supply chains in export agriculture, competition, and...

Supply Chains in Export Agriculture,

Competition, and Poverty in

Sub-Saharan Africa

“The last two decades’ reforms in Africa’s agricultural marketing channels have taken place against a background of relative ignorance of how these markets work. Combining theory (with coverage of complex contractual arrangements like outgrower contracts), household surveys, and in-depth knowledge of local contexts, this masterful book provides the first systematic answer. In their characteristically careful approach, the authors use simulation analysis based on oligopoly theory to isolate and quantify the effect of policy shocks one by one and with synergies, yielding precise orders of magnitude where theory is usually silent. Written in a limpid style, this book is a must-read for academics and sophisticated policy analysts. It will be a reference for years to come.”Olivier Cadot, Professor of International Economics and Director of the Institute of Applied Economics at the University of Lausanne

“This is an innovative and important book. The authors explicitly model the institutions and industrial organization of global trade and commodity exchanges, which have major implications for the efficiency and surplus distribution among the participants in the chain. The combination of theory and empirical analysis across many developing countries is unique and yields important new insights.”Jo Swinnen, Professor of Development Economics at K.U.Leuven, Director of LICOS-Centre for Institutions and Economic Performance at K.U.Leuven and Senior Research Fellow in the Centre for European Policy Studies (CEPS), Brussels

Supply Chains in Export Agriculture, Competition, and Poverty in Sub-Saharan Africa

by: Guido Porto, Nicolas Depetris Chauvin and Marcelo Olarreaga

THE WORLD BANK

THE WORLD BANK

9 781907 142208

SUPPLY CHAINS IN EXPORT AGRICULTURE , COMPET IT ION ,AND POVERTY IN SUB -SAHARAN AFRICA


Copyright © 2011 byThe International Bank for Reconstruction and Development/The World Bank1818 H Street, NW, Washington, DC 20433, USA

ISBN: 978-1-907142-20-8

Typeset by T&T Productions Ltd, London

Published in association with the London Publishing Partnershipwww.londonpublishingpartnership.co.uk

Centre for Economic Policy Research

The Centre for Economic Policy Research is a network of over 700 ResearchFellows and Affiliates, based primarily in European universities. The Centrecoordinates the research activities of its Fellows and Affiliates and communi-cates the results to the public and private sectors. CEPR is an entrepreneur,developing research initiatives with the producers, consumers and sponsorsof research. Established in 1983, CEPR is a European economics research orga-nization with uniquely wide-ranging scope and activities.

The Centre is pluralist and non-partisan, bringing economic research to bearon the analysis of medium- and long-run policy questions. CEPR research mayinclude views on policy, but the Executive Committee of the Centre does notgive prior review to its publications, and the Centre takes no institutional pol-icy positions. The opinions expressed in this report are those of the authorsand not those of the Centre for Economic Policy Research.

CEPR is a registered charity (No. 287287) and a company limited by guaranteeand registered in England (No. 1727026).

Chair of the Board Guillermo de la DehesaPresident Richard PortesChief Executive Officer Stephen YeoResearch Director Mathias DewatripontPolicy Director Richard Baldwin

The World BankThe World Bank Group is a major source of financial and technical assis-tance to developing countries around the world, providing low-interest loans,interest-free credits and grants for investments and projects in areas suchas education, health, public administration, infrastructure, trade, financialand private sector development, agriculture, and environmental and naturalresource management. Established in 1944 and headquartered in Washing-ton, DC, the Group has over 100 offices worldwide. The World Bank’s missionis to fight poverty with passion and professionalism for lasting results and tohelp people help themselves and their environment by providing resources,sharing knowledge, building capacity and forging partnerships in the publicand private sectors.

Supply Chains in ExportAgriculture, Competition, andPoverty in Sub-Saharan Africa

GUIDO PORTO, NICOLAS DEPETRIS CHAUVINAND MARCELO OLARREAGA

Contents

List of Figures ix

List of Tables xiii

About the Authors xv

Preface xvii

Acknowledgments xix

1. Introduction 1

2. Case Studies 9

3. Institutional Arrangements 57

4. Exporters and Farmers: A Model of Supply Chainsin Agriculture 89

5. Supply Chain Simulations and Poverty Analysis inSub-Saharan Africa 163

References 263

List of Figures

A2.1 Per capita expenditure density for Benin. 19

A2.2 Per capita expenditure density for Burkina Faso. 20

A2.3 Per capita expenditure density for Côte d’Ivoire. 21

A2.4 Per capita expenditure density for Ghana. 22

A2.5 Per capita expenditure density for Malawi. 23

A2.6 Per capita expenditure density for Rwanda. 24

A2.7 Per capita expenditure density for Uganda. 25

A2.8 Per capita expenditure density for Zambia. 26

A2.9 Share of income and per capita expenditure for Benin: cotton. 27

A2.10 Share of income and per capita expenditure for BurkinaFaso: cotton. 28

A2.11 Share of income and per capita expenditure for Côted’Ivoire: (a) coffee; (b) cocoa; (c) cotton. 29

A2.12 Share of income and per capita expenditure for Ghana: cocoa. 30

A2.13 Share of income and per capita expenditure for Malawi:(a) tobacco; (b) cotton. 31

A2.14 Share of income and per capita expenditure for Rwanda: coffee. 32

A2.15 Share of income and per capita expenditure for Uganda: coffee. 33

A2.16 Share of income and per capita expenditure for Zambia:(a) tobacco; (b) cotton. 34

3.1 Cotton stylized value chain. 58

3.2 Cocoa stylized value chain. 70

3.3 Coffee stylized value chain. 77

3.4 Tobacco stylized value chain. 84

A4.1 Changes in farm-gate prices and quantities: leader splitsimulations. 112

A4.2 Changes in total and average profits: leader split simulation. 113

A4.3 Farm-gate prices, quantities, and utility: equal market shares. 113

A4.4 Farm-gate prices and complementary factors. 114

A4.5 Changes in farm-gate prices: complementarities. 114

A4.6 Changes in farm-gate prices: standard model and outgrowercontract model. 115

A4.7 Changes in the interest rate. 115

x Supply Chains in Sub-Saharan Africa

A5.1 Income effects under different scenarios for cotton in Zambia. 180

A5.2 Income effects for producers versus income effects for allhouseholds for cotton in Zambia. 180

A5.3 Effects on poor versus nonpoor for (a) producers only and(b) all households for cotton in Zambia. 181

A5.4 Effects on male-headed versus female-headed householdsfor (a) producers only and (b) all households for cotton inZambia. 182

A5.5 Income effect in basic model with and without outgrowercontracts for cotton in Zambia. 183

A5.6 Scenarios under (a) leader split and (b) leaders merge withand without outgrower contracts for cotton in Zambia. 183

A5.7 Income effects in cotton for producer households under(a) leader split and (b) leaders merge. 184

A5.8 Equal market shares income effects in cotton for poor andnonpoor producer households. 184

A5.9 Perfect competition income effects in cotton formale-headed and female-headed producer households. 185

A5.10 Income effect in basic model with and without outgrowercontracts in cotton–Burkina Faso. 185

A5.11 Scenarios under (a) leader split and (b) leaders merge withand without outgrower contracts in cotton–Burkina Faso. 186

A5.12 Income effects for cocoa producer households under (a)exit of the leader, (b) equal market shares, and (c) increasedinternational price. 187

A5.13 Perfect competition income effects in cocoa for poor andnonpoor producer households. 188

A5.14 Leader split income effects in cocoa for male-headed andfemale-headed producer households. 188

A5.15 Scenarios under leader split with and without outgrowercontracts in cocoa–Ghana. 188

A5.16 Income effects for coffee producer households under(a) leader split, (b) leaders merge, and (c) increasedinternational price. 189

A5.17 Equal market shares income effects in coffee for poor andnonpoor producer households. 190

A5.18 Perfect competition income effects in coffee formale-headed and female-headed producer households. 190

A5.19 Income effects in tobacco for producer households under(a) exit of the largest and (b) equal market shares. 191

A5.20 Perfect competition income effects in tobacco for poor andnonpoor producer households. 192

List of Figures xi

A5.21 Leader split income effects in tobacco for male-headed andfemale-headed producer households. 192

A5.22 Scenarios under leader split with and without outgrowercontracts in tobacco–Malawi. 192

List of Tables

A2.1 Agricultural exports in sub-Saharan Africa. 35A2.2 The composition of agricultural exports (2006; all values

are percentages). 37A2.3 List of sub-Saharan Africa household surveys. 38A2.4 Population, age and gender composition. 39A2.5 Demographic decomposition (household size by age group). 40A2.6 Per capita expenditure and income sources in Benin. 41A2.7 Per capita expenditure and income sources in Burkina Faso. 43A2.8 Per capita expenditure and income sources in Côte d’Ivoire. 45A2.9 Per capita expenditure and income sources in Ghana. 47A2.10 Per capita expenditure and income sources in Malawi. 49A2.11 Per capita expenditure and income sources in Rwanda. 51A2.12 Per capita expenditure and income sources in Uganda. 53A2.13 Per capita expenditure and income sources in Zambia. 553.1 Market shares in export supply chains: cotton. 613.2 Market shares in export supply chains: cocoa. 733.3 Market shares in export supply chains: coffee. 803.4 Market shares in export supply chains: tobacco. 85A4.1 Model calibration. 116A4.2 Impact of supply chain changes on the standard model:

(a) Zambia, cotton (percentage changes). 117A4.2 (b) Benin, cotton (percentage changes). 119A4.2 (c) Burkina Faso, cotton (percentage changes). 121A4.2 (d) Côte d’Ivoire, cotton (percentage changes). 123A4.2 (e) Malawi, cotton (percentage changes). 125A4.2 (f) Côte d’Ivoire, cocoa (percentage changes). 126A4.2 (g) Ghana, cocoa (percentage changes). 128A4.2 (h) Côte d’Ivoire, coffee (percentage changes). 130A4.2 (i) Rwanda, coffee (percentage changes). 132A4.2 (j) Uganda, coffee (percentage changes). 134A4.2 (k) Malawi, tobacco (percentage changes). 136A4.2 (l) Zambia, tobacco (percentage changes). 138A4.3 Impact of supply chain changes on the outgrower contract

model: (a) Zambia, cotton (percentage changes). 140A4.3 (b) Benin, cotton (percentage changes). 142

xiv Supply Chains in Sub-Saharan Africa

A4.3 (c) Burkina Faso, cotton (percentage changes). 144A4.3 (d) Côte d’Ivoire, cotton (percentage changes). 146A4.3 (e) Malawi, cotton (percentage changes). 148A4.3 (f) Côte d’Ivoire, cocoa (percentage changes). 149A4.3 (g) Ghana, cocoa (percentage changes). 151A4.3 (h) Côte d’Ivoire, coffee (percentage changes). 153A4.3 (i) Rwanda, coffee (percentage changes). 155A4.3 (j) Uganda, coffee (percentage changes). 157A4.3 (k) Malawi, tobacco (percentage changes). 159A4.3 (l) Zambia, tobacco (percentage changes). 161A5.1 Impact of supply chain changes on the standard model:

(a) Zambia, cotton (household income percentage changes). 193A5.1 (b) Benin, cotton (household income percentage changes). 196A5.1 (c) Burkina Faso, cotton (household income percentage

changes). 199A5.1 (d) Côte d’Ivoire, cotton (household income percentage

changes). 202A5.1 (e) Malawi, cotton (household income percentage changes). 205A5.1 (f) Côte d’Ivoire, cocoa (household income percentage changes). 207A5.1 (g) Ghana, cocoa (household income percentage changes). 210A5.1 (h) Côte d’Ivoire, coffee (household income percentage changes). 213A5.1 (i) Rwanda, coffee (household income percentage changes). 216A5.1 (j) Uganda, coffee (household income percentage changes). 219A5.1 (k) Malawi, tobacco (household income percentage changes). 222A5.1 (l) Zambia, tobacco (household income percentage changes). 225A5.2 Impact of supply chain changes on the outgrower contract

model: (a) Zambia, cotton (household income percentagechanges). 228

A5.2 (b) Benin, cotton (percentage) changes in household income. 231A5.2 (c) Burkina Faso, cotton (household income percentage

changes). 234A5.2 (d) Côte d’Ivoire, cotton (household income percentage

changes). 237A5.2 (e) Malawi, cotton (household income percentage changes). 240A5.2 (f) Côte d’Ivoire, cocoa (household income percentage changes). 242A5.2 (g) Ghana, cocoa (household income percentage changes). 245A5.2 (h) Côte d’Ivoire, coffee (household income percentage changes). 248A5.2 (i) Rwanda, coffee (household income percentage changes). 251A5.2 (j) Uganda, coffee (household income percentage changes). 254A5.2 (k) Malawi, tobacco (household income percentage changes). 257A5.2 (l) Zambia, tobacco (household income percentage changes). 260

About the Authors

Guido Porto is a Professor of Economics at the University of La Plata inArgentina. Before joining the University of La Plata, he was an economist inthe Research Department of the World Bank. He has taught at Universidad andInstituto Di Tella in Argentina and at the University of Maryland as an invitedlecturer. He received a BA in Economics from the University of La Plata inArgentina, a Master’s degree in Economics from Instituto Di Tella in Argentina,and a PhD in Economics from Princeton University. His research focuses onthe econometric estimation of the impacts of trade policies in developingcountries, including impacts on poverty, household welfare, wages, and thedistribution of income as well as on firm behavior. His latest work investigateshow economic agents (households and firms) adjust to trade reform.

Nicolas Depetris Chauvin is Assistant Professor of Public Policy at DubaiSchool of Government and a Research Fellow at the Oxford Centre for theAnalysis of Resource Rich Economies in the Department of Economics at theUniversity of Oxford. Previously he worked as a policy specialist in the Officeof Development Studies at the United Nations Development Programme andas a consultant at the World Bank, the Inter-American Development Bank, andthe Ministry of Economics of Argentina, among others. His fields of interestare international finance, development, macroeconomics, and resource-richeconomies. During his doctoral studies at Princeton University he did researchon debt-relief initiatives for highly indebted poor countries.

Marcelo Olarreaga is Professor of Economics at the University of Geneva andResearch Fellow at the Centre for Economic Policy Research in London. Beforejoining the University of Geneva he worked as an economist in the ResearchDepartment of the World Bank, as well as in the Economics Research Divi-sion of the World Trade Organization. He has also been an invited professorat INSEAD (France), Institute CLAEH (Uruguay), SciencePo-Paris (France), Uni-versidad de la República (Uruguay), and the University of Antwerp (Belgium).He holds an MA from the University of Sussex, and a PhD in Economics fromthe University of Geneva. He is currently doing research on the political econ-omy of trade policy and trade agreements, barriers to developing countries’exports, and distributional and poverty impacts of trade reforms.

Preface

There has been a major shift in the focus of trade research and the develop-ment community in the last decade. Traditionally, much of the attention wason trade policy—tariffs, quotas—and the effects of such policies in creatingan anti-export bias. High tariffs in developing countries act as a disincentiveto export in a number of ways. At the same time, high tariffs in export destina-tions also reduce the ability of developing countries to contest markets. Theanalytical and policy focus of much of the literature was to identify the incen-tive effects of trade policies, with a particular emphasis by the developmentcommunity on the importance of improving the effectiveness of preferentialmarket access schemes. In many countries tariffs were lowered substantiallyin the 1990s and 2000s, and many OECD countries have improved the cover-age and depth of their preferential market access programs for the poorestcountries. Most observers would recognize that the “supply response” to thesepolicy changes in the least developed countries was less than was hoped for.One result has been a shift in focus to analyze more broadly the determi-nants of competitiveness. Trade economists now emphasize the importanceof real trade costs as a barrier to trade, ranging from factors such as badinfrastructure to delays and excessive paperwork in clearing customs andhigh costs of transport, especially for landlocked countries. The effect of suchcosts is to raise prices and make farmers and other producers in low-incomeeconomies less competitive. The indirect effect is to inhibit investment andjob creation.

The focus of the present book is on a neglected source of such “real costs”for farmers in low-income countries: market power along the supply chain.An absence of competition among the providers of key inputs that are usedin farming may result in transfers from farmers (and consumers) to the firmsthat provide intermediates such as seeds and fertilizer and transportationservices and storage. Such firms can raise prices above what they would be ifthe market structure was competitive, to the detriment of farmers, and indeedthe detriment of society as a whole. A lack of competition is likely to resultin inefficiencies—and associated deadweight costs.

As is illustrated in this book, the exercise of market power along the supplychain will have a number of possible effects. Given that the impacts on priceswill imply a redistribution of income across agents in the economy, therewill also be implications for poverty and the attainment of poverty reductionobjectives. In general, smallholders are likely to be poorer than the owners

xviii Supply Chains in Sub-Saharan Africa

of the firms that supply them with services and tangible inputs. The analy-sis shows that the effects of market power on income distribution and theincomes of poor households can be significant. It also demonstrates that thereis much heterogeneity, both across countries for a given crop, and within acountry for different crops.

The bottom line is clear. Measures to increase competition—to make themarket structure less concentrated—will have effects similar to those result-ing from actions to reduce customs clearance times, facilitate the movementof goods across borders, and so forth. The latter are issues that are at theforefront of the policy agenda in many countries and the activities of develop-ment agencies. This is much less the case for the competition-related, marketstructure issues that are the center of attention in this volume. The analysis istherefore not just an interesting analytical exercise—it is of great policy sig-nificance as it suggests that more attention should be devoted to competitionpolicy in low-income countries.

Bernard Hoekman Stephen YeoSector Director, Trade Department Chief Executive OfficerPoverty Reduction and Economic Management CEPRThe World Bank

Acknowledgments

We want to warmly thank Daniel Lederman, Aaditya Mattoo, and BernardHoekman for helping us put this project together and for all of their supportduring its implementation. We thank Jorge Balat, Huberto Ennis, LeonardoGasparini, Sebastian Ludmer, Mariana Marchionni, and Alessando Nicita fortheir contributions to the book, and German Bet, Laura Jaitman, David Jaume,and Cecilia Peluffo for outstanding research assistance. Anil Shamdasani andthe CEPR team did superb publishing work. We would also like to thank PhilipAbbott, Gilles Alfandari, Francois Ruf, Daniel Sarpong, Daniel Sellen, HardwickTchale, and Marcella Vigneri for providing us with important information onthe institutional arrangements for several countries. Silvia Chauvin helpedwith the value chains figures in Chapter 3. Michelle Chester patiently assistedus with the administration of the project.

We have received useful comments and suggestions from Dante Amen-gual, Irene Brambilla, Olivier Cadot, Daniel Lederman, Justin Lin, Will Martin,and Stephen Mink, and from numerous participants in seminar presentations.These comments helped us significantly improve the book. We remain, how-ever, responsible for any omission and errors.

Financial support from the Bank Netherlands Partnership Program at theWorld Bank is gratefully acknowledged. All views expressed here are thoseof the authors and should not be regarded as those of the World Bank or itsclients.

1

Introduction

1 OVERVIEW

The literature on trade and poverty is large and has grown extensively dur-ing the last decade or so (Winters et al. 2004; Goldberg and Pavcnik 2004,2007; Porto 2007, 2010, 2011). Initially, this literature investigated whethertrade reforms, meaning mostly border protection such as tariffs, quotas, orexport taxes, affected poverty. The approach is based on the idea that tradereforms affect domestic prices, and domestic prices affect households via sev-eral channels. As consumers, households benefit from lower prices and arehurt by higher prices. As producers, households benefit from higher pricesand are hurt by lower prices. As income earners, prices can affect wages andemployment and thus trade can affect the labor income of the household.Other channels, such as changes in transfers or capital income, can also playa role. In the end, the impact of trade on poverty is ambiguous: it depends onthe size of the price change (the pass-through), on whether the poor are netproducers or net consumers of the goods affected by the trade reform, on theresponse and nature of the labor markets, and so on (see Deaton 1989, 1997;Hertel and Winters 2006; Hoekman and Olarreaga 2007; Nicita 2009; Porto2005, 2006; Ravallion 1990).

In this book, we study how the internal structure of export markets andthe level of competition affect poverty and welfare in remote rural areas inAfrica. In sub-Saharan Africa, rural poverty is a widespread phenomenon.While most farmers produce for home consumption, some are engaged inhigh-value export agriculture. Here, we focus on export crops such as coffee,cotton, cocoa, and tobacco. For many African countries, these crops, whichare typically produced by smallholders, are a major source of export revenue.In consequence, changes in export prices and in the conditions faced in exportmarkets (both internally and externally) can play a big role in shaping povertyin the region. Traditionally, the literature has focused on how external con-ditions affect poverty, for example by addressing whether agricultural subsi-dies in the developed world affect world prices and how this in turn affectsfarm-gate prices. Our objective in this book is to explore domestic factors. Inparticular, we investigate the role played by the structure of competition inexport agriculture supply chains.

2 Supply Chains in Sub-Saharan Africa

In Africa, commercialization of export agriculture is produced along a sup-ply chain where intermediaries, exporters, and downstream producers inter-act with farmers. Often, the sector is concentrated, with a few firms competingfor the commodities produced by atomistic smallholders. This structure of themarket conduces to oligopsony power: firms have market power over farmersand are able to extract some of the surplus that the export market generates.The extent of oligopsony power depends on the number of competitors andon the relative size of each competitor (the distribution of market shares).Changes in the configuration of the market will thus affect the way the firmsinteract with the farmers. In principle, tighter competition induced by entryor by policies that foster competition (e.g., merger or antitrust policies) canaffect farm-gate prices and therefore household welfare and poverty. This isthe topic of our investigation.

The relationship between firms and farmers in export markets in Africais complex. On top of the standard game-theoretic interrelationship, wherefirms interact with each other and take into account the response of farm-ers when setting prices, many markets are characterized by the presence ofoutgrower contracts. When there are distortions in the economy, or missingmarkets (especially for credit and capital), it may be impossible for farmers tocover any start-up investments related to the production of the export crop.In those scenarios, farmers would not be able to purchase seeds, or the pes-ticides needed for cash crop production and the market for these crops candisappear. In Africa, one way to solve this issue is with outgrower contracts,whereby firms provide inputs on loan at the beginning of the season. Theseloans, and any interest bore, are then recovered at harvest time. While out-grower contracts can be a very useful instrument for making these marketswork, they may fail, sometimes catastrophically, when there are enforceabil-ity problems. Clearly, an inadequate legal system may prevent enforceabil-ity. Equally importantly, the presence of too many players/firms interactingsimultaneously can facilitate side-selling, a situation in which a farmer takesup a loan with one firm, sells to a different one at harvest, and thus defaultson the original loan. In these cases, it is possible for increased competitionto make contract monitoring very costly. Interest payments may become tooburdensome for the farmers. In extreme cases, this may lead to a vicious cyclein which the system fully collapses. Our analysis covers these scenarios.

Our analytical methodology has two main parts. The first is a game-theorymodel of supply chains in cash crop agriculture between many atomisticsmallholders and a few exporters. The model provides the tools needed tosimulate the changes in farm-gate prices of export crops given hypotheticalchanges in the structure of the supply chain. Farm-gate prices are set by firms.The firms buy raw inputs from the farmers (coffee beans, cotton seeds, etc.)and sell them in international markets at given prices. In contrast, the firmsenjoy oligopsony power internally. The oligopsony game delivers the equilib-rium farm-gate prices that the firms offer to the farmers. Given these prices,

Introduction 3

farmers allocate resources optimally and supply raw inputs to the firms andthis supply affects the quantity that firms can supply in the export market. Inequilibrium, firms take into account the supply response of the farmer whenchoosing optimal farm-gate prices.

Once the equilibrium of the model is found, and the solution is calibratedto match key features of the economy, we simulate various changes in compe-tition. Our simulations cover a large number of general settings, from entry toexit. We study the impacts of the entrance of a small competitor, of a hypothet-ical merger between the two leading firms, and of the split of the leader intosmaller competitors (the “small entrant,” “leaders merge,” and “leader split”simulations). We also explore scenarios with tightest, albeit still imperfect,competition (the “equal market shares” simulations) as well as a hypothetical“perfect competition” scenario. In all these cases, given the initial equilibrium,we find the new equilibrium of the model and study the changes in farm-gateprices. We also study simulations of changes in the level of competition withcomplementary policies that affect firms and farmers, and with changes ininternational prices. Finally, we combine all these simulations with a modelthat allows for outgrower contracts. In the end, we run seventy simulationswith seventy corresponding changes in farm-gate prices.

The second component utilizes household surveys to assess the povertyimpacts of those changes in the value chains. We follow a standard first-order effects approach, as in Deaton (1989, 1997). Using the microdata fromthe household surveys, we use income shares derived from the productionof different crops to evaluate the income impacts of a given farm-gate pricechange. We investigate the average impact for all rural households, the distri-bution of these impacts across levels of living (e.g., for poor vis-à-vis nonpoorhouseholds) and the differences in impacts between male-headed and female-headed households.

We explore twelve case studies: the cotton sector in Zambia, Malawi, BurkinaFaso, Côte d’Ivoire, and Benin; the coffee sector in Uganda, Rwanda, and Côted’Ivoire, the tobacco sector in Malawi and Zambia, and the cocoa sector inCôte d’Ivoire and Ghana. We focus on those crops that are plausible vehiclesfor poverty eradication and on those countries where the household surveydata needed for the poverty analysis is available.

2 MAIN RESULTS

In discussing the results from the simulations and the poverty impacts, westudy the case of Zambian cotton in detail. We then briefly explore differencesand similarities with the remaining eleven case studies. The Zambian cottonsector is our leading case because it has undergone several of the transforma-tions our simulations aim to capture (i.e., privatization, entry, and exit) andbecause several outgrowers schemes have been implemented.


The main conclusion of our analysis for the case of cotton in Zambia is thatcompetition among processors is good for farmers as it increases the farm-gate price of the crop and therefore it improves their livelihood. For instance,if the leading firm splits, the increase in income for the average cotton pro-ducer would be equivalent to 2.4 percent of his initial income. On the otherhand, if the two largest firms in the Zambian cotton market merge, the incomeof the average producer would decline by 2.3 percent. The largest possiblegain for the farmers occurs under perfect competition, where farmers wouldenjoy an income gain of 19.3 percent. The implementation of complementarypolicies or a positive international price shock intensify the positive effectsof more competition and mitigate the negative effects of a reduction in thelevel of competition among ginneries. For example, the original increase of2.4 percent in income for producers following the split of the leader becomes2.8 percent under complementary policies for farmers, 3.1 percent when thesepolicies affect only firms, and 3.5 percent when they affect both farmers andfirms. With outgrower contracts, farms need to borrow to finance the pro-duction of export crops. In consequence, we find that, in all simulations, thegains from more competition and the losses from higher market concentra-tion tend to be lower. While possible in theory (since we assume that the costof enforcing these contracts increases with market competition and that thosecosts are transferred to producers through increasing borrowing costs), noneof our simulations uncover a collapse of the market due to increased compe-tition among ginneries.

The general conclusion that more competition among processing andexporting firms is beneficial for smallholders applies to the other case stud-ies. Take, for instance, the case where the firm with the largest market sharessplits. This would lead to an average income increase for producing house-holds of 2.8 percent across all case studies. However, this average masks a lotof variability, with high gains for African cotton farmers and low gains insteadfor African coffee smallholders. For instance, in our baseline scenario, theleader split simulation would increase average households’ income in cottonin Burkina Faso by 12.4 percent but only by 0.1 percent in coffee in Uganda.This does not come as a surprise, however, since the leading cotton firm inBurkina Faso controls 85 percent of the market, while the leading firm inUganda controls only 14.3 percent of the coffee market. Another interest-ing simulation of increased competition is the case of equal market shares,which delivers the upper bound increase in income under imperfect compe-tition. Here, the average effect is much larger than in the case of leader split.The average producer household would see its income grow by 9.1 percentacross our case studies with cotton in Burkina Faso (20.9 percent) and Benin(20.1 percent) showing the largest gains. In contrast, the average householdwould receive less than 1 percent extra income in the equal market sharessimulation in coffee in Uganda and Rwanda.

Introduction 5

The finding that increases in competition among processors benefit farmersneeds to be put into perspective. One important result from our simulationsis that small changes in the level of competition are unlikely to have signif-icant effects on farmers’ livelihoods. This is captured by the small entrantsimulation. Under this scenario, the income of households only increases byan average of a quarter of a percentage point for our case studies. The largesteffects for this simulation are observed in cotton in Malawi (0.94 percent) andtobacco in Zambia (0.74 percent).

We are also interested in assessing the effects of reductions in competitionamong upstream firms. This is done by studying the effects of the mergingof the largest two firms in the market and through the case of the “exit ofthe largest” (and most efficient) firm. In the first simulation, the average lossfor farmers is 1.3 percent of their income. The largest loss is registered inthe case of cotton in Côte d’Ivoire (3.8 percent) where the new merged firmwould control three quarters of the market. In the exit of the largest firmsimulation, the highest income loss for producer households takes place inthe cotton sector of Burkina Faso, where the disappearance of SOFITEX (whichcontrols 85 percent of the market) would lead to a decrease in farm income of15.4 percent. In contrast, the smallest impacts of a reduction in competitionare observed in the coffee sector in Uganda, where the loss would in fact beonly around a tenth of a percentage point in both simulations.

To investigate the role of complementary policies, we compare our baselinescenario with cases where we introduce complementarities affecting farmers(for instance, extension services), affecting firms (infrastructure), or both. Themain finding here is that these policies boost the income of households whenthere is an increase in competition and mitigate (or even revert) the incomeloss when there is a reduction in competition. In all cases, the largest effectsare observed when the complementary factors affect both firms and farmers.Overall, the quantitative effects of these policies depend on the particularcrop, country, and market structure, but they range from as low as a 0.36 per-cent (equal market shares simulation for coffee in Rwanda) to a 2.8 percent(perfect competition simulation for cotton in Côte d’Ivoire) income gains forthe average producer household.

We also study the effects of a 10 percent exogenous increase in the interna-tional prices of the different export crops. Given the current market configu-ration for each crop and country in the study, the exogenous price increaseleads to an average raise in producing households’ income of 6.9 percent.The largest effect takes place in cotton in Burkina Faso (12.8 percent) and thesmallest effect takes place in coffee in Rwanda (2 percent). We also allow forcombinations of higher international prices and different changes in the levelof competition among processors. Increasing competition boosts the positiveeffects of the price change, while a reduction in the competition damps itseffects. For instance, in the case of perfect competition, the income gains forproducing households would range from 68 percent (cotton in Burkina Faso)


to 3.8 percent (coffee in Uganda), with an average effect across case studiesof 29.6 percent. On the other hand, when the largest firm exits the market,the overall income effect of the increase international prices would range from−3.9 percent (cotton in Burkina Faso) and 11.3 percent (cocoa in Côte d’Ivoire)with an average income effect of 3.4 percent for all the case studies.

The survey data allows us to distinguish the effect of the different simula-tions on poor versus nonpoor households and across gender groups. In nineout of the twelve simulations, the benefits of more competition have a largerincome effect in male-headed households than in the female counterpart. Thethree exceptions are the case of cotton in Benin and Côte d’Ivoire and coffee inRwanda. The largest differences between genders are found in Burkina Faso–cotton, where the gains for male-headed households are 217 percent higherthan the income gains of female-headed households, and in Benin–cotton,where instead the gains for female-headed households are 57 percent higherthan the gains for male-headed households. It is noteworthy that in only fourout of the twelve case studies, the increase in competition is pro-poor. This isbecause of the relatively low participation of poor households in the produc-tion of the export crop. The income gains are larger for the poor, on average,in the cases of coffee and cocoa in Côte d’Ivoire, coffee in Rwanda, and cottonin Zambia.

Turning to the models with outgrower contracts, we find that, in all casestudies except for cotton in Burkina Faso, an increase in competition is ben-eficial for farmers and, in turn, a higher market concentration is prejudicial.As in the case of Zambia–cotton, the gains from increasing competition andthe losses from more oligopsony power are lower (because of the borrowingcosts). In general, the discrepancies with the results from models without out-grower contracts are rather small. Interestingly, the cotton sector in BurkinaFaso is the only case where less competition is better for smallholders.

Three of the countries in our study have more than one case study. In Côted’Ivoire we study cotton, coffee, and cocoa. In Malawi and Zambia we coverboth cotton and tobacco. It is interesting, then, to describe how the same sce-nario and simulation has different effects upon crops in the same country.For instance, in Côte d’Ivoire, an increase in competition has a larger effect onproducing households’ income in cotton than in cocoa and coffee. If the leaderfirm in cotton, cocoa, and coffee were to split, the effect on income would bea 4.6 percent, 1.1 percent, and 0.6 percent increase, respectively. In the caseof equal market shares, the increase in the income of households would be14 percent, 10.5 percent, and 5.4 percent, respectively. The effect is also dif-ferent for poor versus nonpoor households and across gender depending onthe crop. Competition in coffee benefits more poor and male-headed house-holds while, in cotton, female-headed and nonpoor households enjoy largergains. In cocoa, competition benefits male-headed households slightly morewhile the effect is about the same among poor and nonpoor households. InMalawi we cannot directly compare the results between cotton and tobacco

Introduction 7

simulations, as the latter are slightly different from the standard simulationthat we run for all the other case studies for reasons we will explain later.However, the overall effects seem to be of about the same magnitude, while,in both crops, male-headed and nonpoor households benefit the most fromthe increase in competition. Finally, in Zambia, the effect of competition hassimilar quantitative effects in cotton and tobacco. The leader split case wouldincrease the income of households by 2.4 percent in cotton and by 3.2 per-cent in tobacco, while the equal market share case would generate a growth inincome of 7.3 percent in cotton and of 7.2 percent in tobacco. In both cropsmale-headed households benefit the most, though only slightly in the caseof cotton. Poor producing households gain more in cotton, while nonpoorbenefit more in the case of increasing competition among tobacco exporters.

3 THE ORGANIZATION OF THE BOOK

In Chapter 2, we introduce the twelve case studies and we document howand why we chose them. We describe the availability of household surveysacross countries and we explore crops that provide an important source ofcash income for the economy. In Chapter 3, we describe the main institutionalarrangements that characterize the supply chains in the each of the twelvecase studies. This description includes the different vertical arrangementsfrom different crops, whether the value chains are characterized by one ormore layers, the structure of competition among firms and farmers in eachlayer, and whether there are market interlinkages. In Chapter 4, we developthe theoretical model of supply chains in export agriculture. The purpose ofthe model is to provide an analytical framework for studying how changes inthe structure of the supply chain affect farm-gate prices. Finally, in Chapter 5,we combine the household survey with the farm-gate price changes to carryout the poverty analysis. We estimate the impact, at the farm level, of changesin the supply chain on household income.

2

Case Studies

In this chapter, we describe the twelve case studies that comprise our inves-tigation on supply chains, agricultural exports, and poverty in sub-SaharanAfrica. The case studies were selected following two criteria. First, the exportcrop had to have the potential to eradicate poverty, and, second, the micro-data needed for the poverty analysis had to be available in the target country.We investigate twelve case studies covering four crops, namely cocoa, cof-fee, cotton, and tobacco, in eight countries, namely Benin, Burkina Faso, Côted’Ivoire, Ghana, Malawi, Rwanda, Uganda, and Zambia.

The objective of this chapter is to document our selection of the case stud-ies. To this end, we begin in Section 1 by looking at export data to assesshow important different export crops are for different countries. Given thenature of our analysis, we only focus on crops that are a major source ofcash income for the economy (thus leaving aside food crops). In Section 2, weexplore the availability of household surveys in sub-Saharan countries and,for those countries where this data is available, we check whether the selectedexport crops are grown by a significantly large number of households andwhether those crops generate a significant share of total household income.To document this, we describe summary statistics (sample size, gender struc-ture, and demographic composition) from the household surveys used in theanalysis. Finally, in Section 3, we use the microdata from these surveys tocharacterize those farmers that produce the export crops.

1 EXPORTS

To get a sense of the overall importance of the export crops for the localeconomy, we begin with a description of the export structure of each tar-get country. Table A2.1 provides the summary statistics. In general, all theselected crops are very important for the economy of at least one of the coun-tries. Cocoa is a crucial foreign exchange generator both in Côte d’Ivoire andGhana, where it raises between 20 and 25 percent of all export revenue. Coffeeexports account for more than 10 percent of the total exports in both Rwandaand Uganda. Cotton accounts for more than one-third of total exports in Beninand Burkina Faso. Finally, tobacco accounts for more than 70 percent of exportearnings in Malawi.


1.1 Benin

Cotton is the main agricultural export of Benin, which, after Mali and Burk-ina Faso, is the largest cotton exporter in West Africa. Historically, cotton hasrepresented around 70 percent of the value of agricultural exports and morethan 30 percent of total export value (Table A2.1). However, the export valuerecently collapsed from $208 million in 2004 to $113 million in 2006. Partof this reduction was due to a decrease of 30 percent in the total volumeexported. This crop generates revenues for some 325,000 small holders. Cot-ton lint constitutes 92 percent of the cotton export value, followed by cottonseed oil (6.8 percent) and seeds (1.1 percent). The main exports markets forBenin in 2007 were China (24.7 percent), India (8.2 percent), Niger (6.6 per-cent), Togo (5.4 percent), Nigeria (5.3 percent), and Belgium (4.6 percent). Thetotal value of exports of goods and services only represents around 13 per-cent of Benin’s GDP. This low number may indicate that there is room forfurther increases in the share of the external sector in Benin’s economy.

1.2 Burkina Faso

Burkina Faso is one of the main producers and exporters of cotton in Africa.In 2006, cotton exports generated $235 million. This accounted for 81 percentof the agricultural exports and for 35 percent of all the revenue generated byexports (Table A2.1). Other important agricultural exports are cattle, sesameseed, and mangoes, but their export value is marginal in comparison with cot-ton. As in the case of Benin, cotton lint constitutes the bulk of cotton exports(96 percent of the export value), followed by cotton seed oil (2.7 percent) andseeds (1.1 percent). The main destination for Burkina Faso’s exports in 2007were China (29.6 percent), Singapore (15.7 percent), Thailand (7.2 percent),Ghana (6.4 percent), and Niger (4.8 percent). Exports of goods and servicesremain low, as they account for 11 percent of GDP on average.

1.3 Côte d’Ivoire

After oil, cocoa is the most important source of foreign revenue in Côted’Ivoire, followed by rubber, coffee, bananas, cashew nuts, and cotton. Cocoaproducts account for more than 60 percent of total agricultural exports andfor more than one-fifth of total export revenue (Table A2.1). This sector hasgenerated around $2 billion per year of export income in recent years. Cocoabeans account for 73 percent of total cocoa exports, followed by cocoa paste(12 percent), cocoa butter (9 percent), husks and shell (3.5 percent), and pow-der and cake (2 percent). Côte d’Ivoire is the largest worldwide cocoa producerand exporter followed by Ghana and Indonesia.

Coffee is the second most important cash crop for the Ivorian economy.Coffee exports amounted to $166 million in 2006 (5.3 percent of total agri-culture exports and 1.8 percent of total exports) up from $130 million in 2004.One-third of the coffee exports are coffee extracts, while the other two-thirds

Case Studies 11

correspond to green coffee. Côte d’Ivoire is the third largest coffee exporterin Africa after Ethiopia and Uganda.

Cotton is also an important cash crop in the Ivorian Economy. The exportsof the sector were $121 million in 2006 (1.3 percent of total export income).Most of the exports are cotton lint (93 percent), followed by cotton seed(5.1 percent) and cotton oil (1.3 percent). Côte d’Ivoire is the fourth largestcotton exporter in sub-Saharan Africa after Burkina Faso, Mali, and Benin.

The external sector is of vital importance for the Ivorian economy, withexports of goods and services accounting for more than 50 percent of thecountry’s GDP in recent years. The main export partners are France (23.7 per-cent), the Netherlands (10.8 percent), the United States (10.2 percent), Nigeria(7.5 percent), and Italy (4.8 percent).

1.4 Ghana

Cocoa has traditionally been the main source of foreign income for the Ghana-ian economy. This may change in the future as the country starts to exportthe recent oil discoveries. In 2006, cocoa generated $1.2 billion in exportsrevenue. This amounted to 79 percent of agricultural exports and one-fourthof all export revenue (Table A2.1). Ghana is the second largest producer andexporter of cocoa after Côte d’Ivoire. Cocoa beans are the main exportingproduct (86.6 percent) followed by cocoa butter and cocoa paste (both around6 percent of cocoa exports).

The exporting sector accounts for more than one-third of total GDP, andthis highlights the importance of the cocoa sector for the Ghanaian economy.The main markets for Ghanaian exports are the Netherlands (12.5 percent),the United Kingdom (8.3 percent), the United States (6.7 percent), Belgium(5.8 percent), France (5.6 percent), and Germany (4.4 percent).

1.5 Malawi

Tobacco is the most important export from Malawi, followed by maize, sugar,tea, and cotton. In 2006, it generated almost three-fourths of all export income($432 million, up from $258 million in 2004, see Table A2.1 for details). Malawiis the third largest exporter of tobacco after Brazil and the United States.However, the main difference between these two countries is that almost allMalawian exports are unmanufactured tobacco.

Cotton is another important cash crop for the Malawian economy, generat-ing $32.6 million in 2006 (around 5 percent of all export income). Cotton lintaccounts for 94 percent of all cotton exports, with the remaining 6 percentcoming from cotton seeds.

Exports are roughly one-fifth of total GDP and the main exports marketsfor Malawi are South Africa (12.6 percent), Germany (9.7 percent), Egypt(9.6 percent), the United States (9.5 percent), Zimbabwe (8.5 percent), Russia(5.4 percent), and the Netherlands (4.4 percent).


1.6 Rwanda

Coffee is the main agriculture export in Rwanda, accounting for 60 percentof all agricultural exports in 2006 and 17.5 percent of total export revenues.The sector generated exports for $48 million in 2006 (Table A2.1). In 2007, thetotal export value decreased to $32 million, remaining ahead of tea, the secondlargest agricultural export with $30 million. Almost all exports of Rwandancoffee are green coffee.

Major export markets for Rwanda in 2007 were the United Kingdom(18.7 percent), Kenya (18.6 percent), Belgium (14 percent), China (12.5 per-cent), and Switzerland (4.7 percent).

1.7 Uganda

As in Rwanda, coffee is the most important cash crop in Uganda. Other impor-tant crops are tobacco and tea. Total coffee exports (99 percent of themgreen coffee) amounted to almost $190 million in 2006. This amount cor-responded to 42 percent of all agricultural exports and 12.5 percent of allexports (Table A2.1). Uganda is the second largest coffee exporter in sub-Saharan Africa.

In 2008, the three largest markets for exports were Sudan (14.3 percent),Kenya (9.5 percent), and Switzerland (9.0 percent).

1.8 Zambia

Cotton and Tobacco are the two most important cash crops in Zambia.Together they generated almost half of the agricultural exports (Table A2.1).However, they only account for less than 2 percent each of total export rev-enues as copper is by far the most important foreign revenue generator inZambia. Almost all tobacco exports ($75 million in 2006) are unmanufactured.Cotton lint exports are the largest exporting cotton item (91.5 percent) fol-lowed by cotton seed (7 percent) and carded and combed cotton (1 percent).

The three main destinations for Zambian exports are Switzerland, SouthAfrica, and Egypt.

1.9 Final Remarks

From this analysis, it follows that most of the agricultural exports in the coun-tries that we study involve little local processing. This is more clearly seen inTable A2.2, which shows the composition of exports of cotton, tobacco, coffee,and cocoa for each country for 2006.

In the top panel, we observe that more than 90 percent of the cottonexported by our target countries is cotton lint. Unfortunately, we do nothave detailed and reliable information on cotton yarn exports. However, thesecountries have a low spinning installed capacity given their cotton potential

Case Studies 13

production. Burkina Faso, Benin, and Côte d’Ivoire are among the top twentyworld exporters of cotton lint but none of them are among the top twentyexporting countries of cotton yarn or apparel.

Almost all of the tobacco exported by the two tobacco-producing coun-tries in our study is unmanufactured (second panel). This is quite commonamong developing countries. In Brazil, the largest exporter, 98 percent of theirtobacco exports are unmanufactured.

The composition of coffee exports is presented in the third panel ofTable A2.2. Green coffee accounts for most coffee exports. This is not unusual,however, since other large coffee exporters (Brazil, Vietnam, and Colombia)also sell mostly green coffee. Roasted and instant coffee is produced in theconsuming countries.

Finally, the bottom panel of Table A2.2 shows the composition of cocoaexports in Côte d’Ivoire and Ghana. Most of their exports are in the rudimen-tary form of beans: 85 percent in Ghana and 75 percent in Côte d’Ivoire incomparison with 69 percent in Indonesia and 1 percent in Brazil (the othertwo big international producers of cocoa). Brazil exports 65 percent of itscocoa production as cocoa butter. None of the top five exporters of cocoa areamong the top twenty exporters of chocolate.

2 THE HOUSEHOLD SURVEYS

In order to perform the poverty analysis, we need household survey data withdetailed information on crop production and income. The available house-hold surveys for the eight target countries in sub-Saharan Africa are listed inTable A2.3.

In the case of Benin, we use the “Questionnaires des indicateurs de basedu bien-être” conducted in 2003. The survey covered 5,350 households outof a population of 1.4 million households. Rural households accounted for61.5 percent of total respondents. In Burkina Faso, we use the “Enquête Burk-inabe sur les conditions de vie des ménages,” also from 2003, which surveyed8,500 households (0.48 percent of the total population) of which 69.4 percentwere located in rural areas. In Côte d’Ivoire, we utilize the “Enquête niveaude vie ménages” studying 10,801 of the existing 3.2 million households inthe country. Households classified as rural were 47.9 percent of the total.In Ghana, we use the “Ghana living standards survey” of 1998. This surveyreviewed the standard of living of 5,998 Ghanaian households, 63.3 percent ofthem residing in rural areas. Information about Malawian households is takenfrom the “Integrated household survey” of 2004. This survey covers 11,280households (coverage rate of 0.42 percent), 87.2 percent of which were in ruralareas. In Rwanda, the most recent available survey is the “Enquête intégralesur les conditions de vie des ménages” from 1998. This survey covers 6,420households amounting to 0.4 percent of the household population. House-holds in rural areas are 82.1 percent of the total households interviewed. The


“Uganda national household survey” of 2005 interviewed 7,425 householdsfrom a population of 5.2 million households. The share of rural householdsis 77.12 percent. Finally, in the case of Zambia, we use the “Living conditionsmonitoring survey III” from 2003. This survey covers 4,837 households (a cov-erage rate of 0.23 percent) of which 47.9 percent were located in rural areas.

Table A2.4 presents a brief demographic characterization of the targetcountries. All of these countries are relatively small in terms of population.Benin is the smallest country with 6.7 million inhabitants and, with a popula-tion of 28.9 million, Uganda is the largest. In all countries the average age ofthe population is very low, ranging from 19.5 years in Uganda to 24.4 yearsin Burkina Faso. Except for the case of Burkina Faso, where they are aboutthe same, in all the other countries the rural population largely surpassesthe urban population. The rural population is on average younger than theurban population, except in Côte d’Ivoire, Malawi, Rwanda, and Zambia. Thenational male share oscillates between 0.464 in Rwanda and 0.498 in Côted’Ivoire. There is no discernable pattern for male shares across urban andrural areas. The male share is larger than the female shares only in urbanareas of Burundi and Malawi.

Average household size and its age composition are presented in Table A2.5at the national, urban, and rural level. Ghana has the smallest average house-hold size, with 4.4 members; Burkina Faso has the largest, with 5.6 members.Rural households are, on average, larger than urban households, except inRwanda and Zambia. Rural household size ranges from 4.7 to 5.9 members(with an average of 5.2) while urban household size ranges from 4.0 to 5.5members (with an average of 4.8). For the countries in the study, the 0–15age group comprises 49.1 percent of all household members in rural areasand 42.5 percent in urban areas. The 16–29 age group represents 22.5 per-cent and 29.8 percent of the rural and urban households, respectively. Thosebetween thirty and forty-nine years of age represent, on average, 17.7 percentand 19.8 percent of the household members in rural and urban areas. Thelast age group, those fifty years old or older, represents only 9.6 percent and7.3 percent of the members of rural and urban households, respectively. Thedemographic age structure of rural and urban households is similar across alltarget countries with the exception of Burkina Faso where the 16–29 age grouphas a larger share than the 0–15 group in the members of urban households.

3 THE DISTRIBUTION OF INCOME AND EXPORT CROP INCOME

In what follows, we use the household survey data to characterize the distri-bution of income in the target countries. Since we are interested in the impacton poverty of changes in the supply chain in export agriculture, we begin hereby plotting densities of per capita expenditures. These densities are estimatednonparametrically with kernel methods (Deaton 1997; Pagan and Ullah 1999).

Case Studies 15

The results are reports in Figures A2.1–A2.8, with one figure for each coun-try in the study. Each figure has three panels. Panel (a) reports the densityfor per capita expenditures for the total population at the national, rural,and urban level. Since we are also interested in gender-specific impacts, weestimate those densities for male-headed and female-headed households inpanels (b) and (c), respectively.

There are three key conclusions that emerge from the examination ofthe per capita expenditure densities. First, there are significant differencesbetween urban and rural households. The urban densities are always shiftedto the right of the rural densities, both in male-headed and female-headedhouseholds. This is particularly evident in the case of Burkina Faso, Ghana,Malawi, and Rwanda. Second, the rural density is similar to the national den-sity in most cases (with the exception of Côte d’Ivoire), again regardless of thegender of the household head. This is consistent with the data in Table A2.4,which shows that the rural population is much larger than the urban popula-tion for most of the countries in our study. Third, male-headed householdstypically enjoy higher levels of per capita expenditure level, both for urbanand rural households, than female-headed households.

We now turn to the analysis of households’ income shares. Results arereported in Tables A2.6–A2.13. We present the descriptive statistics for thetotal population of households (panel (a)) and separately for male-headed(panel (b)) and female-headed households (panel (c)). The bottom panelspresent the summary statistics for the subsample that includes only thosehouseholds that produce at least one of the crops under study. The first col-umn in each table reports statistics from the national sample, the second col-umn reports statistics from the urban sample, and the third column reportsstatistics from the rural sample. Since our focus is mainly on rural house-holds, we also report statistics across quintiles of per capita expenditure forrural households. In each table, the first row reports the average per capitaexpenditures and the following rows report the average income shares (inpercentage). We identify the share of income derived from agriculture and,within this category, we also report the share derived from the export cropunder study. For completeness, we report the share of income derived fromhome-production activities as well as from other sources (wages, nonfarmbusinesses, and transfers).

Data for Benin is in Table A2.6. The share of agricultural income in totalincome for rural households is 34 percent for the whole sample and 56.3 per-cent for those households that produce cotton, the main export crop of thecountry. Cotton generates 6.6 percent of total income for the average ruralhousehold, and 33.8 percent for the average rural cotton producer. This cropis particularly important for the poorest farmers since it generates 10.4 per-cent and 7.1 percent of total income for households in the first and sec-ond quintiles of the income distribution, respectively. Among producers, cot-ton generates about one-third of the total income in each of the quintiles.


Panels (b) and (c) highlight some of the differences in income shares betweenmale-headed and female-headed households. Cotton is mostly an importantsource of income for female-headed households, accounting for 7.6 percentof their total income—in contrast to 1.1 percent in male-headed households—for the total rural population. Conditional of being a cotton producer, cottonincome accounts for 33.3 percent and 21.8 percent of total income in female-headed and male-headed households, respectively.

Table A2.7 presents income shares for the case of Burkina Faso. In ourdata, agricultural sales play a minor role, relative, for example, to incomefrom home production. However, if we consider only the subsample of pro-ducers, agricultural income amounts to 70.5 percent of the total rural house-hold income. The most important crop is cotton, which generates 1.31 per-cent of the income of the average rural household and 56.4 percent of theincome of the average cotton producer. Among producers, cotton is a moreimportant source of income for male-headed households (56.4 percent) thanfor female-headed households (17.7 percent). For male-headed households,cotton generates a similar share of income across quintiles. Instead, female-headed households in the second quintile earn a significantly higher share ofincome from cotton than the rest of the quintiles (31 percent of their totalincome).

The case of Côte d’Ivoire is presented in Table A2.8. Around 52 percent ofthe total income of rural households comes from agriculture; for export cropproducers (any of the three major crops in Côte d’Ivoire), agriculture accountsfor 77 percent of total income. The relevant export crops are cocoa (withan income share of 17.1 percent), coffee (6.8 percent), and cotton (4.2 per-cent). Conditional on being an export crop producer, the income shares are38.5 percent, 15.3 percent, 9.5 percent, for cocoa, coffee, and cotton, respec-tively. The share of cocoa is similar across the first four quintiles, but declinesfor the richest rural households. Coffee is particularly important for house-holds in the first and second quintiles. In contrast, the importance of cottonas an income generator increases with household income: while householdsin the first quintile only derive 0.8 percent of their income from cotton, thecotton share for households in the last quintile is 8.6 percent. On average,rural male-headed households depend more heavily on agricultural income(64.6 percent). Rural female-headed households only get 36.8 percent of theirincome from agriculture, cocoa being the only significant contributor (with4.1 percent of the household income). Conditional on being a producer, agri-cultural income is important for both genders (78.1 percent for male-headedand 63.2 percent for female-headed households).

Rural households in Ghana (Table A2.9) receive 29 percent of their incomefrom agricultural activities (45.9 percent if they are cocoa producer). Agricul-tural income is particularly important for the first two quintiles, where house-holds get one-third of their income from agriculture; these shares declinesharply for households in the last quintile, who get only 23 percent of their

Case Studies 17

income from this activity. Conditional on being a cocoa producer, the averagerural household gets a similar share of income from agriculture regardlessof the quintile. For the average rural household, cocoa contributes 4.8 per-cent of the total income in male-headed households and 2.8 percent of theincome in female-headed households (24.1 percent and 22.2 percent respec-tively for the subsample of producers). For both genders, households in thethird quintile are the ones that more heavily depend on cocoa with 6.3 percent(male-headed) and 4.2 percent (female-headed).

Malawi (Table A2.10) is one of the countries in our study with the low-est income share coming from agricultural sales, with only 11.6 percent forthe average rural household (34.3 percent for producers). This share is evensmaller for female-headed households at 7 percent (29.4 percent for produc-ers). Most of the income of rural households comes from home-productionactivities. This source of income declines with the level of consumption ofthe household and the share of agricultural income increases up to 15.4 per-cent for the average rural household in the last quintile (42.3 percent forproducers). Tobacco (with a share of 3.8 percent) and cotton (with a share of0.5 percent), are the most important crops (the figures are 21.8 percent and2.7 percent respectively for producers).

On average, Rwandan rural households get 18.8 percent of their incomefrom the commercialization of agriculture products (Table A2.11). This shareincreases to 28.3 for the households producing the export crops in our study.The share for male-headed households is slightly higher, at 20.1 percent. Asin the case of Malawi, the share of agriculture income increases with the levelof household consumption. Those in the last quintile get 25.8 percent of theirincome from agriculture. Among producers, the agriculture share does notdrastically change across quintiles. Coffee is the main cash crop, contributingslightly less than 1 percent of total household income. Among producers,instead, coffee—with an average share of 8 percent—is an important sourceof household income, in particular for households in the first quintile (with ashare of 11.57 percent).

The case of Uganda (Table A2.12) is very similar to the case of Rwanda. Onaverage, a rural household generates 15.1 percent of its income from agricul-tural products but this average is very different across quintiles. The share ofagricultural income for rural households in the first quintile is only 5 percent,while, for those in the fifth quintile, the share is 25.7 percent. Among pro-ducers we observe a similar pattern, but the shares are higher. Male-headedhouseholds rely more on agricultural income than female-headed households.Coffee is also the main cash crop here, contributing 2.4 percent of the ruralhousehold income (8.2 percent if they are producers).

The last country in our analysis is Zambia (Table A2.13). Around one-fifthof rural household income comes from the commercialization of agriculturalproducts (36.4 percent among producers). This percentage is similar acrossquintiles, with slightly higher shares for the poorest households in the total


sample, and slightly higher (21.8 percent and 2.7 percent respectively forproducers) shares for the richest households among producers. Cotton andtobacco are the main sources of agricultural income. Cotton contributes, onaverage, 3.2 percent (23.3 percent among producers) of the total income inmale-headed rural households and 2.1 percent (23 percent for producers) inthe case of female-headed households. The contribution of tobacco is 0.8 per-cent and 0.4 percent for male-headed and female-headed rural households(5.9 percent and 4.6 percent, respectively, for the subsample of producers).

To show the importance of the relevant crops for the rural households inthe eight target countries across the entire income distribution more effec-tively, Figures 2.9 to 2.16 display nonparametric regressions of the incomeshares derived from different export crops on the log of per capita householdexpenditures. These regressions are estimated using local polynomials (seePagan and Ullah 1999). For each crop–country pair, we estimate this regres-sion for the total rural sample, and for the subsamples of male-headed andfemale-headed households.

In Benin (Figure A2.9), the share of income coming from cotton declineswith the level of per capita expenditure of the household. The decline is morepronounced in the case of female-headed households. On the other hand, inthe case of Burkina Faso (Figure A2.10), the importance of the income sharederived from cotton grows with the level of per capita expenditure (partic-ularly for male-headed households). The share of coffee and cocoa declinewith per capita expenditure in Côte d’Ivoire (Figure A2.11); cotton shares, incontrast, monotonically increase. In Ghana, the share of cocoa income firstincreases with income, but then declines at the right tail of the income distri-bution (Figure A2.12). Figure A2.13 shows that the share of income generatedby tobacco increases with the level of per capita expenditure of the typicalMalawian rural household, whereas the share of income coming from cottonhas an inverted U-shape. In Rwanda (Figure A2.14), on average, the share ofincome from coffee increases with the level of expenditure of the rural house-hold. We observe a similar pattern for coffee in Uganda (Figure A2.15) butwith a decline in the share of coffee for the richest rural households. In Zam-bia (Figure A2.16) both the share of income from tobacco and cotton increasewith the level of per capita consumption of the rural household.

Case Studies 19

APPENDIX: TABLES AND FIGURES

0

0.2

0.4

6 8 10 12

(a)

00.20.40.6

6 8 10 12

(b)

0

0.2

0.4

6 8 10 12Log per capita expenditure

(c)

Log per capita expenditure


National Rural Urban

Figure A2.1: Per capita expenditure density for Benin.


6 8 10 12

6 8 10 12


0

0.2

0.4

0

0.2

0.4

0

0.2

0.4

(a)

(b)

(c)




Figure A2.2: Per capita expenditure density for Burkina Faso.

Case Studies 21

0

0.2

0.4

7 8 9 10 11 12

(a)

0

0.2

0.4

7 8 9 10 11 12

(b)

0

0.2

0.4

7 8 9 10 11 12Log per capita expenditure

(c)




Figure A2.3: Per capita expenditure density for Côte d’Ivoire.


0

0.2

0.4

9 10 11 12 13

(a)

00.20.40.6

9 10 11 12 13

(b)

0

0.2

0.4

9 10 11 12 13


(c)




Figure A2.4: Per capita expenditure density for Ghana.

Case Studies 23

0

0.2

0.4

5 6 7 8 9

(a)

0

0.2

0.4

5 6 7 8 9

(b)

0

0.2

0.4

5 6 7 8 9Log per capita expenditure

(c)




Figure A2.5: Per capita expenditure density for Malawi.


0

0.2

0.4

4 6 8 10 12

(a)

0

0.2

0.4

4 6 8 10 12

(b)

0

0.2

0.4


(c)




Figure A2.6: Per capita expenditure density for Rwanda.

Case Studies 25

0

0.2

0.4

6 8 10 12

(a)

0

0.2

0.4

6 8 10 12

(b)

0

0.2

0.4


(c)




Figure A2.7: Per capita expenditure density for Uganda.


0

0.4

0.8

10 11 12 13 14

(a)

10 11 12 13 14

(b)

00.20.40.6


(c)

0

0.4

0.8




Figure A2.8: Per capita expenditure density for Zambia.

Case Studies 27

0

5

10

15

6 7 8 9 10 11


Sha

re o

f inc

ome

All Male head Female head

Figure A2.9: Share of income and per capita expenditure for Benin: cotton.


0.5

1.0

1.5

2.0

2.5

10 11

Sha

re o

f inc

ome


06 987


Figure A2.10: Share of income and per capita expenditure for Burkina Faso: cotton.

Case Studies 29

0

5

10

15

7 8 9 10 11 12

Sha

re o

f inc

ome

5

10

15

20

25

7 8 9 10 11 12

Sha

re o

f inc

ome

0

2

4

6

8

10


Sha

re o

f inc

ome

(a)

(b)

(c)




Figure A2.11: Share of income and per capita expenditurefor Côte d’Ivoire: (a) coffee; (b) cocoa; (c) cotton.


0

2

4

6

9 10 11 12 13


Sha

re o

f inc

ome


Figure A2.12: Share of income and per capita expenditure for Ghana: cocoa.

Case Studies 31

0

2

4

6

8

5 6 7 8 9

Sha

re o

f inc

ome

0

0.2

0.4

0.6

0.8


Sha

re o

f inc

ome

(a)

(b)



Figure A2.13: Share of income and per capita expenditurefor Malawi: (a) tobacco; (b) cotton.


0.6

0.7

0.8

0.9

1.0

1.1


Sha

re o

f inc

ome


Figure A2.14: Share of income and per capita expenditure for Rwanda: coffee.

Case Studies 33

0

1

2

3

4

6 8 10 12


Sha

re o

f inc

ome


Figure A2.15: Share of income and per capita expenditure for Uganda: coffee.


0

0.5

1.0

1.5

2.0

9 10 11 12 13

Sha

re o

f inc

ome

1.5

2.0

2.5

3.0

3.5

9 10 11 12 13


Sha

re o

f inc

ome

(a)

(b)



Figure A2.16: Share of income and per capita expenditurefor Zambia: (a) tobacco; (b) cotton.

Case Studies 35

Table A2.1: Agricultural exports in sub-Saharan Africa.

Country – crop 2004 2005 2006

(1) Benin – cotton

Cotton exports (thousands of $) 208,128 181,448 113,306– Percentage of agricultural exports 75 69 33– Percentage of total exports 39 31 N/A

Exports of goods and services as 13 13 N/Apercentage of GDP

(2) Burkina Faso – cotton

Cotton exports (thousands of $) 270,329 213,614 235,290– Percentage of agricultural exports 81 78 81– Percentage of total exports 49 39 35

Exports of goods and services as 11 10 12percentage of GDP

(3) Côte d’Ivoire – cocoa, coffee, cotton

Cocoa exports (thousands of $) 2,124,649 1,988,939 1,946,273– Percentage of agricultural exports 68 66 62– Percentage of total exports 28 24 21

Coffee exports (thousands of $) 130,255 113,436 166,007– Percentage of agricultural exports 4 4 5– Percentage of total exports 2 1 2



(4) Ghana – cocoa

Cocoa exports (thousands of $) 984,034 914,605 1,224,309– Percentage of agricultural exports 77 79 79– Percentage of total exports 28 26 27



Table A2.1: Continued.

Country – crop 2004 2005 2006

(5) Malawi – tobacco, cotton


Tobacco exports (thousands of $) 257,974 320,715 431,787– Percentage of agricultural exports 64 71 73– Percentage of total exports 39 57 73


(6) Rwanda – coffee



(7) Uganda – coffee



(8) Zambia – cotton, tobacco


Tobacco exports (thousands of $) 60,383 63,473 75,205– Percentage of agricultural exports 17 20 23– Percentage of total exports 3 3 2


Source: FAO and WDI.

Case Studies 37

Table A2.2: The composition of agricultural exports(2006; all values are percentages).

Burkina CôteCrops Benin Faso d’Ivoire Ghana Malawi Rwanda Uganda Zambia

(1) Cotton

Carded, 0 0 0 — 0 — — 1combedLint 92 96 93 — 94 — — 92Linter 0 0 0 — 0 — — 0Waste 0 0 0 — 0 — — 0Seed 1 1 5 — 6 — — 7Seed oil 7 3 1 — 0 — — 0

(2) Tobacco

Unmanufac. — — — — 100 — — 100Manufac. — — — — 0 — — 0

(3) Coffee

Extracts — — 33 — — 0 0 —Husks, — — 0 — — 0 0 —skinsRoasted — — 0 — — 0 0 —Substitutes — — 0 — — 0 0 —Green — — 66 — — 100 99 —

(4) Cocoa

Beans — — 73 87 — — — —Butter — — 9 6 — — — —Paste — — 12 6 — — — —Husks, — — 4 0 — — — —shellPowder, — — 2 1 — — — —cake

Source: FAO.


Tab

leA

2.3

:Lis

tof

sub-S

ah

ara

nA

fric

ah

ouse

hol

dsu

rvey

s.

Ru

ral

Sam

ple

shar

esh

are

Hou

seh

old

sC

ou

ntr

yY

ear

Surv

eyH

ou

seh

old

s(%

)(%

)(m

illi

on

s)

Ben

in2003

Qu

esti

on

nai

red

esin

dic

ateu

rsd

eb

ase

5.3

50

61.5

30.3

91.4

du

bie

n-ê

tre

Bu

rkin

aFa

so2003

Enq

uêt

eB

urk

inab

esu

rle

sco

nd

itio

ns

8.5

00

69.4

10.4

81.8

de

vie

des

mén

ages

Côte

d’Ivo

ire

2002

Enq

uêt

en

ivea

ud

evi

em

énag

es10.8

01

47.8

70.3

43.2

Gh

ana

1998

Gh

ana

livi

ng

stan

dar

ds

surv

ey5.9

98

63.3

40.1

44.4

Mal

awi

2004

Inte

gra

ted

hou

seh

old

surv

ey11.2

80

87.2

30.4

22.7

Rw

and

a1998

Enq

uêt

ein

tégra

lesu

rle

sco

nd

itio

ns

6.4

20

82.1

00.4

01.6

de

vie

des

mén

ages

Ugan

da

2005

Ugan

da

nat

ion

alh

ou

seh

old

surv

ey7.4

25

77.1

20.1

45.2

Zam

bia

2003

Livi

ng

con

dit

ion

sm

on

itori

ng

surv

eyII

I4.8

37

47.8

80.2

32.1

Case Studies 39

Tab

leA

2.4

:Pop

ula

tion

,age

an

dgen

der

com

pos

itio

n.

Nat

ion

alU

rban

Ru

ral

︷︸︸

︷︷

︸︸︷

︷︸︸

︷C

ou

ntr

yPop

ula

tion

Age

Mal

ePop

ula

tion

Age

Mal

ePop

ula

tion

Age

Mal

e

Ben

in6.7

21.7

48.7

2.4

22.7

48.7

4.3

21.1

48.7

Bu

rkin

aFa

so9.9

24.4

48.5

1.9

25.2

49.6

8.0

24.2

48.3

Côte

d’Ivo

ire

17.1

22.2

49.8

8.6

21.7

49.8

8.3

22.7

49.8

Gh

ana

19.5

23.4

47.8

6.8

24.1

46.8

12.8

23.0

48.3

Mal

awi

12.5

21.3

49.2

1.4

20.6

51.1

11.1

21.4

49

Rw

and

a8.0

21.0

46.4

0.8

19.9

46.9

7.1

21.2

46.3

Ugan

da

28.9

19.5

48.8

4.6

19.8

47.8

24.3

19.4

49.0

Zam

bia

11.2

20.9

49.0

4.7

20.7

49.2

6.6

21.0

48.9

Sou

rce:

bas

edon

the

hou

seh

old

surv

eys

(Tab

leA

2.3

).


Tab

leA

2.5

:D

emog

raph

icdec

ompos

itio

n(h

ouse

hol

dsi

zeby

age

gro

up).

Nat

ion

alU

rban

Ru

ral

︷︸︸

︷︷

︸︸︷

︷︸︸

︷C

ou

ntr

yA

ll0–1

516–2

930–4

950+

All

0–1

516–2

930–4

950+

All

0–1

516–2

930–4

950+

Ben

in4.9

2.4

1.1

1.0

0.5

4.5

2.0

1.1

0.9

0.4

5.3

2.7

1.0

1.0

0.5

Bu

rkin

aFa

so5.6

2.3

1.5

1.1

0.6

5.0

1.6

1.7

1.1

0.5

5.7

2.5

1.4

1.1

0.7

Côte

d’Ivo

ire

5.4

2.4

1.5

1.1

0.5

4.9

2.0

1.5

1.0

0.3

5.9

2.7

1.4

1.1

0.6

Gh

ana

4.4

2.0

1.0

0.9

0.5

4.0

1.7

1.0

0.8

0.4

4.7

2.3

0.9

0.9

0.6

Mal

awi

4.6

2.2

1.2

0.7

0.5

4.4

1.9

1.5

0.8

0.2

4.7

2.3

1.1

0.7

0.5

Rw

and

a5.0

2.4

1.3

0.9

0.4

5.2

2.3

1.7

0.9

0.3

4.9

2.4

1.2

0.8

0.4

Ugan

da

5.5

2.9

1.3

0.9

0.4

5.0

2.3

1.6

0.8

0.3

5.6

3.0

1.3

0.9

0.4

Zam

bia

5.3

2.5

1.4

0.9

0.4

5.5

2.5

1.7

1.0

0.3

5.2

2.5

1.3

0.8

0.5

Sou

rce:

bas

edon

the

hou

seh

old

surv

eyd

ata

(Tab

leA

2.3

).

Case Studies 41

Tab

leA

2.6

:Per

capit

aex

pen

dit

ure

an

din

com

eso

urc

esin

Ben

in.

Nat

ion

alU

rban

Ru

ral

Sam

ple

tota

lto

tal

tota

lq 1

q 2q 3

q 4q 5

(a)

Nati

onal

Per

cap

ita

exp

end

itu

re14,5

57

21,0

80

10,1

93

2,0

54

4,7

19

7,3

25

11,1

17

26,0

82

Shar

eag

ricu

ltu

re25.8

413.3

234.0

438.9

237.7

035.5

332.9

124.7

8–

Cott

on

5.3

53.4

76.5

910.3

97.0

56.7

14.5

83.9

8Sh

are

hom

e-p

rod

uct

ion

26.7

922.6

429.5

044.3

231.6

027.0

822.5

821.1

3Sh

are

oth

er47.3

764.0

436.4

616.7

630.7

37.3

944.5

154.0

9

(b)

Male

-hea

ded

Per

cap

ita

exp

end

itu

re16,8

50

21,8

37

12,6

64

2,2

78

4,7

46

7,3

22

11,1

61

26,2

17

Shar

eag

ricu

ltu

re11.5

04.4

217.3

627.1

821.8

718.4

618.3

49.2

1–

Cott

on

0.9

00.6

11.1

44.0

71.2

70.3

81.3

20.2

6Sh

are

hom

e-p

rod

uct

ion

18.9

914.3

722.8

137.8

128.3

926.1

415.4

516.7

3Sh

are

oth

er69.5

181.2

159.8

335.0

149.7

455.4

66.2

174.0

6

(c)

Fem

ale

-hea

ded

Per

cap

ita

exp

end

itu

re14,0

99

20,9

03

9,7

53

2,0

33

4,7

15

7,3

25

11,1

08

26,0

43

Shar

eag

ricu

ltu

re28.7

115.4

137.0

140.0

040.2

438.4

035.8

529.3

2–

Cott

on

6.2

44.1

47.5

610.9

77.9

77.7

75.2

45.0

6Sh

are

hom

e-p

rod

uct

ion

28.3

524.5

930.7

044.9

232.1

127.2

424.0

222.4

1Sh

are

oth

er42.9

460

32.2

915.0

827.6

534.3

640.1

348.2

7


Tab

leA

2.6

:C

onti

nu

ed.

Nat

ion

alU

rban

Ru

ral

Sam

ple

tota

lto

tal

tota

lq 1

q 2q 3

q 4q 5

(d)

Pro

du

cers

Per

cap

ita

exp

end

itu

re8,1

39

10,2

51

7,3

25

1,8

90

4,6

99

7,2

98

11,1

06

23,4

07

Shar

eag

ricu

ltu

re56.3

55.5

556.5

351.7

954.7

457.4

64.2

864.8

5–

Cott

on

33.7

636.5

632.8

930.8

632.1

633.9

434.6

336.9

9

(e)

Pro

du

cers

(male

s)

Per

cap

ita

exp

end

itu

re7,9

25

11,0

48

6,1

38

1,9

13

4,8

87

7,2

94

10,3

30

20,9

60

Shar

eag

ricu

ltu

re54.1

744.1

257.2

458.4

258.7

360.5

155.3

634.4

7–

Cott

on

24.0

831.6

321.7

720.8

217.1

811.5

437.2

33.6

1

(f)

Pro

du

cers

(fem

ale

s)

Per

cap

ita

exp

end

itu

re8,1

50

10,1

98

7,3

77

1,8

89

4,6

90

7,2

98

11,1

51

23,4

73

Shar

eag

ricu

ltu

re56.3

956.0

256.5

51.4

454.5

557.3

264.7

165.3

6–

Cott

on

34.1

536.7

633.3

531.3

832.8

934.4

934.5

137.0

4

Sou

rce:

bas

edon

hou

seh

old

surv

eys

(see

Tab

leA

2.3

).

Case Studies 43

Tab

leA

2.7

:Per

capit

aex

pen

dit

ure

an

din

com

eso

urc

esin

Bu

rkin

aFa

so.

Nat

ion

alU

rban

Ru

ral

Sam

ple

tota

lto

tal

tota

lq 1

q 2q 3

q 4q 5

(a)

Nati

onal

Per

cap

ita

exp

end

itu

re12,1

63

26,1

87

8,4

80

1,2

43

2,6

41

4,3

99

7,4

29

25,7

56

Shar

eag

ricu

ltu

re4.1

51.4

84.7

32.8

52.9

83.8

65.8

87.9

1–

Cott

on

1.1

00.1

31.3

11.0

80.9

31.0

41.2

92.2

1Sh

are

hom

e-p

rod

uct

ion

48.5

916.2

155.6

066.6

759.5

755.9

750.3

646.0

8Sh

are

oth

er47.2

682.3

139.6

730.4

837.4

540.1

743.7

646.0

1

(b)

Male

-hea

ded

Per

cap

ita

exp

end

itu

re11,7

69

25,5

93

8,4

70

1,2

50

2,6

47

4,3

94

7,4

41

25,3

59

Shar

eag

ricu

ltu

re4.4

01.6

34.9

33.0

03.1

14.0

06.1

38.1

7–

Cott

on

1.1

90.1

51.3

91.1

70.9

71.1

1.3

72.3

2Sh

are

hom

e-p

rod

uct

ion

50.1

717.7

856.4

267.5

460.4

657.0

451.0

047.0

6Sh

are

oth

er45.4

380.5

938.6

529.4

636.4

338.9

642.8

744.7

7

(c)

Fem

ale

-hea

ded

Per

cap

ita

exp

end

itu

re16,6

12

29,5

79

8,6

29

1,1

57

2,5

57

4,4

68

7,2

23

33,0

17

Shar

eag

ricu

ltu

re1.3

40.7

21.6

90.9

51.1

91.9

01.8

32.9

5–

Cott

on

0.1

00.1

50

0.4

10.1

30

0.2

2Sh

are

hom

e-p

rod

uct

ion

30.6

88.2

143.3

155.7

947.3

840.8

539.8

827.6

4Sh

are

oth

er67.9

891.0

755.0

043.2

651.4

357.2

558.2

969.4

1


Tab

leA

2.7

:C

onti

nu

ed.

Nat

ion

alU

rban

Ru

ral

Sam

ple

tota

lto

tal

tota

lq 1

q 2q 3

q 4q 5

(d)

Pro

du

cers

Per

cap

ita

exp

end

itu

re26,0

34

31,1

36

18,2

45

1,4

51

2,6

93

4,5

11

7,6

24

38,7

53

Shar

eag

ricu

ltu

re70.7

991.0

970.5

271.5

264.2

468.8

671.6

473.1

1–

Cott

on

55.9

287.4

155.5

158.7

952.6

849.7

756.2

57.9

3

(e)

Pro

du

cers

(male

s)

Per

cap

ita

exp

end

itu

re25,6

08

30,8

60

17,8

40

1,4

60

2,6

55

4,5

32

7,6

11

37,6

98

Shar

eag

ricu

ltu

re71.7

291.0

971.4

671.5

265.2

670.6

571.6

474.4

6–

Cott

on

56.7

887.4

156.3

758.7

953.8

451.6

156.2

58.9

6

(f)

Pro

du

cers

(fem

ale

s)

Per

cap

ita

exp

end

itu

re31,4

83

34,1

40

25,2

45

1,3

07

2,9

87

3,8

85

8,0

98

56,6

90

Shar

eag

ricu

ltu

re29.7

5—

29.7

5—

45.1

729.4

6—

14.2

6–

Cott

on

17.7

4—

17.7

4—

31.0

39.1

5—

13.0

7

Sou

rce:

bas

edon

hou

seh

old

surv

eys

(see

Tab

leA

2.3

).

Case Studies 45

Tab

leA

2.8

:Per

capit

aex

pen

dit

ure

an

din

com

eso

urc

esin

Côt

ed’Ivo

ire.

Nat

ion

alU

rban

Ru

ral

Sam

ple

tota

lto

tal

tota

lq 1

q 2q 3

q 4q 5

(a)

Nati

onal

Per

cap

ita

exp

end

itu

re29,9

31

35,7

56

23,3

86

3,6

25

8,1

09

13,5

83

23,0

98

70,9

70

Shar

eag

ricu

ltu

re41.3

629.4

751.9

959.5

357.5

950.1

148.7

041.5

0–

Cott

on

2.8

91.3

64.2

40.8

12.3

64.9

15.5

88.6

1–

Coff

ee5.3

93.7

56.8

111.9

19.0

04.9

54.1

62.8

1–

Coco

a14.6

011.6

917.1

217.5

317.6

718.8

718.4

112.4

8Sh

are

hom

e-p

rod

uct

ion

4.6

42.7

56.3

48.7

48.3

85.8

73.9

24.0

0Sh

are

oth

er54

67.7

841.6

731.7

334.0

344.0

247.3

854.5

(b)

Male

-hea

ded

Per

cap

ita

exp

end

itu

re30,3

81

35,9

67

24,1

12

3,6

41

8,1

51

13,5

36

23,1

31

71,7

27

Shar

eag

ricu

ltu

re45.3

033.1

356.1

164.6

361.3

555.2

952.7

444.2

9–

Cott

on

3.3

81.6

44.9

30.9

32.3

36.0

46.5

99.7

6–

Coff

ee6.3

34.4

07.9

814.4

110.5

55.9

44.7

92.8

9–

Coco

a17.0

213.7

919.8

220.3

820.4

122.5

421.7

413.5

4Sh

are

hom

e-p

rod

uct

ion

4.9

02.9

36.6

69.1

08.8

36.2

34.2

04.2

1Sh

are

oth

er49.8

63.9

437.2

326.2

729.8

238.4

843.0

651.5

(c)

Fem

ale

-hea

ded

Per

cap

ita

exp

end

itu

re27,6

06

34,6

11

19,7

98

3,5

51

7,8

95

13,7

88

22,9

55

65,6

75

Shar

eag

ricu

ltu

re22.4

311.7

432.1

136.7

538.6

128.0

831.0

422.4

5–

Cott

on

0.5

00.0

30.9

20.2

52.5

10.0

91.1

70.7

1–

Coff

ee0.8

70.5

91.1

40.7

21.1

60.7

31.3

92.2

3–

Coco

a2.9

51.5

64.1

24.8

13.8

23.2

83.8

55.2

7Sh

are

hom

e-p

rod

uct

ion

3.4

01.8

84.7

97.1

66.1

04.3

62.6

82.5

3Sh

are

oth

er74.1

786.3

863.1

56.0

955.2

967.5

666.2

875.0

2


Tab

leA

2.8

:C

onti

nu

ed.

Nat

ion

alU

rban

Ru

ral

Sam

ple

tota

lto

tal

tota

lq 1

q 2q 3

q 4q 5

(d)

Pro

du

cers

Per

cap

ita

exp

end

itu

re29,8

92

35,3

54

24,3

30

3,5

61

8,1

54

13,5

46

23,0

01

73,3

79

Shar

eag

ricu

ltu

re77.9

779.3

777.3

675.4

778.6

579.5

576.6

776.3

1–

Cott

on

8.2

35.6

19.5

41.7

45.1

611.2

512.2

721.2

4–

Coff

ee15.3

715.4

315.3

25.6

519.6

611.3

49.1

36.9

3–

Coco

a41.6

348.1

438.4

837.7

838.5

943.2

640.4

630.8

(e)

Pro

du

cers

(male

s)

Per

cap

ita

exp

end

itu

re29,2

15

34,2

54

24,1

54

3,5

44

8,1

64

13,5

18

22,9

47

73,3

23

Shar

eag

ricu

ltu

re78.5

779.7

478.0

776.0

479.3

680.5

277.0

877.2

4–

Cott

on

8.3

35.7

49.6

51.7

24.4

811.7

312.4

22.2

7–

Coff

ee15.5

915.4

515.6

26.4

920.2

511.5

39

6.6

1–

Coco

a41.9

348.4

38.7

437.4

739.1

643.7

540.8

730.8

9

(f)

Pro

du

cers

(fem

ale

s)

Per

cap

ita

exp

end

itu

re37,1

38

45,9

30

26,4

68

3,8

21

8,0

23

13,9

20

23,5

22

74,0

11

Shar

eag

ricu

ltu

re64.2

166.7

963.1

762.4

665.0

558.4

68.5

860.9

2–

Cott

on

5.9

50.8

67.4

22.3

718.1

70.8

89.8

43.9

7–

Coff

ee10.2

914.6

69.2

26.6

68.3

97.2

611.7

112.3

8–

Coco

a34.9

239.1

133.2

844.7

327.6

532.6

132.5

29.2

5

Sou

rce:

bas

edon

hou

seh

old

surv

eys

(see

Tab

leA

2.3

).

Case Studies 47

Tab

leA

2.9

:Per

capit

aex

pen

dit

ure

an

din

com

eso

urc

esin

Gh

an

a.

Nat

ion

alU

rban

Ru

ral

Sam

ple

tota

lto

tal

tota

lq 1

q 2q 3

q 4q 5

(a)

Nati

onal

Per

cap

ita

exp

end

itu

re93,5

15

136,0

77

66,9

40

14,7

53

27,4

26

41,3

44

63,7

63

170,2

49

Shar

eag

ricu

ltu

re22.4

711.6

328.9

633.4

732.4

930.2

126.2

123.1

8–

Coco

a2.9

10.8

44.1

52.3

94.4

25.6

94.2

63.8

2Sh

are

hom

e-p

rod

uct

ion

24.6

28.7

334.1

547.1

040.1

335.2

129.3

121.0

3Sh

are

oth

er52.9

179.6

436.8

919.4

327.3

834.5

844.4

855.7

9

(b)

Male

-hea

ded

Per

cap

ita

exp

end

itu

re93,4

05

141,8

95

65,3

94

14,4

87

27,2

20

41,1

61

63,4

36

170,6

81

Shar

eag

ricu

ltu

re24.1

110.6

631.5

336.0

036.2

432.9

227.7

424.9

2–

Coco

a3.4

41.0

44.7

62.3

95.2

46.3

04.9

84.7

6Sh

are

hom

e-p

rod

uct

ion

25.3

68.4

834.6

747.9

438.0

336.0

529.0

522.8

5Sh

are

oth

er50.5

380.8

633.8

16.0

625.7

331.0

343.2

152.2

3

(c)

Fem

ale

-hea

ded

Per

cap

ita

exp

end

itu

re93,7

40

125,7

74

70,3

96

15,4

80

27,8

95

41,7

98

64,3

76

169,3

75

Shar

eag

ricu

ltu

re19.0

913.3

523.1

426.5

923.9

423.5

323.2

819.5

7–

Coco

a1.8

20.4

92.7

62.4

02.5

64.1

82.8

91.8

8Sh

are

hom

e-p

rod

uct

ion

23.1

29.1

832.9

644.8

244.8

933.1

129.8

217.2

6Sh

are

oth

er57.7

977.4

743.9

28.5

931.1

743.3

646.9

63.1

7


Tab

leA

2.9

:C

onti

nu

ed.

Nat

ion

alU

rban

Ru

ral

Sam

ple

tota

lto

tal

tota

lq 1

q 2q 3

q 4q 5

(d)

Pro

du

cers

Per

cap

ita

exp

end

itu

re83,1

22

124,1

21

71,2

97

15,8

29

27,4

37

40,8

59

63,3

74

192,8

75

Shar

eag

ricu

ltu

re46.2

249.3

445.8

942.5

548.1

444.8

144.2

149.2

7–

Coco

a23.9

927.1

323.6

518.6

222.0

325.4

922.8

927.9

8

(e)

Pro

du

cers

(male

s)

Per

cap

ita

exp

end

itu

re83,5

38

125,6

81

72,3

80

15,7

30

27,1

22

40,7

69

62,9

85

211,3

79

Shar

eag

ricu

ltu

re47.5

148.4

47.4

242.2

550.4

645.9

446.2

550.9

4–

Coco

a24.3

827.3

124.0

717.0

421.9

525.0

525.2

729.9

6

(f)

Pro

du

cers

(fem

ale

s)

Per

cap

ita

exp

end

itu

re81,8

97

120,5

12

67,8

55

16,2

12

28,8

80

41,2

22

64,1

60

143,9

71

Shar

eag

ricu

ltu

re41.5

652.7

840.3

943.7

437.0

540.2

239.5

443.2

4–

Coco

a22.5

626.4

522.1

524.8

222.4

427.3

17.4

520.7

8

Sou

rce:

bas

edon

hou

seh

old

surv

eys

(see

Tab

leA

2.3

).

Case Studies 49

Tab

leA

2.1

0:

Per

capit

aex

pen

dit

ure

an

din

com

eso

urc

esin

Mala

wi.

Nat

ion

alU

rban

Ru

ral

Sam

ple

tota

lto

tal

tota

lq 1

q 2q 3

q 4q 5

(a)

Nati

onal

Per

cap

ita

exp

end

itu

re1,3

85

3,6

43

1,0

78

274

498

742

1,1

29

2,7

09

Shar

eag

ricu

ltu

re10.5

22.8

811.5

56.8

29.6

311.8

913.9

015.4

0–

Tob

acco

3.4

70.8

63.8

21.4

12.6

84.0

64.8

06.1

0–

Cott

on

0.4

10.0

00.4

70.3

30.3

70.4

90.7

20.4

4Sh

are

hom

e-p

rod

uct

ion

40.8

210.0

444.9

756.7

649.7

647.0

841.5

630.0

1Sh

are

oth

er48.6

687.0

843.4

836.4

240.6

141.0

344.5

454.5

9

(b)

Male

-hea

ded

Per

cap

ita

exp

end

itu

re1,4

88

3,7

68

1,1

42

279

499

742

1,1

28

2,7

42

Shar

eag

ricu

ltu

re11.6

42.8

612.9

78.1

110.9

913.2

214.9

816.3

9–

Tob

acco

4.0

90.8

14.5

91.9

23.3

04.8

15.3

86.8

6–

Cott

on

0.4

80.0

00.5

60.4

40.4

50.5

70.8

00.5

0Sh

are

hom

e-p

rod

uct

ion

39.2

59.0

843.7

957.0

949.2

946.2

540.8

629.1

0Sh

are

oth

er49.1

188.0

643.2

434.8

39.7

240.5

344.1

654.5

1

(c)

Fem

ale

-hea

ded

Per

cap

ita

exp

end

itu

re1,0

37

2,9

35

875

265

498

744

1,1

31

2,5

39

Shar

eag

ricu

ltu

re6.7

22.9

97.0

44.2

25.7

67.5

19.9

210.3

2–

Tob

acco

1.3

71.1

41.3

90.3

80.9

31.5

92.6

72.1

8–

Cott

on

0.1

80.0

00.2

00.1

10.1

40.2

10.4

40.1

3Sh

are

hom

e-p

rod

uct

ion

46.1

115.3

948.7

156.0

851.0

949.8

044.1

234.6

5Sh

are

oth

er47.1

781.6

244.2

539.7

43.1

542.6

945.9

655.0

3


Tab

leA

2.1

0:

Con

tin

ued

.

Nat

ion

alU

rban

Ru

ral

Sam

ple

tota

lto

tal

tota

lq 1

q 2q 3

q 4q 5

(d)

Pro

du

cers

Per

cap

ita

exp

end

itu

re1,2

55

2,8

66

1,2

05

292

496

755

1,1

32

2,5

98

Shar

eag

ricu

ltu

re34.3

936.4

234.3

423.4

728.3

533.8

36.2

42.3

5–

Tob

acco

21.9

727.9

721.8

213.6

417.2

921.6

322.3

628.5

2–

Cott

on

2.6

30

2.6

93.2

2.3

92.5

93.3

72.0

6

(e)

Pro

du

cers

(male

s)

Per

cap

ita

exp

end

itu

re1,2

94

3,2

07

1,2

38

294

498

756

1,1

35

2,6

10

Shar

eag

ricu

ltu

re34.9

134.7

934.9

124.9

729.0

134.6

636.1

142.0

7–

Tob

acco

22.3

26.3

222.2

114.6

17.9

222.1

322

28.4

7–

Cott

on

2.6

40

2.7

3.3

72.4

72.6

33.2

72.0

8

(f)

Pro

du

cers

(fem

ale

s)

Per

cap

ita

exp

end

itu

re920

845

923

277

481

745

1,1

07

2,3

98

Shar

eag

ricu

ltu

re29.9

645.6

429.4

115.2

123.6

426.9

636.9

646.9

8–

Tob

acco

19.1

237.2

818.4

97.9

212.7

917.7

25.4

429.4

8–

Cott

on

2.5

20

2.6

12.2

91.8

62.3

14.2

21.7

3

Sou

rce:

bas

edon

hou

seh

old

surv

eys

(see

Tab

leA

2.3

).

Case Studies 51

Tab

leA

2.1

1:

Per

capit

aex

pen

dit

ure

an

din

com

eso

urc

esin

Rw

an

da.

Nat

ion

alU

rban

Ru

ral

Sam

ple

tota

lto

tal

tota

lq 1

q 2q 3

q 4q 5

(a)

Nati

onal

Per

cap

ita

exp

end

itu

re6,4

87

18,0

74

5,2

15

602

1,4

51

2,5

99

5,1

00

16,2

28

Shar

eag

ricu

ltu

re17.2

52.2

618.8

411.1

514.9

119.2

423.6

125.4

1–

Coff

ee0.8

00.0

20.8

80.8

70.7

90.9

60.7

01.0

9Sh

are

hom

e-p

rod

uct

ion

42.8

96.3

546.7

656.5

050.3

047.1

041.6

638.0

4Sh

are

oth

er39.8

691.3

934.4

32.3

534.7

933.6

634.7

336.5

5

(b)

Male

-hea

ded

Per

cap

ita

exp

end

itu

re7,0

60

19,6

60

5,5

85

635

1,4

63

2,6

06

5,1

05

15,5

55

Shar

eag

ricu

ltu

re18.1

81.9

120.0

611.1

416.1

620.4

423.7

225.7

7–

Coff

ee0.8

20.0

20.9

11.0

30.7

80.9

70.7

51.0

6Sh

are

hom

e-p

rod

uct

ion

39.6

04.2

843.6

953.0

846.6

145.2

741.0

035.7

5Sh

are

oth

er42.2

293.8

136.2

535.7

837.2

334.2

935.2

838.4

8

(c)

Fem

ale

-hea

ded

Per

cap

ita

exp

end

itu

re5,2

71

13,9

93

4,4

47

562

1,4

28

2,5

84

5,0

85

18,4

19

Shar

eag

ricu

ltu

re15.2

83.2

516.3

011.1

612.6

316.6

723.2

924.2

3–

Coff

ee0.7

50.0

20.8

10.6

80.8

10.9

60.5

61.1

6Sh

are

hom

e-p

rod

uct

ion

49.9

312.2

153.1

660.5

157.0

151.0

043.6

245.5

2Sh

are

oth

er34.7

984.5

430.5

428.3

330.3

632.3

333.0

930.2

5


Tab

leA

2.1

1:

Con

tin

ued

.

Nat

ion

alU

rban

Ru

ral

Sam

ple

tota

lto

tal

tota

lq 1

q 2q 3

q 4q 5

(d)

Pro

du

cers

Per

cap

ita

exp

end

itu

re6,1

59

9,0

20

6,0

30

623

1,4

38

2,6

28

5,1

15

15,2

25

Shar

eag

ricu

ltu

re28.2

826.1

528.3

29.7

928.2

230.0

124.6

729.0

4–

Coff

ee7.9

62.1

78.0

211.5

28.7

68.0

65.9

97.3

4

(e)

Pro

du

cers

(male

s)

Per

cap

ita

exp

end

itu

re6,2

10

13,0

34

6,0

74

642

1,4

89

2,6

57

5,1

68

13,7

99

Shar

eag

ricu

ltu

re28.4

427.9

828.4

428.2

629.8

30.8

21.7

930.9

3–

Coff

ee7.4

32.4

17.4

810.3

48

7.6

86.2

86.6

7

(f)

Pro

du

cers

(fem

ale

s)

Per

cap

ita

exp

end

itu

re6,0

26

7,0

42

5,9

09

582

1,3

39

2,5

55

4,9

47

21,7

98

Shar

eag

ricu

ltu

re27.8

21.8

27.8

733.6

224.5

728.0

333.9

220.0

9–

Coff

ee9.5

71.6

29.6

714.5

10.5

18.9

95.0

410.5

6

Sou

rce:

bas

edon

hou

seh

old

surv

eys

(see

Tab

leA

2.3

).

Case Studies 53

Tab

leA

2.1

2:

Per

capit

aex

pen

dit

ure

an

din

com

eso

urc

esin

Ugan

da.

Nat

ion

alU

rban

Ru

ral

Sam

ple

tota

lto

tal

tota

lq 1

q 2q 3

q 4q 5

(a)

Nati

onal

Per

cap

ita

exp

end

itu

re26,5

14

53,6

70

20,7

80

3,6

75

9,0

85

13,9

64

21,2

19

54,1

02

Shar

eag

ricu

ltu

re13.1

93.8

815.1

45.0

010.4

414.1

620.0

925.6

9–

Coff

ee2.0

50.4

02.4

01.2

22.3

02.5

73.1

92.6

7Sh

are

hom

e-p

rod

uct

ion

49.0

847.3

549.4

541.4

954.1

954.7

451.8

345.1

6Sh

are

oth

er37.7

348.7

735.4

153.5

135.3

731.1

28.0

829.1

5

(b)

Male

-hea

ded

Per

cap

ita

exp

end

itu

re26,7

95

55,1

37

21,0

42

3,8

50

9,1

12

13,9

70

21,2

64

51,9

16

Shar

eag

ricu

ltu

re14.9

14.5

916.9

95.7

511.2

715.3

121.4

628.4

1–

Coff

ee2.2

90.4

92.6

51.2

02.4

02.8

33.4

43.1

1Sh

are

hom

e-p

rod

uct

ion

49.2

346.7

649.7

341.9

655.3

555.3

051.4

944.1

3Sh

are

oth

er35.8

648.6

533.2

852.2

933.3

829.3

927.0

527.4

6

(c)

Fem

ale

-hea

ded

Per

cap

ita

exp

end

itu

re25,7

49

50,1

18

20,0

49

3,3

44

9,0

08

13,9

43

21,0

71

61,5

58

Shar

eag

ricu

ltu

re8.5

32.2

010.0

13.5

78.3

310.7

915.2

315.9

9–

Coff

ee1.4

10.1

91.6

91.2

82.0

51.8

22.3

31.1

4Sh

are

hom

e-p

rod

uct

ion

48.6

748.7

448.6

640.5

951.2

853.1

353.0

548.8

0Sh

are

oth

er42.8

49.0

641.3

355.8

440.3

936.0

831.7

235.2

1


Tab

leA

2.1

2:

Con

tin

ued

.

Nat

ion

alU

rban

Ru

ral

Sam

ple

tota

lto

tal

tota

lq 1

q 2q 3

q 4q 5

(d)

Pro

du

cers

Per

cap

ita

exp

end

itu

re24,0

12

47,8

73

22,8

82

4,2

60

9,4

02

14,0

66

21,3

67

46,8

36

Shar

eag

ricu

ltu

re24.6

924.8

824.6

817.3

519

23.3

326.1

830.4

4–

Coff

ee8.1

86.9

18.2

38.6

37.5

97.9

68.3

38.6

6

(e)

Pro

du

cers

(male

s)

Per

cap

ita

exp

end

itu

re25,2

93

55,4

03

23,8

88

4,1

13

9,4

54

14,0

74

21,4

69

48,1

77

Shar

eag

ricu

ltu

re26.9

430.0

626.8

118.4

920.3

525.3

829.1

32.0

2–

Coff

ee8.6

48.1

18.6

68.7

7.5

48.4

39.0

99.0

7

(f)

Pro

du

cers

(fem

ale

s)

Per

cap

ita

exp

end

itu

re19,6

57

23,8

07

19,4

50

4,7

21

9,2

78

14,0

36

21,0

42

40,5

93

Shar

eag

ricu

ltu

re17.0

910.5

917.4

13.7

915.7

715.8

516.9

223.0

8–

Coff

ee6.6

43.6

16.7

88.4

47.7

6.2

55.9

46.7

7

Sou

rce:

bas

edon

hou

seh

old

surv

eys

(see

Tab

leA

2.3

).

Case Studies 55

Tab

leA

2.1

3:

Per

capit

aex

pen

dit

ure

an

din

com

eso

urc

esin

Zam

bia

.

Nat

ion

alU

rban

Ru

ral

Sam

ple

tota

lto

tal

tota

lq 1

q 2q 3

q 4q 5

(a)

Nati

onal

Per

cap

ita

exp

end

itu

re80,7

64

107,1

91

66,5

97

25,6

41

43,1

27

60,4

24

86,0

42

169,2

01

Shar

eag

ricu

ltu

re13.5

42.5

419.6

120.3

421.3

819.4

517.2

718.4

1–

Tob

acco

0.4

70

0.7

30.4

30.7

0.7

20.7

1.3

5–

Cott

on

1.9

20.0

22.9

71.9

73.2

53.9

92.5

53.2

7Sh

are

hom

e-p

rod

uct

ion

25.8

62.1

838.9

142.1

739.8

339.2

437.6

232.7

1Sh

are

oth

er60.6

95.2

841.4

837.4

938.7

941.3

145.1

148.8

8

(b)

Male

-hea

ded

Per

cap

ita

exp

end

itu

re81,5

15

108,9

48

66,4

75

25,7

14

43,1

22

60,4

30

85,8

17

166,5

74

Shar

eag

ricu

ltu

re13.9

42.6

920.2

321.4

721.5

920.0

917.4

319.4

7–

Tob

acco

0.5

20

0.8

10.5

40.5

90.8

0.8

61.6

1–

Cott

on

2.0

50.0

13.1

82.1

33.5

44.2

62.6

13.4

2Sh

are

hom

e-p

rod

uct

ion

25.0

81.7

438.1

242.0

739.4

638.2

236.2

931.2

3Sh

are

oth

er60.9

895.5

741.6

536.4

638.9

541.6

946.2

849.3

(c)

Fem

ale

-hea

ded

Per

cap

ita

exp

end

itu

re77,7

39

99,5

65

67,0

66

25,3

89

43,1

46

60,4

00

86,8

63

179,8

98

Shar

eag

ricu

ltu

re11.8

61.8

817.0

616.3

420.4

616.4

216.6

213.7

7–

Tob

acco

0.2

60.0

20.3

90

1.1

80.3

60.0

70.2

2–

Cott

on

1.4

0.0

32.1

11.4

12.0

12.7

12.3

12.6

2Sh

are

hom

e-p

rod

uct

ion

29.1

24.1

342.1

542.5

341.4

44.0

742.8

39.1

4Sh

are

oth

er59.0

293.9

940.7

941.1

338.1

439.5

140.5

847.0

9


Tab

leA

2.1

3:

Con

tin

ued

.

Nat

ion

alU

rban

Ru

ral

Sam

ple

tota

lto

tal

tota

lq 1

q 2q 3

q 4q 5

(d)

Pro

du

cers

Per

cap

ita

exp

end

itu

re68,0

64

81,2

91

65,8

10

24,6

95

43,3

05

60,4

17

87,4

80

153,4

42

Shar

eag

ricu

ltu

re36.2

920.7

936.3

834.9

837.3

536.1

933.2

639.7

5–

Tob

acco

5.6

83.7

15.6

95.0

65.4

4.6

15.5

8.4

8–

Cott

on

23.1

711.6

423.2

323.4

924.8

725.4

720.1

320.4

8

(e)

Pro

du

cers

(male

s)

Per

cap

ita

exp

end

itu

re68,1

31

87,2

24

64,8

85

24,6

54

43,4

21

60,7

69

87,5

68

150,6

09

Shar

eag

ricu

ltu

re36.5

712.6

336.7

135.4

335.9

636.4

134.5

541.5

1–

Tob

acco

5.8

90.6

5.9

25.8

14.0

64.7

87.0

99.6

2–

Cott

on

23.1

99.6

223.2

622.7

524.2

525.5

21.5

620.3

8

(f)

Pro

du

cers

(fem

ale

s)

Per

cap

ita

exp

end

itu

re67,8

07

58,8

48

69,3

46

24,8

47

42,6

83

58,9

87

87,2

57

164,3

77

Shar

eag

ricu

ltu

re34.6

64.8

434.4

31.9

350.3

334.5

729.1

529.4

3–

Tob

acco

4.3

620.5

44.2

60

17.8

93.4

0.4

51.7

9–

Cott

on

23.0

522.5

323.0

528.4

730.5

925.3

315.5

821.1

1

Sou

rce:

bas

edon

hou

seh

old

surv

eys

(see

Tab

leA

2.3

).

3

Institutional Arrangements

This chapter describes the main institutional arrangements in each of thevalue chains to be considered in the analysis. This description includes thedifferent vertical arrangements from different crops, whether the value chainsare characterized by one or more layers, the structure of competition amongfirms and farmers in each layer, and whether there are market interlinkages.For each crop we describe the main characteristics of its world market andthe specific institutional arrangement in the countries being studied. Moreimportantly for our simulations, for each case study we present a list of themain processing/exporting firms and their respective market share.

1 COTTON

The cotton plant is native to tropical countries but cotton production is notlimited to the tropics, as the emergence of new varieties and advances in cul-tivation techniques have led to the expansion of its culture. Whereas the plantis a perennial tree by nature, under extensive cultivation it is mostly grown asan annual shrub. The collected seed cotton goes through the ginning processthat separates the fiber from the cotton seeds. The major end uses for cottonfiber (85 percent of the commercial value of the seed cotton) include wearingapparel, home furnishings, and other industrial uses. Through the spinningprocess the cotton fiber is made into yarns and threads for use in the textileand apparel sectors (wearing apparel accounts for approximately 60 percentof cotton consumption). The cotton seeds provide edible oil and seeds that areused for livestock food (UNCTAD: see http://unctad.org/infocomm/anglais/cotton/chain.htm). Figure 3.1 depicts a stylized value chain for the cottonsector.

The production of cotton is crucially important to several developing coun-tries. In 2006–7, the four main producing countries were China, India, theUnited States, and Pakistan and accounted for approximately three quartersof world output. Although Africa is not the largest cotton exporter (it accountsfor 10–15 percent of world exports), cotton is of critical importance to manyAfrican countries. Cotton is the largest source of export receipts in severalWest and Central African countries. The cotton sector is also key to ruralpoverty reduction, with cotton-related activities accounting for a large share


Smallholders Estate

Ginneries

Cotton seed

Exports

Lint

Nontextile applications

Textileapplications

Oil industries Specialty paperproducers

Yarn spinners

Textile manufacturingAnimal feedproducers

Chemicalmanufacturing Apparel Household

furnishingRetailers

Consumer

Figure 3.1: Cotton stylized value chain.

of rural employment. Almost all exports from West African countries are rawcotton, which means that processing opportunities at the domestic level arenot fully exploited. The four main exporters of cotton lint are the UnitedStates, India, Uzbekistan, and Brazil while, the four largest importers of cot-ton lint are China, Turkey, Bangladesh, and Indonesia.

Global cotton consumption has increased by 2 percent per year since the1940s. This is this despite the fact that cotton share in textile fibers has beendeclining because of the increase in chemical textiles. Regardless of increasinglocal processing (especially in developing countries), cotton is still the maintraded agricultural raw material, with more than 30 percent of cotton pro-duction (approximately 6.3 million tonnes of fiber) traded per annum sincethe beginning of the 1980s. Cotton consumption has shifted to developingcountries, mainly as a reflection of rising wage levels in developed countries.In the textile sector, labor accounts for about one-sixth of production costs.This means that raising labor costs eroded the competitive edge of developedcountries, and contributed to the shifting of cotton processing to low-costeconomies (UNCTAD).

1.1 Cotton in Benin1

Cotton is the main cash crop and the largest source of export receipts forBenin. The average annual production of cotton grain is 350,000 tonnes. The

1This section is largely based on Gergely (2009), Saizonou (2008), and World Bank (2005).

Institutional Arrangements 59

sector generates 45 percent of fiscal revenue (excluding custom duties) and,on average, contributes 13 percent of the national GDP. This crop is of par-ticular importance for rural welfare since cotton-related activities generatemonetary revenue for approximately two million people in 350,000 cottonfarms.2 Cotton accounts approximately for 20 percent of the total cultivatedarea in the country. Two-thirds of the cotton production takes place in thenorth of the country (départements of Borgou and Alibori). At the industriallevel, cotton represents around 60 percent of the industrial tissue throughtwenty ginning companies, five textile plants, three crushing mills, and onecompany producing cotton wool. Gearing activities during a campaign (aroundsix months) create more than 3,500 jobs at the national level. The cotton valuechain generates important spillover effects in other sectors such as trans-portation, retail, and construction.

The origin of cotton production in Benin is similar to other French WestAfrican countries. Cotton emerged as a cash crop in the 1950s, under thedirection of the French parastatal Compagnie Française pour le Développe-ment des Fibres Textiles (CFDT). After independence, cotton production wasshifted to national monopolies, with CFDT retaining a minority share. In Benin,as was generally the case elsewhere, a monopoly marketing board, the SON-APRA, controlled all stages of the production process: distribution of seedsand inputs, provision of credit and other services to producers, ginning, andthe final exports. Once a year and before the growing season, the SONAPRAestablished a single price for seed cotton and inputs, with some adjustmentspossible based on its financial outcome. The cost of the inputs provided bythe organization was reimbursed directly via a deduction from the purchaseprice of cotton (World Bank 2005).

The process of liberalization of the Beninese cotton commodity chainstarted during the 1992–93 campaign. It implied a progressive disengagementof the state from the provisioning and distribution of inputs. Import and dis-tribution operations were taken over gradually by private national operators,whose numbers have increased over time. At the same time, producer organi-zations were given responsibilities so that they could participate fully in theprocess of the transfer of competencies in the domains of input supplies, sup-port to the supervision of farmers, and the commercialization of cotton grain.In the industrial sector, liberalization started in 1994, with the accreditationgranted by the state to private promoters of factories for the production ofcotton grain.

There are three major functions in the Beninese cotton chain: the produc-tion of cotton grain, the provision of inputs, and the production of cottonfiber (shelling). The actors involved in these three functions are structured

2This number comes from the 2002 agricultural census. However, due to the declineobserved in the last few years, some estimate that the number of farms growing cottonhas declined to 120,000 (Gergely 2009).


in organizations or “families”: the family of the producers represented byFUPRO–Benin (Fédération des Unions de Producteurs du Bénin), the family ofthe importers and input distributors represented by the Professional Groupof Agricultural Input Distributors, and the family of cotton grain producersrepresented by the Professional Association of Cotton Grain Shelling factoriesof Benin (APEB). The Cotton Inter-professional Association (AIC) coordinatesthe three families of organizations (Saizonou 2008).

Until 1999, producers’ prices were fixed by the government. From that yearonward, the responsibility was supposedly transferred to the AIC. The newprice mechanism established that seed cotton price was to be determinedthrough negotiations between cotton producers and ginners, with the AICacting as a facilitator. Usually, for the upcoming marketing year, a base priceis set in March–May. The final producer price is then fixed in October, whenthe harvest is about to begin using the cotton world market price as a refer-ence and deducting the customary processing and marketing cost. Accordingto the institutional design, CSPR–GIE is the technical unit of the AIC in chargeof the management of physical and financial flows. It receives support fromthe producer organizations in order to animate primary collection markets forcotton grain. The cotton grain shelling factories ensure transportation of theproducts but are only responsible for the quantities that they have been allo-cated. Since the 2006–7 season, the CSPR–GIE has had total control over thephysical flows, which permits it to ensure 100 percent payment of the actors(Saizonou 2008). Despite the liberalization process and the price mechanismdescribed below, in practice, the government remained a key player in deter-mining the price, since the price-setting mechanism is somewhat vague andthe stakeholders rarely reach a price agreement. Furthermore, the behaviorof producer prices of the last few years seems to show that the price-settingmechanism has not changed to a large extent, as local prices have remainedsticky despite considerable world price fluctuations.

Benin prohibits exports of cotton seeds. The seeds are ground locally andexported as cotton lint or oil. Each cotton company is allocated a quota pro-portional to its installed capacity, which contributes toward segmenting themarket and restricting entry and competition. Ginneries are required to payto the CSRP an advance of 40 percent prior to delivery of the seed cottonas a security (Gergely 2009). The country has an installed ginning capacityof twenty units with a total shelling capacity of up to 587,000 tonnes peryear, which significantly exceeds the average annual production of 350,000tonnes. Ten plants belong to SONAPRA, while private actors, either foreigncompanies (LBC/Aiglon, Louis Dreyfus, Kamsal, IBECO, MCI, and Sodicot) orthe local private sector (Talon and cooperatives) have invested in the privateplants (SONAPRA retained a 35 percent share in each of them). The effectiveallocation of grains often differs from the quota based on the installed capac-ity. In our simulations we will use the allocation of the 2007–8 campaign,where SONAPRA accounted for 55 percent of the market, followed by LCB


Table 3.1: Market shares in export supply chains: cotton.

Benin Burkina Faso Côte d’Ivoire︷︸︸︷︷︸︸︷︷︸︸︷Ranking Firm Share Firm Share Firm Share

1 SONAPRA 55.1 SOFITEX 5.0 Ivoire Coton 45.02 SOCOBE 6.8 SOCOMA 10.0 CIDT 29.03 Ibeco 6.2 Faso Coton 5.0 LCCI 16.04 LCB 10.0 DOPA 6.05 MCI 3.4 SICOSA 4.06 ICB 8.47 CCB 8.58 SODICOT 1.6

Malawi Zambia︷︸︸︷︷︸︸︷Ranking Firm Share Firm Share

1 Great Lakes 50.0 Dunavant 44.02 Clark 50.0 Cargill 32.0

Cotton3 Amaka 13.04 Mulungushi 6.05 Continental 5.06 Mukuba 1.0

with 10 percent, CCB with 8.5 percent, and ICB with 8.4 percent of the market(see Table 3.1 for details).

In theory the cotton industry covers the whole value chain (spinning, weav-ing, printing, and garment making) but the activity in the textile sector hasbeen shrinking in the last decade and now processes less than 2 percent ofthe lint production. The sector that produces for the domestic market and theNigerian market is represented by SOBETEX (a French private group), COTEB(a partnership between the government and European investors), and SITEX(a joint venture between the government and Chinese investors). The sectoris facing increased competition from imports and second-hand garments, inparticular because of the high cost of energy and the low productivity of labor(Gergely 2009). All companies are facing considerable financial difficulties andthe government is in the process of privatizing them. Despite the interest ofthe government in the sector, Benin has so far failed to demonstrate a com-parative advantage in textiles.

Recently, a new strategic plan for the revival of the agricultural sector (andthe cotton subsector in particular) has been adopted. The plan will promotethe development of the agricultural commodity chains within the frameworkof public/private partnerships. The plan retains the structure of the com-modity chains that has been in development since 2000 in the interprofes-sion associations. As part of this plan, SONAPRA has been partially priva-tized in September 2008. The Société Commune de Participation (SCP) was


provisionally declared the successful bidder. As a result, a public–private jointventure called the Société Pour le Développement du Coton (SODECO) willbe created which will retain 33.5 percent ownership. The remaining shareswill be held by the government of Benin (33.5 percent) and the private sector(33 percent).

1.2 Cotton in Burkina Faso3

Cotton is the main cash crop in Burkina Faso, generating income for approxi-mately two million people in the country. It is also the main source of foreignrevenue, accounting for 40 percent of total exports. The production is con-centrated in the western part of the country (Comoé, Kossi, Mouhoum, andKénédougou). While a minority of producers cultivate relatively large areas(up to twenty-five hectares), most of the cotton farms are family owned andsmall scale (typically from three to five hectares).

Cotton production in Burkina Faso is semi-privatized, and is often cited asa model of reform away from the old vertically integrated state-owned cottoncompanies (Hanson 2008). The process of privatization of the sector beganin 1998 when the government sold some of its shares to the producers’ orga-nization (UNPCB). The subsequent partial privatization of the cotton sectorin Burkina Faso created three regional cotton companies. SOFITEX, the coreof the former parastatal operates in the western part of the country, ownsthirteen gins, making up approximately 85 percent of the ginning capacity(Table 3.1). Faso Coton was formed in 2004, operates in the central region witha single gin located in Ouagadougou, and controls 5 percent of the market.SOCOMA, which operates three gins in the eastern region, is the second privatecompany created in 2004 and has a market share of 10 percent. After threeconsecutive years of significant losses from 2005 to 2007, the three cottoncompanies initiated a recapitalization process. DAGRIS—a French parastataland former shareholder in many African cotton and oilseed companies thatwas privatized in January of 2008 (becoming Geo Coton)—did not participatein the recapitalization of its shares in SOFITEX. The government of BurkinaFaso assumed those shares and subsequently created a financial institutioncalled “Fonds Burkinabe de Développement Economique” to broker the saleof the shares to the private sector.4

A distinguishing feature of the cotton sector is the degree of organizationof the producers at the local and regional level. Since 1996, producers have

3This section is largely based on Hanson (2008), WTO (2004a), and Yartey (2008).4The capital structure of the three companies in 2008 was as followed. SOFITEX was

owned by the state (65 percent), UNPCB (30 percent), and private banking (1 percent).SOCOMA’s larger shareholders were Geo Coton (34 percent) and UNPCB (20 percent). FASOCoton was owned by Reinhart AG (31 percent), IPS (29 percent), others (20 percent), andUNPCB (10 percent).


joined together in cotton producers’ groups (GPCs), which took over from vil-lage groups. The GPCs have a cooperative structure that facilitates the supplyof inputs and agricultural machinery, and seeks to ensure proper manage-ment of loans and an increase in crop yields. Recent estimates show around250,000 cotton producers, organized in 8,000 producers’ cooperatives thatform 170 departmental cooperative groups and seventeen regional unions.At the national level, the producers are represented by the National Union ofBurkina Cotton Producers (UNPCB). The state handed over part of the capitalof the SOFITEX in 1998 to the UNPCB to allow producers to take a leading rolein managing the subsector (WTO 2004a).

The ginning companies are in charge of the transportation from the pri-mary markets to the ginning plants. The producers’ associations are paid netof inputs purchased and the proceeds are then distributed among the mem-bers of the association. Most of the seed cotton is ginned and 98 percentof the lint is then exported mainly to Southeast Asia (66 percent) and Europe(20 percent). Burkina Faso also produces and exports seed oil, mainly throughSN-Citec.

Until recently, there was a guaranteed producers’ base price that was setbefore the crop year. The system included the possibility of bonus payments(in case of profit, the producers received a higher price the following sea-son5) similar to the system applied in Côte d’Ivoire and Benin. Since the2007–8 campaign the system has been changed in favor of a market-basedproducer price-setting mechanism. The new mechanism aligns domestic pro-ducer prices with world market prices and thus makes producers share partof the risk. However, to limit excessive price fluctuations for producers, whohave little access to credit, the producer floor price is smoothed by basingit on a five-year centered average of world market prices. A discount is thenapplied to the average to protect the smoothing fund, set up to finance majordeviations of actual prices from the producer floor price. When world pricesare low, the fund makes payments to ginning companies. When world pricesrise, the fund is replenished (Yartey 2008). The system is administrated by theInter-Professional Cotton Association and the fund is financed by assistancefrom the EU, the French Development Agency, and is sustained by paymentsmade by the cotton companies.

1.3 Cotton in Côte d’Ivoire6

Cotton is the third most important crop for the Ivorian economy but ranksfar behind cocoa and coffee in terms of export revenues generated. Despitethis, cotton contributes significantly to livelihood in rural areas. Most of the

5The return premium was divided: 50 percent to growers, 25 percent to the state, and25 percent to the ginning companies.

6This section is largely based on the UNCTAD online report (see references for details).


production takes place in the north of the country and is done by small-scalefarmers, who, on average, own four hectares and receive a low share of worldprices, averaging around 54 percent and reaching 63 percent in recent years.The area planted to cotton has grown steadily since 1960, with a leveling offaround 1989, a jump in area planted at the time of the 1994 devaluation, andsubsequent decline in recent years due to the civil conflict. Seed cotton yieldsalso grew over the 1960s and 1970s, but stagnated and varied erratically untilthe devaluation. As a result, production grew until 1987, and again after thedevaluation, with increased variability. Seed cotton is ginned in the countryand the cotton lint production has mirrored seed cotton production. Most lintis exported and some cotton seed has also been exported since 2000.

Until the late 1990s, a single vertically integrated state enterprise (“Compag-nie Ivoirienne de Développement des Textiles,” or CIDT) was responsible fororganizing virtually all services needed for cotton production and marketing,utilizing the institutional frameworks derived from French colonial heritage(UNCTAD). The privatization of this parastatal company was an objective ofinternational donors, but was resisted by the government. Due to agronomic7

and institutional differences, the cotton sector in Côte d’Ivoire was managedsomewhat differently from cocoa or coffee and the institutional arrangementand its evolution are somewhat different.

The privatization of CIDT began in 1998, when it was broken into regionalcompanies, but each of those held a monopoly over their region, and the statedid not divest a majority interest in those companies until 2002. This did notlead to competition, as the price of seed cotton remained the same for thethree zones; in addition, each company retained exclusive purchasing rightswithin its zone. Three new companies were set up. “CIDT nouvelle” is active inthe south of the country. The government has expressed its readiness to relin-quish its share (a proposed deal was to sell 80 percent of the state’s sharesto producers) but the negotiations on the purchase of CIDT nouvelle havebeen stalled due to the civil conflict affecting the country. The second com-pany is “Cotton-Ivoire,” an equity joint venture active in the northwest of thecountry. The Aga-Khan group and the Swiss-based cotton-trading firm “PaulReinhart” have joined venture interests in the company. The state retains a30 percent share in the venture. The third company is LCCI, a subsidiary ofthe Switzerland-based Aiglon group. LCCI is primarily active in the north-east. Each company is responsible for the purchasing of cotton throughoutits allotted area. This is often implemented by signing contracts, with grow-ers stipulating the area to be planted and the quantity of seed cotton to bedelivered.

The market remained geographically segmented until the introduction oftwo new companies, DOPA and SICOSA. The original three companies felt

7Cotton is more input demanding, requiring fertilizer, pesticides, and variety changesover time.


penalized because they had invested in inputs and extension services for pro-ducers. To resolve this situation, a new system was developed to guaranteeseed cotton supply to the cotton companies and to secure reimbursement oftheir investment in seed cotton inputs. Under this system, the companies signa contract with the producers through their cooperatives for the provision ofinputs and extension services. The extension services could be provided byany private company chosen by the cooperative. In return, via their cooper-atives, the producers are engaged to deliver their seed cotton to the cottoncompany. As of March 2006, the market shares for the ginning companieswere: Ivoire Coton 45 percent, CIDT 29 percent, LCCI 16 percent, DOPA 6 per-cent, and SICOSA 4 percent (Table 3.1).

1.4 Cotton in Malawi8

In recent years, cotton has become an important crop for a Malawian econ-omy looking to broaden an agricultural export base that is heavily relianton tobacco. The cotton sector has about 120,000 smallholder farmers, threeginning companies, and three main input providers. Cotton production hasincreased since the 2003–4 campaign after a long period of decline thatstarted in the late 1980s. This negative trend was the result of several fac-tors, including the structure of the industry, the dominance of the public sec-tor in the purchasing of cotton, decreased productivity, and declining worldprices. On top of that, the integrated cotton, textile, and garment value chainwith intra-sector linkages collapsed with the financial problems of the onlyremaining textile company (David Whitehead and Sons) in the 1990s.

The sector has recovered since the 2003–4 campaign, partly due to theestablishment of the Cotton Development Association (CDA). The CDA pro-vides treated seed and pesticides to cotton farmers under contract farmingarrangements. A further important change was the improved ginning out-turn from 33 percent to 38 percent, which improves the overall crop value.The improvements in the cotton price on the international market have alsocontributed to the recent favorable performance of the sector. Up until 2003–4, cotton yields averaged about 600 kg/ha, but since then, through a num-ber of emerging cotton development initiatives and the slight increase in theginners, average yield has improved to about 900 kg/ha and production hasconsiderably increased to about 50,000 tonnes during the 2007–8 campaignfrom only 14,700 tonnes five years before (Tchale and Keyser 2009).

The estimated production cost for un-ginned seed cotton for Malawi islower than other countries, except for Mozambique and Nigeria. This impliesthat Malawi has some competitive edge against its neighbors in the produc-tion of cotton, and consequently the exportation of lint. Arguably, part of this

8This section is largely based on RATES (2003) and Tchale and Keyser (2009).


competitive advantage is lost due to the government cotton price policy.9

Every year, the government of Malawi sets a minimum seed cotton price with2–3 percent deduction from gross sales for outgrower costs. This price isprobably higher than the one that would exist in a collusion of the two gin-ning companies. This higher price is compensated for by reducing the ginner’sinvestment in outgrower extension and other services, thereby threatening thesustainability of high quality and productivity in the cotton subsector. Theprice-setting policy has recently generated conflict between the governmentand the ginneries as the minimum set price established by the government in2009 was much higher than the price ginneries were willing to pay after theinternational price collapsed following the 2008 crisis.

The two main components of the cotton value chain in Malawi are the cottonproducers, typically small holders, and the three existing ginning companies.The major cotton growing areas are the Lower Shire Valley (50 percent of totalproduction), the Southern region upland areas around Balaka (30 percent), andthe Lakeshore area around Salima (20 percent). Until recently, virtually all thecotton was sold within Malawi to the two ginning companies, Great LakesCotton Company and Clark Cotton Malawi (both subsidiaries of internationalcompanies) that each have half the market (Table 3.1). The market structureis currently changing, as a new company has been established. The Malawiangovernment, in association with China, has created the Malawi Cotton Com-pany. The company is comprised of a cotton ginnery, a textile manufacturingplant and cooking oil extraction. The company will also process cotton seedcake.

Seed cotton is sold to the ginneries in three different ways: through traders,by producer organizations, and directly to ginners. Traders operate in remoteareas, providing transportation to central markets. They often pay cash, andthey often pay it in advance of the announcement of the price for the currentcampaign by the ginneries. In general they offer poor conditions for the farm-ers and the CDA has tried to discourage the sale to middlemen by openingseveral buying points. Sales made through the farmers’ association have beenincreasing over time. The purchases made by these associations are often lim-ited by the amount of available cash. In general, they offer better prices anddeliver other services such as training, organizing inputs, and transportation.Farmers located close to the four ginneries and to the ginners’ own buyingpoints can sell directly to the ginning companies, receiving a better price buthaving to organize and afford the cost of transportation.

Each of the original two ginning companies owns two plants. After the seedcotton is ginned, the ginneries are left with cotton lint and cotton seed. A

9The other reason for the very narrow competitive edge in the lint export market is dueto the low ginning output in Malawi compared with other countries. This is an area thatprovides the greatest scope in terms of improving the ginner’s profit, which could then becascaded to the net farmer profit through investment in required services for the producer(Tchale and Keyser 2009).


proportion of the cotton seed (often around 10 percent) is set aside for theginners to provide the farmers with seeds for the following year. Most of therest is exported to South Africa, undersupplying the two or three active seedcrushing companies that exist in Malawi. Historically a higher proportion ofcotton lint has been sold to the local textile company (David Whitehead andSons) but financial problems led to a drop in its output and local ginneriesstarted to export most of the cotton lint to South Africa and South Asiancountries. The garment industry in the country is small and does not use localtextile as all the fabric for the cut, make, and trim garment firms is imported.

1.5 Cotton in Zambia10

Cotton is one of the main cash crops and it is produced almost entirely bysmall-scale farmers in Zambia. In 2003, around 11 percent of the farmers grewcotton. The largest 20 percent of farmers accounted for half of the productionand sale of this crop. Cotton production is heavily concentrated in the Easternprovince, with over one-third of all households in that province producingthe crop and accounting for about a two-thirds share of national productionduring the 2003 harvest season. Central and Southern provinces follow, with16 percent of farmers growing the crop in the Central province and accountingfor 19 percent of national production, and 12 percent growing in the SouthernProvince and accounting for 13 percent of national production (Tschirley andKabwe 2007).

Until 1994, the sector was dominated by a state monopoly (LINTCO) thatwas responsible for every activity in the industry. The reform period began in1994 when LINTCO was broken up and its ginneries sold to Lonrho (later suc-ceeded by Dunavant) and Clark Cotton (Koyi 2005). Since then, the productionhas gone through four phases: a rapid expansion through 1998, with produc-tion increasing from less than 20,000 tonnes in 1995 to over 100,000 tonnes in1998; a rapid decline in 1999 and 2000, spurred in large measure by a seriouscredit default crisis; production in 2000 falling to less than 50,000 tonnes; asustained and rapid recovery from 2000 to 2006, and a sharp decline in 2007,driven by the kwacha appreciation crisis of the previous year (Tschirley andKabwe 2007).

In the first eight years following the privatization of LINTCO, Zambia’s cot-ton sector operated as a concentrated, market-based system with almost nogovernment involvement, even on a regulatory basis. Extra-market coordina-tion, whether across ginning firms or between ginners, organized farmers,and other stakeholders, was minimal. Since 2002 the Zambian governmenthas developed a more noticeable presence in the sector, and efforts at sector-wide coordination have increased production markedly (Tschirley and Kabwe2007). Starting in 2005, two developments increased the level of effort put into

10This section is largely based on Koyi (2005) and Tschirley and Kabwe (2007, 2009).


sector-wide coordination. First, the Zambia National Farmers’ Union (ZNFU)finalized the creation of the Cotton Association of Zambia (CAZ) to repre-sent farmer interests in the sector, providing the Ginners’ Association withan organized private sector body with whom to communicate on key issues.Second, efforts at revision of the Cotton Act became a focus of intense collabo-ration across stakeholders.11 The government has also launched an initiativeto complement existing private outgrower schemes: the Cotton OutgrowerCredit Fund (COCF).

There has been no government mandated price, nor any pricing guidanceof any kind from government, since liberalization in 1994. Dunavant has typ-ically acted as a price leader, announcing a minimum preplanting price tofarmers, which may be adjusted upwards at the start of the buying season.Cargill typically follows Dunavant’s pricing, while smaller ginners frequentlypay higher prices than Dunavant. New entrants in the market have led toincreased competition among private firms and prices have become a key toolfor attracting buyers. However, there remains a great deal of variability in thelevel of input credit support offered to smallholders by the various ginners.These differences may allow the companies offering less or no support to useprice to attract sellers who may have received input support from anothercompany. The appreciation of the Kwacha in 2006 led to a conflict over thedomestic price of cotton. The government openly tried to influence prices andthe farmers attempted for the first time in an organized way (through CAZ)to negotiate the prices paid by the ginners. The analysis of Tschirley andKabwe (2007) shows that while Zambian companies have paid nominal pricescomparable to those in Tanzania, where more companies compete for the cot-ton crop, a detailed cross-country analysis demonstrates that Zambia pays asubstantially lower share of its realized ex-ginnery price to farmers than inTanzania. The key issue is that Zambia enjoys a very high price premium onworld cotton markets and therefore ginning companies could arguably pay ahigher price than they have been paying.

The value chain in cotton includes the production of cotton seeds at thefarm level, the production of cotton lint, the production of cotton yarn, and,eventually, the production of textiles. In Zambia, most of the production of

11In March 2006 a sub-committee was formed consisting of six members to consult stake-holders: Cotton Ginners Association, Cotton Development Trust, Food Security ResearchProject, Cotton Association of Zambia, Ministry of Agriculture, and co-operatives. Somestakeholders felt that the act was not strong enough to regulate the industry and con-trol the production and marketing of seed cotton. The main concern for processors is theincrease in “side buying” due to the proliferation of companies buying seed cotton andwith over 200,000 farmers producing cotton, prosecution of defaulters would be extremelyexpensive. They asked for an increase in the penalty for purchasing seed cotton withoutproper licensing and for the implementation of a register of producers that have requestedfinancing. The report of this committee was presented to the parliament in February 2009and a new Cotton Act is under consideration.


cotton seeds is devoted to the exports of cotton lint and, to a much lesserextent, of cotton yarn. In the case of exports of cotton lint, the farmers pro-duce cotton, which is purchased by the ginneries to produce cotton lint.12

Cotton lint is then exported to world markets. World markets for cotton lintare best described as competitive. Farmers are atomized and cannot exertmonopoly power when selling the cotton seeds. Instead, it is assumed thatthe ginneries can act monopsonistically over farmers. The number of ginner-ies has increased in recent years with Dunavant remaining as the dominantcompany with 44 percent of the installed ginning capacity. Cargill is the sec-ond largest firm, controlling 32 percent of the market, followed by Amaka(13 percent), Mulungushi (6 percent), Continental (5 percent), and Mukubawith only 1 percent of the total installed capacity (Table 3.1).

Another destination of farm cotton is to produce cotton yarn. In this case,the production of the farmers is processed by the ginneries into cotton lintand then sold as cotton yarn. This may involve another layer down the valuechain. However, Zambia’s spinning industry appears to absorb a small anddeclining share of the country’s lint production. The last available data indi-cates that, in 2002, the country’s four operating spinning mills processedless than 10,000 tonnes of lint, or less than one quarter of lint production inthe country. Swarp (a spinner) estimated in 2002 that 90 percent of Swarp’slint needs are met by purchases from Dunavant and Clark (now Cargill); thebalance appears to come from smaller ginners. Mukuba Textiles and Mulun-gushi Textiles both have gins within their premises and purchase seed cottonfor processing. Starflex, Excel, Mulungushi, and Kafue all experienced seriousfinancial problems in the early 2000s, which led to temporary and sometimesprolonged shutdowns (RATES 2003). The other smaller spinners indicate thatthey periodically import to meet their lint needs when they are unable toreach agreement on price with local ginners. Despite the problems that thesevalue-added sectors have faced, their combined size is not trivial when com-pared with cotton lint: total exports of yarn, woven fabric, and apparel totaledUS$23.5m in 2002 (over US$21m from yarn), compared to US$30m in lintexports (Tschirley and Kabwe 2009).

2 COCOA

Cocoa is grown on trees and the cocoa fruits grow directly on the stems andbranches. In West Africa, where most of the cocoa is produced on small family

12Historically, independent cotton traders—individuals trading cotton who do not ownand are not employed by a ginning company—played a major role as the middlemenbetween farmers and ginneries. However, due to market conditions, they largely disap-peared after 2000.


Smallholders

Trader

Exporter

Roasting and grinding

Confectionery industry Tobacco, cosmetic, andpharmaceutical industry

Retailers

Food industry

Consumer

Cocoabeans

Cocoaliquor

Liquor,butter,

cake, andpowder

Finalproduct

Figure 3.2: Cocoa stylized value chain.

farms,13 they are collected most intensively in the harvest seasons of Decem-ber and June. The cocoa fruits are cut down by hand. Machines cannot be usedbecause it is not possible to harvest all beans at the same time. The seeds arefermented on the ground for around seven days and dried for approximatelythree weeks, before being packed in bags and exported. Figure 3.2 shows asimple representation of the cocoa value chain from the small holder to theconsumer.

West Africa is the primary producer of cocoa today. Côte d’Ivoire, Ghana,Nigeria, and Cameroon produce two-thirds and export three-quarters of totalworld cocoa production. Côte d’Ivoire and Ghana are the largest producers.The third largest producer is Indonesia and other big producers are Brazil,Malaysia, and Ecuador. Most exports are directed to Europe and the UnitedStates, which are the two largest processors and consumers of cocoa.

Over the last ten years the global cocoa production has increased at an aver-age annual growth rate of 2.7 percent, producing 3.7 million tonnes during2007–8. Consumption has shown similar patterns. However, during the sec-ond half of the 1980s, excessive cocoa production led to a divergence betweensupply and demand, caused by excess production of cocoa. Over the last thirtyyears there has been a relative decline in the price of cocoa because of anincreased in the supply of cocoa into world markets. This is due to entry ofnew producer countries and to more efficient processing methods. The pricedecline was partially reversed in 2001 due to changing stock-holding behaviorof the industries, social unrest in Côte d’Ivoire, and lower yields of cocoa.

13This is in contrast to cases like Brazil and Malaysia, where large commercial plantationsdominate.


2.1 Cocoa in Côte d’Ivoire14

Côte d’Ivoire is the largest cocoa producer, accounting for around 40 percentof total cocoa supply. The sector is the most important for the Ivorian econ-omy, contributing with 15 percent of GDP, 35 percent of the total exports, and20 percent of the government revenue in 2007. It employs 700,000 households(35 percent of the total). Despite its importance and resilience, the sector hasbeen greatly affected in recent years by the civil conflict in the country, lowyields, volatile international prices, and excessive taxation.15

The institutional development behind agricultural policy, and indeed allpolicy evolution, was conditioned by Côte d’Ivoire’s experience as a Frenchcolony (Abbott 2007). Following independence the “Caisse de stabilisation desprix des produits agricoles” (Caistab) was established to regulate farm-gateand export prices (both for cocoa and coffee), provide extension service andinputs, and to collect substantial taxes. The Caisse was not directly involvedwith the transportation of cocoa from the farm gate (controlled by privatetraders called traitants) and permitted “private” exporters to operate within asystem of quotas (Losch 2002). The Caisse was relatively successful in insulat-ing farmers from the ample variation of international prices observed betweenthe 1970s and the 1990s.

The reform process started in 1987 and, in the middle of the 1990s, thestate’s control was diminished in order to reduce marketing costs, raise pro-ducer prices, and encourage the creation of producers’ organizations. Thereforms increased production, but did not lead to sufficient changes for farm-ers, which brought about further liberalization reforms in 1999 when theCaistab was disbanded and the producer price was fully liberalized.

The Caistab was replaced by four agencies to manage and monitor thesector. The Autorité de Régulation du Café/Cacao (ARCC) is the regulatoryauthority in charge of defining and enforcing a regulatory framework ensur-ing competition at all levels of the value chain. The Fonds de Régulation et deControle (FRC) is a financial regulation fund managing the price stabilizationsystem through taxes of cocoa exports and forward selling. The Bourse duCafé/Cacao (BCC) is a marketing bourse managed by farmers and exporters,responsible for managing export operations. The Fonds de Développementet de Promotion des Producteurs de Café et Cacao (FDPCC) is a developmentfund established by producers, funded by voluntary levy, to finance demand-driven development programs. Through these agencies the government has

14This section is largely based on Abbott (2007), Losch (2002), and Wilcox and Abbott(2004).

15Ivorian cocoa farmers received the lowest farm-gate prices among a sample of cocoa-producing countries: 40 percent less than farmers in Ghana, 50 percent less than farmers inCameroon or Nigeria, and 60 percent less than a producer in Brazil or Indonesia. The majorfactor behind the low farm-gate prices is taxation. Trading margins, in-country transporta-tion, exporter costs, local processing, and maritime freight are other contributing factors.


tried to strengthen the position of farmers by providing information aboutprices and encouraging farmers to form cooperatives to gain more bargain-ing power.

These four structures have been in charge of the regulation of the cocoa(and coffee) sector since 2001. However, the system has suffered a numberof drawbacks due to external (decline in world market prices) and internalfactors (civil conflict, an excessive tax burden on cocoa farmers who provide$5 billion in fiscal levies and $1.4 billion on para-fiscal levies, and appar-ent corruption cases leading to the arrest of some of the officials in chargeof these agencies). Currently, the system configuration is in the process ofbeing revised. Two committees, formed in September 2008, are reviewing pastreforms and audits of the sector and revisiting the role of the sector’s fouragencies, with a view to formulating a new institutional and regulatory frame-work for the sector.

The elimination of the Caisse de Stabilisation had an impact on the cocoamarket configuration allowing some backward integration by the market’snew entrants: the multinational firms. The state of the sector and the civil con-flict does not allow us to gather recent statistics. However, according to theBureau d’Etudes Techniques et de Développement (BNETD), the market shareenjoyed by cooperatives has decreased from 32 percent during the 1998–99season to 18 percent in the post-liberalization season of 2000–2001, leavingalmost 80 percent to be funneled through middlemen. Once the cocoa arrivesat the port of Abidjan or San Pedro, it is conditioned for export (usinage) andshipped to processors mainly by multinational exporters who, in the cases ofArchers Daniels Midland (ADM), Cargill, and Barry Callebaut, are themselvesprocessors. Therefore the farm-gate price is now the residual of the “c.i.f.”price less transportation, conditioning, taxes, and other associated market-ing costs, which may include rents to exporters (Wilcox and Abbott 2004).Fourteen firms controlled three-quarters of the cocoa that was declared forexport in 1999–2000, the year the Caisse was dissolved. Three years later,these fourteen firms controlled more than 85 percent of the export market,with the top five firms among the sixty-one exporters controlling almost halfof the total exports.16 Table 3.2 provides detailed market shares for thesefourteen companies. The three largest are Cargill West Africa with 16.4 per-cent of the export market, ADM Cocoa Sifca with 11.9 percent, and Tropivalwith 8.3 percent.

16Wilcox and Abbott (2004) note that these reports only allow us to figure out the nom-inal ownership of cocoa exporters as some anecdotal information leads us to believe thatseveral smaller companies are acting on behalf of the larger exporters or that there isoverlapping ownership.


Table 3.2: Market shares in export supply chains: cocoa.

Côte d’Ivoire Ghana︷︸︸︷︷︸︸︷Ranking Firm Share Firm Share

1 ADM Cocoa Sifca 11.9 PBC 32.832 Armajaro 3.3 Akuapo 11.973 Barry Callebaut 3.1 Olam 10.714 Cargill West Africa 16.4 Adwumapa 8.625 Cemoi 3.8 Fed 7.046 Cipexi 6.0 Kuapa 5.917 Cocaf 4.4 Transroyal 5.728 Dafci and IFCO 4.5 Armajaro 5.79 Delbau 4.7 Cocoa Gh 3.17

10 Outspan Ivoire 5.2 Diaby 2.711 Proci 6.0 Others 5.6312 Sifca-Coop 4.913 Tropival 8.314 Zamacom 3.415 Others 14.1

2.2 Cocoa in Ghana17

Ghana provides one-fifth of the total world supply of cocoa beans. The countrywas the world’s leading producer of cocoa by 1911, a position it retained untilthe mid 1970s when it was overtaken by Côte d’Ivoire. The sector has beenof vital importance for the Ghanaian economy, not only for the 1.6 millionsmallholder farmers growing cocoa (production has always been small-farmbased, mostly on plots of three hectares or less, with plantations never havingbeen of much importance) but also for the government and other economicsectors associated with the activity. With the recent discovery of oil fields, itis expected that the sector will become less determinant for the governmentin terms of revenue source.

Until World War II, internal and external marketing were handled by privatefirms, but during the war the colonial government took over the purchase ofcocoa. In 1947 the Cocoa Marketing Board (CMB) was established (after 1979it was renamed The Ghana Cocoa Board or COCOBOD). It was omnipresent inthe cocoa industry and covered extension services, input marketing, and themaintenance and rehabilitation of roads in cocoa-producing villages (Brookset al. 2007). The CMB was the only authorized buyer and exporter of cocoa.The CMB carried out its activities through its subsidiaries the Produce BuyingCompany (PBC) and the Cocoa Marketing Company (CMC). The Quality ControlDivision (QCD) is responsible for ensuring that the overall quality of the beansis kept to the high standard for which Ghanaian cocoa is known worldwide.

17This section is largely based on Brooks et al. (2007), Laven (2007), Lundstedt and Pärssi-nen (2009), and Vigneri and Santos (2007).


While initially set up to protect farmers from price volatility, the CMB grad-ually turned into an instrument of public taxation. Rents were extracted bykeeping producer prices well below the world price, and by using an over-valued exchange rate to make payments to farmers. Between 1967 and 1977,the system for purchasing and marketing cocoa gradually broke down as theeconomic situation deteriorated. An extensive economic recovery programmewas implemented in the mid 1980s. Efforts to improve the efficiency of COCO-BOD led to wide-ranging changes to its structure and activities. Transport ofcocoa shifted to the private sector after 1984. From the 1988–89 campaign,COCOBOD began phasing out input subsidies, and this led to a substantialincrease in input prices. Staff levels were reduced from over 100,000 in theearly 1980s to just over 5,100 staff by 2003 (Brooks et al. 2007). The internalmarketing system was liberalized in 1993 allowing Licensed Buying Compa-nies (LBCs) to compete with the PBC. In addition, the PBC was partly privatizedin 2000 and introduced on the Ghanaian stock exchange. COCOBOD owns40 percent of the stocks directly and another 30 percent indirectly throughits ownership of a major stakeholder (Lundstedt and Pärssinen 2009).

Despite the reforms, the Ghanaian government still plays an important rolein the cocoa sector. Through COCOBOD, the government controls cocoa qual-ity, hands out licenses, finances, and controls activities of private companies.As we will describe below, it also sets producer prices and margins and sellsand exports to manufacturing and processing companies.

The farm-gate price for cocoa in Ghana is determined in a very unique waybecause of the country’s unique marketing arrangements. The ceiling price isdetermined by the international price of cocoa to which the government thennets out a variety of margins to pay for the many layers of its interventionin the sector. The Producer Price Review Committee (PPRC) is very much incharge of how the floor price paid to the farmer is ultimately determined,and this committee is made up of a variety of stockholders ranging from theMinistry of Finance, industry representatives, the COCOBOD, LBCs, farmersrepresentatives, and the University of Ghana. The producer price is a pricefloor, i.e., the LBCs are not allowed to purchase cocoa for less than the pro-ducer price, but, in practice, the price paid to the farmers is not raised abovethis minimum level. This producer price is set at the beginning of each cropyear and is constant throughout the seasons. In addition to setting the pro-ducer price, the PPRC sets a yearly fixed purchase price, i.e., the price thatthe LBCs receive from selling the cocoa to COCOBOD. This price correspondsto the buyer’s margin and is set taking into account average transport costs,commissions paid to purchasing clerks, and other costs faced by the LBC.Each LBC receives the same buyer’s margin.

The system contemplates the existence of a price stabilization mechanism.When there is a discrepancy between the actual and the predicted pricebecause of fluctuations in the world price of cocoa, this implies a surplusor deficit with respect to the target level set at the beginning of the campaign.


The surplus is divided between the government and the farmers (in the formof yearly bonuses after payment), while the deficit is covered by the govern-ment alone (Lundstedt and Pärssinen 2009).

Despite the fact that the number of registered LBCs has increased grad-ually since the liberalization reform,18 the number of companies that areactive players in the local market remains much smaller as fewer than tenof them purchase up to 90 percent of the total harvest. Table 3.2 portrays theranking of LBCs by market shares calculated as the five-year average between2004–5 and 2008–9. PBC has the largest market share with a market shareof around 32.8 percent. The second largest LBC is the domestically ownedcompany Akaufo Adamfo, which has an average market share of 12 percent.Olam, with its approximate market share of 11 percent, is the third largestLBC.

LBCs can be divided into four categories depending on the ownership struc-ture of the company. The first category comprises the former subsidiary ofCOCOBOD, the PBC. The second category of LBCs consists of domesticallyowned LBCs. The third type of companies is the farmer-based fair trade coop-erative Kuapa Kokoo that was established in 1993 by a group of farmers withsupport from a British NGO. The last category of LBCs comprises the two inter-national companies, Singaporean-owned Olam and British-owned Armajaro.19

The international companies have access to foreign capital, an advantage thatmakes them less dependent on the seed fund. When dividing the marketshares into its categories, Lundstedt and Pärssinen (2009) show that domesti-cally owned LBCs have increased their shares over the five-year period, whilethe shares of both Kuapa Kokoo and of Olam and Armajaro have decreased.The PBC strongly decreased its market shares in 2005–6 and 2006–7.

The liberalization process has also meant that the number of LBCs per vil-lage increased by around 30 percent between 2002 and 2004, which impliesthat the potential trading partners of cocoa farmers have increased signifi-cantly over the years (Lundstedt and Pärssinen 2009). This deregulation inthe domestic segment of the supply chain was expected to bring competi-tion among different private buyers and to generate a number of produc-tion incentives to the farmers. Most notably, one would have expected com-petition to emerge by means of price bonuses and/or premiums over theguaranteed price to characterize the new marketing arrangement. One wouldalso have expected this, in turn, to both stimulate farmers’ supply and toincrease traders’ own share of the domestic market. However, what makescocoa Ghana’s cocoa marketing system unique is the virtual absence of any

18Initially, six companies were granted licenses to operate on the internal market whiletoday there are twenty-six active LBCs, including the PBC.

19Both Olam and Armajaro are leading suppliers of cocoa and other commodities (suchas coffee and sugar) on the world market and operate in all main cocoa-producing coun-tries. In Ghana they operate as buying companies, but their expertise includes origination,exporting, and processing of cocoa (Lundstedt and Pärssinen 2009).


price-based competition mechanism. LBC competition for cocoa supplies—which is fierce, if anecdotal evidence is to be believed—is based on the provi-sion of different services (Vigneri and Santos 2007). Examples of this lack ofprice competition include allowing community representatives to select thepurchase clerks or choosing one that is capable, trustworthy, and motivatedto serve farmers’ needs. Incentive packages offered by LBCs may comprisecash payments, bonuses, gifts, rewards, subsidized inputs, credit and train-ing, and other investments looking to maintain durable social relations withtheir suppliers. According to a survey of farmers, the main reasons by farmersfor choosing a particular buyer are cash payments, social relations with thepurchasing clerk, provision of credit, and, in the case of the PBC, accountabil-ity (Laven 2007). For our purposes, it does not matter whether competitionbrings about an increase in prices or a decrease in costs.

The liberalization of the internal market has not led to liberalization of theexternal front. The PBC is the only company that is allowed to export. Origi-nally, the government imposed some minimum volume of purchase require-ments over three consecutive years for the LBCs to be able to export 30 per-cent of their purchases. During the first years of the reform, the LBCs did nothave enough profit margins and volumes of cocoa to cross the export bar.However, now that some of them seem ready to start exporting, the govern-ment has decided not to grant them the required license to export. The reasonadvanced by the government for maintaining the monopsony structure of thesector is to guarantee high quality and contract fulfilment, for which Ghana-ian cocoa receives a price premium on the world market. In contrast, mostLBCs report that they want to enter and would be capable of entering theexport sector and that COCOBOD deliberately hold them back from engagingin external marketing. International processing and manufacturing compa-nies do not oppose the system in Ghana, most likely because Ghana is theonly country in the world offering a consistent supply and relatively low priceof high quality cocoa (Lundstedt and Pärssinen 2009).

3 COFFEE

Coffee is grown in tropical and subtropical regions around the equator. Thetwo types of coffee plants widely cultivated are Robusta and Arabica. Ripecoffee cherries are harvested manually and undergo primary processing inthe producing country before they are exported. The primary processing iscarried out to separate the coffee bean from the skin and pulp of the cherry.There are two alternative methods for doing this: wet and dry. The end prod-ucts of both methods are coffee beans, referred to in the trade as “green”coffee. Wet processing produces “mild” coffee, usually of the Arabica type,and the dry method produces “hard” coffee, either Hard Arabica or HardRobusta. The distinction is important as Mild Arabica, Hard Arabica, and


Smallholders Estate

Cooperative

Curing plant

Trader

Exporter

Roaster

RetailCatering

Consumer

Dry cherry/parchment

Greencoffee

Roasted/instant coffee

Figure 3.3: Coffee stylized value chain.

Hard Robusta coffees are traded separately (Tropical Commodity Coalition(www.teacoffeecocoa.org)). Figure 3.3 shows the value chain for coffee.

Brazil, Vietnam, Colombia, Indonesia, and Ethiopia are the main producersand exporters of green coffee, with Brazil’s share close to one-third of thetotal market. Uganda, Côte d’Ivoire, and Kenya are also among the top twentyexporting nations. Most of the coffee produced is consumed in high-incomecountries. The United States, Germany, Italy, Japan, and France are the top fiveimporters. More than 80 percent of the production is traded internationallyas green coffee, generally packed in 60 kg bags. Green coffee is available tobuyers either directly or via the spot markets in the United States and Europe.International buyers are generally concerned with the uniformity and con-sistency of green coffee and they require information on the type of coffee,the type of primary processing, the country of origin, and the official gradestandard.

Globally, coffee for home consumption is mostly purchased in supermar-kets. The food retail sector is highly concentrated in the United States, the


United Kingdom, and Northern Europe and plays a dominant role in the foodmarketing chain. There is also an important and growing market for special-ties and product differentiation that is been exploited by smaller producers.

3.1 Coffee in Côte d’Ivoire20

Coffee plantations were first established alongside smallholder farms in Côted’Ivoire at the beginning of the 1920s. Historically, coffee was a very importantsource of income for small holders in Côte d’Ivoire and the second source offoreign exchange after cocoa for the government. During the 1960s and 1970sthe country was Africa’s largest coffee exporter. Some 300,000 coffee plantersmanaged 1,200,000 hectares of plantations. However, mismanagement of thesector and the economy and low relative international prices for coffee haveled to a decline of the sector. In the mid 2000s the sector was characterizedby a coffee grove getting old (65 percent of the plantations were more thantwenty-five years old21) and low output figures (around 150,000 tonnes peryear) and yields (250 and 350 kg/ha). The civil conflict has affected the pro-duction of coffee and cocoa less as they are mainly produced in the south ofthe country. However, the conflict has led to the resumption of export taxesand increased trader margins.

The evolution of the institutional setting for the coffee sector has been sim-ilar to that of cocoa. In 1964 the Caisse de Stabilisation des Prix des ProduitsAgricoles (CAISTAB) was created as a price stabilization and support fundfor the cocoa and coffee sector. The CAISTAB was in charge of the primarycollecting and the transportation and export of the crops. It also providedextension and inputs but the state intervened little in the production pro-cess itself. It paid the farmers, through private agents, a preannounced pricefor their crops and sold the output on international markets. The differencebetween these two prices, net of marketing cost, was a surplus that consti-tuted an important part of the government’s revenue. However, from the late1980s onward, international prices dropped below the producer prices andthe surpluses became deficits (Benjamin and Deaton 1993).

The CAISTAB shielded coffee farmers from much of the international pricevariations, with remarkably stable nominal, domestic coffee prices over aperiod of enormous change in international prices (Abbott 2007). However,by the beginning of the 1990s the stabilization lacked reserves, had accu-mulated important debts, and could not continue to guarantee producers’

20This section is largely based on Abbott (2007) and Benjamin and Deaton (1993).21In recent years, the situation has slightly improved, with a program for the regeneration

of the groves that has been set up for 170,000 ha. The diagnostic is that the intensificationof the output through the improvement of the coffee-tree plantations, the rejuvenation ofthe plantations, the maintenance of the groves, and the use of input location is necessaryto regain the interest of the coffee culture in Côte d’Ivoire.


prices.22 This crisis marked the beginning of a gradual liberalization process.International donors pushed for the privatization of the parastatal but theprocess was slow because it was resisted by the government. The CAISTABand the cocoa and coffee sectors were finally privatized in 2000. The BCC(Coffee and Cocoa Marketing Exchange) and ARCC (Coffee and Cocoa Reg-ulatory Authority) took over CAISTAB functions. The reform process contin-ued through the 2000s aiming at improving producer prices and productivity,marketing arrangements, and the monitoring of the sector by government andpublic and private agencies. However, the results of these reforms have beendisappointing as the producers’ price and the competitiveness of the sector asa whole has not improved. As was mentioned in the section regarding cocoa inCôte d’Ivoire, the sector has been undergoing a new set of reforms since late2008. Partially because of this flowing state of the system and also becauseof the civil conflict, it was not possible to find reliable information on coffeeexporters’ market shares. For that reason, in our simulations we decided touse the same market shares that we have for the case of cocoa in Côte d’Ivoire.

3.2 Coffee in Rwanda23

The production of coffee is well spread in Rwanda. Around 400,000 small-holders24 produce coffee on approximately 52,000 hectares of land. There areno large estates producing coffee and only Arabica varieties are grown (WTO2004b). The country’s altitude and rainfall generate excellent agro-ecologicalconditions to cultivate this crop.

Historically, coffee has been one of the main sources of foreign earnings forthe Rwandan economy and several of the state’s public finance crises havebeen associated with strong fluctuation in the international price of coffee.The civil conflict at the beginning of the 1990s greatly affected the produc-tion of coffee, and this has only started to recover in the last few years. Since2000, the volume of production has increased but always below the stipu-lated targets. An important development has been the improvement in thequality of the exported coffee. This has allowed Rwanda to take advantage ofmarket niches favoring the trading of premium blends. The reported quantityand quality of Rwanda’s coffee is affected by unofficial exports and importsof cherries to and from the surrounding region. The restrictions on ordinarycoffee sales during the first period of the coffee season mean that some pro-ducers prefer to sell illegally for higher prices in neighboring countries.

Coffee was introduced at the beginning of the twentieth century and was themain economic commodity during the colonial period. After independence,

22In the 1989–90 crop year, producer prices for cocoa paid by the CAISTAB were reducedby 50 percent, and by 40–50 percent for coffee (Trivedi and Akimaya 1992).

23This section is largely based on Habyalimana (2007), Loveridge et al. (2003), and WTO(2004b).

24Each of them has on average 165 coffee trees.


Table 3.3: Market shares in export supply chains: coffee.

Rwanda Uganda︷︸︸︷︷︸︸︷Ranking Firm Share Firm Share

1 Rwacof 30.4 Ugacof Ltd. 16.22 Rwandex 29.2 Kyagalanyi Coffee Ltd. 14.03 CBC 22.8 Kawacom (U) Ltd. 13.34 Agrocoffee 13.7 Ibero (U) Ltd. 9.35 SICAF 3.9 Job Coffee 8.36 Great Lakes 7.27 Lakeland Holdings Ltd. 7.28 Kampala Domestic Store 5.69 Savannah Commodities 4.9

10 Pan African Impex 4.711 Others 9.4

the Rwanda Coffee Authority, OCIR-Café, was created with the mission ofsupervising coffee-related activities in the country, from production to com-mercialization (OCIR: see www.rwandacafe.com). However, the coffee indus-try was liberalized in the mid 1990s and, consequently, since then the cof-fee board has not been engaged in coffee processing, marketing, or exports.However, OCIR-Café still distributes seedlings and insecticides and providescertification on quality standards. It is remunerated by growers at 3 percentof their export sales price. The organization also issues licenses to private cof-fee traders (WTO 2004b). The liberalization of coffee policies seems to haveincreased yields by taking the poorest fields out of production. The 1990ssaw a large reduction in the proportion of farmers cultivating coffee fields—nationally 55 percent of smallholders grew coffee in 1991 versus only 30 per-cent in 2002 (Loveridge et al. 2003).

In the marketing link of the value chain, middlemen collect coffee beansfrom door to door, bulk them, and deliver them to large buyers, who transportthem to Kigali for hulling and export. Almost every village counts a middle-man, but some middlemen extend their services over more than one village.Large buyers are generally located in Kigali (Habyalimana 2007). At present,secondary processing of coffee is handled mainly by five factory exporters.25

Their market shares in 2005 were: Rwacof 30.4 percent, Rwandex 29.2 percent,CBC 22.8 percent, Agrocoffee 13.7 percent, and SICAF 3.9 percent (Table 3.3).Until recently, the government owned 51 percent of Rwandex. After severalprivatization attempts, the company was divided in two and sold in June 2009to foreign investors. Due to significant private sector investments in coffeewashing stations the amount of fully washed coffee has increased from 1 per-cent to 20 percent of production from 2002 to 2007. However, many washing

25Primary processing is done by the producers themselves using traditional or semi-modern methods. Only a small quantity is processed at modern washing stations.


stations are not profitable because of high operating costs, weak manage-ment, and financial issues. Stakeholders across the industry point out the lackof suitable infrastructure investment as the main constraint for the develop-ment of the coffee sector.

The fix-price policy for producers was abandoned in 1997 and replacedby an indicative weekly price. This “floating” price is announced before theend of each week as a baseline for negotiations between coffee producersand buyers. The calculation of this price is based on a “moving scale” whichtakes into account the various elements related to coffee picking, processing,transportation, and export (WTO 2004b).

3.3 Coffee in Uganda26

Uganda produces two types of coffee: Arabica coffee, which comprises about70 percent of the world’s coffee production but only 10 percent of Uganda’scoffee production; and Robusta coffee, which comprises about 30 percent ofthe world’s production and 90 percent of Uganda’s production. Robusta isgrown in the central part of Uganda in the Lake Victoria crescent, and acrossthe west, southwest, and east of the country. Arabica beans are grown athigher altitude, in the areas of Mount Elgon along Uganda’s western bor-der with Kenya and in southwestern Uganda along the Rwenzori mountainrange. This widespread cultivation places Uganda among the top ten coffee-producing countries in the world. Approximately 500,000 smallholder fam-ilies are engaged directly in its production, with over seven million peopledepending on the crop for their livelihood (Masiga et al. 2007). The crop oncegenerated more than 95 percent of the export income but its importance hasdeclined over time as nontraditional exports have picked up and coffee nowrepresents only around 20 percent of total export earnings. Despite this, thesector potentially has an important poverty reduction role since it occupies amuch larger part of the population than other activities.

Until 1991, the roles of stakeholders in the coffee supply chain were clearlysegregated. The smallholders produced, harvested, and dried their coffee. Thedried cherry was then sold to either primary cooperative societies or privatestores. Primary societies sold their coffee to cooperative unions, while theprivate stores sold the beans either to huller operators who, after hulling, soldthe coffee to the Coffee Marketing Board (CMB). The CMB in turn reprocessedthe crop and exported it as green coffee. The prices paid at each level werepredetermined by the authorities and did not change with movements in theinternational coffee market (Masiga et al. 2007).

The Coffee Marketing Board monopoly was abolished in 1991 and the entitywas split into two entities: the Coffee Marketing Board Limited (CMBL), which

26This section is largely based on Cheyns et al. (2006), Masiga et al. (2007), and VargasHill (2010).


is responsible for export, but on a par with private exporters, and the UgandaCoffee Development Authority (UCDA). The end of the monopoly opened upthe possibility of cooperatives and private operators exporting coffee directlyand nearly all exporters became vertically integrated. The supply chain forexported coffee was dominated by coffee processing and trading compa-nies. Private traders and the old cooperative trading system gradually lostground to private exporters. In recent years, multinational coffee companieshave became also important in Uganda, generating an alternative channel forexporting coffee (Cheyns et al. 2006).

Since the liberalization of the internal coffee market in 1992, farmers havebeen free to decide how and to whom to sell their coffee. For the majority offarmers the price is negotiated at the time of sale and payment is not madeuntil then. Transactions at the farm level are quite small and farmers usuallysell as individuals, only in a few cases selling as a group. Most coffee salesare made at the farm gate to small traders.27 These small-scale traders act asaggregators either for bigger independent traders, who often own a store ormill, or for exporters and their agents (Vargas Hill 2010).

After the coffee has been milled, it is transported to Kampala and sold toexporters. Coffee exporters in Uganda have to be registered with the Ugan-dan Coffee Development Authority (UCDA). The number of registered cof-fee subsector players at post-harvest levels was 324 in 2007–8 comprising30 exporters, 19 export graders, 271 primary processors, and 4 roasters. Atthe export level, over 90 percent of the volume was handled by ten compa-nies, with the largest being Ugacof Ltd. (16.2 percent), Kyagalanyi Coffee Ltd.(14 percent), Kawacom (U) Ltd. (13.3 percent), Ibero (U) Ltd. (9.3 percent), andJob Coffee (8.3 percent; see Table 3.3 for more details). Almost 75 percent ofUgandan coffee is exported to the European Union. More than 50 percent ofUganda coffee was bought by five companies, namely Sucafina (14.3 percent),Decotrade (11.3 percent), Drucafe (9.6 percent), Bernard Rothfos (8.5 percent),and Olam International Ltd. (7.7 percent). This level of concentration in Ugan-dan coffee exports is a reflection of the concentration in the coffee worldmarkets for both traders and roasters.

4 TOBACCO

Tobacco is usually cultivated annually. Harvesting is the first step in thetobacco value chain. This activity is generally labor intensive and involvesremoving the leaves from the plant stalks. Once harvested, the leaves arecured to remove all of the natural sap. Tobacco is stored, aged, blended, con-ditioned, cased, cut, dried, and flavored. After the curing and storage period,

27The majority of Ugandan producers sell their coffee in the form of dry cherries locallyknown as kiboko, which are then milled (the cherry is separated from the husk) by thetraders who buy the coffee.


the tobacco is graded according to color, texture, aroma, and size and soldin auctions or directly through a contract system. The next two steps in thevalue chain (often done in importing countries) are cigarette manufacturingand packaging. The cigarettes are manufactured by machine and put throughquality control checks, wrapping, and printing. They are then inserted intopacks and wrapped to preserve the quality. The last steps are the marketingand distribution. Figure 3.4 represents the value chain of tobacco.

The annual production of tobacco leaf is around seven million tonnes, ofwhich almost 40 percent is produced in China, followed by Brazil and Indiaboth producing around 10 percent of the global production. A significantamount of tobacco is consumed domestically and the rest exported. Brazil,the United States, and Malawi are the top three exporting nations, while Zim-babwe, Mozambique, and Zambia are among the top twenty. Developed coun-tries are still the main importers of tobacco leaf as they are the ones producingand exporting cigarettes and other tobacco products.

4.1 Tobacco in Malawi28

Tobacco is the single most important export crop for Malawi, contributingover 65 percent of foreign earnings. While the exact percentage may changefrom year to year, tobacco typically accounts for 43 percent of the agricul-tural GDP, 13 percent of overall GDP, and 23 percent of the country’s total taxrevenue. Out of a total workforce of about five million people, FAO (2003) esti-mates that around one million people (20 percent of the total labor force) areinvolved in the tobacco industry to some degree, either as producers, as labor-ers on estates or in processing factories, or as buyers or transporters.29 Thenumber of hectares dedicated to the crop generally varies between 120,000and 145,000. During the 2009 marketing campaign, 82 percent of the tobaccosold was burley, while flue-cured tobacco accounted for 15 percent of thetotal proceedings. Malawi is the first tobacco exporter and the second largesttobacco producer in sub-Saharan Africa after Zimbabwe. These two countriesaccount for 75 percent of the total production in sub-Saharan Africa. Theoutstanding position in tobacco export of Malawi is only partially explainedby agro-ecological conditions as they are not outstanding and are not verydifferent from other countries in sub-Saharan Africa. Instead, the two mostimportant factors to explain this phenomenon are the fact that the varieties ofburley tobacco grown in Malawi are relatively low in nicotine and path depen-dency as it was the crop of choice by both of the early European settlers andthe new political elites after independence in 1964 (Poulton et al. 2007).

28This section is largely based on FAO (2003), Jaffe (2003), Poulton et al. (2007), Tchaleand Keyser (2010), van Donge (2002), and World Bank (2008).

29Although some of this engagement is part-time and/or casual, the integrated house-hold survey (IHS) of Malawi records income from tobacco sales as the main source ofhousehold (cash) income in the major growing districts in the country (Poulton et al. 2007).


Smallholders Estate

Tobacco curing

Grading and buying

Primary processing:tobacco is stored, aged, blended, dried, and flavored

Export oftobacco leaf

Cigarettemanufacturing

Packaging

Retail

Consumer

Figure 3.4: Tobacco stylized value chain.

Before 1989, the Tobacco Control Commission closely controlled produc-tion activity. All tobacco producers had to obtain a license from the govern-ment regulatory body. The system was biased against smallholders as onlyestates and landowners were eligible to apply for a production license. More-over, to be allowed to sell tobacco directly on the auction floor, a growerhad to reach a certain production scale. This was the case until early 1995when Malawi embarked on a structural adjustment program that, among otherthings, allowed smallholder farmers to produce cash crops. These measurescontributed significantly to the rapid expansion in tobacco production. Theminimum quantity requirement to sell output in the auction market was over-come with the introduction of “intermediate buyers” for tobacco allowingsmall farmers to produce tobacco (FAO 2003).

The value chain of tobacco in Malawi is relatively simple as most of theexported tobacco is unmanufactured. The intermediate buyers functioned asthe middlemen between small-scale tobacco growers and the auction market,buying tobacco leaf from many small-scale growers at a negotiated price andthem selling them on the auction floor at the market price (FAO 2003). Tobacco


Table 3.4: Market shares in export supply chains: tobacco.

Malawi Zambia︷︸︸︷︷︸︸︷Ranking Firm Share Firm Share

1 AOI 34.0 Zambia Tobacco Leaf 47.612 Universal (LIMBE) 29.0 Alliance One Tobacco 16.343 Africa Leaf 16.0 Tombwe Tobacco 25.214 Premium 13.0 Associated Tobacco 10.845 Malawi Leaf 5.06 ATC 2.07 Wallace 1.0

leaf is generally sold in auction markets30 owned by Auction Holding Lim-ited (four floors: Limbe, Lilongwe, Chinkhoma, and Mzuzu), in which the gov-ernment, through the Agricultural Development and Marketing Corporation(ADMARC), has majority of the shares. Currently, there is a demand amongprivate sectors players that are interested in providing alternative tobaccoauction services. The government has agreed to open the auction system butit has yet to prepare the draft bills for legislation.

The tobacco buyers in Malawi have been described as an “oligopsony” whereeach of the few buyers exerts a disproportionate influence on the market.Increasing competition is one of the key elements in the agenda for the Malaw-ian tobacco sector (van Donge 2002; World Bank 2008). The tobacco buyersare represented by the Tobacco Exporters Association of Malawi (TEAM).31

In 2008 there were only six companies registered to buy tobacco: Limbe LeafTobacco Company Limited, Alliance One Tobacco (AOI), Africa Leaf, PremiumTama, Malawi Leaf, ATC, and RWJC Wallace. Their market shares for burleytobacco, reported in Table 3.4, were: AOI 34 percent, Universal (Limbe) 29 per-cent, Africa Leaf 16 percent, Premium 13 percent, Malawi Leaf 5 percent, ATC2 percent, Wallace 1 percent. The market shares for flue-cured tobacco were:AOI 33 percent, Universal (Limbe) 45 percent, Africa Leaf 14 percent, Premium3 percent, Malawi Leaf 2 percent, ATC 3 percent.

Despite the good conditions for growing burley tobacco in Malawi, the com-petitive edge is quite narrow. This is mainly due to high costs along the valuechain. The key burley tobacco production costs at the farm-level include fer-tilizer, seed, chemicals, and labor. At the assembly level, the major costs that

30There are suggestions to introduce more rural satellite auction markets which willinvariably reduce the congestion at the main auction markets and reduce the participationof intermediate traders.

31This association was formally established in 1930 and it is involved in negotiations anddialogue with all stakeholders in the tobacco industry on behalf of the buyers. Originally,membership in TEAM was restricted to tobacco buyers and exporters but has since beenextended to processing organizations and manufacturers (World Bank 2008).


reduce producers’ net profit include the high cost of transport and intermedi-ation, including the auction floors. In addition to these costs, there are otherhidden costs such as storage charges at the rural depots and the costs relatedto the long waiting time to offload the tobacco at the floors. The costs relatedto levies and taxes were already substantially reduced following the tobaccosector reforms undertaken in 2005 (Jaffe 2003; World Bank 2008). The onlycost element in which significant cost reduction may be possible is in thetobacco transport system. According to sector experts, the tobacco trans-port system is the most costly mainly due to inefficiencies in the tobaccomarketing system. As was previously mentioned, the government is studyingto open satellite auction floors and the deregulation of the auctioning sys-tem to diminish these transportation costs. The authorities also have allowedtobacco contract farming and marketing arrangements that, in principle, by-passed the auctioning system, but this policy was later suspended (Tchale andKeyser 2010).

4.2 Tobacco in Zambia32

Tobacco is, with cotton, the main agriculture exports of Zambia and onethe main sources of income diversification away from copper. Historically,tobacco production was carried out in large-scale commercial farms in thehands of expatriates. After gaining independence in 1964, the participationof smallholders increased. Around 40 percent of the tobacco production auc-tioned is contributed by small-scale farmers. In the 2003–4 season it is esti-mated that about 2,000 small-scale farmers participated in the growing oftobacco on a contract basis, averaging about 0.5–1 ha per farmer (Likulunga2005). The main areas of tobacco production are Mukonchi, in the Kabwe dis-trict of the Central Province; Kalomo, in the Southern Province; Kaoma, in theWestern Province; and several areas in the Eastern Province.

There is not an auction system like in Malawi. Tobacco is bought at buyingstations established by merchants at strategic points in the growing districts.Most of this tobacco is then exported to Malawi, where it is processed andsold to cigarette manufactures that sell in world markets (Balat and Porto2008). A few companies dominate the market. In 2006 the Lusaka floor mar-ket shares, reported in Table 3.4, were: Zambia Leaf Tobacco 47.61 percent,Alliance One Tobacco 16.34 percent, Tombwe Tobacco 25.21 percent, andAssociated Tobacco, 10.84 percent. Given the liberalized nature of the sec-tor, these companies are also able to contract small-scale farmers directly,offering them inputs on credit as well as extension services. The contracts aresigned directly between the small-scale farmers and the tobacco company.

32This section is largely based on Balat and Porto (2007) and Likulunga (2005).


The other key institutional players of the Zambian tobacco sector are theTobacco Board of Zambia and the Tobacco Association of Zambia. The Boardwas created by the Tobacco Act of 1967 with the objective of promoting,controlling, and regulating the production, marketing, and export of tobacco.The Tobacco Association of Zambia is the growers’ association, which ownsand manages the auction floors.

4

Exporters and Farmers: A Model ofSupply Chains in Agriculture

In this chapter, we study a theoretical model of supply chains in export agri-culture. The purpose of the model is to provide an analytical framework forstudying how changes in the structure of the supply chain affect farm-gateprices. These farm-gate price changes will feed into the poverty analysis ofChapter 5.

We present a game-theory model of supply chains in export agriculture.1

There are two main actors in the model: firms and farmers. There is a largenumber of farmers who must choose to produce home-consumption goodsor exportable goods. They are atomistic and face exogenous farm-gate pricesoffered by the firms. These prices and the characteristics of the farmer(land endowment, productivity) determine the allocation of resources of eachfarmer to the “export market” or to home-production activities.

Farm-gate prices are set by firms. The firms buy raw inputs from the farmers(coffee beans, cotton seeds) and sell them in international markets. We assumethat these firms are small in international markets and thus take internationalcommodity prices (for coffee or cotton), as given. In contrast, the firms enjoymonopsony power internally. There are only a few firms in each market, andthey compete in an oligopsony to secure the raw input provided by the farm-ers. The oligopsony game delivers the equilibrium farm-gate prices that thefirms offer to the farmers. Given these prices, farmers allocate resources opti-mally and supply raw inputs to the firms and this supply affects the quantitythat firms can supply in the export market. In equilibrium, firms take intoaccount the supply response of the farmer when choosing optimal farm-gateprices.

The solution to the game depends on various parameters. On the firm side,the equilibrium depends on both the number of firms and on their share ofthe market. In other words, it matters whether the market is characterized

1Our model builds on the ideas and the analytical framework developed by, among manyothers, Salop (1979), Barnum and Squire (1979), Singh et al. (1986), De Janvry et al. (1991),Benjamin (2001), Horn and Levinsohn (2001), McMillan et al. (2003), Taylor and Adelman(2003), Syverson (2004), Sheldon (2006), Sexton et al. (2007), Kranton and Swamy (2008),Ennis (2009), Cadot et al. (2010), and Ludmer (2010).


by symmetrical firms or, instead, by a large dominant firm and many smallcompetitors. Firm characteristics, such as production costs, also matter. Onthe farmer side, the equilibrium depends on factor endowments, preferences,and farm productivity (costs) in export agriculture. These factors determinethe export supply response of the farmers and how this is affected by thestructure of the market. Our model incorporates all of these features.

Once the equilibrium of the model is found, and the solution is calibratedto match key features of the economy, we study comparative static results.The main purpose of these exercises is to compute the changes in farm-gateprices that we need for the poverty analysis. We explore a variety of com-parative static results. Given the initial structure of the market (that is, thenumber of firms and their market shares), we simulate various changes incompetition. Our simulations cover a large number of general settings, fromentry to exit. We study the impacts of the entrance of a small competitor, of ahypothetical merge of the two leading firms, and of the split of the leader intosmaller competitors. In all these cases, given the initial equilibrium, we findthe new equilibrium of the model and study the changes in farm-gate prices,profits, and farmer utility (for different farmers). In the simulations, we takeinto account both firm and farm responses. This means that our comparativestatic results allow firms to adjust prices and quantities separately (implyingthat market shares may change in equilibrium). Farmers also adjust crop sup-ply, and this is, in turn, taken into account by the firms when choosing thenew equilibrium prices. In this sense, while the model is a partial equilibriummodel of the agricultural export markets, it incorporates responses from allagents and their feedback in determining the equilibrium.

We also study a number of additional simulations. Concretely, we focuson the impact of complementary policies that affect firms, complementarypolicies that affect farmers, and changes in international prices. These simu-lations work in the same fashion. For instance, we shock the production costsfaced by firms, the production costs of the farmers, and the internationalcommodity prices, and solve the model for the new equilibrium, allowing forall market interactions.

An important feature of our simulations is that we can study complemen-tarities between domestic policies, international prices, and competition poli-cies. To do this, we explore comparative static results where we change variousparameters both separately and simultaneously and we compare the differ-ent equilibria. These exercises allow us to learn, for instance, whether a givenchange in competition policies (such as entry) can be amplified by comple-mentary factors such as infrastructure.

Finally, we study outgrower contracts. Many markets in Africa are char-acterized by distortions and missing markets and this impedes the optimalallocation of resources. This is critical in export agriculture. If credit is neededupfront to undertake the necessary investments in export cropping (purchaseof seeds, fertilizers, pesticides), then a malfunctioning credit market may

Exporters and Farmers 91

push farmers out of the export market, even in the case of relatively highfarm-gate prices.

To study these issues, we extend our model by including outgrower arrange-ments. In these arrangements, firms cover upfront a fraction of the farmer’scrop production costs. Farmers repay these costs at the harvest time afterpaying an interest rate on the loan. The key feature of the extended modelis that the interest rate charged by the firms may depend on the structure ofthe market, that is, on the number of firms and also on their market shares.This is because the legal system is imperfect and thus we assume that firmscannot perfectly monitor farmers. In consequence, farmers may default onthe loan, and even side-sell to other exporters.

Increases in competition thus have two opposing effects, one effect viahigher farm-gate prices, which encourages export participation and reducespoverty, and another via a potentially higher interest rate, which hindersexport participation and increases poverty. This analysis introduces newinteresting dimensions to the discussion of competition policies and povertyin sub-Saharan Africa.

The chapter is organized as follows. In Section 1, we introduce the basicstructure of the standard model without outgrower contracts. In Section 2,we explain how we set up the simulations and in Section 3, we discuss themain results. In Sections 4 and 5, we extend the standard model to includeoutgrower contracts and we present the simulation results.

1 THE ECONOMY

We study an economy where individuals are endowed with (small) pieces ofproductive land. These agents must choose between being “peasants,” wholive in autarky and consume all their home production, or “farmers” whogrow and sell exportable goods and buy consumption goods from the mar-ket. The main assumption driving our results is simply that market-acquiredconsumption goods are a superior good. In other words, a diversified con-sumption portfolio becomes desirable as the person’s wealth increases.

In terms of behavior, this means that poorer individuals will home-consume100 percent of their endowment, while richer families (in terms of initialendowment) will trade a fraction of their endowment in the market, inexchange for other goods. That is to say, the superior-goods assumption gen-erates a wealth effect driving the peasant/farmer occupation decision in thiseconomy.

The structure of the market for the tradeable good will naturally have astrong impact on the equilibrium prices of the endowments. In particular,perfect competition among buyers of the farmers’ produce will deliver higherequilibrium prices than monopsony or oligopsony situations. In turn, higherprices for the farmers’ produce means higher wealth for individuals. And


through the wealth effect, this means that the more competitive the mar-ket structure is, the more individuals will leave autarky and become farmers.In consequence, as competition increases, the richest peasants will becomefarmers. Autarky behavior will move down along the distribution of income.

1.1 The Model

More formally, this is a one-period endowment economy populated by a mea-sure I of farmers and a finite number n of exporters. Farmers are identicalin preferences but are heterogeneous in the size/productivity of their farms.Specifically, each farmer i is endowed with a farm that can produce ei unitsof crop. ei takes values on an interval [e

¯, e] and its distribution over values

is represented by the continuously differentiable probability function F(e),density f(e).

1.2 Farmers

Individual farmers are identical in their preferences, but are heterogeneousin the size/productivity of their farms. Their Cobb–Douglas utility functiondefined on home consumption h and market goods m is given by

u(h,m) = hα(m+ c)1−α.The constant c is a preference parameter and implies that m = 0 can be

a rational choice—marginal utility of m will be finite even for m = 0. Thelevel c will effectively play the role of imposing a “subsistence” level e of theendowment that must be consumed by farmers. Poor farmers whose initialendowment is lower than the subsistence level will live in autarky. Rich farm-ers, instead, whose endowment ei is larger than the “subsistence” level e willsell part of the “surplus” ei − e to the market (and self-consume the rest). Wewill also show that the cutoff “subsistence” level is decreasing in p. The intu-ition is a wealth effect (or equivalently, in this simple one-period endowmenteconomy, income effect). When p is higher, farmers are richer, and thereforecan afford to diversify their consumption goods.

Each farmer i ∈ I is endowed with a farm with productivity ei. The farmeroperates the farm and its output can be either consumed by the farmer orsold to exporters in the market. The optimization problem is

V(ei;p, r) = maxh,m

hα(m+ c)1−α

subject to m+ ph � (p − λr)ei, h � 0, m � 0,

where ei is individual i’s initial endowment, p is the price for farmers of thecrop, r > 0 is the interest rate. Preferences are parameterized by 0 < α < 1.We now discuss the different pieces of the optimization problem.

We begin with the budget constraint. The farmer produces ei units of crop,of which he will apply h units to his own consumption. The remaining units


will be sold to the exporters at a market price ofp. In addition, we allow for thepossibility of a liquidity constraint affecting the home-market decision. Theliquidity constraint is parameterized by λ. When λ = 0, there is no liquidityconstraint. When λ > 0, the interpretation is that a farmer planning to produceoutput of (ei − h) for the market will need to borrow an amount λ · (ei −h) beforehand. The farmer will then need to pay an interest rate r on theborrowed amount. This possibility of liquidity constraints is introduced tostudy outgrower contracts later on (Sections 4 and 5). For the remainder of thesection, however, we will leave outgrower contracts aside, thereby assumingthat λ = 0.

1.3 Exporters

There are n exporters who sell the crop at an international price of P . Theybuy from farmers at the internal market price of p. These are Cournot oligop-sonists. They choose how much quantity to demand from the market at theprevailing price p, and they understand and correctly anticipate that theirown demand behavior affects p.

The problem faced by an exporter is then to maximize revenues:

Π(P,p, cpj ) = maxqj(P − p − cpj ) · qj,

where qj and cpj are, respectively, the demanded quantity and the unit cost ofproduction of exporter j. In principle, exporters may face different marginalcosts and this determines the equilibrium market shares.

1.4 Markets and Equilibrium

Individuals sell their output to exporters. Each exporter chooses its individualdemand from the farmers. The price P at which exporters sell their output inthe international markets is exogenously given in this model. The domesticprice p earned by farmers is determined in equilibrium, given farmers’ aggre-gate supply and exporters’ aggregate demand. We next define an equilibriumfor this economy.

Definition 4.1. An equilibrium in this economy is a collection of individualdecisions and prices

{(hi)i∈I , (mi)i∈I , {qj}nj=1, p, r}such that:

1. for each farmer i, (hi,mi) maximizes utility given price p and interestrate r ,

2. for each firm j, (qj) is a best-response to the other firms’ decisions(qk)k≠j , to farmers’ aggregate behavior and


3. the goods market clears:

∫i∈I(ei − hi)di =

n∑j=1

qj.

Condition 1 is the standard requirement that a farmer’s behavior be utilitymaximizing given the structure of the problem. Farmers take prices as givenand act accordingly.

The optimality condition for exporters introduces oligopolistic competi-tion. Firms chose demanded quantities in anticipation of their own effect onfarmers’ aggregate behavior, in a context of strategic interaction with otherfirms. Equilibrium in the economy requires that firms’ decisions be a Cournot–Nash equilibrium of the game between firms, given the farmers’ aggregatesupply function.

Condition 3 is a standard market clearing condition requiring that outputsold by farmers to the market coincides with the aggregate demand fromexporters.

1.5 Farmer Solution

We begin with the solution to the problem of the farmers. With λ = 0, theLagrangian and first-order conditions are

L = hα(m+ c)1−α +u1(pei − ph−m)+u2h+u3m,

∂L∂h

: αhα(m+ c)1−α

h−u1p +u2 = 0,

∂L∂m

: (1−α)hα(m+ c)1−αm+ c −u1 +u3 = 0.

The complementary slackness conditions are

u1(pei − ph−m) = 0, u2h = 0, u3m = 0,u1 � 0, u2 � 0, u3 � 0.

It is simple to show that any solution must have h > 0, since marginalutility of h converges to ∞ when h → 0. Therefore, u2 = 0 will always hold.Moreover, we look for a solution with m > 0. This implies µ3 = 0, and thefirst-order conditions become

∂L∂h

: αhα(m+ c)1−α

h−u1p = 0,

∂L∂m

: (1−α)hα(m+ c)1−αm+ c −u1 = 0.

From this we can solve

αhα(m+ c)1−α

h= p(1−α)h

α(m+ c)1−αm+ c ⇒ m = 1−α

αph− c.


Using the budget constraint, we get

pei = ph+m ⇒ pei = ph+ 1−αα

ph− c

⇒ h = αp(pe+ c)

⇒ m = (1−α)pei − cαWe get the standard response functions with Cobb–Douglas utility, for whichit is optimal to assign constant shares of the budget to each consumptiongood.

Note that this can only be a solution provided m � 0. Therefore, we cansolve for the cutoff level of parameters:

m = (1−α)pei − cα � 0 � ei � α1−α

cp.

Define the cutoff levele(p) ≡ α

1−αcp.

For any ei � e(p), the optimal responses are

m = 0, h = ei.For any ei � e(p), the optimal responses are the usual Cobb–Douglas budgetallocation rules:

ph = α(pei + c), m+ c = (1−α)(pei + c).In this sense, we interpret e(p) as a “subsistence” endowment level. Poorfarmers whose ei is lower than this “subsistence” level live in autarky and self-consume 100 percent of their endowment. Note that the cutoff “subsistence”level is decreasing in p. The intuition is an income effect. At higher p, farmersare richer, and therefore can afford to diversify their consumption goods.

The individual farmer’s market supply function is

s(p; e) = max{e− h,0} = max{e−α

(e+ c

p

),0}.

With some algebra, this can be rewritten as

s(p; e) = (1−α)max{e− e(p),0}.The interpretation of this equation is that each farmer supplies a percentage1−α of the “subsistence surplus” e− e(p).

The indirect utility function is

e � e(p) ⇒ V(e;p) = eαc1−α,

e � e(p) ⇒ V(e;p) = αα(1−α)1−αpα

(pe+ c).


This function is strictly increasing. To the left of e(p), it is strictly concave. Ithas a convex kink at e = e(p), and is linear to the right of e(p). To see this,consider the derivative:

e < e(p)

⇒ ∂V(e;p)∂e

= αc1−α

e1−α > 0 and∂2V(e;p)∂e2

= −α(1−α)c1−α

e2−α < 0,

e > e(p)

⇒ ∂V(e;p)∂e

= αα(1−α)1−αp1−α > 0 and∂2V(e;p)∂e2

= 0.

It takes simple algebra to see that the kink at e(p) is convex. The left-handfirst derivative for e(p) is smaller than the right-hand first derivative:

e > e(p) ≡ α1−α

cp

⇒ (1−α)p > αce

⇒ αα(1−α)1−αp1−α > ααα1−α c1−α

e1−α = αc1−α

e1−α .

The shape of the supply function will be relevant for the exporters’ decision.Note that, over the range where s(p; e) > 0, the individual supply function isstrictly increasing and strictly concave:

∂s(p; e)∂p

= cp2α > 0,

∂2s(p; e)∂p2

= −2cp3α < 0.

However, the function has a kink at the level which implies e = e(p). It isglobally weakly increasing but not concave.

We can now easily derive the aggregate supply of export crops that firmswill face. Integrating across individuals, we get

S(p) ≡∫ ee¯

s(p; e)f(e)de

= (1−α)∫ ee¯

max{e− e(p),0}f(e)de

= (1−α)∫ ee(ei − e(p))f (e)de.

Thus, the aggregate supply function is

S(p)1−α =

∫ eeeif (e)de− e(p)(1− F(e(p))).

What is the shape of the aggregate supply function? To avoid carryingaround the term (1−α), which is strictly positive, we look at the shape of

S(p) ≡ S(p)1−α.


First, to see that the aggregate supply function is nondecreasing in p, notethat:

dS(p)dp

= ∂S(p)∂e(p)

∂e(p)∂p

.

The second term is decreasing in p as mentioned before:

∂e(p)∂p

= − α1−α

cp2

= − e(p)p

< 0.

The first term is

∂S(p)∂e(p)

= −e(p)f(e(p))− (1− F(e(p)))+ e(p)f(e(p))

= −(1− F(e(p))) � 0.

This establishes that the aggregate supply function is nondecreasing:

S′(p)1−α = S′(p)

= ∂S(p)∂e(p)

∂e(p)∂p

= −(1− F(e(p)))(− e(p)

p

)

= (1− F(e(p)))(e(p)p

)

� 0.

What about the second derivative?

dS′(p)dp

= cp2

α1−α∂

∂(1− F(e(p)))∂e(p)

∂e(p)∂p

+ (1− F(e(p)))∂(e(p)/p)∂p

= f(e(p))(cp2

α1−α

)2

− (1− F(e(p)))2 cp3

α1−α.

This reduces to

dS′(p)dp

= f(e(p))(e(p)p

)2

− (1− F(e(p)))2 e(p)p2

.

The sign of this derivative is not unambiguous. We can look for conditionsunder which it will be negative:

S′′(p) < 0 � f(e(p))(e(p)p

)2

− (1− F(e(p)))2 e(p)p2

< 0

�f(e(p))e(p)1− F(e(p)) < 2.

Note that there is not a straightforward intuition for this term. This conditionreads as follows: take any p. Then the aggregate supply function will be locally


concave at p if, evaluated at the “subsistence level” e(p) corresponding tosuch p, the distribution function F(·) satisfies

f(e(p))1− F(e(p)) <

2e(p)

.

This condition basically puts a bound on “mass points.” In other words, forS(p) to be locally concave at a given p, the distribution function must sat-isfy the condition that probability does not “grow too fast” (i.e., too high anf(e(p)) relative to the “remaining” probability 1− F(e(p))).

While this condition will hold in the simulations that we run below, it isuseful to discuss the shape of the supply function. This is because this dis-cussion illustrates some of the major issues that drive the economic decisionof the farmers and their participation in export markets. To this end, let’s doa thought experiment in which we start off with a very low price (say, p = 0)for the traded good, and we increase it gradually to see how the economyresponds.

When the market price for the exportable good is very low, all farmers arepoor and they all self-consume their endowment. As the price of the tradeablegood increases, all farmers experience a positive wealth effect. This effectentices them to diversify their consumption portfolio by selling part of theirendowment to the market to buy other goods. Mathematically, as the priceincreases, the “subsistence level” of endowment starts decreasing. However,for sufficiently low prices this “subsistence level” is still higher than each andevery farmer’s endowment (including the rich).

The wealth effect is larger for “richer” farmers, which means that as theendowment value increases, these people are the first to experience the cut-off moment in which the market value of their endowment surpasses their“subsistence level.” Therefore, there is a first, low price p that acts as a trig-ger for rich farmers to start selling part of their endowment to the market inexchange for market consumption goods. After rich farmers have entered themarket, there is a region of prices where rich farmers are selling their goods,but poor farmers are still below their “subsistence levels,” and therefore areoperating in autarky and self-consumption. In such regions where only therich are trading, the slope of the aggregate supply curve is just the slope ofthe rich farmers’ individual supply curves—the individual supply curves ofpoor farmers still has zero slope. Hence, in this region the aggregate supplycurve will be locally concave. Eventually the price raises enough to bring poorfarmers to the market as well. This happens when the value of their endow-ments grows enough to cross the “subsistence level.” At the precise pointwhere the poor farmers enter the market, there will be a convex kink in theaggregate supply curve. The reason is that, at that point, the slope of the poorfarmers’ supply curve switches from zero to strictly positive.

Economically, there is a form of “increasing return” to price increases atthe point where poor farmers, who were previously operating in autonomy,


enter the market and start providing a positive supply of tradeable goods. Inconsequence, one intuition for the shape of the aggregate supply curve (andof the drivers of this shape) is that, in societies with very high inequality ofincome, this type of price regions with “increasing returns” will be present.Tipping points will exist which, when surpassed, suddenly large numbers ofpreviously autarkic farmers enter the market with supply.

1.6 Exporter Solution

We look for a equilibrium for the exporters’ oligopsony game. Exporters cor-rectly understand and anticipate that the market price p depends on theirown actions, other exporters’ actions, and aggregate supply behavior fromfarmers. Let Q ≡ ∑n

j=1 qj denote aggregate demand from exporters, then agiven exporter perceives the following problem:

Π(qk�=j, P , cpj ) = max

qj(P − p − cpj ) · qj such that Q ≡ qj +

∑k�=j

qk.

The state variables are the international price P , and other exporters’ actionsqk�=j . It can be shown that a sufficient condition for the problem to be con-cave is that the aggregate supply function S(p) be concave as well, so thatS′′(p) < 0. As discussed before, this is not guaranteed by concavity of theindividual supply functions s(e;p). In other words, when the aggregate sup-ply function is concave, the exporters’ profit-maximization problem will beconcave in their choice variable. If the aggregate supply function is not con-cave, then the problem may not be concave as well.

If the problem is concave, then the first-order condition ∂π/∂qj = 0 will ofcourse be necessary and sufficient. Moreover, by the Maximum Theorem underconvexity (Stokey and Lucas 1989; Sundaram 1996), the function qj(Q) is welldefined and continuous.

We now turn to the first-order conditions. With n exporters, we have

qj = (1−α)(P − ps(Q)− cpj )(1− F(e(ps(Q))))e(ps(Q))

ps(Q)

⇒ Q = (1−α)(nP −nps(Q)−

n∑j=1

cpj

)(1− F(e(ps(Q))))e(ps(Q))

ps(Q).

2 THE SIMULATIONS

The equations that characterize the equilibrium are a set of the best responsesof the firms and, given the aggregate supply of the farmers, market clearance(total farm supply of raw inputs equal to total firm demand of raw inputs). Thesolution to this problem has to be found numerically and we used Matlab

routines to do this.


The first step in the analysis is the calibration of the parameters of thefarmer model. Note that we need to perform a different calibration for eachof the country–crop case studies. We calibrate α, the parameter of the utilityfunction, the farm supply parameters, and the subsistence cutoff. To do this,we assume that the distribution of endowments follows a lognormal distri-bution with mean µ and standard deviation σ . Then, we use the householdsurvey data (see Chapter 2) and choose the parameters in order to match (asclosely as possible) the observed aggregate shares of income derived from theproduction of the export crop. The calibrated parameters are in Table A4.1.2

As for the solution of the model, we work with aggregate farm supply,

S(p) = (1−α)∫ eeeif (e)de− e(p)(1− F(e(p))),

and the first-order conditions of the oligopsony game. Firms incur differ-ent costs of manufacturing, cpj . In equilibrium, given P , market shares differacross firms. In consequence, we search for farm-gate prices and a structureof firms’ costs so as to match the number of firms and the distribution of mar-ket shares observed in the data. These market shares were described in detailin Chapter 3. To summarize, the search procedure comprises the followingsteps:

1. Given Q, we find an equilibrium price ps(Q) from the aggregate supplyequation.

2. We know the market shares for each firm, shj , so qj = shj ·Q is identi-fied.

3. Finally, we solve for cpj using the best response function of each firm.

This algorithm delivers the solution to the model, given the calibratedparameters. The solution comprises exogenous parameters, the firm costs,and an endogenous quantity, farm-gate prices. Now, given the calibratedparameters and the structure of costs associated with the solution, we cansimulate comparative static results numerically.

We carry out thirty-five simulations for each case study (country–crop pair,as in Chapter 2). The main component of our simulation is a hypotheti-cal change in the structure of the supply chain. To cover different types ofchanges in market structure, we explore the following seven cases:

1. Leader split (with equal marginal costs).

2. Small entrant (with marginal costs equal to smaller firm).

2Note that we calibrate a different set of parameters for each case study. This means thatwe use different parameters for different crops, even in a given country (such as cottonand tobacco in Zambia, for instance). We do this for consistency with the fact that ourmodel is designed to describe one market in isolation. This assumption makes sense if,for instance, different crops are produced by different farmers (because of geography). Amodel with multiple choice of export crops could be an interesting extension of our work.


3. Leaders merge and small entrant (with costs equal to that of the mostefficient merger and that of the smaller incumbent, respectively).

4. Leaders merge (with cost of the most efficient merger).

5. Exit of largest firm.

6. Equal market shares (all firms have the cost of the leader).

7. Perfect competition (ps(Q) equal to P less marginal cost of the mostefficient firm).

Moreover, for each of these “competition” simulations we also allow forchanges in complementary policies. We consider three complementary fac-tors:

1. Complementary policies that affect farmers: 2 percent increase in µ.

2. Complementary policies that affect firms: 2 percent reduction of costscpj .

3. Complementary policies that affect farmers and firms simultaneously:2 percent increase in µ and 2 percent cost reduction.

Finally, we also study simulations where we interact the changes in compe-tition with a 10 percent increase in international prices.

These simulations are performed with the following algorithm. In each ofthe simulations, there is a change in the number of firms,n, and/or a change inthe baseline structure of costs, cpj . To find the solution, we need to findQ andps(Q) that solve the first-order condition for each firm subject to aggregatefarm supply. That is, we solve


ps(Q)∀j ∈ n

subject to

Q = S(p),n∑j=1

qj = (1−α)∫ eeeif (e)de− e(p)(1− F(e(p))).

As a result, we calculate a new qj for firm j ∈ n (now the number of firmscan be different from that in the baseline case, depending on the simulationperformed).

The simulations are designed to compute different changes in farm-gateprices in different scenarios. These price changes feed into the poverty analy-sis of Chapter 5. However, the theoretical model can be used to preliminaryassess changes in farmer’s utility and in industry profits. To explore changesin utility, note that the individual indirect utility function is

e � e(p) ⇒ VNP(e;p) = eαc1−α,

e � e(p) ⇒ VP(e;p) = αα(1−α)1−αpα

(pe+ c).


So, average utility is

V =∫ ee¯

V(e,p)f(e)de

=∫ ee¯

VNP(e;p)f(e)de+∫ eeVP(e;p)f(e)de

=∫ ee¯

eαc1−αf(e)de+∫ ee

αα(1−α)1−αpα

(pe+ c)f(e)de.

Individual profits areπj = (P − ps(Q)− cpj )qj.

Total profits aren∑j=1

πj,

thus average profits are equal to

n∑j=1

πjn.

3 SIMULATION RESULTS

We investigate twelve case studies in the book and, as previously explained,we run a total of thirty-five simulations for each case study.3 Rather than dis-cussing all the possible simulations, we focus on the case of cotton in Zambiaand provide detailed explanations of the results for all the endogenous vari-ables of the model. Then, we summarize the key findings from the remainingcase studies, emphasizing both differences and similarities. We chose cottonin Zambia as our leading case because the cotton sector has undergone severalof the transformations that our simulations aim to capture. Until 1994, thesector was controlled by a state monopoly. Immediately after the privatiza-tion, the sector was dominated by a duopoly, but over the years competitionensued. The Zambian cotton sector has also seen several outgrower schemeswith variable degrees of success (see Sections 4 and 5).

3.1 Zambia Cotton

We begin with the baseline model (without any complementary changes) andwe discuss results from the seven market structure simulations. The resultsare reported in Table A4.2 (in the appendix). The leader split simulation

3We have thirty-five additional simulations in the model with outgrower contracts. SeeSections 4 and 5.


reveals an increase of both farm-gate prices and quantities because it raisescompetition among exporters. In the case of Zambia cotton, the increases infarm-gate prices and quantities are 8.92 and 5.32 percent, respectively. Fig-ure A4.1 shows that prices and quantities increase in all twelve case studies.The largest price increase is observed in the case of cotton in Burkina Fasoand this is because the leader absorbs around 85 percent of the market inthe initial situation. In other cases, such as coffee in Uganda, the split of theleader does not boost competition by much and thus the changes in pricesare small.

The impact of higher prices on welfare is positive, both for producers andfor nonproducers (some of whom become producers). In Zambia, the averageutility gain of cotton producers is equal to 1.56 percent. At the same time, theincrease in prices makes cotton production more profitable for the marginalfarmer and this triggers a supply response. This supply response is, however,small. In the end, the increase in utility for the average nonproducer is only0.06 percent. However, those farmers that do switch enjoy very large gains. Inthe Zambian cotton case, the average gain of the switchers is 26.64 percent.The switchers, nevertheless, are a small group, and consequently the averageimpact on nonproducers is negligible. The unconditional change in total utilityis a weighted average of changes in utility for producers and nonproducers.Given the low participation rate in cotton, this increase is equivalent to only0.11 percent of pre-shock average utility.

In principle, the change in profits in the leader split simulation is ambigu-ous. There are three different discernable patterns (Figure A4.2). In the firstcase, total and average industry profits increase in the same proportion. Thisoccurs when the leader is very efficient compared with its competitors, andthe split into two efficient firms significantly increases total profits and aver-age profits. In the case of Zambian cotton, for instance, average and totalprofits increase by 10.29 percent. In the second case, total industry profitsincrease but average industry profits decrease. This occurs when the leaderis efficient enough to increase total profits as it splits, but not sufficientlyefficient to maintain average profits unaffected. For instance, in the case ofcoffee in Uganda, total profits rise by 3.94 percent but average profits declineby 2.17 percent (Table A4.2(j)). Finally, there are cases where both total andaverage profits decline. This happens when the marginal cost of the leader issimilar to the marginal costs of the competitors. In consequence, while thesplit of the leader does not enhance efficiency, the increase in competitionbrings average and total profits down. For instance, in the Rwandan coffeecase, average and total profits declined by 3.94 percent (Table A4.2(i)).

The cases of Burkina–cotton, Malawi–tobacco, and Benin–cotton are intrigu-ing (Tables A4.2(c), A4.2(k), and A4.2(b), respectively). In these cases, thereare large differences between the market shares of the largest firm and thesmaller competitors and this implies big differences in marginal costs. As aresult, when the largest firm splits, some of the smaller and less efficient firms


cannot compete and they must exit the market. This reduces the number offirms and, in the end, average profits can show a large increase. Note that thisis a compositional effect.

Another market simulation that enhances competition, although on a verydifferent scale, is the small entrant scenario. This means that the impacts onfarm-gate prices, quantities, and average utility (of producers, nonproducers,and switchers) are qualitatively similar as in the leader split simulations, butmuch smaller in magnitude. In the case of Zambian cotton, for instance, pricesand quantities increase by 1.01 and 0.62 percent, respectively; total averageutility increases by only 0.1 percent, and the average utility of the producersincreases by a meager 0.17 percent. These are, by and large, negligible impacts.Note, however, that while the utility of the switchers increases significantly,by over 26 percent, there are only very few switchers.

A small entrant causes profits to decline in all simulations (unlike in theleader split simulation). Recall that changes in market structure bring about acompetition effect and an efficiency effect. A small entrant increases compe-tition and this reduces total and average profits. But, since the small entrantis assumed to have the cost structure of the least efficient firm, the efficiencyeffect disappears. As a result, in all of the case studies, average and total prof-its are lower when a small firm enters. For example, in the Zambian cottoncase study, total industry profits fall by 3.86 percent, and average industryprofits fall by 19.88 percent.

We now discuss the results of the leaders merge simulation, which is theanticompetitive counterpart of the leader split model. Note that, in prac-tical terms, this simulation is equivalent to the elimination of the secondlargest producer. Since competition among firms is now lower, farm-gateprices decline and, therefore, the producers’ average utility also declines.The switchers are, in this scenario, farmers that were producing the exportcrop and retrench into subsistence agriculture as prices decline. The utility ofthe switchers decline significantly. Note, however, that nonproducers’ utilityremains unchanged because these farmers did not participate in the supplychain at the original prices and thus their decisions are unchanged by thelower price of the export crop. Total average utility is just a weighted averageof these changes in utility. In the Zambia–cotton example, farm-gate prices fallby 8.59 percent, the average utility of cotton producers falls by 3.28 percent,the average utility of the switchers falls by 20.24 percent, and total averageutility is reduced by only 0.10 percent.

Due to the coexistence of the competition and efficiency effects, profits caneither increase or decrease. Since competition is actually less intense whenthe leaders merge, the competition effect tends to increase profits. However,the “elimination” of the second largest producer can entail relative efficiencygains or losses. If, for instance, the second firm is relatively efficient (withmarginal costs that are close to those of the leader), then its elimination by themerger can decrease aggregate efficiency (when a lot of its output is diverted


to smaller firms). This pushes industry profits down. In contrast, if the secondlargest producer is relatively inefficient (compared with the leader), then theresulting output reallocation may entail efficiency gains and higher industryprofits. It is therefore unsurprising to observe that profits increase in casessuch as Burkina Faso–cotton or Benin–cotton, where the leader is significantlymore efficient than its merged partner. In contrast, in cases such as Malawi–tobacco, Rwanda–coffee, or Côte d’Ivoire–coffee (Table A4.2(h)), both mergersare relatively similar in efficiency and thus profits tends to decline.

We now turn to the leaders merge and small entrant simulation, which is, infact, a combination of the two previous cases. There are, nevertheless, someinteresting results to highlight. As we explained before, the leaders mergesimulation eliminates the second largest firm from the market, and the smallentrant simulation just duplicates the smaller and less efficient firm. There-fore, there are no efficiency gains because, in practice, in this exercise weare replacing the second most efficient firm with the least efficient one. Thismeans that the efficiency effect is negative. Additionally, the extent of compe-tition is necessarily lower because the anticompetitive effect of the merger (inpractice, the elimination of the second largest firm) more than compensatesfor the procompetitive effect of a small entrant. It follows that the impacton prices, quantities, and profits are negative.4 For instance, in the Zambiancotton case, prices fall by 6.22 percent and total average utility decline by0.07 percent. However, the utility of cotton producers drops by 2.38 percentand the utility of the switchers drops by 20.24 percent. Since prices are goingdown, the utility of nonproducers is not affected.

This simulation delivers interesting results when we look at profits. In threecases, average and total profits increase. This is because the incumbent firmthat merges with the leader is actually similar (in terms of costs and mar-ket shares) as the third largest firm (which now becomes the second firmin terms of market shares). This implies a relatively small efficiency effectso that the impact of the decline in competition prevails. In the other eightcases, average and total profits decline. In these cases, the competition effect(which increases profits) is not large enough to compensate for the efficiencylosses caused by the merge. In the Zambian cotton case, both total and aver-age profits fall by 7.71 percent. Instead, in Benin–cotton, both increase by2.79 percent.

In the next simulation, the exit of the largest firm, we study the effectsthat would take place if the leader leaves the market. Thus, the most efficientfirm, with the smallest marginal cost, disappears and the market is coveredby the remaining (more inefficient) firms. Farm-gate prices and quantities fallbecause the total demand of farm output shrinks—in the Zambian cottoncase, they fall by 11.84 and 7.58 percent, respectively. Total average utility,the average utility of the producers, and the average utility of the switcher

4Note that the effect of this simulation is not the sum of leader split and small entrant.


decline (by 0.14, 4.51, and 20.25 percent, respectively, in the Zambian cottoncase). The utility of the nonproducers remains unchanged.

Surprisingly, there is heterogeneity in the response of profits. In principle,profits should decline because the most efficient firm leaves the market. Infact, this is the case in five case studies. For example, in Zambia–cotton, totalprofits decline by 22.37 percent and average profits decline by 2.96 percent.However, in two cases, total profits fall but average profits increase, prob-ably because the effect of lower competition is enough to compensate forthe efficiency losses caused by the exit of the largest company, though notlarge enough to cause average profits to fall. For example, in Zambia–tobacco(Table A4.2(l)), total profits fall by 21.7 percent but average profits increaseby 5.10 percent. Finally, in the case of Rwanda–coffee, for instance, we findincreases in average and total profits (by 2.17 and 27.72 percent, respectively).In this particular case, the anticompetitive effect is very strong (because onlyfour firms remain and two of them have similar marginal costs to the one thatleft the market) and thus it compensates for the efficiency losses.

We now turn to study more extreme pro-competitive simulations. The firstscenario that we consider is one in which the existing firms are all equallyefficient (and as efficient as the leader). This is the equal market shares sim-ulation. In this model, competition is enhanced and efficiency improves, andboth channels cause large increases in price and quantities. In turn, this hasa positive effect in the average utility of all farmers, both producers and non-producers. A summary of results is reported in Figure A4.3. For example, inthe Zambia cotton case, prices increase by 27.21 percent, total average utilityincreases by 0.34 percent, the average utility of the producers increases by4.89 percent, and the average utility of the switchers increases by 26.78 per-cent. In the majority of the case studies, profits fall because the competitioneffect is stronger than the efficiency effect (as the number of firms remainsunchanged, average and total profits show the same proportional change). Inour leading-case, Zambia–cotton, profits decline by 1.33 percent.5

We end with a discussion of the perfect competition simulation, where weimpose the marginal cost of the larger firms on all incumbents, as in equalmarket shares, and we set farm-gate prices at the difference between theinternational prices and the marginal cost. Clearly, profits drop to zero, whileprices and quantities significantly increase. As a result, utility increases sig-nificantly as well. While this scenario can only be hypothetically realized, itnevertheless provides an interesting baseline for comparison purposes.

3.2 Complementarities

In this section, we summarize the impacts complementary factors to the pro-posed supply chain changes. As we mentioned above, we consider factors and

5There is only one case, Zambia–tobacco, where the efficiency effect is stronger andtotal and average profits increase by 11.26 percent.


policies that affect farmers and firms separately and jointly. Moreover, we alsoexplore the role of complementary increases in international prices. We areinterested not so much in the direct impact of the complementary factorsbut rather in the complementarities between the changes in competition andthe provision of supplementary factors. In other words, we want to assesswhether the impacts of a given change in competition along the supply chaincan be boosted by those complementary factors. All the results are reportedin Table A4.2. The column “respect to original” shows the impact of the com-plementary factors given the initial market structure (that is, without any ofthe simulated competition policies). Figure A4.4 summarizes the effects onfarm-gate prices.

Complementary factors that affect farmers make export crop adoptionmore profitable and hence they trigger a positive supply response. Since mar-ket conditions are kept constant and thus the demand for the crop is alsoconstant, the increase in supply brings farm-gate prices down. Note that thedrop in prices does not entail a utility loss, because the complementary fac-tors allow for productivity gains that are, in the end, welfare enhancing. Whenthe complementary factors affect firms, the productivity of the incumbentsincreases and thus marginal costs decline. This causes an increase in thedemand for the export crop, with farm-gate prices rising as a result. Finally,when there are complementary factors that affect both farmers and firms,prices tend to increase because the increase in demand is not matched by theincrease in supply. Note that this is not a general result: if the complementar-ities affecting farmers were bigger or the complementarities affecting firmswere smaller, equilibrium farm-gate prices could in fact decline. As before,lower prices do not necessarily entail welfare losses for the farmers if theyare generated by productivity enhancements (due to complementary policies).

It is important to note that, in our model, an increase in international priceshas a large impact on farm-gate prices. In Figure A4.4(b), we see that a 10 per-cent increase in international prices brings about increases in farm-gate pricesranging from 15 percent to 25 percent. In principle, we observe that theincrease in farm-gate prices is proportionally higher when the market struc-ture is characterized by tighter competition among firms. Changes in inter-national prices affect farm-gate prices via changes in the derived demand forthe export crop. Note that our results imply large pass-through rates (largerthan expected at least). This happens because in the calibrations the ratio offarm-gate prices to international prices is fairly small and thus even smallchanges in the absolute level of farm-gate prices can generate large propor-tional changes.6

Finally, we turn to explore the magnitude of the complementarities betweenthese complementary factors and the changes in competition along the supply

6In practice, the pass-through rate could be smaller if, for instance, the change in inter-national prices is seen as transitory.


chain. The results are reported in Table A4.2 (where each column showsthe implications of piling up complementary changes with market structurechanges). As before, we focus on the case of farm-gate prices in Zambia–cottonto illustrate the results. In Figure A4.5, we plot the difference between theimpact resulting from a simultaneous change in complementary factors andmarket structure and the impacts resulting from separate simulations of com-plementary factors and market structure. Overall, while we do find traces ofcomplementarities, these are relatively small. In fact, the complementaritieswith policies and factors that affect farms and firms (separately and jointly)are negligible. One reason behind this result is that the complementary poli-cies affect all farmers in the same fashion.7 There are larger complementari-ties with international prices and, interestingly, the larger the extent of com-petition among firms the larger these complementarities become.

4 OUTGROWER CONTRACTS: THEORY

In this section, we extend our standard model to include outgrower contracts.We first solve the model with this addition and we later explain how to adaptthe simulations to deal with these issues. To allow for the possibility of a liq-uidity constraint affecting the home-market decision, we rewrite the farmer’sproblem as follows:

V(ei;p, r) = maxh,m

hα(m+ c)1−α

subject to m+ ph � (p − λr)ei, h � 0, m � 0.

As explained above, the liquidity constraint is parameterized by λ. For ourpurposes, the distinctive feature of the model is that the farmer pays an inter-est rate r on any loan taken from the firms. This interest rate r depends onthe structure of the market.

The model behaves as before, except that we now add a function that deter-mines the interest rate

r = r(r∗, sh1, . . . , shn, J).

The interest rate depends on the exogenous cost of funds for the firms r∗, thenumber of firms n, the share of each firm shj and a parameter J (the “legal”system) that captures how good “institutions” are. For instance, a countrywith a given market structure (say, three firms) may have a well-functioningoutgrower scheme because of good rules of law, while another country withthe same market structure may suffer from a collapse of outgrower schemesbecause of bad institutions. Given these assumptions, we can write

r = r∗ +ϕ(sh1, . . . , shn, J).

7It would be possible to simulate policies that, for instance, only affect nonproducers(thus triggering a large supply response).


Ideally, the form of the function ϕ(·) should be determined as part of theequilibrium game. However, this entails a much more complicated dynamicgame-theoretic oligopsonistic game. Since developing such a model is out-side the scope of our analysis, we will work with functional form assump-tions. While this is a shortcoming, we believe we can still illustrate the maineconomic phenomena that we want to explore.

To operationalize the model, we proceed as follows. First, to capture thenotion that the equilibrium interest rate depends on both the number of firmsand the structure of competition, we assume thatϕ(sh1, . . . , shn, J) is a func-tion of the Herfindahl index

H =n∑j=1

(shj)2.

H ranges from 1/n to 1, where n is the number of firms—if there are n firmsand they are symmetric, then each has a share equal to 1/n and thusH = 1/n.

Also, we wantϕ(·) to depend on the institutional framework. If the marketdoes not have good “institutions,” it could be hard for firms to collect loans.This will be more difficult as n increases. So, we need ϕ to be close to zerowhen there is, say, a monopolist and/or when J is low. At the other extreme,ϕ can be very high when the market tends to perfect competition (if J is notgood enough). Note that ϕ should also depend on N ; in other words, even inthe case of symmetry, it matters if there are two firms, three firms, and so on.

In the end, we assume that

r = r∗ + (1−H) 2r∗

max(1−H0),

where max(1 − H0) is the higher value that (1 − H) could have before thesimulations are performed. That is,

r = r∗ + (1−H) 2r∗

1− 1/n0.

Note that the role of the second parenthesis is a sort of normalization forthe value that r can get. In our normalization, the maximum spread over r∗

is just 2r∗ (so that, in the worst scenario, the interest rate charged to thefarmers will be thrice as high as the cost of capital to the firms).

A key issue to note is that, in the model with outgrower contract, the supplyof the farmer depends on r . This makes sense: if the interest rate that thefarmer pays while producing for the market goes up, then the choice of marketproduction may be affected. This fact requires that we modify the model.

Given r = r∗+ϕ and given λ, the fraction of the investment that is financedwith a loan, the modified cutoff is

e(p) ≡ α(1+ λr)1−α

cp


and this gives a “new” supply function

S(p) = (1−α)∫ eeeif (e)de− e(p)(1− F(e(p))).

Note that, in the end, the formula is the same as before. The difference is that,now, e(p) depends on r and λ and, importantly, r depends among otherthings on the number of firms and on the market shares. In consequence,when we do the simulations and the number of firms n and the shares shrespond endogenously, this affects farm-gate prices and the interest rate and,in turn, both affect the supply of the farmers. This means that the model withoutgrower contract cannot be solved in the same way as the standard model.Instead, we need to solve simultaneously

H =n∑j=1

( qj∑nj=1 qj

)2

, (4.1)

r = r∗ + (1−H) 2r∗

1− 1/n0, (4.2)

e(p) = α(1+ λr)1−α

cp, (4.3)

n∑j=1

qj = (1−α)∫ eeeif (e)de− e(p)(1− F(e(p))), (4.4)


ps(Q)∀j ∈ n. (4.5)

By using this system of nonlinear equations all variables are identified atthe same time. With this model in mind, we run again the simulations andcomplementary changes to see how simulation results are affected by theexistence of outgrower contracts. We discuss these results next.

5 SIMULATION RESULTS WITH OUTGROWER CONTRACTS

The main purpose of the model with outgrower contracts is to assess thepoverty impacts of the interrelationship between the provision of servicesto the farmers (access to credit, seeds, and so on) and the level of competi-tion. We are particularly interested in identifying situations where increasesin competition can jeopardize the market by hindering the success of theoutgrower contracts. For this reason, in this section we briefly focus on howdifferent market structures affect farm-gate prices and interest rates. Resultsare reported in Table A4.3 for all the endogenous variables of the model (in theAppendix). As before, here we use summary graphs to illustrate the results.

In Figure A4.6, we plot the differences in the proportional change in farm-gate prices between the standard model and the model with outgrower con-tracts for the leader split simulation. Interestingly, we find that the differences


in the price effects are tiny. Specifically, they are never larger than 0.2 percent-age points. In most cases, the differences are negative, thus suggesting thatthe increase in farm-gate prices is slightly larger in the model with outgrowercontracts. The reason is that when the leader splits, competition increases.While this pushes prices up, the interest rate increases too, and this reducesfarm supply. In the end, the increase in prices is slightly higher than in thestandard model. Note that there are cases were the leader split simulation pro-duces a more fragmented market and the interest rate falls, so the oppositeresult holds. For example, in the Burkina Faso–cotton case, farm-gate pricesincrease by 0.17 percent more in the outgrower contract model. As can beseen in Table A4.3, the changes in farm-gate prices are very similar in all themarket structure simulations.

In the outgrower contract model, the change in the interest rate is one of themain channels through which farmers are affected. To see the kind of impactsdelivered by our model, in Figure A4.7 we plot the percentage change in theinterest rate for the small entrant and leaders merge and small entrant simu-lations. Here, whereas the standard and the outgrower contract models gener-ate quite similar changes in farm-gate prices, there are sizeable changes in theinterest rate. In the Zambian cotton case, the interest rate would increase byslightly less than 2 percent in the small entrant simulation and would decreaseby over 1 percent in the leaders merge/small entrant case. For our purposes,an increase in the interest rate is akin to a decline in farm-gate prices (or, inother words, to a lower increase in prices). The poverty implications of thesemechanisms are explored in Chapter 5.



0

5

10

15

25

20

BF-CT ZB-TB BN-CT ZB-CT CV-CT ML-TB RN-CF GN-CC CV-CC CV-CF UG-CF

Farm-gate price Quantities

Figure A4.1: Changes in farm-gate prices and quantities: leader split simulations.


−20

0

10

20

30

40

50

60

70

−10

Total industry profits Average industry profits

GN-CC CV-CT ZB-CT CV-CC CV-CF ZB-TB UG-CF RN-CF BN-CT BF-CT ML-TB

Special casesEE > CE EE < CE

Figure A4.2: Changes in total and average profits: leader split simulation.

35

40

10

15

20

25

30

0

5

Farm-gate price Quantities Producers' utility

BF-CT

GN-CC

ZB-CT

CV-CT

ZB-TB

ML-TB

CV-CC

CV-CF

BN-CT

UG-CF

Figure A4.3: Farm-gate prices, quantities, and utility: equal market shares.


−1

0

1

2

3

Affecting farmers Affecting firms Affecting both

BN-CTUG-CFCV-CFGN-CCCV-CCML-TBRN-CFZB-TBCV-CTZB-CT

International prices

GN-CCCV-CFCV-CTBN-CTUG-CFBF-CTRN-CFML-TBZB-CTCV-CC0

5

10

15

20

25

30

(a)

(b)

Figure A4.4: Farm-gate prices and complementary factors.

−1.5

−1.0

−0.5

0

0.5

1.0

1.5

Leadersplit

Smallentrant

Leadersmerge

Leaders merge

and small entrant

Exit of largest

Equal market shares

Affecting farmers

Affectingfirms

Affectingboth

Internationalprices

Figure A4.5: Changes in farm-gate prices: complementarities.


−0.20

−0.16

−0.12

−0.08

−0.04

0

−0.18

−0.14

−0.10

−0.06

−0.02

0.02

BF-CT

BN-CT

ZB-TB

ZB-CT

CV-CT

RN-CF

ML-TB

UG-CF

CV-CF

CV-CC

GN-CC

Figure A4.6: Changes in farm-gate prices: standardmodel and outgrower contract model.

Small entrant Leaders merge and small entrant

8

6

4

2

0

−2

−4

−6

−8

BF-CT

ZB-TB

ZB-CT

CV-CT

RN-CF

BN-CT

GN-CC

ML-TB

CV-CC

CV-CF

UG-CF

Figure A4.7: Changes in the interest rate.


Tab

leA

4.1

:M

odel

calibra

tion

.

Inco

me

Pro

du

cers

’uti

lity

/C

ase

shar

eto

tal

uti

lity

µσ

pf

cα

Bu

rkin

aFa

so,c

ott

on

1.3

11.3

16.0

20

231.8

10,0

00,0

00

0.4

3Z

amb

ia,t

ob

acco

0.7

30.7

34.6

33

2.1

47.4

10,0

00,0

00

0.1

6U

gan

da,

coff

ee2.4

02.4

05.2

16

241.6

3,6

00,0

00

0.3

6C

ôte

d’Ivo

ire,

cott

on

4.2

44.2

44.6

23

232.2

1,0

00,0

00

0.3

0Z

amb

ia,c

ott

on

2.9

72.9

84.2

36

233.7

1,2

00,0

00

0.1

6M

alaw

i,to

bac

co3.8

23.8

24.9

85

238.2

1,8

00,0

00

0.3

7B

enin

,cott

on

6.5

96.5

94.1

88

241.9

600,0

00

0.2

5R

wan

da,

coff

ee0.8

80.8

86.0

71

2.1

37.7

25,0

00,0

00

0.4

5M

alaw

i,co

tton

0.4

70.4

76.0

15

2.1

31.4

30,0

00,0

00

0.3

7C

ôte

d’Ivo

ire,

coco

a17.1

217.1

23.8

67

240.7

150,0

00

0.3

0G

han

a,co

coa

4.1

54.1

44.7

40

237.2

1,3

00,0

00

0.3

2C

ôte

d’Ivo

ire,

coff

ee6.8

16.8

14.3

55

238.8

600,0

00

0.3

0

Th

ein

com

esh

are

com

esfr

om

the

hou

seh

old

surv

eys.µ

den

ote

sm

ean

end

ow

men

t(l

ogn

orm

ald

istr

ibu

tion

),σ

den

ote

sst

and

ard

dev

iati

on

(logn

orm

ald

istr

ibu

tion

),p

fd

enote

sfa

rm-g

ate

pri

ce,a

ndc

andα

den

ote

uti

lity

fun

ctio

np

aram

eter

s(s

eete

xt).


Tab

leA

4.2

:Im

pact

ofsu

pply

chain

chan

ges

onth

est

an

dard

mod

el:(

a)

Zam

bia

,cot

ton

(per

cen

tage

chan

ges

).

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

Lead

ers

and

smal

lof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

mer

ge

entr

ant

larg

est

shar

esco

mp

etit

ion

(A)

Basi

cm

odel

Farm

-gat

ep

rice

0.0

08.9

21.0

1−8

.59

−6.2

2−1

1.8

427.2

171.6

4Q

uan

titi

es0.0

05.3

20.6

2−5

.44

−3.9

0−7

.58

15.3

535.9

1T

ota

lu

tili

ty0.0

00.1

10.0

1−0

.10

−0.0

7−0

.14

0.3

40.9

1Pro

du

cers

’uti

lity

0.0

01.5

60.1

7−3

.28

−2.3

8−4

.51

4.8

913.4

8N

on

pro

du

cers

’uti

lity

0.0

00.0

60.0

10.0

00.0

00.0

00.1

90.5

0Sw

itch

ers’

uti

lity

0.0

026.6

426.6

5−2

0.2

4−2

0.2

4−2

0.2

526.7

827.4

0T

ota

lin

du

stry

pro

fits

0.0

010.2

9−3

.86

−0.3

6−7

.71

−22.3

7−1

.33

−100.0

0A

vera

ge

ind

ust

ryp

rofi

ts0.0

010.2

9−1

9.8

824.5

5−7

.71

−2.9

6−1

.33

−100.0

0

(B)

Com

ple

men

tary

pol

icie

s(f

arm

ers)

Farm

-gat

ep

rice

−0.5

68.2

90.5

1−9

.17

−6.7

3−1

2.3

926.4

371.6

4Q

uan

titi

es14.2

020.1

214.9

38.0

89.8

55.6

931.2

654.5

3T

ota

lu

tili

ty1.4

41.5

61.4

51.3

31.3

61.2

81.8

12.4

6T

ota

lin

du

stry

pro

fits

14.7

126.5

310.0

514.4

85.8

8−1

0.3

914.2

5−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts14.7

126.5

3−8

.29

43.1

05.8

812.0

114.2

5−1

00.0

0


Tab

leA

4.2

:(a

)C

onti

nu

ed. Le

ader

sR

esp

ect

mer

ge

Exit

Equ

alto

Lead

erSm

all

Lead

ers

and

smal

lof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

mer

ge

entr

ant

larg

est

shar

esco

mp

etit

ion

(C)

Com

ple

men

tary

pol

icie

s(fi

rms)

Farm

-gat

ep

rice

2.6

511.5

03.8

5−6

.04

−3.4

2−9

.21

29.2

674.6

2Q

uan

titi

es1.6

26.8

02.3

4−3

.78

−2.1

3−5

.84

16.4

137.1

3T

ota

lu

tili

ty0.0

30.1

40.0

5−0

.07

−0.0

4−0

.11

0.3

60.9

5Pro

du

cers

’uti

lity

0.4

62.0

20.6

7−2

.31

−1.3

2−3

.51

5.2

814.0

7N

on

pro

du

cers

’uti

lity

0.0

20.0

80.0

3−0

.07

−0.0

7−0

.07

0.2

00.5

3T

ota

lin

du

stry

pro

fits

1.0

910.8

7−3

.44

1.3

1−6

.65

−19.7

21.6

7−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts1.0

9−7

.61

−19.5

326.6

3−6

.65

0.3

51.6

7−1

00.0

0

(D)

Com

ple

men

tary

pol

icie

s(f

arm

ers−

firm

s)Fa

rm-g

ate

pri

ce2.0

710.9

23.3

4−6

.63

−3.9

6−9

.78

28.4

774.6

2Q

uan

titi

es16.0

021.8

216.8

59.9

211.8

37.6

332.4

355.8

9T

ota

lu

tili

ty1.4

81.6

01.4

91.3

61.3

91.3

21.8

42.5

1T

ota

lin

du

stry

pro

fits

16.0

126.8

410.5

916.4

37.1

4−7

.32

17.6

9−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts16.0

15.7

0−7

.84

45.5

47.1

415.8

517.6

9−1

00.0

0

(E)

Inte

rnati

onalpri

ces

Farm

-gat

ep

rice

22.3

932.2

324.5

612.4

416.3

09.2

749.1

0103.7

3Q

uan

titi

es12.8

117.9

213.9

67.3

49.5

05.5

226.0

648.2

9T

ota

lu

tili

ty0.2

80.4

00.3

00.1

50.2

00.1

10.6

21.3

4Pro

du

cers

’uti

lity

4.0

05.8

44.4

02.1

92.8

81.6

29.0

619.8

8N

on

pro

du

cers

’uti

lity

0.1

60.2

20.1

70.0

90.1

10.0

60.3

40.7

4T

ota

lin

du

stry

pro

fits

24.2

231.4

216.1

227.0

515.2

06.5

632.5

9−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts24.2

29.5

2−3

.24

58.8

115.2

033.2

032.5

9−1

00.0

0


Tab

leA

4.2

:(b

)Ben

in,c

otto

n(p

erce

nta

ge

chan

ges

).

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

Lead

ers

and

smal

lof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

mer

ge

entr

ant

larg

est

shar

esco

mp

etit

ion

(A)

Basi

cm

odel

Farm

-gat

ep

rice

0.0

09.4

90.2

5−1

.90

−1.4

4−1

0.5

257.3

497.3

9Q

uan

titi

es0.0

05.1

90.1

4−1

.08

−0.8

2−6

.20

27.0

441.4

3T

ota

lu

tili

ty0.0

00.2

60.0

1−0

.05

−0.0

4−0

.29

1.6

32.7

8Pro

du

cers

’uti

lity

0.0

02.2

10.0

6−0

.78

−0.5

9−4

.26

13.8

623.8

1N

on

pro

du

cers

’uti

lity

0.0

00.1

20.0

00.0

00.0

00.0

00.7

31.2

4Sw

itch

ers’

uti

lity

0.0

027.2

327.3

8−2

0.3

0−2

0.2

8−2

0.2

527.9

928.9

6T

ota

lin

du

stry

pro

fits

0.0

038.1

0−0

.95

4.4

72.7

9−6

3.9

3−1

4.7

7−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts0.0

057.8

3−1

1.9

619.4

02.7

9−5

8.7

8−1

4.7

7−1

00.0

0

(B)

Com

ple

men

tary

pol

icie

s(f

arm

ers)

Farm

-gat

ep

rice

−0.4

48.9

4−0

.15

−2.3

7−1

.87

−10.9

456.5

197.3

9Q

uan

titi

es13.6

719.3

813.8

512.4

412.7

66.7

743.5

059.6

8T

ota

lu

tili

ty2.3

62.6

52.3

72.3

02.3

12.0

34.1

65.4

7T

ota

lin

du

stry

pro

fits

13.1

155.9

411.8

818.2

916.2

0−5

8.2

4−1

.74

−100.0

0A

vera

ge

ind

ust

ryp

rofi

ts13.1

178.2

2−0

.55

35.1

916.2

0−5

2.2

8−1

.74

−100.0

0


Tab

leA

4.2

:(b

)C

onti

nu

ed. Le

ader

sR

esp

ect

mer

ge

Exit

Equ

alto

Lead

erSm

all

Lead

ers

and

smal

lof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

mer

ge

entr

ant

larg

est

shar

esco

mp

etit

ion

(C)

Com

ple

men

tary

pol

icie

s(fi

rms)

Farm

-gat

ep

rice

2.0

911.2

62.4

20.1

10.6

7−8

.32

58.2

298.5

7Q

uan

titi

es1.1

76.1

21.3

60.0

60.3

8−4

.86

27.3

941.8

1T

ota

lu

tili

ty0.0

60.3

10.0

70.0

00.0

2−0

.23

1.6

52.8

2Pro

du

cers

’uti

lity

0.4

82.6

30.5

60.0

30.1

6−3

.37

14.0

824.1

0N

on

pro

du

cers

’uti

lity

0.0

30.1

50.0

30.0

00.0

1−0

.11

0.7

41.2

6T

ota

lin

du

stry

pro

fits

−1.9

635.1

0−3

.21

2.6

30.6

1−6

2.4

2−1

3.9

0−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts−1

.96

54.4

1−1

3.9

717.2

90.6

1−5

7.0

6−1

3.9

0−1

00.0

0

(D)

Com

ple

men

tary

pol

icie

s(f

arm

ers−

firm

s)Fa

rm-g

ate

pri

ce1.6

410.7

02.0

1−0

.37

0.2

3−8

.75

57.3

998.5

7Q

uan

titi

es14.9

720.4

115.2

013.7

114.0

98.2

643.8

860.1

0T

ota

lu

tili

ty2.4

22.7

12.4

32.3

62.3

82.1

04.1

95.5

1T

ota

lin

du

stry

pro

fits

10.9

752.6

39.3

916.2

613.8

0−5

6.5

0−0

.75

−100.0

0A

vera

ge

ind

ust

ryp

rofi

ts10.9

774.4

4−2

.77

32.8

713.8

0−5

0.2

9−0

.75

−100.0

0

(E)

Inte

rnati

onalpri

ces

Farm

-gat

ep

rice

19.7

828.8

020.5

317.2

818.3

58.8

076.4

1123.0

3Q

uan

titi

es10.4

514.7

810.8

29.2

09.7

44.8

234.2

449.3

5T

ota

lu

tili

ty0.5

50.8

10.5

80.4

80.5

10.2

52.1

83.5

2Pro

du

cers

’uti

lity

4.6

66.8

54.8

54.0

64.3

22.0

518.6

030.1

6N

on

pro

du

cers

’uti

lity

0.2

50.3

70.2

60.2

20.2

40.1

10.9

71.5

8T

ota

lin

du

stry

pro

fits

6.7

043.7

43.9

412.5

58.7

9−4

9.0

34.8

4−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts6.7

043.7

4−7

.61

28.6

38.7

9−4

1.7

54.8

4−1

00.0

0


Tab

leA

4.2

:(c

)Bu

rkin

aFa

so,c

otto

n(p

erce

nta

ge

chan

ges

).

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

Lead

ers

and

smal

lof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

mer

ge

entr

ant

larg

est

shar

esco

mp

etit

ion

(A)

Basi

cm

odel

Farm

-gat

ep

rice

0.0

022.2

61.0

0−3

.22

−1.2

3−2

7.6

837.6

689.3

8Q

uan

titi

es0.0

020.5

70.9

5−3

.07

−1.1

7−2

7.3

034.1

576.2

6T

ota

lu

tili

ty0.0

00.1

30.0

1−0

.02

−0.0

1−0

.15

0.2

20.5

5Pro

du

cers

’uti

lity

0.0

04.1

20.1

8−1

.34

−0.5

2−1

1.0

17.0

917.2

2N

on

pro

du

cers

’uti

lity

0.0

00.0

70.0

00.0

00.0

00.0

00.1

30.3

3Sw

itch

ers’

uti

lity

0.0

025.6

125.4

3−2

0.2

4−2

0.2

8−2

0.5

225.9

027.2

0T

ota

lin

du

stry

pro

fits

0.0

04.7

1−2

.57

7.2

82.3

9−6

6.3

6−1

0.2

4−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts0.0

057.0

6−2

6.9

360.9

22.3

9−4

9.5

4−1

0.2

4−1

00.0

0

(B)

Com

ple

men

tary

pol

icie

s(f

arm

ers)

Farm

-gat

ep

rice

−0.6

421.1

90.4

7−4

.01

−1.8

6−2

8.2

336.7

089.3

8Q

uan

titi

es25.5

550.3

726.8

521.5

724.1

2−8

.15

67.1

1119.2

7T

ota

lu

tili

ty5.4

35.5

95.4

45.4

15.4

25.2

55.7

06.1

1T

ota

lin

du

stry

pro

fits

25.3

632.6

621.7

734.9

628.2

9−5

6.8

513.9

0−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts25.3

699.0

0−8

.67

102.4

328.2

9−3

5.2

813.9

0−1

00.0

0


Tab

leA

4.2

:(c

)C

onti

nu

ed. Le

ader

sR

esp

ect

mer

ge

Exit

Equ

alto

Lead

erSm

all

Lead

ers

and

smal

lof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

mer

ge

entr

ant

larg

est

shar

esco

mp

etit

ion

(C)

Com

ple

men

tary

pol

icie

s(fi

rms)

Farm

-gat

ep

rice

2.7

523.9

64.0

8−0

.97

1.5

4−2

4.8

839.6

092.2

1Q

uan

titi

es2.6

022.1

03.8

6−0

.92

1.4

7−2

4.4

435.8

278.4

3T

ota

lu

tili

ty0.0

20.1

40.0

2−0

.01

0.0

1−0

.13

0.2

30.5

7Pro

du

cers

’uti

lity

0.4

94.4

40.7

3−0

.40

0.2

8−9

.96

7.4

717.7

7N

on

pro

du

cers

’uti

lity

0.0

10.0

80.0

1−0

.04

0.0

1−0

.04

0.1

40.3

4T

ota

lin

du

stry

pro

fits

0.2

07.8

2−3

.21

8.5

32.4

1−6

3.4

3−7

.55

−100.0

0A

vera

ge

ind

ust

ryp

rofi

ts0.2

061.7

4−2

7.4

062.7

92.4

1−4

5.1

5−7

.55

−100.0

0

(D)

Com

ple

men

tary

pol

icie

s(f

arm

ers−

firm

s)Fa

rm-g

ate

pri

ce2.0

922.8

73.5

3−1

.79

0.9

0−2

5.4

538.6

292.2

1Q

uan

titi

es28.7

352.2

230.4

124.2

027.3

4−4

.64

69.1

4121.8

9T

ota

lu

tili

ty5.4

55.6

05.4

65.4

25.4

45.2

75.7

26.1

3T

ota

lin

du

stry

pro

fits

25.6

436.5

721.0

136.5

528.3

5−5

3.1

417.2

8−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts25.6

4104.8

5−9

.24

104.8

228.3

5−2

9.7

117.2

8−1

00.0

0

(E)

Inte

rnati

onalpri

ces

Farm

-gat

ep

rice

23.0

542.3

325.9

216.8

821.8

4−7

.10

60.2

0122.4

8Q

uan

titi

es21.2

838.1

623.8

415.7

120.2

0−6

.82

53.1

2100.8

1T

ota

lu

tili

ty0.1

30.2

50.1

50.1

00.1

3−0

.04

0.3

60.7

7Pro

du

cers

’uti

lity

4.2

78.0

04.8

23.1

04.0

4−2

.94

11.5

123.6

1N

on

pro

du

cers

’uti

lity

0.0

80.1

50.0

90.0

60.0

7−0

.30

0.2

10.4

6T

ota

lin

du

stry

pro

fits

22.2

942.1

614.5

536.8

124.1

8−4

1.8

523.3

8−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts22.2

942.1

6−1

4.0

9105.2

124.1

8−1

2.7

723.3

8−1

00.0

0


Tab

leA

4.2

:(d

)C

ôte

d’Ivo

ire,

cott

on(p

erce

nta

ge

chan

ges

).

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

Lead

ers

and

smal

lof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

mer

ge

entr

ant

larg

est

shar

esco

mp

etit

ion

(A)

Basi

cm

odel

Farm

-gat

ep

rice

0.0

08.8

50.7

6−7

.37

−5.4

3−1

1.4

626.7

867.2

0Q

uan

titi

es0.0

05.7

70.5

1−5

.05

−3.6

9−7

.95

16.6

337.8

2T

ota

lu

tili

ty0.0

00.1

60.0

1−0

.13

−0.1

0−0

.21

0.4

91.2

6Pro

du

cers

’uti

lity

0.0

01.9

40.1

6−3

.02

−2.2

3−4

.68

6.0

115.4

6N

on

pro

du

cers

’uti

lity

0.0

00.0

80.0

10.0

00.0

00.0

00.2

40.6

1Sw

itch

ers’

uti

lity

0.0

026.2

026.2

8−2

0.2

5−2

0.2

3−2

0.2

626.4

127.2

7T

ota

lin

du

stry

pro

fits

0.0

010.3

4−3

.21

3.2

3−3

.60

−24.5

7−0

.54

−100.0

0A

vera

ge

ind

ust

ryp

rofi

ts0.0

010.3

4−1

9.3

429.0

4−3

.60

−5.7

2−0

.54

−100.0

0

(B)

Com

ple

men

tary

pol

icie

s(f

arm

ers)

Farm

-gat

ep

rice

−0.5

58.2

30.2

7−7

.95

−5.9

3−1

2.0

126.0

367.2

0Q

uan

titi

es16.1

822.7

216.8

110.3

912.0

07.1

034.9

759.3

7T

ota

lu

tili

ty2.9

83.1

73.0

02.8

32.8

72.7

53.5

44.4

3T

ota

lin

du

stry

pro

fits

16.7

528.8

512.7

320.6

812.4

8−1

1.2

717.2

5−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts16.7

528.8

5−6

.06

50.8

512.4

810.9

117.2

5−1

00.0

0


Tab

leA

4.2

:(d

)C

onti

nu

ed. Le

ader

sR

esp

ect

mer

ge

Exit

Equ

alto

Lead

erSm

all

Lead

ers

and

smal

lof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

mer

ge

entr

ant

larg

est

shar

esco

mp

etit

ion

(C)

Com

ple

men

tary

pol

icie

s(fi

rms)

Farm

-gat

ep

rice

2.9

811.6

33.9

4−4

.52

−2.3

3−8

.52

29.1

770.5

7Q

uan

titi

es1.9

87.5

32.6

0−3

.07

−1.5

7−5

.86

18.0

039.4

1T

ota

lu

tili

ty0.0

50.2

10.0

7−0

.08

−0.0

4−0

.15

0.5

41.3

3Pro

du

cers

’uti

lity

0.6

52.5

70.8

6−1

.86

−0.9

6−3

.48

6.5

616.2

6N

on

pro

du

cers

’uti

lity

0.0

30.1

00.0

4−0

.11

−0.1

1−0

.11

0.2

60.6

4T

ota

lin

du

stry

pro

fits

1.6

012.0

7−2

.40

5.3

6−2

.24

−21.2

43.0

6−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts1.6

012.0

7−1

8.6

731.7

0−2

.24

−1.5

43.0

6−1

00.0

0

(D)

Com

ple

men

tary

pol

icie

s(f

arm

ers−

firm

s)Fa

rm-g

ate

pri

ce2.4

110.9

93.4

3−5

.12

−2.8

6−9

.08

28.3

970.5

7Q

uan

titi

es18.4

224.7

019.1

912.6

414.4

19.4

836.5

161.1

6T

ota

lu

tili

ty3.0

53.2

23.0

72.8

92.9

42.8

13.5

94.5

1T

ota

lin

du

stry

pro

fits

18.6

730.9

213.7

323.2

114.1

2−7

.34

21.4

6−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts18.6

730.9

2−5

.22

54.0

114.1

215.8

321.4

6−1

00.0

0

(E)

Inte

rnati

onalpri

ces

Farm

-gat

ep

rice

24.0

933.6

126.0

115.3

018.7

611.3

050.4

0100.7

4Q

uan

titi

es15.0

720.5

016.1

89.8

011.9

17.3

229.4

952.7

8T

ota

lu

tili

ty0.4

40.6

20.4

80.2

80.3

40.2

10.9

41.9

1Pro

du

cers

’uti

lity

5.3

97.5

95.8

33.3

94.1

82.4

911.5

123.3

6N

on

pro

du

cers

’uti

lity

0.2

20.3

00.2

30.1

40.1

70.1

00.4

60.9

3T

ota

lin

du

stry

pro

fits

27.7

936.0

419.7

434.6

322.4

58.4

537.5

4−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts27.7

913.3

7−0

.22

68.2

822.4

535.5

637.5

4−1

00.0

0


Table A4.2: (e) Malawi, cotton (percentage changes).

Smallentrant

Respect with halfto Three Four of the

original firms firms benefits

(A) Basic modelFarm-gate price 0.00 12.15 19.64 6.09Quantities 0.00 11.62 18.62 5.86Total utility 0.00 0.03 0.04 0.01Producers’ utility 0.00 2.18 3.58 1.08Nonproducers’ utility 0.00 0.02 0.03 0.01Switchers’ utility 0.00 25.43 25.52 25.38Total industry profits 0.00 −14.67 −26.52 −16.05Average industry profits 0.00 −43.11 −63.26 −44.03

(B) Complementary policies (farmers)Farm-gate price −0.75 11.49 19.06 5.47Quantities 25.66 40.13 48.86 33.07Total utility 4.62 4.66 4.68 4.64Total industry profits 27.48 8.92 −6.11 6.94Average industry profits 27.48 −27.39 −53.06 −28.70

(C) Complementary policies (firms)Farm-gate price 2.42 14.91 22.60 8.39Quantities 2.34 14.21 21.36 8.05Total utility 0.01 0.03 0.05 0.02Producers’ utility 0.42 2.69 4.14 1.49Nonproducers’ utility 0.00 0.02 0.03 0.01Total industry profits 5.30 −10.12 −22.59 −11.23Average industry profits 5.30 −40.08 −61.29 −40.82

(D) Complementary policies (farmers − firms)Farm-gate price 1.65 14.23 22.00 7.76Quantities 28.53 43.30 52.21 35.76Total utility 4.63 4.66 4.68 4.65Total industry profits 34.16 14.67 −1.14 13.00Average industry profits 34.16 −23.55 −50.57 −24.66

(E) International pricesFarm-gate price 21.33 36.44 45.81 26.42Quantities 20.18 33.93 42.22 24.86Total utility 0.05 0.08 0.10 0.06Producers’ utility 3.90 6.81 8.65 4.87Nonproducers’ utility 0.03 0.05 0.06 0.03Total industry profits 51.40 29.49 11.70 32.17Average industry profits 51.40 −13.67 −44.15 −11.89


Tab

leA

4.2

:(f

)C

ôte

d’Ivo

ire,

coco

a(p

erce

nta

ge

chan

ges

).

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

Lead

ers

and

smal

lof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

mer

ge

entr

ant

larg

est

shar

esco

mp

etit

ion

(A)

Basi

cm

odel

Farm

-gat

ep

rice

0.0

01.9

50.3

7−1

.88

−1.3

8−2

.18

18.2

437.2

1Q

uan

titi

es0.0

00.8

50.1

6−0

.83

−0.6

1−0

.97

7.4

414.1

0T

ota

lu

tili

ty0.0

00.1

40.0

3−0

.14

−0.1

0−0

.16

1.3

32.7

0Pro

du

cers

’uti

lity

0.0

00.5

50.1

0−0

.78

−0.5

8−0

.91

5.2

210.6

9N

on

pro

du

cers

’uti

lity

0.0

00.0

50.0

10.0

00.0

00.0

00.5

00.9

9Sw

itch

ers’

uti

lity

0.0

030.1

230.2

4−2

0.2

4−2

0.3

5−2

0.2

930.2

330.5

5T

ota

lin

du

stry

pro

fits

0.0

06.9

6−2

.91

−3.5

4−7

.10

−8.6

5−1

.24

−100.0

0A

vera

ge

ind

ust

ryp

rofi

ts0.0

00.2

8−8

.98

3.3

5−7

.10

−2.1

3−1

.24

−100.0

0

(B)

Com

ple

men

tary

pol

icie

s(f

arm

ers)

Farm

-gat

ep

rice

−0.3

41.6

20.0

4−2

.23

−1.7

2−2

.53

17.8

037.2

1Q

uan

titi

es11.5

712.5

111.7

610.6

610.9

110.5

119.6

327.0

3T

ota

lu

tili

ty2.9

23.0

82.9

52.7

72.8

12.7

54.3

85.9

2T

ota

lin

du

stry

pro

fits

12.7

620.2

59.4

09.0

24.9

43.4

312.5

1−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts12.7

612.7

32.5

616.8

14.9

410.8

212.5

1−1

00.0

0


Tab

leA

4.2

:(f

)C

onti

nu

ed. Le

ader

sR

esp

ect

mer

ge

Exit

Equ

alto

Lead

erSm

all

Lead

ers

and

smal

lof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

mer

ge

entr

ant

larg

est

shar

esco

mp

etit

ion

(C)

Com

ple

men

tary

pol

icie

s(fi

rms)

Farm

-gat

ep

rice

2.5

34.4

82.9

30.6

41.1

70.3

420.3

839.8

1Q

uan

titi

es1.1

01.9

41.2

70.2

80.5

10.1

58.2

414.9

5T

ota

lu

tili

ty0.1

80.3

30.2

10.0

50.0

90.0

31.4

82.8

9Pro

du

cers

’uti

lity

0.7

21.2

70.8

30.1

80.3

30.1

05.8

311.4

4N

on

pro

du

cers

’uti

lity

0.0

70.1

20.0

80.0

20.0

30.0

10.5

51.0

6T

ota

lin

du

stry

pro

fits

1.7

17.9

6−1

.44

−1.2

8−5

.07

−6.1

11.9

4−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts1.7

11.2

1−7

.60

5.7

7−5

.07

0.6

01.9

4−1

00.0

0

(D)

Com

ple

men

tary

pol

icie

s(f

arm

ers−

firm

s)Fa

rm-g

ate

pri

ce2.1

74.1

32.5

90.2

80.8

3−0

.02

19.9

339.8

1Q

uan

titi

es12.7

713.6

812.9

611.8

712.1

311.7

320.5

027.9

5T

ota

lu

tili

ty3.1

33.2

83.1

62.9

73.0

22.9

54.5

56.1

3T

ota

lin

du

stry

pro

fits

14.7

621.4

611.1

311.6

37.2

96.3

516.1

2−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts14.7

613.8

74.1

919.6

17.2

913.9

516.1

2−1

00.0

0

(E)

Inte

rnati

onalpri

ces

Farm

-gat

ep

rice

22.0

524.2

222.6

619.9

320.6

919.6

340.0

563.9

4Q

uan

titi

es8.8

59.6

49.0

78.0

78.3

57.9

615.0

222.1

1T

ota

lu

tili

ty1.6

11.7

61.6

51.4

51.5

11.4

32.9

14.6

1Pro

du

cers

’uti

lity

6.3

26.9

46.4

95.7

15.9

25.6

211.5

118.3

8N

on

pro

du

cers

’uti

lity

0.6

00.6

60.6

10.5

40.5

60.5

31.0

61.6

7T

ota

lin

du

stry

pro

fits

26.6

730.8

521.9

825.5

820.1

721.2

033.1

7−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts26.6

722.6

814.3

634.5

520.1

729.8

533.1

7−1

00.0

0


Tab

leA

4.2

:(g

)G

han

a,c

ocoa

(per

cen

tage

chan

ges

).

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

Lead

ers

and

smal

lof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

mer

ge

entr

ant

larg

est

shar

esco

mp

etit

ion

(A)

Basi

cm

odel

Farm

-gat

ep

rice

0.0

03.4

30.2

8−1

.46

−1.0

5−4

.02

29.2

747.4

8Q

uan

titi

es0.0

02.3

50.2

0−1

.02

−0.7

2−2

.81

18.7

129.0

5T

ota

lu

tili

ty0.0

00.0

60.0

1−0

.03

−0.0

2−0

.07

0.5

30.8

6Pro

du

cers

’uti

lity

0.0

00.7

40.0

6−0

.60

−0.4

3−1

.65

6.5

610.7

7N

on

pro

du

cers

’uti

lity

0.0

00.0

30.0

00.0

00.0

00.0

00.2

60.4

2Sw

itch

ers’

uti

lity

0.0

026.0

526.1

5−2

0.2

6−2

0.3

2−2

0.2

426.3

326.7

0T

ota

lin

du

stry

pro

fits

0.0

029.4

7−2

.17

3.6

80.6

6−3

7.5

6−7

.07

−100.0

0A

vera

ge

ind

ust

ryp

rofi

ts0.0

018.6

8−1

0.3

214.0

50.6

6−3

1.3

2−7

.07

−100.0

0

(B)

Com

ple

men

tary

pol

icie

s(f

arm

ers)

Farm

-gat

ep

rice

−0.2

93.1

50.0

2−1

.77

−1.3

3−4

.31

28.8

647.4

8Q

uan

titi

es17.0

719.7

717.3

115.8

916.2

413.8

438.3

650.3

8T

ota

lu

tili

ty3.3

03.3

73.3

13.2

73.2

83.2

23.9

14.3

0T

ota

lin

du

stry

pro

fits

17.4

251.0

614.7

321.8

418.1

4−2

5.5

510.7

6−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts17.4

238.4

75.1

734.0

218.1

4−1

8.1

010.7

6−1

00.0

0


Tab

leA

4.2

:(g

)C

onti

nu

ed. Le

ader

sR

esp

ect

mer

ge

Exit

Equ

alto

Lead

erSm

all

Lead

ers

and

smal

lof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

mer

ge

entr

ant

larg

est

shar

esco

mp

etit

ion

(C)

Com

ple

men

tary

pol

icie

s(fi

rms)

Farm

-gat

ep

rice

3.0

66.4

83.3

91.5

62.0

4−0

.94

31.7

150.3

7Q

uan

titi

es2.1

04.4

02.3

21.0

81.4

0−0

.65

20.1

530.6

1T

ota

lu

tili

ty0.0

50.1

10.0

60.0

30.0

4−0

.02

0.5

70.9

2Pro

du

cers

’uti

lity

0.6

61.4

10.7

40.3

40.4

4−0

.39

7.1

211.4

4N

on

pro

du

cers

’uti

lity

0.0

30.0

60.0

30.0

10.0

2−0

.11

0.2

80.4

5T

ota

lin

du

stry

pro

fits

0.9

928.5

6−1

.55

4.9

31.5

3−3

4.1

8−3

.66

−100.0

0A

vera

ge

ind

ust

ryp

rofi

ts0.9

917.8

5−9

.75

15.4

21.5

3−2

7.6

0−3

.66

−100.0

0

(D)

Com

ple

men

tary

pol

icie

s(f

arm

ers−

firm

s)Fa

rm-g

ate

pri

ce2.7

66.1

93.1

11.2

51.7

4−1

.24

31.2

950.3

7Q

uan

titi

es19.4

622.1

019.7

418.2

818.6

716.3

139.9

952.1

5T

ota

lu

tili

ty3.3

73.4

43.3

73.3

33.3

43.2

83.9

64.3

6T

ota

lin

du

stry

pro

fits

18.6

750.1

015.5

523.4

019.2

5−2

1.5

314.7

8−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts18.6

737.5

95.9

235.7

319.2

5−1

3.6

914.7

8−1

00.0

0

(E)

Inte

rnati

onalpri

ces

Farm

-gat

ep

rice

24.8

928.6

025.4

823.0

823.8

620.5

653.8

976.6

6Q

uan

titi

es16.0

918.3

216.4

514.9

915.4

613.4

432.4

943.9

7T

ota

lu

tili

ty0.4

50.5

20.4

60.4

10.4

30.3

70.9

81.4

1Pro

du

cers

’uti

lity

5.5

56.4

05.6

95.1

45.3

24.5

712.2

617.5

7N

on

pro

du

cers

’uti

lity

0.2

20.2

50.2

20.2

00.2

10.1

80.4

80.6

9T

ota

lin

du

stry

pro

fits

25.0

649.3

420.4

530.9

525.2

3−6

.18

29.6

5−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts25.0

636.8

910.4

244.0

525.2

33.2

029.6

5−1

00.0

0


Tab

leA

4.2

:(h

)C

ôte

d’Ivo

ire,

coff

ee(p

erce

nta

ge

chan

ges

).

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

Lead

ers

and

smal

lof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

mer

ge

entr

ant

larg

est

shar

esco

mp

etit

ion

(A)

Basi

cm

odel

Farm

-gat

ep

rice

0.0

01.5

20.2

9−1

.47

−1.0

8−1

.71

14.2

527.5

7Q

uan

titi

es0.0

00.9

00.1

7−0

.88

−0.6

4−1

.02

8.0

815.0

0T

ota

lu

tili

ty0.0

00.0

40.0

1−0

.04

−0.0

3−0

.05

0.4

10.8

1Pro

du

cers

’uti

lity

0.0

00.3

60.0

7−0

.60

−0.4

4−0

.70

3.4

06.6

6N

on

pro

du

cers

’uti

lity

0.0

00.0

20.0

00.0

00.0

00.0

00.1

90.3

6Sw

itch

ers’

uti

lity

0.0

026.8

527.1

2−2

0.2

5−2

0.3

2−2

0.2

326.9

127.0

9T

ota

lin

du

stry

pro

fits

0.0

06.7

8−2

.95

−3.2

2−6

.89

−8.2

5−5

.87

−100.0

0A

vera

ge

ind

ust

ryp

rofi

ts0.0

00.1

1−9

.02

3.6

9−6

.89

−1.6

9−5

.87

−100.0

0

(B)

Com

ple

men

tary

pol

icie

s(f

arm

ers)

Farm

-gat

ep

rice

−0.2

51.2

70.0

5−1

.73

−1.3

3−1

.97

13.9

527.5

7Q

uan

titi

es14.6

615.6

714.8

613.6

713.9

413.5

123.6

831.5

8T

ota

lu

tili

ty2.9

12.9

62.9

22.8

62.8

82.8

63.3

83.8

3T

ota

lin

du

stry

pro

fits

15.8

823.3

412.3

612.4

28.0

96.7

710.1

8−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts15.8

815.6

35.3

420.4

58.0

914.3

910.1

8−1

00.0

0


Tab

leA

4.2

:(h

)C

onti

nu

ed. Le

ader

sR

esp

ect

mer

ge

Exit

Equ

alto

Lead

erSm

all

Lead

ers

and

smal

lof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

mer

ge

entr

ant

larg

est

shar

esco

mp

etit

ion

(C)

Com

ple

men

tary

pol

icie

s(fi

rms)

Farm

-gat

ep

rice

2.9

54.4

73.2

71.4

81.9

01.2

416.9

030.6

2Q

uan

titi

es1.7

42.6

21.9

20.8

71.1

20.7

49.5

116.5

1T

ota

lu

tili

ty0.0

90.1

30.0

90.0

40.0

50.0

40.4

90.9

0Pro

du

cers

’uti

lity

0.7

01.0

60.7

70.3

50.4

50.2

94.0

57.4

1N

on

pro

du

cers

’uti

lity

0.0

40.0

60.0

40.0

20.0

30.0

20.2

20.4

0T

ota

lin

du

stry

pro

fits

2.5

58.5

8−0

.69

−0.0

7−4

.02

−4.8

2−1

.75

−100.0

0A

vera

ge

ind

ust

ryp

rofi

ts2.5

51.8

0−6

.89

7.0

6−4

.02

1.9

8−1

.75

−100.0

0

(D)

Com

ple

men

tary

pol

icie

s(f

arm

ers−

firm

s)Fa

rm-g

ate

pri

ce2.6

94.2

23.0

21.2

01.6

40.9

716.5

930.6

2Q

uan

titi

es16.6

017.5

916.8

215.6

215.9

115.4

725.2

633.2

6T

ota

lu

tili

ty3.0

13.0

63.0

22.9

62.9

72.9

53.4

63.9

3T

ota

lin

du

stry

pro

fits

18.9

025.5

115.0

616.1

311.4

710.7

914.9

7−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts18.9

017.6

67.8

724.4

211.4

718.7

014.9

7−1

00.0

0

(E)

Inte

rnati

onalpri

ces

Farm

-gat

ep

rice

24.3

626.0

624.8

422.6

823.2

922.4

538.4

855.5

9Q

uan

titi

es13.3

814.2

413.6

312.5

312.8

412.4

120.2

827.9

3T

ota

lu

tili

ty0.7

10.7

60.7

30.6

60.6

80.6

51.1

31.6

4Pro

du

cers

’uti

lity

5.8

76.2

95.9

95.4

65.6

15.4

09.3

613.6

1N

on

pro

du

cers

’uti

lity

0.3

20.3

40.3

30.3

00.3

10.2

90.5

00.7

3T

ota

lin

du

stry

pro

fits

33.0

736.8

627.9

832.6

126.6

828.2

034.5

8−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts33.0

728.3

019.9

842.0

826.6

837.3

634.5

8−1

00.0

0


Tab

leA

4.2

:(i)

Rw

an

da,c

offee

(per

cen

tage

chan

ges

).

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

Lead

ers

and

smal

lof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

mer

ge

entr

ant

larg

est

shar

esco

mp

etit

ion

(A)

Basi

cm

odel

Farm

-gat

ep

rice

0.0

04.4

40.5

6−5

.84

−4.2

5−6

.08

8.7

131.9

1Q

uan

titi

es0.0

04.2

10.5

4−5

.61

−4.0

7−5

.84

8.2

129.2

8T

ota

lu

tili

ty0.0

00.0

20.0

0−0

.02

−0.0

2−0

.02

0.0

30.1

2Pro

du

cers

’uti

lity

0.0

00.7

90.1

0−2

.42

−1.7

6−2

.52

1.5

75.9

3N

on

pro

du

cers

’uti

lity

0.0

00.0

10.0

00.0

00.0

00.0

00.0

20.0

7Sw

itch

ers’

uti

lity

0.0

025.3

925.4

6−2

0.2

2−2

0.2

4−2

0.2

525.4

125.7

4T

ota

lin

du

stry

pro

fits

0.0

0−3

.93

−3.9

23.8

4−5

.05

2.1

7−4

.32

−100.0

0A

vera

ge

ind

ust

ryp

rofi

ts0.0

0−3

.93

−19.9

429.8

0−5

.05

27.7

2−4

.32

−100.0

0

(B)

Com

ple

men

tary

pol

icie

s(f

arm

ers)

Farm

-gat

ep

rice

−0.4

04.0

30.2

1−6

.29

−4.6

3−6

.52

8.2

631.9

1Q

uan

titi

es26.1

131.3

026.8

419.1

221.1

018.8

336.2

162.6

4T

ota

lu

tili

ty5.7

15.7

35.7

15.6

85.6

95.6

85.7

55.8

7T

ota

lin

du

stry

pro

fits

27.5

722.8

622.2

032.6

921.1

030.6

122.7

6−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts27.5

72.3

81.8

365.8

721.1

063.2

722.7

6−1

00.0

0


Tab

leA

4.2

:(i)

Con

tin

ued

. Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

Lead

ers

and

smal

lof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

mer

ge

entr

ant

larg

est

shar

esco

mp

etit

ion

(C)

Com

ple

men

tary

pol

icie

s(fi

rms)

Farm

-gat

ep

rice

2.6

77.1

23.3

7−3

.26

−1.4

9−3

.50

11.1

935.0

2Q

uan

titi

es2.5

36.7

23.2

0−3

.12

−1.4

3−3

.34

10.5

232.0

2T

ota

lu

tili

ty0.0

10.0

30.0

1−0

.01

−0.0

1−0

.01

0.0

40.1

4Pro

du

cers

’uti

lity

0.4

71.2

80.6

0−1

.36

−0.6

2−1

.45

2.0

26.5

2N

on

pro

du

cers

’uti

lity

0.0

10.0

20.0

1−0

.02

−0.0

2−0

.02

0.0

20.0

8T

ota

lin

du

stry

pro

fits

3.7

3−0

.47

−1.1

48.3

9−1

.47

6.7

70.4

0−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts3.7

3−1

7.0

6−1

7.6

235.4

9−1

.47

33.4

60.4

0−1

00.0

0

(D)

Com

ple

men

tary

pol

icie

s(f

arm

ers−

firm

s)Fa

rm-g

ate

pri

ce2.2

56.7

43.0

1−3

.72

−1.8

9−3

.95

10.7

335.0

2Q

uan

titi

es29.2

334.4

530.1

122.1

824.3

521.9

139.0

466.0

1T

ota

lu

tili

ty5.7

25.7

45.7

35.7

05.7

05.6

95.7

65.8

8T

ota

lin

du

stry

pro

fits

32.3

326.7

925.7

738.4

925.6

736.4

728.7

6−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts32.3

35.6

54.8

173.1

125.6

770.5

928.7

6−1

00.0

0

(E)

Inte

rnati

onalpri

ces

Farm

-gat

ep

rice

22.9

028.0

624.2

616.0

118.6

915.7

731.5

360.6

8Q

uan

titi

es21.2

325.8

622.4

614.9

717.4

114.7

528.9

453.9

1T

ota

lu

tili

ty0.0

90.1

10.0

90.0

60.0

70.0

60.1

20.2

5Pro

du

cers

’uti

lity

4.2

15.1

94.4

72.9

23.4

22.8

75.8

511.4

8N

on

pro

du

cers

’uti

lity

0.0

50.0

60.0

50.0

40.0

40.0

40.0

70.1

4T

ota

lin

du

stry

pro

fits

43.0

934.9

532.9

452.5

136.2

350.7

943.2

9−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts43.0

912.4

610.7

890.6

436.2

388.4

943.2

9−1

00.0

0


Tab

leA

4.2

:(j

)U

gan

da,c

offee

(per

cen

tage

chan

ges

).

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

Lead

ers

and

smal

lof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

mer

ge

entr

ant

larg

est

shar

esco

mp

etit

ion

(A)

Basi

cm

odel

Farm

-gat

ep

rice

0.0

00.9

60.1

3−0

.96

−0.7

6−1

.08

9.1

617.8

9Q

uan

titi

es0.0

00.7

70.1

1−0

.77

−0.6

1−0

.87

7.1

813.7

6T

ota

lu

tili

ty0.0

00.0

10.0

0−0

.01

−0.0

1−0

.01

0.1

00.1

9Pro

du

cers

’uti

lity

0.0

00.1

90.0

3−0

.40

−0.3

2−0

.45

1.8

53.6

6N

on

pro

du

cers

’uti

lity

0.0

00.0

10.0

00.0

00.0

00.0

00.0

50.1

0Sw

itch

ers’

uti

lity

0.0

025.6

525.7

1−2

0.2

8−2

0.2

2−2

0.2

225.6

825.7

6T

ota

lin

du

stry

pro

fits

0.0

03.9

4−2

.10

−1.6

7−4

.51

−4.7

1−1

4.3

8−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts0.0

0−2

.17

−7.8

64.8

9−4

.51

1.6

5−1

4.3

8−1

00.0

0

(B)

Com

ple

men

tary

pol

icie

s(f

arm

ers)

Farm

-gat

ep

rice

−0.1

80.7

9−0

.04

−1.1

5−0

.94

−1.2

78.9

617.8

9Q

uan

titi

es20.3

521.2

620.4

819.4

419.6

319.3

228.7

636.6

5T

ota

lu

tili

ty3.9

13.9

23.9

13.8

93.9

03.8

94.0

24.1

3T

ota

lin

du

stry

pro

fits

21.5

726.0

418.9

019.8

016.2

216.2

25.2

8−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts21.5

718.6

311.9

027.7

816.2

223.9

65.2

8−1

00.0

0


Tab

leA

4.2

:(j

)C

onti

nu

ed. Le

ader

sR

esp

ect

mer

ge

Exit

Equ

alto

Lead

erSm

all

Lead

ers

and

smal

lof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

mer

ge

entr

ant

larg

est

shar

esco

mp

etit

ion

(C)

Com

ple

men

tary

pol

icie

s(fi

rms)

Farm

-gat

ep

rice

2.8

13.7

82.9

61.8

42.0

61.7

211.7

820.7

7Q

uan

titi

es2.2

32.9

92.3

51.4

71.6

41.3

79.1

815.8

8T

ota

lu

tili

ty0.0

30.0

40.0

30.0

20.0

20.0

20.1

20.2

2Pro

du

cers

’uti

lity

0.5

60.7

60.5

90.3

70.4

10.3

42.3

94.2

7N

on

pro

du

cers

’uti

lity

0.0

20.0

20.0

20.0

10.0

10.0

10.0

70.1

2T

ota

lin

du

stry

pro

fits

2.7

06.0

90.2

91.5

1−1

.64

−1.3

8−1

0.1

7−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts2.7

0−0

.15

−5.6

18.2

8−1

.64

5.2

0−1

0.1

7−1

00.0

0

(D)

Com

ple

men

tary

pol

icie

s(f

arm

ers−

firm

s)Fa

rm-g

ate

pri

ce2.6

23.5

92.7

81.6

51.8

81.5

311.5

720.7

7Q

uan

titi

es22.9

623.8

623.1

222.0

622.2

721.9

531.1

039.1

3T

ota

lu

tili

ty3.9

43.9

53.9

43.9

33.9

33.9

34.0

54.1

7T

ota

lin

du

stry

pro

fits

24.9

228.7

221.8

623.7

219.7

620.3

110.4

1−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts24.9

221.1

514.6

931.9

719.7

628.3

310.4

1−1

00.0

0

(E)

Inte

rnati

onalpri

ces

Farm

-gat

ep

rice

23.8

124.8

924.0

722.7

123.0

522.5

932.9

244.0

7Q

uan

titi

es18.0

918.8

818.2

817.3

017.5

417.2

124.5

732.1

9T

ota

lu

tili

ty0.2

50.2

60.2

60.2

40.2

40.2

40.3

50.4

8Pro

du

cers

’uti

lity

4.9

15.1

54.9

74.6

84.7

54.6

66.8

69.2

7N

on

pro

du

cers

’uti

lity

0.1

30.1

40.1

40.1

30.1

30.1

30.1

90.2

5T

ota

lin

du

stry

pro

fits

36.0

237.5

531.7

136.8

131.5

834.0

027.2

2−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts36.0

229.4

623.9

645.9

331.5

842.9

327.2

2−1

00.0

0


Tab

leA

4.2

:(k

)M

ala

wi,

tobacc

o(p

erce

nta

ge

chan

ges

).

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

Lead

ers

and

smal

lof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

mer

ge

entr

ant

larg

est

shar

esco

mp

etit

ion

(A)

Basi

cm

odel

Farm

-gat

ep

rice

0.0

05.4

50.1

4−5

.12

−4.3

7−6

.00

21.3

946.1

2Q

uan

titi

es0.0

03.9

60.1

0−3

.82

−3.2

6−4

.49

14.9

630.5

2T

ota

lu

tili

ty0.0

00.0

90.0

0−0

.08

−0.0

7−0

.10

0.3

50.7

7Pro

du

cers

’uti

lity

0.0

01.1

60.0

3−2

.09

−1.7

9−2

.45

4.6

410.1

8N

on

pro

du

cers

’uti

lity

0.0

00.0

40.0

00.0

00.0

00.0

00.1

80.3

8Sw

itch

ers’

uti

lity

0.0

025.8

426.1

0−2

0.2

3−2

0.2

3−2

0.2

526.0

026.4

8T

ota

lin

du

stry

pro

fits

0.0

05.0

0−0

.88

−6.5

0−1

0.3

4−1

5.9

3−1

4.5

4−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts0.0

022.5

0−1

3.2

79.0

9−1

0.3

4−1

.92

−14.5

4−1

00.0

0

(B)

Com

ple

men

tary

pol

icie

s(f

arm

ers)

Farm

-gat

ep

rice

−0.4

04.9

7−0

.22

−5.5

4−4

.75

−6.4

220.8

746.1

2Q

uan

titi

es18.4

622.9

918.6

214.0

014.7

013.2

335.6

953.9

9T

ota

lu

tili

ty3.9

24.0

23.9

23.8

23.8

43.8

14.3

34.8

2T

ota

lin

du

stry

pro

fits

18.8

925.3

217.5

611.5

36.7

30.5

33.0

0−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts18.8

946.2

02.8

630.1

26.7

317.2

83.0

0−1

00.0

0


Tab

leA

4.2

:(k

)C

onti

nu

ed. Le

ader

sR

esp

ect

mer

ge

Exit

Equ

alto

Lead

erSm

all

Lead

ers

and

smal

lof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

mer

ge

entr

ant

larg

est

shar

esco

mp

etit

ion

(C)

Com

ple

men

tary

pol

icie

s(fi

rms)

Farm

-gat

ep

rice

2.6

47.8

42.8

9−2

.50

−1.6

3−3

.36

23.5

848.8

7Q

uan

titi

es1.9

45.6

72.1

1−1

.85

−1.2

1−2

.50

16.4

132.1

5T

ota

lu

tili

ty0.0

40.1

30.0

5−0

.04

−0.0

3−0

.05

0.3

90.8

1Pro

du

cers

’uti

lity

0.5

61.6

70.6

1−1

.02

−0.6

7−1

.38

5.1

210.8

0N

on

pro

du

cers

’uti

lity

0.0

20.0

60.0

2−0

.09

−0.0

9−0

.09

0.1

90.4

1T

ota

lin

du

stry

pro

fits

0.9

16.9

3−0

.59

−4.7

8−9

.20

−13.7

9−1

1.4

9−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts0.9

124.7

5−1

3.0

211.0

9−9

.20

0.5

8−1

1.4

9−1

00.0

0

(D)

Com

ple

men

tary

pol

icie

s(f

arm

ers−

firm

s)Fa

rm-g

ate

pri

ce2.2

37.3

42.5

1−2

.93

−2.0

2−3

.79

23.0

548.8

7Q

uan

titi

es20.7

024.9

520.9

416.2

817.0

615.5

337.3

655.8

6T

ota

lu

tili

ty3.9

74.0

73.9

73.8

73.8

93.8

64.3

74.8

7T

ota

lin

du

stry

pro

fits

20.0

427.6

717.9

713.6

38.1

53.1

36.6

4−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts20.0

448.9

53.2

232.5

68.1

520.3

26.6

4−1

00.0

0

(E)

Inte

rnati

onalpri

ces

Farm

-gat

ep

rice

22.7

027.9

723.4

116.9

218.3

816.0

543.8

574.5

0Q

uan

titi

es15.8

319.2

716.3

011.9

612.9

511.3

729.1

646.5

2T

ota

lu

tili

ty0.3

70.4

60.3

90.2

80.3

00.2

60.7

31.2

6Pro

du

cers

’uti

lity

4.9

36.1

05.0

93.6

53.9

73.4

69.6

616.5

8N

on

pro

du

cers

’uti

lity

0.1

90.2

30.1

90.1

40.1

50.1

30.3

60.6

3T

ota

lin

du

stry

pro

fits

24.6

531.5

520.1

221.5

313.8

112.7

319.0

2−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts24.6

515.1

15.1

141.7

913.8

131.5

219.0

2−1

00.0

0


Tab

leA

4.2

:(l)

Zam

bia

,tob

acc

o(p

erce

nta

ge

chan

ges

).

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

Lead

ers

and

smal

lof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

mer

ge

entr

ant

larg

est

shar

esco

mp

etit

ion

(A)

Basi

cm

odel

Farm

-gat

ep

rice

0.0

09.6

32.2

0−7

.33

−3.4

4−1

3.9

021.4

264.4

5Q

uan

titi

es0.0

06.9

51.6

2−5

.52

−2.5

6−1

0.6

415.0

541.4

0T

ota

lu

tili

ty0.0

00.0

30.0

1−0

.02

−0.0

1−0

.04

0.0

70.2

1Pro

du

cers

’uti

lity

0.0

01.4

60.3

3−2

.89

−1.3

6−5

.40

3.3

310.6

4N

on

pro

du

cers

’uti

lity

0.0

00.0

20.0

00.0

00.0

00.0

00.0

40.1

3Sw

itch

ers’

uti

lity

0.0

025.6

225.6

2−2

0.2

4−2

0.2

2−2

0.2

525.6

926.2

3T

ota

lin

du

stry

pro

fits

0.0

07.6

4−7

.98

12.0

8−0

.05

−21.1

711.2

6−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts0.0

0−1

3.8

9−2

6.3

849.4

4−0

.05

5.1

011.2

6−1

00.0

0

(B)

Com

ple

men

tary

pol

icie

s(f

arm

ers)

Farm

-gat

ep

rice

−0.5

29.1

31.7

4−7

.91

−3.9

3−1

4.4

220.7

564.4

5Q

uan

titi

es16.9

524.9

718.8

710.5

414.0

24.6

834.1

364.7

1T

ota

lu

tili

ty1.5

11.5

41.5

11.4

81.4

91.4

61.5

81.7

5T

ota

lin

du

stry

pro

fits

17.9

026.5

28.3

332.1

617.7

2−6

.49

31.7

4−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts17.9

01.2

2−1

3.3

476.2

117.7

224.6

831.7

4−1

00.0

0


Tab

leA

4.2

:(l)

Con

tin

ued

. Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

Lead

ers

and

smal

lof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

mer

ge

entr

ant

larg

est

shar

esco

mp

etit

ion

(C)

Com

ple

men

tary

pol

icie

s(fi

rms)

Farm

-gat

ep

rice

1.3

410.9

63.6

8−6

.10

−2.0

2−1

2.5

322.3

265.7

4Q

uan

titi

es0.9

97.8

82.6

9−4

.58

−1.5

0−9

.55

15.6

442.1

2T

ota

lu

tili

ty0.0

00.0

30.0

1−0

.02

−0.0

1−0

.04

0.0

70.2

2Pro

du

cers

’uti

lity

0.2

01.6

60.5

5−2

.41

−0.8

1−4

.88

3.4

710.8

6N

on

pro

du

cers

’uti

lity

0.0

00.0

20.0

1−0

.01

−0.0

1−0

.01

0.0

40.1

4T

ota

lin

du

stry

pro

fits

0.9

27.7

9−7

.48

13.2

50.6

7−1

8.9

312.8

6−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts0.9

2−1

3.7

6−2

5.9

851.0

00.6

78.1

012.8

6−1

00.0

0

(D)

Com

ple

men

tary

pol

icie

s(f

arm

ers−

firm

s)Fa

rm-g

ate

pri

ce0.8

110.4

53.2

1−6

.69

−2.5

3−1

3.0

621.6

465.7

4Q

uan

titi

es18.0

826.0

320.1

011.6

215.2

35.9

234.8

165.5

3T

ota

lu

tili

ty1.5

11.5

51.5

21.4

91.5

01.4

61.5

91.7

6T

ota

lin

du

stry

pro

fits

19.0

126.7

48.9

433.5

618.6

0−3

.84

33.6

2−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts19.0

11.3

9−1

2.8

478.0

818.6

028.2

233.6

2−1

00.0

0

(E)

Inte

rnati

onalpri

ces

Farm

-gat

ep

rice

16.0

526.6

019.2

47.3

712.6

60.9

037.2

487.3

3Q

uan

titi

es11.4

118.4

713.5

85.3

59.0

70.6

625.2

753.7

7T

ota

lu

tili

ty0.0

50.0

80.0

60.0

20.0

40.0

00.1

20.3

0Pro

du

cers

’uti

lity

2.4

64.1

72.9

71.1

11.9

30.1

35.9

414.7

1N

on

pro

du

cers

’uti

lity

0.0

30.0

50.0

40.0

10.0

20.0

00.0

70.1

8T

ota

lin

du

stry

pro

fits

24.0

528.7

112.2

539.7

422.9

26.4

141.0

5−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts24.0

52.9

7−1

0.2

086.3

222.9

241.8

841.0

5−1

00.0

0


Tab

leA

4.3

:Im

pact

ofsu

pply

chain

chan

ges

onth

eou

tgro

wer

con

tract

mod

el:(

a)

Zam

bia

,cot

ton

(per

cen

tage

chan

ges

).

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

Lead

ers

and

smal

lof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

mer

ge

entr

ant

larg

est

shar

esco

mp

etit

ion

(A)

Basi

cm

odel

Farm

-gat

ep

rice

0.0

08.9

31.0

1−8

.62

−6.2

2−1

1.8

427.2

871.6

4Q

uan

titi

es0.0

05.2

40.5

4−5

.21

−3.8

5−7

.59

14.8

535.3

2T

ota

lu

tili

ty0.0

00.0

90.0

1−0

.08

−0.0

6−0

.12

0.2

60.7

4Pro

du

cers

’uti

lity

0.0

01.5

90.0

9−2

.57

−2.0

3−4

.03

4.6

413.8

3N

on

pro

du

cers

’uti

lity

0.0

00.0

40.0

00.0

00.0

00.0

00.1

20.3

3Sw

itch

ers’

uti

lity

0.0

018.5

618.5

4−1

4.8

9−1

4.8

6−1

4.9

218.1

118.8

9T

ota

lin

du

stry

pro

fits

0.0

010.1

9−3

.93

−0.1

0−7

.65

−22.3

8−1

.93

−100.0

0A

vera

ge

ind

ust

ryp

rofi

ts0.0

010.1

9−1

9.9

424.8

7−7

.65

−2.9

8−1

.93

−100.0

0In

tere

stra

te0.0

01.8

21.6

7−5

.36

−1.2

60.3

811.0

111.0

1

(B)

Com

ple

men

tary

pol

icie

s(f

arm

ers)

Farm

-gat

ep

rice

−0.5

68.3

00.5

2−9

.20

−6.7

3−1

2.3

926.5

171.6

4Q

uan

titi

es14.1

820.0

114.8

38.3

19.8

95.6

630.7

053.8

8T

ota

lu

tili

ty1.4

21.5

21.4

31.3

41.3

61.2

91.7

12.2

6T

ota

lin

du

stry

pro

fits

14.6

926.4

09.9

514.7

55.9

2−1

0.4

213.5

7−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts14.6

926.4

0−8

.38

43.4

45.9

211.9

813.5

7−1

00.0

0In

tere

stra

te0.3

72.1

32.1

0−5

.00

−0.8

50.5

611.0

111.0

1

Exporters and Farmers 141T

able

A4.3

:(a

)C

onti

nu

ed. Le

ader

sR

esp

ect

mer

ge

Exit

Equ

alto

Lead

erSm

all

Lead

ers

and

smal

lof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

mer

ge

entr

ant

larg

est

shar

esco

mp

etit

ion

(C)

Com

ple

men

tary

pol

icie

s(fi

rms)

Farm

-gat

ep

rice

2.3

211.1

63.5

1−6

.38

−3.7

6−9

.52

29.0

174.1

4Q

uan

titi

es1.3

76.4

82.0

0−3

.81

−2.3

4−6

.09

15.7

336.3

5T

ota

lu

tili

ty0.0

20.1

10.0

3−0

.06

−0.0

4−0

.10

0.2

70.7

6Pro

du

cers

’uti

lity

0.3

71.9

70.4

8−1

.88

−1.2

9−3

.30

4.9

914.3

6N

on

pro

du

cers

’uti

lity

0.0

10.0

50.0

2−0

.06

−0.0

6−0

.06

0.1

30.3

5T

ota

lin

du

stry

pro

fits

0.7

110.4

5−3

.84

1.1

3−6

.96

−20.1

60.5

7−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts0.7

1−7

.96

−19.8

626.4

2−6

.96

−0.2

00.5

7−1

00.0

0In

tere

stra

te1.0

42.7

82.8

8−4

.22

−0.0

20.9

811.0

111.0

1

(D)

Com

ple

men

tary

pol

icie

s(f

arm

ers−

firm

s)Fa

rm-g

ate

pri

ce1.7

410.5

93.0

0−6

.97

−4.2

9−1

0.0

928.2

274.1

4Q

uan

titi

es15.7

021.4

316.4

59.8

711.5

77.3

431.6

855.0

2T

ota

lu

tili

ty1.4

51.5

51.4

61.3

61.3

81.3

11.7

32.2

9T

ota

lin

du

stry

pro

fits

15.5

526.3

410.1

216.2

16.7

6−7

.84

16.4

4−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts15.5

55.2

8−8

.23

45.2

76.7

615.2

016.4

4−1

00.0

0In

tere

stra

te1.3

73.1

83.2

7−3

.89

0.3

41.1

511.0

111.0

1

(E)

Inte

rnati

onalpri

ces

Farm

-gat

ep

rice

20.7

630.5

322.8

410.8

714.6

57.7

247.5

6101.3

2Q

uan

titi

es11.7

316.7

512.7

36.4

88.4

14.4

924.7

946.8

1T

ota

lu

tili

ty0.2

10.3

00.2

20.1

10.1

40.0

70.4

71.0

7Pro

du

cers

’uti

lity

3.7

75.6

14.0

42.1

42.6

01.3

08.7

920.1

4N

on

pro

du

cers

’uti

lity

0.1

00.1

40.1

00.0

50.0

70.0

40.2

20.4

9T

ota

lin

du

stry

pro

fits

21.9

729.3

114.0

824.8

413.0

84.0

529.1

6−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts21.9

77.7

6−4

.93

56.0

413.0

830.0

729.1

6−1

00.0

0In

tere

stra

te4.2

36.3

16.5

8−0

.78

3.6

82.7

911.0

111.0

1


Tab

leA

4.3

:(b

)Ben

in,c

otto

n(p

erce

nta

ge

chan

ges

).

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

Lead

ers

and

smal

lof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

mer

ge

entr

ant

larg

est

shar

esco

mp

etit

ion

(A)

Basi

cm

odel

Farm

-gat

ep

rice

0.0

09.5

40.2

5−1

.91

−1.4

5−1

0.4

657.4

797.3

9Q

uan

titi

es0.0

04.7

20.1

3−0

.99

−0.7

6−6

.83

26.2

540.5

6T

ota

lu

tili

ty0.0

00.1

90.0

0−0

.04

−0.0

3−0

.33

1.3

72.4

1Pro

du

cers

’uti

lity

0.0

01.7

40.0

4−0

.58

−0.4

5−4

.95

13.6

824.0

4N

on

pro

du

cers

’uti

lity

0.0

00.0

80.0

00.0

00.0

00.0

00.5

10.8

9Sw

itch

ers’

uti

lity

0.0

019.8

720.6

4−1

5.8

9−1

5.8

4−1

5.9

020.4

721.5

9T

ota

lin

du

stry

pro

fits

0.0

037.5

6−0

.97

4.5

62.8

5−6

4.2

7−1

5.5

8−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts0.0

057.2

1−1

1.9

719.5

02.8

5−5

9.1

6−1

5.5

8−1

00.0

0In

tere

stra

te0.0

012.2

10.4

3−2

.49

−1.7

117.6

719.3

519.3

5

(B)

Com

ple

men

tary

pol

icie

s(f

arm

ers)

Farm

-gat

ep

rice

−0.4

39.0

0−0

.15

−2.3

8−1

.87

−10.8

856.6

597.3

9Q

uan

titi

es13.6

418.8

513.8

012.5

112.8

06.0

642.6

358.7

3T

ota

lu

tili

ty2.3

32.5

42.3

42.2

92.3

01.9

73.8

75.0

5T

ota

lin

du

stry

pro

fits

13.0

955.3

211.8

418.3

616.2

4−5

8.6

2−2

.66

−100.0

0A

vera

ge

ind

ust

ryp

rofi

ts13.0

977.5

1−0

.59

35.2

716.2

4−5

2.7

1−2

.66

−100.0

0In

tere

stra

te0.7

412.6

71.2

2−1

.73

−0.9

117.7

219.3

519.3

5


able

A4.3

:(b

)C

onti

nu

ed. Le

ader

sR

esp

ect

mer

ge

Exit

Equ

alto

Lead

erSm

all

Lead

ers

and

smal

lof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

mer

ge

entr

ant

larg

est

shar

esco

mp

etit

ion

(C)

Com

ple

men

tary

pol

icie

s(fi

rms)

Farm

-gat

ep

rice

1.8

111.0

52.1

4−0

.16

0.3

9−8

.53

58.0

998.2

1Q

uan

titi

es0.9

55.4

81.1

2−0

.06

0.2

2−5

.67

26.5

040.8

3T

ota

lu

tili

ty0.0

40.2

20.0

50.0

00.0

1−0

.28

1.3

92.4

3Pro

du

cers

’uti

lity

0.3

72.0

80.4

2−0

.01

0.0

9−4

.24

13.8

324.2

5N

on

pro

du

cers

’uti

lity

0.0

20.0

90.0

2−0

.09

0.0

0−0

.09

0.5

10.9

0T

ota

lin

du

stry

pro

fits

−2.0

834.4

6−3

.33

2.5

60.5

4−6

2.9

7−1

4.9

8−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts−2

.08

53.6

7−1

4.0

717.2

10.5

4−5

7.6

8−1

4.9

8−1

00.0

0In

tere

stra

te1.6

013.0

12.1

4−0

.80

0.0

617.7

719.3

519.3

5

(D)

Com

ple

men

tary

pol

icie

s(f

arm

ers−

firm

s)Fa

rm-g

ate

pri

ce1.3

710.4

91.7

4−0

.64

−0.0

4−8

.96

57.2

698.2

1Q

uan

titi

es14.7

019.6

914.9

013.5

513.8

97.3

642.9

059.0

2T

ota

lu

tili

ty2.3

82.5

82.3

92.3

32.3

52.0

23.8

85.0

7T

ota

lin

du

stry

pro

fits

10.7

951.9

09.2

316.1

613.7

0−5

7.1

3−1

.97

−100.0

0A

vera

ge

ind

ust

ryp

rofi

ts10.7

973.6

0−2

.91

32.7

513.7

0−5

1.0

0−1

.97

−100.0

0In

tere

stra

te2.2

813.4

32.8

6−0

.10

0.7

917.8

219.3

519.3

5

(E)

Inte

rnati

onalpri

ces

Farm

-gat

ep

rice

18.4

327.4

419.1

515.9

717.0

07.5

375.2

3121.2

3Q

uan

titi

es9.4

914.0

39.8

28.3

58.8

23.4

232.9

647.9

4T

ota

lu

tili

ty0.4

40.6

90.4

50.3

90.4

10.1

11.8

33.0

4Pro

du

cers

’uti

lity

4.2

86.8

44.4

23.7

73.9

80.9

318.2

930.2

0N

on

pro

du

cers

’uti

lity

0.1

70.2

60.1

70.1

50.1

60.0

60.6

71.1

2T

ota

lin

du

stry

pro

fits

5.8

543.1

83.2

11.6

68.0

2−5

0.6

12.4

6−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts5.8

543.1

8−8

.22

7.6

18.0

2−4

3.5

52.4

6−1

00.0

0In

tere

stra

te6.8

62.6

07.7

14.7

85.8

418.1

219.3

519.3

5


Tab

leA

4.3

:(c

)Bu

rkin

aFa

so,c

otto

n(p

erce

nta

ge

chan

ges

).

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

Lead

ers

and

smal

lof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

mer

ge

entr

ant

larg

est

shar

esco

mp

etit

ion

(A)

Basi

cm

odel

Farm

-gat

ep

rice

0.0

022.4

41.0

1−3

.28

−1.2

5−2

7.5

937.9

389.3

8Q

uan

titi

es0.0

018.5

60.6

8−2

.25

−0.8

8−2

8.6

530.4

071.4

1T

ota

lu

tili

ty0.0

00.1

00.0

0−0

.01

0.0

0−0

.15

0.1

60.4

5Pro

du

cers

’uti

lity

0.0

03.2

40.0

3−0

.59

−0.2

5−1

1.4

55.5

215.8

2N

on

pro

du

cers

’uti

lity

0.0

00.0

50.0

00.0

00.0

00.0

00.0

90.2

4Sw

itch

ers’

uti

lity

0.0

020.8

021.6

1−1

7.8

9−1

7.8

8−1

8.3

620.2

721.7

9T

ota

lin

du

stry

pro

fits

0.0

02.6

9−2

.81

8.2

22.6

8−6

7.0

7−1

3.1

9−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts0.0

054.0

4−2

7.1

162.3

32.6

8−5

0.6

0−1

3.1

9−1

00.0

0In

tere

stra

te0.0

039.2

85.6

3−1

8.4

7−6

.28

38.8

267.1

367.1

3

(B)

Com

ple

men

tary

pol

icie

s(f

arm

ers)

Farm

-gat

ep

rice

−0.6

321.3

70.4

9−4

.07

−1.8

7−2

8.1

436.9

789.3

8Q

uan

titi

es25.4

247.9

326.3

722.4

824.3

2−9

.82

62.5

7113.4

0T

ota

lu

tili

ty5.4

35.5

45.4

35.4

25.4

25.2

45.6

25.9

8T

ota

lin

du

stry

pro

fits

25.2

330.1

721.3

336.0

028.5

0−5

7.7

410.2

4−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts25.2

395.2

5−9

.01

103.9

928.5

0−3

6.6

110.2

4−1

00.0

0In

tere

stra

te2.3

539.2

88.4

7−1

6.7

1−3

.72

38.8

367.1

367.1

3


able

A4.3

:(c

)C

onti

nu

ed. Le

ader

sR

esp

ect

mer

ge

Exit

Equ

alto

Lead

erSm

all

Lead

ers

and

smal

lof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

mer

ge

entr

ant

larg

est

shar

esco

mp

etit

ion

(C)

Com

ple

men

tary

pol

icie

s(fi

rms)

Farm

-gat

ep

rice

2.5

323.9

33.8

5−1

.22

1.3

1−2

5.0

039.6

491.8

7Q

uan

titi

es2.0

819.8

82.9

9−0

.51

1.2

1−2

6.0

531.8

573.2

8T

ota

lu

tili

ty0.0

10.1

00.0

10.0

00.0

1−0

.14

0.1

70.4

6Pro

du

cers

’uti

lity

0.3

03.5

40.3

50.0

40.2

3−1

0.5

75.8

616.3

2N

on

pro

du

cers

’uti

lity

0.0

10.0

60.0

1−0

.03

0.0

0−0

.03

0.0

90.2

5T

ota

lin

du

stry

pro

fits

−0.3

05.3

9−3

.92

8.9

42.1

7−6

4.4

2−1

0.9

0−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts−0

.30

58.0

8−2

7.9

463.4

12.1

7−4

6.6

3−1

0.9

0−1

00.0

0In

tere

stra

te6.3

639.2

813.2

2−1

3.4

90.7

638.8

767.1

367.1

3

(D)

Com

ple

men

tary

pol

icie

s(f

arm

ers−

firm

s)Fa

rm-g

ate

pri

ce1.8

822.8

53.3

2−2

.03

0.6

7−2

5.5

738.6

791.8

7Q

uan

titi

es27.9

849.5

329.2

024.6

126.8

9−6

.62

64.3

2115.6

6T

ota

lu

tili

ty5.4

45.5

55.4

45.4

35.4

35.2

65.6

36.0

0T

ota

lin

du

stry

pro

fits

24.9

133.5

420.0

036.9

427.9

2−5

4.3

813.1

2−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts24.9

1100.3

1−1

0.0

0105.4

227.9

2−3

1.5

713.1

2−1

00.0

0In

tere

stra

te8.4

939.2

815.7

8−1

1.8

83.0

838.8

867.1

367.1

3

(E)

Inte

rnati

onalpri

ces

Farm

-gat

ep

rice

21.9

741.3

324.7

515.8

720.7

4−8

.02

59.3

8120.7

9Q

uan

titi

es18.9

034.8

820.8

114.6

918.0

1−9

.46

48.0

894.2

9T

ota

lu

tili

ty0.1

00.2

00.1

10.0

80.1

0−0

.06

0.2

80.6

3Pro

du

cers

’uti

lity

3.5

57.0

23.8

13.0

43.4

2−4

.52

9.8

322.0

0N

on

pro

du

cers

’uti

lity

0.0

60.1

10.0

60.0

50.0

5−0

.27

0.1

50.3

5T

ota

lin

du

stry

pro

fits

19.6

438.1

511.8

135.0

421.7

4−4

4.4

017.6

5−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts19.6

438.1

5−1

6.1

5102.5

621.7

4−1

6.5

917.6

5−1

00.0

0In

tere

stra

te25.4

639.8

935.6

11.8

621.8

039.0

267.1

367.1

3


Tab

leA

4.3

:(d

)C

ôte

d’Ivo

ire,

cott

on(p

erce

nta

ge

chan

ges

).

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

Lead

ers

and

smal

lof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

mer

ge

entr

ant

larg

est

shar

esco

mp

etit

ion

(A)

Basi

cm

odel

Farm

-gat

ep

rice

0.0

08.8

60.7

6−7

.40

−5.4

4−1

1.4

626.8

567.2

0Q

uan

titi

es0.0

05.6

80.4

4−4

.79

−3.5

9−8

.01

16.0

837.1

7T

ota

lu

tili

ty0.0

00.1

40.0

1−0

.11

−0.0

8−0

.19

0.4

11.1

0Pro

du

cers

’uti

lity

0.0

02.0

10.1

1−2

.50

−1.9

6−4

.44

5.9

115.9

2N

on

pro

du

cers

’uti

lity

0.0

00.0

60.0

00.0

00.0

00.0

00.1

70.4

4Sw

itch

ers’

uti

lity

0.0

019.5

519.4

8−1

5.8

6−1

5.8

3−1

5.9

219.3

420.4

3T

ota

lin

du

stry

pro

fits

0.0

010.2

4−3

.27

3.5

4−3

.49

−24.6

3−1

.17

−100.0

0A

vera

ge

ind

ust

ryp

rofi

ts0.0

010.2

4−1

9.3

929.4

3−3

.49

−5.7

9−1

.17

−100.0

0In

tere

stra

te0.0

01.8

51.3

5−5

.79

−2.1

71.2

910.8

410.8

4

(B)

Com

ple

men

tary

pol

icie

s(f

arm

ers)

Farm

-gat

ep

rice

−0.5

48.2

40.2

8−7

.97

−5.9

4−1

2.0

026.0

967.2

0Q

uan

titi

es16.1

622.6

016.7

110.6

712.0

97.0

334.3

558.6

4T

ota

lu

tili

ty2.9

73.1

32.9

82.8

42.8

72.7

53.4

44.2

4T

ota

lin

du

stry

pro

fits

16.7

328.7

012.6

321.0

112.5

7−1

1.3

416.5

2−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts16.7

328.7

0−6

.14

51.2

612.5

710.8

216.5

2−1

00.0

0In

tere

stra

te0.3

92.1

81.8

1−5

.39

−1.7

11.4

510.8

410.8

4


able

A4.3

:(d

)C

onti

nu

ed. Le

ader

sR

esp

ect

mer

ge

Exit

Equ

alto

Lead

erSm

all

Lead

ers

and

smal

lof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

mer

ge

entr

ant

larg

est

shar

esco

mp

etit

ion

(C)

Com

ple

men

tary

pol

icie

s(fi

rms)

Farm

-gat

ep

rice

2.6

211.2

83.5

6−4

.89

−2.7

0−8

.86

28.8

870.0

6Q

uan

titi

es1.6

87.1

72.2

3−3

.10

−1.7

8−6

.18

17.2

438.5

3T

ota

lu

tili

ty0.0

40.1

80.0

5−0

.07

−0.0

4−0

.15

0.4

51.1

5Pro

du

cers

’uti

lity

0.5

62.5

60.7

1−1

.60

−0.9

9−3

.49

6.4

116.6

4N

on

pro

du

cers

’uti

lity

0.0

20.0

70.0

2−0

.09

−0.0

9−0

.09

0.1

80.4

7T

ota

lin

du

stry

pro

fits

1.1

611.5

0−2

.84

5.1

8−2

.56

−21.7

81.8

7−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts1.1

611.5

0−1

9.0

331.4

8−2

.56

−2.2

21.8

7−1

00.0

0In

tere

stra

te1.0

92.6

82.6

5−4

.56

−0.7

91.8

210.8

410.8

4

(D)

Com

ple

men

tary

pol

icie

s(f

arm

ers−

firm

s)Fa

rm-g

ate

pri

ce2.0

510.6

53.0

7−5

.48

−3.2

2−9

.42

28.1

170.0

6Q

uan

titi

es18.0

724.2

818.7

412.5

914.1

49.1

035.6

660.1

6T

ota

lu

tili

ty3.0

13.1

73.0

32.8

92.9

22.8

03.4

74.2

9T

ota

lin

du

stry

pro

fits

18.1

430.2

313.2

022.9

713.7

3−7

.98

20.0

7−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts18.1

430.2

3−5

.67

53.7

113.7

315.0

320.0

7−1

00.0

0In

tere

stra

te1.4

32.9

73.0

7−4

.20

−0.3

81.9

610.8

410.8

4

(E)

Inte

rnati

onalpri

ces

Farm

-gat

ep

rice

22.3

231.7

724.1

613.6

116.9

89.6

248.7

298.2

2Q

uan

titi

es13.7

919.1

314.7

58.8

110.6

66.0

928.0

051.0

5T

ota

lu

tili

ty0.3

60.5

10.3

80.2

30.2

70.1

50.7

81.6

5Pro

du

cers

’uti

lity

5.1

67.3

95.4

93.3

23.9

12.1

211.3

223.6

2N

on

pro

du

cers

’uti

lity

0.1

50.2

10.1

60.0

90.1

10.0

60.3

20.6

7T

ota

lin

du

stry

pro

fits

25.1

933.6

317.4

132.0

920.0

55.5

433.6

9−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts25.1

911.3

6−2

.16

65.1

120.0

531.9

333.6

9−1

00.0

0In

tere

stra

te4.3

46.1

66.5

2−0

.93

3.2

53.3

510.8

410.8

4


Table A4.3: (e) Malawi, cotton (percentage changes).

Smallentrant

Respect with halfto Three Four of the

original firms firms benefits

(A) Basic modelFarm-gate price 0.00 12.25 19.77 6.16Quantities 0.00 9.80 15.75 4.49Total utility 0.00 0.01 0.02 0.01Producers’ utility 0.00 1.28 2.23 0.31Nonproducers’ utility 0.00 0.01 0.01 0.00Switchers’ utility 0.00 17.71 17.23 17.89Total industry profits 0.00 −16.27 −28.59 −17.27Average industry profits 0.00 −44.18 −64.29 −44.84Interest rate 0.00 22.22 33.33 17.38

(B) Complementary policies (farmers)Farm-gate price −0.75 11.59 19.19 5.55Quantities 25.66 37.90 45.35 31.37Total utility 4.62 4.64 4.65 4.63Total industry profits 27.48 6.92 −8.70 5.41Average industry profits 27.48 −28.72 −54.35 −29.73Interest rate 0.00 22.22 33.33 17.54

(C) Complementary policies (firms)Farm-gate price 2.07 14.61 22.30 8.13Quantities 2.00 11.98 18.05 6.38Total utility 0.00 0.02 0.03 0.01Producers’ utility 0.41 1.76 2.75 0.73Nonproducers’ utility 0.00 0.01 0.02 0.01Total industry profits 4.52 −12.46 −25.32 −13.18Average industry profits 4.52 −41.64 −62.66 −42.12Interest rate 0.00 22.22 33.33 16.97

(D) Complementary policies (farmers − firms)Farm-gate price 1.30 13.93 21.71 7.50Quantities 28.11 40.58 48.16 33.69Total utility 4.63 4.64 4.66 4.63Total industry profits 33.18 11.73 − 4.57 10.54Average industry profits 33.18 −25.51 −52.28 −26.31Interest rate 0.00 22.22 33.33 17.14

(E) International pricesFarm-gate price 19.58 34.57 43.82 24.82Quantities 18.56 30.07 37.07 22.12Total utility 0.04 0.06 0.07 0.04Producers’ utility 3.96 5.89 7.22 4.31Nonproducers’ utility 0.02 0.03 0.04 0.02Total industry profits 46.79 23.24 5.31 26.35Average industry profits 46.79 −17.84 −47.34 −15.77Interest rate 0.00 22.22 33.33 13.63


Tab

leA

4.3

:(f

)C

ôte

d’Ivo

ire,

coco

a(p

erce

nta

ge

chan

ges

).

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

Lead

ers

and

smal

lof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

mer

ge

entr

ant

larg

est

shar

esco

mp

etit

ion

(A)

Basi

cm

odel

Farm

-gat

ep

rice

0.0

01.9

50.3

7−1

.88

−1.3

8−2

.18

18.2

537.2

1Q

uan

titi

es0.0

00.8

60.1

5−0

.83

−0.6

2−0

.98

7.3

714.0

4T

ota

lu

tili

ty0.0

00.1

40.0

2−0

.13

−0.1

0−0

.15

1.2

32.5

3Pro

du

cers

’uti

lity

0.0

00.6

10.1

0−0

.76

−0.5

8−0

.90

5.4

711.2

7N

on

pro

du

cers

’uti

lity

0.0

00.0

40.0

10.0

00.0

00.0

00.3

60.7

2Sw

itch

ers’

uti

lity

0.0

022.7

722.7

7−1

5.4

9−1

5.5

4−1

5.4

522.8

923.3

6T

ota

lin

du

stry

pro

fits

0.0

06.9

7−2

.92

−3.5

4−7

.11

−8.6

7−1

.34

−100.0

0A

vera

ge

ind

ust

ryp

rofi

ts0.0

00.2

9−8

.99

3.3

5−7

.11

−2.1

4−1

.34

−100.0

0In

tere

stra

te0.0

0−0

.22

0.2

30.0

10.3

10.3

31.7

71.7

7

(B)

Com

ple

men

tary

pol

icie

s(f

arm

ers)

Farm

-gat

ep

rice

−0.3

41.6

20.0

4−2

.23

−1.7

2−2

.53

17.8

137.2

1Q

uan

titi

es11.5

712.5

111.7

510.6

510.8

910.4

919.5

726.9

6T

ota

lu

tili

ty2.8

93.0

42.9

12.7

42.7

82.7

24.2

45.7

1T

ota

lin

du

stry

pro

fits

12.7

620.2

59.3

89.0

24.9

23.4

112.4

0−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts12.7

612.7

42.5

516.8

14.9

210.8

012.4

0−1

00.0

0In

tere

stra

te0.0

8−0

.12

0.3

20.0

70.3

80.3

81.7

71.7

7


Tab

leA

4.3

:(f

)C

onti

nu

ed. Le

ader

sR

esp

ect

mer

ge

Exit

Equ

alto

Lead

erSm

all

Lead

ers

and

smal

lof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

mer

ge

entr

ant

larg

est

shar

esco

mp

etit

ion

(C)

Com

ple

men

tary

pol

icie

s(fi

rms)

Farm

-gat

ep

rice

2.1

74.1

22.5

70.2

90.8

1−0

.01

20.0

339.3

8Q

uan

titi

es0.9

41.7

81.1

00.1

20.3

4−0

.02

8.0

414.7

4T

ota

lu

tili

ty0.1

50.2

80.1

70.0

20.0

5−0

.01

1.3

52.6

8Pro

du

cers

’uti

lity

0.6

51.2

50.7

60.0

80.2

2−0

.03

6.0

111.9

3N

on

pro

du

cers

’uti

lity

0.0

40.0

80.0

50.0

10.0

2−0

.24

0.3

90.7

6T

ota

lin

du

stry

pro

fits

1.3

07.6

1−1

.83

−1.7

3−5

.50

−6.5

81.3

1−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts1.3

00.8

8−7

.96

5.2

9−5

.50

0.0

91.3

1−1

00.0

0In

tere

stra

te0.1

6−0

.01

0.4

10.1

50.4

50.4

31.7

71.7

7

(D)

Com

ple

men

tary

pol

icie

s(f

arm

ers−

firm

s)Fa

rm-g

ate

pri

ce1.8

13.7

72.2

3−0

.08

0.4

7−0

.37

19.5

839.3

8Q

uan

titi

es12.5

913.5

112.7

811.6

911.9

411.5

420.2

927.7

3T

ota

lu

tili

ty3.0

53.2

03.0

82.9

12.9

42.8

84.3

85.8

7T

ota

lin

du

stry

pro

fits

14.2

921.0

510.6

811.1

26.8

05.8

115.4

0−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts14.2

913.4

93.7

719.0

66.8

013.3

715.4

0−1

00.0

0In

tere

stra

te0.2

30.0

80.4

80.2

00.5

10.4

71.7

71.7

7

(E)

Inte

rnati

onalpri

ces

Farm

-gat

ep

rice

20.2

722.4

220.8

618.1

718.9

117.8

738.3

061.7

7Q

uan

titi

es8.1

78.9

78.3

87.3

97.6

67.2

714.3

921.4

4T

ota

lu

tili

ty1.3

81.5

31.4

21.2

41.2

91.2

22.6

04.2

0Pro

du

cers

’uti

lity

6.1

66.8

26.3

25.5

25.7

35.4

211.6

018.7

4N

on

pro

du

cers

’uti

lity

0.4

00.4

40.4

10.3

60.3

70.3

50.7

41.1

9T

ota

lin

du

stry

pro

fits

24.2

628.6

619.7

022.9

817.7

118.5

430.1

1−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts24.2

620.6

212.2

231.7

617.7

127.0

030.1

1−1

00.0

0In

tere

stra

te0.6

90.6

70.9

60.5

70.8

90.7

61.7

71.7

7


Tab

leA

4.3

:(g

)G

han

a,c

ocoa

(per

cen

tage

chan

ges

).

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

Lead

ers

and

smal

lof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

mer

ge

entr

ant

larg

est

shar

esco

mp

etit

ion

(A)

Basi

cm

odel

Farm

-gat

ep

rice

0.0

03.4

30.2

8−1

.47

−1.0

5−4

.01

29.2

847.4

8Q

uan

titi

es0.0

02.4

90.1

8−0

.97

−0.7

1−3

.02

18.4

028.7

1T

ota

lu

tili

ty0.0

00.0

60.0

0−0

.02

−0.0

2−0

.07

0.4

60.7

6Pro

du

cers

’uti

lity

0.0

00.9

50.0

5−0

.51

−0.3

8−1

.79

6.8

011.3

0N

on

pro

du

cers

’uti

lity

0.0

00.0

20.0

00.0

00.0

00.0

00.1

90.3

1Sw

itch

ers’

uti

lity

0.0

019.5

519.6

2−1

5.8

3−1

5.7

5−1

5.7

819.5

620.0

5T

ota

lin

du

stry

pro

fits

0.0

029.6

4−2

.19

3.7

30.6

8−3

7.7

3−7

.41

−100.0

0A

vera

ge

ind

ust

ryp

rofi

ts0.0

018.8

4−1

0.3

414.1

00.6

8−3

1.5

0−7

.41

−100.0

0In

tere

stra

te0.0

0−2

.63

0.3

4−0

.83

−0.3

44.0

25.4

15.4

1

(B)

Com

ple

men

tary

pol

icie

s(f

arm

ers)

Farm

-gat

ep

rice

−0.2

93.1

50.0

2−1

.77

−1.3

3−4

.30

28.8

747.4

8Q

uan

titi

es17.0

619.9

017.2

715.9

316.2

513.6

038.0

049.9

9T

ota

lu

tili

ty3.2

93.3

63.2

93.2

63.2

73.2

03.8

14.1

7T

ota

lin

du

stry

pro

fits

17.4

151.2

214.6

921.8

818.1

5−2

5.7

510.3

6−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts17.4

138.6

25.1

434.0

718.1

5−1

8.3

210.3

6−1

00.0

0In

tere

stra

te0.2

3−2

.26

0.5

8−0

.60

−0.1

04.0

55.4

15.4

1


Tab

leA

4.3

:(g

)C

onti

nu

ed. Le

ader

sR

esp

ect

mer

ge

Exit

Equ

alto

Lead

erSm

all

Lead

ers

and

smal

lof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

mer

ge

entr

ant

larg

est

shar

esco

mp

etit

ion

(C)

Com

ple

men

tary

pol

icie

s(fi

rms)

Farm

-gat

ep

rice

2.6

76.0

73.0

01.1

71.6

5−1

.31

31.3

449.9

1Q

uan

titi

es1.8

04.2

32.0

00.8

21.1

2−1

.13

19.6

130.0

2T

ota

lu

tili

ty0.0

40.1

00.0

50.0

20.0

3−0

.03

0.4

90.8

1Pro

du

cers

’uti

lity

0.6

01.5

40.6

60.2

90.3

8−0

.77

7.3

111.9

0N

on

pro

du

cers

’uti

lity

0.0

20.0

40.0

20.0

10.0

1−0

.09

0.2

00.3

3T

ota

lin

du

stry

pro

fits

0.6

028.4

0−1

.91

4.5

61.1

7−3

4.8

2−4

.56

−100.0

0A

vera

ge

ind

ust

ryp

rofi

ts0.6

017.7

0−1

0.0

915.0

21.1

7−2

8.3

0−4

.56

−100.0

0In

tere

stra

te0.5

5−1

.78

0.9

1−0

.27

0.2

54.0

95.4

15.4

1

(D)

Com

ple

men

tary

pol

icie

s(f

arm

ers−

firm

s)Fa

rm-g

ate

pri

ce2.3

75.7

92.7

20.8

61.3

5−1

.62

30.9

249.9

1Q

uan

titi

es19.1

121.8

919.3

617.9

818.3

315.7

639.3

851.4

8T

ota

lu

tili

ty3.3

33.4

03.3

43.3

13.3

23.2

53.8

54.2

2T

ota

lin

du

stry

pro

fits

18.2

049.8

815.1

022.9

418.8

1−2

2.2

913.7

3−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts18.2

037.3

95.5

135.2

318.8

1−1

4.5

213.7

3−1

00.0

0In

tere

stra

te0.7

5−1

.44

1.1

3−0

.06

0.4

74.1

25.4

15.4

1

(E)

Inte

rnati

onalpri

ces

Farm

-gat

ep

rice

22.9

526.6

323.5

121.1

721.9

118.6

551.9

874.3

6Q

uan

titi

es14.7

817.0

915.1

013.7

314.1

612.0

031.1

342.5

0T

ota

lu

tili

ty0.3

70.4

30.3

70.3

40.3

50.2

90.8

41.2

2Pro

du

cers

’uti

lity

5.4

26.4

05.5

45.0

35.1

84.2

512.4

117.9

6N

on

pro

du

cers

’uti

lity

0.1

50.1

70.1

50.1

40.1

40.1

20.3

40.5

0T

ota

lin

du

stry

pro

fits

22.6

847.4

418.2

628.4

422.9

0−9

.13

26.1

4−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts22.6

835.1

58.4

041.2

822.9

0−0

.04

26.1

4−1

00.0

0In

tere

stra

te2.2

00.8

32.6

61.4

22.0

24.3

15.4

15.4

1


Tab

leA

4.3

:(h

)C

ôte

d’Ivo

ire,

coff

ee(p

erce

nta

ge

chan

ges

).

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

Lead

ers

and

smal

lof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

mer

ge

entr

ant

larg

est

shar

esco

mp

etit

ion

(A)

Basi

cm

odel

Farm

-gat

ep

rice

0.0

01.5

20.2

9−1

.47

−1.0

8−1

.70

14.2

627.5

7Q

uan

titi

es0.0

00.9

10.1

6−0

.88

−0.6

6−1

.04

8.0

014.9

1T

ota

lu

tili

ty0.0

00.0

40.0

1−0

.04

−0.0

3−0

.05

0.3

70.7

3Pro

du

cers

’uti

lity

0.0

00.4

00.0

6−0

.57

−0.4

4−0

.68

3.5

97.0

9N

on

pro

du

cers

’uti

lity

0.0

00.0

10.0

00.0

00.0

00.0

00.1

30.2

6Sw

itch

ers’

uti

lity

0.0

019.6

819.8

5−1

5.4

6−1

5.4

2−1

5.4

419.7

220.0

0T

ota

lin

du

stry

pro

fits

0.0

06.8

0−2

.97

−3.2

3−6

.90

−8.2

6−5

.98

−100.0

0A

vera

ge

ind

ust

ryp

rofi

ts0.0

00.1

2−9

.03

3.6

9−6

.90

−1.7

1−5

.98

−100.0

0In

tere

stra

te0.0

0−0

.25

0.2

30.0

30.3

30.3

41.7

71.7

7

(B)

Com

ple

men

tary

pol

icie

s(f

arm

ers)

Farm

-gat

ep

rice

−0.2

51.2

70.0

5−1

.73

−1.3

2−1

.97

13.9

527.5

7Q

uan

titi

es14.6

615.6

814.8

413.6

613.9

213.4

923.5

831.4

8T

ota

lu

tili

ty2.8

92.9

32.9

02.8

42.8

52.8

43.3

03.7

2T

ota

lin

du

stry

pro

fits

15.8

723.3

512.3

412.4

18.0

76.7

410.0

6−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts15.8

715.6

45.3

220.4

48.0

714.3

710.0

6−1

00.0

0In

tere

stra

te0.0

8−0

.15

0.3

10.0

90.3

90.3

91.7

71.7

7


Tab

leA

4.3

:(h

)C

onti

nu

ed. Le

ader

sR

esp

ect

mer

ge

Exit

Equ

alto

Lead

erSm

all

Lead

ers

and

smal

lof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

mer

ge

entr

ant

larg

est

shar

esco

mp

etit

ion

(C)

Com

ple

men

tary

pol

icie

s(fi

rms)

Farm

-gat

ep

rice

2.5

54.0

72.8

71.0

81.5

00.8

516.5

130.1

7Q

uan

titi

es1.5

02.3

91.6

70.6

30.8

60.4

89.2

116.1

9T

ota

lu

tili

ty0.0

70.1

10.0

70.0

30.0

40.0

20.4

30.8

0Pro

du

cers

’uti

lity

0.6

41.0

50.7

10.2

70.3

60.1

94.1

87.7

8N

on

pro

du

cers

’uti

lity

0.0

20.0

40.0

30.0

10.0

10.0

10.1

60.2

9T

ota

lin

du

stry

pro

fits

2.0

58.1

4−1

.17

−0.6

2−4

.55

−5.4

0−2

.48

−100.0

0A

vera

ge

ind

ust

ryp

rofi

ts2.0

51.3

8−7

.35

6.4

8−4

.55

1.3

6−2

.48

−100.0

0In

tere

stra

te0.1

8−0

.02

0.4

10.1

70.4

70.4

51.7

71.7

7

(D)

Com

ple

men

tary

pol

icie

s(f

arm

ers−

firm

s)Fa

rm-g

ate

pri

ce2.2

93.8

22.6

20.8

11.2

40.5

816.2

030.1

7Q

uan

titi

es16.3

317.3

316.5

315.3

515.6

215.1

824.9

332.9

1T

ota

lu

tili

ty2.9

63.0

12.9

72.9

22.9

32.9

13.3

73.7

9T

ota

lin

du

stry

pro

fits

18.3

124.9

914.4

915.4

810.8

510.1

214.1

2−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts18.3

117.1

77.3

323.7

310.8

517.9

814.1

2−1

00.0

0In

tere

stra

te0.2

50.0

70.4

90.2

30.5

30.5

01.7

71.7

7

(E)

Inte

rnati

onalpri

ces

Farm

-gat

ep

rice

22.3

924.0

822.8

620.7

321.3

220.5

036.5

353.3

1Q

uan

titi

es12.3

413.2

012.5

711.4

911.7

811.3

619.2

626.8

6T

ota

lu

tili

ty0.5

90.6

40.6

00.5

50.5

60.5

40.9

71.4

2Pro

du

cers

’uti

lity

5.7

96.2

35.8

95.3

65.4

95.2

89.4

613.9

1N

on

pro

du

cers

’uti

lity

0.2

10.2

30.2

20.2

00.2

00.1

90.3

50.5

1T

ota

lin

du

stry

pro

fits

30.0

634.0

925.1

429.3

823.6

224.9

230.9

3−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts30.0

625.7

117.3

138.6

223.6

233.8

430.9

3−1

00.0

0In

tere

stra

te0.7

20.6

90.9

80.6

00.9

20.7

91.7

71.7

7


Tab

leA

4.3

:(i)

Rw

an

da,c

offee

(per

cen

tage

chan

ges

).

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

Lead

ers

and

smal

lof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

mer

ge

entr

ant

larg

est

shar

esco

mp

etit

ion

(A)

Basi

cm

odel

Farm

-gat

ep

rice

0.0

04.4

50.5

7−5

.85

−4.2

5−6

.09

8.7

231.9

1Q

uan

titi

es0.0

04.0

10.4

6−5

.33

−4.0

0−5

.58

7.8

828.8

7T

ota

lu

tili

ty0.0

00.0

10.0

0−0

.02

−0.0

1−0

.02

0.0

30.1

1Pro

du

cers

’uti

lity

0.0

00.7

60.0

6−2

.05

−1.6

0−2

.15

1.5

46.2

6N

on

pro

du

cers

’uti

lity

0.0

00.0

10.0

00.0

00.0

00.0

00.0

10.0

6Sw

itch

ers’

uti

lity

0.0

019.8

320.0

0−1

6.6

6−1

6.6

6−1

6.6

419.8

220.3

7T

ota

lin

du

stry

pro

fits

0.0

0−4

.15

−4.0

14.1

8−4

.98

2.4

8−4

.67

−100.0

0A

vera

ge

ind

ust

ryp

rofi

ts0.0

0−4

.15

−20.0

030.2

2−4

.98

28.1

0−4

.67

−100.0

0In

tere

stra

te0.0

02.7

11.1

0−3

.95

−0.9

2−3

.68

4.3

44.3

4

(B)

Com

ple

men

tary

pol

icie

s(f

arm

ers)

Farm

-gat

ep

rice

−0.4

04.0

30.2

2−6

.30

−4.6

3−6

.53

8.2

831.9

1Q

uan

titi

es26.1

031.0

426.7

219.4

521.1

619.1

435.7

962.1

4T

ota

lu

tili

ty5.7

15.7

25.7

15.6

85.6

95.6

85.7

45.8

5T

ota

lin

du

stry

pro

fits

27.5

522.5

922.0

733.1

021.1

730.9

922.3

2−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts27.5

52.1

61.7

366.3

821.1

763.7

322.3

2−1

00.0

0In

tere

stra

te0.1

62.7

91.3

4−3

.81

−0.7

3−3

.55

4.3

44.3

4


Tab

leA

4.3

:(i)

Con

tin

ued

. Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

Lead

ers

and

smal

lof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

mer

ge

entr

ant

larg

est

shar

esco

mp

etit

ion

(C)

Com

ple

men

tary

pol

icie

s(fi

rms)

Farm

-gat

ep

rice

2.3

06.7

52.9

9−3

.62

−1.8

6−3

.85

10.8

434.5

7Q

uan

titi

es2.1

66.1

42.7

1−3

.20

−1.7

5−3

.44

9.8

431.2

1T

ota

lu

tili

ty0.0

10.0

20.0

1−0

.01

−0.0

1−0

.01

0.0

30.1

2Pro

du

cers

’uti

lity

0.4

31.2

00.5

2−1

.20

−0.7

0−1

.31

1.9

66.8

1N

on

pro

du

cers

’uti

lity

0.0

00.0

10.0

0−0

.02

−0.0

2−0

.02

0.0

20.0

6T

ota

lin

du

stry

pro

fits

3.0

7−1

.27

−1.8

08.0

0−2

.04

6.3

5−0

.66

−100.0

0A

vera

ge

ind

ust

ryp

rofi

ts3.0

7−1

7.7

2−1

8.1

735.0

0−2

.04

32.9

4−0

.66

−100.0

0In

tere

stra

te0.3

93.0

31.6

7−3

.59

−0.4

3−3

.34

4.3

44.3

4

(D)

Com

ple

men

tary

pol

icie

s(f

arm

ers−

firm

s)Fa

rm-g

ate

pri

ce1.8

96.3

72.6

3−4

.08

−2.2

5−4

.31

10.3

834.5

7Q

uan

titi

es28.7

533.7

129.5

022.0

723.9

521.7

738.2

165.0

1T

ota

lu

tili

ty5.7

25.7

35.7

25.6

95.7

05.6

95.7

55.8

6T

ota

lin

du

stry

pro

fits

31.4

825.7

624.9

137.9

724.9

335.9

227.4

1−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts31.4

84.8

04.0

972.4

624.9

369.9

027.4

1−1

00.0

0In

tere

stra

te0.5

43.1

91.8

9−3

.46

−0.2

5−3

.22

4.3

44.3

4

(E)

Inte

rnati

onalpri

ces

Farm

-gat

ep

rice

21.0

926.2

022.4

014.2

816.8

814.0

429.7

458.4

0Q

uan

titi

es19.4

623.8

220.4

813.5

815.6

713.3

526.9

651.5

6T

ota

lu

tili

ty0.0

70.0

90.0

80.0

50.0

60.0

50.1

00.2

1Pro

du

cers

’uti

lity

4.1

75.0

94.3

52.9

63.3

32.9

15.8

211.7

4N

on

pro

du

cers

’uti

lity

0.0

40.0

50.0

40.0

30.0

30.0

30.0

50.1

1T

ota

lin

du

stry

pro

fits

39.1

631.1

429.3

648.5

632.5

046.8

238.6

9−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts39.1

69.2

97.8

085.7

032.5

083.5

338.6

9−1

00.0

0In

tere

stra

te1.6

04.3

03.4

0−2

.49

1.0

7−2

.32

4.3

44.3

4


Tab

leA

4.3

:(j

)U

gan

da,c

offee

(per

cen

tage

chan

ges

).

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

Lead

ers

and

smal

lof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

mer

ge

entr

ant

larg

est

shar

esco

mp

etit

ion

(A)

Basi

cm

odel

Farm

-gat

ep

rice

0.0

00.9

60.1

4−0

.96

−0.7

6−1

.08

9.1

717.8

9Q

uan

titi

es0.0

00.7

80.1

0−0

.77

−0.6

2−0

.88

7.0

613.6

3T

ota

lu

tili

ty0.0

00.0

10.0

0−0

.01

−0.0

1−0

.01

0.0

80.1

7Pro

du

cers

’uti

lity

0.0

00.2

20.0

2−0

.37

−0.3

1−0

.43

1.9

43.9

2N

on

pro

du

cers

’uti

lity

0.0

00.0

00.0

00.0

00.0

00.0

00.0

40.0

7Sw

itch

ers’

uti

lity

0.0

019.1

419.2

2−1

5.9

8−1

5.8

5−1

5.8

719.0

919.2

6T

ota

lin

du

stry

pro

fits

0.0

03.9

5−2

.11

−1.6

7−4

.52

−4.7

2−1

4.5

0−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts0.0

0−2

.16

−7.8

74.8

9−4

.52

1.6

4−1

4.5

0−1

00.0

0In

tere

stra

te0.0

0−0

.12

0.1

5−0

.02

0.1

90.1

61.7

91.7

9

(B)

Com

ple

men

tary

pol

icie

s(f

arm

ers)

Farm

-gat

ep

rice

−0.1

80.7

9−0

.04

−1.1

5−0

.94

−1.2

78.9

617.8

9Q

uan

titi

es20.3

421.2

620.4

719.4

319.6

119.3

128.6

236.5

0T

ota

lu

tili

ty3.8

93.9

03.8

93.8

83.8

83.8

83.9

94.0

9T

ota

lin

du

stry

pro

fits

21.5

726.0

518.8

819.7

916.2

016.2

05.1

3−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts21.5

718.6

311.8

827.7

816.2

023.9

45.1

3−1

00.0

0In

tere

stra

te0.0

8−0

.02

0.2

40.0

50.2

60.2

21.7

91.7

9


Tab

leA

4.3

:(j

)C

onti

nu

ed. Le

ader

sR

esp

ect

mer

ge

Exit

Equ

alto

Lead

erSm

all

Lead

ers

and

smal

lof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

mer

ge

entr

ant

larg

est

shar

esco

mp

etit

ion

(C)

Com

ple

men

tary

pol

icie

s(fi

rms)

Farm

-gat

ep

rice

2.4

23.3

92.5

71.4

61.6

71.3

411.4

020.3

4Q

uan

titi

es1.9

12.6

82.0

21.1

51.3

11.0

58.7

615.4

4T

ota

lu

tili

ty0.0

20.0

30.0

20.0

10.0

20.0

10.1

00.1

9Pro

du

cers

’uti

lity

0.5

20.7

40.5

50.3

10.3

50.2

82.4

54.4

8N

on

pro

du

cers

’uti

lity

0.0

10.0

10.0

10.0

10.0

10.0

10.0

50.0

8T

ota

lin

du

stry

pro

fits

2.1

55.5

9−0

.23

0.9

3−2

.21

−1.9

8−1

0.9

2−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts2.1

5−0

.62

−6.1

07.6

6−2

.21

4.5

6−1

0.9

2−1

00.0

0In

tere

stra

te0.1

70.0

80.3

30.1

30.3

40.2

91.7

91.7

9

(D)

Com

ple

men

tary

pol

icie

s(f

arm

ers−

firm

s)Fa

rm-g

ate

pri

ce2.2

33.2

12.4

01.2

61.4

91.1

411.1

920.3

4Q

uan

titi

es22.5

823.4

922.7

221.6

821.8

821.5

630.6

138.6

2T

ota

lu

tili

ty3.9

23.9

33.9

23.9

13.9

13.9

14.0

24.1

2T

ota

lin

du

stry

pro

fits

24.2

428.1

021.2

223.0

019.0

719.5

89.4

9−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts24.2

420.5

714.0

931.2

019.0

727.5

59.4

9−1

00.0

0In

tere

stra

te0.2

40.1

70.4

10.1

90.4

10.3

41.7

91.7

9

(E)

Inte

rnati

onalpri

ces

Farm

-gat

ep

rice

21.8

822.9

622.1

320.7

921.1

220.6

731.0

041.9

5Q

uan

titi

es16.6

417.4

216.8

115.8

616.0

815.7

623.0

930.6

2T

ota

lu

tili

ty0.2

10.2

20.2

10.2

00.2

00.1

90.2

90.4

0Pro

du

cers

’uti

lity

4.8

95.1

44.9

44.6

54.7

14.6

26.9

49.4

8N

on

pro

du

cers

’uti

lity

0.0

90.1

00.0

90.0

90.0

90.0

90.1

30.1

8T

ota

lin

du

stry

pro

fits

32.7

334.4

628.5

933.3

228.2

830.4

823.3

9−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts32.7

326.5

521.0

342.2

128.2

839.1

823.3

9−1

00.0

0In

tere

stra

te0.7

00.7

20.8

90.5

70.8

10.6

81.7

91.7

9


Tab

leA

4.3

:(k

)M

ala

wi,

tobacc

o(p

erce

nta

ge

chan

ges

).

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

Lead

ers

and

smal

lof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

mer

ge

entr

ant

larg

est

shar

esco

mp

etit

ion

(A)

Basi

cm

odel

Farm

-gat

ep

rice

0.0

05.4

60.1

4−5

.12

−4.3

7−6

.00

21.4

346.1

2Q

uan

titi

es0.0

03.9

00.0

9−3

.78

−3.2

9−4

.55

14.3

629.9

4T

ota

lu

tili

ty0.0

00.0

80.0

0−0

.07

−0.0

7−0

.09

0.2

90.6

8Pro

du

cers

’uti

lity

0.0

01.2

10.0

2−1

.93

−1.7

2−2

.36

4.5

110.4

9N

on

pro

du

cers

’uti

lity

0.0

00.0

30.0

00.0

00.0

00.0

00.1

30.2

9Sw

itch

ers’

uti

lity

0.0

019.7

719.8

5−1

6.2

5−1

6.2

5−1

6.2

319.5

620.3

1T

ota

lin

du

stry

pro

fits

0.0

04.9

3−0

.89

−6.4

5−1

0.3

7−1

5.9

8−1

5.1

2−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts0.0

022.4

1−1

3.2

89.1

4−1

0.3

7−1

.98

−15.1

2−1

00.0

0In

tere

stra

te0.0

01.1

40.2

4−0

.79

0.5

41.0

110.2

78.6

5

(B)

Com

ple

men

tary

pol

icie

s(f

arm

ers)

Farm

-gat

ep

rice

−0.4

04.9

7−0

.22

−5.5

4−4

.74

−6.4

120.9

146.1

2Q

uan

titi

es18.4

422.9

018.5

714.0

314.6

413.1

535.0

053.3

3T

ota

lu

tili

ty3.9

03.9

93.9

13.8

23.8

33.8

04.2

44.7

0T

ota

lin

du

stry

pro

fits

18.8

625.2

217.5

211.5

66.6

70.4

52.3

0−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts18.8

646.0

92.8

330.1

56.6

717.1

92.3

0−1

00.0

0In

tere

stra

te0.3

51.3

50.6

4−0

.51

0.8

91.2

310.2

88.6

5


Tab

leA

4.3

:(k

)C

onti

nu

ed. Le

ader

sR

esp

ect

mer

ge

Exit

Equ

alto

Lead

erSm

all

Lead

ers

and

smal

lof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

mer

ge

entr

ant

larg

est

shar

esco

mp

etit

ion

(C)

Com

ple

men

tary

pol

icie

s(fi

rms)

Farm

-gat

ep

rice

2.3

07.5

12.5

3−2

.84

−1.9

8−3

.70

23.2

848.4

4Q

uan

titi

es1.6

45.3

41.7

9−2

.10

−1.5

4−2

.84

15.5

831.3

2T

ota

lu

tili

ty0.0

30.1

10.0

3−0

.04

−0.0

3−0

.06

0.3

20.7

1Pro

du

cers

’uti

lity

0.4

91.6

60.5

3−1

.09

−0.8

5−1

.51

4.9

511.0

4N

on

pro

du

cers

’uti

lity

0.0

10.0

50.0

2−0

.07

−0.0

7−0

.07

0.1

40.3

0T

ota

lin

du

stry

pro

fits

0.5

26.3

9−0

.95

−5.1

7−9

.60

−14.2

7−1

2.5

7−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts0.5

224.1

2−1

3.3

310.6

4−9

.60

0.0

2−1

2.5

7−1

00.0

0In

tere

stra

te0.8

11.6

11.1

7−0

.09

1.3

71.5

510.2

78.6

5

(D)

Com

ple

men

tary

pol

icie

s(f

arm

ers−

firm

s)Fa

rm-g

ate

pri

ce1.8

97.0

22.1

6−3

.27

−2.3

7−4

.12

22.7

548.4

4Q

uan

titi

es20.3

324.5

620.5

415.9

816.6

615.1

336.4

054.9

1T

ota

lu

tili

ty3.9

44.0

33.9

43.8

53.8

73.8

44.2

84.7

4T

ota

lin

du

stry

pro

fits

19.5

527.0

117.5

213.1

47.6

52.5

45.3

5−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts19.5

548.1

72.8

332.0

07.6

519.6

35.3

5−1

00.0

0In

tere

stra

te1.1

31.8

01.5

40.1

71.6

91.7

510.2

88.6

5

(E)

Inte

rnati

onalpri

ces

Farm

-gat

ep

rice

20.9

826.2

221.6

615.2

516.6

614.3

842.1

972.3

2Q

uan

titi

es14.4

817.9

014.8

910.7

011.5

510.0

427.4

844.7

5T

ota

lu

tili

ty0.3

10.3

80.3

10.2

20.2

40.2

10.6

11.0

9Pro

du

cers

’uti

lity

4.7

55.9

84.8

83.4

63.7

03.2

09.4

616.7

4N

on

pro

du

cers

’uti

lity

0.1

30.1

60.1

30.0

90.1

00.0

90.2

60.4

6T

ota

lin

du

stry

pro

fits

22.3

229.4

518.0

419.0

211.5

010.1

115.5

1−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts22.3

213.2

73.2

938.8

611.5

028.4

615.5

1−1

00.0

0In

tere

stra

te3.3

83.7

34.0

72.0

73.9

63.2

010.2

88.6

5


Tab

leA

4.3

:(l)

Zam

bia

,tob

acc

o(p

erce

nta

ge

chan

ges

).

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

Lead

ers

and

smal

lof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

mer

ge

entr

ant

larg

est

shar

esco

mp

etit

ion

(A)

Basi

cm

odel

Farm

-gat

ep

rice

0.0

09.6

52.2

1−7

.38

−3.4

4−1

3.9

021.4

664.4

5Q

uan

titi

es0.0

06.7

71.4

1−5

.03

−2.4

4−1

0.5

414.6

040.8

6T

ota

lu

tili

ty0.0

00.0

20.0

0−0

.01

−0.0

1−0

.03

0.0

50.1

7Pro

du

cers

’uti

lity

0.0

01.4

10.1

2−1

.82

−1.0

1−4

.54

3.1

911.1

2N

on

pro

du

cers

’uti

lity

0.0

00.0

10.0

00.0

00.0

00.0

00.0

30.0

9Sw

itch

ers’

uti

lity

0.0

017.2

517.1

6−1

4.7

2−1

4.7

3−1

4.7

717.0

717.7

7T

ota

lin

du

stry

pro

fits

0.0

07.4

4−8

.18

12.7

30.0

8−2

1.0

710.7

2−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts0.0

0−1

4.0

5−2

6.5

450.3

00.0

85.2

310.7

2−1

00.0

0In

tere

stra

te0.0

03.1

83.7

1−9

.50

−2.2

2−1

.90

7.5

27.5

2

(B)

Com

ple

men

tary

pol

icie

s(f

arm

ers)

Farm

-gat

ep

rice

−0.5

29.1

51.7

6−7

.95

−3.9

4−1

4.4

220.7

964.4

5Q

uan

titi

es16.9

424.7

418.6

211.0

914.1

44.7

933.6

264.1

0T

ota

lu

tili

ty1.5

01.5

31.5

11.4

91.4

91.4

61.5

61.7

0T

ota

lin

du

stry

pro

fits

17.8

826.2

78.0

832.8

917.8

5−6

.38

31.1

2−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts17.8

81.0

2−1

3.5

477.1

917.8

524.8

331.1

2−1

00.0

0In

tere

stra

te0.2

33.4

83.9

8−9

.25

−1.9

3−1

.86

7.5

27.5

2


Tab

leA

4.3

:(l)

Con

tin

ued

. Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

Lead

ers

and

smal

lof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

mer

ge

entr

ant

larg

est

shar

esco

mp

etit

ion

(C)

Com

ple

men

tary

pol

icie

s(fi

rms)

Farm

-gat

ep

rice

1.0

910.7

13.4

3−6

.38

−2.2

8−1

2.7

622.1

165.3

8Q

uan

titi

es0.7

87.4

92.2

6−4

.29

−1.6

1−9

.65

15.0

441.3

9T

ota

lu

tili

ty0.0

00.0

20.0

1−0

.01

0.0

0−0

.03

0.0

50.1

7Pro

du

cers

’uti

lity

0.1

51.5

60.2

9−1

.52

−0.6

6−4

.17

3.3

011.3

0N

on

pro

du

cers

’uti

lity

0.0

00.0

10.0

0−0

.01

−0.0

1−0

.01

0.0

30.0

9T

ota

lin

du

stry

pro

fits

0.5

47.2

6−8

.01

13.4

60.4

3−1

9.2

311.8

8−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts0.5

4−1

4.2

0−2

6.4

151.2

90.4

37.6

911.8

8−1

00.0

0In

tere

stra

te0.4

93.7

54.2

9−8

.91

−1.5

7−1

.79

7.5

27.5

2

(D)

Com

ple

men

tary

pol

icie

s(f

arm

ers−

firm

s)Fa

rm-g

ate

pri

ce0.5

610.2

02.9

7−6

.96

−2.7

9−1

3.2

921.4

365.3

8Q

uan

titi

es17.8

325.5

619.5

911.9

415.1

05.8

234.1

264.7

0T

ota

lu

tili

ty1.5

01.5

31.5

11.4

91.5

01.4

71.5

61.7

0T

ota

lin

du

stry

pro

fits

18.5

526.0

98.3

133.7

818.3

0−4

.20

32.4

8−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts18.5

50.8

7−1

3.3

578.3

718.3

027.7

332.4

8−1

00.0

0In

tere

stra

te0.7

04.0

34.5

4−8

.68

−1.2

9−1

.76

7.5

27.5

2

(E)

Inte

rnati

onalpri

ces

Farm

-gat

ep

rice

14.8

325.3

217.9

66.2

011.4

2−0

.24

36.0

785.5

7Q

uan

titi

es10.4

417.2

712.3

34.9

28.1

7−0

.10

24.0

452.2

9T

ota

lu

tili

ty0.0

40.0

60.0

40.0

20.0

30.0

00.0

90.2

3Pro

du

cers

’uti

lity

2.3

83.9

92.6

51.4

91.8

90.0

35.8

215.1

7N

on

pro

du

cers

’uti

lity

0.0

20.0

30.0

20.0

10.0

2−0

.11

0.0

50.1

2T

ota

lin

du

stry

pro

fits

21.9

126.5

510.1

638.0

420.9

74.2

538.0

0−1

00.0

0A

vera

ge

ind

ust

ryp

rofi

ts21.9

11.2

4−1

1.8

784.0

620.9

739.0

038.0

0−1

00.0

0In

tere

stra

te2.1

75.7

56.2

7−6

.91

0.6

7−1

.46

7.5

27.5

2

5

Supply Chain Simulations and PovertyAnalysis in Sub-Saharan Africa

In this chapter, we estimate the impact on household income, at the farm level,of changes in the supply chain. In the previous chapter, using our theoreticalmodel, we identified the farm-gate price changes generated by shocks to thelevel of competition in the value chain for our twelve case studies. We simu-lated seven alternative market configurations for the baseline model and wealso allowed for complementary factors affecting farms, firms, or both andalso for shocks to the international price of the crop. We also ran the sameset of simulations under the extended model with outgrower contracts. Inthe end, we have seventy simulations with seventy corresponding changes infarm-gate prices, the main variable of interest for our analysis.

We now want to use these price changes and the household survey datadescribed in Chapter 2 to carry out a comprehensive analysis of the impactsof changes in value chains on poverty and welfare. Using standard methods toapproximate welfare changes with first-order effects (see Section 1), we esti-mate the impact on average household income for the total population as wellas for the subset of export crop producers. Furthermore, the household dataallows us to differentiate the effect among poor and nonpoor households, adistinction that will help us to understand under which circumstances com-mercial agriculture can work as an effective vehicle for poverty alleviation. Wealso explore gender issues by looking at results for male-headed and female-headed households and advance some explanations for any potential differ-ential impacts.

In Section 2, we present the welfare implications of our simulations for eachof the twelve country–crop case studies. Rather than providing a detailed dis-cussion of the seventy simulations per case study, we focus, as in Chapter 4,on the case of cotton in Zambia. We then summarize the main findings anddiscrepancies for the other eleven cases grouped by crop to facilitate the com-parison. We present all the tables with the simulation results for the interestedreader.

1 THE METHODOLOGY

Our task is to estimate the welfare effects of the changes in farm-gate pricesand in input costs due to complementary policies or changes in the conditions


under which outgrower contracts are implemented. We adopt the standardfirst-order approach advanced by Deaton (1989).1

1.1 Calculation of Income Changes without Outgrower Contracts

To derive the formulas needed in the analysis, we start from the income–expenditure equation. This equation indicates that, in equilibrium, expendi-tures need to be covered with income (we can allow for transfers, savings, andso on). Suppose for simplicity that the farmer produces only two crops, theexportable crop q1 and the subsistence crop q2. Then we can write:

p · c + r1v1 = p1q1 + p2q2 + x0. (5.1)

In (5.1), r1v1 is the “expenditure” in investment in sector 1 (we could alsoinclude a similar term for the second crop, but we do not need it for theanalysis). The term r1v1 includes expenditures in seeds, fertilizer, pesticides,and also interest payments on loans. The term p · c is total expenditure ingoods and services. Finally, p1q1 and p2q2 are gross income from sales ofproduct 1 and 2, respectively, and x0 is an exogenous source of income.

We are first interested in studying the first-order effect assuming nochanges in production costs. The welfare effect of a price change is definedas −dx0/y , where y = p1q1 + p2q2. Assume that there is an increase in p1,keeping v1 and p2 constant for the moment. We then have:

−dx0

y= s1 d lnp1.

This means that the proportional change in income dy/y is the product ofthe income share s1 and the proportional change in prices (these are the pricechanges from the different simulations in the previous chapter). For example,if a household earns 50 percent of its income from cotton and the price ofcotton increases by 10 percent, then the impact effect for the household wouldbe equivalent to 5 percent of its initial income.

Assume now that the change in farm-gate price comes along with a changein the input cost due, for example, to complementary policies. Now, we have

−dx0

y= s1 d lnp1 − r1v1

ydr1

r1.

One practical problem we face is that we do not observe input expenditures inthe data. So, we will assume that the input expenditure is a constant fraction

1This approach has been extensively utilized in the literature. Early examples includeDeaton (1989b), Budd (1993), Benjamin and Deaton (1993), Barret and Dorosh (1996), andSahn and Sarris (1991). More recent examples include Ivanic and Martin (2008) and Wodonet al. (2008). Deaton (1997) provides an account of the early use of these techniques indistributional analysis of pricing policies.

Supply Chain Simulations and Poverty Analysis 165

of the total gross sales of the product r1v1 = δ∗p1q1. In this case, we havethat

−dx0

y= s1 d lnp1 − δs1 d ln r1.

In practical terms, when simulating the impact of changes in the value chaintogether with complementary factors that affect farmers, we use d lnp1 calcu-lated from the simulations, and we let d ln r1 = −0.02 (a decline of 2 percentin production costs). We assume δ = 0.5 in our numeric simulations.

There is an important caveat to this approach. Our framework works wellfor things like seeds, fertilizer, etc., but not for labor. Suppose that theincrease in competition for output also increases wages in the sector. Arethere additional welfare effects? The answer depends on how farmers allo-cate their labor supply and how we measure welfare. Suppose the farmer onlyworks on her farm. If wages increase, we have a welfare effect because nowshe earns more money on wages, but this is just a cost of production in ourmodel. In our analysis, we will not deal with these effects.

1.2 Calculation of Income Changes with Outgrower Contracts

The model is the same as above:

p · c + r1v1 = p1q1 + p2q2 + x0.

The difference is that now when there is a change in the supply chains, thereare effects on output prices p1 and also on the interest rate charged on inputs.Input expenditures are r1v1, which, in turn, we assume equals a fraction δ ofsales p1q1

r1v1 = δp1q1.As in the theoretical model, we will now assume that farmers finance a fractionλ of their expenditures in inputs with outgrower contracts. This means thatthe amount being financed is

λr1v1 = λδp1q1.

The farmer needs to pay interest on this equal to r∗λδp1q1. Hence,

p · c + r1v1 + r∗λr1v1 = p1q1 + p2q2 + x0.

To calculate the welfare effects, allow changes in p1 and in r

−dx0

y= s1 d lnp1 − δλs1 d ln r .

As before, we also want to consider the case of changes in farm-gate pricesthat come along with a change in the input cost due, for example, to comple-mentary policies. Taking this and the outgrower scheme into consideration,we have

−dx0

y= s1 d lnp1 − δλs1 d ln r − δs1 d ln r1.

Once again we assume δ = 0.5 for our calculations.


2 WELFARE SIMULATIONS

Using the price changes generated by each of the seventy simulations of Chap-ter 4 and the methodology described in the previous section, here we studythe effects of changes in the value chain on farmers’ income for our twelvecase studies. We explore the Zambian cotton case in detail and then we sum-marize the major findings from the remaining eleven case studies.

2.1 Cotton in Zambia

Most of the cotton seeds in Zambia are devoted to the exports of cotton lint.Atomistic farmers produce cotton which is purchased by the ginneries to pro-duce cotton lint that later is exported to world markets. While two ginneriescontrol 72 percent of the market and therefore can exercise monopsonisticpower over farmers, their share in the world market is insignificant and con-sequently take the international price as given.

Baseline Model

Table A5.1(a) presents the simulation results for changes in household incomefor each of the seven market configurations and five scenarios in our leadingcase. The outcomes are also presented for different rural population groups.The exercise in the previous chapter showed that the change in farm-gate cot-ton prices ranged from −12 percent (in the case of exit of the largest firmunder complementary policies for the farmers) to 104 percent (in the case ofperfect competition under an increase of 10 percent in international prices).The overall impact of these prices changes on average household incomedepends on the share of cotton on total household income. In Chapter 2 weshowed that most households in the survey do not produce cotton, or, whenthey do, in general they do not specialize in its production. For the averagerural household in Zambia, cotton generates less than 3 percent of its totalincome. Among producers, the cotton share in income increases to 23 percent.

The main conclusion from the simulations is that, in our baseline model,competition among ginneries is good for the cotton farmers because theyfetch a higher farm-gate price and therefore enjoy a higher level of income.For example, if Dunavant (the leader firm) splits, the increase in income for theaverage producer would be equivalent to 2.4 percent of its initial income. Onthe other hand, if the two largest firms (Dunavant and Cargill) were to merge,the income of the average producer would decline by 2.3 percent. The largestpossible gain for the farmers comes under perfect competition, where farmerswould enjoy an income gain of 19.3 percent. The upper bound increase inincome under imperfect competition is 7.3 percent, and this takes place inthe equal market share simulation. Another evident conclusion from our basicmodel is that small changes in the level of competition among ginneries arenot likely to generate important impacts on farmers’ income. For instance, a


small firm entering the market would generate only an increase of one quarterof a percentage point in producers’ income.

One concern often encountered in practice is to understand the implica-tions of exit, in particular of the largest firm. The exit of Dunavant would implya reduction in competition among the remaining firms that would impact neg-atively in the farm-gate price for cotton in Zambia. In addition, in our model,the largest firm is also the most efficient one (smallest marginal cost) so theexit would imply a reduction in the total demand for cotton, further depress-ing the farm-gate price. In our basic simulation, this is the worst scenario forproducers, with an average income loss of 3.2 percent.

We also estimate the income effect under different complementary poli-cies.2 Figure A5.1 shows the income effect for producers in Zambia under thefive different scenarios when we increase competition (leader split) and whenwe reduce it (leaders merge). The implementation of complementary policiesand the positive international price shock intensify the positive effects ofmore competition and moderate the negative effects of a reduction in thelevel of competition among ginneries. The original increase of 2.4 percent inincome for producers following the split of the leader becomes 2.8 percentunder complementary policies for farmers, 3.1 percent when these policiesaffect firms, and 3.5 percent when they affect both farmers and firms. If thesplit of the leader takes place concurrently with an increase of 10 percent ininternational prices, the average producer earns 8.7 percent more income. Onthe other hand, if the leaders merge, a complementary policy affecting bothfarmers and firms will cut the income loss for producer households from2.32 percent to 1.25 percent.

It should be noted that we are estimating only the first-order effects ofthe price changes and, in consequence, only farmers that were initially pro-ducers are affected. The nonproducers are in fact isolated from the changesin the supply chain, meaning both that they do not enjoy the benefits ofincreased competition, if any, or the losses from higher oligopsony power.In Table A5.1(a), we report the income changes for households that producesome cotton versus the whole population of rural households. Figure A5.2illustrates the difference in income impacts for the two groups for differentshocks to the level of competition for our basic model. For instance, in thecase of equal market shares, producers would enjoy a gain of 7.3 percentwhile the gain for the whole rural population is only 0.8 percent. The qualita-tive results are the same for the complementary policies and increase in theinternational price scenarios.

Nonproducers are not affected because we are not incorporating estimatesof second-order effects. The main reason to do this is that we do not have

2The literature on complementary policies to trade shocks includes Deininger andOlinto (2000), Eswaran and Kotwal (1986), Balat and Porto (2007), Key et al. (2000), andMcKay et al. (1997).


a model to estimate those effects that can be convincingly utilized with sub-Saharan data. Estimates of second-order effects require estimates of supplyresponses, which in turn require some evidence on farm supply elasticities.Even if these elasticities were available, the estimated second-order welfareimpacts would nevertheless be small because, in the margin, the returns todifferent economic activities are equalized. This may not necessarily be thecase in the presence of distortions or market imperfection that generate awedge between the marginal return to factors allocated to export crops andto subsistence crops. This type of effect can be seen in the model of Chap-ter 4, because there is discrete increase in utility for those farmers that switchactivities and adopt export crops when prices increase. But, as we also showedin Chapter 4, these welfare effects are very small, on average. This is mostlybecause initial farmer participation in the export supply chain is very limitedand thus the majority of households are nonproducers. In consequence, evenif the switchers enjoy sizeable gains, there are only a few of them in any givensimulation. In the end, these gains are averaged out across many nonpartic-ipants, thereby creating negligible welfare effects. In short, the addition ofthose supply responses is unlikely to affect our welfare and poverty analysis.This feature of the analysis is a general result, not a property of our data (see,for example, Cadot et al. 2009; McMillan et al. 2003; Heltberg and Tarp 2002;Key et al. 2000; Lopez et al. 1995).

With the household survey data, we can also distinguish differential effectsfor poor and nonpoor rural households. Given a farm-gate price change, theresults among the two groups will depend entirely on the initial income inci-dence of cotton across groups of households. Our microdata shows that,among cotton producers, cotton is relatively more important for poor thanfor nonpoor households. However, for the whole rural population, the oppo-site happens. Parts (a) and (b) of Figure A5.3 show these two results. Forinstance, in panel (a), an increase in competition represented by the split ofthe leader increases the income of poor producers by 2.6 percent, and of non-poor producers by 2.3 percent. For the sample of all households (panel (b)),the increase in the income of the average poor household is only 0.23 percentand the increase in the average of the nonpoor is 0.29 percent. Once again,we do not discuss the differential impact for the poor and the nonpoor acrossdifferent market and policy configurations because the result is proportionalto the change in price and this change is the same for all households. Thisis a limitation of the model (it could well be the case that a particular com-plementary policy has a differentiated effect on farm-gate prices received bypoor versus nonpoor farmers because different access to markets, informa-tion, technologies, or inputs) that is partially driven by the restriction imposedby the available data.

An important result to discuss is the presence of gender-specific impacts,that is, differential impacts for male-headed and female-headed households.As before, since our theoretical model delivers a common price change that


applies to all producers, the differences in the poverty impacts will be drivenby the share of cotton in total income across households. Figures A5.4(a)and A5.4(b) display the effects of different shocks to the level of competi-tion in the ginning sector among male-headed and female-headed householdsboth for the sample of producers and for the total population. For producers,the share of income among male-headed and female-headed households issimilar and therefore the results of the simulations do not differ significantlyacross genders. In the case of equal market shares, the average income of amale-headed producer household increases by 7.36 percent, while it increasesby 7.17 percent in the case of the average female-headed producer household.If we consider all households, the income effects are negligible, but the differ-ence between male-headed and female-headed households is proportionallylarger than in the case of the producers’ subsample. For the equal marketshare simulation, male-headed household income increases on average by0.87 percent, while it only increases by 0.57 percent for the female-headedcounterparts. It should be mentioned that we are not considering second-order effects, nor are we allowing the complementary policies to have a dif-ferent impact based on gender considerations.3

Outgrower Contracts

In the previous analysis we assumed that farmers have access to workingcapital and that the structure of the market does not affect the cost of thoseinputs. However, in the absence of enforcement mechanisms, processors maybe reticent to advance the inputs needed for production, or, if they do, chargea premium to compensate for the possibility that the contracts are not hon-ored. In our analysis, we have assumed that the borrowing cost for the farm-ers increases with the level of competition. This modification to the basicmodel does not generate important effects in the equilibrium farm-gate pricebut it changes the production cost for the farmers and therefore affectstheir income. The results from the simulations are reported in parts (a)–(l)of Table A5.1.

3This is a simplification of the model, which, once again, it is driven by data constraints.Note that the literature points out several constraints that particularly affect female farm-ers and their ability to improve yield, profit, and efficiency in agriculture production. Someof these constraints are women’s legal and cultural status, which affects the degree ofcontrol women have over productive resources, inputs, and the benefits which flow fromthem (Olawoye 1989); property rights and inheritance laws, which govern access to anduse of land and other natural resources (Jiggins 1989a); the relationship among ecologicalfactors such as the seasonality of rainfall and availability of fuelwood, economic factorssuch as product market failures, and gender-determined responsibilities such as feedingthe family, which trade off basic household self-provisioning goals and care of the fam-ily against production for the market (Jiggins 1989b; Horenstein 1989); and the way thatagricultural services are staffed, managed, and designed (FAO 1993; Saito and Weidemann1990; Gittinger et al. 1990). Given these constraints, changes in the level of competitionand complementary policies may have different effects among female farmers.


Figure A5.5 illustrates the effects on producing farmers’ income of the intro-duction of outgrower contracts and liquidity constraints. We plot the changein average income for cotton producers due to changes in market configura-tion in the models with and without outgrower contracts. Despite the fact thatthe differences in levels seems to be minor, the percentage changes amongthe two models are economically important. All the simulations where mar-ket competition increases show lower gains for farmers in the model withoutgrower contracts. These gains are reduced to 5.8 percent in the leadersplit case, 11.9 percent in the equal market shares, and to almost half in thesmall entrant simulation. On the other hand, in the simulations generatingmarket concentration, the losses under outgrower contracts are smaller dueto a reduction in the borrowing costs (for instance, 18.5 percent lower in theleaders merge simulation). The exit of the largest firm is an interesting caseas it reduces market competition but nevertheless increases the borrowingcosts for the farmers and their income falls further.

In the simulations that we implemented above, the presence of outgrowercontracts affects the magnitude of the impacts, but it does not affect the sign.In principle, however, it could happen that an increase in competition breaksdown the whole contractual agreement, leading to a collapse of the market.The case of the cotton sector in Zambia in the early 2000s is an exampleof this type of effect (see, for example, Brambilla and Porto 2011) and theseimplications should thus be taken into account when designing competitionpolicies in the sub-Saharan cash-crop sector.

Now we move to analyze the effects of outgrower contracts under the differ-ent scenarios of complementary policies and exogenous increase in interna-tional prices. Figure A5.6(a) shows the average income change for producersin the event of an increase in competition (leader split) in the basic model,the three scenarios with complementary policy, and the increase of 10 per-cent in the international price of cotton for the model with and without out-grower contracts. Figure A5.6(b) presents the same simulations for the case ofa reduction in the level of competition (leaders merge). When the leader splits,the results under outgrower contracts are qualitatively the same but the levelof income gains is lower. This is due to the additional cost for the farmers ofa higher interest rate. This additional cost is increasing in the size of the gaingenerated by the complementary policy. While the interest rate paid by thefarmers is 2.13 percent higher in the case of complementary policies support-ing farmers, it increases to 3.18 percent in the case of policies assisting bothfarmers and firms. Despite having the highest interest rate, this last policystill maximizes the gains for the household producing cotton in Zambia. Theaverage income effect is 3.13 percent, while, in the model without outgrowercontracts, it was 3.48 percent, which implies a reduction of 10 percent. Inthe case of leaders merging, the reduction in competition reduces the inter-est costs for farmers, partially compensating the reduction of income dueto a lower farm-gate price. This offsetting factor varies across the different


complementary policies. The average income losses are 20.7 percent lowerwhen complementary policies focus on farmers, while they are only 15.3 per-cent lower when the policies center on the firms. In the case of an increase ininternational prices, the existence of outgrower contracts reduces the averageincome gain for farmers at about the same order of magnitude for both anincrease and a reduction in the competition level among ginneries.

2.2 Other Case Studies

Cotton

Besides our lead case of Zambia, we study the effects of competition amongcotton processors on farm-gate prices and household income for another foursub-Saharan African countries. For Benin, Burkina Faso, and Côte d’Ivoire werun the same set of simulation we had for Zambia. In the case of Malawi, wefollow a different approach as the country currently has two ginneries eachcontrolling 50 percent of the market. For that reason some of our simulationsdo not make much sense and we decided to study the effect of splitting themarket among three and later four firms, and we also allow for the entrance ofa small firm. We apply the model both with and without outgrower contractsand we study the effect of complementary policies and exogenous variationin the international price of cotton.

Qualitatively, we found the same result that we found in the Zambian case:competition is good for the farmers. The income effects, as expected, aremuch larger for those households producing the crop than for the typicalrural household that may or may not produce cotton (five times larger inBenin and more than forty times in the case of Burkina Faso). For that reason,we discuss mainly the result of the income simulation for cotton producers.The income effect depends both on the magnitude of the price change andon the importance of cotton in the total income of the average producer-household. Figure A5.7(a) displays the effect of increased competition (leadersplit) under the five different scenarios. The income effects are, on average,3.3 percent in Benin, 4.6 percent in Côte d’Ivoire, and 12.4 percent in Burk-ina Faso (all these impacts are larger than in the case of Zambia, which wasslightly over 2 percent). In the Burkina Faso case, part of the result is dueto the fact that the leader, SOFITEX, controls 85 percent of the market. Inthe other countries, the differences in income impact are mainly driven byincome shares as the farm-gate price changes when the leader splits is of thesame order of magnitude (around 9 percent) for the three countries. In thecase of Malawi, moving from two to three firms in the market causes incomegains of only 1.9 percent, while moving to four firms causes income gains of3 percent. Complementary policies also have a positive impact on farmers’income, with larger impacts in cases where those policies affect both farmersand firms. For instance, the income gains with both types of complementarypolicies are 12 percent in Burkina Faso, 34 percent in Benin, and 47 percent in


Côte d’Ivoire—higher than under the baseline scenario. In Figure A5.7(b) weshow the income effects of a merger between the two leading companies. Thistype of competition simulations brings about welfare losses for the farmers.The largest losses are observed in Côte d’Ivoire, where the average producerhousehold would lose 3.9 percent of income in comparison with 2.3 percentin Zambia, 1.8 in Burkina Faso, and only 0.7 in Benin. On the positive side, theimplementation of complementary policies can reduce these negative impactsand sometimes even can attain an overall positive effect, as in the case of Burk-ina Faso and Benin. In the case of Burkina Faso this only happens when wecombine complementary policies supporting firms and farmers. Instead, inthe case of Benin, it would be enough with firms’ oriented complementarypolicies to obtain a positive (but negligible) effect on farmers’ income underthis scenario. The overall effect of the complementary policies on farmers’welfare depends on the combined farm-gate price and quantity effect. Typi-cally, farmer support would depress the farm-gate price but would increasethe supply of crops while complementary policies supporting firms wouldhave a positive impact in farm-gate price but a minor impact in the overallsupply from farmers. Policies affecting both farmers and firms would gener-ate a price effect that lies in between the previous two cases and a quantityeffect that would be higher than the combination of the two individual supplyeffects.

The analysis of the impact on poor versus nonpoor households shows adifferent pattern than in the case of Zambia. As we can see in Figure A5.8nonpoor households benefit more than poor household from an increase incompetition (equal market shares in this case) in Benin, Burkina Faso, andCôte d’Ivoire. (This also happens in Malawi, not shown in the graph). The dif-ferences, however, are very small. For instance, in our baseline case, underequal market shares, nonpoor households in Burkina Faso will earn 21.4 per-cent more than under the actual market configuration, while poor householdswould see their income rise by 20.1 percent.

Turning now to gender issues, we find that male-headed households benefitrelatively more than female-headed households in Burkina Faso and Malawi,while the opposite happens in Benin and Côte d’Ivoire. The gender differencesare not trivial. For example, in the case of perfect competition in Burkina Faso,male-headed households would earn 50.4 percent more, while female-headedhouseholds would earn 15.9 more (see Figure A5.9). On the contrary, female-headed households in Benin would benefit from a 34.7 percent increase inincome increases versus a 22 percent for the average male-headed household.

The last issue we want to discuss for the cotton sector is the analysis of theplausible negative effects of competition when we introduce the need of out-grower contracts that may not be perfectly enforceable. The cases of Benin,Côte d’Ivoire, and Malawi are similar to the Zambia case. In general, competi-tion is still good in our calibrations but the benefits for farmers are slightlyoffset by increasing borrowing cost. On the other hand, the case of Burkina


Faso merits a thorough discussion. Figure A5.10 displays the seven shocksto the level of competition in cotton processors in Burkina Faso for the basicmodel. Contrary to what we found in the other cotton case studies, the bene-fits of tighter competition are greatly offset by the increasing costs of fundsin the model with outgrower contracts. While farm-gate prices and quantitieschanges are about the same in the model with and without outgrower con-tracts, the interest rate greatly increases in the model with outgrower con-tracts. For example, a leader split situation would generate an increase inthe interest rate of 39.3 percent in Burkina Faso but only 1.8 percent in thecase of Zambia. That would reduce gains for producing households in Burk-ina Faso from 12.4 percent to 6.5 percent in the baseline case with outgrowercontracts. Figures A5.11(a) and 5.11b show that the gap between the incomeeffects in the two models does not shrink when we bring in the possibilityof complementary policy. The 40 percent difference in income gains betweenthe two models in the baseline case slightly increases to 41 percent when weintroduce complementary variations aiming to reduce costs for both firmsand farmers. If the two leading companies were to merge in the model with-out outgrower contracts, households would lose 1.8 percent of their incomebut would actually win an extra 1.3 percent in the model with outgrower con-tracts. This is a case where the reduction in competition generates savingsin borrowing costs that are larger than the reduction in the price paid to thefarmers. Another interesting feature of this simulation, and a difference withthe previous cases, is that, with outgrower contracts, the complementary pol-icy applied only to firms results in a worse outcome for farmers than the oneapplied only to the farmers.

Cocoa

We now study the effects of competition in farm-gate prices and rural house-hold income in the two largest cocoa producers, Côte d’Ivoire and Ghana.Like our previous cases, increase in competition among exporting companiesbenefits producers and a decrease in competition reduces the price paid tofarmers. Figure A5.12(a) represents this last case, where a reduction in com-petition due to the exit of the largest firm decreases the income received byproducing households in both countries. However, the effect is small in com-parison with other case studies, with the typical Ivorian producing householdlosing 1.3 percent of its income and the Ghanaian counterpart receiving only1 percent less. This result is both a combination of prices not varying muchdue to a relatively low market power concentration and a moderate share ofcocoa in income due to producers’ diversification. If the reduction in com-petition came along with the implementation of complementary policies, thereductions in income are partially compensated or even reverted in the caseof policies benefiting both firms and farmers.

Figure A5.12(b) shows the effects of an increase in competition where wedisplay the case of equal market shares. This would lead to an increase in


farm-gate prices of 18.2 percent in Côte d’Ivoire and 29.3 percent in Ghana.Despite the significant price differential in favor of Ghanaian producers, theoverall impact on income is larger in Côte d’Ivoire because the average ruralhousehold cocoa income share is around twice that of Ghana. In our standardscenario, equal market shares would lead to an increase in income of 10.5 per-cent in Côte d’Ivoire and 7.3 percent in Ghana. Once again, if combined withcomplementary policies, this increase would be larger with households receiv-ing an extra 2.1 percent and 1 percent of income in Côte d’Ivoire and Ghana,respectively.

An increase in the international price of cocoa (Figure A5.12(c)) wouldhave a larger effect on farmers’ income in Côte d’Ivoire than in Ghana. Onceagain, this result is mainly driven by the differences in cocoa income sharesamong the two countries. In the current market configuration, producinghouseholds in Côte d’Ivoire would see their income increase by 12.7 percentagainst 6.2 percent in Ghana. This difference varies, however, with the levelof competition. Increases in competition would make the income gains differ-ence between the two countries larger, while reduction in competition wouldreduce the gap.

The income impact of more or less competition among exporters on poorand nonpoor cocoa-producing households (Figure A5.13) is about the same inboth countries. In the simulation of perfect competition in our baseline model,poor Ivorian households would gain 21.6 percent while nonpoor householdswould witness an increase in income of 21.3 percent. On the other hand, inGhana, nonpoor household are the ones that benefit the most. However, theincome gain is only marginally higher than for poor households. In both coun-tries, on average, male-headed households benefit more than female-headedhouseholds from an increase in competition (Figure A5.14 shows the case ofleader split). The gender difference is slightly higher for Côte d’Ivoire.

The introduction of outgrower contracts does not generate importantincome effects in cocoa like in the previous case of cotton in Burkina Faso.In both countries the effects are modest, even in the extreme cases of equalmarket shares and perfect competition. The only significant difference withthe cotton cases reviewed above is that, in the case of cocoa in both Ghanaand Côte d’Ivoire, when the leader splits, the interest rate decreases insteadof increasing and, therefore, households enjoy a larger income effect in themodel that has outgrower contracts. This difference is important in the caseof Ghana for the basic model, but the differential shrinks when we incorporatecomplementary policies (Figure A5.15). On the other hand, an increase in inter-national prices generates higher income effects for rural producing house-holds in the model without outgrower contracts than in the model with them.

Coffee

Increasing competition benefits coffee smallholder producers in Côte d’Ivoire,Rwanda, and Uganda. In Figure A5.16(a) we show the result of a leader split in


the income of the average producing household. The effects are modest forthe three countries, with a larger effect in Côte d’Ivoire and Rwanda than inUganda. The increase is larger when accompanied with complementary poli-cies and, as is the case in most simulations, policies focusing on firms have ahigher impact than the ones supporting the farmers.

Figure A5.16(b) shows the simulation for the case of a reduction of compe-tition due to the merging of the two leading firms. The farmers’ income effectsare again modest, with the most negative effect taking place in Rwanda, wherethe average producing household loses slightly more than half a percentagepoint of their income.

Another issue we are interested in studying is the effect of an increasein the international price of coffee. Figure A5.16(c) displays the effects ofa 10 percent increase in the international price on the income received bythe producing households. In the current market configuration, farmers inCôte d’Ivoire would receive 6.5 percent more, while in Rwanda and Ugandathe increase would only be around 2 percent. Increasing competition amongexporters would drive these income gains up, but only modestly for Rwandaand Uganda. Côte d’Ivoire producers would get 12 percent more income if theincrease in international prices takes place in a market where all firms havethe same market share.

Figures A5.17 and A5.18 show the income effects of more competitionamong processors on poor and nonpoor households (equal market share sim-ulation) and on male-headed and female-headed households (perfect compe-tition simulation). In Côte d’Ivoire and Rwanda, poor households benefit moreon average than nonpoor households. These differences are significant since,in the first case, the income gain is 57.7 percent higher, and in the second caseit is 43.3 percent higher. On the other hand, the effect is about the same forpoor and nonpoor households in Uganda. Male-headed households benefit onaverage more in Côte d’Ivoire and Uganda, while female-headed householdbenefit more on average in Rwanda. Once again, the differences are signif-icant as, for instance, in Côte d’Ivoire, the male-headed household incomeincrease in the case of perfect competition is 67 percent larger than for thefemale-headed households.

The effects of outgrower contracts are similar to what was discussed aboveand therefore we do not report the simulations here. The interested readercan check the specific results in parts (h)–(j) of Table A5.1.

Tobacco

Tobacco is the last crop in our study and we perform the analysis for thecases of Malawi and Zambia. As before, we find positive effects of competition.We illustrate this with two of the multiple simulations. In Figure A5.19(a) wepresent the “exit of the largest” for our basic model and the different scenar-ios. The negative impact of the loss of competition is felt strongest by farmersin Zambia, where the average producing household loses 4.7 percent of its


income in the baseline case. This is almost three times the effect in Malawiand is mostly due to the price effect. The largest firm in tobacco in Zambiacontrols almost half of the market while in Malawi the leading firm controlsone-third of it. The introduction of complementary policies at the firm andthe farm levels somehow reduces this negative impact but never turns it pos-itive, as was the case in some of the previous cases. Figure A5.19(b) showsthe case of equal market shares. This increase in competition generates size-able increases in income in both countries. Producing households in Malawiobtain on average 5.7 percent more income, while in Zambia the increment is7.2 percent. Complementary policies would enhance these income gains butthe increase is not as substantial as it was in other cases.

Nonpoor households profit more than poor households from increases incompetition among domestic tobacco buyers. Figure A5.20 presents the caseof perfect competition. In this scenario, nonpoor farmers would gain 41 per-cent and 47 percent more in Malawi and Zambia, respectively, than poor pro-ducing households in the same countries. The income effect is larger in bothcountries for male-headed household. In Figure A5.21 we show the effectsupon the income of households of the leader splitting simulation in the base-line scenarios for both genders. As can be seen from the figure, the genderdifference is larger in Zambia, where male-headed households would profitfrom a 34 percent extra income increase, in comparison with female-headedhouseholds.

In both case studies for tobacco, the incorporation of outgrower contractsinto the model reduces the farmers’ income gains from further competitionamong exporters. Figure A5.22 illustrates the results of leader split with andwithout outgrower contracts in Malawi. In the basic model, the improvementin producing households’ income is 44.8 percent lower in the model with out-grower contracts. This difference is reduced to 29.4 percent when we incor-porate complementary policies affecting both farmers and firms.

3 SUMMARY OF FINDINGS

In this chapter we studied the effects on the income of households of increas-ing competition among processors in twelve case studies covering four cashcrops in eight sub-Saharan African countries. The main conclusion of theanalysis is that competition among processors is good for farmers as itincreases the farm-gate price of the crop. Take, for instance, the case wherethe firm with the largest market share splits. This would lead to an averageincome increase for producing households of 2.8 percent in our case stud-ies. This average, however, masks a great variability, with cotton-producinghouseholds as the top earners and the smallholders in the coffee sector withthe lowest gains. For instance, in our baseline scenario, the leader split simu-lation would increase average households’ income for cotton in Burkina Faso


by 12.4 percent but only 0.1 percent in coffee in Uganda. This does not comeas a surprise, however, as the leading firm in cotton in Burkina Faso con-trols 85 percent of the market but only 14.3 percent of the market in thecase of coffee in Uganda. Another interesting simulation showing an increasein competition is the case of “equal market shares.” This would give us theupper bound increase in income under imperfect competition. Here the aver-age effect is much larger than in the case of leader split. The average pro-ducing household in our study would see their income grow by 9.1 percent,with cotton in Burkina Faso once again experiencing the largest impact with20.9 percent, followed by Benin with 20.1 percent and Côte d’Ivoire with a14 percent increase, both in cotton. At the other end of the spectrum, theaverage household gains less than 1 percent in the equal market shares sim-ulation in coffee in Uganda and Rwanda.

The conclusion from the previous paragraph that an increase in compe-tition among processors is good for the farmers needs to be put into per-spective. One of the findings from our simulations is that small changes tothe level of competition are unlikely to have significant effects on farmers’livelihoods. This is captured by the small entrant simulation. Under this sce-nario, the income of households only increases by an average of a quarter ofa percentage point for our case studies. The largest effects for this simula-tion are observed in cotton in Malawi (0.94 percent) and tobacco in Zambia(0.74 percent).

We were also interested in assessing the effects on farmers’ income of areduction in competition among upstream firms. This was done by studyingthe effects of the merging of the largest two firms in the market and throughthe case of the exit of the largest (and most efficient) firm. In the first simula-tion, the average loss for producing households is 1.3 percent of their income,with the largest loss registered in the case of cotton in Côte d’Ivoire (3.8 per-cent), where the new firm would control three quarters of the market. In theexit of the largest firm simulation, the worst income loss for producing house-holds would take place in the cotton sector of Burkina Faso, where the dis-appearance of SOFITEX, which controls 85 percent of the market, would leadto a reduction in income of 15.4 percent. On the other hand, the reductionin competition arising from these two simulations would affect the averageproducing household in the coffee sector in Uganda the least, where the losswould only be around a tenth of a percentage point in both simulations.

Another issue we wanted to incorporate into the analysis is the effects ofcomplementary policies. We did so by comparing our baseline scenario withcases where we introduced a complementary policy affecting farmers (2 per-cent of increase in yields), affecting firms (2 percent reduction of processingcost), or both. The main finding here is that these policies help to increasethe income of households when there is an increase in competition and canmitigate or even revert income loss when there is a loss because of a reduc-tion in competition. In all cases, the effect is the largest when applied to both


firms and farmers. In the case of only farmers versus only firms’ complemen-tary policies, the latter have the largest effect on the livelihood of farmers inall cases except tobacco in Zambia. The quantitative effects of these policiesdepend on the particular crop, country, and market structure but it goes fromas little as an extra 0.2 percent more up to 2.7 percent extra income for theaverage producing household.

We also studied the effects on households’ livelihood of a 10 percent exoge-nous increase in the international price of the crop under consideration. Giventhe current market configuration for each crop and country, that exogenousprice increase leads to an average raise in producing households’ income of6.9 percent, with the largest effect taking place for cotton in Burkina Faso(12.8 percent) and the lowest for coffee in Rwanda (2 percent). We thenallow for combinations of the international crop price increase with differentchanges to the level of competition among processors. Increasing competitionwill boost the positive effects of the price change, while a reduction in thecompetition will dampen its effects. For instance, in the case of perfect com-petition, the income gains for producing households would range between68 percent (cotton in Burkina Faso) and 3.8 percent (coffee in Uganda), withan average effect across case studies of 29.6 percent. On the other hand, whenthe largest firm exits the market, the overall income effect of the internationalprice increase would range between −3.9 percent (cotton in Burkina Faso) and11.3 percent (cocoa in Côte d’Ivoire), with an average income effect of 3.4 per-cent for all the case studies.

The survey data allows us to distinguish the effect of the different simu-lations on poor versus nonpoor households and across gender groups. Herethe results depend on the income share of the crop in each country for eachgroup, as the price simulations are unique. A richer model could incorporatepolicies or market changes that affect poor or female-headed households ina different way from nonpoor and male-headed households, but this is notthe case in our simulations. In nine out of the twelve simulations, the benefitsof more competition have a larger income effect in male-headed householdsthan in the female counterpart. The three exceptions were the case of cot-ton in Benin and Côte d’Ivoire, and coffee in Rwanda. The largest differencesamong genders were registered in Burkina Faso cotton, where male-headedhouseholds received 217 percent more income increase than female-headedhouseholds, and in Benin cotton, where female-headed households received57 percent more than the male equivalent. In only four out of the twelve casestudies, the increase in competition has been pro-poor. The income gains onaverage benefitted more poor households in the case of coffee and cocoa inCôte d’Ivoire, coffee in Rwanda, and cotton in Zambia.

We also present the results for a model that incorporates outgrower con-tracts. Small farmers can receive financing from processors in exchange forfuture output sales through outgrower schemes. We assume that the cost ofenforcing these contracts increases with market competition and that those


costs are transferred to producers through increasing borrowing costs. Wetherefore run the same set of simulations taking this feature into consider-ation and compare it with our original set of simulations. We find that, withoutgrower contracts, the benefits of increasing competition and the negativeeffects of a more concentrated market are both reduced. The effect is, how-ever, rather small except for the case of cotton in Burkina Faso. In this lastcase study, the merging of the largest two firms would reduce farmers’ incomeby 1.8 percent in the basic model without outgrower contracts but it wouldactually increase it by 1.3 percent in the model with these types of contracts.This is an atypical case in which less competition is better for smallholders.

Three of the countries in our study have more than one case study. In Côted’Ivoire we study cotton, coffee, and cocoa. In Malawi and Zambia we coverboth cotton and tobacco. It is interesting, then, to describe how the same sce-nario and simulation has different effects across crops in the same country.For instance, in Côte d’Ivoire, an increase in competition has a larger effecton producing households’ income in cotton than in cocoa and coffee. If theleader firm in cotton, cocoa, and coffee were to split, the effect on incomewould be a 4.6, 1.1, and 0.6 percent increase, respectively. In the case ofequal market shares, the increase in the income of households would be 14,10.5, and 5.4 percent, respectively. The effect is also different for poor versusnonpoor households, and across gender depending on the crop. Competitionin coffee benefits more poor and male-headed households, while, in cotton,female-headed and nonpoor households are the ones that obtain larger gains.In cocoa, competition benefits male-headed households slightly more, whilethe effect is about the same among poor and nonpoor households. In Malawiwe cannot directly compare the results from the cotton and the tobacco sim-ulations, since the latter are slightly different to the standard simulation werun for all the other case studies. However, the overall effects seem to beof about the same order of magnitude and, in both crops, male and nonpoorhouseholds benefit the most from the increase in competition. Finally, in Zam-bia, the effect of competition has similar quantitative effects on cotton andtobacco. The leader split case would increase the income of households by2.4 percent for cotton and by 3.2 percent for tobacco, while the equal mar-ket share case would generate a growth in income of 7.3 percent in cottonand 7.2 percent in tobacco. In both crops, male-headed households benefitthe most, though only slightly in the case of cotton. Poor producing house-holds gain more in cotton, while nonpoor benefit more in the case of increasedcompetition among tobacco exporters.



Basicmodel

CP:farmers

CP:firms

CP:both

Internationalprice

Leader split Leaders merge10

8

6

4

2

0

−2

−4

Figure A5.1: Income effects under different scenarios for cotton in Zambia.

−4

−2

0

2

4

6

8

Leadersmerge

Exitof thelargest

Smallentrant

Leadersplit

Equalmarketshares

Only producersAll households

Figure A5.2: Income effects for producers versus incomeeffects for all households for cotton in Zambia.


10

8

6

4

2

0

−2

−4

−6Leadersmerge

Exit of thelargest

Smallentrant

Leadersplit

Equalmarketshares

PoorNonpoor

(a)

PoorNonpoor

Leadersmerge

Exit of thelargest

Smallentrant

Leadersplit

Equalmarketshares

−0.6

−0.2

−0.4

0

0.2

0.4

0.6

0.8

1.0(b)

Figure A5.3: Effects on poor versus nonpoor for (a) producers onlyand (b) all households for cotton in Zambia.


8

6

4

2

0

−2

−4Leadersmerge

Exit of thelargest

Smallentrant

Leadersplit

Equalmarketshares

−0.6

−0.2

−0.4

0

0.2

0.4

0.6

0.8

1.0

(a)

(b)

MaleFemale

MaleFemale

Leadersmerge

Exit of thelargest

Smallentrant

Leadersplit

Equalmarketshares

Figure A5.4: Effects on male-headed versus female-headed households for(a) producers only and (b) all households for cotton in Zambia.


−5

0

5

10

15

20

25Without outgrower contractsWith outgrower contracts

Leaderssplit

Smallentrant

Leadersmerge

andsmall

entrant

Exit of the

largest

Equalmarketshares

Perfectcompetition

Leadersmerge

−5.8%−48% +6.6% +18.5% −0.9%

−11.9%

−4.6%

Figure A5.5: Income effect in basic model with andwithout outgrower contracts for cotton in Zambia.

0123456789

10

−3

−2

−1

0

1

2

3

4

Basicmodel

CP:farmers

CP:firms

CP:both

Internationalprice

Basicmodel

CP:farmers

CP:firms

CP:both

Internationalprice

Without outgrowercontracts

With outgrowercontracts

Without outgrowercontracts

With outgrowercontracts

(a)

(b)

−5.8% −6.1%−10.3% −10.1%

−11.2%

+18.5% +20.7% +15.3% +18.4%

−10.6%

Figure A5.6: Scenarios under (a) leader split and (b) leaders mergewith and without outgrower contracts for cotton in Zambia.


25

20

15

10

5

0Basicmodel

CP:farmers

CP:firms

CP:both

Internationalprice

BeninBurkina FasoCote d'Ivoire ˆ

Basicmodel

CP:farmers

CP:firms

CP:both

Internationalprice

10

6

4

2

0

BeninBurkina FasoCote d'Ivoire ˆ

(a)

(b)

Figure A5.7: Income effects in cotton for producer householdsunder (a) leader split and (b) leaders merge.

25

20

15

10

5

0Benin Burkina Faso Cote d'Ivoire ˆ

PoorNonpoor

Figure A5.8: Equal market shares income effects in cottonfor poor and nonpoor producer households.


0

10

20

30

40

50

60MaleFemale

Benin Burkina Faso Cote d'Ivoire ˆ

Figure A5.9: Perfect competition income effects in cotton formale-headed and female-headed producer households.

60

40

20

0

−20

Leaderssplit

Smallentrant

Leadersmerge

and smallentrant

Exit ofthe

largest

Equalmarketshares

Perfectcompetition

Leadersmerge

Without outgrowercontractsWith outgrowercontracts

(a)

−47%

−169% +151% +170% −42%

−53%

−23%

Figure A5.10: Income effect in basic model with and withoutoutgrower contracts in cotton–Burkina Faso.


10

8

6

4

2

0

−2

−4Basicmodel

CP:farmers

CP:firms

CP:both

Internationalprice

0

5

10

15

20

25

Basicmodel

CP:farmers

CP:firms

CP:both

Internationalprice



(a)

(b)

Figure A5.11: Scenarios under (a) leader split and (b) leaders mergewith and without outgrower contracts in cotton–Burkina Faso.


1.5

1.0

0.5

0

−0.5

−1.0

−1.5Basicmodel

CP:farmers

CP:firms

CP:both

14

12

10

8

6

4

2

0Basicmodel

CP:farmers

CP:firms

CP:both

(b)

25

20

15

10

5

0Current

configurationLeader split Leaders

mergeExit of

thelargest

Equalmarketshares

(c)

Côte d'Ivoire Ghana



(a)

Figure A5.12: Income effects for cocoa producer households under (a) exit of theleader, (b) equal market shares, and (c) increased international price.


PoorNonpoor

25

20

15

10

5

0Cote d'Ivoire ˆ Ghana

Figure A5.13: Perfect competition income effects in cocoafor poor and nonpoor producer households.

1.2

1.0

0.8

0.6

0.4

0.2

0Cote d'Ivoire ˆ Ghana

MaleFemale

Figure A5.14: Leader split income effects in cocoa formale-headed and female-headed producer households.

Basicmodel

CP:farmers

CP:firms

CP:both

Internationalprice

+23.3%

+13.2% +2.5% +0.5%

−7.8%Without outgrower contracts

With outgrower contracts

8

7

6

5

4

3

2

1

0

Figure A5.15: Scenarios under leader split with andwithout outgrower contracts in cocoa–Ghana.


2.0

1.6

1.2

0.8

0.4

0Basicmodel

CP:farmers

CP:firms

CP:both

Cote d'Ivoire ˆ

Rwanda

Uganda

Cote d'Ivoire ˆ

Rwanda

Uganda

0

0.4

0.8

−0.4

(a)

Basicmodel

CP:farmers

CP:firms

CP:both

(b)

(c)14

12

10

8

6

4

2

0

Cote d'Ivoire ˆ

Rwanda

Uganda

Currentconfiguration

Leadersmerge

Equalmarketshares

Figure A5.16: Income effects for coffee producer households under (a) leader split,(b) leaders merge, and (c) increased international price.


7

6

5

4

3

2

1

0Cote d'Ivoire ˆ Rwanda Uganda

PoorNonpoor

Figure A5.17: Equal market shares income effects in coffeefor poor and nonpoor producer households.

12

10

8

6

4

2

0Cote d'Ivoire ˆ Rwanda Uganda

MaleFemale

Figure A5.18: Perfect competition income effects in coffeefor male-headed and female-headed producer households.


4

2

0

−2

−4

−6

6

Basicmodel

CP: farmers

CP:firms

CP:both

Internationalprice

Basicmodel

CP: farmers

CP:firms

CP:both

Internationalprice

MalawiZambia

(a)

MalawiZambia

14

12

10

8

6

4

2

0

(b)

Figure A5.19: Income effects in tobacco for producer householdsunder (a) exit of the largest and (b) equal market shares.


30

25

20

15

10

5

0Malawi Zambia

PoorNonpoor

Figure A5.20: Perfect competition income effects in tobaccofor poor and nonpoor producer households.

4.0

3.5

3.0

2.5

2.0

1.5

1.0

0.5

0Malawi Zambia

MaleFemale

Figure A5.21: Leader split income effects in tobacco formale-headed and female-headed producer households.

8

7

6

5

4

3

2

1

0Basicmodel

CP:farmers

CP:firms

CP:both

Internationalprice

Without outgrower contractsWith outgrower contracts

−44.8% −35.1% −35.1%−29.4%

−10.4%

Figure A5.22: Scenarios under leader split with andwithout outgrower contracts in tobacco–Malawi.


Tab

leA

5.1

:Im

pact

ofsu

pply

chain

chan

ges

onth

est

an

dard

mod

el:(

a)

Zam

bia

,cot

ton

(hou

seh

old

inco

me

per

cen

tage

chan

ges

).

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

and

smal

lLe

ader

sof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

entr

ant

mer

ge

larg

est

shar

esco

mp

etit

ion

(A)

Basi

cm

odel

On

lyp

rod

uce

rsT

ota

l0.0

02.4

00.2

71.6

82.3

23.1

97.3

319.3

0Poor

0.0

02.6

00.2

91.8

22.5

13.4

67.9

520.9

2N

on

poor

0.0

02.2

90.2

61.6

02.2

13.0

46.9

918.4

0M

ale-

hea

ded

0.0

02.4

10.2

71.6

82.3

23.2

07.3

619.3

7Fe

mal

e-h

ead

ed0.0

02.3

50.2

61.6

42.2

63.1

27.1

718.8

7

All

hou

seh

old

sT

ota

l0.0

00.2

70.0

30.1

80.2

60.3

50.8

12.1

3Poor

0.0

00.2

30.0

30.1

60.2

20.3

10.7

01.8

6N

on

poor

0.0

00.2

90.0

30.2

00.2

80.3

90.8

92.3

5M

ale-

hea

ded

0.0

00.2

80.0

30.2

00.2

70.3

80.8

72.2

8Fe

mal

e-h

ead

ed0.0

00.1

90.0

20.1

30.1

80.2

50.5

71.5

1

(B)

Com

ple

men

tary

pol

icie

s(f

arm

ers)

On

lyp

rod

uce

rsT

ota

l0.3

92.7

70.6

81.2

71.9

32.8

07.6

619.8

4Poor

0.4

23.0

00.7

31.3

82.0

93.0

38.3

021.5

1N

on

poor

0.3

72.6

40.6

41.2

21.8

42.6

77.3

018.9

1M

ale-

hea

ded

0.3

92.7

80.6

81.2

81.9

42.8

17.6

919.9

1Fe

mal

e-h

ead

ed0.3

82.7

10.6

61.2

51.8

92.7

47.4

919.3

9

All

hou

seh

old

sT

ota

l0.0

40.3

10.0

70.1

40.2

10.3

10.8

52.1

9Poor

0.0

40.2

70.0

60.1

20.1

90.2

70.7

41.9

1N

on

poor

0.0

50.3

40.0

80.1

60.2

40.3

40.9

32.4

1M

ale-

hea

ded

0.0

50.3

30.0

80.1

50.2

30.3

30.9

12.3

4Fe

mal

e-h

ead

ed0.0

30.2

20.0

50.1

00.1

50.2

20.6

01.5

5


Tab

leA

5.1

:(a

)C

onti

nu

ed.

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

and

smal

lLe

ader

sof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

entr

ant

mer

ge

larg

est

shar

esco

mp

etit

ion

(C)

Com

ple

men

tary

pol

icie

s(fi

rms)

On

lyp

rod

uce

rsT

ota

l0.7

13.1

01.0

40.9

21.6

32.4

87.8

820.1

1Poor

0.7

73.3

61.1

31.0

01.7

62.6

98.5

521.8

0N

on

poor

0.6

82.9

50.9

90.8

81.5

52.3

67.5

219.1

7M

ale-

hea

ded

0.7

23.1

11.0

40.9

31.6

32.4

97.9

120.1

8Fe

mal

e-h

ead

ed0.7

03.0

31.0

10.9

01.5

92.4

27.7

119.6

5

All

hou

seh

old

sT

ota

l0.0

80.3

40.1

10.1

00.1

80.2

70.8

72.2

2Poor

0.0

70.3

00.1

00.0

90.1

60.2

40.7

61.9

3N

on

poor

0.0

90.3

80.1

30.1

10.2

00.3

00.9

62.4

5M

ale-

hea

ded

0.0

80.3

70.1

20.1

10.1

90.2

90.9

32.3

8Fe

mal

e-h

ead

ed0.0

60.2

40.0

80.0

70.1

30.1

90.6

21.5

8

(D)

Com

ple

men

tary

pol

icie

s(f

arm

ers−

firm

s)

On

lyp

rod

uce

rsT

ota

l1.1

03.4

81.4

40.5

31.2

52.1

08.2

120.6

4Poor

1.1

93.7

81.5

60.5

71.3

52.2

78.9

022.3

8N

on

poor

1.0

53.3

21.3

70.5

01.1

92.0

07.8

319.6

8M

ale-

hea

ded

1.1

03.5

01.4

40.5

31.2

52.1

08.2

420.7

2Fe

mal

e-h

ead

ed1.0

73.4

01.4

10.5

21.2

22.0

58.0

220.1

8

All

hou

seh

old

sT

ota

l0.1

20.3

80.1

60.0

60.1

40.2

30.9

12.2

8Poor

0.1

10.3

30.1

40.0

50.1

20.2

00.7

91.9

8N

on

poor

0.1

30.4

20.1

70.0

60.1

50.2

51.0

02.5

1M

ale-

hea

ded

0.1

30.4

10.1

70.0

60.1

50.2

50.9

72.4

4Fe

mal

e-h

ead

ed0.0

90.2

70.1

10.0

40.1

00.1

60.6

41.6

2


Tab

leA

5.1

:(a

)C

onti

nu

ed.

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

and

smal

lLe

ader

sof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

entr

ant

mer

ge

larg

est

shar

esco

mp

etit

ion

(E)

Inte

rnati

onalpri

ces

On

lyp

rod

uce

rsT

ota

l6.0

38.6

96.6

24.3

93.3

52.5

013.2

327.9

5Poor

6.5

49.4

27.1

74.7

63.6

32.7

114.3

430.3

0N

on

poor

5.7

58.2

86.3

14.1

93.1

92.3

812.6

126.6

4M

ale-

hea

ded

6.0

68.7

26.6

44.4

13.3

62.5

113.2

828.0

5Fe

mal

e-h

ead

ed5.9

08.4

96.4

74.2

93.2

82.4

412.9

327.3

2

All

hou

seh

old

sT

ota

l0.6

70.9

60.7

30.4

80.3

70.2

81.4

63.0

8Poor

0.5

80.8

40.6

40.4

20.3

20.2

41.2

72.6

9N

on

poor

0.7

31.0

60.8

00.5

30.4

10.3

01.6

13.4

0M

ale-

hea

ded

0.7

11.0

30.7

80.5

20.4

00.2

91.5

63.3

0Fe

mal

e-h

ead

ed0.4

70.6

80.5

20.3

40.2

60.2

01.0

42.1

9


Tab

leA

5.1

:(b

)Ben

in,c

otto

n(h

ouse

hol

din

com

eper

cen

tage

chan

ges

).

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

and

smal

lLe

ader

sof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

entr

ant

mer

ge

larg

est

shar

esco

mp

etit

ion

(A)

Basi

cm

odel

On

lyp

rod

uce

rsT

ota

l0.0

03.3

30.0

90.5

10.6

73.6

920.1

134.1

6Poor

0.0

03.2

50.0

90.4

90.6

53.6

019.6

433.3

6N

on

poor

0.0

03.4

70.0

90.5

30.7

03.8

520.9

935.6

6M

ale-

hea

ded

0.0

02.1

50.0

60.3

30.4

32.3

812.9

722.0

2Fe

mal

e-h

ead

ed0.0

03.3

80.0

90.5

10.6

83.7

520.4

234.6

8

All

hou

seh

old

sT

ota

l0.0

00.6

30.0

20.1

00.1

30.6

93.7

86.4

2Poor

0.0

00.7

90.0

20.1

20.1

60.8

74.7

58.0

7N

on

poor

0.0

00.4

60.0

10.0

70.0

90.5

12.7

94.7

4M

ale-

hea

ded

0.0

00.1

10.0

00.0

20.0

20.1

20.6

51.1

1Fe

mal

e-h

ead

ed0.0

00.7

20.0

20.1

10.1

40.8

04.3

37.3

6

(B)

Com

ple

men

tary

pol

icie

s(f

arm

ers)

On

lyp

rod

uce

rsT

ota

l0.5

53.8

40.6

50.0

50.1

33.1

320.5

234.8

6Poor

0.5

43.7

50.6

30.0

40.1

33.0

620.0

434.0

4N

on

poor

0.5

74.0

00.6

80.0

50.1

43.2

721.4

236.3

9M

ale-

hea

ded

0.3

52.4

70.4

20.0

30.0

82.0

213.2

322.4

8Fe

mal

e-h

ead

ed0.5

63.8

90.6

60.0

50.1

33.1

820.8

335.3

9

All

hou

seh

old

sT

ota

l0.1

00.7

20.1

20.0

10.0

20.5

93.8

66.5

5Poor

0.1

30.9

10.1

50.0

10.0

30.7

44.8

58.2

4N

on

poor

0.0

80.5

30.0

90.0

10.0

20.4

42.8

54.8

4M

ale-

hea

ded

0.0

20.1

20.0

20.0

00.0

00.1

00.6

71.1

4Fe

mal

e-h

ead

ed0.1

20.8

30.1

40.0

10.0

30.6

84.4

27.5

1


Tab

leA

5.1

:(b

)C

onti

nu

ed.

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

and

smal

lLe

ader

sof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

entr

ant

mer

ge

larg

est

shar

esco

mp

etit

ion

(C)

Com

ple

men

tary

pol

icie

s(fi

rms)

On

lyp

rod

uce

rsT

ota

l0.7

33.9

50.8

50.2

40.0

42.9

220.4

234.5

8Poor

0.7

13.8

60.8

30.2

30.0

42.8

519.9

433.7

6N

on

poor

0.7

64.1

20.8

90.2

50.0

43.0

521.3

236.0

9M

ale-

hea

ded

0.4

72.5

50.5

50.1

50.0

31.8

813.1

722.2

9Fe

mal

e-h

ead

ed0.7

44.0

10.8

60.2

40.0

42.9

620.7

335.1

0

All

hou

seh

old

sT

ota

l0.1

40.7

40.1

60.0

40.0

10.5

53.8

46.5

0Poor

0.1

70.9

30.2

00.0

60.0

10.6

94.8

28.1

7N

on

poor

0.1

00.5

50.1

20.0

30.0

10.4

12.8

34.8

0M

ale-

hea

ded

0.0

20.1

30.0

30.0

10.0

00.1

00.6

71.1

3Fe

mal

e-h

ead

ed0.1

60.8

50.1

80.0

50.0

10.6

34.4

07.4

5

(D)

Com

ple

men

tary

pol

icie

s(f

arm

ers−

firm

s)

On

lyp

rod

uce

rsT

ota

l1.2

84.4

51.4

10.7

80.5

72.3

720.8

335.2

8Poor

1.2

54.3

51.3

70.7

60.5

62.3

120.3

434.4

5N

on

poor

1.3

34.6

51.4

70.8

20.6

02.4

721.7

436.8

2M

ale-

hea

ded

0.8

22.8

70.9

10.5

00.3

71.5

313.4

322.7

4Fe

mal

e-h

ead

ed1.2

94.5

21.4

30.7

90.5

82.4

021.1

535.8

1

All

hou

seh

old

sT

ota

l0.2

40.8

40.2

60.1

50.1

10.4

43.9

16.6

3Poor

0.3

01.0

50.3

30.1

90.1

30.5

64.9

28.3

3N

on

poor

0.1

80.6

20.2

00.1

10.0

80.3

32.8

94.9

0M

ale-

hea

ded

0.0

40.1

50.0

50.0

30.0

20.0

80.6

81.1

5Fe

mal

e-h

ead

ed0.2

70.9

60.3

00.1

70.1

20.5

14.4

97.6

0


Tab

leA

5.1

:(b

)C

onti

nu

ed.

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

and

smal

lLe

ader

sof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

entr

ant

mer

ge

larg

est

shar

esco

mp

etit

ion

(E)

Inte

rnati

onalpri

ces

On

lyp

rod

uce

rsT

ota

l6.9

410.1

07.2

06.4

46.0

63.0

926.8

043.1

6Poor

6.7

79.8

67.0

36.2

95.9

23.0

126.1

742.1

4N

on

poor

7.2

410.5

47.5

26.7

26.3

33.2

227.9

745.0

4M

ale-

hea

ded

4.4

76.5

14.6

44.1

53.9

11.9

917.2

827.8

2Fe

mal

e-h

ead

ed7.0

410.2

57.3

16.5

36.1

53.1

327.2

043.8

1

All

hou

seh

old

sT

ota

l1.3

01.9

01.3

51.2

11.1

40.5

85.0

38.1

1Poor

1.6

42.3

91.7

01.5

21.4

30.7

36.3

310.2

0N

on

poor

0.9

61.4

01.0

00.8

90.8

40.4

33.7

25.9

9M

ale-

hea

ded

0.2

30.3

30.2

30.2

10.2

00.1

00.8

71.4

1Fe

mal

e-h

ead

ed1.4

92.1

81.5

51.3

91.3

10.6

65.7

89.3

0


Tab

leA

5.1

:(c

)Bu

rkin

aFa

so,c

otto

n(h

ouse

hol

din

com

eper

cen

tage

chan

ges

).

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

and

smal

lLe

ader

sof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

entr

ant

mer

ge

larg

est

shar

esco

mp

etit

ion

(A)

Basi

cm

odel

On

lyp

rod

uce

rsT

ota

l0.0

012.3

60.5

50.6

81.7

9−1

5.3

620.9

149.6

1Poor

0.0

011.8

50.5

30.6

61.7

1−1

4.7

420.0

547.5

9N

on

poor

0.0

012.6

70.5

70.7

01.8

3−1

5.7

521.4

450.8

7M

ale-

hea

ded

0.0

012.5

50.5

60.6

91.8

1−1

5.6

021.2

350.3

9Fe

mal

e-h

ead

ed0.0

03.9

50.1

80.2

20.5

74.9

16.6

815.8

6

All

hou

seh

old

sT

ota

l0.0

00.2

90.0

10.0

20.0

40.3

60.4

91.1

7Poor

0.0

00.2

20.0

10.0

10.0

30.2

80.3

80.9

0N

on

poor

0.0

00.3

60.0

20.0

20.0

50.4

40.6

01.4

3M

ale-

hea

ded

0.0

00.3

10.0

10.0

20.0

40.3

90.5

21.2

4Fe

mal

e-h

ead

ed0.0

00.0

30.0

00.0

00.0

00.0

40.0

60.1

3

(B)

Com

ple

men

tary

pol

icie

s(f

arm

ers)

On

lyp

rod

uce

rsT

ota

l0.7

612.8

71.3

70.0

81.1

2−1

4.5

621.4

850.7

2Poor

0.7

312.3

51.3

20.0

81.0

7−1

3.9

620.6

148.6

5N

on

poor

0.7

813.2

01.4

10.0

81.1

4−1

4.9

322.0

352.0

1M

ale-

hea

ded

0.7

713.0

71.3

90.0

81.1

3−1

4.7

821.8

251.5

1Fe

mal

e-h

ead

ed0.2

44.1

20.4

40.0

30.3

64.6

56.8

716.2

2

All

hou

seh

old

sT

ota

l0.0

20.3

00.0

30.0

00.0

30.3

40.5

11.2

0Poor

0.0

10.2

30.0

20.0

00.0

20.2

60.3

90.9

2N

on

poor

0.0

20.3

70.0

40.0

00.0

30.4

20.6

21.4

7M

ale-

hea

ded

0.0

20.3

20.0

30.0

00.0

30.3

70.5

41.2

7Fe

mal

e-h

ead

ed0.0

00.0

30.0

00.0

00.0

00.0

40.0

60.1

4


Tab

leA

5.1

:(c

)C

onti

nu

ed.

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

and

smal

lLe

ader

sof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

entr

ant

mer

ge

larg

est

shar

esco

mp

etit

ion

(C)

Com

ple

men

tary

pol

icie

s(fi

rms)

On

lyp

rod

uce

rsT

ota

l1.5

313.3

02.2

60.8

60.5

4−1

3.8

121.9

851.1

8Poor

1.4

612.7

62.1

70.8

20.5

2−1

3.2

521.0

949.1

0N

on

poor

1.5

613.6

42.3

20.8

80.5

5−1

4.1

622.5

452.4

8M

ale-

hea

ded

1.5

513.5

12.3

00.8

70.5

5−1

4.0

222.3

351.9

8Fe

mal

e-h

ead

ed0.4

94.2

50.7

20.2

70.1

74.4

17.0

316.3

6

All

hou

seh

old

sT

ota

l0.0

40.3

10.0

50.0

20.0

10.3

30.5

21.2

1Poor

0.0

30.2

40.0

40.0

20.0

10.2

50.4

00.9

3N

on

poor

0.0

40.3

80.0

70.0

20.0

20.4

00.6

41.4

8M

ale-

hea

ded

0.0

40.3

30.0

60.0

20.0

10.3

50.5

51.2

8Fe

mal

e-h

ead

ed0.0

00.0

40.0

10.0

00.0

00.0

40.0

60.1

4

(D)

Com

ple

men

tary

pol

icie

s(f

arm

ers−

firm

s)

On

lyp

rod

uce

rsT

ota

l2.2

713.8

03.0

71.6

10.1

2−1

3.0

222.5

552.2

9Poor

2.1

813.2

42.9

51.5

40.1

1−1

2.4

921.6

350.1

6N

on

poor

2.3

314.1

53.1

51.6

50.1

2−1

3.3

523.1

253.6

2M

ale-

hea

ded

2.3

114.0

23.1

21.6

30.1

2−1

3.2

222.9

053.1

1Fe

mal

e-h

ead

ed0.7

34.4

10.9

80.5

10.0

44.1

67.2

116.7

2

All

hou

seh

old

sT

ota

l0.0

50.3

30.0

70.0

40.0

00.3

10.5

31.2

4Poor

0.0

40.2

50.0

60.0

30.0

00.2

40.4

10.9

5N

on

poor

0.0

70.4

00.0

90.0

50.0

00.3

80.6

51.5

1M

ale-

hea

ded

0.0

60.3

50.0

80.0

40.0

00.3

30.5

71.3

1Fe

mal

e-h

ead

ed0.0

10.0

40.0

10.0

00.0

00.0

40.0

60.1

4


Tab

leA

5.1

:(c

)C

onti

nu

ed.

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

and

smal

lLe

ader

sof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

entr

ant

mer

ge

larg

est

shar

esco

mp

etit

ion

(E)

Inte

rnati

onalpri

ces

On

lyp

rod

uce

rsT

ota

l12.8

023.4

914.3

912.1

29.3

73.9

433.4

267.9

8Poor

12.2

722.5

313.8

011.6

38.9

93.7

832.0

565.2

1N

on

poor

13.1

224.0

914.7

512.4

39.6

14.0

434.2

769.7

1M

ale-

hea

ded

13.0

023.8

614.6

112.3

19.5

24.0

033.9

469.0

5Fe

mal

e-h

ead

ed4.0

97.5

14.6

03.8

83.0

01.2

610.6

821.7

3

All

hou

seh

old

sT

ota

l0.3

00.5

60.3

40.2

90.2

20.0

90.7

91.6

1Poor

0.2

30.4

30.2

60.2

20.1

70.0

70.6

01.2

3N

on

poor

0.3

70.6

80.4

20.3

50.2

70.1

10.9

71.9

6M

ale-

hea

ded

0.3

20.5

90.3

60.3

00.2

40.1

00.8

41.7

1Fe

mal

e-h

ead

ed0.0

30.0

60.0

40.0

30.0

30.0

10.0

90.1

8


Tab

leA

5.1

:(d

)C

ôte

d’Ivo

ire,

cott

on(h

ouse

hol

din

com

eper

cen

tage

chan

ges

).

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

and

smal

lLe

ader

sof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

entr

ant

mer

ge

larg

est

shar

esco

mp

etit

ion

(A)

Basi

cm

odel

On

lyp

rod

uce

rsT

ota

l0.0

04.6

20.4

02.8

43.8

55.9

913.9

935.1

1Poor

0.0

04.4

50.3

82.7

33.7

15.7

613.4

733.7

9N

on

poor

0.0

04.6

90.4

02.8

83.9

16.0

814.2

135.6

6M

ale-

hea

ded

0.0

04.5

80.3

92.8

13.8

25.9

413.8

734.8

0Fe

mal

e-h

ead

ed0.0

06.0

40.5

23.7

05.0

37.8

318.2

845.8

8

All

hou

seh

old

sT

ota

l0.0

00.3

80.0

30.2

30.3

10.4

91.1

42.8

5Poor

0.0

00.2

00.0

20.1

20.1

70.2

60.6

11.5

3N

on

poor

0.0

00.5

60.0

50.3

50.4

70.7

31.7

14.2

9M

ale-

hea

ded

0.0

00.4

40.0

40.2

70.3

60.5

71.3

23.3

2Fe

mal

e-h

ead

ed0.0

00.0

80.0

10.0

50.0

70.1

10.2

50.6

2

(B)

Com

ple

men

tary

pol

icie

s(f

arm

ers)

On

lyp

rod

uce

rsT

ota

l0.7

65.3

51.1

92.0

63.1

15.2

314.6

436.1

6Poor

0.7

35.1

51.1

41.9

82.9

95.0

314.0

934.8

0N

on

poor

0.7

75.4

31.2

12.0

93.1

65.3

114.8

736.7

2M

ale-

hea

ded

0.7

55.3

01.1

82.0

43.0

85.1

814.5

135.8

3Fe

mal

e-h

ead

ed0.9

96.9

91.5

52.6

94.0

66.8

319.1

347.2

5

All

hou

seh

old

sT

ota

l0.0

60.4

30.1

00.1

70.2

50.4

21.1

92.9

4Poor

0.0

30.2

30.0

50.0

90.1

40.2

30.6

41.5

8N

on

poor

0.0

90.6

50.1

50.2

50.3

80.6

41.7

94.4

2M

ale-

hea

ded

0.0

70.5

00.1

10.1

90.2

90.4

91.3

83.4

1Fe

mal

e-h

ead

ed0.0

10.0

90.0

20.0

40.0

50.0

90.2

60.6

4


Tab

leA

5.1

:(d

)C

onti

nu

ed.

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

and

smal

lLe

ader

sof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

entr

ant

mer

ge

larg

est

shar

esco

mp

etit

ion

(C)

Com

ple

men

tary

pol

icie

s(fi

rms)

On

lyp

rod

uce

rsT

ota

l1.5

66.0

72.0

61.2

22.3

64.4

515.2

436.8

7Poor

1.5

05.8

51.9

81.1

72.2

74.2

814.6

735.4

8N

on

poor

1.5

86.1

72.0

91.2

32.4

04.5

215.4

837.4

4M

ale-

hea

ded

1.5

46.0

22.0

41.2

12.3

44.4

115.1

036.5

4Fe

mal

e-h

ead

ed2.0

37.9

42.6

91.5

93.0

95.8

219.9

148.1

8

All

hou

seh

old

sT

ota

l0.1

30.4

90.1

70.1

00.1

90.3

61.2

43.0

0Poor

0.0

70.2

60.0

90.0

50.1

00.1

90.6

61.6

1N

on

poor

0.1

90.7

40.2

50.1

50.2

90.5

41.8

64.5

1M

ale-

hea

ded

0.1

50.5

70.1

90.1

10.2

20.4

21.4

43.4

8Fe

mal

e-h

ead

ed0.0

30.1

10.0

40.0

20.0

40.0

80.2

70.6

5

(D)

Com

ple

men

tary

pol

icie

s(f

arm

ers−

firm

s)

On

lyp

rod

uce

rsT

ota

l2.3

06.7

92.8

40.4

51.6

33.7

015.8

837.9

2Poor

2.2

26.5

32.7

30.4

31.5

73.5

615.2

836.4

9N

on

poor

2.3

46.8

92.8

80.4

51.6

63.7

616.1

338.5

0M

ale-

hea

ded

2.2

86.7

32.8

10.4

41.6

23.6

715.7

437.5

8Fe

mal

e-h

ead

ed3.0

18.8

73.7

10.5

82.1

34.8

420.7

549.5

4

All

hou

seh

old

sT

ota

l0.1

90.5

50.2

30.0

40.1

30.3

01.2

93.0

8Poor

0.1

00.3

00.1

20.0

20.0

70.1

60.6

91.6

5N

on

poor

0.2

80.8

30.3

50.0

50.2

00.4

51.9

44.6

3M

ale-

hea

ded

0.2

20.6

40.2

70.0

40.1

50.3

51.5

03.5

8Fe

mal

e-h

ead

ed0.0

40.1

20.0

50.0

10.0

30.0

70.2

80.6

7


Tab

leA

5.1

:(d

)C

onti

nu

ed.

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

and

smal

lLe

ader

sof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

entr

ant

mer

ge

larg

est

shar

esco

mp

etit

ion

(E)

Inte

rnati

onalpri

ces

On

lyp

rod

uce

rsT

ota

l12.5

917.5

613.5

99.8

07.9

95.9

026.3

352.6

4Poor

12.1

116.9

013.0

89.4

37.6

95.6

825.3

450.6

6N

on

poor

12.7

817.8

313.8

09.9

68.1

25.9

926.7

453.4

5M

ale-

hea

ded

12.4

717.4

013.4

79.7

27.9

25.8

526.1

052.1

7Fe

mal

e-h

ead

ed16.4

422.9

417.7

612.8

110.4

57.7

134.4

168.7

8

All

hou

seh

old

sT

ota

l1.0

21.4

31.1

00.8

00.6

50.4

82.1

44.2

8Poor

0.5

50.7

70.5

90.4

30.3

50.2

61.1

52.3

0N

on

poor

1.5

42.1

51.6

61.2

00.9

80.7

23.2

26.4

3M

ale-

hea

ded

1.1

91.6

61.2

80.9

30.7

50.5

62.4

94.9

7Fe

mal

e-h

ead

ed0.2

20.3

10.2

40.1

70.1

40.1

00.4

60.9

3


Table A5.1: (e) Malawi, cotton (household income percentage changes).

SmallRespect entrant with

to Three Four half oforiginal firms firms the benefits

(A) Basic model

Only producers Total 0.00 1.87 3.02 0.94Poor 0.00 1.49 2.40 0.75Nonpoor 0.00 2.12 3.43 1.06Male-headed 0.00 1.89 3.05 0.95Female-headed 0.00 1.68 2.72 0.84

All households Total 0.00 0.06 0.09 0.03Poor 0.00 0.04 0.07 0.02Nonpoor 0.00 0.07 0.11 0.03Male-headed 0.00 0.07 0.11 0.03Female-headed 0.00 0.02 0.04 0.01

(B) Complementary policies (farmers)



(C) Complementary policies (firms)




Table A5.1: (e) Continued.



(D) Complementary policies (farmers − firms)



(E) International prices




Tab

leA

5.1

:(f

)C

ôte

d’Ivo

ire,

coco

a(h

ouse

hol

din

com

eper

cen

tage

chan

ges

).

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

and

smal

lLe

ader

sof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

entr

ant

mer

ge

larg

est

shar

esco

mp

etit

ion

(A)

Basi

cm

odel

On

lyp

rod

uce

rsT

ota

l0.0

01.1

30.2

10.6

40.9

11.2

610.5

321.4

9Poor

0.0

01.1

30.2

10.6

40.9

21.2

710.6

021.6

2N

on

poor

0.0

01.1

20.2

10.6

30.9

11.2

510.4

621.3

3M

ale-

hea

ded

0.0

01.1

40.2

20.6

50.9

21.2

710.6

421.7

0Fe

mal

e-h

ead

ed0.0

00.9

20.1

70.5

20.7

51.0

38.6

017.5

5

All

hou

seh

old

sT

ota

l0.0

00.3

30.0

60.1

90.2

70.3

73.1

26.3

7Poor

0.0

00.3

50.0

70.2

00.2

90.4

03.3

16.7

5N

on

poor

0.0

00.3

10.0

60.1

80.2

50.3

52.9

25.9

6M

ale-

hea

ded

0.0

00.3

90.0

70.2

20.3

10.4

33.6

17.3

7Fe

mal

e-h

ead

ed0.0

00.0

80.0

20.0

50.0

70.0

90.7

51.5

3

(B)

Com

ple

men

tary

pol

icie

s(f

arm

ers)

On

lyp

rod

uce

rsT

ota

l0.9

62.0

91.1

80.3

20.0

40.3

111.4

422.6

5Poor

0.9

62.1

01.1

90.3

20.0

40.3

111.5

122.7

8N

on

poor

0.9

52.0

71.1

70.3

20.0

40.3

111.3

522.4

8M

ale-

hea

ded

0.9

72.1

11.1

90.3

30.0

40.3

111.5

522.8

7Fe

mal

e-h

ead

ed0.7

81.7

10.9

60.2

60.0

30.2

59.3

418.5

0

All

hou

seh

old

sT

ota

l0.2

80.6

20.3

50.1

00.0

10.0

93.3

96.7

1Poor

0.3

00.6

60.3

70.1

00.0

10.1

03.5

97.1

1N

on

poor

0.2

70.5

80.3

30.0

90.0

10.0

93.1

76.2

8M

ale-

hea

ded

0.3

30.7

20.4

00.1

10.0

10.1

13.9

27.7

7Fe

mal

e-h

ead

ed0.0

70.1

50.0

80.0

20.0

00.0

20.8

21.6

2


Tab

leA

5.1

:(f

)C

onti

nu

ed.

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

and

smal

lLe

ader

sof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

entr

ant

mer

ge

larg

est

shar

esco

mp

etit

ion

(C)

Com

ple

men

tary

pol

icie

s(fi

rms)

On

lyp

rod

uce

rsT

ota

l1.4

62.5

91.6

90.8

30.5

40.2

011.7

722.9

9Poor

1.4

72.6

01.7

00.8

40.5

40.2

011.8

423.1

3N

on

poor

1.4

52.5

71.6

80.8

30.5

30.2

011.6

822.8

2M

ale-

hea

ded

1.4

72.6

11.7

10.8

40.5

40.2

011.8

823.2

2Fe

mal

e-h

ead

ed1.1

92.1

11.3

80.6

80.4

40.1

69.6

118.7

8

All

hou

seh

old

sT

ota

l0.4

30.7

70.5

00.2

50.1

60.0

63.4

96.8

2Poor

0.4

60.8

10.5

30.2

60.1

70.0

63.7

07.2

2N

on

poor

0.4

00.7

20.4

70.2

30.1

50.0

53.2

66.3

7M

ale-

hea

ded

0.5

00.8

90.5

80.2

90.1

80.0

74.0

47.8

9Fe

mal

e-h

ead

ed0.1

00.1

80.1

20.0

60.0

40.0

10.8

41.6

4

(D)

Com

ple

men

tary

pol

icie

s(f

arm

ers−

firm

s)

On

lyp

rod

uce

rsT

ota

l2.4

13.5

42.6

51.7

91.4

81.1

412.6

624.1

5Poor

2.4

23.5

62.6

71.8

01.4

91.1

512.7

424.3

0N

on

poor

2.3

93.5

22.6

31.7

71.4

71.1

312.5

723.9

7M

ale-

hea

ded

2.4

33.5

82.6

81.8

11.4

91.1

512.7

924.3

8Fe

mal

e-h

ead

ed1.9

72.8

92.1

71.4

61.2

10.9

310.3

419.7

2

All

hou

seh

old

sT

ota

l0.7

11.0

50.7

90.5

30.4

40.3

43.7

57.1

6Poor

0.7

61.1

10.8

30.5

60.4

60.3

63.9

87.5

9N

on

poor

0.6

70.9

80.7

40.5

00.4

10.3

23.5

16.6

9M

ale-

hea

ded

0.8

31.2

20.9

10.6

10.5

10.3

94.3

48.2

9Fe

mal

e-h

ead

ed0.1

70.2

50.1

90.1

30.1

10.0

80.9

01.7

2


Tab

leA

5.1

:(f

)C

onti

nu

ed.

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

and

smal

lLe

ader

sof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

entr

ant

mer

ge

larg

est

shar

esco

mp

etit

ion

(E)

Inte

rnati

onalpri

ces

On

lyp

rod

uce

rsT

ota

l12.7

413.9

913.0

912.1

111.6

811.3

423.1

336.9

3Poor

12.8

114.0

713.1

712.1

811.7

511.4

123.2

737.1

5N

on

poor

12.6

413.8

812.9

912.0

211.5

911.2

522.9

636.6

5M

ale-

hea

ded

12.8

614.1

213.2

112.2

211.7

911.4

523.3

537.2

9Fe

mal

e-h

ead

ed10.4

011.4

210.6

99.8

99.5

49.2

618.8

930.1

6

All

hou

seh

old

sT

ota

l3.7

84.1

53.8

83.5

93.4

63.3

66.8

610.9

5Poor

4.0

04.3

94.1

13.8

03.6

73.5

67.2

711.6

0N

on

poor

3.5

33.8

83.6

33.3

63.2

43.1

46.4

110.2

4M

ale-

hea

ded

4.3

74.8

04.4

94.1

54.0

13.8

97.9

412.6

7Fe

mal

e-h

ead

ed0.9

11.0

00.9

30.8

60.8

30.8

11.6

52.6

3


Tab

leA

5.1

:(g

)G

han

a,c

ocoa

(hou

seh

old

inco

me

per

cen

tage

chan

ges

).

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

and

smal

lLe

ader

sof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

entr

ant

mer

ge

larg

est

shar

esco

mp

etit

ion

(A)

Basi

cm

odel

On

lyp

rod

uce

rsT

ota

l0.0

00.8

60.0

70.2

60.3

71.0

17.3

311.9

0Poor

0.0

00.8

20.0

70.2

50.3

50.9

67.0

111.3

7N

on

poor

0.0

00.9

10.0

70.2

80.3

91.0

67.7

512.5

7M

ale-

hea

ded

0.0

00.8

90.0

70.2

70.3

81.0

47.5

612.2

6Fe

mal

e-h

ead

ed0.0

00.7

70.0

60.2

30.3

30.9

06.5

710.6

7

All

hou

seh

old

sT

ota

l0.0

00.1

40.0

10.0

40.0

60.1

71.2

11.9

7Poor

0.0

00.1

40.0

10.0

40.0

60.1

61.1

91.9

4N

on

poor

0.0

00.1

40.0

10.0

40.0

60.1

71.2

42.0

1M

ale-

hea

ded

0.0

00.1

60.0

10.0

50.0

70.1

91.3

92.2

6Fe

mal

e-h

ead

ed0.0

00.0

90.0

10.0

30.0

40.1

10.8

11.3

1

(B)

Com

ple

men

tary

pol

icie

s(f

arm

ers)

On

lyp

rod

uce

rsT

ota

l0.4

31.2

90.5

00.1

70.0

60.5

87.7

312.4

0Poor

0.4

11.2

30.4

80.1

60.0

60.5

57.3

911.8

4N

on

poor

0.4

51.3

60.5

30.1

80.0

60.6

18.1

713.1

0M

ale-

hea

ded

0.4

41.3

30.5

20.1

70.0

60.6

07.9

712.7

8Fe

mal

e-h

ead

ed0.3

81.1

60.4

50.1

50.0

50.5

26.9

311.1

2

All

hou

seh

old

sT

ota

l0.0

70.2

10.0

80.0

30.0

10.1

01.2

82.0

5Poor

0.0

70.2

10.0

80.0

30.0

10.0

91.2

62.0

2N

on

poor

0.0

70.2

20.0

90.0

30.0

10.1

01.3

02.0

9M

ale-

hea

ded

0.0

80.2

50.1

00.0

30.0

10.1

11.4

72.3

6Fe

mal

e-h

ead

ed0.0

50.1

40.0

60.0

20.0

10.0

60.8

51.3

7


Tab

leA

5.1

:(g

)C

onti

nu

ed.

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

and

smal

lLe

ader

sof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

entr

ant

mer

ge

larg

est

shar

esco

mp

etit

ion

(C)

Com

ple

men

tary

pol

icie

s(fi

rms)

On

lyp

rod

uce

rsT

ota

l0.7

71.6

20.8

50.5

10.3

90.2

37.9

512.6

2Poor

0.7

31.5

50.8

10.4

90.3

70.2

27.5

912.0

6N

on

poor

0.8

11.7

10.9

00.5

40.4

10.2

58.3

913.3

3M

ale-

hea

ded

0.7

91.6

70.8

80.5

30.4

00.2

48.1

913.0

1Fe

mal

e-h

ead

ed0.6

91.4

50.7

60.4

60.3

50.2

17.1

211.3

2

All

hou

seh

old

sT

ota

l0.1

30.2

70.1

40.0

80.0

60.0

41.3

22.0

9Poor

0.1

20.2

60.1

40.0

80.0

60.0

41.2

92.0

6N

on

poor

0.1

30.2

70.1

40.0

90.0

70.0

41.3

42.1

3M

ale-

hea

ded

0.1

50.3

10.1

60.1

00.0

70.0

41.5

12.4

0Fe

mal

e-h

ead

ed0.0

80.1

80.0

90.0

60.0

40.0

30.8

81.3

9

(D)

Com

ple

men

tary

pol

icie

s(f

arm

ers−

firm

s)

On

lyp

rod

uce

rsT

ota

l1.1

92.0

51.2

80.9

40.8

10.1

98.3

413.1

2Poor

1.1

41.9

61.2

20.9

00.7

80.1

87.9

712.5

3N

on

poor

1.2

62.1

71.3

50.9

90.8

60.2

08.8

113.8

6M

ale-

hea

ded

1.2

32.1

11.3

20.9

70.8

40.2

08.6

013.5

2Fe

mal

e-h

ead

ed1.0

71.8

41.1

50.8

40.7

30.1

77.4

811.7

6

All

hou

seh

old

sT

ota

l0.2

00.3

40.2

10.1

60.1

30.0

31.3

82.1

7Poor

0.1

90.3

30.2

10.1

50.1

30.0

31.3

62.1

4N

on

poor

0.2

00.3

50.2

20.1

60.1

40.0

31.4

12.2

1M

ale-

hea

ded

0.2

30.3

90.2

40.1

80.1

50.0

41.5

82.4

9Fe

mal

e-h

ead

ed0.1

30.2

30.1

40.1

00.0

90.0

20.9

21.4

5


Tab

leA

5.1

:(g

)C

onti

nu

ed.

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

and

smal

lLe

ader

sof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

entr

ant

mer

ge

larg

est

shar

esco

mp

etit

ion

(E)

Inte

rnati

onalpri

ces

On

lyp

rod

uce

rsT

ota

l6.2

47.1

76.3

85.9

85.7

85.1

513.5

019.2

1Poor

5.9

66.8

56.1

05.7

15.5

34.9

212.9

018.3

5N

on

poor

6.5

97.5

76.7

46.3

16.1

15.4

414.2

620.2

9M

ale-

hea

ded

6.4

37.3

86.5

86.1

65.9

65.3

113.9

219.8

0Fe

mal

e-h

ead

ed5.5

96.4

25.7

25.3

65.1

94.6

212.1

117.2

2

All

hou

seh

old

sT

ota

l1.0

31.1

91.0

60.9

90.9

60.8

52.2

33.1

8Poor

1.0

21.1

71.0

40.9

70.9

40.8

42.2

03.1

3N

on

poor

1.0

51.2

11.0

81.0

10.9

80.8

72.2

83.2

4M

ale-

hea

ded

1.1

91.3

61.2

11.1

41.1

00.9

82.5

73.6

5Fe

mal

e-h

ead

ed0.6

90.7

90.7

00.6

60.6

40.5

71.4

92.1

2


Tab

leA

5.1

:(h

)C

ôte

d’Ivo

ire,

coff

ee(h

ouse

hol

din

com

eper

cen

tage

chan

ges

).

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

and

smal

lLe

ader

sof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

entr

ant

mer

ge

larg

est

shar

esco

mp

etit

ion

(A)

Basi

cm

odel

On

lyp

rod

uce

rsT

ota

l0.0

00.5

80.1

10.3

30.4

70.6

55.4

211.0

6Poor

0.0

00.6

70.1

30.3

80.5

40.7

56.2

612.7

7N

on

poor

0.0

00.4

20.0

80.2

40.3

40.4

83.9

78.1

0M

ale-

hea

ded

0.0

00.5

90.1

10.3

40.4

80.6

65.5

311.2

7Fe

mal

e-h

ead

ed0.0

00.3

50.0

70.2

00.2

90.4

03.3

16.7

5

All

hou

seh

old

sT

ota

l0.0

00.1

30.0

30.0

80.1

10.1

51.2

42.5

3Poor

0.0

00.1

90.0

40.1

10.1

50.2

11.7

43.5

6N

on

poor

0.0

00.0

70.0

10.0

40.0

60.0

80.6

91.4

2M

ale-

hea

ded

0.0

00.1

60.0

30.0

90.1

30.1

71.4

62.9

7Fe

mal

e-h

ead

ed0.0

00.0

20.0

00.0

10.0

20.0

20.2

10.4

2

(B)

Com

ple

men

tary

pol

icie

s(f

arm

ers)

On

lyp

rod

uce

rsT

ota

l0.4

91.0

80.6

10.1

70.0

20.1

65.8

911.6

5Poor

0.5

71.2

40.7

00.1

90.0

20.1

86.7

913.4

5N

on

poor

0.3

60.7

90.4

40.1

20.0

10.1

24.3

18.5

3M

ale-

hea

ded

0.5

01.1

00.6

20.1

70.0

20.1

66.0

011.8

8Fe

mal

e-h

ead

ed0.3

00.6

60.3

70.1

00.0

10.1

03.5

97.1

1

All

hou

seh

old

sT

ota

l0.1

10.2

50.1

40.0

40.0

00.0

41.3

52.6

7Poor

0.1

60.3

50.2

00.0

50.0

10.0

51.8

93.7

5N

on

poor

0.0

60.1

40.0

80.0

20.0

00.0

20.7

51.4

9M

ale-

hea

ded

0.1

30.2

90.1

60.0

40.0

10.0

41.5

83.1

3Fe

mal

e-h

ead

ed0.0

20.0

40.0

20.0

10.0

00.0

10.2

30.4

5


Tab

leA

5.1

:(h

)C

onti

nu

ed.

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

and

smal

lLe

ader

sof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

entr

ant

mer

ge

larg

est

shar

esco

mp

etit

ion

(C)

Com

ple

men

tary

pol

icie

s(fi

rms)

On

lyp

rod

uce

rsT

ota

l0.7

51.3

30.8

70.4

30.2

80.1

06.0

611.8

3Poor

0.8

71.5

41.0

10.5

00.3

20.1

26.9

913.6

6N

on

poor

0.5

50.9

70.6

40.3

10.2

00.0

74.4

38.6

6M

ale-

hea

ded

0.7

71.3

60.8

90.4

40.2

80.1

06.1

712.0

6Fe

mal

e-h

ead

ed0.4

60.8

10.5

30.2

60.1

70.0

63.7

07.2

2

All

hou

seh

old

sT

ota

l0.1

70.3

00.2

00.1

00.0

60.0

21.3

92.7

1Poor

0.2

40.4

30.2

80.1

40.0

90.0

31.9

53.8

0N

on

poor

0.1

00.1

70.1

10.0

60.0

40.0

10.7

81.5

2M

ale-

hea

ded

0.2

00.3

60.2

30.1

20.0

70.0

31.6

33.1

8Fe

mal

e-h

ead

ed0.0

30.0

50.0

30.0

20.0

10.0

00.2

30.4

5

(D)

Com

ple

men

tary

pol

icie

s(f

arm

ers−

firm

s)

On

lyp

rod

uce

rsT

ota

l1.2

41.8

21.3

60.9

20.7

60.5

96.5

212.4

3Poor

1.4

32.1

01.5

81.0

60.8

80.6

87.5

214.3

5N

on

poor

0.9

11.3

31.0

00.6

70.5

60.4

34.7

79.1

0M

ale-

hea

ded

1.2

61.8

61.3

90.9

40.7

80.6

06.6

412.6

7Fe

mal

e-h

ead

ed0.7

61.1

10.8

30.5

60.4

60.3

63.9

87.5

8

All

hou

seh

old

sT

ota

l0.2

80.4

20.3

10.2

10.1

70.1

31.4

92.8

5Poor

0.4

00.5

90.4

40.3

00.2

40.1

92.1

04.0

0N

on

poor

0.1

60.2

30.1

70.1

20.1

00.0

80.8

41.5

9M

ale-

hea

ded

0.3

30.4

90.3

70.2

50.2

00.1

61.7

53.3

4Fe

mal

e-h

ead

ed0.0

50.0

70.0

50.0

40.0

30.0

20.2

50.4

8


Tab

leA

5.1

:(h

)C

onti

nu

ed.

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

and

smal

lLe

ader

sof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

entr

ant

mer

ge

larg

est

shar

esco

mp

etit

ion

(E)

Inte

rnati

onalpri

ces

On

lyp

rod

uce

rsT

ota

l6.5

57.2

06.7

36.2

36.0

15.8

311.9

019.0

0Poor

7.5

78.3

17.7

77.1

96.9

46.7

313.7

421.9

4N

on

poor

4.8

05.2

74.9

34.5

64.4

04.2

78.7

213.9

2M

ale-

hea

ded

6.6

87.3

46.8

66.3

56.1

35.9

512.1

319.3

7Fe

mal

e-h

ead

ed4.0

04.3

94.1

13.8

03.6

73.5

67.2

611.6

0

All

hou

seh

old

sT

ota

l1.5

01.6

51.5

41.4

31.3

81.3

42.7

34.3

5Poor

2.1

12.3

12.1

72.0

01.9

31.8

83.8

36.1

1N

on

poor

0.8

40.9

20.8

60.8

00.7

70.7

51.5

32.4

4M

ale-

hea

ded

1.7

61.9

31.8

11.6

71.6

11.5

73.2

05.1

0Fe

mal

e-h

ead

ed0.2

50.2

80.2

60.2

40.2

30.2

20.4

60.7

3


Tab

leA

5.1

:(i)

Rw

an

da,c

offee

(hou

seh

old

inco

me

per

cen

tage

chan

ges

).

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

and

smal

lLe

ader

sof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

entr

ant

mer

ge

larg

est

shar

esco

mp

etit

ion

(A)

Basi

cm

odel

On

lyp

rod

uce

rsT

ota

l0.0

00.3

90.0

50.3

80.5

20.5

40.7

72.8

2Poor

0.0

00.4

90.0

60.4

70.6

40.6

70.9

63.5

1N

on

poor

0.0

00.3

40.0

40.3

30.4

50.4

70.6

72.4

7M

ale-

hea

ded

0.0

00.3

70.0

50.3

50.4

80.5

00.7

22.6

4Fe

mal

e-h

ead

ed0.0

00.4

70.0

60.4

50.6

10.6

40.9

13.3

5

All

hou

seh

old

sT

ota

l0.0

00.0

40.0

00.0

40.0

50.0

50.0

80.2

8Poor

0.0

00.0

40.0

00.0

40.0

50.0

50.0

70.2

7N

on

poor

0.0

00.0

40.0

10.0

40.0

50.0

60.0

80.2

9M

ale-

hea

ded

0.0

00.0

40.0

10.0

40.0

50.0

60.0

80.2

9Fe

mal

e-h

ead

ed0.0

00.0

40.0

00.0

30.0

50.0

50.0

70.2

6

(B)

Com

ple

men

tary

pol

icie

s(f

arm

ers)

On

lyp

rod

uce

rsT

ota

l0.1

40.5

30.2

00.2

30.3

80.4

00.9

12.9

9Poor

0.1

80.6

60.2

40.2

90.4

70.5

01.1

33.7

3N

on

poor

0.1

20.4

70.1

70.2

00.3

30.3

50.7

92.6

2M

ale-

hea

ded

0.1

30.5

00.1

80.2

20.3

50.3

70.8

52.8

0Fe

mal

e-h

ead

ed0.1

70.6

30.2

30.2

80.4

50.4

81.0

83.5

6

All

hou

seh

old

sT

ota

l0.0

10.0

50.0

20.0

20.0

40.0

40.0

90.3

0Poor

0.0

10.0

50.0

20.0

20.0

40.0

40.0

90.2

9N

on

poor

0.0

10.0

60.0

20.0

20.0

40.0

40.0

90.3

1M

ale-

hea

ded

0.0

10.0

60.0

20.0

20.0

40.0

40.0

90.3

1Fe

mal

e-h

ead

ed0.0

10.0

50.0

20.0

20.0

30.0

40.0

80.2

8


Tab

leA

5.1

:(i)

Con

tin

ued

.

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

and

smal

lLe

ader

sof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

entr

ant

mer

ge

larg

est

shar

esco

mp

etit

ion

(C)

Com

ple

men

tary

pol

icie

s(fi

rms)

On

lyp

rod

uce

rsT

ota

l0.2

40.6

30.3

00.1

30.2

90.3

10.9

93.0

9Poor

0.2

90.7

80.3

70.1

60.3

60.3

81.2

33.8

5N

on

poor

0.2

10.5

50.2

60.1

20.2

50.2

70.8

72.7

1M

ale-

hea

ded

0.2

20.5

90.2

80.1

20.2

70.2

90.9

32.9

0Fe

mal

e-h

ead

ed0.2

80.7

50.3

50.1

60.3

40.3

71.1

83.6

8

All

hou

seh

old

sT

ota

l0.0

20.0

60.0

30.0

10.0

30.0

30.1

00.3

1Poor

0.0

20.0

60.0

30.0

10.0

30.0

30.0

90.2

9N

on

poor

0.0

20.0

70.0

30.0

10.0

30.0

30.1

00.3

2M

ale-

hea

ded

0.0

20.0

70.0

30.0

10.0

30.0

30.1

00.3

2Fe

mal

e-h

ead

ed0.0

20.0

60.0

30.0

10.0

30.0

30.0

90.2

8

(D)

Com

ple

men

tary

pol

icie

s(f

arm

ers−

firm

s)

On

lyp

rod

uce

rsT

ota

l0.3

80.7

70.4

40.0

10.1

50.1

71.1

23.2

7Poor

0.4

70.9

60.5

50.0

10.1

90.2

11.4

04.0

7N

on

poor

0.3

30.6

80.3

90.0

10.1

30.1

50.9

82.8

6M

ale-

hea

ded

0.3

50.7

20.4

10.0

10.1

40.1

61.0

53.0

6Fe

mal

e-h

ead

ed0.4

50.9

20.5

30.0

10.1

80.2

11.3

43.8

9

All

hou

seh

old

sT

ota

l0.0

40.0

80.0

40.0

00.0

20.0

20.1

10.3

3Poor

0.0

40.0

70.0

40.0

00.0

10.0

20.1

10.3

1N

on

poor

0.0

40.0

80.0

50.0

00.0

20.0

20.1

20.3

4M

ale-

hea

ded

0.0

40.0

80.0

50.0

00.0

20.0

20.1

20.3

4Fe

mal

e-h

ead

ed0.0

30.0

70.0

40.0

00.0

10.0

20.1

00.3

0


Tab

leA

5.1

:(i)

Con

tin

ued

.

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

and

smal

lLe

ader

sof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

entr

ant

mer

ge

larg

est

shar

esco

mp

etit

ion

(E)

Inte

rnati

onalpri

ces

On

lyp

rod

uce

rsT

ota

l2.0

22.4

82.1

41.6

51.4

11.3

92.7

85.3

6Poor

2.5

23.0

92.6

72.0

61.7

61.7

33.4

76.6

7N

on

poor

1.7

72.1

71.8

81.4

51.2

41.2

22.4

44.6

9M

ale-

hea

ded

1.8

92.3

22.0

11.5

51.3

21.3

02.6

15.0

2Fe

mal

e-h

ead

ed2.4

12.9

52.5

51.9

61.6

81.6

63.3

16.3

7

All

hou

seh

old

sT

ota

l0.2

00.2

50.2

10.1

60.1

40.1

40.2

80.5

3Poor

0.1

90.2

40.2

00.1

60.1

30.1

30.2

70.5

1N

on

poor

0.2

10.2

60.2

20.1

70.1

50.1

40.2

90.5

5M

ale-

hea

ded

0.2

10.2

60.2

20.1

70.1

50.1

40.2

90.5

5Fe

mal

e-h

ead

ed0.1

90.2

30.2

00.1

50.1

30.1

30.2

60.4

9


Tab

leA

5.1

:(j

)U

gan

da,c

offee

(hou

seh

old

inco

me

per

cen

tage

chan

ges

).

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

and

smal

lLe

ader

sof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

entr

ant

mer

ge

larg

est

shar

esco

mp

etit

ion

(A)

Basi

cm

odel

On

lyp

rod

uce

rsT

ota

l0.0

00.0

90.0

10.0

60.0

80.1

00.8

11.5

9Poor

0.0

00.0

90.0

10.0

60.0

80.1

00.8

11.5

9N

on

poor

0.0

00.0

90.0

10.0

60.0

80.1

00.8

11.5

9M

ale-

hea

ded

0.0

00.0

90.0

10.0

60.0

80.1

00.8

71.7

0Fe

mal

e-h

ead

ed0.0

00.0

70.0

10.0

50.0

60.0

80.6

41.2

5

All

hou

seh

old

sT

ota

l0.0

00.0

20.0

00.0

20.0

20.0

30.2

20.4

3Poor

0.0

00.0

20.0

00.0

10.0

20.0

20.1

70.3

3N

on

poor

0.0

00.0

30.0

00.0

20.0

20.0

30.2

60.5

1M

ale-

hea

ded

0.0

00.0

30.0

00.0

20.0

20.0

30.2

40.4

7Fe

mal

e-h

ead

ed0.0

00.0

20.0

00.0

10.0

10.0

20.1

60.3

0

(B)

Com

ple

men

tary

pol

icie

s(f

arm

ers)

On

lyp

rod

uce

rsT

ota

l0.1

60.2

50.1

70.1

00.0

80.0

60.9

71.7

7Poor

0.1

60.2

50.1

70.1

00.0

80.0

60.9

71.7

7N

on

poor

0.1

60.2

50.1

70.1

00.0

80.0

60.9

71.7

7M

ale-

hea

ded

0.1

70.2

60.1

90.1

10.0

90.0

71.0

41.8

9Fe

mal

e-h

ead

ed0.1

30.1

90.1

40.0

80.0

70.0

50.7

61.3

9

All

hou

seh

old

sT

ota

l0.0

40.0

70.0

50.0

30.0

20.0

20.2

60.4

8Poor

0.0

30.0

50.0

40.0

20.0

20.0

10.2

00.3

7N

on

poor

0.0

50.0

80.0

60.0

30.0

30.0

20.3

10.5

6M

ale-

hea

ded

0.0

50.0

70.0

50.0

30.0

30.0

20.2

90.5

3Fe

mal

e-h

ead

ed0.0

30.0

50.0

30.0

20.0

20.0

10.1

90.3

4


Tab

leA

5.1

:(j

)C

onti

nu

ed.

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

and

smal

lLe

ader

sof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

entr

ant

mer

ge

larg

est

shar

esco

mp

etit

ion

(C)

Com

ple

men

tary

pol

icie

s(fi

rms)

On

lyp

rod

uce

rsT

ota

l0.2

50.3

40.2

60.1

90.1

70.1

51.0

51.8

5Poor

0.2

50.3

40.2

60.1

90.1

70.1

51.0

51.8

5N

on

poor

0.2

50.3

40.2

60.1

90.1

70.1

51.0

51.8

5M

ale-

hea

ded

0.2

70.3

60.2

80.2

00.1

80.1

61.1

21.9

7Fe

mal

e-h

ead

ed0.2

00.2

60.2

10.1

50.1

40.1

20.8

21.4

5

All

hou

seh

old

sT

ota

l0.0

70.0

90.0

70.0

50.0

50.0

40.2

80.5

0Poor

0.0

50.0

70.0

50.0

40.0

40.0

30.2

20.3

8N

on

poor

0.0

80.1

10.0

80.0

60.0

50.0

50.3

30.5

9M

ale-

hea

ded

0.0

70.1

00.0

80.0

60.0

50.0

50.3

10.5

5Fe

mal

e-h

ead

ed0.0

50.0

60.0

50.0

40.0

30.0

30.2

00.3

5

(D)

Com

ple

men

tary

pol

icie

s(f

arm

ers−

firm

s)

On

lyp

rod

uce

rsT

ota

l0.4

10.5

00.4

30.3

50.3

30.3

11.2

12.0

2Poor

0.4

10.5

00.4

30.3

50.3

30.3

11.2

12.0

2N

on

poor

0.4

10.5

00.4

30.3

50.3

30.3

11.2

12.0

2M

ale-

hea

ded

0.4

40.5

30.4

50.3

80.3

60.3

31.2

92.1

6Fe

mal

e-h

ead

ed0.3

20.3

90.3

30.2

80.2

60.2

50.9

51.5

9

All

hou

seh

old

sT

ota

l0.1

10.1

30.1

10.1

00.0

90.0

80.3

30.5

5Poor

0.0

90.1

00.0

90.0

70.0

70.0

60.2

50.4

2N

on

poor

0.1

30.1

60.1

40.1

10.1

10.1

00.3

80.6

4M

ale-

hea

ded

0.1

20.1

50.1

30.1

10.1

00.0

90.3

60.6

0Fe

mal

e-h

ead

ed0.0

80.0

90.0

80.0

70.0

60.0

60.2

30.3

9


Tab

leA

5.1

:(j

)C

onti

nu

ed.

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

and

smal

lLe

ader

sof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

entr

ant

mer

ge

larg

est

shar

esco

mp

etit

ion

(E)

Inte

rnati

onalpri

ces

On

lyp

rod

uce

rsT

ota

l2.1

22.2

12.1

42.0

62.0

32.0

12.9

33.9

2Poor

2.1

22.2

12.1

42.0

62.0

32.0

12.9

33.9

2N

on

poor

2.1

22.2

12.1

42.0

62.0

32.0

12.9

33.9

2M

ale-

hea

ded

2.2

62.3

62.2

82.2

02.1

62.1

43.1

24.1

8Fe

mal

e-h

ead

ed1.6

61.7

41.6

81.6

11.5

91.5

82.3

03.0

7

All

hou

seh

old

sT

ota

l0.5

70.6

00.5

80.5

50.5

50.5

40.7

91.0

6Poor

0.4

40.4

60.4

40.4

30.4

20.4

20.6

10.8

1N

on

poor

0.6

70.7

00.6

80.6

50.6

40.6

40.9

31.2

4M

ale-

hea

ded

0.6

30.6

60.6

40.6

10.6

00.6

00.8

71.1

7Fe

mal

e-h

ead

ed0.4

00.4

20.4

10.3

90.3

90.3

80.5

60.7

5


Tab

leA

5.1

:(k

)M

ala

wi,

tobacc

o(h

ouse

hol

din

com

eper

cen

tage

chan

ges

).

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

and

smal

lLe

ader

sof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

entr

ant

mer

ge

larg

est

shar

esco

mp

etit

ion

(A)

Basi

cm

odel

On

lyp

rod

uce

rsT

ota

l0.0

01.4

50.0

41.1

61.3

61.5

95.6

812.2

4Poor

0.0

01.1

30.0

30.9

11.0

61.2

54.4

49.5

7N

on

poor

0.0

01.6

00.0

41.2

81.5

01.7

66.2

613.5

0M

ale-

hea

ded

0.0

01.4

70.0

41.1

81.3

81.6

15.7

512.4

0Fe

mal

e-h

ead

ed0.0

01.2

70.0

31.0

21.2

01.4

05.0

010.7

8

All

hou

seh

old

sT

ota

l0.0

00.2

10.0

10.1

70.2

00.2

30.8

21.7

6Poor

0.0

00.1

20.0

00.0

90.1

10.1

30.4

60.9

9N

on

poor

0.0

00.2

80.0

10.2

30.2

60.3

11.1

02.3

8M

ale-

hea

ded

0.0

00.2

50.0

10.2

00.2

30.2

80.9

82.1

2Fe

mal

e-h

ead

ed0.0

00.0

80.0

00.0

60.0

70.0

80.3

00.6

4

(B)

Com

ple

men

tary

pol

icie

s(f

arm

ers)

On

lyp

rod

uce

rsT

ota

l0.4

21.8

50.4

70.7

30.9

41.1

76.0

712.7

8Poor

0.3

31.4

50.3

70.5

70.7

30.9

24.7

59.9

9N

on

poor

0.4

72.0

40.5

20.8

01.0

41.2

96.6

914.0

8M

ale-

hea

ded

0.4

31.8

70.4

80.7

40.9

51.1

96.1

512.9

4Fe

mal

e-h

ead

ed0.3

71.6

30.4

20.6

40.8

31.0

35.3

511.2

5

All

hou

seh

old

sT

ota

l0.0

60.2

70.0

70.1

00.1

40.1

70.8

71.8

4Poor

0.0

30.1

50.0

40.0

60.0

80.1

00.4

91.0

4N

on

poor

0.0

80.3

60.0

90.1

40.1

80.2

31.1

82.4

8M

ale-

hea

ded

0.0

70.3

20.0

80.1

30.1

60.2

01.0

52.2

1Fe

mal

e-h

ead

ed0.0

20.1

00.0

20.0

40.0

50.0

60.3

20.6

7


Tab

leA

5.1

:(k

)C

onti

nu

ed.

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

and

smal

lLe

ader

sof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

entr

ant

mer

ge

larg

est

shar

esco

mp

etit

ion

(C)

Com

ple

men

tary

pol

icie

s(fi

rms)

On

lyp

rod

uce

rsT

ota

l0.7

02.0

80.7

70.4

30.6

60.8

96.2

612.9

8Poor

0.5

51.6

30.6

00.3

40.5

20.7

04.9

010.1

5N

on

poor

0.7

72.3

00.8

50.4

80.7

30.9

86.9

014.3

1M

ale-

hea

ded

0.7

12.1

10.7

80.4

40.6

70.9

06.3

413.1

4Fe

mal

e-h

ead

ed0.6

21.8

30.6

70.3

80.5

80.7

95.5

111.4

3

All

hou

seh

old

sT

ota

l0.1

00.3

00.1

10.0

60.1

00.1

30.9

01.8

7Poor

0.0

60.1

70.0

60.0

40.0

50.0

70.5

11.0

5N

on

poor

0.1

40.4

00.1

50.0

80.1

30.1

71.2

22.5

2M

ale-

hea

ded

0.1

20.3

60.1

30.0

70.1

10.1

51.0

82.2

4Fe

mal

e-h

ead

ed0.0

40.1

10.0

40.0

20.0

30.0

50.3

30.6

8

(D)

Com

ple

men

tary

pol

icie

s(f

arm

ers−

firm

s)

On

lyp

rod

uce

rsT

ota

l1.1

22.4

81.2

00.0

10.2

50.4

86.6

513.5

1Poor

0.8

81.9

40.9

40.0

00.1

90.3

75.2

010.5

6N

on

poor

1.2

42.7

41.3

20.0

10.2

70.5

27.3

314.8

9M

ale-

hea

ded

1.1

42.5

11.2

10.0

10.2

50.4

86.7

413.6

8Fe

mal

e-h

ead

ed0.9

92.1

91.0

60.0

10.2

20.4

25.8

611.8

9

All

hou

seh

old

sT

ota

l0.1

60.3

60.1

70.0

00.0

40.0

70.9

61.9

4Poor

0.0

90.2

00.1

00.0

00.0

20.0

40.5

41.0

9N

on

poor

0.2

20.4

80.2

30.0

00.0

50.0

91.2

92.6

2M

ale-

hea

ded

0.1

90.4

30.2

10.0

00.0

40.0

81.1

52.3

3Fe

mal

e-h

ead

ed0.0

60.1

30.0

60.0

00.0

10.0

20.3

50.7

1


Tab

leA

5.1

:(k

)C

onti

nu

ed.

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

and

smal

lLe

ader

sof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

entr

ant

mer

ge

larg

est

shar

esco

mp

etit

ion

(E)

Inte

rnati

onalpri

ces

On

lyp

rod

uce

rsT

ota

l6.0

37.4

36.2

24.8

84.4

94.2

611.6

419.7

8Poor

4.7

15.8

14.8

63.8

23.5

13.3

39.1

015.4

7N

on

poor

6.6

48.1

96.8

55.3

84.9

54.7

012.8

321.8

1M

ale-

hea

ded

6.1

07.5

26.3

04.9

44.5

54.3

211.7

920.0

4Fe

mal

e-h

ead

ed5.3

16.5

45.4

74.3

03.9

63.7

510.2

517.4

2

All

hou

seh

old

sT

ota

l0.8

71.0

70.8

90.7

00.6

50.6

11.6

82.8

5Poor

0.4

90.6

00.5

00.4

00.3

60.3

50.9

41.6

0N

on

poor

1.1

71.4

41.2

10.9

50.8

70.8

32.2

63.8

4M

ale-

hea

ded

1.0

41.2

81.0

70.8

40.7

80.7

42.0

13.4

2Fe

mal

e-h

ead

ed0.3

10.3

90.3

20.2

50.2

30.2

20.6

11.0

3


Tab

leA

5.1

:(l)

Zam

bia

,tob

acc

o(h

ouse

hol

din

com

eper

cen

tage

chan

ges

).

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

and

smal

lLe

ader

sof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

entr

ant

mer

ge

larg

est

shar

esco

mp

etit

ion

(A)

Basi

cm

odel

On

lyp

rod

uce

rsT

ota

l0.0

03.2

50.7

41.1

62.4

84.6

97.2

321.7

5Poor

0.0

02.5

20.5

80.9

01.9

23.6

45.6

116.8

6N

on

poor

0.0

03.7

10.8

51.3

22.8

35.3

68.2

624.8

4M

ale-

hea

ded

0.0

03.3

70.7

71.2

02.5

64.8

67.4

922.5

3Fe

mal

e-h

ead

ed0.0

02.5

10.5

70.9

01.9

13.6

35.5

916.8

2

All

hou

seh

old

sT

ota

l0.0

00.0

70.0

20.0

30.0

50.1

00.1

60.4

7Poor

0.0

00.0

50.0

10.0

20.0

40.0

70.1

10.3

2N

on

poor

0.0

00.0

90.0

20.0

30.0

70.1

30.2

00.5

9M

ale-

hea

ded

0.0

00.0

80.0

20.0

30.0

60.1

10.1

70.5

2Fe

mal

e-h

ead

ed0.0

00.0

40.0

10.0

10.0

30.0

50.0

80.2

5

(B)

Com

ple

men

tary

pol

icie

s(f

arm

ers)

On

lyp

rod

uce

rsT

ota

l0.5

03.7

61.2

60.6

51.9

94.1

97.6

822.4

3Poor

0.3

92.9

10.9

80.5

11.5

53.2

55.9

517.3

9N

on

poor

0.5

74.2

91.4

40.7

52.2

84.7

98.7

725.6

1M

ale-

hea

ded

0.5

23.8

91.3

10.6

82.0

74.3

47.9

523.2

3Fe

mal

e-h

ead

ed0.3

92.9

00.9

80.5

01.5

43.2

45.9

417.3

4

All

hou

seh

old

sT

ota

l0.0

10.0

80.0

30.0

10.0

40.0

90.1

70.4

8Poor

0.0

10.0

50.0

20.0

10.0

30.0

60.1

10.3

3N

on

poor

0.0

10.1

00.0

30.0

20.0

50.1

10.2

10.6

1M

ale-

hea

ded

0.0

10.0

90.0

30.0

20.0

50.1

00.1

80.5

4Fe

mal

e-h

ead

ed0.0

10.0

40.0

10.0

10.0

20.0

50.0

90.2

6


Tab

leA

5.1

:(l)

Con

tin

ued

.

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

and

smal

lLe

ader

sof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

entr

ant

mer

ge

larg

est

shar

esco

mp

etit

ion

(C)

Com

ple

men

tary

pol

icie

s(fi

rms)

On

lyp

rod

uce

rsT

ota

l0.4

53.7

01.2

40.6

82.0

64.2

37.5

322.1

9Poor

0.3

52.8

70.9

60.5

31.6

03.2

85.8

417.2

0N

on

poor

0.5

24.2

21.4

20.7

82.3

54.8

38.6

025.3

3M

ale-

hea

ded

0.4

73.8

31.2

90.7

12.1

34.3

87.8

022.9

8Fe

mal

e-h

ead

ed0.3

52.8

60.9

60.5

31.5

93.2

75.8

217.1

6

All

hou

seh

old

sT

ota

l0.0

10.0

80.0

30.0

10.0

40.0

90.1

60.4

8Poor

0.0

10.0

50.0

20.0

10.0

30.0

60.1

10.3

2N

on

poor

0.0

10.1

00.0

30.0

20.0

60.1

10.2

00.6

0M

ale-

hea

ded

0.0

10.0

90.0

30.0

20.0

50.1

00.1

80.5

3Fe

mal

e-h

ead

ed0.0

10.0

40.0

10.0

10.0

20.0

50.0

90.2

6

(D)

Com

ple

men

tary

pol

icie

s(f

arm

ers−

firm

s)

On

lyp

rod

uce

rsT

ota

l0.9

54.2

01.7

60.1

81.5

83.7

37.9

822.8

6Poor

0.7

43.2

61.3

60.1

41.2

32.8

96.1

817.7

3N

on

poor

1.0

84.8

02.0

10.2

01.8

14.2

69.1

126.1

0M

ale-

hea

ded

0.9

84.3

51.8

20.1

91.6

43.8

78.2

623.6

8Fe

mal

e-h

ead

ed0.7

33.2

51.3

60.1

41.2

22.8

96.1

717.6

8

All

hou

seh

old

sT

ota

l0.0

20.0

90.0

40.0

00.0

30.0

80.1

70.4

9Poor

0.0

10.0

60.0

30.0

00.0

20.0

50.1

20.3

3N

on

poor

0.0

30.1

10.0

50.0

00.0

40.1

00.2

20.6

2M

ale-

hea

ded

0.0

20.1

00.0

40.0

00.0

40.0

90.1

90.5

5Fe

mal

e-h

ead

ed0.0

10.0

50.0

20.0

00.0

20.0

40.0

90.2

6


Tab

leA

5.1

:(l)

Con

tin

ued

.

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

and

smal

lLe

ader

sof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

entr

ant

mer

ge

larg

est

shar

esco

mp

etit

ion

(E)

Inte

rnati

onalpri

ces

On

lyp

rod

uce

rsT

ota

l5.4

28.9

86.4

94.2

72.4

90.3

012.5

729.4

8Poor

4.2

06.9

65.0

33.3

11.9

30.2

49.7

422.8

5N

on

poor

6.1

910.2

57.4

14.8

82.8

40.3

514.3

533.6

6M

ale-

hea

ded

5.6

19.3

06.7

34.4

22.5

80.3

213.0

230.5

3Fe

mal

e-h

ead

ed4.1

96.9

45.0

23.3

01.9

20.2

49.7

222.7

9

All

hou

seh

old

sT

ota

l0.1

20.1

90.1

40.0

90.0

50.0

10.2

70.6

4Poor

0.0

80.1

30.0

90.0

60.0

40.0

00.1

80.4

3N

on

poor

0.1

50.2

40.1

80.1

20.0

70.0

10.3

40.8

0M

ale-

hea

ded

0.1

30.2

20.1

60.1

00.0

60.0

10.3

00.7

1Fe

mal

e-h

ead

ed0.0

60.1

00.0

80.0

50.0

30.0

00.1

50.3

4


Tab

leA

5.2

:Im

pact

ofsu

pply

chain

chan

ges

onth

eou

tgro

wer

con

tract

mod

el:

(a)

Zam

bia

,cot

ton

(hou

seh

old

inco

me

per

cen

tage

chan

ges

).

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

and

smal

lLe

ader

sof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

entr

ant

mer

ge

larg

est

shar

esco

mp

etit

ion

(A)

Basi

cm

odel

On

lyp

rod

uce

rsT

ota

l0.0

02.2

60.1

41.5

71.8

93.2

26.4

618.4

1Poor

0.0

02.4

50.1

51.7

12.0

53.4

97.0

119.9

6N

on

poor

0.0

02.1

50.1

31.5

01.8

03.0

76.1

617.5

5M

ale-

hea

ded

0.0

02.2

70.1

41.5

81.9

03.2

36.4

918.4

8Fe

mal

e-h

ead

ed0.0

02.2

10.1

41.5

41.8

53.1

56.3

218.0

0

All

hou

seh

old

sT

ota

l0.0

00.2

50.0

20.1

70.2

10.3

60.7

12.0

3Poor

0.0

00.2

20.0

10.1

50.1

80.3

10.6

21.7

7N

on

poor

0.0

00.2

70.0

20.1

90.2

30.3

90.7

92.2

4M

ale-

hea

ded

0.0

00.2

70.0

20.1

90.2

20.3

80.7

62.1

8Fe

mal

e-h

ead

ed0.0

00.1

80.0

10.1

20.1

50.2

50.5

11.4

4

(B)

Com

ple

men

tary

pol

icie

s(f

arm

ers)

On

lyp

rod

uce

rsT

ota

l0.3

62.6

00.5

11.2

11.5

32.8

46.7

918.9

5Poor

0.3

92.8

20.5

51.3

11.6

63.0

87.3

620.5

4N

on

poor

0.3

42.4

80.4

81.1

51.4

62.7

16.4

718.0

7M

ale-

hea

ded

0.3

62.6

10.5

11.2

11.5

42.8

56.8

219.0

2Fe

mal

e-h

ead

ed0.3

52.5

40.5

01.1

81.5

02.7

86.6

418.5

2

All

hou

seh

old

sT

ota

l0.0

40.2

90.0

60.1

30.1

70.3

10.7

52.0

9Poor

0.0

30.2

50.0

50.1

20.1

50.2

70.6

51.8

2N

on

poor

0.0

40.3

20.0

60.1

50.1

90.3

50.8

32.3

1M

ale-

hea

ded

0.0

40.3

10.0

60.1

40.1

80.3

40.8

02.2

4Fe

mal

e-h

ead

ed0.0

30.2

00.0

40.0

90.1

20.2

20.5

31.4

8


Tab

leA

5.2

:(a

)C

onti

nu

ed.

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

and

smal

lLe

ader

sof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

entr

ant

mer

ge

larg

est

shar

esco

mp

etit

ion

(C)

Com

ple

men

tary

pol

icie

s(fi

rms)

On

lyp

rod

uce

rsT

ota

l0.5

42.7

80.7

11.0

11.3

82.6

46.9

319.0

9Poor

0.5

93.0

20.7

71.1

01.4

92.8

77.5

120.6

9N

on

poor

0.5

22.6

50.6

80.9

71.3

12.5

26.6

018.2

0M

ale-

hea

ded

0.5

42.7

90.7

21.0

21.3

82.6

56.9

519.1

6Fe

mal

e-h

ead

ed0.5

32.7

20.7

00.9

91.3

52.5

96.7

718.6

6

All

hou

seh

old

sT

ota

l0.0

60.3

10.0

80.1

10.1

50.2

90.7

62.1

1Poor

0.0

50.2

70.0

70.1

00.1

30.2

50.6

71.8

4N

on

poor

0.0

70.3

40.0

90.1

20.1

70.3

20.8

42.3

2M

ale-

hea

ded

0.0

60.3

30.0

80.1

20.1

60.3

10.8

22.2

6Fe

mal

e-h

ead

ed0.0

40.2

20.0

60.0

80.1

10.2

10.5

41.5

0

(D)

Com

ple

men

tary

pol

icie

s(f

arm

ers−

firm

s)

On

lyp

rod

uce

rsT

ota

l0.9

03.1

31.0

80.6

51.0

22.2

77.2

519.6

3Poor

0.9

73.4

01.1

70.7

01.1

12.4

67.8

621.2

8N

on

poor

0.8

62.9

91.0

30.6

20.9

82.1

76.9

118.7

1M

ale-

hea

ded

0.9

03.1

51.0

90.6

51.0

32.2

87.2

819.7

0Fe

mal

e-h

ead

ed0.8

83.0

61.0

60.6

31.0

02.2

27.0

919.1

8

All

hou

seh

old

sT

ota

l0.1

00.3

50.1

20.0

70.1

10.2

50.8

02.1

7Poor

0.0

90.3

00.1

00.0

60.1

00.2

20.7

01.8

9N

on

poor

0.1

10.3

80.1

30.0

80.1

20.2

80.8

82.3

9M

ale-

hea

ded

0.1

10.3

70.1

30.0

80.1

20.2

70.8

62.3

2Fe

mal

e-h

ead

ed0.0

70.2

50.0

80.0

50.0

80.1

80.5

71.5

4


Tab

leA

5.2

:(a

)C

onti

nu

ed.

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

and

smal

lLe

ader

sof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

entr

ant

mer

ge

larg

est

shar

esco

mp

etit

ion

(E)

Inte

rnati

onalpri

ces

On

lyp

rod

uce

rsT

ota

l5.2

57.7

25.6

23.6

52.9

91.8

511.9

326.4

1Poor

5.6

98.3

76.1

03.9

63.2

42.0

112.9

328.6

3N

on

poor

5.0

17.3

65.3

63.4

82.8

51.7

711.3

725.1

8M

ale-

hea

ded

5.2

77.7

55.6

43.6

63.0

01.8

611.9

726.5

1Fe

mal

e-h

ead

ed5.1

37.5

45.5

03.5

72.9

21.8

111.6

625.8

2

All

hou

seh

old

sT

ota

l0.5

80.8

50.6

20.4

00.3

30.2

01.3

22.9

1Poor

0.5

00.7

40.5

40.3

50.2

90.1

81.1

52.5

4N

on

poor

0.6

40.9

40.6

80.4

40.3

60.2

31.4

53.2

1M

ale-

hea

ded

0.6

20.9

10.6

60.4

30.3

50.2

21.4

13.1

2Fe

mal

e-h

ead

ed0.4

10.6

00.4

40.2

90.2

30.1

50.9

32.0

7


Tab

leA

5.2

:(b

)Ben

in,c

otto

n(p

erce

nta

ge)

chan

ges

inh

ouse

hol

din

com

e.

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

and

smal

lLe

ader

sof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

entr

ant

mer

ge

larg

est

shar

esco

mp

etit

ion

(A)

Basi

cm

odel

On

lyp

rod

uce

rsT

ota

l0.0

03.8

00.0

40.3

30.4

15.5

318.1

232.1

3Poor

0.0

03.7

10.0

40.3

20.4

05.4

017.7

031.3

7N

on

poor

0.0

03.9

70.0

40.3

40.4

35.7

718.9

233.5

3M

ale-

hea

ded

0.0

02.4

50.0

30.2

10.2

63.5

611.6

820.7

1Fe

mal

e-h

ead

ed0.0

03.8

60.0

40.3

30.4

25.6

118.4

032.6

1

All

hou

seh

old

sT

ota

l0.0

00.7

10.0

10.0

60.0

81.0

43.4

06.0

3Poor

0.0

00.9

00.0

10.0

80.1

01.3

14.2

87.5

9N

on

poor

0.0

00.5

30.0

10.0

50.0

60.7

72.5

24.4

6M

ale-

hea

ded

0.0

00.1

20.0

00.0

10.0

10.1

80.5

91.0

5Fe

mal

e-h

ead

ed0.0

00.8

20.0

10.0

70.0

91.1

93.9

16.9

2

(B)

Com

ple

men

tary

pol

icie

s(f

arm

ers)

On

lyp

rod

uce

rsT

ota

l0.4

74.2

30.5

20.1

40.0

54.9

818.5

432.8

3Poor

0.4

64.1

30.5

10.1

40.0

54.8

618.1

032.0

5N

on

poor

0.4

94.4

10.5

50.1

50.0

55.2

019.3

534.2

6M

ale-

hea

ded

0.3

02.7

30.3

40.0

90.0

33.2

111.9

521.1

6Fe

mal

e-h

ead

ed0.4

84.2

90.5

30.1

40.0

55.0

518.8

233.3

2

All

hou

seh

old

sT

ota

l0.0

90.7

90.1

00.0

30.0

10.9

43.4

86.1

7Poor

0.1

11.0

00.1

20.0

30.0

11.1

84.3

87.7

6N

on

poor

0.0

70.5

90.0

70.0

20.0

10.6

92.5

74.5

6M

ale-

hea

ded

0.0

20.1

40.0

20.0

00.0

00.1

60.6

01.0

7Fe

mal

e-h

ead

ed0.1

00.9

10.1

10.0

30.0

11.0

73.9

97.0

7


Tab

leA

5.2

:(b

)C

onti

nu

ed.

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

and

smal

lLe

ader

sof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

entr

ant

mer

ge

larg

est

shar

esco

mp

etit

ion

(C)

Com

ple

men

tary

pol

icie

s(fi

rms)

On

lyp

rod

uce

rsT

ota

l0.4

74.1

70.5

30.1

30.0

34.8

618.3

432.4

1Poor

0.4

64.0

70.5

10.1

30.0

34.7

517.9

131.6

5N

on

poor

0.4

94.3

50.5

50.1

40.0

35.0

819.1

433.8

3M

ale-

hea

ded

0.3

02.6

90.3

40.0

90.0

23.1

311.8

220.9

0Fe

mal

e-h

ead

ed0.4

74.2

30.5

30.1

30.0

34.9

418.6

232.9

0

All

hou

seh

old

sT

ota

l0.0

90.7

80.1

00.0

20.0

10.9

13.4

46.0

9Poor

0.1

10.9

90.1

20.0

30.0

11.1

54.3

37.6

6N

on

poor

0.0

60.5

80.0

70.0

20.0

00.6

82.5

54.5

0M

ale-

hea

ded

0.0

20.1

40.0

20.0

00.0

00.1

60.6

01.0

6Fe

mal

e-h

ead

ed0.1

00.9

00.1

10.0

30.0

11.0

53.9

56.9

9

(D)

Com

ple

men

tary

pol

icie

s(f

arm

ers−

firm

s)

On

lyp

rod

uce

rsT

ota

l0.9

44.6

01.0

10.6

00.4

94.3

218.7

533.1

2Poor

0.9

24.5

00.9

90.5

90.4

84.2

118.3

132.3

4N

on

poor

0.9

84.8

01.0

50.6

30.5

14.5

019.5

734.5

6M

ale-

hea

ded

0.6

12.9

70.6

50.3

90.3

12.7

812.0

921.3

5Fe

mal

e-h

ead

ed0.9

64.6

71.0

20.6

10.4

94.3

819.0

333.6

1

All

hou

seh

old

sT

ota

l0.1

80.8

60.1

90.1

10.0

90.8

13.5

26.2

2Poor

0.2

21.0

90.2

40.1

40.1

21.0

24.4

37.8

2N

on

poor

0.1

30.6

40.1

40.0

80.0

70.6

02.6

04.6

0M

ale-

hea

ded

0.0

30.1

50.0

30.0

20.0

20.1

40.6

11.0

8Fe

mal

e-h

ead

ed0.2

00.9

90.2

20.1

30.1

00.9

34.0

47.1

4


Tab

leA

5.2

:(b

)C

onti

nu

ed.

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

and

smal

lLe

ader

sof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

entr

ant

mer

ge

larg

est

shar

esco

mp

etit

ion

(E)

Inte

rnati

onalpri

ces

On

lyp

rod

uce

rsT

ota

l5.7

49.3

55.9

15.3

55.1

00.7

424.3

540.4

9Poor

5.6

19.1

35.7

75.2

24.9

80.7

223.7

839.5

3N

on

poor

5.9

99.7

66.1

75.5

85.3

20.7

725.4

242.2

6M

ale-

hea

ded

3.7

06.0

33.8

13.4

53.2

90.4

715.7

026.1

0Fe

mal

e-h

ead

ed5.8

39.4

96.0

05.4

35.1

80.7

524.7

241.1

0

All

hou

seh

old

sT

ota

l1.0

81.7

61.1

11.0

00.9

60.1

44.5

77.6

1Poor

1.3

62.2

11.4

01.2

61.2

00.1

75.7

59.5

7N

on

poor

0.8

01.3

00.8

20.7

40.7

10.1

03.3

85.6

2M

ale-

hea

ded

0.1

90.3

00.1

90.1

70.1

70.0

20.7

91.3

2Fe

mal

e-h

ead

ed1.2

42.0

21.2

71.1

51.1

00.1

65.2

58.7

3


Tab

leA

5.2

:(c

)Bu

rkin

aFa

so,c

otto

n(h

ouse

hol

din

com

eper

cen

tage

chan

ges

).

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

and

smal

lLe

ader

sof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

entr

ant

mer

ge

larg

est

shar

esco

mp

etit

ion

(A)

Basi

cm

odel

On

lyp

rod

uce

rsT

ota

l0.0

06.5

10.3

80.3

51.2

6−2

1.7

89.8

738.4

3Poor

0.0

06.2

50.3

60.3

41.2

0−2

0.8

99.4

736.8

6N

on

poor

0.0

06.6

80.3

90.3

61.2

9−2

2.3

310.1

239.4

1M

ale-

hea

ded

0.0

06.6

10.3

80.3

61.2

8−2

2.1

210.0

339.0

3Fe

mal

e-h

ead

ed0.0

02.0

80.1

20.1

10.4

06.9

63.1

612.2

9

All

hou

seh

old

sT

ota

l0.0

00.1

50.0

10.0

10.0

30.5

20.2

30.9

1Poor

0.0

00.1

20.0

10.0

10.0

20.3

90.1

80.7

0N

on

poor

0.0

00.1

90.0

10.0

10.0

40.6

30.2

91.1

1M

ale-

hea

ded

0.0

00.1

60.0

10.0

10.0

30.5

50.2

50.9

6Fe

mal

e-h

ead

ed0.0

00.0

20.0

00.0

00.0

00.0

60.0

30.1

0

(B)

Com

ple

men

tary

pol

icie

s(f

arm

ers)

On

lyp

rod

uce

rsT

ota

l0.3

76.9

40.0

30.6

91.6

4−2

0.9

710.4

539.5

4Poor

0.3

66.6

60.0

30.6

71.5

7−2

0.1

210.0

337.9

3N

on

poor

0.3

87.1

20.0

30.7

11.6

8−2

1.5

110.7

240.5

5M

ale-

hea

ded

0.3

87.0

50.0

30.7

01.6

6−2

1.3

010.6

240.1

6Fe

mal

e-h

ead

ed0.1

22.2

20.0

10.2

20.5

26.7

03.3

412.6

4

All

hou

seh

old

sT

ota

l0.0

10.1

60.0

00.0

20.0

40.5

00.2

50.9

4Poor

0.0

10.1

30.0

00.0

10.0

30.3

80.1

90.7

2N

on

poor

0.0

10.2

00.0

00.0

20.0

50.6

10.3

01.1

4M

ale-

hea

ded

0.0

10.1

70.0

00.0

20.0

40.5

30.2

60.9

9Fe

mal

e-h

ead

ed0.0

00.0

20.0

00.0

00.0

00.0

60.0

30.1

1


Tab

leA

5.2

:(c

)C

onti

nu

ed.

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

and

smal

lLe

ader

sof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

entr

ant

mer

ge

larg

est

shar

esco

mp

etit

ion

(C)

Com

ple

men

tary

pol

icie

s(fi

rms)

On

lyp

rod

uce

rsT

ota

l0.3

47.1

50.0

60.6

01.5

7−2

0.3

510.8

239.8

2Poor

0.3

36.8

50.0

60.5

81.5

0−1

9.5

210.3

838.1

9N

on

poor

0.3

57.3

30.0

60.6

21.6

1−2

0.8

611.1

040.8

3M

ale-

hea

ded

0.3

57.2

60.0

60.6

11.5

9−2

0.6

710.9

940.4

4Fe

mal

e-h

ead

ed0.1

12.2

80.0

20.1

90.5

06.5

13.4

612.7

3

All

hou

seh

old

sT

ota

l0.0

10.1

70.0

00.0

10.0

40.4

80.2

60.9

4Poor

0.0

10.1

30.0

00.0

10.0

30.3

70.2

00.7

2N

on

poor

0.0

10.2

10.0

00.0

20.0

50.5

90.3

11.1

5M

ale-

hea

ded

0.0

10.1

80.0

00.0

20.0

40.5

10.2

71.0

0Fe

mal

e-h

ead

ed0.0

00.0

20.0

00.0

00.0

00.0

60.0

30.1

1

(D)

Com

ple

men

tary

pol

icie

s(f

arm

ers−

firm

s)

On

lyp

rod

uce

rsT

ota

l0.7

47.5

80.3

20.9

71.9

6−1

9.5

611.4

040.9

3Poor

0.7

17.2

70.3

10.9

31.8

8−1

8.7

610.9

339.2

6N

on

poor

0.7

67.7

70.3

30.9

92.0

1−2

0.0

511.6

941.9

7M

ale-

hea

ded

0.7

57.7

00.3

30.9

81.9

9−1

9.8

611.5

741.5

7Fe

mal

e-h

ead

ed0.2

42.4

20.1

00.3

10.6

36.2

53.6

413.0

8

All

hou

seh

old

sT

ota

l0.0

20.1

80.0

10.0

20.0

50.4

60.2

70.9

7Poor

0.0

10.1

40.0

10.0

20.0

40.3

50.2

10.7

4N

on

poor

0.0

20.2

20.0

10.0

30.0

60.5

60.3

31.1

8M

ale-

hea

ded

0.0

20.1

90.0

10.0

20.0

50.4

90.2

91.0

3Fe

mal

e-h

ead

ed0.0

00.0

20.0

00.0

00.0

10.0

50.0

30.1

1


Tab

leA

5.2

:(c

)C

onti

nu

ed.

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

and

smal

lLe

ader

sof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

entr

ant

mer

ge

larg

est

shar

esco

mp

etit

ion

(E)

Inte

rnati

onalpri

ces

On

lyp

rod

uce

rsT

ota

l7.9

516.3

07.8

17.8

88.5

0−1

0.9

521.7

855.8

7Poor

7.6

315.6

37.4

97.5

68.1

5−1

0.5

020.8

953.5

9N

on

poor

8.1

616.7

18.0

18.0

88.7

2−1

1.2

322.3

357.2

8M

ale-

hea

ded

8.0

816.5

57.9

38.0

18.6

3−1

1.1

222.1

256.7

4Fe

mal

e-h

ead

ed2.5

45.2

12.5

02.5

22.7

23.5

06.9

617.8

6

All

hou

seh

old

sT

ota

l0.1

90.3

90.1

80.1

90.2

00.2

60.5

21.3

2Poor

0.1

40.2

90.1

40.1

40.1

50.2

00.3

91.0

1N

on

poor

0.2

30.4

70.2

30.2

30.2

50.3

20.6

31.6

1M

ale-

hea

ded

0.2

00.4

10.2

00.2

00.2

10.2

70.5

51.4

0Fe

mal

e-h

ead

ed0.0

20.0

40.0

20.0

20.0

20.0

30.0

60.1

5


Tab

leA

5.2

:(d

)C

ôte

d’Ivo

ire,

cott

on(h

ouse

hol

din

com

eper

cen

tage

chan

ges

).

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

and

smal

lLe

ader

sof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

entr

ant

mer

ge

larg

est

shar

esco

mp

etit

ion

(A)

Basi

cm

odel

On

lyp

rod

uce

rsT

ota

l0.0

04.3

40.1

92.5

02.9

66.1

912.3

333.4

1Poor

0.0

04.1

70.1

82.4

12.8

55.9

611.8

632.1

5N

on

poor

0.0

04.4

00.1

92.5

43.0

06.2

812.5

233.9

3M

ale-

hea

ded

0.0

04.3

00.1

92.4

82.9

36.1

312.2

233.1

1Fe

mal

e-h

ead

ed0.0

05.6

70.2

43.2

73.8

78.0

916.1

143.6

6

All

hou

seh

old

sT

ota

l0.0

00.3

50.0

20.2

00.2

40.5

01.0

02.7

1Poor

0.0

00.1

90.0

10.1

10.1

30.2

70.5

41.4

6N

on

poor

0.0

00.5

30.0

20.3

10.3

60.7

61.5

14.0

8M

ale-

hea

ded

0.0

00.4

10.0

20.2

40.2

80.5

81.1

63.1

6Fe

mal

e-h

ead

ed0.0

00.0

80.0

00.0

40.0

50.1

10.2

20.5

9

(B)

Com

ple

men

tary

pol

icie

s(f

arm

ers)

On

lyp

rod

uce

rsT

ota

l0.7

05.0

10.9

11.7

92.2

85.4

512.9

834.4

6Poor

0.6

74.8

20.8

71.7

22.1

95.2

512.4

933.1

6N

on

poor

0.7

15.0

90.9

21.8

22.3

15.5

413.1

834.9

9M

ale-

hea

ded

0.6

94.9

70.9

01.7

72.2

65.4

012.8

634.1

5Fe

mal

e-h

ead

ed0.9

16.5

51.1

92.3

42.9

77.1

216.9

645.0

2

All

hou

seh

old

sT

ota

l0.0

60.4

10.0

70.1

50.1

80.4

41.0

52.8

0Poor

0.0

30.2

20.0

40.0

80.1

00.2

40.5

71.5

0N

on

poor

0.0

90.6

10.1

10.2

20.2

80.6

71.5

94.2

1M

ale-

hea

ded

0.0

70.4

70.0

90.1

70.2

10.5

11.2

33.2

5Fe

mal

e-h

ead

ed0.0

10.0

90.0

20.0

30.0

40.1

00.2

30.6

1


Tab

leA

5.2

:(d

)C

onti

nu

ed.

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

and

smal

lLe

ader

sof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

entr

ant

mer

ge

larg

est

shar

esco

mp

etit

ion

(C)

Com

ple

men

tary

pol

icie

s(fi

rms)

On

lyp

rod

uce

rsT

ota

l1.2

05.4

71.4

51.2

91.8

44.9

113.3

934.9

1Poor

1.1

55.2

71.3

91.2

41.7

74.7

312.8

833.5

9N

on

poor

1.2

25.5

61.4

71.3

01.8

74.9

913.6

035.4

5M

ale-

hea

ded

1.1

95.4

21.4

31.2

71.8

24.8

713.2

734.5

9Fe

mal

e-h

ead

ed1.5

67.1

51.8

91.6

82.4

06.4

217.4

945.6

1

All

hou

seh

old

sT

ota

l0.1

00.4

40.1

20.1

00.1

50.4

01.0

92.8

4Poor

0.0

50.2

40.0

60.0

60.0

80.2

10.5

81.5

2N

on

poor

0.1

50.6

70.1

80.1

60.2

20.6

01.6

44.2

7M

ale-

hea

ded

0.1

10.5

20.1

40.1

20.1

70.4

61.2

63.3

0Fe

mal

e-h

ead

ed0.0

20.1

00.0

30.0

20.0

30.0

90.2

40.6

1

(D)

Com

ple

men

tary

pol

icie

s(f

arm

ers−

firm

s)

On

lyp

rod

uce

rsT

ota

l1.8

96.1

42.1

70.5

81.1

64.1

814.0

335.9

5Poor

1.8

25.9

12.0

80.5

61.1

24.0

313.5

034.6

0N

on

poor

1.9

26.2

42.2

00.5

91.1

84.2

514.2

536.5

1M

ale-

hea

ded

1.8

86.0

92.1

50.5

71.1

54.1

513.9

135.6

3Fe

mal

e-h

ead

ed2.4

78.0

32.8

30.7

61.5

25.4

718.3

346.9

8

All

hou

seh

old

sT

ota

l0.1

50.5

00.1

80.0

50.0

90.3

41.1

42.9

2Poor

0.0

80.2

70.0

90.0

30.0

50.1

80.6

11.5

7N

on

poor

0.2

30.7

50.2

60.0

70.1

40.5

11.7

14.3

9M

ale-

hea

ded

0.1

80.5

80.2

00.0

50.1

10.4

01.3

23.3

9Fe

mal

e-h

ead

ed0.0

30.1

10.0

40.0

10.0

20.0

70.2

50.6

3


Tab

leA

5.2

:(d

)C

onti

nu

ed.

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

and

smal

lLe

ader

sof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

entr

ant

mer

ge

larg

est

shar

esco

mp

etit

ion

(E)

Inte

rnati

onalpri

ces

On

lyp

rod

uce

rsT

ota

l10.9

815.6

311.6

08.3

67.2

64.5

023.7

649.6

2Poor

10.5

715.0

411.1

68.0

56.9

84.3

322.8

647.7

5N

on

poor

11.1

515.8

711.7

88.4

97.3

74.5

724.1

250.3

9M

ale-

hea

ded

10.8

815.4

911.5

08.2

97.1

94.4

623.5

449.1

8Fe

mal

e-h

ead

ed14.3

520.4

315.1

610.9

39.4

85.8

831.0

464.8

4

All

hou

seh

old

sT

ota

l0.8

91.2

70.9

40.6

80.5

90.3

71.9

34.0

3Poor

0.4

80.6

80.5

10.3

60.3

20.2

01.0

42.1

6N

on

poor

1.3

41.9

11.4

21.0

20.8

90.5

52.9

06.0

7M

ale-

hea

ded

1.0

41.4

81.1

00.7

90.6

90.4

32.2

44.6

9Fe

mal

e-h

ead

ed0.1

90.2

80.2

00.1

50.1

30.0

80.4

20.8

7


Table A5.2: (e) Malawi, cotton (household income percentage changes).



(A) Basic model



(B) Complementary policies (farmers)



(C) Complementary policies (firms)




Table A5.2: (e) Continued.



(D) Complementary policies (farmers − firms)



(E) International prices




Tab

leA

5.2

:(f

)C

ôte

d’Ivo

ire,

coco

a(h

ouse

hol

din

com

eper

cen

tage

chan

ges

).

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

and

smal

lLe

ader

sof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

entr

ant

mer

ge

larg

est

shar

esco

mp

etit

ion

(A)

Basi

cm

odel

On

lyp

rod

uce

rsT

ota

l0.0

01.1

60.1

70.6

60.8

81.3

210.2

321.1

8Poor

0.0

01.1

70.1

70.6

60.8

91.3

310.3

021.3

1N

on

poor

0.0

01.1

60.1

70.6

50.8

71.3

110.1

621.0

3M

ale-

hea

ded

0.0

01.1

80.1

70.6

60.8

91.3

310.3

321.3

9Fe

mal

e-h

ead

ed0.0

00.9

50.1

40.5

40.7

21.0

88.3

617.3

0

All

hou

seh

old

sT

ota

l0.0

00.3

50.0

50.1

90.2

60.3

93.0

36.2

8Poor

0.0

00.3

70.0

50.2

10.2

80.4

13.2

16.6

5N

on

poor

0.0

00.3

20.0

50.1

80.2

40.3

72.8

45.8

7M

ale-

hea

ded

0.0

00.4

00.0

60.2

20.3

00.4

53.5

17.2

7Fe

mal

e-h

ead

ed0.0

00.0

80.0

10.0

50.0

60.0

90.7

31.5

1

(B)

Com

ple

men

tary

pol

icie

s(f

arm

ers)

On

lyp

rod

uce

rsT

ota

l0.9

42.1

11.1

30.2

90.0

60.3

711.1

422.3

4Poor

0.9

52.1

21.1

30.3

00.0

60.3

711.2

022.4

8N

on

poor

0.9

42.0

91.1

20.2

90.0

60.3

711.0

522.1

7M

ale-

hea

ded

0.9

52.1

31.1

40.3

00.0

60.3

811.2

422.5

6Fe

mal

e-h

ead

ed0.7

71.7

20.9

20.2

40.0

50.3

09.1

018.2

5

All

hou

seh

old

sT

ota

l0.2

80.6

30.3

30.0

90.0

20.1

13.3

06.6

2Poor

0.3

00.6

60.3

50.0

90.0

20.1

23.5

07.0

2N

on

poor

0.2

60.5

90.3

10.0

80.0

20.1

03.0

96.1

9M

ale-

hea

ded

0.3

20.7

20.3

90.1

00.0

20.1

33.8

27.6

6Fe

mal

e-h

ead

ed0.0

70.1

50.0

80.0

20.0

00.0

30.7

91.5

9


Tab

leA

5.2

:(f

)C

onti

nu

ed.

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

and

smal

lLe

ader

sof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

entr

ant

mer

ge

larg

est

shar

esco

mp

etit

ion

(C)

Com

ple

men

tary

pol

icie

s(fi

rms)

On

lyp

rod

uce

rsT

ota

l1.2

22.3

81.4

10.5

80.3

40.0

811.2

622.4

4Poor

1.2

32.3

91.4

20.5

90.3

40.0

811.3

322.5

7N

on

poor

1.2

12.3

61.4

00.5

80.3

40.0

811.1

822.2

7M

ale-

hea

ded

1.2

32.4

01.4

30.5

90.3

40.0

811.3

722.6

6Fe

mal

e-h

ead

ed1.0

01.9

41.1

50.4

80.2

80.0

79.2

018.3

2

All

hou

seh

old

sT

ota

l0.3

60.7

10.4

20.1

70.1

00.0

23.3

46.6

5Poor

0.3

80.7

50.4

40.1

80.1

10.0

33.5

47.0

5N

on

poor

0.3

40.6

60.3

90.1

60.0

90.0

23.1

26.2

2M

ale-

hea

ded

0.4

20.8

20.4

80.2

00.1

20.0

33.8

67.7

0Fe

mal

e-h

ead

ed0.0

90.1

70.1

00.0

40.0

20.0

10.8

01.6

0

(D)

Com

ple

men

tary

pol

icie

s(f

arm

ers−

firm

s)

On

lyp

rod

uce

rsT

ota

l2.1

63.3

22.3

61.5

31.2

70.8

612.1

623.5

9Poor

2.1

83.3

42.3

81.5

41.2

80.8

612.2

323.7

4N

on

poor

2.1

53.2

92.3

41.5

11.2

60.8

512.0

723.4

2M

ale-

hea

ded

2.1

83.3

52.3

81.5

41.2

90.8

712.2

823.8

2Fe

mal

e-h

ead

ed1.7

72.7

11.9

31.2

51.0

40.7

09.9

319.2

7

All

hou

seh

old

sT

ota

l0.6

40.9

80.7

00.4

50.3

80.2

53.6

06.9

9Poor

0.6

81.0

40.7

40.4

80.4

00.2

73.8

27.4

1N

on

poor

0.6

00.9

20.6

50.4

20.3

50.2

43.3

76.5

4M

ale-

hea

ded

0.7

41.1

40.8

10.5

20.4

40.2

94.1

78.0

9Fe

mal

e-h

ead

ed0.1

50.2

40.1

70.1

10.0

90.0

60.8

71.6

8


Tab

leA

5.2

:(f

)C

onti

nu

ed.

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

and

smal

lLe

ader

sof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

entr

ant

mer

ge

larg

est

shar

esco

mp

etit

ion

(E)

Inte

rnati

onalpri

ces

On

lyp

rod

uce

rsT

ota

l11.5

912.8

311.8

810.9

510.5

910.1

921.8

135.3

7Poor

11.6

612.9

111.9

611.0

110.6

510.2

521.9

535.5

8N

on

poor

11.5

012.7

411.7

910.8

710.5

110.1

121.6

535.1

0M

ale-

hea

ded

11.7

012.9

612.0

011.0

510.6

910.2

922.0

235.7

1Fe

mal

e-h

ead

ed9.4

710.4

89.7

08.9

48.6

58.3

217.8

128.8

9

All

hou

seh

old

sT

ota

l3.4

43.8

03.5

23.2

53.1

43.0

26.4

710.4

9Poor

3.6

44.0

33.7

33.4

43.3

33.2

06.8

511.1

1N

on

poor

3.2

13.5

63.2

93.0

42.9

42.8

36.0

59.8

1M

ale-

hea

ded

3.9

84.4

04.0

83.7

63.6

33.5

07.4

812.1

3Fe

mal

e-h

ead

ed0.8

30.9

20.8

50.7

80.7

60.7

31.5

62.5

2


Tab

leA

5.2

:(g

)G

han

a,c

ocoa

(hou

seh

old

inco

me

per

cen

tage

chan

ges

).

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

and

smal

lLe

ader

sof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

entr

ant

mer

ge

larg

est

shar

esco

mp

etit

ion

(A)

Basi

cm

odel

On

lyp

rod

uce

rsT

ota

l0.0

01.0

60.0

50.2

40.3

01.3

16.9

311.4

9Poor

0.0

01.0

10.0

40.2

30.2

91.2

56.6

210.9

8N

on

poor

0.0

01.1

20.0

50.2

50.3

21.3

87.3

212.1

4M

ale-

hea

ded

0.0

01.0

90.0

50.2

40.3

11.3

57.1

411.8

4Fe

mal

e-h

ead

ed0.0

00.9

50.0

40.2

10.2

71.1

76.2

110.3

0

All

hou

seh

old

sT

ota

l0.0

00.1

70.0

10.0

40.0

50.2

21.1

51.9

0Poor

0.0

00.1

70.0

10.0

40.0

50.2

11.1

31.8

7N

on

poor

0.0

00.1

80.0

10.0

40.0

50.2

21.1

71.9

4M

ale-

hea

ded

0.0

00.2

00.0

10.0

40.0

60.2

51.3

22.1

8Fe

mal

e-h

ead

ed0.0

00.1

20.0

10.0

30.0

30.1

40.7

61.2

7

(B)

Com

ple

men

tary

pol

icie

s(f

arm

ers)

On

lyp

rod

uce

rsT

ota

l0.4

11.4

60.4

60.1

80.1

00.8

87.3

311.9

9Poor

0.3

91.3

90.4

40.1

70.1

00.8

47.0

011.4

6N

on

poor

0.4

41.5

40.4

90.1

90.1

10.9

37.7

412.6

7M

ale-

hea

ded

0.4

21.5

00.4

80.1

80.1

10.9

17.5

512.3

6Fe

mal

e-h

ead

ed0.3

71.3

10.4

10.1

60.0

90.7

96.5

710.7

5

All

hou

seh

old

sT

ota

l0.0

70.2

40.0

80.0

30.0

20.1

51.2

11.9

8Poor

0.0

70.2

40.0

80.0

30.0

20.1

41.1

91.9

5N

on

poor

0.0

70.2

50.0

80.0

30.0

20.1

51.2

42.0

2M

ale-

hea

ded

0.0

80.2

80.0

90.0

30.0

20.1

71.3

92.2

8Fe

mal

e-h

ead

ed0.0

50.1

60.0

50.0

20.0

10.1

00.8

11.3

2


Tab

leA

5.2

:(g

)C

onti

nu

ed.

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

and

smal

lLe

ader

sof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

entr

ant

mer

ge

larg

est

shar

esco

mp

etit

ion

(C)

Com

ple

men

tary

pol

icie

s(fi

rms)

On

lyp

rod

uce

rsT

ota

l0.6

31.6

60.6

80.3

90.3

10.6

47.4

512.1

0Poor

0.6

01.5

80.6

50.3

80.3

00.6

17.1

111.5

6N

on

poor

0.6

61.7

50.7

20.4

20.3

30.6

77.8

712.7

8M

ale-

hea

ded

0.6

51.7

10.7

00.4

10.3

20.6

67.6

712.4

7Fe

mal

e-h

ead

ed0.5

61.4

80.6

10.3

50.2

80.5

76.6

810.8

5

All

hou

seh

old

sT

ota

l0.1

00.2

70.1

10.0

70.0

50.1

11.2

32.0

0Poor

0.1

00.2

70.1

10.0

60.0

50.1

01.2

11.9

7N

on

poor

0.1

10.2

80.1

20.0

70.0

50.1

11.2

62.0

4M

ale-

hea

ded

0.1

20.3

10.1

30.0

70.0

60.1

21.4

12.3

0Fe

mal

e-h

ead

ed0.0

70.1

80.0

80.0

40.0

30.0

70.8

21.3

3

(D)

Com

ple

men

tary

pol

icie

s(f

arm

ers−

firm

s)

On

lyp

rod

uce

rsT

ota

l1.0

42.0

61.1

00.8

10.7

20.2

17.8

412.6

0Poor

0.9

91.9

71.0

50.7

70.6

90.2

07.4

912.0

4N

on

poor

1.1

02.1

81.1

60.8

50.7

60.2

38.2

813.3

1M

ale-

hea

ded

1.0

72.1

21.1

30.8

30.7

40.2

28.0

812.9

9Fe

mal

e-h

ead

ed0.9

31.8

50.9

80.7

20.6

50.1

97.0

311.3

0

All

hou

seh

old

sT

ota

l0.1

70.3

40.1

80.1

30.1

20.0

41.3

02.0

9Poor

0.1

70.3

40.1

80.1

30.1

20.0

31.2

82.0

5N

on

poor

0.1

80.3

50.1

90.1

40.1

20.0

41.3

22.1

2M

ale-

hea

ded

0.2

00.3

90.2

10.1

50.1

40.0

41.4

92.3

9Fe

mal

e-h

ead

ed0.1

10.2

30.1

20.0

90.0

80.0

20.8

61.3

9


Tab

leA

5.2

:(g

)C

onti

nu

ed.

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

and

smal

lLe

ader

sof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

entr

ant

mer

ge

larg

est

shar

esco

mp

etit

ion

(E)

Inte

rnati

onalpri

ces

On

lyp

rod

uce

rsT

ota

l5.5

86.6

15.6

95.3

45.2

04.3

512.6

218.2

3Poor

5.3

36.3

15.4

45.1

04.9

64.1

512.0

517.4

1N

on

poor

5.9

06.9

86.0

15.6

45.4

94.5

913.3

319.2

5M

ale-

hea

ded

5.7

66.8

15.8

75.5

05.3

64.4

813.0

018.7

8Fe

mal

e-h

ead

ed5.0

15.9

35.1

04.7

94.6

63.9

011.3

116.3

4

All

hou

seh

old

sT

ota

l0.9

21.0

90.9

40.8

80.8

60.7

22.0

93.0

2Poor

0.9

11.0

80.9

30.8

70.8

50.7

12.0

62.9

7N

on

poor

0.9

41.1

10.9

60.9

00.8

80.7

32.1

33.0

7M

ale-

hea

ded

1.0

61.2

61.0

81.0

10.9

90.8

32.4

03.4

6Fe

mal

e-h

ead

ed0.6

20.7

30.6

30.5

90.5

70.4

81.3

92.0

1


Tab

leA

5.2

:(h

)C

ôte

d’Ivo

ire,

coff

ee(h

ouse

hol

din

com

eper

cen

tage

chan

ges

).

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

and

smal

lLe

ader

sof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

entr

ant

mer

ge

larg

est

shar

esco

mp

etit

ion

(A)

Basi

cm

odel

On

lyp

rod

uce

rsT

ota

l0.0

00.6

00.0

90.3

40.4

50.6

85.2

710.9

0Poor

0.0

00.6

90.1

00.3

90.5

20.7

86.0

812.5

9N

on

poor

0.0

00.4

40.0

70.2

50.3

30.5

03.8

67.9

8M

ale-

hea

ded

0.0

00.6

10.0

90.3

40.4

60.6

95.3

711.1

1Fe

mal

e-h

ead

ed0.0

00.3

70.0

50.2

10.2

80.4

13.2

16.6

5

All

hou

seh

old

sT

ota

l0.0

00.1

40.0

20.0

80.1

00.1

61.2

12.5

0Poor

0.0

00.1

90.0

30.1

10.1

50.2

21.6

93.5

1N

on

poor

0.0

00.0

80.0

10.0

40.0

60.0

90.6

71.4

0M

ale-

hea

ded

0.0

00.1

60.0

20.0

90.1

20.1

81.4

12.9

3Fe

mal

e-h

ead

ed0.0

00.0

20.0

00.0

10.0

20.0

30.2

00.4

2

(B)

Com

ple

men

tary

pol

icie

s(f

arm

ers)

On

lyp

rod

uce

rsT

ota

l0.4

91.0

90.5

80.1

50.0

30.1

95.7

311.5

0Poor

0.5

61.2

50.6

70.1

70.0

30.2

26.6

213.2

7N

on

poor

0.3

60.8

00.4

20.1

10.0

20.1

44.2

08.4

2M

ale-

hea

ded

0.5

01.1

10.5

90.1

50.0

30.2

05.8

411.7

2Fe

mal

e-h

ead

ed0.3

00.6

60.3

50.0

90.0

20.1

23.5

07.0

2

All

hou

seh

old

sT

ota

l0.1

10.2

50.1

30.0

30.0

10.0

41.3

12.6

3Poor

0.1

60.3

50.1

90.0

50.0

10.0

61.8

43.7

0N

on

poor

0.0

60.1

40.0

70.0

20.0

00.0

20.7

31.4

7M

ale-

hea

ded

0.1

30.2

90.1

60.0

40.0

10.0

51.5

43.0

9Fe

mal

e-h

ead

ed0.0

20.0

40.0

20.0

10.0

00.0

10.2

20.4

4


Tab

leA

5.2

:(h

)C

onti

nu

ed.

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

and

smal

lLe

ader

sof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

entr

ant

mer

ge

larg

est

shar

esco

mp

etit

ion

(C)

Com

ple

men

tary

pol

icie

s(fi

rms)

On

lyp

rod

uce

rsT

ota

l0.6

31.2

20.7

30.3

00.1

70.0

45.7

911.5

5Poor

0.7

31.4

10.8

40.3

50.2

00.0

56.6

913.3

3N

on

poor

0.4

60.9

00.5

30.2

20.1

30.0

34.2

48.4

5M

ale-

hea

ded

0.6

41.2

50.7

40.3

10.1

80.0

45.9

111.7

7Fe

mal

e-h

ead

ed0.3

80.7

50.4

40.1

80.1

10.0

33.5

47.0

5

All

hou

seh

old

sT

ota

l0.1

40.2

80.1

70.0

70.0

40.0

11.3

32.6

4Poor

0.2

00.3

90.2

30.1

00.0

60.0

11.8

63.7

1N

on

poor

0.0

80.1

60.0

90.0

40.0

20.0

10.7

41.4

8M

ale-

hea

ded

0.1

70.3

30.2

00.0

80.0

50.0

11.5

63.1

0Fe

mal

e-h

ead

ed0.0

20.0

50.0

30.0

10.0

10.0

00.2

20.4

4

(D)

Com

ple

men

tary

pol

icie

s(f

arm

ers−

firm

s)

On

lyp

rod

uce

rsT

ota

l1.1

11.7

11.2

10.7

90.6

50.4

46.2

612.1

4Poor

1.2

81.9

71.4

00.9

10.7

60.5

17.2

214.0

1N

on

poor

0.8

11.2

50.8

90.5

70.4

80.3

24.5

88.8

9M

ale-

hea

ded

1.1

31.7

41.2

40.8

00.6

70.4

56.3

812.3

7Fe

mal

e-h

ead

ed0.6

81.0

40.7

40.4

80.4

00.2

73.8

27.4

1

All

hou

seh

old

sT

ota

l0.2

50.3

90.2

80.1

80.1

50.1

01.4

32.7

8Poor

0.3

60.5

50.3

90.2

50.2

10.1

42.0

13.9

0N

on

poor

0.1

40.2

20.1

60.1

00.0

80.0

60.8

01.5

6M

ale-

hea

ded

0.3

00.4

60.3

30.2

10.1

80.1

21.6

83.2

6Fe

mal

e-h

ead

ed0.0

40.0

70.0

50.0

30.0

30.0

20.2

40.4

7


Tab

leA

5.2

:(h

)C

onti

nu

ed.

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

and

smal

lLe

ader

sof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

entr

ant

mer

ge

larg

est

shar

esco

mp

etit

ion

(E)

Inte

rnati

onalpri

ces

On

lyp

rod

uce

rsT

ota

l5.9

66.6

06.1

15.6

35.4

55.2

411.2

218.2

0Poor

6.8

87.6

27.0

66.5

06.2

96.0

512.9

621.0

1N

on

poor

4.3

74.8

44.4

84.1

33.9

93.8

48.2

213.3

3M

ale-

hea

ded

6.0

86.7

36.2

35.7

45.5

55.3

411.4

418.5

5Fe

mal

e-h

ead

ed3.6

44.0

33.7

33.4

43.3

23.2

06.8

511.1

1

All

hou

seh

old

sT

ota

l1.3

71.5

11.4

01.2

91.2

51.2

02.5

74.1

7Poor

1.9

22.1

21.9

71.8

11.7

51.6

93.6

15.8

5N

on

poor

0.7

60.8

50.7

80.7

20.7

00.6

71.4

42.3

3M

ale-

hea

ded

1.6

01.7

71.6

41.5

11.4

61.4

13.0

14.8

9Fe

mal

e-h

ead

ed0.2

30.2

50.2

30.2

20.2

10.2

00.4

30.7

0


Tab

leA

5.2

:(i)

Rw

an

da,c

offee

(hou

seh

old

inco

me

per

cen

tage

chan

ges

).

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

and

smal

lLe

ader

sof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

entr

ant

mer

ge

larg

est

shar

esco

mp

etit

ion

(A)

Basi

cm

odel

On

lyp

rod

uce

rsT

ota

l0.0

00.3

20.0

20.3

50.4

10.4

40.6

62.7

0Poor

0.0

00.4

00.0

30.4

40.5

10.5

50.8

23.3

7N

on

poor

0.0

00.2

80.0

20.3

10.3

60.3

90.5

72.3

7M

ale-

hea

ded

0.0

00.3

00.0

20.3

30.3

90.4

10.6

12.5

3Fe

mal

e-h

ead

ed0.0

00.3

80.0

20.4

20.4

90.5

20.7

83.2

1

All

hou

seh

old

sT

ota

l0.0

00.0

30.0

00.0

40.0

40.0

40.0

70.2

7Poor

0.0

00.0

30.0

00.0

30.0

40.0

40.0

60.2

6N

on

poor

0.0

00.0

30.0

00.0

40.0

40.0

50.0

70.2

8M

ale-

hea

ded

0.0

00.0

30.0

00.0

40.0

40.0

50.0

70.2

8Fe

mal

e-h

ead

ed0.0

00.0

30.0

00.0

30.0

40.0

40.0

60.2

5

(B)

Com

ple

men

tary

pol

icie

s(f

arm

ers)

On

lyp

rod

uce

rsT

ota

l0.1

40.4

60.1

60.2

10.2

80.3

10.7

92.8

8Poor

0.1

70.5

70.2

00.2

70.3

50.3

80.9

93.5

9N

on

poor

0.1

20.4

00.1

40.1

90.2

40.2

70.6

92.5

2M

ale-

hea

ded

0.1

30.4

30.1

50.2

00.2

60.2

90.7

42.7

0Fe

mal

e-h

ead

ed0.1

60.5

50.1

90.2

50.3

30.3

60.9

43.4

2

All

hou

seh

old

sT

ota

l0.0

10.0

50.0

20.0

20.0

30.0

30.0

80.2

9Poor

0.0

10.0

40.0

20.0

20.0

30.0

30.0

80.2

7N

on

poor

0.0

10.0

50.0

20.0

20.0

30.0

30.0

80.3

0M

ale-

hea

ded

0.0

10.0

50.0

20.0

20.0

30.0

30.0

80.3

0Fe

mal

e-h

ead

ed0.0

10.0

40.0

10.0

20.0

30.0

30.0

70.2

6


Tab

leA

5.2

:(i)

Con

tin

ued

.

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

and

smal

lLe

ader

sof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

entr

ant

mer

ge

larg

est

shar

esco

mp

etit

ion

(C)

Com

ple

men

tary

pol

icie

s(fi

rms)

On

lyp

rod

uce

rsT

ota

l0.1

90.5

20.2

20.1

50.2

20.2

50.8

42.9

4Poor

0.2

40.6

40.2

70.1

90.2

80.3

11.0

53.6

6N

on

poor

0.1

70.4

50.1

90.1

30.2

00.2

20.7

42.5

7M

ale-

hea

ded

0.1

80.4

80.2

10.1

40.2

10.2

40.7

92.7

5Fe

mal

e-h

ead

ed0.2

30.6

10.2

60.1

80.2

70.3

01.0

03.4

9

All

hou

seh

old

sT

ota

l0.0

20.0

50.0

20.0

20.0

20.0

30.0

80.2

9Poor

0.0

20.0

50.0

20.0

10.0

20.0

20.0

80.2

8N

on

poor

0.0

20.0

50.0

20.0

20.0

20.0

30.0

90.3

0M

ale-

hea

ded

0.0

20.0

50.0

20.0

20.0

20.0

30.0

90.3

0Fe

mal

e-h

ead

ed0.0

20.0

50.0

20.0

10.0

20.0

20.0

80.2

7

(D)

Com

ple

men

tary

pol

icie

s(f

arm

ers−

firm

s)

On

lyp

rod

uce

rsT

ota

l0.3

30.6

50.3

60.0

20.0

90.1

20.9

83.1

1Poor

0.4

10.8

20.4

50.0

20.1

10.1

51.2

23.8

8N

on

poor

0.2

90.5

70.3

10.0

10.0

80.1

00.8

62.7

3M

ale-

hea

ded

0.3

10.6

10.3

40.0

10.0

90.1

10.9

22.9

2Fe

mal

e-h

ead

ed0.3

90.7

80.4

30.0

20.1

10.1

41.1

63.7

0

All

hou

seh

old

sT

ota

l0.0

30.0

70.0

40.0

00.0

10.0

10.1

00.3

1Poor

0.0

30.0

60.0

30.0

00.0

10.0

10.0

90.3

0N

on

poor

0.0

30.0

70.0

40.0

00.0

10.0

10.1

00.3

2M

ale-

hea

ded

0.0

30.0

70.0

40.0

00.0

10.0

10.1

00.3

2Fe

mal

e-h

ead

ed0.0

30.0

60.0

30.0

00.0

10.0

10.0

90.2

9


Tab

leA

5.2

:(i)

Con

tin

ued

.

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

and

smal

lLe

ader

sof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

entr

ant

mer

ge

larg

est

shar

esco

mp

etit

ion

(E)

Inte

rnati

onalpri

ces

On

lyp

rod

uce

rsT

ota

l1.8

22.2

01.8

91.4

61.3

31.3

02.5

15.0

4Poor

2.2

72.7

42.3

51.8

21.6

51.6

23.1

36.2

8N

on

poor

1.5

91.9

31.6

51.2

81.1

61.1

42.2

04.4

2M

ale-

hea

ded

1.7

02.0

61.7

71.3

71.2

41.2

22.3

54.7

2Fe

mal

e-h

ead

ed2.1

72.6

22.2

51.7

41.5

81.5

52.9

96.0

0

All

hou

seh

old

sT

ota

l0.1

80.2

20.1

90.1

50.1

30.1

30.2

50.5

0Poor

0.1

70.2

10.1

80.1

40.1

30.1

20.2

40.4

8N

on

poor

0.1

90.2

30.2

00.1

50.1

40.1

30.2

60.5

2M

ale-

hea

ded

0.1

90.2

30.2

00.1

50.1

40.1

30.2

60.5

2Fe

mal

e-h

ead

ed0.1

70.2

00.1

70.1

30.1

20.1

20.2

30.4

6


Tab

leA

5.2

:(j

)U

gan

da,c

offee

(hou

seh

old

inco

me

per

cen

tage

chan

ges

).

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

and

smal

lLe

ader

sof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

entr

ant

mer

ge

larg

est

shar

esco

mp

etit

ion

(A)

Basi

cm

odel

On

lyp

rod

uce

rsT

ota

l0.0

00.0

90.0

10.0

60.0

70.1

00.7

71.5

4Poor

0.0

00.0

90.0

10.0

60.0

70.1

00.7

71.5

4N

on

poor

0.0

00.0

90.0

10.0

60.0

70.1

00.7

71.5

4M

ale-

hea

ded

0.0

00.0

90.0

10.0

60.0

80.1

10.8

21.6

5Fe

mal

e-h

ead

ed0.0

00.0

70.0

10.0

50.0

60.0

80.6

01.2

1

All

hou

seh

old

sT

ota

l0.0

00.0

20.0

00.0

20.0

20.0

30.2

10.4

2Poor

0.0

00.0

20.0

00.0

10.0

10.0

20.1

60.3

2N

on

poor

0.0

00.0

30.0

00.0

20.0

20.0

30.2

40.4

9M

ale-

hea

ded

0.0

00.0

30.0

00.0

20.0

20.0

30.2

30.4

6Fe

mal

e-h

ead

ed0.0

00.0

20.0

00.0

10.0

10.0

20.1

50.2

9

(B)

Com

ple

men

tary

pol

icie

s(f

arm

ers)

On

lyp

rod

uce

rsT

ota

l0.1

60.2

50.1

70.1

00.0

90.0

60.9

31.7

2Poor

0.1

60.2

50.1

70.1

00.0

90.0

60.9

31.7

2N

on

poor

0.1

60.2

50.1

70.1

00.0

90.0

60.9

31.7

2M

ale-

hea

ded

0.1

70.2

70.1

80.1

10.0

90.0

60.9

91.8

4Fe

mal

e-h

ead

ed0.1

30.2

00.1

30.0

80.0

70.0

50.7

31.3

5

All

hou

seh

old

sT

ota

l0.0

40.0

70.0

50.0

30.0

20.0

20.2

50.4

6Poor

0.0

30.0

50.0

30.0

20.0

20.0

10.1

90.3

6N

on

poor

0.0

50.0

80.0

50.0

30.0

30.0

20.2

90.5

5M

ale-

hea

ded

0.0

50.0

70.0

50.0

30.0

30.0

20.2

80.5

1Fe

mal

e-h

ead

ed0.0

30.0

50.0

30.0

20.0

20.0

10.1

80.3

3


Tab

leA

5.2

:(j

)C

onti

nu

ed.

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

and

smal

lLe

ader

sof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

entr

ant

mer

ge

larg

est

shar

esco

mp

etit

ion

(C)

Com

ple

men

tary

pol

icie

s(fi

rms)

On

lyp

rod

uce

rsT

ota

l0.2

10.3

00.2

20.1

50.1

40.1

10.9

71.7

6Poor

0.2

10.3

00.2

20.1

50.1

40.1

10.9

71.7

6N

on

poor

0.2

10.3

00.2

20.1

50.1

40.1

10.9

71.7

6M

ale-

hea

ded

0.2

20.3

20.2

30.1

60.1

50.1

21.0

31.8

8Fe

mal

e-h

ead

ed0.1

70.2

30.1

70.1

20.1

10.0

90.7

61.3

8

All

hou

seh

old

sT

ota

l0.0

60.0

80.0

60.0

40.0

40.0

30.2

60.4

7Poor

0.0

40.0

60.0

50.0

30.0

30.0

20.2

00.3

7N

on

poor

0.0

70.0

90.0

70.0

50.0

40.0

40.3

10.5

6M

ale-

hea

ded

0.0

60.0

90.0

70.0

50.0

40.0

30.2

90.5

2Fe

mal

e-h

ead

ed0.0

40.0

60.0

40.0

30.0

30.0

20.1

80.3

4

(D)

Com

ple

men

tary

pol

icie

s(f

arm

ers−

firm

s)

On

lyp

rod

uce

rsT

ota

l0.3

70.4

60.3

80.3

10.3

00.2

71.1

21.9

4Poor

0.3

70.4

60.3

80.3

10.3

00.2

71.1

21.9

4N

on

poor

0.3

70.4

60.3

80.3

10.3

00.2

71.1

21.9

4M

ale-

hea

ded

0.3

90.4

90.4

10.3

30.3

20.2

91.2

02.0

7Fe

mal

e-h

ead

ed0.2

90.3

60.3

00.2

40.2

30.2

10.8

81.5

2

All

hou

seh

old

sT

ota

l0.1

00.1

20.1

00.0

80.0

80.0

70.3

00.5

2Poor

0.0

80.0

90.0

80.0

60.0

60.0

60.2

30.4

0N

on

poor

0.1

20.1

50.1

20.1

00.0

90.0

90.3

60.6

2M

ale-

hea

ded

0.1

10.1

40.1

10.0

90.0

90.0

80.3

40.5

8Fe

mal

e-h

ead

ed0.0

70.0

90.0

70.0

60.0

60.0

50.2

10.3

7


Tab

leA

5.2

:(j

)C

onti

nu

ed.

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

and

smal

lLe

ader

sof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

entr

ant

mer

ge

larg

est

shar

esco

mp

etit

ion

(E)

Inte

rnati

onalpri

ces

On

lyp

rod

uce

rsT

ota

l1.9

32.0

21.9

41.8

71.8

41.8

22.7

13.6

8Poor

1.9

32.0

21.9

41.8

71.8

41.8

22.7

13.6

8N

on

poor

1.9

32.0

21.9

41.8

71.8

41.8

22.7

13.6

8M

ale-

hea

ded

2.0

62.1

62.0

71.9

91.9

71.9

42.8

93.9

3Fe

mal

e-h

ead

ed1.5

11.5

91.5

21.4

61.4

51.4

32.1

22.8

9

All

hou

seh

old

sT

ota

l0.5

20.5

50.5

20.5

00.5

00.4

90.7

30.9

9Poor

0.4

00.4

20.4

00.3

90.3

80.3

80.5

60.7

6N

on

poor

0.6

10.6

40.6

20.5

90.5

90.5

80.8

61.1

7M

ale-

hea

ded

0.5

70.6

00.5

80.5

60.5

50.5

40.8

11.1

0Fe

mal

e-h

ead

ed0.3

70.3

90.3

70.3

60.3

50.3

50.5

20.7

0


Tab

leA

5.2

:(k

)M

ala

wi,

tobacc

o(h

ouse

hol

din

com

eper

cen

tage

chan

ges

).

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

and

smal

lLe

ader

sof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

entr

ant

mer

ge

larg

est

shar

esco

mp

etit

ion

(A)

Basi

cm

odel

On

lyp

rod

uce

rsT

ota

l0.0

00.8

00.0

21.2

01.3

01.6

74.8

711.5

6Poor

0.0

00.6

30.0

10.9

41.0

11.3

13.8

19.0

4N

on

poor

0.0

00.8

80.0

21.3

31.4

31.8

45.3

712.7

4M

ale-

hea

ded

0.0

00.8

10.0

21.2

21.3

11.6

94.9

411.7

1Fe

mal

e-h

ead

ed0.0

00.7

00.0

21.0

61.1

41.4

74.2

910.1

8

All

hou

seh

old

sT

ota

l0.0

00.1

20.0

00.1

70.1

90.2

40.7

01.6

6Poor

0.0

00.0

60.0

00.1

00.1

10.1

40.3

90.9

4N

on

poor

0.0

00.1

60.0

00.2

30.2

50.3

20.9

52.2

4M

ale-

hea

ded

0.0

00.1

40.0

00.2

10.2

20.2

90.8

42.0

0Fe

mal

e-h

ead

ed0.0

00.0

40.0

00.0

60.0

70.0

90.2

50.6

0

(B)

Com

ple

men

tary

pol

icie

s(f

arm

ers)

On

lyp

rod

uce

rsT

ota

l0.4

01.2

00.4

20.8

00.9

01.2

75.2

612.0

9Poor

0.3

10.9

40.3

30.6

30.7

00.9

94.1

29.4

5N

on

poor

0.4

41.3

20.4

70.8

80.9

91.4

05.8

013.3

3M

ale-

hea

ded

0.4

01.2

20.4

30.8

10.9

11.2

95.3

312.2

4Fe

mal

e-h

ead

ed0.3

51.0

60.3

70.7

00.7

91.1

24.6

410.6

4

All

hou

seh

old

sT

ota

l0.0

60.1

70.0

60.1

20.1

30.1

80.7

61.7

4Poor

0.0

30.1

00.0

30.0

60.0

70.1

00.4

30.9

8N

on

poor

0.0

80.2

30.0

80.1

60.1

70.2

51.0

22.3

5M

ale-

hea

ded

0.0

70.2

10.0

70.1

40.1

60.2

20.9

12.0

9Fe

mal

e-h

ead

ed0.0

20.0

60.0

20.0

40.0

50.0

70.2

70.6

3


Tab

leA

5.2

:(k

)C

onti

nu

ed.

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

and

smal

lLe

ader

sof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

entr

ant

mer

ge

larg

est

shar

esco

mp

etit

ion

(C)

Com

ple

men

tary

pol

icie

s(fi

rms)

On

lyp

rod

uce

rsT

ota

l0.5

51.3

50.5

80.6

30.7

51.1

05.3

612.1

7Poor

0.4

31.0

50.4

50.5

00.5

80.8

64.1

99.5

2N

on

poor

0.6

01.4

90.6

40.7

00.8

21.2

25.9

113.4

2M

ale-

hea

ded

0.5

51.3

70.5

90.6

40.7

61.1

25.4

312.3

3Fe

mal

e-h

ead

ed0.4

81.1

90.5

10.5

60.6

60.9

74.7

210.7

2

All

hou

seh

old

sT

ota

l0.0

80.1

90.0

80.0

90.1

10.1

60.7

71.7

5Poor

0.0

40.1

10.0

50.0

50.0

60.0

90.4

30.9

9N

on

poor

0.1

10.2

60.1

10.1

20.1

40.2

11.0

42.3

6M

ale-

hea

ded

0.0

90.2

30.1

00.1

10.1

30.1

90.9

32.1

0Fe

mal

e-h

ead

ed0.0

30.0

70.0

30.0

30.0

40.0

60.2

80.6

4

(D)

Com

ple

men

tary

pol

icie

s(f

arm

ers−

firm

s)

On

lyp

rod

uce

rsT

ota

l0.9

41.7

50.9

80.2

30.3

50.7

05.7

512.7

0Poor

0.7

41.3

70.7

70.1

80.2

70.5

54.5

09.9

3N

on

poor

1.0

41.9

21.0

80.2

60.3

90.7

76.3

414.0

0M

ale-

hea

ded

0.9

51.7

71.0

00.2

40.3

60.7

15.8

312.8

7Fe

mal

e-h

ead

ed0.8

31.5

40.8

70.2

00.3

10.6

25.0

711.1

9

All

hou

seh

old

sT

ota

l0.1

40.2

50.1

40.0

30.0

50.1

00.8

31.8

3Poor

0.0

80.1

40.0

80.0

20.0

30.0

60.4

71.0

3N

on

poor

0.1

80.3

40.1

90.0

50.0

70.1

41.1

22.4

7M

ale-

hea

ded

0.1

60.3

00.1

70.0

40.0

60.1

20.9

92.2

0Fe

mal

e-h

ead

ed0.0

50.0

90.0

50.0

10.0

20.0

40.3

00.6

6


Tab

leA

5.2

:(k

)C

onti

nu

ed.

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

and

smal

lLe

ader

sof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

entr

ant

mer

ge

larg

est

shar

esco

mp

etit

ion

(E)

Inte

rnati

onalpri

ces

On

lyp

rod

uce

rsT

ota

l5.3

06.6

65.4

34.1

13.8

83.5

610.3

818.5

1Poor

4.1

55.2

14.2

43.2

13.0

42.7

98.1

214.4

8N

on

poor

5.8

47.3

55.9

84.5

34.2

83.9

311.4

520.4

1M

ale-

hea

ded

5.3

76.7

55.5

04.1

63.9

33.6

110.5

218.7

6Fe

mal

e-h

ead

ed4.6

75.8

74.7

83.6

23.4

23.1

49.1

416.3

0

All

hou

seh

old

sT

ota

l0.7

60.9

60.7

80.5

90.5

60.5

11.4

92.6

6Poor

0.4

30.5

40.4

40.3

30.3

10.2

90.8

41.5

0N

on

poor

1.0

31.2

91.0

50.8

00.7

50.6

92.0

23.5

9M

ale-

hea

ded

0.9

21.1

50.9

40.7

10.6

70.6

21.7

93.2

0Fe

mal

e-h

ead

ed0.2

80.3

50.2

80.2

10.2

00.1

90.5

40.9

7


Tab

leA

5.2

:(l)

Zam

bia

,tob

acc

o(h

ouse

hol

din

com

eper

cen

tage

chan

ges

).

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

and

smal

lLe

ader

sof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

entr

ant

mer

ge

larg

est

shar

esco

mp

etit

ion

(A)

Basi

cm

odel

On

lyp

rod

uce

rsT

ota

l0.0

02.9

30.3

70.9

41.5

34.5

06.4

820.9

9Poor

0.0

02.2

70.2

90.7

31.1

93.4

95.0

316.2

7N

on

poor

0.0

03.3

50.4

21.0

71.7

55.1

47.4

023.9

7M

ale-

hea

ded

0.0

03.0

40.3

80.9

71.5

84.6

66.7

221.7

4Fe

mal

e-h

ead

ed0.0

02.2

70.2

90.7

31.1

83.4

85.0

116.2

3

All

hou

seh

old

sT

ota

l0.0

00.0

60.0

10.0

20.0

30.1

00.1

40.4

5Poor

0.0

00.0

40.0

10.0

10.0

20.0

70.0

90.3

1N

on

poor

0.0

00.0

80.0

10.0

30.0

40.1

20.1

80.5

7M

ale-

hea

ded

0.0

00.0

70.0

10.0

20.0

40.1

10.1

60.5

0Fe

mal

e-h

ead

ed0.0

00.0

30.0

00.0

10.0

20.0

50.0

70.2

4

(B)

Com

ple

men

tary

pol

icie

s(f

arm

ers)

On

lyp

rod

uce

rsT

ota

l0.4

83.4

10.8

60.4

61.0

74.0

16.9

321.6

7Poor

0.3

72.6

40.6

70.3

60.8

33.1

15.3

716.8

0N

on

poor

0.5

43.8

90.9

90.5

31.2

24.5

77.9

124.7

4M

ale-

hea

ded

0.4

93.5

30.9

00.4

81.1

14.1

57.1

822.4

4Fe

mal

e-h

ead

ed0.3

72.6

40.6

70.3

60.8

33.1

05.3

616.7

5

All

hou

seh

old

sT

ota

l0.0

10.0

70.0

20.0

10.0

20.0

90.1

50.4

7Poor

0.0

10.0

50.0

10.0

10.0

20.0

60.1

00.3

2N

on

poor

0.0

10.0

90.0

20.0

10.0

30.1

10.1

90.5

9M

ale-

hea

ded

0.0

10.0

80.0

20.0

10.0

30.1

00.1

70.5

2Fe

mal

e-h

ead

ed0.0

10.0

40.0

10.0

10.0

10.0

50.0

80.2

5


Tab

leA

5.2

:(l)

Con

tin

ued

.

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

and

smal

lLe

ader

sof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

entr

ant

mer

ge

larg

est

shar

esco

mp

etit

ion

(C)

Com

ple

men

tary

pol

icie

s(fi

rms)

On

lyp

rod

uce

rsT

ota

l0.3

23.2

40.7

20.6

11.2

54.1

36.7

021.3

1Poor

0.2

52.5

10.5

60.4

70.9

73.2

05.2

016.5

2N

on

poor

0.3

63.6

90.8

30.7

01.4

34.7

17.6

524.3

3M

ale-

hea

ded

0.3

33.3

50.7

50.6

31.2

94.2

76.9

422.0

7Fe

mal

e-h

ead

ed0.2

52.5

00.5

60.4

70.9

73.1

95.1

816.4

7

All

hou

seh

old

sT

ota

l0.0

10.0

70.0

20.0

10.0

30.0

90.1

40.4

6Poor

0.0

00.0

50.0

10.0

10.0

20.0

60.1

00.3

1N

on

poor

0.0

10.0

90.0

20.0

20.0

30.1

10.1

80.5

8M

ale-

hea

ded

0.0

10.0

80.0

20.0

10.0

30.1

00.1

60.5

1Fe

mal

e-h

ead

ed0.0

00.0

40.0

10.0

10.0

10.0

50.0

80.2

5

(D)

Com

ple

men

tary

pol

icie

s(f

arm

ers−

firm

s)

On

lyp

rod

uce

rsT

ota

l0.7

93.7

11.2

20.1

30.7

93.6

37.1

521.9

8Poor

0.6

22.8

80.9

40.1

00.6

22.8

25.5

417.0

4N

on

poor

0.9

14.2

41.3

90.1

50.9

14.1

58.1

625.1

0M

ale-

hea

ded

0.8

23.8

41.2

60.1

40.8

23.7

67.4

022.7

7Fe

mal

e-h

ead

ed0.6

12.8

70.9

40.1

00.6

12.8

15.5

317.0

0

All

hou

seh

old

sT

ota

l0.0

20.0

80.0

30.0

00.0

20.0

80.1

50.4

7Poor

0.0

10.0

50.0

20.0

00.0

10.0

50.1

00.3

2N

on

poor

0.0

20.1

00.0

30.0

00.0

20.1

00.1

90.6

0M

ale-

hea

ded

0.0

20.0

90.0

30.0

00.0

20.0

90.1

70.5

3Fe

mal

e-h

ead

ed0.0

10.0

40.0

10.0

00.0

10.0

40.0

80.2

5


Tab

leA

5.2

:(l)

Con

tin

ued

.

Lead

ers

Res

pec

tm

erge

Exit

Equ

alto

Lead

erSm

all

and

smal

lLe

ader

sof

mar

ket

Per

fect

ori

gin

alsp

lit

entr

ant

entr

ant

mer

ge

larg

est

shar

esco

mp

etit

ion

(E)

Inte

rnati

onalpri

ces

On

lyp

rod

uce

rsT

ota

l4.7

97.9

75.4

33.7

92.7

90.0

711.4

128.1

2Poor

3.7

16.1

74.2

12.9

42.1

70.0

58.8

521.8

0N

on

poor

5.4

69.0

96.2

04.3

23.1

90.0

813.0

332.1

1M

ale-

hea

ded

4.9

68.2

55.6

23.9

22.8

90.0

711.8

229.1

3Fe

mal

e-h

ead

ed3.7

06.1

64.2

02.9

32.1

60.0

58.8

221.7

4

All

hou

seh

old

sT

ota

l0.1

00.1

70.1

20.0

80.0

60.0

00.2

50.6

1Poor

0.0

70.1

20.0

80.0

60.0

40.0

00.1

70.4

1N

on

poor

0.1

30.2

20.1

50.1

00.0

80.0

00.3

10.7

6M

ale-

hea

ded

0.1

10.1

90.1

30.0

90.0

70.0

00.2

70.6

8Fe

mal

e-h

ead

ed0.0

60.0

90.0

60.0

40.0

30.0

00.1

30.3

2

References

Abbott, P. (2007). Distortions to agricultural incentives in Côte d’Ivoire. AgriculturalDistortions Working Paper 46. World Bank.

Balat, J., and G. Porto (2007). Globalization and complementary policies: povertyimpacts in rural Zambia. In Globalization and Poverty (ed. A. Harrison). Universityof Chicago Press for the NBER.

Balat, J., and G. Porto (2008). Changes in the structure of the value chain in agriculturein Zambia: impacts on farm income. Mimeo, World Bank.

Barnum, H., and L. Squire (1979). A model of an agricultural household: theory andevidence. World Bank Occasional Paper 27.

Barrett, C., and P. A. Dorosh (1996). Farmers’ welfare and changing food prices: non-parametric evidence from rice in Madagascar. American Journal of Agricultural Eco-nomics 78, 656–669.

Benjamin, D. (1992). Household composition, labor markets, and labor demand: test-ing for separation in agricultural household models. Econometrica 60(2), 287–322.

Benjamin, D., and A. Deaton (1993). Household welfare and the pricing of cocoa andcoffee in Côte d’Ivoire: lessons from the living standards surveys. The World BankEconomic Review 7, 293–318.

Brambilla, I., and G. Porto (2011). Market structure, outgrower contracts and farmoutput. Evidence from cotton reforms in Zambia. Oxford Economic Papers, forth-coming.

Brooks, J., A. Croppenstedt, and E. Aggrey-Fynn (2007). Distortions to agriculturalincentives in Ghana. Agricultural Distortions Working Paper 47. World Bank.

Budd, J. W. (1993). Changing food prices and rural welfare: a nonparametric examina-tion of the Côte d’Ivoire. Economic Development and Cultural Change 41, 587–603.

Cadot, O., L. Dutoit, and J. de Melo (2009). The elimination of Madagascar’s VanillaMarketing Board, 10 years on. Journal of African Economies 18(3), 388–430.

Cheyns, E., H. A. Mrema, and B. Sallée (2006). Socio-Economic Study of the Ugandan Cof-fee Chain. Centre de Coopération Internationale en Recherche Agronomique pourle Développement.

De Janvry, A., M. Fafchamps, and E. Sadoulet (1991). Peasant household behaviour withmissing markets: some paradoxes explained. Economic Journal 101, 1400–1417.

Deaton, A. (1989a). Rice prices and income distribution in Thailand: a non-parametricanalysis. Economic Journal 99(395), 1–37.

Deaton, A. (1989b). Household survey data and pricing policies in developing coun-tries. The World Bank Economic Review 3, 183–210.

Deaton, A. (1997). The Analysis of Household Surveys: A Microeconometric Approachto Development Policy. Johns Hopkins University Press for the World Bank.

Deininger, K., and P. Olinto (2000). Why liberalization alone has not improved agri-cultural productivity in Zambia: the role of asset ownership and working capitalconstraints. World Bank Working Paper 2302.

Ennis, H. (2009). Agriculture, value chains, and income distribution in rural Africa.Mimeo, Universidad Carlos III Madrid.


Eswaran, M., and A. Kotwal (1986). Access to capital and agrarian production organi-zation. Economic Journal 96, 482–498.

FAO (1993). Agricultural Extension and Farm Women in the 1980s. Rome: FAO.FAO (2003). Issues in the Global Tobacco Economy: Selected Case Studies. Rome: FAO.Friedman, J., and J. Levinsohn (2002). The distributional impacts of Indonesia’s finan-

cial crisis on household welfare: a “rapid response” methodology. World Bank Eco-nomic Review 16, 397–423.

Gergely, N. (2009). The cotton sector in Benin. Africa Region Working Paper Series 125.World Bank.

Gittinger, J. P., S. Chernick, N. R. Horenstein, and K. Saito (1990). Household FoodSecurity and the Role of Women. World Bank.

Glover, D. J. (1984). Contract farming and smallholder outgrower schemes in less-developed countries. World Development 12(11–12), 1143–1157.

Glover, D. J. (1990). Contract farming and outgrower schemes in East and SouthernAfrica. Journal of Agricultural Economics 41(3), 303–315.

Goldberg, P., and N. Pavcnik (2004). Trade, inequality, and poverty: what do we know?Evidence from recent trade liberalization episodes in developing countries. Brook-ings Trade Forum 2004, 223–269.

Goldberg, P., and N. Pavcnik (2007). Distributional effects of globalization in develop-ing countries. Journal of Economic Literature 45(1), 39–82.

Habyalimana, S. (2007). Access to finance through value chains: lessons from Rwanda.AEGIS European Conference on African Studies, July 11–14, 2007, Leiden, TheNetherlands.

Hanson, S. (2008). C-4 cotton reform and production update. GAIN Report SG8020.USDA Foreign Agricultural Service.

Heltberg, R., and F. Tarp (2002). Agriculture supply response and poverty in Mozam-bique. Food Policy 27(2), 103–24.

Hertel, T., and A. L. Winters (eds) (2006). Poverty Impacts of the Doha DevelopmentAgenda. Palgrave–McMillan for the World Bank.

Hoekman, B., and M. Olarreaga (eds) (2007). Global Trade and Poor Nations: PovertyImpacts and Policy Implications of Liberalization. Brookings Institution Press.

Horenstein, N. R. (1989). Women and Food Security in Kenya. World Bank.Horn, H., and J. Levinsohn (2001). Merger policies and trade liberalisation. Economic

Journal 111, 244–277.International Coffee Organization (ICO). Country profiles. URL: www.ico.org.Ivanic, M., and W. Martin (2008). Implications of higher global food prices for poverty

in low-income countries. World Bank Policy Research Working Paper Series 4594.Jaffe, S. M. (2003). Malawi’s tobacco sector: standing on one strong leg is better than

on none. Africa Region Working Paper Series 55. World Bank.Jiggins, J. (1989a). How poor women earn income in sub-Saharan Africa and what

works against them. World Development 17(7), 953–963.Jiggins, J. (1989b). Agricultural technology: impact, issues and action. In The Women

and International Development Annual (ed. R. S. Gallin, M. Aronoff, and A. Ferguson),Volume 1, Part 1, Chapter 1. Boulder, CO: Westview Press.

Key, N., E. Sadoulet, and A. De Janvry (2000). Transactions costs and agriculturalhousehold supply response. American Journal of Agricultural Economics 82, 245–259.

Koyi, G. (2005). The textiles and clothing industry in Zambia. Paper prepared for theConference on the Future of African Textiles Industry, Cape Town, October 10–12,2005.

Kranton, R., and A. Swamy (2008). Contracts, hold-up, and exports: textiles and opiumin colonial India. American Economic Review 98(3), 967–989.

References 265

Laven, A. (2007). Marketing Reforms in Ghana’s Cocoa Sector. Overseas DevelopmentDepartment.

Likulunga, M. (2005). The status of contract farming and contractual arrangements inZambian agriculture and agribusiness. Report prepared for the Food, Agricultureand Natural Resources Policy Analysis Network (FARNPAN).

Lopez, R., J. Nash, and J. Stanton (1995). Adjustment and poverty in Mexican agri-culture: how farmers’ wealth affects supply response. World Bank Policy ResearchWorking Paper 1494.

Losch, B. (2002). Global restructuring and liberalization: Côte d’Ivoire and the end ofthe international cocoa market? Journal of Agrarian Change 2(2), 206–227.

Loveridge, S., J. B. Nyarwaya, and E. Shingiro (2003). Decaffeinated? Situation, trendsand prospects for smallholder coffee production in Rwanda. Rwanda Food SecurityResearch Project/MINAGRI. URL: http://aec.msu.edu/fs2/rwanda/index.htm.

Ludmer, S. (2010). A model of supply chains in export agriculture. Mimeo, UniversidadNacional de La Plata, Argentina.

Lundstedt, H., and S. Pärssinen (2009). Cocoa is Ghana, Ghana is cocoa: evaluatingreforms of the Ghanaian cocoa sector. Masters Thesis, Lund University.

Masiga, M., A. Ruhweza, and YOMA Consultants (2007). Commodity Revenue Manage-ment: Coffee and Cotton in Uganda. International Institute for Sustainable Develop-ment.

McKay, A., O. Morrissey, and C. Vaillant (1997). Trade liberalization and agricul-tural supply response: issues and some lessons. European Journal of DevelopmentResearch 9(2), 129–147.

McMillan, M., D. Rodrik, and K. H. Welch (2003). When economic reform goes wrong:cashews in Mozambique. Brookings Trade Forum 2003, 97–151.

Nicita, A. (2009). The price effect of tariff liberalization: measuring the impact onhousehold welfare. Journal of Development Economics 89(1), 19–27

Olawoye, J. E. (1989). Women in extension. In Extension for Development: Concepts andIssues (ed. W. M. Rivera and T. O. Ogunfiditimi), Chapter 14. Ibadan: Lidato Press.

Pagan, A., and A. Ullah (1999). Nonparametric Econometrics. Cambridge UniversityPress.

Porter, G., and K. Phillips-Howard (1997). Comparing contracts: an evaluation of con-tract farming schemes in Africa. World Development 25(2), 227–238.

Porto, G. (2005). Informal export barriers and poverty. Journal of International Eco-nomics 66, 447–470.

Porto, G. (2006). Using survey data to assess the distributional effects of trade policy.Journal of International Economics 70, 140–160.

Porto, G. (2007). Globalization and poverty in Latin America: some channels and someevidence. The World Economy 30(9), 1430–1456.

Porto, G. (2010). Commodity prices: impact and adjustment. Mimeo, UniversidadNacional de La Plata, Argentina.

Porto, G. (2011). On trade, poverty, and prices in developing countries. PREM Note,World Bank, forthcoming.

Poulton, C., J. Kydd, and D. Kabambe (2007). Case study on Malawi tobacco. All-AfricaReview of Experiences with Commercial Agriculture. World Bank.

RATES (2003). Cotton–textile–apparel value chain report Malawi.Ravallion, M. (1990). Rural welfare effects of food price changes under induced wage

responses: theory and evidence for Bangladesh. Oxford Economic Papers 42, 574–585.

Sahn, D., and A. Sarris (1991). Structural adjustment and the welfare of rural small-holders: a comparative analysis from sub-Saharan Africa. World Bank EconomicReview 5, 259–289.


Saito, K., and C. J. Weidemann (1990). Agricultural Extension for Women Farmers inAfrica. World Bank.

Saizonou, J. (2008). The Interprofessional Cotton Association in Benin. URL: www.inter-reseaux.org/ IMG/pdf/Fiche_AIC_premiere_version_english.pdf.

Salop, S. (1979). Monopolistic competition with outside goods. Bell Journal of Eco-nomics 10(1), 141–156.

Sexton, R., I. Sheldon, S. McCorriston, and H. Wang (2007). Agricultural trade liberaliza-tion and economic development: the role of downstream market power. AgriculturalEconomics 36, 253–270.

Sheldon, I. (2006). Market structure, industrial concentration, and price transmission.Mimeo, Ohio State University.

Singh, I., L. Squire, and J. Strauss (eds) (1986). Agricultural Household Models: Exten-sions, Applications and Policy. Baltimore, MD: Johns Hopkins Press for the WorldBank.

Stokey, N., and R. E. Lucas (1989). Recursive Methods in Economic Dynamics. Cam-bridge, MA: Harvard University Press.

Sundaram, R. K. (1996). A First Course in Optimization Theory. Cambridge UniversityPress.

Syverson, C. (2004). Market structure and productivity: a concrete example. Journal ofPolitical Economy 112(6), 1181–1222.

Taylor, E., and H. Adelman (2003). Agricultural household models: genesis, evolutionand extensions. Review of Economics of the Household 1(1), 33–58.

Tchale, H., and J. Keyser (2009). Malawi Country Economic Memorandum. SeizingOpportunities for Growth Through Trade. Volume II: Background Papers. Quanti-tative Value Chain Analysis. World Bank.

Tchale, H., and J. Keyser (2010). Quantitative value chain analysis: an application toMalawi. Policy Research Working Paper 5242, World Bank.

Trivedi, P., and T. Akiyama (1992). A framework for evaluating the impact of pricingpolicies for cocoa and coffee in Côte d’Ivoire. The World Bank Economic Review 6(2),307–330.

Tschirley, D., and S. Kabwe (2007). Cotton in Zambia: 2007 assessment of its organi-zation, performance, current policy initiatives, and challenges for the future. FoodSecurity Collaborative Working Papers 54485.

Tschirley, D., and S. Kabwe (2009). The cotton sector of Zambia. Africa Region WorkingPaper Series 124, World Bank.

Tschirley, D., C. Poulton, and P. Labaste (2009). Organization and Performance of Cot-ton Sectors in Africa: Learning from Reform Experience. World Bank.

van Donge, J. (2002). Disordering the market: the liberalisation of burley tobacco inMalawi in the 1990s. The Journal of Southern African Studies 28(1), 89–115.

Vargas Hill, R. (2010). Liberalisation and producer price risk: examining subjectiveexpectations in the Ugandan coffee market. Journal of African Economies 19(4),433–458.

Vigneri, M., and P. Santos (2007). Ghana and the Cocoa Marketing Dilemma: What HasLiberalization without Price Competition Achieved? Overseas Development Institute.

Warning, M., and N. Key (2002). The social performance and distributional conse-quences of contract farming: an equilibrium analysis of the Arachide de Boucheprogram in Senegal. World Development 30(2), 255–263.

Wilcox, J. M., and P. Abbott (2004). Market power and structural adjustment: the caseof West African cocoa market liberalization. American Agricultural Economics Asso-ciation Annual Meeting.

Winters, A., N. McCulloch, and A. McKay (2004). Trade liberalization and poverty: theevidence so far. Journal of Economic Literature 42, 72–115.

References 267

Wodon, Q., C. Tsimpo, P. Backiny-Yetna, G. Joseph, F. Adoho, and H. Coulombe (2008).Potential impact of higher food prices on poverty: summary estimates for a dozenWest and Central African countries. World Bank Policy Research Working PaperSeries 4745.

World Bank (2005). Diagnostic Trade Integration Study of Benin. World Bank.World Bank (2008). Pathways to greater efficiency in the Malawi tobacco industry.

Poverty Reduction and Economic Management, AFTP1. World Bank, Africa Region.World Trade Organization (2004a). Burkina Faso Trade Policy Review. Geneva: World

Trade Organization.World Trade Organization (2004b). Rwanda Trade Policy Review. Geneva: World Trade

Organization.Yartey, C. (2008). Tackling Burkina Faso’s cotton crisis. IMF Survey Magazine. URL:

www.imf.org/external/pubs/ft/survey/so/2008/CAR022508B.htm.

Supply Chains in Export Agriculture,

Competition, and Poverty in

Sub-Saharan Africa

“The last two decades’ reforms in Africa’s agricultural marketing channels have taken place against a background of relative ignorance of how these markets work. Combining theory (with coverage of complex contractual arrangements like outgrower contracts), household surveys, and in-depth knowledge of local contexts, this masterful book provides the first systematic answer. In their characteristically careful approach, the authors use simulation analysis based on oligopoly theory to isolate and quantify the effect of policy shocks one by one and with synergies, yielding precise orders of magnitude where theory is usually silent. Written in a limpid style, this book is a must-read for academics and sophisticated policy analysts. It will be a reference for years to come.”Olivier Cadot, Professor of International Economics and Director of the Institute of Applied Economics at the University of Lausanne

“This is an innovative and important book. The authors explicitly model the institutions and industrial organization of global trade and commodity exchanges, which have major implications for the efficiency and surplus distribution among the participants in the chain. The combination of theory and empirical analysis across many developing countries is unique and yields important new insights.”Jo Swinnen, Professor of Development Economics at K.U.Leuven, Director of LICOS-Centre for Institutions and Economic Performance at K.U.Leuven and Senior Research Fellow in the Centre for European Policy Studies (CEPS), Brussels


by: Guido Porto, Nicolas Depetris Chauvin and Marcelo Olarreaga

THE WORLD BANK

THE WORLD BANK

9 781907 142208

supply chains in export agriculture, competition, and...

Documents