research Institute, 1993.

STATA Corporation. Statistics, Data Management and Graphics. Stata Statistical
Software, Release 5. Stata Press, College Station, Texas, 1997.

98

Swinton, S.M., N.M. Owens, and ED. van Ravenswaay. “Health Risk Information to
Reduce Water Pollution.” Flexible Incentives for the Adoption of Environmental
Technologies in Agriculture. Frank Casey, Andrew Schmitz, Scott Swinton, and
David Zilberman, eds. Kluwer Academic Publishers, 1999.

Swinton, S.M. “Less is More: Why Agro-Chemical Use Will Decline in Industrialized
Countries.” Economics of A gro-Chemicals: An International Overview of Use
Patterns, Technical and Institutional Determinants, Policies and Perspectives.

Brookfield. USA. 1996.

Swinton, S.M., and MB. Williams. “Assessing the Economic Impacts of Integrated Pest
Management: Lessons from the Past, Directions for the Future.” Stafl Paper 98-

12. Department of Agricultural Economics, Michigan State University. East
Lansing. USA. 1998.

Thomas, J .K., H. Ladewig, and Wm. A. McIntosh. “The Adoption of Integrated Pest
Management Practices Among Texas Cotton Growers.” Rural Sociologi. 55,

3(1990):395-410.

Tjomhom, J .D., G.W. Norton, K.L. Heong, N.S. Talekar, and VP. Gapud. “Determinants
of Pesticide Misuse in Philippines Onion Production.” Phillip. Ent. 11(1997):139-

49.
Vandeman A, J. Femandez-Cornejo, S. Jans, and B. Lin. Adoption of Integrated Pest

Management in US Agriculture. United States Department of Agriculture.
Economic Research Service. Agriculture Information Bulletin Number 707.

Washington DC. USA. 1994.

Waibel, H. G., Fleischer, P.E. Kenmore and G. Feder. “Evaluation of IPM Programs-
Concepts and Methodologies.” Pesticide Policy Project in Cooperation with FAO.
Workshop an Evaluation of IPM Programs, Hannover, Germany, March 16-18,

1998.

Weber, C. “Pesticide Use in Cotton Production.” Pesticide Action Network Germany,
Cotton Connection, Towards Sustainable Cotton Production: Proceedings of the
International Conference “Cotton Connection- For Ecologically, Socially and
Economically Sustainable Cotton Production” 25"1-26th November, 1994,

Hamburg, Germany.

Wozniak, G.D. “The Adoption of Interrelated Innovations: A Human Capital Approach.”

99

The Review of Economics and Statistics, 66,1(1984):70-79.

World Bank. Integrated Pest Management: Strategies and Policies for Eflective
Implementation. Environmentally Sustainable Development Studies and

Monographs Series No.13. Washington DC. USA. 1997.

Yee, Y., and W. Ferguson. “Cotton Pest Management Strategies and Related Pesticide
Use and Yield.” Journal of Production Agriculture. 12(1999):6l 8-23.

100

Appendices

Table A3.1: Cotton IPM Practice Probit Model Results for District, 1998/99

cial Rotation Planting Spraying Chemicals Chemicals Crops
Pests Dates F

armer
Characteristics

Health Risks

Awareness &

L.L. Ratio
value

 

101

Table A3.2: Cotton IPM Practice Probit Model Results for Chipinge District, 1998/99

Ad ust ust te
Rotation Planting Spraying Scouting Chemicals
Dates F

Awareness

c

L.L. Ratio

value

Source: ,
Note: Probit model for buming stover (field sanitation) was insignificant. No farmer used beneficial insects
and trap crops as pest management strategies while less than 10% reported using safer chemicals in their
pest management decisions, thus making it statistically impossible to estimate probit models based on such
rare practices in Chipinge District.

 

102

Table A3.3: List of Cotton IPM-related Practices Used in Survey Areas, 1989/99

COTTON IPM-RELATED PRACTICES SANYATI CHIPINGE
DISTICT (%) DISTRICT (%)
A.
l
C

. le sanitation

C
.A just antin
scouting

 

Source: Field survey, 1999

103

CHAPTER 4

COTTON SUPPLY RESPONSE TO MARKET LIBERALIZATION IN
ZIMBABWE

4.1 Introduction

Structural change in the agricultural and food system has been receiving increased
attention in the last couple of years. The motivation for structural adjustment programs
(SAPS) is to address problems of declining agricultural output, limited commercial import
capacity, and stagnating economic growth due to distorted pricing and exchange rate
policies leading to loss of competitiveness in international markets (Lele, 1990; Jaeger
1992; Ngeno, 1996; Rodrik, 1990). The cornerstone of the reforms is the promotion of
private sector participation and the liberalization of agricultural pricing and marketing.

Much debate has focused on SAPS failure to eradicate the problems of deepening
poverty on the African continent (Gulhati, 1989; Husain, 1994). The economic hardship
facing the poor provides ample evidence that trickle down effects never occurred with the
adoption of SAPS (Longhurst, 1988;Quarcoo, 1990). Most SAPS are integrated packages
and slippage in one area could have profound impact on the outcome in other areas. In
addition, each reform has different impact on the various segments of the population
creating both winners and losers (Bates, 1993; Husain, 1994). Consequently, the effects
of SAPS, and the constraints arising from them, require careful analysis. Moreover,
country experiences Show a diversified pattern in terms of the content, implementation
process and impacts of SAPS, as other factors such as agrarian structure, frequency of
natural disasters and level of prior investments affect performance outcomes (Kayizzi-

Mugerwa et al., 1994).

104

Cotton is a strategic export crop in Sub-Saharan Africa (SSA) and in Zimbabwe it
contributed 11% in total exports in 1997 (Zimbabwe Ministry of Agriculture (ZMOA),
1999). The most serious challenges facing cotton industries in SSA are likely to result
from the liberalization of markets and the removal of subsidies following the adoption of
structural adjustment programs (SAPS) and the abolition of government marketing boards
in a number of countries (World Bank, 1995).

The key question this paper will address is this: What has been the response of
Zimbabwe’s cotton supply to SAPS, including the end of the Cotton Marketing Board
(CMB) monopoly on seed cotton marketing and processing? The empirical analysis
examines the response of cotton acreage to changes in relative prices, macroeconomic
policies (i.e. exchange rates and interests rates), geo-political and institutional factors.
The effect of drought, a major shock in the sub-region is also analyzed. To explore the
differences in farm sizes, capital intensity, specialization and financial flexibility that
make supply response between large scale commercial (LSC) and smallholder growers
heterogeneous, we estimate two separate acreage response models.

There have been several recent attempts to understand the effect of economic
policies on aggregate agricultural performance in SSA (Ogbu and Gbetibouo, 1990;
Mlarnbo and Kayizz—Mugerwa, 1991). But, there are still few examples of studies that
focus on supply response of individual cash crops in Africa (Bond, 1983; Jaeger, 1992;
Savadogo et al., 1995;Makanda, 1996). In addition, uncertainty still prevails regarding
what SAPS entail for the future of large versus smallholder cotton growers in Africa.
Townsend and Thirtle (1998) looked at long run relationships between the major

macroeconomic variables and agriculture in South Africa. Recently, Amdt et al., ( 2000)

105

qualitatively analyzed stabilization and structural adjustment in Mozarnbique’s economy.
Magadhlela (1997) adopts a similar approach but focuses narrowly on smallholder
irrigators in Zimbabwe. To the extent that previous research in Africa does not provide
much illiunination about reform impacts by scale of farm, the results of this paper have a
bearing on these issues. Moreover, economists are concerned about SAPS as their impact
has been mixed and it differs across countries (Putterrnan, 1995; Weissman, 1990). There
is an active debate on the welfare impacts of the reforms leading to growing consensus on
the need for human centered adjustment and adjustment with transformation (Kayizzi-
Mugerwa et al., 1994; Longhurst et al., 1988; Economic Commission for Africa, 1991).
Not surprising, detailed studies on SAPS in East and Southern Africa have focused on the
staple food crop maize (Masters, 1993; Argwings-Khodhelk, et al., 1993; Rubey, 1995).
But more work remains to be done on the effect of SAPS on cash crops. Policy impact
studies provide a road map for expanding agricultural supplies and are imperative
whenever policies change significantly (Ramaswanmi, 1993; Chen and Ito, 1992).
Masters (1993) provides an ex-ante critical analysis of proposed maize reforms in
Zimbabwe and their implications for Short run competition and long run efficiency goals.
What is different about our study is that we take an ex-post analysis of the reforms in
cotton, a leading cash crop for smallholders. Instead of focusing on SAPS alone, we
extend our analysis to include other external Shocks that could derail intended outcomes.
Zimbabwe’s agriculture is dualistic24 with highly sophisticated LSC farms average size

2,000 hectares (ha) co-existing with numerous smallholder farms that average about 2ha.

 

2" Zimbabwe’s LSC farmers have title deeds and can use their land as collateral for bank loans while
smallholder farms have communal ownership of land; they have usufructory rights but do not have rights to
sell, lease or mortgage the land.

106

In this vein, the two acreage response models help us characterize the current structure of
Zimbabwe’s cotton production more accurately. Highlighting the diversity in SAP
outcomes is important for two reasons. First, it allows the identification of successes and
failures and it brings to the fore the fact that aggregate trends can hide significant
differences in policy outcomes. Second, awareness of the distributional impacts of SAPS
helps policy makers set priorities, concentrating action where they are most needed. As
Zimbabwe undergoes SAPS, given its agriculturally based economy, sustained cotton
supply is critical to three areas of government objectives in agriculture: 1) expansion of
smallholder income earning opportunities 2) improvement of rural infrastructure and
employment opportunities and 3) reduction of balance of payment deficits given that
about 70% of the cotton produced in Zimbabwe is exported (Sithole and Attwood, 1990;

Zimbabwe National Development Plan, 1991).

4.1.1 Overview of Cotton Production Challenges in Africa

Cotton production in Africa dates as far back as the late 18th century when
European colonial authorities introduced it in East and Southern Africa in order to reduce
dependency on South American imports (Isaacman, 1996; Gibbon, 1998). In West
Africa, the evolution of cotton production is associated with the pioneering work by the
French dating back to the second half of the 20‘h century (Fok, undated). Cotton is now
successfully grown in more than 30 African countries and the annual market value of
African lint exports is approximately US$1 billion. SSA accounts for 4.5% of world
cotton production and its share of internationally traded cotton lint was 14% in 1995
(Lele, 1989; Gibbon, 1998). Although cotton has been one of SSA’S few exports success

stories in recent years, export diversification and food self-sufficiency have often sparked

107

huge policy debates. Consequently, cotton has undergone metamorphosis in the African
development literature moving from ascent to decline and to “rehabilitation” (Eicher and
Rusike, 1995). Cotton came under severe attack in the 19605 and early 19705 during
Africa’s food crisis leading to the ‘food first paradigm’ sharply critical of cash crop
expansion. Growing consensus that the “food first” lobby greatly overstated its case
against cash crops has now emerged (Maxwell and Fernando, 1989). Further, poor
economic growth rates combined with the debt crisis partly attributed to overvalued
exchange rates led more countries to Shift to export-oriented strategies in the mid-19805
(Dibley et al., 1996; Smith, 1989; Godfrey 1983). In the late 19805 and 19905, cotton
returned to the spotlight with more scientific work advocating its critical role in raising
farm incomes, food security and national foreign exchange generation (Bingen, 1998;
Dione, 1989; Gibbon, 1998; Jayne, 1992; Stringfellow, 1996; Strasberg et al., 1997).
Cotton industries in Afiica are facing several strategic issues that will affect both the
profitability and sustainability of cotton even in countries with comparative advantage in
cotton production (World Bank, 1994). The loss of market share for SSA countries, the
demand depressing financial crisis in Asia in 19905, and the huge production swings in
China and India have added momentum to the volatility in world prices raising new
concerns about marketing and financial risks crippling the cotton industry (F A0, 1997;
World Bank, 1995;Lele, 1989). Highly volatile prices and market share contraction has
lead to undesirable fluctuations in farm incomes, foreign exchange revenues and the raw
material supply of textile industries in Africa. Above all, the effect of SAPS on the

competitive advantage of cotton industries in Africa still remains unclear and uncertain.

108

Questions still linger about how reforms can revive African agriculture (Kelly, 2000;

Anandajayasekerum, 2000).

4.1.2 Contemporary Issues in Cotton Production in Zimbabwe

In Zimbabwe, cotton production is at a turning point. The rapid growth in post-
independence smallholder cotton production and marketing at a time when LSC sector is
diversifying into greenhouse production of horticultural and floricultural crops for export
market requires close scrutiny. Zimbabwe’s cotton production has historically been
concentrated in the Natural regions25 III, and IV although new varieties suitable for
cultivation in higher elevations more than 1200m above sea level have recently been
developed (Zimbabwe Cotton Handbook (ZCH), 1998).

Zimbabwe’s first post-independence cotton revolution occurred between 1980
and 1985. During this period smallholder cultivated area increased dramatically by 120%
(Figure 4.1). The return to peaceful conditions in the rural areas after a protracted war of
liberation, and the lifting up of international sanctions provided a favorable environment
for the dramatic growth in cultivated area (Chimedza-Mabeza, 1999; Rorhbach, 1989;
Mlambo, 1997). Further, key factors responsible for the dramatic increase in smallholder
cotton cultivation in the early 19805 are the upsurge in the number of rural cotton depots
and the clearance of tsetse flies from Gokwe, a high potential cotton area (Govereh, 1999;
Rukuni, 1994). As a result, the numbers of officially registered cotton growers increased
rapidly from about 80,000 farmers to more than 300,000 between 1983 and 1992

(Blackie, 1983; Jowa, 1996).

109

The smallholder acreage growth momentum slowed down between 1985 and

1995, and it’s generally attributed to two major droughts in 1986/87 and 1991/92. The
slow implementation pace of the resettlement program due to government budgetary
constraints could be a critical factor given that previous efforts to resettle the landless
have concentrated in Natural Regions III, IV and V where cotton is the most dependable
crop Mariga (1994). Kupfuma (1998) cites a severe reduction in support services ear-
marked for smallholders between 1986 and 1994; with Agricultural Finance Corporation
(AFC) credit loans allocated mostly to farmers in high rainfall areas leaving the bulk
without access to essential credit.

However, a major recovery (i.e. the second revolution) occurred after 1995 as the
area cultivated increased by about 60% between 1995 and 1999 (Figure 4.1). Perhaps a
crucial factor explaining the second revolution is the end of the civil war in Mozambique
and the signing of the peace accord26 in 1992 (Amdt et al., 2000). The end of hostilities
stimulated massive cotton expansion in the South East Lowveld that borders the Central
Region of Mozambique. Further, the region’s unique Vertisols27 enabled smallholders to
grow cotton without fertilizer, allowing significant savings in input costs.

A different pattern seems to characterize the LSC cotton acreage trends. Before
1980, LSC growers dominated cotton production with an insignificant amount coming
from the smallholders. By 1985/86, LSC farmers contributed 60% of the national share of

cotton output, down from about 90% in 1980. More significant contraction in LSC cotton

 

25 Agro-ecological zones demarcated on the basis of rainfall pattern, soil type and agricultural production
potential. Zimbabwe has 5 such regions, and the risk of crop failure due to drought increases as one moves
from natural region I to V.

26 Following negotiations in 1991, peace was agreed on in 1992, and it was accompanied in October 1994
by free and democratic elections whereby Joachim Chissano was elected President (Amdt et al., 2000).

2 Vertisols constitute about 5 percent of the land area or 1.2 million ha in Zimbabwe. The largest area is
found in the South East Lowveld in natural region IV. These soils are mainly derived from basalt and are

110

occurred when cultivated area reached an average of 64,000 hectares in the 19905
compared to an average of 80,000 hectares in the 19805 (Figure 1). As a result, LSC
average annual output dropped to 77,000 tons in the 19905 compared to 150,000 tons in
the 19805. In addition to the poor farm-gate prices and rising production costs, other
causes for the major decline in LSC cotton production includes adverse weather
conditions, emergence of verticillium wilt as a serious disease in some areas, poor quality
planting seed in 1989, and loss of confidence in the Cotton Marketing Board (CMB)
(Gillham, 1993). Further, the introduction of minimum wage legislation and stringent
labor regulations may have contributed to stagnation in LSC cotton acreage (Billing,
1985; Mlambo, 1997; ZCH, 1995).

During the past two decades, the structure of cotton production in Zimbabwe has
changed radically. In the late 19805, LSC farmers diversified into unregulated crops such
as tobacco and horticulture while smallholders now provide the bulk of national output

(67% in 1998) thus making cotton the most important cash crop in the communal areas.

4.1.3 Adoption of Structural Adjustment Programs (SAPs) In Zimbabwe

Zimbabwe adopted the World Bank/IMF-inspired SAPS in 1990, in response to
severe external shocks and emerging macroeconomic imbalances. The SAP package
consisted of:(1) flexible exchange rate policy, (2) state-owned enterprise restructuring
and reform, (3) marketing and pricing policy liberalization and (4) export promotion and
diversification (Sithole, 1996). Zimbabwe’s reform program resembles Similar programs

in many other African countries.

 

highly fertile due to their strong nutrient and water holding properties Nyamudeza et al. (undated).

lll

In cotton, the CMB had a GOZ-mandated monopoly in terms of seed cotton
purchases and distribution of cottonseed. Throughout the 19805, extensive controls
existed on cotton prices and marketing. Pan-territorial (i.e. no geographic price
differential) and pan-seasonal (i.e. no seasonal price differential) pricing policy operated
for most controlled crops including cotton, and producer prices were set pre-harvest
(Takavarasha, 1992). Nonetheless, CMB began to face financial problems as a pattern set
in of high budget deficits, weak management and motivation, delayed payment for
farmers and high interest rates. The next major challenge was to sustain the CMB by
making it more efficient, transparent and internally accountable.

The introduction of SAPS led to the subsequent demise of the CMB monopoly in
1993/94 and it gave birth to COTTCO, a privately owned company quoted on the
Zimbabwe Stock Exchange (ZSE) in Harare. The transformed company has as its
Shareholders, farmers, workers and government. The adoption of SAPS allowed an era of
competition and eliminated a financial burden for the GOZ after having subsidized the
losses of the CMB for more than a decade. Liberalization enabled the entry of COTPRO
and CARGILL Inc.(a large US based agro-business company) as new players in the
cotton industry in 1994/95 and 1995/96 respectively. Sheltered for decades from
competition, CMB has lost some market share to the emerging competitors although it
still operates as a regional monopoly in some geographic locations, a factor attributed to
the limited ginning capacity of rival firms. This study recognizes that reviving and
sustaining the “white revolution” in Zimbabwe requires compelling evidence on the

implications of SAPS on the performance of both LSC and smallholder cotton growers. It

112

draws lessons for other SSA countries along with some concluding remarks for further

research.

4.2 Methodology and Data
4.2.1 Theoretical Model for Cotton Supply Response

Assume the goal of Zimbabwe cotton farmers is profit maximization and let the
concave cotton profit function be 7r(p, r1..r,,)= py(x)- Zrix rT F C for i=1...n and the
resulting indirect expected profit function is given by 7r*= rim, r1..r,,, B). Using
Hotelling’s lemma, the first derivative of profit function gives us the cotton supply
function dn/ fli= y*( p, r1,..r,,,) where p is the product price, and r1..r,, are the prices of
variable inputs used in producing cotton output y and T F C is total fixed cost associated
with a vector of fixed inputs B. Nerlove’s na'l've functional form of acreage response to
prices states that desired acreage is a linear function of expected prices:

(1 ) A,*=ao+a,1>, *+a22,+u,
where A; is the expected cotton cultivated area in period t, P,* is the expected price in
period t, The vector P; comprises prices expected by the farmer in future periods. If
farmers rely on past information of the series when planning production to maximize
profits, then the expected prices are functions of past prices P,*=/1(P,, P,. 1, P,.2,. ..)
(Sarmiento and Allen, 2000, Martin, 1992). Additional supply shifters Z, in period t
include prices of substitute and complementary crops, government policy, technological
change and weather, and u, is the stochastic error term. Desired acreage behavior is
hypothesized as:

(I) At 7414:5641. ' At—I)

113

where LHS represents actual change in acreage between two periods, Aft-AH is the
desired change between two periods, and 6 is known as the coefficient of adjustment with
range 0< 5<1. Therefore, rearranging equation (2) we have:

(3) A, =5A,* +(1-6)A,.,
Substituting the desired acreage (1) above and the Nerlove price expectations P,*=P,.1
into (3) gives us the acreage response model:

(4) A,= 60+ 5, AM +5; P,.1+§3 Z, + u,
Estimates of price responses have also been obtained from functional forms based on
actual yield equations presented below:

(5) Yi=flo+flrYp +fl2PH+fl3Zi+ei
where Y, is actual yield at time t and Y ” is potential yield at time t given by Y ,” =Za V4,..-
for i=1...k where k =1 for annual crops (Bond, 1983). Past studies have demonstrated the
role of farm sizes in supply response analysis (Adelaja, 1991; Sedjo and Lyon, 1996). AS
already noted, the dichotomy in Zimbabwe’s cotton production in terms of external input
use, scale of operation, yields, and property rights allows us to model cotton supply
response as two separate functions. Also, the recommended long-staple irrigated LSC
cotton varieties differ considerably from the rain-fed medium staple varieties grown
mostly by smallholders, which further justify our disaggregation approach (ZCH, 1998).
The basic differences between these varieties create grade differences that fetch distinct
prices leading to different price responses. Further, because LSC farmers tend to have
irrigation, drought has differential impacts across the two sectors. In addition to the risk

management ability attributed to variations in capital intensity and economic status of the

114

growers, the adjustment process is influenced by factors like differential access to
productive land, and to credit and product markets.

Although the GOZ traditionally announces post planting or pre-harvest prices
during the month of March or April, at planting time in October farmers must anticipate
the future government-determined cotton price. Given the GOZ’s policy of annual review
of agricultural prices, we assume that cotton growers’ price expectations are
systematically formed in a forward-looking manner, suggesting that Zimbabwe’s cotton
farmers have rational rather than na'i've price expectations”. After SAPS, the Zimbabwe
Maize Commodity Exchange (ZIMACE) and COTTCO have offered more opportunities
for forward fixed price contracts to growers further justifies the rational expectations
argument (COTTCO, 1996). The formation of price expectations has been reviewed
extensively in the literature (Askari and Cummings, 1976). We expect that cotton acreage
will be positively and inversely correlated with its own price and price of substitute crops
respectively.

In SSA, weather plays an important role, as most agricultural production is rain-
fed. It is estimated that at least 60% of SSA is vulnerable to drought and up to 30% is
highly vulnerable (Benson and Clay, 1998). The role of risk in supply response has been
noted (Antonovitz and Green, 1990). Other work has used residuals from yield trend
model as a proxy for quantifying climatic risks (Jaeger, 1992). In addition to the negative
effect due to climatic risks, a number of African countries have been coping with

conflicts and civil strife which have devastating effects on their agriculture (Lele, 1990;

 

28 We assume that expectations are formed in a rational or forward-looking manner rather than an adaptive
(backward-looking or na‘lve) manner. Rational expectations assume economic agents use current and
relevant information in forming expectations and do not solely rely on the past (Gujarati, 1995; Soludo,
1998)

115

Messer et a1 1998). Civil war is linked to destruction of physical, human, and social
capital including disruption of markets and other forms of social and economic order
(World Bank, 2001). External factors like trade deficits and the debt burden have also
been hostile to growth and development in Africa (Ajayi, 1994). In addition to addressing
the structural imbalances, SAPS have serious social and economic consequences. In
Zimbabwe’s case, we expect exogenous shocks like ESAP and peace in Mozambique
MOZPEACE to have positive supply response effect among the cotton growers.

Some SSA countries such as Malawi, Tanzania and Mozambique have had a
history of forced cotton acreage although this has not been the experience of Zimbabwe
(Isaacman, 1996; Kydd, 1982; Shapiro, 1974). In the absence of forced cotton acreage,
we assume independent cotton acreage decisions are a good proxy for the aggregate
cotton supply in the respective subsectors (Ogbu and Gbetibouo, 1990; Bond 1983).
Also, area planted is under the control of the farmer to a much greater degree than output,
which is subject to drought, pests and other exogenous shocks (Townsend and Thirtle,
1997). As in conventional supply analysis, current period acreage is highly correlated
with acreage in previous years (Martin, 1998, Parhusip, 1976). Considering this, the
dependent variable is lagged one time period as an independent variable. The lagged
acreage for the LSC LLSAREA and smallholder farmers LCAREA respectively are expected
to have a positive effect on acreage response.

Agricultural output prices, input costs, exchange rates, wages, interest rates,
market imperfections and information are central to most discussions on responsiveness
of agriculture to policy and should be taken into account (Jaeger, 1992; Townsend and

Thirtle, 1998). The annual government guaranteed prices are used as output and input

116

price expectations. The prices are assessed in nominal terms as this represents what the
farmers use in their land allocation decisions. Since tobacco has not been a controlled
commodity, we used the average annual price at the Zimbabwe Tobacco Auction Sales
Floors. Maize and sorghum are the two major crops that compete with cotton for land and
labor in the smallholder sector while tobacco is considered the leading competitor in LSC
agriculture. The input price used in the model is ammonium nitrate fertilizer prices
ANFPRICE. We assume that a rise in fertilizer price induces a reduction in cotton acreage.
Macroeconomic variables that form part of our analysis are nominal exchange
rates EXCHRA TE and interest rates INTEREST. The exchange rate is a measure of
agricultural competitiveness (Dibley et al., 1996). It also captures the financial risk
induced by international price distortions in both factor and product markets. Zimbabwe
exports about 70% of her hand picked cotton to niche markets in Europe, the Far East,
South Africa and South America. Also, exchange rate devaluation29 is a critical
determinant of the distribution of income in a country as it influences the returns to
producers or consumers of tradable goods. Devaluation has a dual effect on the
profitability equation. On the revenue side, it raises the dollar price of internationally
tradable export commodities like cotton relative to non-tradable but on the cost side it
increases the price of imported machinery, and it spurs domestic inflation which can
“undo” its potential benefits (Dibley, et al., 1996; Poulton, 1998). Smith“ (1989) also

notes that devaluation timing, export taxes, and unequal market bargaining power may

 

29 Exchange rate devaluation is a reduction in the nominal value of the local currency relative to another
currency or group of currencies and it represents an increase in the cost of foreign exchange when paid for
in local currency (Dibley et al., 1996).

117

“cream off” potential producer price increase from devaluation. Its effect on acreage
response cannot be determined a priori.

Access to capital is critical in any supply response model and interest rates are a
proxy for LSC growers’ access to productive capital in support of cotton cultivation. We
also expect the level of legislated minimum wage MIWAGE to affect cotton acreage
response. It is reasonable given that cotton cultivators use hired labor especially during
labor—peak periods like weeding and harvesting. The price of labor used in the final
model is the official minimum wage in Zimbabwe. We expect a rise in wages to increase
production costs and hence lower the cotton acreage. Higher wages increase the
attractiveness of off-farm employment and thus offer disincentives to smallholder cotton
acreage expansion (Kydd, 1992). However, the negative effects are easily diluted by the
practice of voluntary labor exchange among the rural households.

Key non-price factors such as physical infrastructure, credit, extension and
research also play a major role (Lele, 1989; Savadogo et al., 1995). Change in technology
are a dominant factor in explaining supply response although there is no direct measure
of “changes in technology” hence the use of a proxy such as time which assumes a
smooth change in technology of equal amounts each period (Townsend and Thirtle,
1997). Comia et al., (1987) observes that supply capacity in Africa is severely
constrained by technical training. This study uses the annual number of graduates at the
Cotton Training Center as a proxy for technology development and transfer. We expect
the trained graduates to make more informed choices about land allocation, and to
implement innovative ideas in choosing cotton varieties, pest scouting, soil and water

management, and record keeping. The training variable is expected to positively

118

influence both smallholder and LSC cotton acreage. Other important non-price supply
shifters analyzed include size of the smallholder cattle herd that represents access to draft
power, a form of rural capital (Eicher and Baker, 1982; Savadogo et al., 1995). Draft
power variable was excluded in the final model, as it was statistically insignificant.

Due to limited data range, we could not estimate models with a rich lag structure,
as it would have reduced both the degrees of freedom and the measurement power of the
estimated model (Sarmiento and Allen, 2000). Given that the range for our data is 17
years, we decided to adopt a more parsimonius model. Variables included in the final
model are grouped into four categories; farm size, product and factor prices,
macroeconomic and institutional variables, and exogenous factors.

The LSC and smallholder supply models differ in terms of Z, values and relative
prices. Macroeconomic variables such as EXCHRA TE and INTEREST are deliberately
excluded from the smallholder model. Exchange rate movements are more likely to affect
acreage decisions and hence export revenue for LSC farmers who face competing needs
for land for diverse export crops such as tobacco, flowers, paprika and soybeans. More
importantly, changes in the exchange rate affect the rental price of imported machinery
and machinery parts. We expect the exchange rate to influence the LSC cotton growers’
diversification strategies more than smallholders who use hoe-cultivation. The latter’s
acreage decisions are more likely to be influenced by household consumption decisions
suggesting that land allocation between food and cash crops is a paramount factor which
is captured by the inclusion of the relative price of the substitute crops. Further, the
information gap makes it highly unlikely for them to use the exchange rate in their

acreage decisions.

119

Nearly all LSC cotton growers have collateral and hence ready access to financial
credit. It is therefore reasonable to assume that fluctuations in interest rates affects LSC
farmer’s access to capital, and is a critical ingredient for their acreage decisions. Since
LSC cotton grower’s operations are highly mechanized, we consider the variable cattle as
irrelevant in the Specification of their acreage response model. Instead, cattle is
considered a proxy for capital in smallholder agriculture but was eventually dropped from
the final model as it turned out insignificant. Supply response model Specification
requires a clear understanding of the farming system. As already highlighted, food crops
compete with cotton for land and labor in Zimbabwe’s smallholder agriculture while in
the same vein tobacco competes with cotton on LSC farms. This explains why the
relative prices are different in the two models. It is also important to note that
autoregressive supply response models are subject to specification errors and diagnostic

tests are needed to correct any such problems (Greene, 1997; Ferris 1998).

4.2.2 Empirical Econometric Model

From the fore-going considerations, we present the general structure of the auto-
regressive cotton acreage response model estimated:
(5) (LSC cotton acreage), =f(lagged acreage, product prices, input prices,
macroeconomic factors, institutional factors, exogenous shocks)
The smallholder acreage supply model is similar, except that macroeconomic variables
are excluded as largely irrelevant.
( 6) (Smallholder cotton acreage)Ff(lagged acreage, product prices, input
prices, institutional factors, exogenous shocks)

The empirical models are Specified mathematically as follows:

120

( 7) LSAREA= ,1 (LLSAREA, COTTONP, TOBACCOP, ANFPRICE, WAGE, TRAINING
EXCHRA TE, INTEREST, LDROUGH T, MOZPEACE, ESAP) + e ,
(8) CAREA = ¢ (LCAREA, COTTONP, MAIZEP, SOGH UMP, ANFPRICE, WAGE,

TRAINING, LDROUGH T, MOZPEA CE, ESAP) + u ,

4.2.3. Data Considerations

The data used in the analysis were obtained from various publications (Table 4.1)
including the Zimbabwe Ministry of Agriculture (ZMOA, 1999), Central Statistical
Office (C80) (1989), Reserve Bank of Zimbabwe (RBZ), (RBZ, 1995), Cotton Training
Center (J arachara, personal communication), and the National Employment Council

(N EC). These sources provided quantitative data on annual cultivated cotton area, yields,
crop producer prices, fertilizer prices, interest rates, wages, exchange rates and training
levels at CTC. Most of the data covers the period from independence in 1980 to 1997.
Our analysis focuses on this period first, because of the extensive controls that existed in
cotton pricing and marketing policy before 1990. Second, because we are interested in
analyzing the impact of SAPS implemented in1990 by the GOZ. Third, because of the
reliability of aggregate level data collected after the war period, such data offers the
background for understanding cotton’s acreage historical transformation and its growth
potential.

4.3. Empirical Results

Several diagnostic tests30 were conducted to address the problem of model

misspecification. Both the LSC and smallholder cotton acreage series were stationary and

 

3° Inspection of corellograms and preliminary unit root tests were made on all variables included in the final
model. Both the Dickey Fuller (DF) unit root tests and corellograms showed no evidence of non-
stationarity in the variables. The reported standard errors were corrected for a general form of

121

did not show any trends or random walk components. Since degrees of freedom were a
premium, we estimated a structural model with parsimonious set of parameters. The
supply response models use data in levels rather than logarithms based on results of
MWD and RESET tests. Supply response modeled as differenced equations were rejected
due to theoretical consistency problems. Secondly, potential loss of long-term
relationships between cotton acreage and relevant economic variables resulting from use
of differenced model enabled us to work with the levels model [Gujarati, 1995]. The
explanatory power of the models is high as both regressions are highly significant. The
coefficients of acreage adjustment for LSC and smallholders are respectively 0.78 and
0.61 although the former is insignificant. Based on the evidence, we infer that 61% of the
discrepancy between the desired and actual smallholder cotton acreage is eliminated in a

year.

4.3.1 Large Scale Commercial Cotton Supply Response Results

Results show that contrary to policy makers’ expectations, the implementation of
SAPS has contributed to a significant reduction in LSC cotton acreage in Zimbabwe
(Table 4.2). The coefficient for SAP is negative and significant at 5% level. The result
suggests evidence of structural change in LSC cotton cultivation due to SAPS. Expected
SAP-induced cotton acreage expansion may have been undone by rising inflation rates,

electricity prices, railway tariffs and the high cost of manufactured inputs resulting in a

 

heteroscedasticity using the Huber-White method. Based on CRDW test for co-integration, neither model
exhibits spurious regressions, as all the explanatory variables included in the respective models were
cointegated with acreage at 5% and higher, a positive sign for stable long run relationship (Gujarati, 1995).
The RESET test for functional form misspecification yields a value of 6.25 (LSC) and 0.45 (smallholder)
less than the respective critical values of 7.71 and 5.99, indicating failure to reject the hypotheses of correct
functional form. The MWD test identifies the levels model as suitable functional forms for both farmer
groups.

122

negative impact on LSC cotton acreage. SAPS may also have added momentum for LSC
growers to diversify out of cotton production as illustrated by a major downturn in cotton
acreage after 1989. Given that SAPS are still ongoing under the Zimbabwe Program for
Economic and Social Transformation3 I (ZIMPREST), their full impact can only be felt at
the end of this program.

Among LSC farmers, there is evidence of negative acreage response to the
Mozambique peace process although it barely achieves statistical significance. We
postulate that the peace settlement which saw the repatriation of Mozambican refugees
back to their country may have resulted in the disappearance of previously available
cheap labor making it more difficult for LSC farmers to find cotton pickers (Magadela,
1997). Further, the development of machine pickable varieties by the Cotton Research
Institute is in response to decreasing hand picking labor supply following the departure of
Mozambique refugees (ZCH, 1995).

As expected, the coefficient for minimum wage MIWAGE has a negative
significant impact on acreage. The negative sign provides evidence of economic hardship
facing LSC growers following the legislation of minimum wages in agriculture soon after
independence. As labor costs escalated, it became increasing difficult for LSC growers to
hire contract laborers for harvesting cotton leading to a contraction in cotton acreage. The
official minimum wage rose annually by about 32 % during the first decade of

independence and it rose further by 21% per annum between 1990 and 1995.

 

3 ' The main objectives of ZIMPREST are synonymous with the GOZ’s 1981 “Growth With Equity Policy”
which are to: restore macroeconomic stability (low inflation and interest rates, stable exchange rates),
facilitate public and private sector savings and investment, promote economic empowerment and poverty
alleviation through employment creation and to strengthen safety nets for the disadvantaged groups
(ZIMPREST, 1998).

123

Ammonium nitrate fertilizer price ANFPRICE has an unexpected positive and
significant effect on cotton acreage. A number of explanations can give rise to this
scenario. Lack of viable alternative fertilizers to Ammonium Nitrate could be a factor. A
possible explanation also, is that a positive coefficient suggests the presence of
substitution effects between fertilizer and land in Zimbabwe’s commercial cotton
production.

The cotton price coefficient for LSC acreage model is positive and significant at
5% level. The result indicates that prices do matter and can be powerfiil engines of
transformation (Berg, 1995). This highlights LSC growers responsiveness to econorrric
incentives brought about by higher cotton prices. The coefficient for tobacco prices
TOBACCOP is significant and has the expected negative sign. The relative price effect is
theoretically consistent as we expect tobacco to compete for land and labor inputs with
cotton.

Among the macroeconomic variables, the coefficient for interest rate INTEREST is
significant while that for exchange rate EXCHRA TE is insignificant. The coefficient for
interest rate INTEREST has a statistically significant negative effect on LSC cotton supply
response. Interest rates negatively affect the financial position of growers as well as the
level of capital deployment. Zimbabwe’s nominal interest rates increased from 7.5% to
35.5% between 1980 and 1995 choking off potential investment in productive assets in
support of cotton cultivation. High interest rates also exacerbate the level of farm
indebtedness. The insignificance of the exchange rate variable should be put into context.
Prior to SAPS, Zimbabwe had rigid controls in foreign exchange and the exchange rate

was subjected to erratic depreciation in 1982, 1983,1988 and 1989. The over-valued

124

exchange rates in the 1980s meant that cotton growers were explicitly taxed as they were
paid less than the farm-gate export parity prices. More systematic exchange rate
devaluation only started after liberalization when the Z8 fell from U$1.00: Z$2.64, in
1990 down to U$l.00:Z$38.00 in 1999. Whether the devaluations brought the exchange
rate to equilibrium is debatable. However, realignment of the exchange rate is expected to
generate positive incentives to expand cotton acreage. Nonetheless, the evidence clearly

Shows that LSC farmers are highly responsive to agricultural output prices.

4.3.2 Smallholder Cotton Supply Response Results

The smallholder model results Shows that institutional factors such as resolving
land constraints and improved knowledge through access to training matters most in
efforts to stimulate cotton acreage expansion in Zimbabwe. However, the coefficient for
SAP displays a statistically insignificant negative Sign for smallholders.

A key result is that the coefficient for Mozambique peace MOZPEACE is positive
and statistically significant at the 1% level. The result suggests that the peaceful
conditions prevailing in the South East Lowveld after a protracted civil war in
Mozambique enabled previously displaced farmers to return to their old farms in the
villages and produce cotton. Cotton expansion in the South East Lowveld accelerated in
the 19905 and is attributed to the “peace dividend”. The cotton boom in the Lowveld has
benefited from the huge belt of fertile vertisols that make it easier for farmers to grow
cotton without any fertilizer applications. The non-use of fertilizer presents a major
comparative cost advantage to Lowveld cotton growers at a time when the post-
liberalization upsurge in fertilizer prices and the tightening of credit supply has caused a

severe decline in the aggregate use of external inputs throughout SSA (Kelly, 2000;

125

Poulton et al., 1998; World Bank, 1994). In addition, the peace process in Mozambique
resulted not only in the cotton boom in the Lowveld, but it also helped eliminate costly
transport diversions and improve the flow of exports from land-locked Zimbabwe to the
port of Beira, the country’s shortest and most vital link to the ocean. Similar evidence of
the Mozambique civil war raising external transportation costs, displacing thousands of
refugees and drastically changing the economic situation in neighboring countries has
been cited (Kydd, 1986;Lele, 1990, Cromwell, 1992).

It is striking to note that the coefficients for input and output prices including that
of cotton COTTONP are statistically insignificant. This suggests that agricultural pricing
policy did not provide sufficient incentives for smallholder cotton acreage expansion. To
Speed up price response, there is need to raise farmer awareness and understanding of
seed cotton pricing and payment options through the provision of rural market
infrastructure such as market information system that disseminate price news and lower
entry barriers in all major cotton growing areas to eliminate regional monpsony and thus
encourage greater competition and efficiency (Masters, 1993; Poulton, 1998). Lele,
(1989) argues that effectiveness of price policies in ensuring a supply response is itself
heavily conditioned by the quality of the institutions that conduct cotton research,
extension, input supply and commercialization.

An important question is whether the seed cotton grading process allows farmers
to obtain the highly priced grades. A major complaint from Zimbabwe smallholders is
that they receive low grades and prices for their cotton. Analysis of marketed cotton
output in Sanyati and Chipinge indicates that less than 50% of the farmers’ output

achieve first grade (A) in both districts while second grade (B) seems to dominate most

126

of the output in Chipinge. Further, Zimbabwe’s cotton farmers have historically faced
late payment problems and that has major disincentive effects. Lack of trust in the
marketing system increases marketing costs, restricts use of markets, and thus limits
opportunities (Shaffer etal., 1985). The positive and significant lagged cotton acreage
LCAREA response suggests a cob-web supply relationship, implying that smallholder’s
planting decisions are influenced by previous year’s acreage.

The positive coefficient for training is statistically significant at 5% level. This
indicates a profound relationship between smallholder cotton expansion and CTC training
of growers. The result highlights the pivotal role CTC has played in the transfer of
technology among smallholders. A total of 16, 014 trainees (i.e. comprising mostly
farmers and extension workers) attended various cotton courses offered by CTC in the
19805 compared to 12,023 in the 19905. The smallholder model results shows that
institutional factors are the centerpiece in efforts to stimulate cotton acreage expansion in

Zimbabwe.

4.4 Conclusion

The clearest pattern emerging from our study is that economic reforms generate
unintended negative effects for both LSC and smallholder farmers. It seems fair to
conclude that, in the short run, SAPS generate negative impacts among Zimbabwe cotton
growers. The rapid upfront costs of the deflationary policies are working against cotton
acreage expansion. The depressing effect on cotton growers is particularly worrisome
given that cotton is the second leading cash crop in terms of foreign exchange generation
after tobacco. The negative effect of the SAP coefficient on LSC cotton growers suggests

that compensatory measures are needed to limit their vulnerability to adverse effects of

127

SAPS and the resulting diversification out of cotton production. The results present
unexpected challenges to addressing the equity objective through adjustment process, as
it appears that SAPS have negatively affected the wealthier LSC farmers. The real
challenge is to craft policies that improve smallholders’ access to SAP-induced benefits
without hurting the LSC growers. The basic argument is that such pro-poor policies have
a public good impact.

An important lesson supported by our findings is that liberalization does not result
in automatic production incentives (Putterman, 1995). A growing number of independent
researchers have come to the same conclusion that in the African context, SAPS have
resulted in modest or disappointing supply responses (Ajayi, 1994;Bigsten and Ndung’u,
1992). Additionally, SAPS are being blamed for their deleterious effects on human and
physical infrastructure in Africa (Browne, 1992). Consequently, the appropriateness of
liberalization as an agricultural development strategy in a situation of missing and
imperfect markets, and poor infrastructure is now being questioned (United Nations,
1999). Moreover, there is a convergence of ideas on the fact that poverty worsened in
Africa in the 19805 following the adoption of SAPS (Berg, 1995; Cornia et al., 1988).
Other work cites not only increased incidence of poverty but also the creation of “new”
poverty following adjustment, which threaten the productive capacity of poor farmers
(Stewart, 1995).

A striking result from our analysis is that the end of the Mozambique civil war
and the signing of the peace accord had positive Spillover benefits for smallholders in
neighboring Zimbabwe. In other words, after accounting for SAP and drought, a net gain

occurred in land area cultivated to cotton highlighting the economic benefit of

128

investments in peace. The policy implication is that resolving hostilities is critical for
agricultural growth in Africa. Our results also Show that price incentives play a
significant role in determining cotton acreage response for LSC growers. However, the
insignificant coefficients for both product and factor prices suggests that there is no
evidence of the expected rapid switching of acreage into cotton production due to
agricultural price liberalization among the smallholder growers. Structural rigidities are
inferred by smallholder’s limited capacity to respond to price incentives without assured
relevant production inputs and credit support. The evidence underlines the fact that
smallholder supply response is strongly influenced by non-price factors such as access to
land, investment in training and peace, but not economic incentives.

Future work should investigate more Specific elements of the reform agenda such
as 1) the move from monopsonistic to oligopolistic marketing in cotton purchases 2)
credit disbursement under market liberalization, 3) implications of reforms in cotton
research and development, especially the recent elimination of the cotton varietal purity
policy (one variety in one region), and 4) liberalization of cotton seed sector and the
potential impact of the introduction of cotton biotechnology. Further research effort
should be devoted to diagnosing the implications of land reforms for commercial
cottonseed production and seed cotton supply response in Zimbabwe in general. The
results obtained here provide ample evidence that scale of operation influence Zimbabwe
cotton growers’ response to the generally improving incentive environment brought about
by SAPS. Although it seems reasonable to argue that smallholders are relatively more
vulnerable to the dire effects of SAPS due to their low income and inability to save and

accumulate assets, the bottom-line is that so far, Zimbabwe LSC cotton growers have

129

benefited Specifically from cotton price liberalization while smallholder’s response to the
economic incentives offered by SAPS has been muted in the short run.

Future gains from SAPS depend upon the cotton growers’ capacity to obtain
marketing and other services on more favorable terms than those originally offered by the
state. In Zimbabwe’s case, it could be that new opportunities offered by liberalized
markets are still in their formative stages or are concentrated in regions with better
infrastructure. The missing or imperfect market argument seems plausible given the
initial limited geographical presence of rival firms in the cotton industry, inaccessible
roads in the new cotton zones, and the vacuum in credit service provision created by the
gradual retreat of the GOZ. Signs of overall negative initial effects of SAPS on
Zimbabwe’s cotton growers suggest that its future benefits are uncertain.

The immediate challenge for the GOZ is to redesign and administer SAPS that do
not create new resource gaps but rather reduce the flaws depicted by the early generation
of reforrrrs. Compensatory measures to reduce, mitigate and cope with the unintended
negative short-run impacts of SAPS are needed if the elusive goal of eradicating income
poverty through broad-based expansion of cotton production is to be realized. Key
features of a future strategy for cotton expansion are; targeted training of smallholders,
secure access to land, job creation schemes to re-deploy retrenched workers and
improvements in smallholders’ access to markets and information. Increasing training
and market opportunities must be complemented by investment in rural infrastructure to
eradicate barriers due to geographic remoteness of the smallholder cotton growers. A
clear lesson is that getting prices right in an institutional vacuum will not generate the

needed positive supply response from Zimbabwe’s smallholder cotton growers.

130

Table 4.1 Description of variables used in Zimbabwe’s Cotton Supple Response Models, 1980-1997

 

 

Variable Name Units Data Mean Standard
Source Deviation
Dependent variable(s)
Communal area acreage (CAREA) Hectares ZMOA 153,926.50 71,446.19
Large-scale commercial acreage (LSAREA) Hectares ZMOA 65,068.95 21,144.77
Scale of operation (Farm Size)
Lagged communal acreage (LCAREA) Hectares ZMOA 153,926.50 71,446.19
Lagged LSC acreage (LLSAREA) Hectares ZMOA 65,068.95 21,144.77
Product and factor prices
Maize producer price (MAIZEP) ZS/ton ZMOA 540.28 71 1.71
Sorghum producer price (SOGHUMP) ZS/ton ZMOA 405.56 481.30
Ammonium nitrate price (ANFPRICE) ZS/ton ZMOA 1043.28 1,206.25
Cotton producer price (COTTONP) Z$/ton ZMOA 1,914.47 2,014.45
Tobacco average price (TOBACCOP) ZS/ton ZMOA 15,288.91 18,496.90
Minimum wage (MIWAGE) Z$/month NECZ 363.23 437.63
Macroeconomic and institutional variables
Exchange rate (EXCHRATE) Z$/U$ ZMOA 9.84 14.25
Interest rate (INTEREST) (%) RBZ 19.75 12.17
Cotton Training Center (TRAINING) Number CTC 1,465.68 385.35
Exogenous shocks
Lagged drought (LDROGHT) (0,1) ZMOA 0.25 0.44
Mozambique peace (MOZPEACE) (0,1) Amdt, C et a1. 0.43 0.51
Structural Adjustment Program (SAP) (0,1) Mlambo, A.S. 0.52 0.51

 

131

Table 4.2: Zimbabwe Cotton Acreage Response Model Results, 1980-1997

Dependent variable: cultivated cotton acreage (ha)

 

 

 

 

Large Scale Commercial Model Smallholder Model
Parameter Estimate z-value Estimate z-value
Scale of operation (farm size)
LAGGED AREA 2.10 0.06 *0.39 1.87
Product and factor prices
COTTONP “230.00 2.08 35.00 0.23
TOBACCOP **-4.10 -2.01 - -
ANF PRICE * 120.00 1.66 -120.00 -1.21
MIWAGE *-1400.00 -1.72 430.00 0.89
MAIZEP - - -91.00 -0.32
SOGHUMP - - 170.00 0.47
Macroeconomic and institutional
EXCHRATE 38,000.00 1.34 - -
INTEREST **-6,200.00 -2.41 - -
TRAINING -2,100.00 -1.07 “34.00 2.09
Exogenous Shocks
LDROUGHT 9,600.00 1.55 12,000.00 1.25
MOZPEACE -94,000.00 -1.39 ***33,000.00 2.42
SAP **-380,000.00 -2.02 -32,000 -0.27
N 17 17
Log likelihood chi-square -162.82 -184.72
x2-p value 0.0000 0.0000

 

***=significance at 1% level, ”=significance at 5% level,*=significance at 10% level

132

LIST OF FIGURES

 

Cultivated Cotton Acreage Trends, 1980-1999

350000.00
300000.00 ’:
250000.00
200000.00 »
150000.00 g
100000.00
50000.00

Acreage [ha]

 

 

 

L {—0— National + Communal + CommercLaTl

 

 

 

Figure 4.1: Zimbabwe Trends in Cotton Area Cultivated, 1980-2000

133

References

Adelaja, A.O. “Price Changes, Supply Elasticities, Industry Organization, and Dairy
Output Distribution.” American Journal of A gricultural Economics, 73(1991):89-
102.

Ajayi, S, I. “The State of Research on the Macroeconomic Effectiveness of Structural
Adjustment Programs in Sub-Saharan Africa.” Structural Adjustment and Beyond
in Sub-Saharan Africa: Research and Policy Issues. Rolph Van der Hoeven, and
Fred Van Der Kraaij, eds. James Currey, London, United Kingdom, 1994.

Anandajayasekerum, P., A.B.Torkelsson, and J. Dixon. “Structural Adjustment
Programs and Their Implications To Smallholder Producers: Lessons Learned
From Eastern and Southern Africa.” International Farming Systems Association
Workshop, Nairobi, Kenya, 2000.

ArgwingS-Kodhek G., M. Mukumbe and E. Monke. “The Impacts of Maize Market
Liberalization in Kenya.” Food Research Institute Studies. Stanford University,
22(1993):227-51.

Amdt C., H.T. Jensen, and F .Tarp. “Stabilization and Structural Adjustment in

Mozambique: An Appraisal.” Journal of International Development,
12(2000):299-323.

Bates, R.H. “ Politics of Economic Policy Reform: A Review Article.” Journal of
African Economies, 2(1993):418-33.

Benson, C., and E. Clay. The Impact of Drought 0n Sub-Saharan African Economies: A
Preliminary Examination. World Bank Technical Paper Number, 401 ,
Washington DC. 1998.

Bigstem, A., and NS. Ndung’u. Kenya. Structural Adjustment and The African
Farmer. Alex Duncan and John Howell, eds. Overseas Development Institute,
London, 1992.

Berg, E. “African Adjustment Programs: False Attacks and True Dilemmas.” Structural

134

Adjustment: Retrospect and Prospect. Daniel M. Schydlowsky, eds. Praeger,
Westport, 1995.

Billing, K.J. Zimbabwe and the CGIAR Centers: A Study of T heir Collaboration in
Agricultural Research. Study Paper Number 6, 1985.

Binswanger, HP, and Townsend, RE, “The Growth Performance of Agriculture in Sub-
Saharan Afi'ica.” American Journal of Agricultural Economics. 82:(1075-1086),
2000.

Bond , M.E. Agricultural Responses to Prices in Sub-Saharan African Countries. IMF
Staff Papers, Washington DC. (1983):703-26.

Brown, R.S. “Alternative Policy Frameworks for African Development in the 19905.”
Beyond Structural Adjustment In A fiica: The Political Economy of Sustainable
and Democratic Development. Julius E. Nyang’oro and Timothy M. Shaw, eds.
New York, 1992.

Central Statistical Office, Statistical Year Book. Government Printers, Harare,
Zimbabwe, 1989.

Chen, D,T., and S, Ito., “Modeling Supply Response with Implicit Revenue Functions: A
Policy-Switching Procedure for Rice.” American Journal of Agricultural
Economics, 74(1992):1 86-96.

Chen, Z., and R. Lent. “Supply Analyis in an Oligopsony Model.” American Journal of
Agricultural Economics. 74(1992):973-79.

Comia, G.A., R. Jolly and F. StewartAdjustment with a Human Face. Volume 1,
Country Case Studies, Oxford University Press, Oxford, 1987.

Comia, G.A., R. Jolly and F. Stewart. Adjustment with a Human Face. Volume 11,
Country Case Studies, Oxford University Press, Oxford, 1988.

Cotton Company of Zimbabwe Private Limited. Annual Reports. Harare, 1995, 1996,
1997, and 1998.

135

Cotton Company of Zimbabwe Private Limited. Cotton-0n. The House Journal of The
Cotton Company of Zimbabwe, Harare, Zimbabwe, 1999.

Cromwell, E. Malawi, Structural Adjustment and The African Farmer. Alex
Duncan and John Howell, eds. Overseas Development Institute, London, 1992.

Dercon, S. “Peasant Supply Response and Macroeconomic Policies: Cotton in Tanzania.”
Journal of African Economies, 2(1993):158-94.

Dibley, D., T. Reardon, and J, Staatz. How Does a Devaluation Affect an Economy?
Lessons From Africa, Asia, and Latin America. Department of Agricultural
Economics, Staff paper Number 96-105, Michigan State University, East Lansing,
1996.

Dione, J. Informing Food Security Policy in Mali: Interactions Between Technology,
Institutions and Market Reforms.” Ph.D. Dissertation, Michigan State University,
East Lansing, 1989.

Economic Commission for Africa. African Alternative Framework to Structural
Adjustment Programs for Socio-Economic Recovery and Transformation: A
Popular Version. Addis Ababa, 1991.

Eicher, OK, and D. Baker. Research on Agricultural Development in Sub-Saharan
Africa: A Critical Survey. MSU International Development Paper Number 1, East
Lansing, 1982.

Eicher, GK, and J .Rusike. “Agribusiness in Eastern and Southern Africa.” Africa and
Rural and Urban Studies. Michigan State University Press, East Lansing, Volume
2, Numbers 2-3, p7-28, 1995.

Ferris, J .N. Agricultural Prices and Commodity Market Analysis. McGraw Hill, Boston,
Massachusetts, 1998. '

Fok, A.C.M. “Progress in the Cotton Sector in Africa: Green Revolution or Institutional
Evolution?” Annual Crops Division-CIRAD, Paris, undated.

136

Gibbon, P. “Peasant Cotton Cultivation and Marketing Behavior in Tanzania Since
Liberalization.” Working Paper Sub-series on Globalization and Economic
Restructuring in Africa Number 1, Center for Development Research,
Copenhagen, 1998.

Gillham, E. M. Case Study Report on Cotton Research and Development Work in South
Afiica, Tanzania, Uganda and Zimbabwe. International Cotton Advisory
Committee, World Bank, Washington DC. 1993.

Godfrey, M. “Export Orientation and Structural Adjustment in Sub-Saharan Africa.” IDS
Bulletin, 14(1983):39-44.

Goetz, S.J., “Economies of Scope and the Cash Crop-Food Crop Debate in Senegal.”
World Development, 20(1992):727-34.

Govereh, J. “Impacts of Tsetse Control on Migration and Capital Accumulation:
Zambezi Valley, Zimbabwe.” Ph.D Dissertation, Michigan State University, East
Lansing, USA, 1999.

Greene, W.H. Econometric Analysis. Third Edition, Prentice Hall, New Jersey, 1997.

Gujarati, D.N. Basic Econometrics. Third Edition, McGraw Hill Inc., 1995.

Gulhati, R. Malawi: Promising Reforms, Bad Luck Economic Development Institute of
the World Bank, Analytical Case Studies, Number 3, 1989. Washington DC.
1989.

Husain, I. The Macroeconomics of Aajustment in Sub-Saharan Afiican Countries:
Results and Lessons. Policy Research Working Paper 1365, The World Bank,
Washington DC, 1994.

J aeger, W.K. The Effects of Economic Policies on African Agriculture. The World Bank
Discussion Papers, Africa Technical Department Series, 147, Washington DC.
1992.

137

Kayizzi-Mugerwa, S., and J. Levin. “Adjustment and Poverty: A Review of the African
Experience.” African Development Review, 6(1994):1-38.

Kelly, V. “Sahelian Input Markets: Recent Progress and Remaining Challenges.” Staff
Paper 00-36. Department of Agricultural Economics, Michigan State University,
East Lansing, Michigan, 2000.

Kupfuma, B. “Economic Efficiency and Returns to Scale of Communal Area Agriculture
in Zimbabwe and Implications for Agrarian Reform.” Ph.D Dissertation,
Michigan State University, East Lansing, 1998.

Kydd, J. “The Effectiveness of Structural Adjustment Lending: Initial Evidence from
Malawi.” World Development, l4(1986):347-65.

Kydd, J ., and R. Christiansen. “Structural Change in Malawi Since Independence:
Consequences of Development Strategy Based on Large-Scale Agriculture.”
World Development, 10(1 982):355-3 75.

Isaacman, A. Cotton is the Mother of Poverty: Peasants, Work, and Rural Struggle in
Colonial Mozambique, 1938-1961. James Currey, London, 1996.

Jayne, T.S. “Cash Cropping Incentives, Food Market Performance and the Divergence
Between National and Household Comparative Advantage: Evidence From
Zimbabwe.” Working Paper AEE 3/92. Department of Agricultural Economics
and Extension, University of Zimbabwe, Harare, 1992.

Jowa, P. “IPM on Cotton in Zimbabwe. IPM Implementation Workshop of East, Central
and Southern Africa.” Harare, 1996.

Lele, U. Cotton in A fiica: An Analysis of Diflerences in Performance. Managing
Agricultural Development in Africa. Madia Discussion Paper No.7, World Bank,
Washington DC. 1989.

Lele, U. “Structural Adjustment, Agricultural Development and the Poor: Some Lessons
from the Malawian Experience.” World Development, 18(1990):1207-19.

138

Longhurst, R., S. Kamara, and J., Mensurah. “Structural Adjustment and Vulnerable
Groups in Sierra Leone.”IDS Bulletin, 19(1988):25-30.

Mabeza-Chimedza, R. “Zimbabwe’s Smallholder Agriculture Miracle.” Food Policy,
23(1998):529-37.

Magadlela, D. “The Social Impact of Structural Adjustment Programs on the Smallholder
Irrigation Farmers in Zimbabwe.” Structural Adjustment, Reconstruction and
Development in Afiica. Kempe Ronald Hope, Sr., ed. Ashgate, England, 1997.

Makanda, D.W. “Wheat Policy in Kenya.” PhD Dissertation, Michigan State University,
East Lansing, USA, 1996.

Marsh, J .M. “Estimating Inter-temporal Supply Response in the Fed Beef Market.”
American Journal of Agricultural Economics. 76(1994):444-53.

Martin, C.A. “Supply Response in Argentina: Aggregate Planted Area in Crop
Production.” Plan B MSc Paper, Department of Agricultural Economics,
Michigan State University, 1998.

Mariga, I. K. “Cotton Research and Development, 1920-1990.” Zimbabwe ’s Agricultural
Revolution. Mandivamba Rukuni and Carl K. Eicher, eds. University of
Zimbabwe Publications, Harare, 1994.

Masters, WA. “The Scope and Sequence of Maize Market Reform in Zimbabwe.” Food
Research Institute Studies, Stanford University, 22(1993):227-51,1993.

Maumbe, B.M. and SM. Swinton. “Why Do Smallholder Cotton Growers in Zimbabwe
Adopt IPM?: The Role of Pesticide-related Health Risks and Technology
Awareness.” Selected Paper American Agricultural Economists Association,
Tampa, Florida, July 30t — August 2nd , 2000.

Maxwell, 8., and A., Fernando. “Cash Crops in Developing Countries: The Issues, the
Facts, the Policies.” World Development, l7(1989):l677-1708.

139

Messer, E., M.J. Cohen, and J. D’Costa. “Food From Peace: Breaking the Link Between
Conflict and Hunger.” International Food Policy Research Institute, 2020 Brief
Number 50, 1998.

Mlambo, AS. The Economic Structural Adjustment Program: The Case of Zimbabwe
[990-1995. University of Zimbabwe Publications, Harare, 1997.

Mlambo, K., and S. Kayizzi-Mugerwa. “The Macroeconomics of Transition: Zimbabwe
in the 19805.” African Development Review, 3(1991):47-67.

Mudhara, M., P. Anandajayasekeram, B. Kupfuma, and E. Mazhangara. Impact
Assessment of Cotton Research and the Enabling Environment in Zimbabwe,
[970-1995. Southern Africa Center for Cooperation in Agricultural Research,
Gaborone, 1995.

Ngeno, N.K. Comparative Analysis of Economic Reform and Structural Adjustment
Programs in Eastern Africa: With Emphasis on Trade Policies. Office of
Sustainable Development Bureau for Africa. Technical Paper Number 20,
Nairobi, Kenya, 1996.

Nyamudeza, P., J. Hussein and B. Matibiri. “Zimbabwe Vertisols: Their Properties and
Land Use.” Zimbabwe, undated.

Ogbu, O. M., and M.Gbetibouo. “Agricultural Supply Response in Sub-Saharan Africa:
A Critical Review of the Literature.” African Development Review, 2(1990): 83-
99.

Parhusip, U. “Supply and Demand Analysis of Rice in Indonesia.” Master of Science
Research Paper (Plan B), Department of Agricultural Economics, Michigan State
University, East Lansing, 1976.

Poulton, C., A. Dorward, and J, Kydd. “The Revival of Smallholder Cash Crops in
AfricazPublic and Private Roles in the Provision of Finance.” Journal of
International Development, 10(1998): 85-103.

Poulton, C. “Cotton Production and Marketing in Northern Ghana: The Dynamics of

140

Competition in a System of Interlocking Transactions.” Smallholder Cash Crop
Production Under Market Liberalization: A New Institutional Economics
Perspective. A. Dorward, J. Kydd and C., Poulton, eds. CAB International,
Wallingford, United Kingdom, 1998.

Putterrnan, L. “Economic Reform and Smallholder Agriculture in Tanzania: A
Discussion of Recent Market Liberalization, Road Rehabilitation, and Technology
Dissemination Efforts.” World Development. 23(1995):311-326.

Quarcoo, P.K. “Structural Adjustment Programs in Sub-Saharan Africa: Evolution of
Approaches.” Afiican Development Review, 2(1990):l-26.

Ramaswami, B. “Supply Response to Agricultural Insurance: Risk Reduction and Moral
Hazard Effects.” American Journal of Agricultural Economics.75(l 993):914-45.

Reserve Bank of Zimbabwe. Quarterly Economic and Statistical Review. Harare,

Zimbabwe, 1 6(1 995): 1 -46.

Rohrbach, D.D. The Economics of Smallholder Maize Production in Zimbabwe:
Implications for Food Security. MSU International Development Papers, Number
11, Department of Agricultural Economics, East Lansing, Michigan, 1989.

Rodrik, D. “How Should Structural Adjustment Programs be Designed?” World
Development, 18(1990):933-47.

Rubey, L. “The Impact of Policy Reform on Small-Scale Agribusiness.” Africa and
Rural and Urban Studies. Michigan State University Press, East Lansing,
2(1995):93-119.

Rukuni, M. “The Evolution of Agricultural Policy: 1890-1990.”Zimbabwe ’s
Agricultural Revolution. Mandivamba Rukimi and Carl K. Eicher, eds. University
of Zimbabwe Publications, 1994.

Rukuni, M., and CK. Eicher. “Zimbabwe’s Agricultural Revolution: Lessons for
Southern Africa.” Department of Agricultural Economics, Michigan State
University, East Lansing, Staff Paper No. 93-1, 1993.

141

Sarmiento, C., and PG. Allen. “Dynamics of Beef Supply in the Presence of
Cointegration: A New Test of the Backward-Bending Hypothesis.” Review of
Agricultural Economics. 22(2000):421-37.

Savadogo, K., T. Reardon, and K. Pietola. “Mechanization and Agricultural Supply
Response in the Sahel: A Farm-Level Profit Function Analysis.” Journal of
African Economies. 4(1995)336-77.

Sedjo, RA, and K.S.Lyon. “Timber Supply Model 96: A Global Timber Supply Model
With a Pulpwood Component.” Discussion Paper 96-15, Resources for the Future,
Washington DC, 1996.

Shaffer, J .D., M. Weber, H. Riley, and J. Staatz. “Influencing the Design of Marketing
Systems to Promote Development in Third World Countries: Agricultural
Markets in Semi-arid Tropics.” Proceedings of the International Workshop,

Pantacheru, ICRISAT, 1985.

Shapiro, K. “Efficiency and Modernization in Afi‘ican Agriculture.” Ph. D
Dissertation, Food Research Institute, Stanford University, 1974.

Sithole, G. Analysis of Policy Reform and Structural Adjustment Programs in Zimbabwe:
With Emphasis on Agriculture and Trade. Ministry of Agriculture, Harare,
Zimbabwe, 1996.

Sithole, G ., and EA. Attwood. Structural Aay'ustment and the Agricultural Sector, In
Integrating Food, Nutrition and Agricultural Policy in Zimbabwe: Proceeding of
the First National Consultative Workshop Juliasdale, Zimbabwe, July, 1990.

Smith, L.D. “Structural Adjustment, Price Reform and Agricultural Performance in Sub-
Saharan Africa.” Journal of A gricultural Economics, 40(1989):21-31.

Smith, J .L.G. General Report on the Committee of Inquiry into the Administration of
Parastatals. Government Printers, Harare, Zimbabwe, 1989.

Soludo, C.C. Macroeconomic Policy Modeling of African Economies, Aeena
Publishers, Enugu, Nigeria, 1998.

142

Stewart, F. Adjustment and Poverty: Options and Choices. Routledge, New York, 1995.

Strasberg, P.J. “Smallholder Cash-Cropping, F cod-Cropping and Food Security in
Northern Mozambique: Summary, Conclusions, and Policy Recommendations.”
Working Paper Number 25. MAP/Michigan State University Research Team,
Ministry of Agriculture and Fisheries, Directorate of Economics, Mozambique,
1997.

Stringfellow, R. “Smallholder Out-grower Schemes in Zambia.” Research Report, ODA
Crops Post Harvest Program, Project Number A0439, 1996.

Takavarasha, T. “Agricultural Pricing Policy for Zimbabwe.” Food Security Research in
Southern Afiica: Policy Implications. J.B. Wyckoff and Mandivamba Rukuni,
eds. Univeresity of Zimbabwe/MSU Food Security Research in Southern Africa
Project, 1992.

Townsend, RF, and C. Thirtle. “Dynamic Acreage Response: An Error Correction
Model For Maize and Tobacco in Zimbabwe.” Issues in Agricultural
Competitiveness: Markets and Policies. Roger Rose, Carolyn Tanner and Margot
A. Bellamy, eds. International Association of Agricultural Economists
Occassional Paper Number 7, 1997.

Townsend, RF, and C. Thirtle. “The Effects of Macroeconomic Policy on South African
Agriculture: Implications For Exports, Prices and Farm Incomes.” Journal of
International Development, 10(1998): l 1 7-128.

United Nations. “Agricultural Policies, Prices and Production.” African Development
in A Comparative Perspective. Africa World Press, Inc.Trenton, New Jersey,
1999.

Weissman, S.R. “Structural Adjustment in Africa: Insights from the Experiences of
Ghana and Senegal.” World Development, 18(1990):l621-34.

World Bank. Cotton Production Prospects for the Decade to 2005: A Global Overview.
World Bank Technical Paper Number 231, Washington DC. 1994.

143

World Bank. Cotton Production Prospects for the Next Decade. World Bank Technical
Paper Number 287, Washington DC. 1995.

World Bank. World Development Report 2000/2001: Attacking Poverty. Washington
DC. 2001.

Zimbabwe Cotton Handbook. Commercial Cotton Growers Association, Harare,
Zimbabwe, 1998.

Zimbabwe Government. ZIMPREST: Zimbabwe Program For Economic and Social
Transformation, [996-2000. Harare, 1998.

Zimbabwe National Employment Council. Statutory Instruments for Collective
Bargaining for the Agricultural Industry. Harare, Zimbabwe, 1993-2000,

Zimbabwe. Second F ive- Year National Development Plan, 1 991-1995. Government
Printers, Harare, 1991.

Zimbabwe Ministry of Agriculture. The Agricultural Sector of Zimbabwe, Statistical
Bulletin. Government Printers, Harare, 1999.

144

Appendices

TableA4.1: Results of Diagnostic Tests on the Cotton Acreage Response Models

 

 

 

Test Test Statistic Value(s)
Smallholder Large Scale Critical
Model Commercial Value @ 5% significance
Heteroscedasticity
White test 10.75 14.75 12(9)=16.91;x2(1 1)=19.68
Park test 0.406 -2.35 1.746
Glejser test 0185 0.254 1.753
Arch effects 3.10 13.50 x2(6)=18.54
Functional Form
RESET test 0.48 0.37 5.32; 5.99
MWD test -5. 15 -4.28 1.75
Autocorrelation
Ljung-Box:8 lags 52.19 87.88 x2(l6)=7.96
Q statistic32 :8 lags 21.98 24.41 12(16)=7.96
Breusch-Godfrey (80) test 12.46 11.25 5.99
Berenblutt-Webb g-test33 - 1.09 (d, du)=(0.11; 3.44)
Berenblutt-Webb test 1 .16 - (dL,dU)=(O.22;3 .09)
Cointegration
CRDW test 1.51-1.78 1.53-2.00 0.386

 

 

’2 The Q-statistic is given by Q=nZka (for k=1....m) where n= sample size, m=lag length and pk =
autocorrelation function (ACF]. The Ljung-Box (LB) statistic is defined as LB=n(n+2) Z(p,.2/n-k)~x2,,,
(for k=1....m) (Gujarati, 1995]. The computed Q and LB statistics exceed the critical chi-square values and

we reject the null hypothesis that all pk =0; at least some are none zero.

33 The g statistic is used to test for the presence of perfect positive serial correlation and is computed as
follows: g=Z',=,e,2/X,=,u,2 where ut and e, are OLS residuals from the original and first differenced
models respectively. Since observed g>dL, reject null hypothesis that true p=1 i.e. no perfect positive serial
correlation of first order in both models.

145

Table A4.2: Zimbabwe Cotton Yield Response Model Results, 1980-1997

Dependent variable: Natural log of cotton yield (kg/ha)

 

 

 

 

Large Scale Commercial Smallholder Model
Model

Parameter Estimate z-value Estimate z-value
Scale of operation (farm size)
LAGGED AREA ***-0.01 -4.51 -0.01 -1.06
Product and factor prices
COTTONP ***0.29 3.77 0.02 0.34
TOBACCOP ***-0.34 -5.21 - -
ANFPRICE 0.38 0.71 1“-0.33 -1.94
MIWAGE *3.40 1.76 ***1.92 2.06
MAIZEP - - 0.03 0.96
Macroeconomic and institutional
INTEREST ***-48.00 -3.82 - -
TRAINING 0.06 1.00 *0.10 1.79
Exogenous Shocks
DROUGHT ***-190.00 -2.93 ***-290.00 -9.78
ESAP ***-440.00 -3.25 ***-200.00 -4.27
N 16 17
Log likelihood chi-square -95.04 -90.61
xz—p value 0.0000 0.0000

 

Weance at 1% level, "=significance at 5% leve1,*=significance at 10% level

146

Table A4.3 Agricultural Policy Changes in Zimbabwe, 1980- 2000

 

Year

Policy Change

Effects of Policy

 

Prior to 1980

1982-1985

1986-1990

1990-1995

1996-1999

Cotton marketing controlled
by Cotton Marketing Board;
Growth With Equity policy

unveiled at independence.

Transitional National Development

Plan.

First Five-Year National Development

Plan.

Economic Structural Adjustment
Program (ESAP);export diversification
& promotion strategy; parastatal reform;

privatization and commercialization.

Land Acquisition Act passed.

Second phase of adjustment.

Further market reforms under
ZIMPREST adopted in 1995.
Parastatal, financial & civil

service reforms. Economic downturn;
Fast track land reform.

Millenium Recovery Plan.

GOZ regulated annual cotton producer-
prices. Large-scale commercial

sector dominate seed cotton supply.
Promote more equitable ownership of
natural resources especially land; achieve
sustained economic growth.

Aimed at increasing exports, investments
and employment Formation of SFCS under
AFC; targeted lending to smallholders;
Formation of the Resettlement Farm Sector
(RFS); smallholders assume leading role as
seed cotton producers in the country.
Establishment of several rural cotton depots.
Target growth 5.1% and 28,000 new jobs
per annum.

Use of flexible exchange rate.
Restructuring parastatals CMB, CSC

DMB etc. Formation of COTTCO

Private Limited. Liberalization of
agricultural pricing policy. ZIMACE
commodity clearing house formed.
Oligopolistic cotton marketing e.g.,
CARGILL Inc. COTPRO and
CHIPANGAYI.

Foreign currency shortages. Interest

rate liberalization. Fuel shortages;

Slow growth in agriculture sector.

Cyclone in cotton zones S.E. Lowveld

 

147

Table A4.4: Zimbabwe Data for Cotton Supply Response Analysis, 1980-1997

YEAR LSC SMALL- TRAlNlNG EXCHANGE INTEREST WAGES
ACREAGE HOLDER RATE RATE (ZS)
(HA) ACREAGE (ZS/USS) (%)
(HA)

VD

153,000
1 ,000
l ,000
181,150
1 ,760
217,620
67,5
SOURCE ZMOA

—a__~__—a—l—~
DJ --

 

Table A4.5: Zimbabwe Data for Cotton Supply Response Analysis Continued, 1980-1997

CO N TOBACCO MAIZE SORGHUM FERTILIZER
PRICE (21:) PRICE (ZS) PRICE (ZS) PRICE (ZS) PRICE (AN)
7. . .00 .
1

91. l .
,426.00 140.
4 00.

4, 57.00

,915.00

5,

14,300.
1990/91 18,141.00
1991 14,976.00
1 . 5,
199 . 28,
1 . 60, 1.

1995 . , .
1 . 21,02.
R A&

 

148

Table 1.4.6 Dick-

Variable

/

Smallholder c
Large-scale 3
Nominal colt
Nominal ma’
Nominal tob
Nominal sor
Nominal A.‘
Nominal ex
Nominal 1m
Nominal m
Cotton Trai
Note: lndeper

Table A4.7:
Prices

W
Nominal r
Nominal 5
Nominal .
N0minal 1
Interest r;
EXChange

Table A4.6 Dickey Fuller (DF) unit root test (constant without trend)

 

 

Variable Calculated Critical Value Unit Root

t-value @5% Level Status
Smallholder cotton acreage (ha) 10.61 -2.90 REJECT
Large-scale acreage (ha) 5.74 -2.90 REJECT
Nominal cotton price (ZS/ton) 18.96 -2.90 REJECT
Nominal maize price (ZS/ton) 7.35 -2.90 REJECT
Nominal tobacco price (ZS/ton) 4.27 ~2.90 REJECT
Nominal sorghum price (ZS/ton) 10.36 -2.90 . REJECT
Nominal AN fertilizer prices (ZS/ton) 32.60 -2.90 REJECT
Nominal exchange rate (ZS/USS) 14.63 -2.90 REJECT
Nominal interest rates (%) 7.65 -2.90 REJECT
Nominal minimum wages (Z$) 39.15 -2.90 REJECT
Cotton Training Center graduates 3.77 -2.90 REJECT

 

Note: Independent variables are stationary (or stable) so the series do not exhibit unit root problem

Table A4.7: Zimbabwe average annual growth rates of official prices, 1980-1995

 

 

 

Prices Pre—ESAP Pre-ESAP Post-ESAP Post-ESAP

Nominal (%) Real (%) Nominal (%) Real (%)
Nominal cotton price (ZS/ton) 13.28 -2. 16 30.22 3.34
Nominal maize price (ZS/ton) 7.92 -4.25 61.11 4.91
Nominal sorghum price (ZS/ton) 10.48 -3.43 31.48 -3.28
Nominal AN fertilizer prices (ZS/ton) 15.74 8.07 52.42 42.95
Nominal minimum wages (ZS) 31.68 4.04 21.38 -5.31
Interest rates 6.33 -4.51 33.69 -1.66
Exchange rates 25.97 31.36 50.64 3.47
Annual inflation rate 2.21 15.19

 

149

 

 

 

 

APPENDICES

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Communal Resettlement Small-scale Commercial
Farmers Farmers commercial ‘ Farmers
farmers
Zimbabwe 1 Commercial
Farmer’s Farmer’s
Union (ZFU] I Union (CFU]
L l l 1
Extension Marketing & Processing Cotton Research Cotton
Agritex , Chemical COTTCO, COTPRO & Institute (CRI) Training
Companies Cargill Pvt Ltd Center (CTC)
Domestic
Cotton Consumption
Exports
Local Textile Domestic Oil Animal Feed
Companies Expressors Manufacturers

 

 

 

 

Figure A4.1. Structure of the Cotton Industry in Zimbabwe

 

 

Source: Adapted from “Impact Assessment of Cotton Research and Enabling
Environment in Zimbabwe, 1970- 1 995”

150

 

 

Yield [kg/ha]

Cotton Yleld Trends, 1980-1999

2500.00 .
2000.00 I
1500.00
1000.00 ,

 

Q (1’ 0t Q) Q) Q (I; V (3 %
Q) Q) Q) Q) 9.) g 9 o.) Q
'9 '9 '9 r9 ’9 v9 '3’ \Q’ '9 '99

Years

 

+Communal —l—Large Scale Commercial +National 1

 

 

 

Figure A 4.2: Zimbabwe Cotton Yield Trends, 1980-1999

 

 

Productlon [mt]

Seed Cotton Productlon, 1980-1999

400000.00 .

 

9%

c n, hi '0
as to 9 q
.9 .9 '3 .9 .9

Year

 

 

l-O—Communal +Large Scale Commercial +National '

 

 

Figure A4.3: Zimbabwe Seed Cotton Production Trends, 1980-199

151

 

Relative Real Producer Price Ratio Trends

 

 

 

 

 

Ratio
9%.”.9’99'53’.“
00000000

00000000

 

o 0.. o \ ti '5 9.)
9”W®”»9”r9 qud‘ge'gekge 999.9 to QQQQQQ§NQ¢§§

Years

 

+Cotton-maize +Cotton-tobacco
+Cotton-groundnut +Cotton-sorghum

 

 

Figure A4.4: Zimbabwe Trends in Relative Real Agricultural Producer Prices, 1980-1996

 

Nominal crop price trends

Prlces (Zslton)

1 > .\ »‘» .\.. '1 ' . -
e .. ... ,1. ...1 k
,_ mu‘m‘I-C‘I‘lul‘taA—I‘F 7‘17 AE,‘_

 

 

 

[=0— Cotton + Maize + Sorghum

 

Figure A4.5: Zimbabwe Nominal Crop Producer Prices, 1980-1998

152

 

 

 

Price (ZS/ton)

Zimbabwe real producer price trends

3000 7
2500 ;
2000
1500 ‘
1000 1

 

 

l-o—Cotton —I—Sorghum +Maize l

 

Figure A4.6: Zimbabwe Real Crop Producer Prices, 1980-1998

 

Prices (ZS/ton)

Local and world nominal cotton prices

10000 f
8000
6000 I
4000

 

Q (I; ht b %
’99 v99 '9‘” '99 '99

Year

 

l—o—Zimbabwe cotton prices +Wor1d cotton prices 1

Figure A4.7: Comparison of local and world cotton prices, 1980-1997

153

 

 

 

 

Percentage (%)

Trends In major macroeconomic variables, 1980-1998

    

e3” 49“” 3“” 3"" 49‘5” 49°? 0°“ 49““ 49°? e9” «s°°

Year

 

I—O—Exchange rate +interest rate +inflation rate I

 

 

Figure A4.8: Zimbabwe macroeconomic trends, 1980-1998

 

 

Prlce (ls/ton)

Cotton price trends

    

x " 1‘.

0'1! '2) W ‘b
o‘” e‘b $59., eéefi’pegeqheTeq

Years

{—0— Nominal -l— Real

 

Figure A4.9: Zimbabwe Cotton Producer Price Trends, 1980-1997

154

 

 

 

Maize Producer Price Trends

    

 

 

 

 

    

3500.00 ,.
= 3000.00
§ 2500.00 j
{I 2000.00 ‘
3 1500.00 ;
-‘=’ 1000.00 -
“- 500.00 - .
0.00 "1 ‘ ‘ ‘ - .-
Q 1'0
e“ 3”” e9? .81” of” 433’? 49‘” 43°? 65°
Year
I—O— Nominal —l-Real
Figure A4.10: Zimbabwe Maize Producer Price Trends, 1980-1998
I Minimum wage trends
2000 ,
5 ii
8 1500
g r
K} 1000 .
8
at 500
g t , M, . "»»- r: i"
0 3 ..- -- *5 :: : 9; 5.; =*~._~ r -
Q ‘1- ‘1-
e” e: Estate’s taster”
Years

 

 

1—0— Real wage —I— Nominal wage I

 

Figure A4.11: Zimbabwe minimum wage trends, 1980-1999

155

 

 

 

 

 

 

Percent
09
0

Inflation rate trends

    

“i:

123456789101112131415161718192021
Year

 

{g—gdp deflator +year—on-year cpi_|

 

Figure A4.12: Zimbabwe inflation rate trends, 1980-2000

 

Percent

Real Interest rate tends

 

 

WWWW s

—0— real interest rate

Figure A4.13: Zimbabwe real interest rate (nominal-inflation) 1980-1997

&,§>§

 

156

 

CHAPTER 5
SUMMARY AND CONCLUSION

Africa’s cotton production is facing tremendous challenges at the dawn of the new
century. A number of significant changes are taking place both at the farm and national
policy levels that require close attention from agricultural economists and farmers alike.
Cotton is a strategic export crop throughout Sub-Saharan Africa (SSA) and its successfiil
cultivation is critical for foreign exchange earnings, income diversification and
employment generation. As the continent has been plagued by poor economic
performance, it is in Africa’s best interest to expand cotton production and restore the
buoyancy of its export crop sector. Above all, most of the SSA countries face the major
problem of poverty; the sustained development of cotton production has an important
contribution towards the solution of this dilemma.

The study analyzed both the farm and policy level constraints to expanding cotton
production in Zimbabwe. In Africa, indiscriminate use of pesticides is threatening the
health of poor farmers. Unless efforts are directed at resolving health risk of exposure to
pesticides, long-term investments in the cotton industry in SSA will not payoff. In
addition, cotton productivity has been eroded by the use of ineffective traditional
calendar based spray techniques. The relevance of using fixed spray schedules
irrespective of pest status is now being questioned in Africa. Part of the problem is that
pesticide dependency and misuse has created pest resistance and secondary pest outbreak,
especially the red spider mite in Zimbabwe despite the implementation of the acaricide

rotation scheme.

157

The United Nation’s Food and Agriculture Organization (F A0) in collaboration
with the government of Zimbabwe (GOZ) has embarked on Farmer Field School (FPS)-
based Integrated Pest and Production Management (IPPM) extension efforts as a means
to tackle the “pesticide treadmill” dilemma crippling cotton production in Africa. Little is
known about conditions under which integrated pest management (IPM) innovations can
be diffused in Africa. This study is a leading effort to understand the motivational factors
required to provide the adoption stimulus for cotton IPM. In addition, structural
adjustment programs (SAPS) implemented in the 19905 have raised new questions about
whether or not the reforms are beneficial to Zimbabwe’s cotton farmers. In particular, the
distributional impact of SAPs has not received serious attention in Africa. Despite heavy
reliance on pesticides, declining cotton yields (Figure A4.2), coupled with inadequate
knowledge about the distributional impact of SAPS on large and small-scale farmers who
cultivate one of Africa’s strategic export crop is detrimental to the long term
sustainability of productivity and rural poverty alleviation. This chapter summarizes the
main findings and suggests policy implications to revive cotton production in Africa.

To address the pesticide-induced farmer health risks, pest management and policy
constraints confronting cotton production in Africa, a field research was conducted
between June and December 1999 to collect primary data in Zimbabwe’s two leading
cotton producing districts namely Sanyati and Chipinge. Supporting secondary data was
obtained from the Zimbabwe Ministry of Agriculture (ZMOA) publications, the Cotton
Training Center (CTC) and the Reserve Bank of Zimbabwe (RBZ). The challenges

facing sustainable cotton production in Zimbabwe provides lessons for other African

158

 

 

countries going through the same cycle of pest management and structural adjustment
reforms.

Chapter 2 presented an analysis of the acute health risks attributed to pesticides
use among smallholder cotton growers in Zimbabwe. Using primary data from Sanyati
and Chipinge district, our study demonstrates that pesticide use in smallholder cotton
production has significant adverse effects on farmer health. Pesticide-related acute health
symptoms impose opportunity costs in the form of lost sick days and increased
expenditure on treating health problems which average Z$180 and Z$316 per household
head in 1998/99 in Sanyati and Chipinge respectively (1998/99 exchange rate: Z$38 = 1
US$). For every Zimbabwe dollar spent on pesticides, a minimum of 0.45 and 0.83
dollars is incurred in the form of health costs in Sanyati and Chipinge respectively.

The main result from the cost of illness models is that farmer-reported acute
health symptoms is the key factor in predicting health costs among cotton growers. The
results from the Poisson acute symptom incidence models show that acute health effects
are associated with the use of toxic chemicals, leaking sprayers, label illiteracy, personal
habits like smoking and alcohol intake, and taking meals in the fields after spraying.
Exposure averting and mitigating strategies that significantly reduce the incidence of
acute symptoms are protective clothing and knowledge of first aid, respectively. Contrary
to our expectation, exposure to F FS-based IPM training does not significantly reduce
acute symptom incidences. Surprisingly, farmers who favor reforming calendar-based
spraying are less likely to suffer acute pesticide symptoms.

In terms of policy, we recommend that the key to successful improvement of

farmers’ health in the short-term is limited use of hazardous pesticides by smallholders in

159

order to minimize their negative side effects. Emphasis should be placed on substituting
the most toxic pesticides with ones with greater selectivity and high safety for human
health and the environment. Secondly, evidence of adverse health effects among the
cotton growers suggests the need for government policy intervention in farm worker
safety. Reinforcing farmer use of both relevant protective clothing and equipment, and
safe pesticide storage and disposal practices is a critical intervention strategy to preserve
farmer health in the short-run. Thirdly, labeling policy reform is needed to help label-
illiterate farmers understand better the health hazards of toxic cotton pesticides.
Educating illiterate cotton growers about the negative health impacts of toxic pesticides
will help minimize occupational exposure to pesticides. Fourth, improving farmer health
services should entail changing the current mode of health delivery. Existing agricultural
extension services should help disseminate pesticide-related health information, given
that there is minimal contact between farmers and the formal health care system in
Zimbabwe. Increased efforts should also focus on public policies designed to change
perceptions that view chemical interventions as panacea for all pest problems.

Future research effort should focus on cotton farmers for longer time periods to
establish the true adverse effects of the observed acute and chronic health effects. Our
results suggest that spurring pest management innovations should not be emphasized at
the expense of endangering farmer health. The findings from this study provide a basis
for discussions on the merits of policy reforms in pesticide use in smallholder agriculture
and health care delivery in rural Zimbabwe. Future pest management policy design
should be cast in a much broader context to include both agricultural policies and farmer

health issues.

160

 

In Chapter 3, factors affecting cotton IPM adoption are analyzed. The results
show that awareness of technical practices resulting from participation in FFS extension
training is significantly associated with the adoption of IPM practices. In addition, scale
of production is positively related to IPM practice adoption. Scale provides a relative
measure of potential economic gain from adoption. This suggests scale economies in
cotton IPM use. Pesticide-related health risks however had no significant influence on the
adoption of IPM technology. This evidence implies that the GOZ should expand its use
of F FS and other farmer-to-farmer approaches that diffuse IPPM awareness. Further, the
findings underline the fact that investment in pest management information and
knowledge is important for successful cotton production in Zimbabwe.

An important conclusion from Chapter 4 is that SAPS have led to negative short-
run impacts on LSC cotton growers in Zimbabwe. The results suggest that positive
effects of liberalization are yet to be demonstrated. The initial evidence of the negative
impact of SAPS on LSC cotton growers so far highlight the importance for policy makers
to pay closer attention to crafting compensatory mechanism to alleviate the vulnerability
of the growers. This requires targeted policies to ease the burdens that reforms impose on
the farmers. Realizing that SAPs are multifaceted process, further studies ought to break
down and analyze the impact of specific elements of adjustment packages on supply
response for individual food and cash crops. Such an approach would help policy makers
eliminate any cross contamination from different policy instruments. Moreover, our study
is conducted in aggregate terms but analysis of responses to economic decline at the

household level would further illuminate SAP impacts.

161

The second key result is that the signing of the Mozambique peace accord appears
to have stimulated smallholder cotton acreage in Zimbabwe. The evidence underlines the
vital importance of investment in peace and the spillover benefits in agricultural
production and marketing in Africa. The results also suggest that much of the positive
response in Zimbabwe’s smallholder cotton production is attributed to shifis in
geographic location of production rather than the intensification of existing farming
systems. From policy standpoint, the evidence highlights the fact that peace can be a
major ingredient for progress in agricultural production expansion in Africa.

Finally, the overall research results indicate that there are good reasons to refine
the current reforms in Africa to engender a more vigorous positive supply response. Key
strategies to stimulate cotton production and diminish the risk of crop failure in the short
term requires a major focus on institutional factors such as Farmer Field School-based
IPM training, investment in regional peace efforts, and rural infrastructure development
to lower market access barriers. This is particularly important as opportunities for
expanding production through extending the land frontier become limiting in Africa.
Crafting policies that ameliorate SAP-induced negative effects among the vulnerable
farmers is central for generating a positive cotton supply response in future.

Although our study did not specifically address the benefits and costs of IPM use
under smallholder conditions, future research ought to focus closely on this subject.
Demonstrable IPM benefits could provide incentives for the future uptake of IPM by the
farmers.

The insignificance of pesticide-related health risks in IPM adaption should not be

excuses for policy inaction, but rather a challenge for extension re-design. Serious

I62

thought must be given by policy makers to farmer health risks associated 'with cotton
cultivation practices in Africa. Short term solutions to pesticide-induced acute illnesses
demand raising farmer awareness about health risks and improving on-farm safety as a
precautionary measure when using chemicals. Alternative pest management approaches
that emphasize less use of toxic pesticides should provide the basis for a future pest
management strategy that does not compromise farmer health and the sustainability of
cotton production. Despite the several challenges confronting the continent, SSA remains
a key supplier of world cotton and peace will provide a vital foundation for agricultural

growth in the twenty-first century.

163

r

QUESTION NAIRES

SMALLHOLDER COTTON PRODUCTION AND PEST MANAGEMENT IN ZIMBABWE:

SERIAL NUMBER

LOWVELD AND MIDDLEVELD SURVEY 1999-2000

APPENDIX A]: HOUSEHOLD LEVEL QUESTIONNAIRE

Date of Interview

 

Name of District

 

 

 

Name of Enumerator

DISTRICT
Name of Ward NWARD
Name of Village VILNAME
PROJECT SPONSORSHIP

JOINTLY SPONSORED BY ROCKEFELLER FOUNDATION AND WK KELLOGG

FOUNDATION

164

.3

SECTION 1: HOUSEHOLD DEMOGRAPHIC CHARACTERISTICS

1. Demographic characteristics of HOUSEHOLD HEAD and SPOUSE (S)

 

FAMILY MEMBER AGE GENDER EDUCATIONAL HIGHEST
(YEARS) (M/F) LEVEL LEVEL
(YEARS) EDUCATION

 

HEAD OF 1
HOUSEHOLD

 

FIRST WIFE

 

SECOND WIFE

 

THIRD WIFE

 

WIDOWER

 

WIDOW

 

 

 

 

 

 

 

 

FAMILY MEMBER CODES

 

1: HEAD OF HOUSEHOLD 5=WIDOWER
2= FIRST WIFE 6= WIDOW
3= SECOND WIFE 7=OTHER (SPECIFY)

4= THIRD WIFE

EDUCATIONAL LEVEL CODES

 

1=NO EDUCATION 6=COLLEGE DIPLOMA TRAINING
2=PRIMARY EDUCATION 7=UNIVERSITY DEGREE
3=COMPLETED ZIMBABWE JUNIOR CERTIFICATE 8=POST GRADUATE DEGREE
4=COMPLETED O-LEVEL 9=OTHER (SPECIFY)

5=COMPLETED A-LEVEL

2.How many people are above the age of 14 years in your household? BVADULTS
3.How many people are below the age of 14 years in your household? BLWCHDN
4. Have you received any training related to cotton production in the past 5 years?

1=YES =NO

165

5. If YES, please name the source, nature of training and year of training?

 

SOURCE OF TRAINING NATURE OF YEAR OF
TRAINING TRAINING

 

AGRITEX

 

COTTON TRAINING CENTER (CTC)

 

COTTON COMPANY OF ZIMBABWE
(COTTCO)

 

LOCAL FARMER CLUBS

 

CARGILL PVT LTD

 

COTPRO

 

 

 

 

 

CODES FOR TRAINING IN COTTON AGRONOMIC PRACTICES

1=COTTON PEST SCOUTING

2=PESTICICE SPRAYING

3=RECORD KEEPING/CREDIT MANAGEMENT
4=COTTON PICKING

5=SOIL CONSERVATION

CODES FOR SOURCE OF TRAINING

1=AGRITEX 4.CARGILL
2=COTTON TRAINING CENTER [CTC] 5.COTPRO
3=COTTON COMPANY OF ZIMBABWE 6.LOCAL FARMER CLUBS

7.0THER (SPECIFY)
SECTION 2: OFF-FARM INCOME EARNING ACTIVITIES IN THE HOUSEHOLD

1.15 the head of household formally employed elsewhere? FORMEMPL
1=YES 0=NO (If NO please GO TO QUESTION 5)

2. If YES, state your primary occupation OCCUPTN

 

3.What is your annual gross income from your primary occupation? OCCUPINC

 

. LESS THAN 2310 000

. BETWEEN Z$10 001 AND 2320 000
. BETWEEN Z$20 001 AND Z$30 000
. BETWEEN Z$30 001 AND Z$40 000
. BETWEEN Z$40 001 AND Z$50 000
. BETWEEN 2350 001 AND Z60 000

. MORE THAN Z$60 000

\IQUIJibJN—

166

4. Household income from off-farm activities from Sept lst, 1998 to August 3 lst, I999

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

< ------- l998-----) (- 1999 9
HOUSEHOLD OFF- S O N D J F M A M J J A TOTAL
MEMBER FARM E C O E A E A P A U U U INCOME
ACTIVITY PTVCNBR R YNLG(Z$)
IHead of
Household
2First Wife
3Second Wife
4Third Wife
5.Widow
6.Widower
7. Adult
Child
8. Relative
9.
10.
OFF-FARM ACTIVITIES CODES
1=BASKET WEAVING l 1=ROOF THATCHING
2=BEER BREWING 12=TAILORING
3=BICYCLE REPAIRER l3=SHOE MAKING
4=BRICK MAKING 14= STONE CARVING
5=BUILDING ACTIVITIES 15=TRADING STORE
6=BUTCHER l6=TRADlTIONAL HEALING
7=CARPENTRY 17= WOOD CARVING
8=KNITTING l8=COTTON PICKING
9=MIDWIF E l9=TEA PICKING

10=NEIGHBOURING FARM 20=OTHER (SPECIFY)

167

5. Do you own the following assets?

 

ASSET
DESCRIPTION QUANTITY VALUE [Z5]

 

1. Bicycle

 

2. Bore-hole

 

3. Lorry

 

4. Passenger Vehicle

 

5. Radio

 

6. Television

 

7. Trading store

 

8. Tractor

 

9. Trailer

 

10. Van

 

l 1. Water cart

 

12. Planter

 

l3. Cultivator

 

14. Mould-board Plough

 

15. Wheel burrow

 

 

 

 

 

6. Do you receive remittances in CASH or in KIND from employed relatives and children?
1=YES 0=NO REMTANCE

NOTE TO THE EN UMERATOR: PLEASE CALCULATE THE DOLLAR AMOUNT IF FARMER
RECEIVEDANY REMITTANCES IN KIND IN THE PAST YEAR.

 

7. Estimate amount received in the past year from Sept 1St 1998 to August 31St 1999 [in ZS]
TOTREM

{— 1998 —)<— 1999 —>

 

 

 

 

ll SEPT OCT NOV DEC JAN FEB MAR APR MAY JUNE JUL AUG ll

l l

 

 

 

 

 

 

 

 

 

 

 

 

 

168

 

 

SECTION 3: ON-F ARM MANAGEMENT OF COTTON PRODUCTION

l.Who makes the major decisions about cotton production in this household? HDECISN

I=HEAD OF HOUSEHOLD
2= SPOUSE

3=ADULT CHILD
4=OTHER (SPECIFY)

2. Indicate your cropping program for the past three years?

 

COTTON FIELD I998/99 I997/98

l996/97

 

l. Homestead field

 

2. Cotton field 1

 

3. Cotton field 2

 

4. Cotton field 3

 

 

 

 

 

 

CODES FOR CROP ROTATIONS
l=MAIZE-COTTON
2=SORGHUM-COTTON
3=GROUNDNUTS-COTTON
4=MAIZE-COTTON-SORGUM
5=MAIZE-COTTON-SUNF LOWER
6=MAIZE-COTTON-GROUNDNUTS
7=FALLOW

8= OTHER (SPECIFY)

 

3. Did you cultivate all your fields in 1998/99?
1=YES 0=NO

4.1fNO, did you leave any land lying fallow in 1998/99?
l= YES 0=NO

5. When did you plant your cotton crop in 1998/99?

. SECOND WEEK OF OCTOBER

. THIRD WEEK OF OCTOBER

. LAST WEEK OF OCTOBER

. FIRST WEEK OF NOVEMBER

. SECOND WEEK OF NOVEMBER
. THIRD WEEK OF NOVEMBER

. FOURTH WEEK OF NOVEMBER
. OTHER (SPECIFY)

OOQQLAALQN-d

6. Did you use any hired labor in 1998/99?
1=YES 0=NO

169

FARMCULT

FFALLOW -

DATEPLN T

HIREDLAB

7. If YES, Specify your source of hired labor in 1998/99 HLABSORC

l.SCHOOL CHILDREN

2. LABORERS FROM NEIGHBOURING VILLAGES
3. FORMER “REF UGEES” FROM MOZAMBIQUE
4. LOCAL NON-COTTON FARMERS

5. OTHER (SPECIFY).

8. How would you characterize the availability of hired labor in your community?
LABAVAIL
l ABUNDANT LABOUR FORCE
2.SCARCE LABOUR FORCE
3.NOT AVAILABLE
4.0THER (SPECIFY)

9. Did you use chemicals to control cotton pests in 1998/99? CHEMUSE
1=YES 0=NO

 

10. If NO, what are the reasons for NOT using chemicals? RNOCHEM

1. LACK OF MONEY TO BUY CHEMICALS

2. CHEMICALS ARE DANGEROUS TO FAMILY HEALTH
3. DISCOURAGED 1N IPPM TRAINING

4. DON’T BELIEVE IN CHEMICAL USE

5. OTHER (SPECIFY)

1 1. Have you used chemicals in the past two years to control pests? PASTCHEM
1=YES =NO

12. Number of years you have grown cotton? COTYEARS

1-”
a

170

SECTION 3: SOIL CONSERVATION AND TILLAGE PRACTICES IN COTTON PRODUCTION

I. Do you own any livestock? LIVESTOK
l= YES 0=NO

 

2. What type of animals and how many did you own at the end Of the past season,
1998/99?

 

ANIMAL CLASS NUMBER

 

[Beef Cattle

 

2.Donkeys

 

3 .Goats

 

4.Pigs

 

3. Sheep

 

4. Chickens

 

5. Rabbits
6. Ducks

 

 

 

 

 

3. What was your source of draft power in 1998/99? DRAFT P80

I. OWN DRAFT ANIMALS

2. OWN TRACTOR

3. HIRED “TRACTOR” SERVICES
4. HIRED DRAFT ANIMALS

5. BORROWED DRAFT ANIMALS
6. OTHER (SPECIFY)

NOTE TO THE ENUMERATOR: IF F ARMER HIRED DRAFT ANIMALS OR TRACTOR
PLEASE COMPLETE QUESTION 4 BELOW, IF NOT GO TO QUESTION 5.

 

4. Indicate the source of HIRED DRAFT POWER, number of animal/tractor days and the cost?

DRAFT WER SOURCE DRAFT AN IMAL/ TRACTOR COST OF
DAYS ZS/ACRE

   

l.
2. ANlMA
3. TH R (SPEC

171

5. Among the following TILLAGE PRACTICES, please indicate below the ones you used on

 

 

your cotton crop in 1998/99?
TILLAGE PRACTICE PRACTICE
ADOPTION
1=YES 0=NO
l. MINIMUM/ZERO TILLAGE

 

2. CONTOUR RIDGES

 

3. TIED RIDGES

 

4. WINTER PLOUGH

 

5.CONVENTIONAL PLOUGHING

 

6. MULCHING

 

7.0THER (SPECIFY)

 

 

 

 

6. Did you burn cotton stover in 1998/99?

1 =YES 0=NO

7. If YES, please indicate date of slashing and destruction of cotton stover

TILPRAC

BURN STOV

 

COTTON FIELD

SLASHING DATE DESTRUCTION DATE

 

HOMESTEAD FIELD

 

COTTON FIELD l

 

COTTON FIELD 2

 

COTTON FIELD 3

 

OTHER (SPECIFY)

 

 

 

 

 

172

_.J.fl*_'

SECTION 4. PEST MANAGEMENT PRACTICES

1. Do you have a master farmer ceItificate? FARMCERT
1=YES 0=NO (IF NO, GO TO QUESTION 3)

2.State the year you received this certificate? YEARCERT
3.How many COTTON EXTENSION MEETINGS did you attend in l998/99? COTEXMTG

4. Number of direct contacts with the COTTON EXTENSION WORKER in l998/99?
CONTACTS

5. Who provides you with COTTON EXTENSION ADVICE in this area? SOPEXT

 

l.AGRlTEX
2.COTTON TRAINING CENTER (CTC)
3.COTI‘ON COMPANY OF ZIMBABWE (COTTCO)
4. CARGILL
5. COTPRO
6.LOCAL FARMERS
7. NON-GOVERNMENTAL ORGANIZATION (NGO)
4.0THER (SPECIFY)
6. What year did you FIRST receive pest management advice? F RSTPMGT
7. Who do you consider as your main source of PEST MANAGEMENT INFORMATION?
PESTINFO
l.AGRlTEX
2.COTTON TRAINING CENTER (CTC)
3.COTTON COMPANY OF ZIMBABWE (COTTCO)
4. CARGILL
5. COTPRO
6.LOCAL FARMERS
7. NON-GOVERNMENTAL ORGANIZATION (NGO)
8.AGRICURA
9.RADIO AND TELEVISION (MEDIA)
10.0THER (SPECIFY)
8. Did you practice professional pest scouting in cotton in 1998/99? PROSCOUT
1=YES 0=NO

 

173

9. If YES, describe who scouted, the pest pressure in 1998/99 and your experience scouting for the

 

same pest.
PEST TYPE SCOUTING WHO PEST NUMBER OF
(1=YES 0=NO) SCOUTS PRESSURE YEARS OF
SCOUTING
INSECTS

 

 

1.Heliothis Boll Worm

 

2.Red-Boll Worm

 

3.Pink Boll Worm

 

4.Spiny Boll Worm

 

5.Stainers

1'

 

6.Aphids

 

DISEASE

 

7. Bacterial Blight

 

8. Damping Off

 

9. Boll rot

 

10. WEEDS

 

l 1.RODENTS

 

MITES

 

12.Red Spider mite

 

l 3 .Terrnites

 

 

 

 

 

 

 

COTTON SCOUTING CODES

1=HOUSEHOLD HEAD 2=SPOUSE
4=INPUT SUPPLIER EXPERT 5=EXTENSION WORKER

CODES FOR PEST PRESSURE

 

l.SEVERE 3. LIGHT
2.MEDIUM 4. OTHER (SPECIFY)

174

3=OTHER FAMILY MEMBER
6=OTHER (SPECIFY)

10. How many times did you SCOUT your cotton crop for INSECTS in 1998/99?

INSCOUT
I I. How Often did you SCOUT your cotton crop for DISEASES in 1998/99?

DISCOUT
12. DO you believe scouting for pests is beneficial to you? SCOUTBFT

l= YES 0=NO

ENUMERATOR INSTRUCTION: IF RESPONSE IS YES, GO TO QUESTION 13 AND IF NO, GO
TO QUESTION 14 BELOW.

13. What do you consider the main ADVANTAGE in pest scouting? SCOUTADV

1. MINIMIZES PEST DAMAGE

2. MINIMIZES AMOUNT OF CHEMICALS USED
3. MAXIMIZES NET FARM INCOME

4. REDUCES LABOR COSTS

5. OTHER (SPECIFY)

14. What is the main DISADVANTAGE with pest scouting? SCOUTDIS

l. LABOR INTENSIVE

2. NO SIGNIFICANT YIELD GAIN
3.TOO MANY PESTS TO SCOUT
4. DO NOT UNDERSTAND IT

5. NO DISADVANTAGE

6. OTHER (SPECIFY)

15. How do you determine when to apply INSECTICIDES (strategy for pest management)?
CHEMTIMG

1.WEEKLY

. EVERY TWO WEEKS

. ONLY AFTER SCOUTING FOR PESTS

. SPECIFIC GROWTH STAGES OF THE PLANT

. BASED ON PREVAILING PEST PROBLEMS

. BASED ON RECOMMENDATION FROM INPUT SUPPLIERS
. ABITRARILY

. OTHER (SPECIFY)

OOQO‘Ut-bbJN

16. What is your main source of water for cotton spraying? WATPRO

I. OWN WELL

2. OWN BOREHOLE

3. COMMUNAL BOREHOLE
4. WATER HARVESTING

5. NEARBY RIVER

6. DAMS

7. OTHER (SPECIFY)

175

17. How far is your water source from your main cotton field (meters)?

18. State how you transport water to the cotton fields.

l.OWN WATER CART
2. HIRE WATER CART
3. WHEEL BARROW
4. OTHER (SPECIFY)

19. Who made chemical application most of he time in 1998/99?

1. HEAD OF HOUSEHOLD (FARMER).

2. SPOUSE

3. ADULT FAMILY MEMBER
4. HIRED LABOR

5. OTHER [SPECIFY]

20. Do you OWN any sprayers?

1=YES 0=NO (IF NO, GO TO QUESTION 24)

WATERDST

WATRAN SP

APLCATOR

OSPRAYER___

21. Describe the type of sprayer, number in each category owned and cost of purchase.

 

TYPE OF SPRAYER QUANTITY YEAR OF
PURCHASE

PRICE OF TOTAL
SPRAYER (ZS) COST (ZS)

 

1 =Knap-sack

 

2=Ultra-Low Volume (ULV)

 

 

 

 

 

3=Other (specify)

 

 

22 Did you RENT any sprayers from other farmers in 1998/99?
1=YES 0=NO

176

SPRYRENT

 

 

23.1fYE

TYPE
SPRA
HIRE]

 

lKna]

 

2.L’ltrz
Volurr

‘ I

 

 

TOTA

 

 

24. \Vhat

 

 

 

 

25. D0 }
dIsease i
1=YES

 

23. If YES, indicate below the type of sprayers, number hired and the hiring cost in 1998/99.

 

TYPE OF
SPRAYER
HIRED

NUMBER OF
SPRAYERS
HIRED (1)

NUMBER OF
DAYS HIRED

(2)

RATE PAID
[Z$/ DAY]
(3)

TOTAL COST [ZS] (4)
(l)*(2)*(3)

 

l .Knap-sack

 

2.Ultra-Low
Volume

 

3. Other

 

 

TOTAL COST

 

 

 

 

 

24. What methods did you use to control WEEDS in your cotton crop last year?

 

WEED CONTROL METHOD

METHOD USE (YES=I/NO=O)

 

1= Hand hoe

 

2= Ox-drawn cultivator

 

3=Pre-emergence herbicides

 

 

4=Post-emergence herbicides

 

 

25. Do you know any TRADITIONAL METHODS that are used to control cotton pests (weeds, insects or

disease).
I =YES

0=NO

IKNOW

177

 

 

26. Indigenous control methods used to suppress pests, diseases and weeds

 

PEST TYPE DESCRIBE TRADITIONAL/ INDIGENOUS CONTROL
METHOD

 

l. Weed type

 

2. Insects type

 

3. Disease type

 

4. Rodents type

 

5. Nematodes type

 

6. Other [specify]

 

 

 

 

SECTION 5: ACCESS TO, AND USE OF CREDIT AND FERTILIZER IN COTTON

 

PRODUCTION

I. Did you apply for agricultural credit in l998/99? CREDTAPP

1=YES 0=N0 .
2. Did you receive agricultural credit in l998/99? CREDTRCV l-ir
1=YES 0=NO (GO TO QUESTION 6) l

178

. Who provided you with credit and describe the type of credit received in I998/99?

 

CREDIT SOURCE

CREDIT RECEIVED 1=CASH 2=KIND
3=BOTH

CREDIT VALUE
(Z$)

 

l=Agricultural Finance
Corporation (AFC)

 

2=Cotton Company of
Zimbabwe (CO'I'TCO)

 

3=Agricura

 

4=Cargill Zimbabwe Pvt. Ltd.

 

5=Savings club

 

6=Cotpro

 

7=Cooperative

 

8=Relatives

 

9=Local Rural Traders

 

 

I O=Other (specify)

 

 

 

. State the Specific type of credit received in 1998/99

1. SEASONAL INPUTS (FERTILIZER. CHEMICALS AND SEEDS)

CREDITWHY

2. MACHINERY AND IMPLEMENTS (TRACTORS. PLOUGH, SPRAYERS ETC)
3. WATER DEVELOPMENT (BOREHOLE, WATER CART ETC)
4. FARM DEVELOPMENT (FENCING MATERIAL ETC)

. Did you receive your SEASONAL INPUTS on time in l998/99?

I. ALWAYS ON TIME

2. SOME INPUTS RECEIVED ON TIME

3. LATE DELIVERY

4. NEVER DELIVERED

5. OTHER (SPECIFY)

179

SEASOINP

 

6. Indicate other years that you have received cotton credit in the past 5 years?

 

YEAR CREDIT RECEIVED 1=YES 0=NO CREDIT SOURCE

 

1997/98

 

l996/97

 

1995/96

 

l 994/95

 

1992/93

 

 

 

 

 

 

CODES FOR CREDIT SOURCES

1=AGRICULTURAL FINANCE CORPORATION 6=COTPRO ,
2=COTION COMPANY OF ZIMBABWE (COTTCO) 7=COOPERATIVE

3=AGRICURA 8=RELATIVES

4=CARGILL ZIMBABWE PVT. LTD. 9=LOCAL RURAL TRADERS

5=SAVINGS CLUB lO=OTHER (SPECIFY)

7. State for which additional crop(s) you obtained credit in 1998/99? CREDTUSE

 

 

 

 

 

CROP ENTERPRISE CREDIT OBTAINED VALUE OF l
1=YES 0=NO CREDIT
OBTAINED [Z5]
1. MAIZE
2. SORGHUM
3. SUNFLOWER

 

4. GROUNDNUT

 

 

 

6.

 

 

 

 

 

8. Do you have any problems with your current input contract (credit) scheme?
Explain

 

 

 

 

180

SECTION 6:SMALLHOLDER COTTON AND FOOD CROP-INTERACTIONS

1.Describe other crops grown on the farm and the estimated area planted to each crop in 1998./99.

 

CROP ENTERPRISE ESTIMATED AREA PLANTED (HA)

 

l=Groundnuts

 

2=Maize

 

3=Sorghum

 

4=Sunflower

 

5=Tobacco

 

6=Other (Specify)

 

 

 

 

 

2. If you had additional land, what crop would you plant? PLTADD

 

l .COTTON
2.MAIZE

3 .SORGHUM
4.GROUNDNUTS
5.SUNFLOWER

3.Estimate the distance of your home to the nearest cotton depot [km]? CMKTDIST
4. How did you transport your cotton crop to the market in l998/99? COTRANSP
I. OWN TRANSPORT
2. HIRED PRIVATE TRANSPORT
3. COTTON COMPANY OF ZIMBABWE (COTTCO)
4. CARGILL
5. COTPRO
6. OTHER (SPECIFY)
5. What was the average cost of transporting cotton to the market per bale in l998/99 (ZS)?

COTCOST

181

6. What is the amount of income obtained from other cropping activities?

 

 

 

 

 

 

 

 

 

 

 

 

 

CROP SALES QUANTITY QUANTITY AVERAGE PRICE TOTAL
HARVESTED SOLD (KG) RECEIVED REVENUE
(KG)(BAGS) (BAGS) (ZS/BAG) (ZS)
l. Maize
2. Sorghum
3. Groundnut
4. Sunflower
5. Tobacco
6. Other
7. Did you have storage problems with your COTTON in l998/99? STOPROB
1=YES 0=NO
8. Describe the nature of the storage problem. STOPROB

l. STORAGE PESTS PROBLEMS
2. LACK OF ADEQUATE ON-FARM STORAGE SPACE
3. LACK OF STORAGE CONTAINERS (BAGS, BALES ETC?)

4. THEFT

5. COTTON ACCIDENTALLY BURNT IN STORE-ROOM
6. OTHER (SPECIFY).

182

 

SECTION 7: FUTURE STRATEGIES IN COTTON PRODUCTION

I.ldentify what you consider as TWO LEADING fiIture strategies in cotton production from list below?

 

FUTURE STRATEGY LEADING
STRATEGY
(1=YES 0=NO)

 

[Increase cultivated area under cotton

 

2.Receive training on IPPM techniques

 

3.Hire more labor for cotton production

 

4.Reduce use of pesticides in cotton production

 

5. Build own well/bore-hole

 

6. Acquire own transport

 

7. Acquire own draft power

 

8. Use more credit in future

 

9. Use more fertilizer on cotton

 

IO. Adopt land conservation practices in cotton

 

l 1. Invest in cotton storage shed

 

12. Adopt/abandon crop rotations

 

13. Expand cultivated area under food crops eg maize and or sorghtim

 

I4. Abandon cotton production

 

 

 

 

2. What do you consider as the main problem you face in cotton production today?
COTPROB
I. PESTS MANAGEMENT
2. LACK OF CREDIT
3. DROUGHT
4. POOR VARIETIES
5. POOR MARKET PRICES
6. EXPLOITATION BY CONTRACTORS
7. OTHER (SPECIFY)

183

3. What are your views regarding the improvement of pest management strategies in smallholder cotton
production? PMGTVIEW

l. MAINTAIN CURRENT STRATEGIES

2. NEED NEW AND BETTER METHODS

3. MORE PEST MANAGEMENT TRAINING FOR FARMERS
4. NO IDEA

5. OTHER (SPECIFY)

4.Would you be willing to grow a new variety of cotton [Bt] that does not require the use of pesticides?
BTCOTTON

I. WOULD TRY IT IMMEDIATELY

2. NOT SURE

3. CURRENT VARIETIES ARE SATISFACTORY
4. PESTICIDES CONTROL MORE EFFECTIVE

5. If farmer wants to try Bt cotton, does farmer consider any of the following a major deterrent.
BTPROBLM
I. PLANTING Bt COTTON & CONVENTIONAL COTTON SIDE BY SIDE

2. LONG TERM EFFECT OF RED BOLL WORM ON Bt COTTON UNKNOWN
3. OTHER [SPECIFY]

THANK YOU VERY MUCH

=EN D OF HOUSEHOLD QUESTIONNAIRE:

184

APPENDIX A2: HOUSEHOLD MEMBER HEALTH QUESTIONNAIRE

SERIAL NUMBER

SMALLHOLDER COTTON PRODUCTION AND PEST MANAGEMENT IN ZIMBABWE:
SANYATI AND CHIPINGE DISTRICT SURVEY 1999-2000

Date of Interview

 

Name of Enumerator

Name of Farmer

 

 

 

Name of District DISTRICT

Name of Ward NWARD

Name of Village VILNAME
PROJECT SPONSORSHIP

JOINTLY SPONSORED BY ROCKEFELLER FOUNDATION AND WK KELLOGG
FOUNDATION

185

SECTION 1: HEALTH EFFECTS OF PESTICIDE USE IN SMALLHOLDER COTTON
PRODUCTION

1. Has any one in your family ever experienced the following SYMPTOMS afier spraying
pesticides. SPRAYSYM

NAUSEA
VOMITING
ABDOMINAL PAINS
DIARRHEA
BLURRED VISION
DlZZINESS

NASAL BLEEDING
NONE

WNQVPP‘NT‘

2. Has anyone in your family ever had PESTICIDE STOMACH POISONING problem in the past year?

PPOISON
0=NO 1=YES 2=DON’T RECALL

186

3. If YES, indicate in table below the individual in your family who has been affected by PESTICIDE
STOMACH POISONING in the past year?

 

 

 

 

 

 

 

 

 

 

 

 

 

5= OTHER (SPECIFY)

HOUSEHOLD STOMACH POISONING NUMBER OF INCIDENCES IN
MEMBER AFFECTED SEVERITY l998/99
TOTAL
. HOUSEHOLD MEMBER CODES CODES FOR SEVERITY OF PESTICIDE
WIN—G

1: HEAD OF HOUSEHOLD 1=DEADLY

2= SPOUSE 2=VERY SEVERE

3= CHILDREN 3=MILD TO MODERATE

4= RELATIVE 4=DO NOT KNOW

5=OTHER (SPECIFY)

4. Has anyone in your household ever had SKIN IRRITATIONS in the past year after using chemicals?

1 =YES 0=NO

SKIN IRRIT

187

 

5. If YES please indicate in table below the individual in your household who has been affected
by SKIN IRRITATIONS in the past year as well as the severity and number of incidences
of this condition.

HOUSEHOLD SKIN N UMBER OF
MEMBER IRRITATIONS IN CIDENC ES
SEVERITY

 

 

 

TOTAL

HOUSEHOLD MEMBER CODES CODES FOR SEVERITY OF SKIN IRRITATIONS
1: HEAD OF HOUSEHOLD I=DEADLY

2= SPOUSE 2=VERY SEVERE

3= CHILDREN 3=MILD TO MODERATE

4= RELATIVE 4=DO NOT KNOW

5= OTHER (SPECIFY) 5=OTHER (SPECIFY)

6. Have you had EYE IRRITATIONS in the past year after using pesticides?
1=YES 0=N0 EYEIRRIT

7. If YES, please indicate in table below the individual in your household who has been affected
by EYE IRRITATIONS in the past year as well as the severity and number of incidences
of this condition.

H SEHOLD EYE NUMB F
MEMBER IRRITATION INCIDENCES
SEVERITY

 

 

 

HOUSEHOLD MEMBER CODES CODES FOR SEVERITY OF SKIN IRRITATIONS
l= HEAD OF HOUSEHOLD I=DEADLY

2= SPOUSE 2=VERY SEVERE

3= CHILDREN 3=MILD TO MODERATE

4= RELATIVE 4=DO NOT KNOW

5= OTHER (SPECIFY) 5=OTHER (SPECIFY)

188

NOTE TO ENUMERATOR: Please complete the following table if household member was sick for a
number of days and or received treatment from nearby health center or hospital for above reported

symptoms.

8. Indicate the symptoms, number of days sick, treatment received and number of visits and cost of
treatment in the past year.

 

HOUSEHOLD PESTICIDE NO TREATED NO. OF COSTNISIT TOTAL

MEMBER AILMENT OF 1=YES CLINIC (ZS) COST (ZS)
DAYS 0=NO VISITS
SICK

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

NOTE TO ENUMERATOR: Request farmer to check one’s medical record cards if available to
verify information on treatment costs..

 

 

HOUSEHOLD MEMBER CODES CODES FOR PESTICIDE AILMENTS.

1= HEAD OF HOUSEHOLD l=SKlN IRRITATION

2= SPOUSE 2=EYE IRRITATION

3= CHILDREN 3=STOMACH POISONING

4= RELATIVE 4=BACK PROBLEMS (CARRYING KNAPSACK FOR

PROLONGED PERIODS, WATER)

189

9.Does any one in your family have the following medical condition (long term medical effects)

 

 

 

 

 

 

 

 

 

 

 

 

HOUSEHOLD LONG TERM MEDICAL DURATION OF ILLNESS
MEMBER CONDITION
HOUSEHOLD MEMBER CODES CODES FOR LONG TERM PESTICIDE RELATED
ILLNESS.
l= HEAD OF HOUSEHOLD I= CANCER
2= SPOUSE 2= BLINDNESS
3= CHILDREN 3= BACK PROBLEMS
4= RELATIVE 4=LUNG DAMAGE
5= OTHER SPECIFY
10. Do your sprayers LEAK on your back when you apply chemicals? SPLEAK
1=YES 0=NO

l I. What is your main source of HEALTH OR SAFETY INFORMATION on pesticides use?

HSINFO

 

 

D
l. AGRICURA 7. VILLAGE HEALTH WORKER
2. AGRITEX 8. NATIONAL RADIO
3. COTTCO 9. LOCAL TELEVISION
4. CARGILL 10. ClBA-GEIGY
5. OTHER FARMERS I 1. OTHER (SPECIFY)
6. LOCAL NEWSPAPERS
12. Does anyone in your household consume alcohol? ALCOHOL
1=YES 0=NO

190

13. If YES, please indicate below the individual, amounts consumed as well as duration of alcohol
consumption.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

HOUSEHOLD ALCOHOL ALCOHOL DURATION OF
MEMBER CONSUMPTION CONSUMED ALCOHOL
PER WEEK CONSUMPTION
[LITRES] [YEARS]
HOUSEHOLD MEMBER CODES CODES FOR ALCOHOL CONSUMPTION
’ l= HEAD OF HOUSEHOLD I=LARGER (CASTLE, PILSNER, LION, BLACK LABEL)
2= SPOUSE 2=SPIRITS (WHISKY, KACHASU)
3= CHILDREN 3=TRADITIONAL BREW (CHIBUKU OR SCUDS)
4= RELATIVE 4=OTHER
14. Does anyone in this household smoke? HHSMOKE
1=YES 0=NO

15. Please provide information regarding SMOKING by household members.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

HOUSEHOLD TYPE OF CIGARETTES CIGARETTES/WEEK SMOKING
MEMBER SMOKED NUMBER DURATION
- [YEARS]
HOUSEHOLD MEMBER CODES CODES FOR CIGARETTERS
1: HEAD OF HOUSEHOLD 1=MADISON
2= SPOUSE 2=KINGSGATE
3= CHILDREN 3=EVEREST
4= RELATIVE 4=BERKELY

5=SHAMROCK

191

15. Indicate below any member of the household who used to drink alcohol in the past PDRINK

l= HEAD OF HOUSEHOLD
2= SPOUSE

3= CHILDREN

4= RELATIVE

l6. Identify any member of the household who used to smoke in the past. PSMOKE

1= HEAD OF HOUSEHOLD

2= SPOUSE
3= CHILDREN
4= RELATIVE

SECTION 2: PESTICIDE SAFETY AND HANDLING

17. Where do you normally store your agricultural chemicals CHEMGT

l.STOREROOM FOR CHEMICALS ONLY
2.STOREROOM FOR FOOD CROPS
3.KITCHEN

4.BEDROOM

5.0THER (SPECIFY)

18. Do you own any of the following protective clothing?

 

NOTE TO ENUMERATOR: You may ask the farmer to show you some of these protective clothes in
order to prove beyond any doubt that the farmer has the required protective gear.

192

19. Please rank on the basis of color of the triangle which is the MOST DANGEROUS and LEAST
DANGEROUS pesticide used in cotton pest management.

ARNING SIG

 

COIES FOR COLOR CATEGORIZATION OF PESTICIDES
l.VERY DANGEROUS PESTICIDE

2.DANGEROUS PESTICIDE

3.AVERAGELY DANGEROUS PESTICIDE

4. LEAST DANGEROUS PESTICIDE

 

20. Do you sometimes have meals in the fields on the same day that you spray chemicals in your
cotton fields? EATSPRY

1=YES =NO

21. Do you sometimes smoke in the fields while you spray your cotton crop? SMOKSPRY

1=YES 0=NO
22. What do you do to the pesticide containers after use? PESTDSPO
1.DESTROY OR BURN
2.USE AS CONTAINERS FOR DRINKING WATER
3.RETURN TO CHEMICAL DEALERS
4.USED FOR FOOD STORAGE
S.NOTHING
6.0THER (SPECIFY)
23. Do you understand FIRST AID? . FAID
1=YES 0=NO
24. Do you have a FIRST AID KIT in your household? FKTAID

l =YES 0=NO

25. How far are you from the nearest health center DISTHCTR

THANK YOU VERY MUCH

=—-END OF HEALTH QUESTIONNAIRE“

193

 

APPENDIX A3: FIELD LEVEL QUESTIONNAIRE
SERIAL NUMBER
SMALLHOLDER COTTON PRODUCTION AND PEST MANAGEMENT IN ZIMBABWE:
CHIPINGE AND SANYATI DISTRICT SURVEY 1999-2000

Name of Enumerator

 

Date of Interview

 

Name of Farmer

 

 

 

Name of District DISTRICT

Name of Ward NWARD

Name of Village VILNAME
PROJECT SPONSORSHIP

JOINTLY SPONSORED BY ROCKEFELLER FOUNDATION AND WK KELLOGG
FOUNDATION

194

 

 

 

 

SECTION I: FIELD LEVEL MANAGEMENT OF COTTON PRODUCTION

1. Indicate the cotton variety grown and area planted to cotton in l998/99 in each field?

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1=YES

0=NO

195

COTTON FIELD COTTON ESTIMATE ESTIMATE REPLANTED FDXNTED
VARIETY AREA(HA) AREA (HA) COTTON ? COTTON?
PLANTED FALLOW YES=l/NO=0 1=YES/0=NO
l=HOMESTEAD
2=FIELD # I
3=FIELD # 2
4=FIELD # 3
CODES FOR COTTON VARIETIES GROWN IN ZIMBABWE
1=ALBAR 0502 10. CY 889
2=ALBAR K603 I l. ALBAR EU8910
3=DELMAC 12. ALBAR DF 885
4=ALBAR GSOI 13. ALBAR FQ902
5=ALBAR K602 l4. ALBAR FQ904
6=ALBAR 888714 15. LONG STAPLE 797
=ALBAR AG4869 l6. ALBAR SZ 93—14
8=ALBAR BC853 17. AF (88)4
9=ALBAR EU883GSOZ
2. Describe the planting densities for each cotton field
COTTON COTTON IN - COTTON QUANTITY COST OF
FIELDS ROW SPACING BETWEEN ROW COTTON COTTON
(CM) SPACING SEEDS SEED -
(CM) APPLIED (Z$)
(KG)
l.HOMESTAED
2.FIELD # l
3.FIELD # 2
4.FIELD #3
3. Did you use any hired labor in l998/99? HIREDLAB

 

 

4. Did you practice LABOR SHARING in cotton production in 1998/99? LABSHARE
1=YES 0=NO

NOTE TO THE NUMERATOR:
I.PLEASE COMPLETE SHEET NUMBER 1, 2, 3 AND 4 FOR LABOUR USE IN COTTON
FIELDS.

 

2.IF LABOR IS REWARDED IN KIND CALCULATE THE EQUIVALENT DOLLAR VALUE OF
OUTPUT PAID AS WAGE FOR WORK RENDERED USING l998/99-PRODUCER PRICE FOR
THE PARTICULAR CROP.

5. Indicate the amount of family and any hired labor used in cotton production as well as the TOTAL wage
paid to hired labor for the listed field operations in l998/99?

HOMESTAED FIELD: l998/99 OPERATIONS SHEET #1

 

TASK TIME FAMILY HIRED HIRED SHARED
(DAYS) LABOUR LABOUR LABOUR LABOUR
NUMBER NUMBER TOTAL NUMBER?
WAGE
(ZS)

 

l= Land preparation

 

2=Planting cotton

 

3=Basal fertilizer application

 

4=Top dressing fertilizer

 

5=Scouting for pests

 

6=Mechanical weeding

 

7=Pre-emergence herbicides

 

8=Post emergence herbicides

 

9=lnsecticides application

 

10=Rodenticides application

 

l 1=Cotton Picking

 

12=Buming Stover

 

 

 

 

 

 

 

 

196

 

COTTON FIELD # l: l998/99 OPERATIONS SHEET #2
TASK TIME FAMILY HIRED HIRED SHARED
(DAYS) LABOUR LABOUR LABOUR LABOUR
NUMBER NUMBER TOTAL NUMBER?
WAGE
(ZS)

 

l= Land preparation

 

2=Planting cotton

 

3=Basal fertilizer application

 

4=Top dressing fertilizer

 

5=Scouting for pests

 

6=Mechanical weeding

 

7=Pre-emergence herbicides

 

8=Post emergence herbicides

 

9=lnsecticides application

 

10=Rodenticides application

 

1 l=Cotton Picking

 

 

12=Buming Stover

 

 

 

 

 

 

 

197

COTTON FIELD # 2: 1998/99 OPERATIONS SHEET #3

 

TASK TIME FAMILY HIRED HIRED SHARED
(DAYS) LABOUR LABOUR LABOUR LABOUR
NUMBER NUMBER TOTAL NUMBER?
WAGE
(Z3)

 

I= Land preparation

 

2=Planting cotton

 

3=Basal fertilizer application

 

4=Top dressing fertilizer

 

5=Scouting for pests

 

6=Mechanical weeding [

 

7=Pre-emergence herbicides

 

8=Post emergence herbicides

 

9=lnsecticides application

 

10=Rodenticides application

 

ll=Cotton Picking

 

12=Buming Stover

 

 

 

 

 

 

 

 

 

198

COTTON FIELD # 3: l998/99 OPERATIONS SHEET #4

 

TASK TIME FAMILY HIRED HIRED SHARED
(DAYS) LABOUR LABOUR LABOUR LABOUR
NUMBER NUMBER TOTAL NUMBER?
WAGE
(Z$)

 

1: Land preparation

 

2=Planting cotton

 

3=Basal fertilizer application

 

4=Top dressing fertilizer

 

5=Scouting for pests

 

6=Mechanical weeding

 

7=Pre-emergence herbicides

 

8=Post emergence herbicides

 

9=lnsecticides application

 

10=Rodenticides application

 

l 1=Cotton Picking

 

12=Buming Stover

 

 

 

 

 

 

 

 

6. Did you use fertilizer on your cotton crop in l998/99? COTFERT
1=YES 0=NO

199

7.Describe fertilizer use on your cotton crop in l998/99?

HOMESTAED FIELD: I998/99 FERTILIZER APPLICATION BY FIELD

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

TYPE OF AREA NUMBER OF PRICE TOTAL
FERTILIZER APPLIED FERTILIZER (ZS/BAG) COST (ZS)
APPLIED (HA) BAGS APPLIED
l=Ammonium Nitrate
(AN)
2=Compound K
3=Compound L
4= Lime
5= Urea
6
7.
COTTON FIELD #l: I998/99 FERTILIZER APPLICATION BY FIELD
TYPE OF AREA NUMBER OF PRICE TOTAL
FERTILIZER APPLIED FERTILIZER (ZS/BAG) COST (ZS)
APPLIED (HA) BAGS APPLIED

 

l=Ammonium Nitrate
(AN)

 

2=Compound K

 

3=Compound L

 

4= Lime

 

5= Urea

 

 

 

 

 

 

 

 

 

200

COTTON FIELD #2: I998/99 FERTILIZER APPLICATION BY FIELD

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

TYPE OF AREA NUMBER OF PRICE TOTAL
FERTILIZER APPLIED FERTILIZER (ZS/BAG) COST (ZS)
APPLIED (HA) BAGS APPLIED
l=Ammonium Nitrate
(AN)
2=Compound K
3=Compound L
4: Lime
5= Urea
6.
7
COTTON FIELD #3: l998/99 FERTILIZER APPLICATION BY FIELD
TYPE OF AREA NUMBER OF PRICE TOTAL
FERTILIZER APPLIED FERTILIZER (ZS/BAG) COST (ZS)
APPLIED (HA) BAGS APPLIED

 

l=Ammonium Nitrate
(AN)

 

2=Compound K

 

3=Compound L

 

4: Lime

 

5= Urea

 

 

 

 

 

 

 

 

201

 

 

8. Did you apply organic manure to your cotton crop in l998/99 season?

1=YES 0=NO

OGMANURE

9. If YES, please indicate the field and amount of organic manure applied in I998/99?

 

COTTON
FIELD

ORGANIC MANURE APPLIED
[1=YES/ 0=NO]

QUANTITY
APPLIED

 

l=HOMESTEAD FIELD

 

2=FIELD #l

 

3=FlELD #2

 

4=FIELD #3

 

 

 

 

202

 

10. What are the types of chemicals that you applied on your cotton crop in l998/99?

HOMESTEAD: AGRO—CHEMICAL USE IN 1998/99 SHEET # l

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

CHEMICAL TARGET NUMBER OF QUANTITY OF UNIT COST PER
NAME PEST APPLICATIONS CONCENTRATE SIZE UNIT (ZS)
(KG/ML)
l. Carbaryl
2. Thiodan
3. Larvin
4. Fenvalerate
5. Agrithrin super
6. Karate
7. Dimethoate
8. Rogor
9. Marshall
10. Pfumo
I l.
Monocrotophos
HERBICIDES
CODES FOR PESTICIDES
Herbicides Aca ricides Fungicides
12. Trif l3. Cucacron I4. Calirus
15 Stomp I6. Hosthion l7. Tecto
l8 Bladex l9. Azodrin/Nuvacron 20. Brassicol
21 Gesagard 22. Tedion/Tetradifon 23. Rizolex
24 Dual 25. Mitac 26. Monceren combi
27 Planavin 28. Secure 29. Vitavax
30 Cotoran 31. Kelthane 32 Quintozene
33 Cotogard 34. Omite
35 Gesapax
36 Diuron
37 Igran

203

 

COTTON FIELD #l: AGRO-CHEMICAL USE IN 1998/99 SHEET # 2

 

 

 

 

CHEMICAL TARGET NUMBER OF QUANTITY OF UNIT COST PER

NAME PEST APPLICATIONS CONCENTRATE SIZE UNIT (ZS)
(KG/ML)

1. Carbaryl

2. Thiodan

3. Larvin

 

4. Fenvalerate

 

5. Agrithrin super

 

6. Karate
7. Dimethoate
8

 

 

 

. Rogor

 

9. Marshall ,

 

10. Pfumo

l l.
Monocrotophos

HERBICIDES

 

 

 

 

 

 

 

 

 

 

 

 

 

CODES FOR PESTICIDES

 

Herbicides Acaricides Fungicides

12. Trif l3. Cucacron I4. Calirus

16 Stomp l6. Hosthion 17. Tecto

18 Bladex l9. Azodrin/Nuvacron 20. Brassicol :-
21 Gesagard 22.Tedion/Tetradifon 23.Rizolex I
24 Dual 25. Mitac 26. Monceren combi F
27 Planavin 28. Secure 29. Vitavax L
30 Cotoran 31. Kelthane 32 Quintozene ‘
33 Cotogard 34.0mite

35 Gesapax

36 Diuron

37 Igran

204

 

COTTON FIELD #2: AGRO-CHEMICAL USE IN I998/99 SHEET # 3

 

 

 

 

CHEMICAL TARGET NUMBER OF QUANTITY OF UNIT COST PER

NAME PEST APPLICATIONS CONCENTRATE SIZE UNIT (ZS)
(KG/ML)

l. Carbaryl

2. Thiodan

3. Larvin

 

4. Fenvalerate

 

5. Agrithrin super

 

6. Karate
7. Dimethoate

 

 

8. Rogor

 

9. Marshall

 

10. Pfumo

I I.
Monocrotophos

HERBICIDES

 

 

 

 

 

 

 

 

 

 

 

 

 

CODES FOR PESTICIDES

 

 

Herbicides Acaricides Fungicides
12. Trif l3. Cucacron l4. Calirus

17 Stomp 16. Hosthion 17. Tecto

l8 Bladex l9. Azodrin/Nuvacron 20. Brassicol
21 Gesagard 22. Tedion/Tetradifon 23. Rizolex
24 Dual 25.Mitac 26. Monceren combi
27 Planavin 28. Secure 29.Vitavax
30 Cotoran 31. Kelthane 32 Quintozene
33 Cotogard 34. Omite

35 Gesapax

36 Diuron

37 Igran

205

COTTON FIELD #3: AGRO—CHEMICAL USE IN l998/99 SHEET # 4

 

 

 

 

CHEMICAL TARGET NUMBER OF QUANTITY OF UNIT COST PER

NAME PEST APPLICATIONS CONCENTRATE SIZE UNIT (ZS)
(KG/ML)

l. Carbaryl

2. Thiodan

3. Larvin

 

4. Fenvalerate

 

5. Agrithrin super

 

6. Karate
7. Dimethoate
8

 

 

 

. Rogor

 

9. Marshall

 

10. Pfumo

I I.
Monocrotophos

HERBICIDES

 

 

 

 

 

 

 

 

 

 

 

 

 

CODES FOR PESTICIDES

 

Herbicides Acaricides Fungicides
12. Trif 13. Cucacron l4. Calirus

l8 Stomp l6. Hosthion l7. Tecto

18 Bladex 19. Azodrin/Nuvacron 20. Brassicol
21 Gesagard 22. Tedion/Tetradifon 23. Rizolex
24 Dual 25.Mitac 26.Monceren combi
27 Planavin 28. Secure 29.Vitavax
30 Cotoran 31. Kelthane 32 Quintozene
33 Cotogard 34. Omite

36 Gesapax

36 Diuron

37 Igran

206

 

NOV
YOL

12.1

stra

 

 

I 1 Imagine a poor, average and good season, how much yield would expect to get on your farm

N ECTED YIELD (KG/HA)

Poor
verage

 

NOW WE WANT TO ASK YOU QUESTIONS ABOUT PEST MANAGEMENT PRACTICES
YOU MAY HAVE USED ON YOUR COTTON CROP IN l998/99.

12. Have you ever received training on IPPM?
1=YES 0=NO

NOTE TO ENUMERATOR: Provide definition of IPPM if farmer is not clear about its meaning:

 

. IPPM is the control of pests using multiple management tactics, which include cultural, biological,

strategic and chemical control. IPPM knowledge is disseminated through Farmer Field Schools.

13. Where did you receive your training? IPMEDUC

1. AGRITEX

2. COTTON TRAINING CENTER (CTC)
3. IPPM TRAINED FARMERS

4. MASTER FARMERS

5. OTHER (SPECIFY)

207

l4. Indicate which pest management practices you used on your farm to control cotton INSECTS in
1998/99 season?

 

PEST MANAGEMENT PRACTICE FOR COTTON INSECTS

ADOPTION
STATUS
1=YES 0=NO

 

1=Use of pest resistant cotton varieties

 

2=Pest scouting and use of economic thresholds in making pesticide treatment.

 

3=Buming cotton stover (Field Sanitation)

 

4=Crop rotations

 

5=Mulching

 

6=Soil testing for pests

 

7=Use of less toxic and safer chemicals

 

8=A1ternating pesticides to slow development of pest resistance to pesticides.

 

9= Adjusting planting dates to lessen pest problems

 

10=Use of beneficial organisms that prey on pests (e.g. predator insects or mites).

 

1 l= Adjusting application rates, timing and frequency of pesticide use to protect
beneficial organisms

 

 

l2= Use of pheromones such as ants or moths that excrete poisonous chemicals
externally for trapping .

 

 

15.

What do you think about IPPM? IPMFPP

1. LABOR INTENSIVE

2. IPPM IS EFFECTIVE PEST CONTROL METHODS

3. IPPM METHODS ARE INEFFECTIVE

4. NOT SUPERIOR TO CONVENTIONAL CHEMICAL USE
5. OTHER (SPECIFY)

208

 

 

16. How many bales did you harvest from the different cotton fields in l998/99?

 

COTTON FIELD NO OF BALES

 

1=HOMESTAED FIELD

 

2=FIELD l

 

3=FlELD 2

 

4=FIELD 3

 

 

 

 

1?. Indicate how many bales where sold to the following markets in l998/99

MARKETING UTLET NUMB O BALES
l.
. A

4.L AL R
5.NEI RS
6.

 

18. Please indicate the grade distribution of cotton harvested in 1998/99?

 

GRADE NUMBER OF BALES PRICE PER KG TOTAL REVENUE
CATEGORY

 

A

 

B

 

C

 

D

 

 

REJ EC T

 

 

 

 

 

209

EL; .‘fi

.111

SOIL CONSERVATION AND TILLAGE PRACTICES IN COTTON FIELDS

 

 

20.Predominant soil texture in the village SOILTXT
I=SANDY LOAM SOIL
2=BLACK CLAY SOIL (VERTISOLS)
3=RED CLAY SOIL
4: OTHER (SPECIFY)
21. Cotton field characteristics
COTTON FIELD HOMESTEAD FIELD # l FIELD # 2 FIELD # 3
CHARACTERISTICS FIELD

 

Dominant soil texture

 

Walking time fi'om homestead
(hours/minutes)

 

Type of conservation practice

 

Soil Test Conducted?
(1=YES/0=NO)

 

 

 

 

 

 

 

CODES FOR SOIL TEXTURE CODES FOR CONSERVATION PRACTICES

I=SANDY SOIL
2=BLACK SOIL (VERTISOL)
3=RED SOIL

4=CLAY SOIL

5=OTHER (SPECIFY)

1=TIED RIDGES

2=CONTOURS

3=WINDBREAKS

4=MULCHING
5=MINIMUM TILLAGE

210

22. COTTON FIELD NUMBER

SIDE NUMBER

23. COTTON FIELD NUMBER

:FIELD AREA MEASUREMENTS

ANGLE (DEGREES) LENGTH (METERS)

 

:FIELD AREA MEASUREMENTS

 

SIDE NUMBER

ANGLE EGREES) LENGTH

 

211

M IIIIIIIIIIIIIIIIIIIIIIIIIIIIII

lllllllllllllllll[If

(l
l
3 1293 o 3