39044 W O R L D B A N K W O R K I N G P A P E R N O . 8 8 Conflict, Livelihoods, and Poverty in Guinea-Bissau Edited by Barry Boubacar-Sid, Edward G. E. Creppy, Estanislao Gacitua-Mario, and Quentin Wodon THE WORLD BANK W O R L D B A N K W O R K I N G P A P E R N O . 8 8 Conflict, Livelihoods, and Poverty in Guinea-Bissau Edited by Boubacar-Sid Barry, Edward G. E. Creppy, Estanislao Gacitua-Mario, and Quentin Wodon THE WORLD BANK Washington, D.C. Copyright © 2007 The International Bank for Reconstruction and Development / The World Bank 1818 H Street, N.W. Washington, D.C. 20433, U.S.A. All rights reserved Manufactured in the United States of America First Printing: January 2007 printed on recycled paper 1 2 3 4 5 10 09 08 07 World Bank Working Papers are published to communicate the results of the Bank's work to the devel- opment community with the least possible delay. The manuscript of this paper therefore has not been pre- pared in accordance with the procedures appropriate to formally-edited texts. Some sources cited in this paper may be informal documents that are not readily available. The findings, interpretations, and conclusions expressed herein are those of the author(s) and do not necessarily reflect the views of the International Bank for Reconstruction and Development/The World Bank and its affiliated organizations, or those of the Executive Directors of The World Bank or the gov- ernments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank of the legal status of any territory or the endorsement or acceptance of such boundaries. The material in this publication is copyrighted. Copying and/or transmitting portions or all of this work without permission may be a violation of applicable law. The International Bank for Reconstruction and Development/The World Bank encourages dissemination of its work and will normally grant per- mission promptly to reproduce portions of the work. For permission to photocopy or reprint any part of this work, please send a request with complete information to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA, Tel: 978-750-8400, Fax: 978-750-4470, www.copyright.com. All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, The World Bank, 1818 H Street NW, Washington, DC 20433, USA, Fax: 202-522-2422, email: pubrights@worldbank.org. ISBN-10: 0-8213-7021-9 ISBN-13: 978-0-8213-7021-6 eISBN: 978-0-8213-7022-3 ISSN: 1726-5878 DOI: 10.1596/978-0-8213-7021-6 Boubacar-Sid Barry is Economist in the Poverty Reduction and Economic Management office of the Africa Region at the World Bank. Edward G. E. Creppy is Economist in the same office. Estanislao Gacitua-Mario is Senior Social Scientist in the Social Development Department at the World Bank. Quentin Wodon is Lead Poverty Specialist in the Poverty Reduction and Economic Management office of the Africa Region at the World Bank. Library of Congress Cataloging-in-Publication Data Conflict, livelihoods, and poverty in Guinea-Bissau / edited by Boubacar-Sid Barry . . . [et al.]. p. cm.--(World Bank working paper ; no. 88) Includes bibliographical references and index. ISBN-13: 978-0-8213-7021-6 ISBN-10: 0-8213-7021-9 eISBN-13: 978-0-8213-7022-3 1. Guinea-Bissau--Economic conditions--1974- 2. Guinea-Bissau--Social conditions. 3. Guinea-Bissau--Politics and government--1974- I. Barry, Boubacar-Sid, 1969- II. World Bank. HC1080.C67 2007 330.96657--dc22 2006101052 Contents Foreword vii Abstract ix Acknowledgments xi Acronyms and Abbreviations xiii 1. Conflict, Livelihoods, and Poverty in Guinea-Bissau: A Brief Overview 1 By Boubacar-Sid Barry, Estanislao Gacitua-Mario, and Quentin Wodon Conflict and Political Instability in Guinea-Bissau 3 Conflict, Livelihoods, and Poverty: Contributions in This Volume 4 Next Steps 7 References 8 2. Conflict, Growth, and Poverty in Guinea-Bissau 11 By Boubacar-Sid Barry and Quentin Wodon Guinea-Bissau's Conflict and Economic Performance: A Brief Review 12 Impact of the Conflict on GDP Per Capita and Poverty 16 Conclusion 21 References 22 3. Institutions, Social Networks, and Conflicts in Guinea-Bissau: Results from a 2005 Survey 23 By Estanislao Gacitua-Mario, Sigrun Aasland, Hakon Nordang, and Quentin Wodon Governance and Recent Political Developments: A Brief Review 25 Household Perceptions: Results from a Recent Survey 30 Conclusion 40 References 40 4. Poverty and Its Determinants in Guinea-Bissau 43 By Edward G. E. Creppy and Quentin Wodon Poverty Estimates and Income Sources 44 Poverty Profile and Determinants of Poverty 49 Conclusion 56 References 57 iii iv World Bank Working Paper 5. Livelihoods in Guinea-Bissau 59 By Estanislao Gacitua-Mario, Hakon Nordang, and Quentin Wodon Livelihoods in Rural Areas 60 Livelihoods in Urban Areas 63 Income Sources, Cashew Nuts, and Fishing 66 Conclusion 73 References 74 6. Cashew Production, Taxation, and Poverty in Guinea-Bissau 77 By Boubacar-Sid Barry, Edward G. E. Creppy, and Quentin Wodon Guinea-Bissau's Cashew Sector: A Brief Review 78 Distributional Issues in Cashew Production, Pricing, and Taxation 84 Conclusion 88 References 88 LIST OF TABLES 2.1 Impact of the Conflict on Per Capita GDP 18 2.2 Impact of the 1998 Conflict on Poverty 21 3.1 Selected ICGR Indicators of Cohesion in 2005 26 3.2 Subjective Perceptions of Well-Being, 2005 32 3.3 Factors Contributing to Inequality in 2005 34 3.4 The Institutions Most Trusted by Citizens, 2005 35 3.5 Perceptions of the Security Situation and Sources of Conflict, 2005 37 4.1 Expenditure Shares by Quintile, 2002 45 4.2 Population-based Poverty Measures by Region and Areas, 2002 46 4.3 Number of Years Required to Achieve Cumulative Growth and Poverty Targets 49 4.4 Revenue Shares by Quintile of Per Capita Consumption, 2002 50 4.5 Revenue Shares by Type of Activity and by Region, 2002 50 4.6 Repartition of Poverty by Group and Headcount Index for Selected Characteristics, 2002 52 4.7 Impact of Household Characteristics on Welfare, 2002 54 5.1 Main Sources of Income for Households by Tercile, 2005 68 6.1 Cashew Sales by Tercile, 2005 85 6.2 Impact of Change in Cashew Prices on Poverty 87 LIST OF FIGURES 2.1 Visualization of Hypothetical AO, TC, and LS Outliers 17 2.2 Impact of the 1998 Conflict on Per Capita GDP 19 3.1 Selected Governance and Accountability Indicators, 2004 25 Contents V 4.1 Problem Satisfying the Food Needs of Households by Quintile, 2002 47 4.2 Mean Values of Well-Being Indicators by Consumption Quintiles, 2002 53 6.1 Real GDP and Cashew Production Volumes, 1999­2003 79 6.2 Breakout of Cashew Export Costs 80 6.3 Trend in Cashew Prices, 1979­2003 81 6.4 Cashew Nut Production, 1979­2003 83 LIST OF BOXES 3.1 Land Regulation in Guinea-Bissau--A Review 39 Foreword T his book seeks to provide an integrated poverty and social analysis of postconflict Guinea-Bissau. The rationale for going beyond a traditional quantitative analysis of poverty is the conviction that understanding social and political processes is essential to induce and support changes leading to poverty alleviation, equitable growth, and good governance. The book relies in part on Country Social Analysis (CSA), an approach developed to identify policy recommendations for removing the barriers to socioeconomic mobility and inclusion as well as for strengthening the equitable impacts of policy reforms. The analysis involves two basic building blocks: (1) the examination of social diversity and economic opportunities, or what impact social diversity has on the economic opportunities (liveli- hoods) of the different social groups (whether based on sex, age, ethnicity, location, or cul- ture); and (2) analysis of power relationships, local institutions, and their governance--or what the relationship is between policymaking and resource reallocation. On the basis of this analysis, the CSA approach aims to inform macro-level policy dialogue and draw rec- ommendations on three levels: (1) removing institutional and group-based constraints to access assets, services, and public goods; (2) increasing accountability of institutions and participation in decisionmaking; and (3) reducing structural and development-induced risks in Bank operations. The papers presented in this volume were originally prepared for the Guinea-Bissau Integrated Poverty and Social Assessment (IPSA), which used the CSA approach to reach a broader understanding of the pervasive nature of poverty in Guinea-Bissau. The analytical work was implemented by a joint team from the World Bank's Social Development Depart- ment and the Poverty Reduction and Economic Management unit in the African region. The objective was to build knowledge of the interconnections between socioeconomic and insti- tutional factors, instability, poverty, and growth. Most of the papers presented in this vol- ume integrate economic and social analysis to evaluate the macroeconomic and fiscal framework as well as policies in the rural and social sectors, taking into account the policies' impacts on sociopolitical stability. The papers rely on existing, nationally representative household survey data as well as additional data collected through a small-scale survey on household perceptions, livelihoods, land access, and the cashew economy, designed specif- ically for this project. The quantitative data collection and analysis was complemented by a range of focus groups and participatory rural appraisals. Overall, the papers presented here provide an overview of key institutional dimensions of Guinea-Bissau's postconflict context, and an in-depth analysis of poverty and livelihoods. Sudhir Shetty Steen Lau Jorgensen Director Director Poverty Reduction and Economic Management Social Development Africa Region Vice Presidency Sustainable Development Network vii Abstract C onflict and political instability have considerably weakened Guinea-Bissau's productive infrastructure during the past three decades. This situation has contributed to an in- crease in the degree of vulnerability of the population, especially in rural areas where most economic activities continue to take place. As growth has been weak, poverty levels have remained high. This book provides a collection of papers on conflict, livelihoods, and poverty in Guinea-Bissau, based on both a nationally representative 2002 household survey and a small-scale 2004 survey with both quantitative and qualitative components. The chapters deal with growth and poverty, institutions and social networks, determinants of poverty, means of livelihood, and cashew production and taxation. ix Acknowledgments T he papers included in this book were prepared by a team composed of Boubacar-Sid Barry, Edward G. E. Creppy, Estanislao Gacitua-Mario, and Quentin T. Wodon, as well as Hakon Nordang and Sigrun Aasland. The papers were originally prepared as a contribution to the World Bank's Integrated Poverty and Social Assessment (IPSA) for Guinea-Bissau. Some of the papers build on qualitative and quantitative data collected by the National Institute of Research and Census, under the leadership of Mamadú Jao, the director general of the National Institute of Research and Census (INEP). The papers also build on collaborations with the National Institute of Statistics (INEC) and officials of the Ministry of Education (MOE). The INEC team was composed of Adulai Jalo (economist) and Bessa Vitor da Silva (demographer). The team from the MOE included Mario Nosolini (director general) as well as Mamadu Saliu Jassi, Ibrahima Diallo, and Adulai So (education specialists). The IPSA report as well as the background papers included in this volume benefited from financial support from the trust fund managed by the Environmentally and Socially Sustainable Development (ESSR) Network in the World Bank's Africa Poverty Reduction and Economic Management (PREM) front office. Robert R. Blake and Marcelo Andrade provided analytical advice and ensured quality control and management support for the task as a whole. Madani M. Tall and Iradj A. Alikhani offered conceptual guidance and sup- ported the process. The views expressed here are those of the authors, and do not necessar- ily represent those of the World Bank, its executive directors, or the countries they represent. xi Acronyms and Abbreviations CBOs Community-based Organizations CEPAL Communidade dos Paises de Lingua Portuguesa CFAF Franc of the African Financial Community CMB Municipal Council of Bissau CNE Commission Nationale Electorale CSOs Civil Society Organizations CWIQ Core Welfare Indicator Survey ECOWAS Economic Community of West African States EEMP Emergency Economic Management Plan EU European Union FAO Food and Agriculture Organization GDP Gross Domestic Product GNU Government of National Union HDI Human Development Index HIPC Heavily Indebted Poor Countries ICOF Full-Fledged Household Survey ILAP Light Household Survey IMF International Monetary Fund INEC National Institute of Statistics INEP National Institute of Research and Census IPSA Integrated Poverty and Social Assessment LICUS Low-Income Country Under Stress NGOs Non-Governmental Organizations NRRP National Reconciliation and Rehabilitation Program PAIGC Partido Africano para a Independência da Guiné-Bissau (African Party for the Independence of Guinea-Bissau) PRGF Poverty Reduction and Growth Facility PRS Partido de Renovação Social (Social Renovation Party) PRSP Poverty Reduction Strategy Paper PSR Public Sector Reform SAGB Cotton Society of Guinea-Bissau TIPS Trade and Investment Promotion Support UNDP United Nations Development Programme USAID U.S. Agency for International Development US$ U.S. Dollars xiii CHAPTER 1 Conflict, Livelihoods, and Poverty in Guinea-Bissau A Brief Overview Boubacar-Sid Barry, Estanislao Gacitua-Mario, and Quentin Wodon Conflict and political instability have considerably weakened Guinea-Bissau's productive infrastructure during the past three decades. This situation has contributed to an increase in the degree of vulnerability of the population, especially in rural areas. As growth has been weak, poverty levels have remained high. This book provides a collection of papers on conflict, livelihoods, and poverty in Guinea-Bissau based on both a nationally representative 2002 household survey and a small-scale survey with both quantitative and qualita- tive components implemented in 2004. After a brief review of political develop- ments since independence, this overview summarizes the report's main findings in the areas of conflict, growth and poverty, institutions and social networks, the characteristics of the poor and determinants of poverty, the population's means of livelihood, and, finally, cashew production and taxation. T his book is about the social and economic impacts of conflict in Guinea-Bissau. There has been a substantial body of research in recent years on conflict, growth, and poverty; however, relatively little has been written on Guinea-Bissau, in large part because good data sources on the country have been lacking. Today, the availability of new household survey data makes it possible to conduct a meaningful analysis of the impact of conflict on Guinea-Bissau, with a focus on the poor, tensions that still remain, and several key reforms that could improve the future of its citizens. Before discussing the specific case of Guinea-Bissau, it is worth emphasizing that many of the broad lessons provided in the literature on conflicts apply to the specific case of Guinea-Bissau. According to Rodrick (1999), conflicts are important to understanding why economic growth is inconsistent and why the growth of so many countries has col- lapsed since the mid-1970s. The countries that experienced the sharpest drops in growth 1 2 World Bank Working Paper after 1975 were those with divided societies (as measured by indicators of inequality and ethnic fragmentation) and with weak institutions of conflict management (as measured by indicators such as the quality of governmental institutions, rule of law, democratic rights, and social safety nets). These findings were also confirmed by Alesina and others (1996), who concluded that growth is significantly lower than normal in countries and time periods with a high propensity for government collapse. Studies also point to spillover effects of regional conflicts on the economic performance of countries (Sambanis 2003). Conflicts in neighboring countries also heighten the risk that the country itself will fall into civil war (Collier and others 2003),1 a fact that is well known in West Africa. It is often argued that the risks of conflicts are explained primarily by economic factors (such as initial income, dependence on commodity exports, and the economic growth of the preceding period), but they are also dependent on non-economic factors (such as social fragmentation). Recent research has established that inefficient institutions lead to ineffi- cient policies, which can, in turn, cause conflicts. According to Acemoglu (2006), economic institutions determine the framework for policymaking, placing heavy constraints on pol- icy. Groups with political power, or the elite, often choose policies that transfer resources from the rest of the society to themselves, which leads institutions to be inefficient from the point of view of society at large. Resource transfers can be obtained in a variety of ways, such as through revenue extraction, factor price manipulation, and political consolidation.2 Linking institutional inefficiency to conflict, Yartey (2004) found that, based on a cross-country analysis, weak institutions and low economic growth are major causes of conflicts in Sub-Saharan Africa. Along the same lines, Ghura and Mercereau (2004) pro- vided empirical evidence that the propensity for political instability and conflict in the Cen- tral African Republic has been increased by low tax revenues and deteriorations in trade. In contrast, Campos and others (1999)--in a study on the Middle East and North Africa over the period 1970­90--suggested that domestic political instability can have positive effects on investment by forcing governments to adopt more efficient policies. However, external political instability tends to reduce domestic political instability and thus makes policies more inefficient. In the specific case of Guinea-Bissau, conflict and political instability have weakened the country's productive infrastructure during the past three decades, and this has led to an increase in the vulnerability of the population, especially in rural areas. As growth has been weak, poverty levels have remained high. In this context, the objective of this book is to provide an analysis of the links between conflict, livelihood, and poverty using recent household survey data (the nationally representative 2002 household survey as well as a small-scale survey with both quantitative and qualitative components implemented in 2004). After a brief review of political developments since independence, this chapter sum- marizes the main findings of the chapters that deal with conflict, growth and poverty, insti- tutions and social networks, the characteristics of the poor and determinants of poverty, citizens' means of livelihood, and finally, cashew production and taxation. Several areas for future research are highlighted in the conclusion. 1. The subject is further covered by Gudmundsson (2004), Elbadawi and Sambanis (2000), Fosu (2001), Collier (1999), Collier and Hoeffler (2002), and Easterly and Levine (1997). See also Sachs and Warner (1997) on growth in Africa. 2. See also Acemoglu (2002, 2003, 2006a, 2006b). Conflict, Livelihoods, and Poverty in Guinea-Bissau 3 Conflict and Political Instability in Guinea-Bissau Following a decade-long freedom war with Portugal, Guinea-Bissau became independent in 1974, though the country remained plagued by political instability and conflicts. The government of the first president, Luis Cabral, was overthrown in a bloodless coup led by the prime minister and former armed forces commander João Bernardo "Nino" Vieira in 1980. Several coup attempts against the Vieira government followed in 1983, 1985, and 1993. In 1994, the country's first multiparty legislative and presidential elections were held. President Vieira of the ruling Party for the Independence of Guinea-Bissau and Cape- Verde (PAIGC) was re-elected with a slight margin over Kumba Yala of the Party of Social Restoration (PRS). However, an army uprising against the Vieira government in June 1998 triggered a bloody civil war that displaced over 300,000 members of the civilian population. Failed governance, breakdown of the rule of law, and limited accountability and trans- parency of public sector management were among the causes of the 1998 conflict, which lasted nearly a year. By June 1999 the situation, although still fragile, improved steadily. A Government of National Union was put in place and charged with the tasks of preparing the National Reconciliation and Rehabilitation Program and organizing elections. Free and fair parlia- mentary elections were held on November 28, 1999, and the presidential runoff vote took place on January 16, 2000. Opposition parties scored a landslide victory in both elections. Kumba Yala of the PRS was elected president, and took office on February 17, 2000. Shortly after, a broad coalition cabinet, including members of various parties represented in the national assembly, was appointed. Following the elections, parliament approved a new constitution in April 2001, but it was neither vetoed nor promulgated by the president. The resulting ambiguity undermined the rule of law, and a prolonged friction developed between the executive and legislative branches of the government over the choice of the new cabinet chief. As a result, the country operated without an effective government for several weeks, and the announcement of yet another failed coup attempt in mid-April 2001 demonstrated once again the fragility of the political and democratic processes. The prevailing political instability, which affected all institutions, prevented an economic recovery and postconflict reconstruction. In December 2001, the government allegedly thwarted a coup attempt by army offi- cers. Between December 2001 and December 2002, President Yala carried out two cabinet reshuffles, dissolved the parliament, and called for legislative elections, which were repeat- edly postponed. In the ensuing discontent among the population, the army was able to seize power in a bloodless coup in September 2003. President Yala then announced his voluntary resignation and was placed under house arrest. The military, led by the chief of defense, Gen. Verissimo Seabra,3 quickly restored power to civilian control, and a civilian caretaker government was appointed with the objective of restoring the rule of law and reasserting control over public finances. In this climate, a businessman, Henrique Rosa, was sworn in as interim president, with the support of almost all political parties as well as the civil society. By the end of 2003, with the support of the international community, the newly appointed government prepared an Emergency Economic Management Plan (EEMP) and a budget for 2004. Legislative elections were held in March 2004 and an 3. General Seabra was killed along with other officers during a mutiny on October 6, 2004. 4 World Bank Working Paper orderly transition to a government elected on the basis of a broad reform platform took place in May 2004. The transition government, led by the leader of the Partido Africano para la Independência da Guiné-Bissau (PAIGC), Carlos Gomes Júnior was appointed with the objective of ensuring the functioning of public administration and preparing presidential elections by mid-2005. Despite the difficult economic situation, the government drafted a Poverty Reduction Strategy Paper and significantly improved budgetary discipline during the second half of 2004. The government also endorsed a reform program, with the agreement of the inter- national community; prepared the 2005 budget; and successfully organized a free, fair, and transparent presidential election. This election was won by the independent candidate and former president, Vieira.4 The successful elections put an end to the postconflict transition that began after the civil war and brought a return to constitutional and democratic rule in Guinea-Bissau following years of political instability and administrative chaos. At the end of October 2005, President Vieira dismissed the PAIGC government of Prime Minis- ter Gomes Júnior. In early November, Vieria swore in a five-party coalition government led by PAIGC dissident Aristides Gomes. The PAIGC continues to claim the legitimacy of the new cabinet although several parliamentarians are believed to have defected from its ranks. In early 2006, the supreme court of the country validated the nomination of the new prime minister, and the national assembly approved the government's program. These developments provided a window of opportunity for scaling up reforms and economic reconstruction. Conflict, Livelihoods, and Poverty: Contributions in this Volume This publication seeks to provide an integrated poverty and social analysis of postconflict Guinea-Bissau. The rationale for going beyond a traditional quantitative analysis of poverty is the conviction that understanding social and political processes is essential to induce and support the changes needed for poverty alleviation, equitable growth, and good governance. The publication consists of six chapters. Following this brief introduction, the second chapter measures the impact of the conflict on per capita GDP in Guinea-Bissau. It is worth emphasizing that the country performed in a satisfactory manner before the 1987­98 con- flict. Following a decade of state-controlled economic policy, authorities embarked on a comprehensive reform program in 1987, and agreed upon a three-year enhanced struc- tural adjustment facility arrangement with the International Monetary Fund (IMF) in 1995. Broad liberalization of the economy contributed to a reduction in macroeconomic imbalances. Between 1996 and 1997, the public sector primary surplus increased from 3 to 5 percent, and the external current account deficit (excluding official transfers) decreased from 18 to 14 percent. Inflation decreased from a record 50 percent at the end of 1997 to 4. The first-round poll had seven candidates, including three former presidents: Malam Bacai Sanhá of the PAIGC, independent candidate João Bernardo "Nino" Vieira, and Kumba Yala of the PRS. The second-round poll included Sanhá and Vieira. The latter was declared the winner in the second round, with 52 percent of the votes and a strong backing from Yala's supporters. Eight teams of international observers, including an 80-strong mission from the European Union (EU), issued a joint statement saying the first- and second-round polls were "free, fair, and transparent." Conflict, Livelihoods, and Poverty in Guinea-Bissau 5 8 percent in early 1998, owing to restrained public sector demand and stability regained by becoming a member of the West African Economic and Monetary Union. Domestic production, particularly in agriculture, accelerated with average real GDP growth reaching more than 5 percent from 1987 to 1997. Many of these gains vanished with the 1998 conflict. The disruption of the economy was significant as real GDP contracted by about 28 percent in 1998 because of a sharp decline of production activities in all sectors. The conflict also adversely affected the country's fiscal and balance of payments positions. The external current account deficit (excluding official transfers) remained high at nearly 16 percent of GDP and the current primary budget balance, moved from a surplus of about 5 percent of GDP in 1997 to a deficit of about 7 percent of GDP in 1998. As a result, domestic and external debt service arrears piled up to reach nearly 15 percent of the GDP. After the war, the real annual GDP grew by nearly 8 percent from 1999 to 2000, but macroeconomic performance deteriorated again from 2001 to 2004. In Chapter 2, the impact of the conflict on GDP per capita, and on poverty, is assessed using a technique to identify and correct outliers in a time series that was applied previ- ously in an assessment of conflicts' impact on GDP per capita in Rwanda (and, later, a number of Sub-Saharan African countries) by Lopez and Wodon (2005). The estimates suggest that GDP per capita would be roughly 42 to 43 percent higher today if the conflict had not occurred. In turn, assuming that the conflict has had no impact on inequality (good data does not exist on the potential impact on inequality), the authors find that about one-third of the population now in poverty can be said to be poor because of the conflict. In Chapter 3, the focus shifts to popular perceptions of general well-being and insti- tutions. In Guinea-Bissau, boundaries between the public and private sectors are often poorly defined. The financial accountability system remains weak, and the disclosure of information on public accounts is limited. This has contributed to the erosion of public confidence in the government and, thus, citizens' willingness to pay taxes or support gov- ernment policies. As a result, indicators on the rule of law, government effectiveness, and the overall risk rating of the International Country Risk Guide for Guinea-Bissau are among the lowest in Sub-Saharan Africa. The weak governance situation dates back to the colonial era, when the state pro- moted a limited government presence in rural areas. Traditional authorities, therefore, played a leading institutional role in local communities under colonial rule, and the postindependence state largely inherited the weak institutional legacy of the colonial era. Traditional authorities, particularly in rural areas, continued to wield power in rural com- munities, either by dominating the state-sanctioned village committees or by sidelining them. In the end, the village committee system gradually became defunct, leaving the state with little presence at the village level. At the same time, social and political leaders estab- lished strong paternalistic relationships with their constituencies without building broader coalitions or strengthening the legitimacy of existing state institutions. In recent years, the limited progress that has been made in social service delivery, notably in education and health, are largely due to the active role played by civil society organizations and the sup- port of the international community. Citizens' lack of confidence in many institutions is confirmed by data analysis in Chapter 3. The authors provide recent survey results that give insights into popular opinions 6 World Bank Working Paper about, among other issues, changes in well-being over time, trust in various institutions, sources of conflict at the local level, and ways of dealing with conflict. There is a clear per- ception among the population that there has been a reduction in well-being as a result of the conflict, as well as a lack of improvement in the conditions of the poor, who are highly vulnerable to the effects of conflict. The data also suggest a lack of improvement in per- ception regarding postconflict security. Local conflicts often emerge because of the com- petition for scarce productive resources, but poorer households deal with these conflicts in a different way than wealthier households. Importantly, public perceptions of different institutions indicate that local schools and health posts are the most trustworthy institu- tions. In contrast, the populace has little trust in national institutions such as the army, the police, the judicial system, and the central government. Nongovernmental organizations (NGOs) and other civil society or community-based organizations often supplement the weak presence of the government in rural areas. Chapter 4 estimates the share of the population in poverty in 2002, predicts how it may have evolved since then, and assesses the level of growth required to reduce poverty in the future. Within the general population, the estimated poverty and extreme poverty head counts are 65.7 percent and 21.6 percent, respectively. Findings from the small- scale survey, completed in 2005, suggest that most households believe that their living situation has worsened not only compared to the preconflict period, but also in years since. This is confirmed by simulations of changes in poverty after 2002 as related to growth in GDP per capita. These simulations suggest that the number of those in poverty may have increased from 65.7 percent in early 2002 to 72.3 percent by the end of 2005. This means that it will be difficult to reach the first target of the Millennium Develop- ment Goals. For example, if the economy were to grow at 4 percent per year, it would take 11 years to achieve a cumulative growth rate of 25 percent, which would, in turn, reduce the share of the population in poverty from the estimated level of 72.3 percent in 2005 only to 58.8 percent by 2020. The chapter also provides a brief poverty profile and an analysis of the determinants of poverty using the nationally representative survey of 2002. Geographic location, demographic structure (both household size and headship), employment (both in terms of sector and type), education, and migration all have potentially large effects on the consumption level and thereby the poverty of households. For example, the incidence of poverty among household heads with tertiary education is less than half of that for heads with no level of education. Poverty incidence is also higher among households with heads working in agriculture, compared to government and private sector workers. The fifth chapter analyzes modes of livelihood among Guinea-Bissau's population. The country is poor and highly dependent on agriculture and extractive activities such as fishing. Yet, beyond such general statements, it is useful to describe in qualitative terms the main activities that take place both in rural areas at the regional level, and in urban areas through mostly informal employment (as already discussed in Lourenço-Lindell, 2002). This is done on the basis of fieldwork conducted in the country in early 2005. In addition, the chapter provides a brief quantitative, empirical analysis of the main income sources of households, and a more detailed analysis of two key sectors of the economy that could form the basis of a growth strategy: cashew nuts and fishing. In addition, Chapter 5 suggests that most urban families essentially live off of activi- ties in the informal sector. For urban men, casual work is increasingly the main activity and Conflict, Livelihoods, and Poverty in Guinea-Bissau 7 livelihood strategy. This represents a serious cause of vulnerability as formal wage oppor- tunities become limited. As for women, they have historically played an important role in small trading, domestic work, dressmaking, weaving, the extraction of cashew juice and wine, and so on. During periods of harvests, some people, especially younger women and youth, seasonally migrate to rural areas to offer their labor to farmers, particularly cashew producers. In rural areas, subsistence agriculture represents the main livelihood strategy. Rice remains the most widely grown staple crop, but cashews have played an increasingly important role, providing an important source of income for many households. The liveli- hood strategies of the rural poor tend to differ across regions and are also influenced by the social characteristics of households and individual household members (such as ethnicity, family structure, sex, religion, age, and so on). For example, agriculture and cattle are the main livelihood means in the northern and eastern regions and in part of the south (rice), while fishing is common in the south and the islands. The sixth and last chapter provides a review of the cashew sector in Guinea-Bissau, as well as estimates of the likely impact of changes in farm-gate prices and export taxes on poverty among cashew producers and in the country as a whole. Relatively high transac- tion costs represent an implicit tax on cashew producers. Reducing these costs would likely increase the farm-gate price for cashew producers, and therefore improve the consump- tion of goods and services. The available evidence suggests that a 15 percent increase in farm-gate prices for cashew nuts could result in an increase in consumption of 9.5 percent for the extreme poor, and 3.3 percent for the poor. In contrast, an increase of about 10 per- cent in the export tax on unprocessed cashew nuts potentially could generate (with full pass-through to farmers) a reduction in the farm-gate price of close to 15 percent, assum- ing that farmers get 70 percent of the export price. The chapter also notes that over the last three decades, the production of rice has decreased significantly in favor of cashews. This situation may, under some circumstances, represent a threat to food security. For the rural sector to ensure food security and create new jobs, policymakers would need to adopt a coherent agrarian development strategy in the context of the Poverty Reduction Strategy, which would aim to rehabilitate and encourage rice production while promoting the pro- cessing of raw cashews into exportable cashew kernels in order to generate more added value in the cashew sector. Next Steps The analysis presented in this volume is only a beginning, as several key policy areas have not been dealt with that warrant specific attention. Of special importance is the need for Guinea-Bissau's authorities to restore fiscal sustainability through public sector reform. This need is partly because of recent conflicts and the resulting increase in the size of the security forces, but also because the public wage bill has become excessive following recent increases, leading to large budget deficits and the scarcity of resources that should be allo- cated to poverty reduction. There is a serious need to bring the public sector wage bill, exclud- ing education and health care, to a sustainable level, and to reform the security sector. Apart from freeing resources for attacking poverty, preliminary evidence suggests that a reduction in the wage bill could also be good for growth. Indeed, if real government con- sumption (namely wage spending) were reduced by 2 percent (5 percent in nominal terms 8 World Bank Working Paper corresponding to 1 percentage point of the GDP) annually over 2006­08, the estimates sug- gest that real GDP could increase by about 1.6 and 2.9 percentage points in 2007 and 2008, respectively, compared to baseline growth targets for the same years. Public sector downsizing will, however, not be sufficient for achieving medium-term fis- cal sustainability. The government will need to pursue sound expenditure management and improve the efficiency of revenue collection by reinforcing and sustaining the ongoing reforms in key departments of the ministries of finance and fisheries to achieve a current surplus. There is also a need to address the size of the pension system and the ethnic composition of the armed forces over the medium- and long-term. The sustainability of Guinea-Bissau's external debt also remains a concern, as medium-term debt sustainability will depend on continued fiscal adjustment and support from the international community. Finally, persistent internal political tensions, the weak institutional and governance capacity of the state, and the vulnerability of cashew nuts to external price fluctuations represent ongoing challenges. References Acemoglu, Daron. 2003. "The Form of Property Rights: Oligarchic Versus Democratic Societies." NBER Working Paper No. 10037. ------. 2006. "Modeling Inefficient Institutions." Forthcoming in R. Blundell, W. Newey, and T. Persson, eds., Advances in Economic Theory and Econometrics: Proceedings of 2005 World Congress of the Econometric Society. Acemoglu, Daron, and J. A. Robinson. 2006a. Economic Origins of Dictatorship and Democracy. Cambridge University Press. ------. 2006b. "Economic Backwardness in Political Perspective." American Political Science Review 100:115­31. Acemoglu, D., S. Johnson, and J. A. Robinson. 2002. "Reversal of Fortune: Geography and Institutions in the Making of the Modern World Income Distribution." Quarterly Journal of Economics 117 (November):1231­94. Alesina, A, S. Ozler, N. Roubini, and P. Swagel. 1996. "Political instability and economic growth." Journal of Economic Growth 1:189­211. Campos, N. F., J. B. Nugent, and J. A. Robinson. 1999. "Can Political Instability be Good for Growth? The Case of Meddle East and North Africa." Department of Economics, University of Southern California, Los Angeles, California. Collier, P. 1999. "On the Economic Consequences of Civil War." Oxford Economic Papers 51 (January):163­83. Collier, P., and A. Hoeffler. 2002. "Greed and Grievance in African Civil Wars." CSAE Working Paper No. WPS/2002­01. Centre for African Economies, Oxford. Collier, P., V. L. Elliott, H. Hegre, A. Hoeffler, M. Reynal-Querol, and N. Sambanis. 2003. Breaking the Conflict Trap: Civil War and Development Policy. Washington, D.C.: The World Bank and Oxford University Press. Easterly, W., and R. Levine. 1997. "Africa's Growth Tragedy: Policies and Ethnic Divisions." Quarterly Journal of Economics 112 (November):1203­50. Elbadawi, I., and N. Sambanis. 2000. "Why Are There So Many Civil Wars in Africa? Understanding and Preventing Violent Conflict." Journal of African Economies 9 (October):244­69. Conflict, Livelihoods, and Poverty in Guinea-Bissau 9 Fosu, A. K. 2001 "Political Instability and Economic Growth in Developing Economies: Some Specification Empirics." Economics Letters 70 (February):289­94. Ghura and Mercereau. 2004. "Political Instability and Growth: The Central African Repub- lic." International Monetary Fund Working Paper No. 04/80, IMF, Washington, DC. Gudmundsson, R. 2004. "Alleviating Poverty in Post Conflict Cases: a Survey of the Liter- ature." In Jean A. P. Clément, ed., From Conflict to Reconstruction--Main Lessons and Challenges for Sub-Saharan Africa--The Case of the Democratic Republic of Congo. Forthcoming, Washington, D.C.: International Monetary Fund. Lopez, H., and Q. Wodon. 2005. "The Economic Impact of Armed Conflict in Rwanda," Journal of African Economies 14(4):586­602. Lourenço-Lindell, I. 2002. Walking the Tight Rope: Informal Livelihoods and Social Networks in a West African City. Stockholm: Stockholm University. Rodrick, D. 1999. "Where Did All the Growth Go? External Shocks, Social Conflict, and Growth Collapses." Journal of Economic Growth 4(December):385­412. Sachs, J., and A. Warner. 1997. "Sources of Slow Growth in African Economies." Journal of African Economies 6(3):335­76. Sambanis, N. 2003. "Using Case Studies to Expand the Theory of Civil War." CPR Working Paper No. 5. The World Bank, Washington, D.C. Yartey, C. 2004. "The Economics of Civil Wars in Sub-Saharan Africa." In Jean A. P. Clément, ed., From Conflict to Reconstruction--Main Lessons and Challenges for Sub-Saharan Africa--The Case of the Democratic Republic of Congo. Forthcoming, Washington, D.C.: International Monetary Fund. CHAPTER 2 Conflict, Growth, and Poverty in Guinea-Bissau Boubacar-Sid Barry and Quentin Wodon5 Conflicts and political instability have been serious constraints to growth in Guinea-Bissau. Of special concern was the civil war of 1998, which lasted 11 months and led to substantial loss of life as well as to a massive decrease in GDP per capita. Based on research on the economic cost of conflicts in Sub-Saharan Africa conducted by Lopez and Wodon (2005) and using a technique to identify outliers in time series and to correct the series for such outliers, this chapter estimates that GDP per capita today could have been more than 40 percent higher if there had been no conflict in 1998. In turn, one in three persons living in poverty today might not be poor had it not been for the conflict. C onflicts have had a devastating impact on the economies of affected countries in Sub-Saharan Africa (Yartey 2004, World Bank 2000a, Easterly and Levine 1999, Rodrick 1999). In Guinea-Bissau, conflicts and political instability have been the main constraints for growth and poverty reduction over the past three decades. Rather than being massive and pervasive, conflicts and instability in Guinea-Bissau have become cyclical events confined, to some extent, to the capital, but with economic and social con- quences for the country as a whole. While the continued instability has not often caused a significant loss of human life and physical destruction (with the notable exception of the 1998 conflict, as discussed below), the instability has been detrimental to institution building and has not produced an environment conducive to investment, growth, and poverty reduction. As many as six major military-related incidents concerning political power have taken place since independence in 1974. Thereafter, the country experienced two decades of a state-controlled system before organizing its first democratic elections in 1994. The fledgling democratic process was halted by the 1998 civil conflict that led 5. The authors are grateful to Humberto Lopez for help in estimating the impact of Guinea-Bissau's conflict on per capita GDP. 11 12 World Bank Working Paper to a drop in GDP per capita of roughly 30 percent, and to the displacement of up to 350,000 people--or about one-fifth of the population. A state of persistent political instability prevailed after the war, and it was only in 2001 that new democratic elections brought a new civilian government to power. However, the democratic process stalled again as the elected government failed to restore political stability and good governance. Following several cabinet reshuffles, alleged coup attempts, and the dissolution of the parliament during 2001­02, the military staged a bloodless coup d'état in September 2003, and a civilian transition government was charged with restoring order and rebuilding the public administration. Since mid-2004, the authorities have made sig- nificant progress in restoring fiscal discipline despite the highly polarized and still fragile political environment. The objective of this chapter is to measure the impact of the conflict on per capita GDP in Guinea-Bissau. This was done using a technique developed by Tsay (1988, 1986) and Gomez, Maravall, and Peña (1997) that identified and corrected outliers in time series. This was then applied to the assessment of the impact of conflicts on GDP per capita in Sub- Saharan African countries by Lopez and Wodon (2005).6 The estimates suggest that GDP per capita would be roughly 42 to 43 percent higher today if the conflict had not occurred. In turn, assuming that the conflict had no impact on inequality (we don't have data on that potential impact), we found that about one-third of people now in poverty can be said to be poor because of the conflict. The chapter is organized as follows: the first section briefly reviews the political context and recent economic performance of the country, including a discussion of the various impacts of the conflict; and the second section provides econometric estimates of the loss in per capita GDP attributed to the conflict and the likely impact that this economic loss has had, in turn, on poverty. A brief conclusion follows. Guinea-Bissau's Conflict and Economic Performance: A Brief Review Political Context Guinea-Bissau has experienced a long history of conflict and political instability. Following a decade-long freedom war with Portugal, Guinea-Bissau became independent in 1974. As noted in Chapter 1, the government of Luis Cabral was overthrown in a bloodless coup led by the prime minister and former armed forces commander, João Bernardo "Nino" Vieira, in 1980. There were several coup attempts against the Vieira government in 1983, 1985, and 1993. In 1994, the country's first multiparty legislative and presidential elections were held. President Vieira, of the ruling Party for the Independence of Guinea-Bissau and Cape-Verde (PAIGC), was re-elected with a slight margin over Kumba Yala of the Party of Social Restoration (PRS). An army uprising against the Vieira government in June 1998 triggered a bloody civil war that displaced over 300,000 members of the civilian population. Failed 6. Lopez and Wodon (2005) noted that this approach differs from that used by Murdoch and Sandler (2001) who used cross-country panel data to measure the average effect of conflicts on growth. The main issue with cross-country results is that indices of war tend to be discrete variables with low cross- country variability. As a result, the averaging involved in estimating the impact of conflicts may under- or overestimate impacts where conflicts are particularly severe or instead relatively mild, respectively. Conflict, Livelihoods, and Poverty in Guinea-Bissau 13 governance, breakdown of the rule of law, and limited accountability and transparency of public sector management were among the causes of this conflict, which lasted nearly a year. Following the conflict, attention was turned to the task of reconciliation and rehabil- itation. By June 1999, the situation, although still fragile, calmed down and improved steadily. A Government of National Union (GNU) was put in place and charged with the tasks of preparing a National Reconciliation and Rehabilitation Program (NRRP) and orga- nizing elections. Free and fair parliamentary elections were held on November 28, 1999, and a presidential runoff vote took place on January 16, 2000. Opposition parties scored a landslide victory in both elections. President-elect Kumba Yala of the PRS took office on February 17, 2000. Shortly after, a broad coalition cabinet, including members of various parties in the national assembly, was appointed. The task of national reconciliation and reconstruction was, however, undermined by the lack of political consensus. Following the elections, the parliament approved a new con- stitution in April 2001, which was neither vetoed nor promulgated by the president. The resulting ambiguity undermined the rule of law, and a prolonged friction developed between the executive and the legislative branches of the government over the choice of the new cabinet chief. As a result, the country operated without an effective government for several weeks, and the announcement of yet another foiled coup attempt in mid-April 2001 demonstrated once again the fragility of the political and democratic processes. The pre- vailing political instability, which affected all institutions, was not conducive to economic recovery and postconflict reconstruction. In December 2001, the government allegedly thwarted a coup attempt by army officers. Delays in holding parliamentary elections led to a military coup and the appointment of a transition government in September 2003. Between December 2001 and December 2002, President Yala carried out two cabinet reshuffles, dissolved the parliament, and called for legislative elections, which were repeatedly postponed. The ensuing discontent among the population led the army to seize power in a bloodless coup in September 2003. President Yala announced his resignation and was placed under house arrest. The mili- tary, led by the chief of defense, Gen. Verrisimo Seabra (who was later killed along with other officers during the October 6, 2004 mutiny), quickly restored power to civilian con- trol, and a civilian caretaker government was appointed with the objective of restoring the rule of law and reasserting control over public finances. In this context, the businessman Henrique Rosa was sworn in as interim president, with the support of almost all political parties as well as that of the civil society. By the end of 2003, the newly appointed gov- ernment prepared an Emergency Economic Management Plan (EEMP) and a budget for 2004 with the support of the international community. Parliamentary elections were successfully organized in early 2004, and the postelection government endorsed the EEMP reform agenda. Legislative elections were held in March 2004 and an orderly transition to a government elected on the basis of a broad reform plat- form took place in May 2004. The transition government, led by the leader of the PAIGC, Carlos Gomes Júnior was appointed with the objective of ensuring the functioning of the public administration and preparing presidential elections by mid-2005. Despite the dif- ficult economic situation, the government prepared a draft of the country's PRSP and significantly improved budgetary discipline during the second half of 2004. The govern- ment also endorsed the reform program, which was agreed to by the international com- munity; prepared the 2005 budget; and successfully organized a free, fair, and transparent 14 World Bank Working Paper presidential election that was won by the independent candidate and former president, João Bernardo "Nino" Vieira. Recurrent military interventions on the political scene, as experienced in the recent past, still remain a major obstacle to stability and democracy. So far, the military has kept their promise not to intervene in the political process during and after the presidential elections. However, the widespread poverty in the barracks and the highly polarized political envi- ronment remain ingredients for future interventions of the military. Moreover, the major- ity of soldiers in Guinea-Bissau's army are from the Balante ethnic group. There is a critical need for reforming the security sector to prevent further instability and conflict. Consequences of the 1998 Conflict The human losses and economic disruption created by the 11-month conflict that took place in 1998 were widespread. Between 2,000 and 6,000 lives were lost and up to 350,000 persons were internally displaced, including 7,000 who were generally better educated and sought refuge in Senegal, Cape Verde, the Gambia, Guinea, and Portugal, thereby depleting Guinea- Bissau's limited human capital. While solidarity within the population helped to avoid a major humanitarian crisis, deprivation increased dramatically and food stocks were depleted (on vulnerability in Guinea-Bissau, see for example Lourenço-Lindell, 2002, and World Bank 2000b and 2006). Real GDP declined by 28 percent because of a sharp decline of production in all sectors. A significant deceleration in the production of cashew nuts and cereals probably helped cause a 17 percent decline in agriculture production. Most of the formal industrial, trade, and service activities were interrupted in the last seven months of 1998. The closing of banks and nonpayment of most wages repressed the impact of the war on prices. However, the armed conflict adversely affected the country's fiscal position and its balance of payments. Marketing of cashew nuts, the main source of public revenue, was severely disrupted and merchandise exports and imports (in U.S. dollars) decreased by about 45 and 30 percent, respectively, over the period 1997­98. As a result, the external current account deficit (excluding official transfers) remained at a high of nearly 16 percent of the GDP, and the current primary budget balance moved from a surplus of about 5 percent of the GDP in 1997 to a deficit of about 7 percent of the GDP in 1998. After June 1998, external debt ser- vice arrears piled up that were equivalent to almost 10 percent of the GDP. Arrears on salaries, goods and services, pensions, and transfers also represented 5 percent of the GDP. Apart from a reduction in income flows as represented by the GDP, capital destruction was also widespread. Damages to public infrastructure (including the airport, energy and water systems, health and education facilities, roads, markets, public enterprises, and admin- istrative buildings) amounted to $25­$30 million. Some 5,000 houses were damaged in the capital city of Bissau, and losses of up to $90 million were incurred by businesses because of destruction, requisition, confiscation, and looting. As the private sector was hit, the bank- ing system was also weakened from an already fragile position. Investor confidence faded, leading to a reduction in investments (both foreign and domestic). Furthermore, the imple- mentation of necessary reforms in the areas of taxation, civil service, public enterprises, and the social sectors were interrupted as the administration could not implement the 1998­2001 Policy Framework Paper that had been agreed upon in May 1998, and the con- flict led to a temporary halt in donor funding. Conflict, Livelihoods, and Poverty in Guinea-Bissau 15 Postconflict Economic Performance The economy bounced back during 1999­2000 because of a return to peace and good agricultural performance. During the two years following the conflict, economic perfor- mance improved as a result of the enhanced security situation and gradual resettlement of displaced populations. Nearly all of the 300,000 to 350,000 persons initially displaced by the conflict in 1998 returned to their homes. Agricultural production, which repre- sents nearly 60 percent of the GDP, grew on average by 6 percent in real term per annum during the period 1999­2000. Cashew nut production, the leading agricultural subsector, averaged 70,000 metric tons over the indicated period (compared to 50,000 metric tons in 1997). Performance in agriculture had positive spillover effects on the economy as a whole (as evidenced by the close correlation between trends of real GDP growth and cashew nut production). Services grew on average by 15 percent in real terms, largely because of the close interrelationship between transportation activities and the boom in the cashew sector. Industrial activities suffered from heavy damage in infrastructure and electricity supply. Industrial growth first decelerated by 1 percent in 1999 and then picked up by 10 percent in the following year, thanks to donor support to the government's reconstruction program. Real annual GDP grew by an average of nearly 8 percent from 1999 to 2000, leading to a growth in per capita GDP of close to 10 percent over two years. At the same time, gov- ernment revenue reached 18 percent of the GDP, and the overall budget deficit (excluding grants) increased from 14 percent to 25 percent of the GDP as a result of increased post- conflict reconstruction activities. Late in 1999, the government and the International Mon- etary Fund agreed upon a macroeconomic framework for the period 2000­01 that would consolidate and sustain economic performance, reduce inflation, and limit fiscal and exter- nal gaps to financially sustainable levels. Progress continued to be encouraging on the macroeconomic front through December 2000, and the country completed its interim PRSP, reached the decision point under the HIPC Initiative, and agreed to a three-year PRGF program with the IMF for the period 2001­03. However, inadequate public policies prevented the conclusion of the IMF's first PRGF review in 2001, resulting in payment difficulties. Early in 2001, the macroeconomic pro- gram was found to be substantially off track owing to large unbudgeted expenditures mainly on defense that were financed by credit from the banking system and promissory notes. Late in 2001 a string of political crises at the northern border with Senegal in the Casamance region added to the worsening security situation. Military crackdowns ordered by the authorities led to a significant expenditure overrun, and the primary fiscal deficit reached 3.7 percent of the GDP in 2001 compared with the targeted surplus of 0.8 percent under the PRGF program. Consequently, the first review of the PRGF by the IMF was not concluded as scheduled in 2001, and the program was declared off track. After two unsuccessful short- term, staff-monitored programs, efforts to revive the program were abandoned by mid- 2002, and the arrangement expired at the end of 2003 without the completion of a review. In view of the disappointing developments, the IMF established that it would be unlikely that Guinea-Bissau would be able to meet PRGF standards in the near future. Under these conditions, donor support (equivalent to 4 percent of the GDP) did not materialize as expected, causing severe payment difficulties. External budgetary supports decreased from $12 million in 2000 to $1 million in 2001. The balance of payment and fiscal difficulties were further fueled by a record 37 percent decline in world market prices 16 World Bank Working Paper of cashew nuts in 2001. Consequently, export volume growth plummeted from 26 per- cent to -1 percent between 1999 and 2002, and government budgetary revenues as a share of GDP decreased from 19 percent to 15 percent during the same period. In this context, the external account and budget deficits (excluding external grants) reached 13 percent and 26 percent of the GDP, respectively, by the end of 2001. The economy became recessionary in 2001 and further stagnated through 2002­03. Real GDP grew at a minor rate of 0.2 percent by the end of 2001, and contracted by 7.2 per- cent by the end of 2002. The main contributing factors to the economic contraction in 2002 were the 15 percent fall of the production of cashew nuts, a lower than expected agricul- tural performance due to unfavorable weather conditions, and the continued suspension of donor-funded policy lending representing about 7 percent of the GDP. The fiscal situ- ation continued to be precarious because of the sharp decline in economic activity and the implementation of a number of policy decisions outside the emergency financial manage- ment framework that was agreed upon with the IMF and the Bank. Economic performance remained sluggish in 2003. Although the agriculture sector grew by about 5 percent, real GDP grew by only 0.6 percent because of a significant con- traction in the other sectors of the economy. Simultaneously, fiscal management further deteriorated when a substantial diversion of resources to expenditures took place outside the legally established budgetary procedures. A key issue in public finance management was the decision by President Yala to increase the salaries of the military in 2003 by more than tenfold. Consequently, the public sector wage bill increased from 8 percent of GDP in 2003 to about 11 percent in 2004.7 Thanks to an enhanced fiscal management and increased donor support to the EEMP, economic performance started to recover in 2004. During the second half of 2004, the tran- sition government formed after the September 2003 military coup took decisive measures aimed at restoring fiscal discipline. The treasury committee, established in 2003, was reinstated and tasked with implementing a strict cashflow management system. As a result, budgetary revenues increased from 15.5 percent in 2003 to 17.5 percent by the end of 2004. It is worth noting that this performance was largely attributed to improved tax collection in the informal sector and illegal fishing activities. The government, however, continued to rely on external assistance for the implementation of the EEMP in 2004. The economy started to recover in 2004 with real GDP growth reaching 2.2 percent by the end of 2004. While growth was not expected to change substantially in 2005 because of capacity con- straints and continued political uncertainties, performance was expected to improve in the medium term if adequate fiscal and capacity building reforms were carried out (for a recent review of developments in Guinea-Bissau, see Economist Intelligence Unit, 2005). Impact of the Conflict on GDP Per Capita and Poverty Impact of the Conflict on Per Capita GDP This section assesses the impact of the conflict on per capita GDP using a procedure pro- posed by Tsay (1988, 1986) for detecting outliers in time series. The basic idea is to empir- ically test and correct the data for the presence of outliers in the GDP per capita time series. 7. Most public sector wages remained unpaid in the first nine months of 2003 prior to the military coup. Conflict, Livelihoods, and Poverty in Guinea-Bissau 17 If the outliers are observed at the time of conflicts (for example, the year 1998 for Guinea- Bissau), then the corrected series for GDP per capita can be assumed to represent the counter- factual path that the economy would have followed if there had been no conflict. Details on the estimation methodology are provided in Lopez and Wodon (2005) who conducted similar work on Rwanda. Here we provide the basic estimation strategy and the key empir- ical results. The results for Guinea-Bissau are compared to a broader set of results for Sub-Saharan countries that had a recent conflict. The variable e(t) can be understood as a white noise sequence with a zero mean and constant variance 2. L is the lag operator (i.e., Le(t) = e(t-1)), and (L) is a polynomial in L with j j < and roots of (z) = 0 outside the unit circle. GDP per capita, denoted by Y(t), is represented through a univariate time series with: Y(t) = X (t) + Z(t), with (1) X (t) = µ + (L)e(t), and (2) Z(t) = w (L)o(t) (3) In the above, X(t) is a stationary and invertible series for which the polynomial (L) can be approximated by (L) = (L)/(L), with (L) and (L) finite order polynomials in L (that is, X(t) would follow an ARMA process). Z(t) is defined so that o(t) = 1 at t = s, if at time s an outlier occurs and o(t) = 0 otherwise. The structure of w(L) defines various poten- tial types of outliers, which are represented visually in Figure 2.1. If w(L) = , Z(t) defines an additive outlier (AO), which accounts for a "one shot" change in GDP in a given year. If w(L) = /(1-L), we obtain a level shift (LS), which implies a permanent impact on the series. Finally, if w(L) = /(1-L), the estimation suggests a transitory change (TC), which vanishes over time. The empirical strategy to estimate these equations is discussed in Lopez and Wodon (2005). The method can be generalized to control for additional variables that may affect the GDP per capita, which is done here to consider the potential impact of changes in terms of trade. Figure 2.1 Visualization of Hypothetical AO, TC, and LS Outliers 2.5 AO TC LS 2 1.5 1 0.5 0 1 12 23 34 45 56 67 78 89 100 111 122 133 144 155 166 177 188 199 Source: Lopez and Wodon 2005. 18 World Bank Working Paper The procedure applied to Guinea-Bissau using the data on GDP per capita in constant local currency units as well as data on the changes in the terms of trade reveals the existence of a single additive ("one shot") outlier for the year 1998, which corresponds to the civil war. The estimates provided here are from comparative data for Guinea-Bissau and various other Sub-Saharan African countries that a conflict. According to Table 2.1, GDP per capita would have been 42 or 43 percent higher in 2004 (depending on the model) than it actually was if there had been no conflict, which is a very large impact (also see Figure 2.2). As a point of comparison, in Rwanda, where the human, social, and economic costs of the genocide have been staggering, the esti- mates suggest that the per capita GDP would be between 25 and 30 percent higher today than if the genocide had not taken place (Lopez and Wodon 2005), which suggests a smaller economic impact than in Guinea-Bissau. Much of the difference between the impacts observed in Guinea-Bissau and Rwanda stems from the fact that two outliers were identified in Rwanda. First, in 1994, the year of the genocide, there was a loss of about 40 percent of the GDP. In the following year, however, there was a rebound (another additive, but this time a positive outlier) so that part of the negative impact of Table 2.1 Impact of the Conflict on Per Capita GDP GDP Per GDP per capita GDP capita Index, index, controlling per capita controlling for the conflict Year index for the conflict and terms of trade 1970* 100.0 100.0 100.0 1980* 80.9 80.9 80.9 1990 102.6 102.6 102.6 1991 104.1 104.1 104.1 1992 101.8 101.8 101.8 1993 100.9 100.9 100.9 1994 101.0 101.0 101.0 1995 102.4 102.4 102.4 1996 111.0 111.0 111.0 1997 115.1 115.1 115.1 1998 80.6 114.7 115.2 1999 84.5 120.3 120.8 2000 88.3 125.7 126.2 2001 86.0 122.4 122.9 2002 77.6 110.4 110.9 2003 75.8 108.0 108.4 2004 76.8 109.3 109.8 GDP cost (%) in 2002 -- -42.3 -42.9 Source: Authors' estimates. Note: The value of 42.4 percent is obtained as (109.3/76.8) - 1. While estimates are only provided for 1970 and 1980, the actual estimations were done using annual GDP data from before 1990. Conflict, Livelihoods, and Poverty in Guinea-Bissau 19 Figure 2.2 Impact of the 1998 Conflict on Per Capita GDP 140 Counterfactual 130 120 index 110 GDP 100 Actual capita 90 Per 80 70 60 50 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 Source: Authors' estimates using the outlier identification technique. the conflict was offset. Such a rebound was not observed in Guinea-Bissau, leading to a larger economic impact of the conflict even though the loss in human life was substan- tially less severe. It is also important to note that losses in GDP were observed in Guinea-Bissau as in Rwanda, but the conflicts did not seem to have a negative impact on the growth rate of GDP per capita. These results could be interpreted as an indication that these conflicts gen- erated a reduction in the stock values of key economic variables, such as human and phys- ical capital, without changing the returns on these stocks. Impact of the Conflict on Poverty Once we have estimated what level of GDP per capita would have been achieved without the conflict, it is straightforward to estimate what the poverty level would have been with- out the conflict, provided we are willing to make a number of (rather strong) assumptions. In the absence of good information on the distributional consequences of a conflict (that is, who gets affected most and why), these assumptions are necessary to translate our esti- mate of the impact of the conflict on GDP per capita into an estimate of how consumption per equivalent adult may have changed as measured in household surveys. To explain these assumptions, it is useful to briefly review how poverty was estimated in Guinea-Bissau within the framework of the Poverty Reduction Strategy prepared by authorities. The official estimates of poverty for Guinea-Bissau are based on a somewhat ad hoc methodology, in part because of weaknesses in the scope of the data collected in the house- hold survey used to measure poverty. The only available nationally-representative survey is the 2002 ILAP, a light household survey that was implemented following the core wel- fare indicator survey (CWIQ) methodology. The survey includes a consumption module, but this module is not very detailed, making it difficult to properly estimate poverty lines based on traditional approaches, such as the cost of basic needs framework. Instead, the authorities adopted an extreme poverty line of $1 per person per day after adjusting for 20 World Bank Working Paper purchasing power parity (Sylla 2004). This resulted in a unique extreme poverty line for the country as a whole (without adjusting for differences in the cost of living between urban and rural areas, for example) of CFAF 108,000 per person per year. A second poverty line of CFAF 216,000 per person per year was obtained, which corresponded to $2 per person per day. An estimate of the consumption of households per equivalent adult was then derived from the survey data to identify those in poverty and compute a range of poverty and extreme poverty measures. The results from our GDP per capita simulations for poverty measurement assumes that the GDP per capita growth, as measured in the national accounts, is essentially per- fectly correlated with the average growth in the consumption per adult equivalent at the household level. That is, we will use our estimates of the impact of the conflict on per capita GDP as our best guess for the impact of the conflict on the mean adult equivalent per household consumption. A second assumption is that we can rely on the poverty lines used for measuring poverty in the 2002 household survey to assess the impact of the conflict. We do not change the poverty lines for our counterfactual poverty measures without the conflict because we assume that the conflict did not affect relative prices and consumption patterns in such a way that other poverty lines would have had to be used in the absence of conflict. A third assumption is that inequality in per adult equivalent consumption was not affected by the conflict, so that we only need to incorporate the impact of the conflict on mean consumption for our poverty simulations. We only have one survey in Guinea- Bissau that does not have comparable preconflict household level data, so we simply can- not assess the impact of the conflict on inequality. It is best to assume that the inequality has remained unchanged. If we accept these assumptions, the procedure for assessing the impact of the conflict on poverty is very simple. We first compute poverty using the 2002 household survey data, and then we compute our counterfactual poverty measures after scaling up the adult equiv- alent consumption aggregate for all households in the survey by a factor equal to the ratio of the estimated per capita GDP without the conflict to the observed per capita GDP at the time of the survey. The poverty measures used here are the standard first three poverty measures of the so-called FGT class (Foster and others, 1984), namely the head count, the poverty gap, and the squared poverty gap. Denoting the poverty line by z, the number of households (pop- ulation weighted) by n, the number of poor households by q, and the adult equivalent con- sumption aggregate of household k by ck, then the three measures are defined, taking a value of zero for the head count, one for the poverty gap, and two for the squared poverty gap in the following expression: Pa =1 (4) n q k=1 z - ck z Today, as shown in Table 2.2 and following the methodology used in the PRSP for poverty measurement, we find that 65.7 percent of the population is poor; that is, with a level of consumption per equivalent adult below the $2 poverty line in purchasing power parity terms, with 21.6 percent being the extreme poor. If the conflict had not taken place, so that per capita GDP today would be 42.3 percent higher than actually observed (the results would be very similar after correcting for terms of trade, hence are not provided here), Conflict, Livelihoods, and Poverty in Guinea-Bissau 21 Table 2.2 Impact of the 1998 Conflict on Poverty Poverty Poverty Poverty reduction, observed without the Poverty reduction percentage versus in 2002 survey conflict in percentage points baseline (1) (2) (2)-(1) [(2)-(1)]/(1) Poverty Head count index 65.7 43.0 -22.7 -34.6% Poverty gap 25.7 13.2 -12.5 -48.6% Squared poverty gap 12.9 5.7 -7.2 -55.8% Extreme poverty Head count index 21.6 7.6 -14.0 -64.8% Poverty gap 5.5 1.8 -3.7 -67.3% Squared poverty gap 2.2 0.7 -1.5 -68.2% Source: Authors' estimates using 2002 ILAP. poverty would be much lower. Table 2.2 suggests that the share of the population in poverty today could be around 43.0 percent instead of 65.7 percent, while the share of the population in extreme poverty could be at 7.6 percent instead of 21.6 percent. Note that the percentage reduction in poverty following the 42.3 percent reduction in GDP per capita suggests that the elasticity of poverty reduction to growth is roughly around one (it is lower for some poverty measures, and higher for others), which is a reasonable estimate according to international experience. As expected, the corresponding elasticities for the extreme poverty measures are larger, (these elasticities are a function of the poverty lines--the higher the poverty line, the lower the elasticity). Overall, one in three persons is in poverty today in Guinea-Bissau because of the conflict, with an even higher share for extreme poverty measures. Conclusion This paper has estimated the economic cost of the 1998 Guinea-Bissau conflict on poverty. Following previous work by Lopez and Wodon (2005), a methodology for the identifica- tion and correction of outliers in time series was used to estimate the counterfactual GDP per capita that would have been observed in 2002 if the conflict had not taken place. The results suggest that the per capita GDP could have been between 42 to 43 percent higher in 2002 than it was without the conflict, assuming that the trend in GDP as available in national accounts statistics adequately represents changes in economic activity. Next, rely- ing on a number of (admittedly) strong assumptions, we estimated the counterfactual poverty measures that would have prevailed in the absence of the conflict. Our estimates suggest that one in three persons today in poverty might not have been poor if the conflict had not taken place, and the proportion is even higher for extreme poverty. While all these results could be sensitive to some of our assumption, they do provide an idea of the large impact that the conflict has had on poverty in Guinea-Bissau. 22 World Bank Working Paper References Economist Intelligence Unit (EIU). 2005. Country Report--Guinea-Bissau, April 2005. London: EIU. Easterly, W., and R. Levine. 1997. "Africa's Growth Tragedy: Policies and Ethnic Divisions." Quarterly Journal of Economics 112 (November):1203­50. Foster, J., J. Greer and E. Thorbecke. 1984. A Class of Decomposable Poverty Measures." Econometrica 52(3): 761­66. Gomez, V., A. Marvall, and D. Peña. 1997. "Missing Observation in ARIMA Models: Skipping Strategy versus Additive Outlier Approach." Working Paper no. WP9701, Bank of Spain. Lopez, H. and Q. Wodon. 2005. "The Economic Impact of Armed Conflict in Rwanda." Journal of African Economies 14(4):586­602. Lourenço-Lindell, I. 2002. Walking the Tight Rope: Informal Livelihoods and Social Networks in a West African City. Stockholm: Stockholm University. Murdoch, J. and T. Sandler. 2001. "Economic Growth, Civil Wars, and Spatial Spillovers." University of Texas. Rodrick, D. 1999. "Where Did All the Growth Go? External Shocks, Social Conflict, and Growth Collapses." Journal of Economic Growth 4:385­412. Sylla, M. B. 2004. Evaluation de la pauvrete en Guinée Bissau (2001­2002). Bissau. Tsay, R. S. 1986. "Time Series Model Specification in the Presence of Outliers." Journal of the American Statistical Association 81: 132­141. ------1988. "Outliers, Level Shifts, and Variance Changes in Time Series." Journal of Forecasting 7:1­20. World Bank. 2000a. Can Africa Claim the 21st Century? Washington, D.C.: The World Bank. ------. 2000b. Report and Recommendation of the President of the International Develop- ment Association to the Executive Directors on a Proposed Economic Rehabilitation and Recovery Credit in an Amount of SDR 18 Million to the Republic of Guinea-Bissau. World Bank, Washington, D.C. ------. 2006. "Guinea-Bissau: Integrated Poverty and Social Assessment." World Bank Report No. 34553-GW. Yartey, C. 2004. "The Economics of Civil Wars in Sub-Saharan Africa." In Jean A. P. Clément, ed., From Conflict to Reconstruction--Main Lessons and Challenges for Sub-Saharan Africa--The Case of the Democratic Republic of Congo. Forthcoming, Washington D.C.: International Monetary Fund. CHAPTER 3 Institutions, Social Networks, and Conflicts in Guinea-Bissau Results from a 2005 Survey Estanislao Gacitua-Mario, Sigrun Aasland, Hakon Nordang, and Quentin Wodon8 Guinea-Bissau endured a major conflict in 1998 and has suffered from persistent political instability since independence. After a brief review of indicators of governance in Guinea-Bissau and recent political developments, the objective of this Chapter is to provide results from a recent survey that gives insights into the opinions of the population, among others, about changes in well-being over time, trust in various institutions, sources of conflicts at the local level, and ways to deal with conflicts. There is a clear perception among citizens that there has been a decline in well-being as a result of the conflict, as well as a lack of improvement since then, at least for those in poverty who are highly vulnerable. The data suggest an increase in the lack of security after the conflict and no clear sign of improvement. The population has little trust in national institutions such as the army, the police, the judicial system, and the central government. Local conflicts often emerge because of the competition for scarce productive resources, but poorer households deal with these conflicts differently than wealthier households. G uinea-Bissau is one of the poorest countries in the world. According to the HDI index the country is ranked 172nd (among 177) in the 2004 Human Develop- ment Report (UNDP 2004). It is estimated that 65.7 percent of the population live in poverty and 21.6 percent in extreme poverty. Guinea-Bissau's economy is based on farming and fishing, which represent about 63 percent of GDP. The agricultural sector 8. The authors are grateful to Momar Sylla for providing the Guinea-Bissau 2002 survey data, Clarence Tsimpo and Edward Creppy for processing part of the data from the 2005 IPSA survey, and finally to Ilda Lourenço-Lindell for sharing research on livelihoods in Guinea-Bissau. 23 24 World Bank Working Paper generates 80 percent of the employment, and contributes about 50 percent to the PIB and 90 percent of exports (through cashew nuts, essentially). The official development assistance received represents a very large share of the gross national income. Since its independence from Portugal in 1974, Guinea-Bissau has been characterized by the increasing isolation of state bureaucracies and continuing political disarray. Gov- ernance has suffered from a gap between state and civil society, which marked the transi- tion from colonial administration to independent democratic rule (Forrest 2003). As a result, public administration has been characterized by hypertrophy, disorganization of its components, and incapacity to fulfill its functions and provide public services. After independence, the country was ruled under a one-party system. The first multiparty elections were held in 1994. Nevertheless, even after those elections, political leadership used author- itarian strategies to hold on to power and internal divisions led to growing factionalism and institutional paralysis. The persistence of political instability has arisen in part because of social fragmentation. The population is highly diverse. The country has more than 20 different ethnic groups, with no group having an absolute majority. The five largest ethnic groups are Balanta (28 percent), Fula (23 percent), Mandinga (13 percent), Manjaco (11 percent), and Pepel (7 percent). The other 15 or so ethnic groups represent about 18 percent of the population. The urban/rural distribution of the groups suggests that Balanta and Pepel are more concentrated in rural areas, whereas the Manjaco, Fula, and Mandinga tend to reside in urban areas. Most of the population is Muslim (40 percent)--mainly drawn from the Eastern groups, such as the Fula and Mandinga--or Animist (37 percent), followed by Catholic (17 percent), and other religions (6 percent). In general, different stakeholders, such as various political forces, ethnic groups, and the military have evolved into a closed network isolated from each other. The pursuit of politicized and/or personalized interests without a clear strategic focus contradicts popular interests. Eventually, an army uprising in 1998 led to the president's ouster and a civil war. After the signing of a peace agreement, presidential elections were held in 2000. In September 2003, a coup took place and a transitional national government was appointed. Parlia- mentary elections were held in March 2004, and, in October of the same year, an army mutiny resulted in the killing of the chief of staff. Despite some stabilization in 2005, the political situation is still volatile and the elections that took place on June 19, 2005, will not necessarily ensure stability and lead to an improvement of the living conditions of the poor and vulnerable. In short, Guinea-Bissau is still a country in which political instability and lack of progress in governance contribute to the deterioration of potential economic and social opportunities.9 The main objective of this chapter is to provide the results from a recent small-scale household survey that gives insights into the opinions of the population about changes in its level of well-being over time, its trust in various local and national authorities, sources of conflicts at the local level, and ways with which the population deals with these conflicts. The first section provides a general background on the current state of governance as mea- sured by various indicators and recent political developments. The detailed results from the analysis are provided in the second section. A brief conclusion follows. 9. On Guinea-Bissau's history and the transition to democracy, see among others Mendy (1994), Pélissier (1989), and Rudebeck (1999, 2004). See also more generally Berry (1993) and Mamdami (1996). Conflict, Livelihoods, and Poverty in Guinea-Bissau 25 Governance and Recent Political Developments: A Brief Review Indicators of Governance: International Comparisons Institutions in Guinea-Bissau are characterized by a persistent tension that exist between for- mal and informal institutions as well as community-based (rural) and state-based organiza- tions (Forrest 2003). The result of this fragility is that most institutions do not respond to the public interest in an efficient, effective, and fair way. This disarray, as well as poor institu- tional capacity, is further exacerbated by corruption (Government of Guinea-Bissau 2002). While in some indicators of governance and accountability, Guinea-Bissau performs close to the average for Sub-Saharan countries (see Figure 3.1), rule of law and government effectiveness are among the lowest in the region. Most state institutions have been unable to generate or command long-term loyalties. This is partly explained by their lack of capac- ity to deliver services and benefits. In addition, social and political leaders have established strong paternalistic relationships with their constituencies without reaching out or bridg- ing to other social groups, and thereby without building broader coalitions or strength- ening the legitimacy of the existing state institutions. In the absence of strong state institutions, the population has strengthened their local networks and has learned how to count on their own forces to search for solutions to their problems. Local institutions (from neighborhood associations to the local schools and the traditional authorities) there- fore perform many functions (mutual help, leisure, collective security, conflict resolution, and so on), some of which could be handled by the state. The consequences of the 1998 conflict are still reflected across the country in the dam- aged infrastructure as well as in the weak public sector, existing insecurity, political insta- bility, and crime. Again, while at the local level communities work together to resolve some immediate problems, the capacity to work together beyond the local dimension and to resolve differences in Guinea-Bissau is poor. Political risks and ethnic tensions are building Figure 3.1 Selected Governance and Accountability Indicators, 2004 Control of Sub -Saharan Africa Corruption Liberia Rule of Law Niger Regulatory Guinea Bissau Quality Government Effectiveness Political Stability Voice and Accountability 0 10 20 30 40 50 Source: Governance and Anti-Corruption Indicators database (World Bank). Available at: http://info.worldbank.org/governance/kkz2004/. 26 World Bank Working Paper Table 3.1 Selected ICGR Indicators of Cohesion in 2005 Average Government Internal Military Religious Ethnic risk stability conflicts politics tensions tensions rating Guinea-Bissau 5.5 7.0 1.5 5.0 3.0 52.5 Sub-Saharan Africa 8.9 8.5 2.5 4.1 3.2 56.2 Somalia (worst) 7.0 4.5 1.0 3.0 2.0 26.0 Botswana (best) 11.0 10.5 6.0 5.0 4.5 78.5 Source: International Country Risk Guide (ICRG). Available at http://www.icrgonline.com. up. Over the past 10 years, there has been an increasing politicization of ethnicity, which has been conveniently used by self-serving leaders seeking to increase their legitimacy and power. As shown in Table 3.1, overall risk rating for Guinea-Bissau (according to the International Country Risk Guide database) is below the average for Sub-Saharan countries. Thus, existing indicators suggest that there is a weak sense of public institutions and nation-state building in Guinea-Bissau. The accountability of formal institutions is rather limited and cohesion is poor. Social capital functions at the local level to establish networks that contribute to achieve discrete goals of various population groups, helping to provide services and support livelihoods. Within those networks, accountability is higher and reflects the paternalistic nature of the relationships between local leadership and the com- munity or groups for which these leaders act as protectors, and hinges on intermediaries with larger social institutions. However, public institutions are not effective at the national level and have failed to gain public credibility. At the same time there is an increased frac- ture or disarticulation between national level institutions and leaderships and local com- munities and their authorities. As stressed in Chapter 1,10 it is important to realize that weaknesses in governance have impacts on the strategic management of key sectors of the economy. This can be illustrated through the case of the fishing sector. The sector's governance is constrained by its weak regulatory and legal framework, low institutional capacity, and modest qualified human resources. In terms of official policies, Guinea-Bissau has not had a coherent set of sector objectives since the fisheries management plan prepared in 1996. In the mid-1990s, the USAID-funded Trade and Investment Promotion Services (TIPS) program provided rec- ommendations to the ministry of fisheries that led to the creation of annual fisheries management plans. The plans also included long-term programs aimed at moving the "off- shore" industrial fishery gradually onshore. Since then, Guinea-Bissau has not developed official fisheries management plans or a coherent policy for the sector. In practice, the fisheries sector policies of Guinea-Bissau have focused exclusively on maximizing public financial benefits from the sector (through fishing licenses and com- pensations collected from mainly foreign fleets and governments). Monitoring, control, and supervision (MCS); development of artisanal fisheries; and maintaining and improv- 10. Acemoglu (2003 and 2006), Acemoglu and Robinson (2006a, 2006b), and Acemoglu, Johnson, and Robinson (2002), Fosu (2001), and Rodrick (1999). Conflict, Livelihoods, and Poverty in Guinea-Bissau 27 ing existing infrastructure have been lesser objectives, and have been largely unattainable given the institutional and financial constraints that the sector and the country face. However, this heavy reliance of the government on fisheries sector revenues under- scores both the importance and difficulty of strengthening the governance of the sector. This reliance undermines the bargaining power and negotiating options of the government vis-à-vis foreign governments seeking access agreements, potentially reducing the overall environmental and social benefits to the country of such agreements. Another equally important aspect of this dependence relates to the country's capacity and systems to collect, handle, allocate, and use natural resource rents for poverty reduction and social cohesion purposes. Given past concerns and allegations about corruption, attention toward strength- ening mechanisms and systems for fiscal transparency and accountability is required. Recent Political Developments The recent presidential election that took place in June and July 2005 was meant to end the transitional arrangements that had been put in place since the ousting of former president Kumba Yala in September 2003. It was also meant to put an end to the postconflict transi- tion that began after the civil war of 1998­9 and bring a return to constitutional and demo- cratic rule in Guinea-Bissau following years of political instability and administrative chaos. Some 22 presidential candidates, drawn from as many political parties, contested the first round of elections held on June 19, 2005. The three front-runners were all former presidents--João Bernardo "Nino" Vieira, previously the head of the ruling African Party for the Independence of Guinea-Bissau and Cape Verde (PAIGC), who as the former mil- itary ruler and the country's first democratically elected president ruled the country from 1980 when he seized power through a coup against then president Luis Cabral, until 1999 when he was ousted because of clashes with the Army; Malam Bacai Sanhá, the PAIGC's official presidential candidate, who was the interim president following Vieira's departure; and Kumba Yala of the Partido de Renovação Social (PRS), who was the democratically elected president between 2000 and September 2003 until he was toppled through a blood- less coup. Sanhá and Vieira headed to the second round, with Yala, the third-place candidate in the first round, declaring on July 2, 2005, that he would support the candidacy of Nino Vieira. With the backing of Kumba Yala--who is from the largest ethnic group comprising more than 30 percent of the population, the Balanta, and who is claimed to have the sup- port and loyalty of key figures within the Army--Nino Vieira went on to win the second set of elections held on July 26, 2005. On July 28, 2005, the electoral commission, Com- missão Nacional de Eleições (CNE), announced the victory of Nino Vieira, who had just returned to Guinea-Bissau after six years in exile in Portugal where he had been forced to flee because of indictments for corruption and human rights abuses during his 19-year- long rule of Guinea-Bissau. Nino Vieira won narrowly, receiving 52.4 percent of the vote, a little over 19,000 votes more than Sanhá. Rivals clashed in the capital shortly following the announcement, and Sanhá, the runner-up, and the ruling PAIGC, rejected the results by claiming electoral fraud11 and referred the case first to the CNE for a recount and later 11. PAIGC rejected the results because of alleged electoral fraud, including the voting of foreigners in Gabu, Bafatá, and Quínara and the use of lists with unregistered voters. 28 World Bank Working Paper to the supreme court.12 Although some minor voting irregularities allegedly occurred, international monitors13 declared the elections as free, fair, and transparent. Nevertheless, the run-up to the elections and the period between the first and second rounds of voting was marred by fear and uncertainty. In May 2005, Kumba Yala, arguing that his own ousting from power in 2003 had been illegal, claimed he was still president and threatened to take power by other means. He backed his words up with action when he, together with a group of armed men, stormed the presidential palace for a few hours. Closer to the second round of elections, on July 17, 2005, two police officers died and 10 others were injured when armed men attempted to attack the presidential palace and the ministry of interior. This was allegedly carried out by supporters of the PRS and appears to have been motivated by the desire to avenge the death of four PRS supporters at a demonstration in June (EIU 2005). The attack reflects a general pattern of violent actions that have defined politics in Guinea-Bissau, particularly since the civil war in 1998­9. Moreover, following the dismissal by the supreme court of the challenge by the PAIGC, the political climate has been characterized by political paralysis which was only recently resolved when the current prime minister, Carlos Gomes Júnior, also of the PAIGC, finally decided to accept the results of the presidential elections. The presidential inauguration finally took place on October 1, 2005, more than two months after the presidential elections, after a month-long delay orchestrated by the ruling party and the prime minister, who refused to recognize Vieira's electoral victory. While the ceremony itself went smoothly, the inauguration was marked by fear, amidst rumors-- published by Senegalese newspaper Wal Fadjri--that a small group of soldiers had attacked a police station in Farim, in the Oio region, with the apparent intention of disrupting the inauguration. Moreover, none of the 19 heads of state invited to attend Vieira's swearing- in ceremony attended, heavily armed soldiers patrolled the streets, and many of the coun- try's leading politicians--including the presidential runner-up, Sanhá--stayed home, raising fears of a return to years of instability. President Vieira, however, aware that the legacy of his rule from 1980 to 1999 had made his re-election controversial and divisive has, so far, tried to counter these fears by continuing his electoral campaign mantra of reconciliation and national unity, promising to uphold the law, modernize the economy, defend freedom of expression, and reform the army. He has also confirmed that he is willing to cooperate with the current government and the prime minister, despite the ruling party's previous attempt to have his victory annulled. Vieira also seems to have gained the support of the military, with Armed Forces chief of staff Baptista Tagme Na Wai, and Navy chief of staff Bubo Na Tchuto, and Aniceto Na Fla, all pledging their support to the country's democratically elected civilian leaders. Nevertheless, the return to electoral democratic rule still leaves many issues unresolved which will continue to plague the political situation in Guinea-Bissau. Ultimately, many of the causes underlying the political instability and insecurity which has defined Guinea- Bissau during the last seven to eight years remain. 12. The supreme court rejected the case and issued a ruling on August 19 stating that Sanhá's appeal had taken place outside the timeframe established by law. 13. International monitors included groups from the European Union, the African Union, the United States, the Economic Community of West African States (ECOWAS), and the Comunidade dos Países de Língua Portuguesa (CEPAL). Conflict, Livelihoods, and Poverty in Guinea-Bissau 29 An immediate cause for concern relates to the current constellation of, and competi- tion for, power between the executive and the legislative. Although the immediate dispute between Vieira and the PAIGC has been overcome, tensions between the president and the prime minister will surely remain--as a political analyst recently remarked, "with Gomes seeking to limit Vieira's powers and prerogatives, and Vieira seeking to undermine the pre- mier's position within the PAIGC--the party Vieira once led."14 This situation has been further complicated during the electoral campaign by an increasingly visible split within the PAIGC, between, on the one hand, its leadership, many of whom refused to recognize Vieira's victory and insisting the official PAIGC candidate Sanhá was the rightful winner, and, on the other hand, other party members--mainly drawn from the "old guard" who are still committed to their old party leader. So far, it seems Vieira has the upper hand. With the backing of Kumba Yala--who Gomes Júnior has suggested may have been promised the role of prime minister in a new government in return for his support during the electoral campaign--Vieira has quickly moved to cement his own position and influence within the armed forces by confirming the reappointment of Armed Forces chief of staff Gen. Wai, who fought alongside Vieira during the anticolonial struggle but who also belongs to the Balanta ethnic group. More- over, it is also within Vieira's powers to form a new government immediately or to call early elections in a few months time, which could undermine the role and relative power of the ruling PAIGC and the current prime minister. Already there seems to be some evidence that the new president is moving to consol- idate power, by insisting that he, instead of the prime minister, will now preside over the meetings of the council of ministers, suggesting a more activist stance than the interim president who preceded him, Henrique Pereira Rosa, took. Nevertheless, there are also speculations that Gomes Júnior and the PAIGC took their time to accept the results from the presidential elections to gain political concessions from Vieira--in terms of promises of ministerial posts in a likely new government--and financial concessions from some of Vieira's external backers.15 Despite the president's pledges to adhere to and build national unity, this fierce com- petition over power reflects a lack of unity between the different political groups in the coun- try. Political power in Guinea-Bissau is concentrated in an executive subject to limited checks and balances from weak democratic institutions. The state also retains main control over access to the majority of resources in the country. As a result, the stakes are high for the presidency and key political positions, undermining efforts at building national unity in a country with more than 20 different ethnic groups, widespread poverty, and great dis- parities between people. The lack of mechanisms for social and political accountability and representation extending beyond the capital of Bissau to the regions where the majority of the population resides further exacerbates this problem. There is still a concern that the military, which has played a central role in the conflicts and political instability of the past, will continue to interfere (or be manipulated into inter- fering) in national politics and civilian affairs. Despite recent pledges by Wai to abide by the rule of law and the country's elected civilian leaders, the evidence of past actions, including 14. Comment by Chris Melville, analyst with Global Insight, in Menezes (2005). 15. Comment by Chris Melville, analyst with Global Insight, in Menezes (2005). 30 World Bank Working Paper its role during the recent electoral campaign suggests that the military will continue to remain a source of instability in the future. Indeed, while this has in the past been partly a result of political manipulation of and equipping an inflated, unpaid army often sharing popular discontent, conditions have not sufficiently changed to expect that the military will refrain from intervening in the future. Moreover, the tense and volatile relationship between civilian power and the military is at its roots merely a reflection of a much deeper problem of governance. Despite the recent elections, therefore, there is much to be done before Guinea-Bissau will complete its postconflict transition. Perhaps most importantly, the state continues to lack legitimacy and credibility amongst most members of the population. This is because of a variety of factors, especially its continued inability to provide basic services and infra- structure, particularly outside of the capital of Bissau, and the perceived persistence of cor- ruption and mismanagement in the public sector. This contributes to a history of weak citizen-state relations, dating back to colonial times (Forrest 2003, Lourenço-Lindell 2002). Moreover, while political freedom/rights ratings have improved over the last decade since the process of political liberalization and transition to multiparty democratic politics in the early 1990s,16 other estimates of governance have either worsened or largely remained the same (Kaufmann, Kraay, and Mastruzzi 2005), suggesting that the lack of government capacity and credibility--unless addressed by the new administration--will continue to pose a principal risk factor and source of potential instabilities in the future. Household Perceptions: Results from a Recent Survey Subjective Perceptions of Well-Being Recent data on the perceptions of the population regarding the security situation and their institutions are available from a small-scale survey conducted in 2005 for the Integrated Poverty and Social Analysis (IPSA) report on Guinea-Bissau by the World Bank (2006). The survey was conducted in both urban and rural areas, with about 400 households par- ticipating. In this section, our objective is to present the results on the perception of the population of the quality of life and the extent to which it trusts institutions within the country. Table 3.2 provides data on the level of confidence that household heads have in various types of organizations. An assets index classifies the households according to their level of wealth. The main variables used to create the index include variables related to the dwelling of the household (home ownership, number of rooms, number of bedrooms, type of toilet facility, principal source of water, principal source of energy for cooking, roof material, and wall material), various types of assets (generator, battery, gas or electrical stove, traditional stove, refrigerator, bed, mattress, mat, suitcase, television set, radio, cell phone, bicycle, motorbike, car, canoe, outboard motor for canoe, and various other types of furniture), productive assets (this included fishing tools, rice cleaners, cashew presses, tools to work the soil, and sewing machines), and various types of livestock. On the basis of this wealth 16. See the Freedom House Country Report for Guinea-Bissau available at http://www.freedom house.org. Conflict, Livelihoods, and Poverty in Guinea-Bissau 31 index, households were ranked in three terciles. The poorest households are in the first ter- cile and the richest are in the third, with each of the three terciles accounting for one third of the households. Roughly speaking, the first tercile can be seen as representing the very poor, while the bottom two terciles account for the poor. Table 3.2 also provides data on subjective perceptions of welfare in the sample. Clearly, an overwhelming majority of households declare that they are worse off today compared to before the war, with few variations between terciles. In addition, a majority also believes that they are worse off today than one year ago, but here the proportion is especially high among the poor or extreme poor. This suggests that the continuous state of instability may be more detrimental for poorer than wealthier households. The last four items in Table 3.2 are measures of subjective poverty and information on the number of meals per day, whether households feel that they are lacking food, and if they do, when this occurs. These data suggest that those classified in the bottom terciles see themselves as comparatively poorer than households in the top tercile; they have fewer meals per day on average, and they are also more likely to lack food (this affects a very large share of the poor), especially during the rainy season. Role of Social Networks17 The data in Table 3.2 suggest that there is a high degree of poverty and vulnerability in Guinea-Bissau. Yet social networks often seem to provide crucial sources of support for poor households, particularly during times of stress and coping. Indeed the resilience of the population in the face of hardship could probably not be explained if it had not been for the minimal protection offered to many households by "safety nets" that poor house- holds build (Lourenço-Lindell 2002). Nevertheless, it is important not to romanticize the notion and capacity of social networks to sustain poor households. The poorest of the poor are sometimes excluded from such networks--social assets, like other assets, may be resources they do not have access to. Other networks are based on hierarchical and nonegalitarian patron-client relations that while providing some source of security to the poorer client, the relationship may still be abusive. Hence, social networks are not necessarily a panacea to solve the welfare problems of the poor. In Guinea-Bissau there exist a range of social networks, the most common of which are those based on kinship and family, village-membership, religion, and neighborhood ties. Lourenço-Lindell (2002) noted that the networks differ in terms of their durability, the types of claims that can be made, and in terms of the power balance that they appear to exhibit (kin and market-based, instrumental and altruistic, hierarchical and egalitarian, and so on). Kinship ties are often important sources of food and money, both for con- sumption purposes as well as for market activities. Neighbors in urban areas living within the same house or compound provide a similar role. Also in urban areas, relations of assis- tance develop at the market place in a variety of ways. There were partnerships between sellers in the same trade who assist each other, sometimes pooling resources and contacts. Some traders count on credit from suppliers. Abotas, the rotating savings groups referred to earlier, is another key resource. Another kind of cooperation consisted of organized 17. This section relies in large part on Lourenço-Lindell (2002). 32 World Bank Working Paper Table 3.2 Subjective Perceptions of Well-Being, 2005 Tercile Area 1 2 3 Urban Rural Total Change in well-being versus before the war Got better 14.18 12.06 15.79 12.18 15.06 13.98 Same 9.22 9.93 7.52 7.05 10.04 8.92 Got worse 75.89 77.30 76.69 80.77 74.13 76.63 Not applicable 0.71 0.71 0.00 0.00 0.77 0.48 Total 100.00 100.00 100.00 100.00 100.00 100.00 Change in well-being versus last year Got better 28.47 27.66 40.14 33.73 31.03 32.08 Same 18.06 15.60 9.86 10.84 16.86 14.52 Got worse 53.47 56.74 50.00 55.42 52.11 53.40 Total 100.00 100.00 100.00 100.00 100.00 100.00 Household situation versus community Very poor 33.83 29.69 16.92 23.33 29.05 26.85 Poor 4.51 3.13 4.62 4.00 4.15 4.09 Medium 45.11 57.03 50.77 50.67 51.04 50.90 Rich 15.79 9.38 26.92 21.33 14.94 17.39 The most rich 0.75 0.78 0.77 0.67 0.83 0.77 Total 100.00 100.00 100.00 100.00 100.00 100.00 How many meals do you take in a day? One 43.97 29.41 26.09 25.31 38.34 33.25 Two 37.59 43.38 44.20 46.30 38.74 41.69 Three or more 18.44 27.21 29.71 28.40 22.92 25.06 Total 100.00 100.00 100.00 100.00 100.00 100.00 Lack of food Yes 93.71 95.74 73.24 79.52 92.69 87.56 No 6.29 4.26 26.76 20.48 7.31 12.44 Total 100.00 100.00 100.00 100.00 100.00 100.00 If lack of food, when? End of dry season 0.00 4.51 2.00 1.57 2.52 2.19 Rainy season 60.61 55.64 45.00 42.52 60.92 54.52 Beginning of rainy season 7.58 6.77 5.00 2.36 8.82 6.58 Beginning and end of rainy season 3.03 0.75 11.00 8.66 2.10 4.38 Dry season 7.58 8.27 8.00 8.66 7.56 7.95 Rainy season and dry season 3.79 0.75 1.00 1.57 2.10 1.92 Others/nonresponse 1.52 0.00 1.00 0.79 0.84 0.82 All the time 15.91 23.31 27.00 33.86 15.13 21.64 Total 100.00 100.00 100.00 100.00 100.00 100.00 Source: Authors' estimation from 2005 IPSA survey. Conflict, Livelihoods, and Poverty in Guinea-Bissau 33 groups among day workers (surni), where men share job opportunities and redistribute incomes. At the village level, there has been a revival of traditional social clubs (mand- juandades), during the last two decades that provide recreational and some level of eco- nomic assistance to their members. These informal, communal groups, which may include both women and men, are structured according to age. Within a tabanca, a new group is established about every three years, and is often named after the major event occurring in the tabanca that year. In Caió, the groups go under the name uran; the latest group was established there in 2004 and is called uran telemobel, since this was the year cell phones became popular in Guinea-Bissau. Finally, some households participate in local religious groupings--Catholic, Evangelical, and Islamic congregations. Some of these, however, may be more important in terms of providing personal acquaintances and contacts for cooperation in other social settings, such as the market, than in terms of the volume of material assistance they provide to needy members. Individual or household vulnerability, therefore, is partly a reflection of which com- binations or types of social networks--if any--one has access to. Indeed, households and individuals vary in their access to social networks and are therefore not equally protected against shocks and unexpected events. The more vulnerable individuals and groups are those that are more isolated, with a very small number of sources that they can rely on for assistance. Secondly, many households or individuals also do not seem to be in a position to combine different sorts of these "safety nets." Some remain reliant on one type of net- work, seemingly unable to select those networks they prefer and to break free from those they deem oppressive and disadvantageous. This vulnerability may also be exacerbated during periods of general economic downturn--the time when social assistance is often needed the most. Economic hardship seems to erode the collaborative efforts among the poor as well as the material basis of their support networks. As a result, many are forced to rely on their own resources, as they per- ceive other members of their networks as being as bad off as they are. The redistribution of groups among casual workers, for example, have over the last couple of years become less and less reliable, because of a continued influx of unemployed youth into the labor mar- ket, a reduction in employment opportunities, and declining incomes. In a situation of increasing impoverishment, the assistance provided by religious groups is also put under pressure. And, as noted earlier, participation in the abotas also becomes difficult for those unable to comply with regular contributions in times of uncertain and irregular incomes. As such, it may seem that those that rely exclusively on market-based networks--while having more flexibility and choice--may end up being more vulnerable than those that also, or even only, have access to kinship-based sources of support, which may be more tolerant of the inability to reciprocate during long periods of time. Given this, it would be worth asking the question of whether it may be likely that those who through their social status had limited access to productive resources--single, widowed, or abandoned women, (unmarried youth) and nonnative migrants--may also have lesser access to kinship-based networks. Finally, it is worth noting that there is a high level of inequality in Guinea-Bissau. This inequality, as well as to some extent diversity itself, are likely to undermine the formation or maintenance of social capital. The country comprises some 20 different ethnic groups, underscoring the fact that social capital formation may not always be durable and sustain- able, particularly in urban settings where inequality and diversity is at greater display. Yet 34 World Bank Working Paper at the same time, it is not possible to associate inequality with any one particular issue. In fact, when asked about potential sources of inequality in the country, households men- tioned that religion, age, and gender were more important factors at the source of existing inequalities than ethnicity. Geographic location played a less important role in the mindset of the population, even though it can be demonstrated that it actually plays an important role in explaining differences in well-being between households. Table 3.3 Factors Contributing to Inequality in 2005 Tercile Area 1 2 3 Urban Rural Total Ethnicity A lot 4.65 7.41 14.39 14.10 5.42 8.84 A little 25.58 31.11 34.85 35.90 27.08 30.56 No 69.77 61.48 50.76 50.00 67.50 60.61 Total 100.00 100.00 100.00 100.00 100.00 100.00 Religion A lot 13.99 12.86 34.31 31.87 13.08 20.24 A little 25.17 28.57 22.63 21.88 27.69 25.48 No 60.84 58.57 43.07 46.25 59.23 54.29 Total 100.00 100.00 100.00 100.00 100.00 100.00 Age A lot 35.11 19.26 26.52 21.15 30.58 26.88 A little 14.50 22.22 30.30 30.77 16.94 22.36 No 50.38 58.52 43.18 48.08 52.48 50.75 Total 100.00 100.00 100.00 100.00 100.00 100.00 Gender A lot 10.85 14.81 20.74 20.63 12.13 15.54 A little 34.11 25.93 36.30 32.50 31.80 32.08 No 55.04 59.26 42.96 46.88 56.07 52.38 Total 100.00 100.00 100.00 100.00 100.00 100.00 Geographic location A lot 3.88 10.53 16.79 18.83 5.02 10.43 A little 34.11 27.07 32.82 31.82 30.96 31.30 No 62.02 62.41 50.38 49.35 64.02 58.27 Total 100.00 100.00 100.00 100.00 100.00 100.00 Source: Authors' estimation from 2005 IPSA survey. Trust in Different Institutions The 2005 IPSA survey also provided data on public perceptions of different institutions and confirmed that local-level institutions are more trustworthy than central or local gov- Conflict, Livelihoods, and Poverty in Guinea-Bissau 35 ernments, the military, or political parties. While we do not know the level of trust of various institutions, the data enables us to test which institutions benefit from the highest level of trust among households. Table 3.4 shows that local schools and health posts are perceived as the most trustworthy institutions. This trust is explained by the services provided by these institutions and the level of participation that the community has in parent associa- tions and health-post teams. This trust appears to be particularly strong in rural areas, where the local school and the health post are the only institutions that are present and have some legitimacy for the community. The judiciary, the police, religious organizations, and traditional authorities have much lower levels of trust. Note that there are differences between terciles. Among the poor, trust in local institutions dealing with education and health as well as religious issues is even stronger, relatively speaking (that is, as compared Table 3.4 The Institutions Most Trusted by Citizens, 2005 Tercile Area 1 2 3 Urban Rural Total First-most trusted organization (%) Religious organizations 12.21 5.65 9.30 8.78 9.32 9.11 Political parties 1.53 0.00 1.55 1.35 0.85 1.04 Local government 0.00 1.61 5.43 4.73 0.85 2.34 NGOs 6.87 0.00 17.83 15.54 3.81 8.33 Traditional authorities 2.29 4.03 1.55 0.00 4.24 2.60 Schools 42.75 37.10 31.01 31.76 40.25 36.98 Health centers 29.01 43.55 26.36 31.08 33.90 32.81 Army 0.76 0.00 0.00 0.68 0.00 0.26 Police 2.29 1.61 0.78 0.68 2.12 1.56 Justice 0.76 4.84 6.20 4.73 3.39 3.91 Central government 1.53 1.61 0.00 0.68 1.27 1.04 All 100.00 100.00 100.00 100.00 100.00 100.00 Second-most trusted organization (%) Religious organizations 7.27 3.54 13.40 11.50 5.80 7.81 Political parties 1.82 0.88 4.12 1.77 2.42 2.19 Local government 1.82 2.65 4.12 4.42 1.93 2.81 NGOs 0.91 0.88 6.19 4.42 1.45 2.50 Traditional authorities 6.36 5.31 2.06 0.88 6.76 4.69 Schools 30.00 38.94 23.71 28.32 32.85 31.25 Health centers 45.45 35.40 36.08 36.28 40.58 39.06 Army 1.82 0.00 0.00 0.00 0.97 0.63 Police 3.64 5.31 5.15 7.08 3.38 4.69 Justice 0.00 6.19 4.12 5.31 2.42 3.44 Central government 0.91 0.88 1.03 0.00 1.45 0.94 All 100.00 100.00 100.00 100.00 100.00 100.00 Source: Authors' estimation from 2005 IPSA survey. 36 World Bank Working Paper to trust in other institutions), than among the better off. Indeed, at least some of the better off tend to have higher levels of trust than the very poor in NGOs, the justice system, and the local government. In all three terciles, very few find that the police, the army, political parties, and the central government are very trustworthy. Note that there are also some differences between urban and rural areas. For exam- ple, trust in traditional authorities is higher among the rural population, while trust in NGOs is more prevalent among urban dwellers. It is interesting to highlight these dif- ferences because they are related to the different roles these institutions have. In urban settings, trust in NGOs appears to be related to the role NGOs have in service delivery and access to goods and services (such as concession of credit to members and so on). Trust in traditional authorities, while higher in rural areas, is still low. This is partly explained by the fact that traditional authorities are perceived as controlling access to information and mediating the relationship with outsiders. At the same time, in rural as well as urban settings, there are multiple local community-based organizations that inspire trust and encourage people to participate and engage in activities that play an important role in the survival of the community and its families. Security, Sources of Conflict at the Local Level, and the Issue of Land Table 3.5 provides data on perceptions regarding changes in the security situation and sources of conflicts at the local level. Only half of the households consider that the security situation has improved over the last year and that the share of the population that believes that the situation has deteriorated is higher among the bottom tercile. Data on the sources of local conflicts are also revealing. In a country such as Guinea-Bissau, which has experi- enced a great deal of population movements across the whole subregion over centuries and recently was engaged in a civil war that led to the displacement of thousands of people, there is a risk that segments of the population end up in situations with limited access to productive resources, and hence, in a state of poverty and vulnerability. This also means that competition for productive resources will be important, and indeed, Table 3.5 sug- gests that apart from internal family issues, most local conflict situations seem to arise from issues related to land, water, and livestock, especially in rural areas. Table 3.5 also shows that very poor and poor households tend to rely mostly on traditional authorities to solve conflicts, while the better off rely more on the police. Chapter 5 will discuss issues related to the livelihoods of the population across the country's regions. Here, it is worth discussing in more detail the issue of land. In fact, although this cannot be tested here, in many cases, issues related to access to land and other productive resources involve a family component because of differences in access between genders. Specifically, access to land and labor in Guinea-Bissau, like in most other African countries, requires membership in social institutions, in particular autochthonous kinship and domestic groups.18 Critical productive resources are vested within these groups, and, in most cases, these resources are principally controlled by senior men. Marriage is the foundation of these groups and is the institution through which men are allocated land. Women gain use rights over land through the land allo- 18. Autochthonous groups refer to those descended from the original founders of the community. Conflict, Livelihoods, and Poverty in Guinea-Bissau 37 Table 3.5 Perceptions of the Security Situation and Sources of Conflict, 2005 Tercile Area 1 2 3 Urban Rural Total Change in security situation (%) Improvement 45.83 42.25 57.75 54.82 44.66 48.60 No difference 9.72 11.97 11.27 12.65 9.92 10.98 Deterioration 42.36 44.37 30.28 31.33 43.89 39.02 No data 2.08 1.41 0.70 1.20 1.53 1.40 Total 100.00 100.00 100.00 100.00 100.00 100.00 Sources of conflict (%) Land 14.60 21.8 17.42 18.95 17.27 17.91 Water 14.60 13.53 16.67 9.15 18.47 14.93 Family 20.44 12.03 32.58 31.37 15.66 21.64 Livestock 34.31 27.07 17.42 18.95 30.92 26.37 Fishing 0.00 1.50 2.27 1.96 0.80 1.24 Extraction activities 0.00 0.00 0.76 0.65 0.00 0.25 Participation to community live 1.46 1.50 7.58 8.50 0.40 3.48 Religion 0.73 1.50 0.00 0.65 0.80 0.75 Wealth 0.73 1.50 0.00 1.31 0.40 0.75 Others 13.14 19.55 5.31 8.49 15.26 12.69 Total 100.00 100.00 100.00 100.00 100.00 100.00 Whom do you turn to in case of conflict? Police 23.57 15.33 64.54 64.24 15.42 34.69 Traditional authority 56.43 52.55 17.73 20.00 56.52 42.11 Police and traditional authority 9.29 3.65 5.67 3.64 7.91 6.22 Family 5.00 16.79 4.26 3.64 11.86 8.61 Others 5.71 11.68 7.80 8.48 8.29 8.37 Total 100.00 100.00 100.00 100.00 100.00 100.00 Source: Authors' estimation from 2005 IPSA survey. cated to their husbands.19 They are also primarily able to call upon and mobilize labor through the institution of marriage. Unmarried women or women in female-headed households, unmarried male youth, and "outsiders"--that is, those that have not descended from the community founders-- are also restricted in their access to productive resources. The access to productive resources and participation in social life by nonautochthones depends on the acceptance 19. In rare cases, women can also own land in their own right. Widowed women might inherit land in trust for their sons, which is then allocated to the sons upon marriage. Women who are single or divorced and widows who have lost all their sons may be forced to return to their families and might be given a small area to farm. Some widowed or otherwise single women have to negotiate access to land every season. 38 World Bank Working Paper and goodwill of senior, male "patron" autochthones--frequently, the Chefe de Tabanca-- and normally can only occur by marrying into a local autochthonous family. The "client" outsider would be expected to return the favor in a number of ways, although these rela- tionships are never equal and work in favor of the more powerful patron who has greater bargaining power. Field work carried for this study picked up on the problem of access to land by nonnatives (nonautochthones)--including immigrants from the Casamance region of Senegal to the north of the country and migrants internal to Guinea-Bissau, who as a result were either left without the ability to farm (in some cases) or gained limited and conditional access to land. Those that gained access to land often had to abide by a set of conditions, such as being limited to cultivating rice, corn, or cassava, and not being allowed to use the land for fruit and cashew plantations. Access rights were also only temporary, meaning that nonnative households had to renegotiate the access to and use of land on an annual basis. Land received was also often of lesser quality, which in some parts of the country (for example, the less fertile east) can mean chronic food shortages. Women's access to land and other productive resources was also seen to be affected by this (see Box 3.1). As women only gain access to land through their husbands, which cre- ates a certain level of dependency of women on men, women who had been abandoned, divorced, or widowed would in some cases find themselves without access to land. While women were, in most cases, still provided with land, though mainly on a temporary basis, the more serious problem in these cases was their access to and ability to mobilize labor to help work and cultivate the land. The extent of this problem seemed to vary according to ethnicity and religion, with women in the east seeming to be left more vulnerable if single, widowed, or abandoned than in other areas. Anecdotal evidence suggesting an increase in the abandonment of women because of their reported enhanced income-earning status provides a cause for concern. The issue of access to land is of immediate importance given the current national and legislative debate on reforming land access policies. A new land law was adopted in 1998, but the regulations required to implement the law are still under discussion. While the new law makes a grand leap forward in recognizing traditional authorities' role in managing and distributing communal land, the discussion here underscores the importance of also addressing the specific needs of women, youth, and migrants in accessing land. Moreover, the regulations also need to take seriously the many latent yet intensifying conflicts over access to land, particularly those between livestock breeders and agriculturalists as well as between traditional communities and private investors (including ponteiros). Conflicts over land in urban areas are also becoming more serious, most of which have been between urban settlers and the Municipal Council of Bissau (CMB). The CMB has taken on the responsibility of managing land in Bissau and claims ownership of much of the land that is currently being inhabited by poor dwellers, particularly in the outskirts of the city of Bis- sau. The local government has also encroached extensively on customary land in Bissau and nearby Biombo, literally expropriating the Pepel (native owners of these lands) of large shares of their land. An area in Alto Bandim, for example, which was originally used by local farmers for cashew production, but which was expropriated by the municipal government without any compensation and later sold to private persons or turned into luxury res- idences, is the recently built Ministers' Quarters on Alto Bandim (Lourenço-Lindell 2002). Perhaps most importantly, efforts need to be made to ensure the implementation capacity Conflict, Livelihoods, and Poverty in Guinea-Bissau 39 Box 3.1 Land Regulation in Guinea-Bissau--A Review Under colonial administration, two types of property rights had been conceptualized under Decree 43893 of 1961: the land right of the state and that of the communities called reserved areas. The reserved areas had covered areas actually under cultivation and residential areas, yet excluded the areas under fallow and extraction activities (fishing and forestry). The Land Law 4/75 adopted after the independence in 1975 transferred land property to the state. The new Land Law (Law No. 5/98) adopted in 1998 articulated three objectives: (1) guarantee land rights to local communities for economic use; (2) acknowledge customary land regime by rep- resentative institutions; and (3) stimulate land investment by creating market value. Contrary to the previous legislation, the new law has authorized and introduced permanent or temporary land concessions both in urban and rural areas. In addition, the law has cre- ated a tax mechanism to increase land-use efficiency aimed at dismantling large-scale farms (latifundios) unable to demonstrate income generating capacities. It has also permitted the formation of an administration commission of land whose objective is to safeguard the law enforcement and consorted interventions in land use. The land in Guinea-Bissau has been declared a property of the state constituting a common property of all the people (Article 20, Section 1). Yet, all citizens have a right to private use of land, and the state can grant the right of private use to individuals and collective units and national and foreign entity for economic and social development objectives (Article 40, Section 1). Such land rights for private use are recognized either by: (1) customary use (uso consuetudinário), as envisaged land use by tradi- tional local communities, and (2) concession (Article 40, Section 3). The present law therefore recognizes land-use rights based on customary use practiced by traditional local communities. This is an important instrument to recognize and guarantee traditional rights in a modern legislation context. The private land use right under customary use is a permanent title and can be exercised in rural and urban land including reserved areas for local communities (Article 16). The land under the customary use covers cultivated and inhabited zones as well as unexplored zones and resources attributed to local community residents through their representatives. The administration and distribution of the land under customary use by the local community res- idents shall be exercised according to the custom and practices of each community, and any omitted part shall be governed by this law (Article 17, Section 1, 2). The law clearly stipulates in Article 17, Section 3, that the state recognizes the residents' right for community admin- istration and exploration of the land, forest, and other natural resources according to local custom and practices. The private land-use right under customary use is guaranteed by the state independent of written contract or register, obliging the land commission to promote and constantly update the registration services (Article 17, Section 4). This right shall be transferred by successive inheritance and transferable to other resident(s) within the local community free of charge, according to the local practices independent of endorsement by contract and register (Article 18). Transfer of the private land-use right under customary use to the third party, other than those referred to in Article 18, shall be made in agreement with the state and the local communi- ties and follow the process to be applied in the concession. In such cases, the population of the local communities retains the right to freely and directly negotiate the terms of transfer in accordance with the conditions specified in Article 23 (Article 19). The private land-use right under customary use can be converted into the private land-use right under concession (Article 21). The private land-use right under concession is granted by administrative contract and has a duration of 90 years to be automatically renewed if not revoked (Article 22). This right is transferable by contract or inheritance upon request for official authorization. The process of grant, transfer of, and conversion into the private land-use right under concession involves tax obligations and the right is subject to annual land tax payment (Article 38). The land tax revenues shall be distributed in the following manner; (1) 60 percent to the public treasury, (2) 20 percent to local communities, (3) 10 percent to regional and sectoral admin- istrative authorities, and (4) 10 percent to the Land Commissions. Source: World Bank (2006). 40 World Bank Working Paper of the new law and its regulations, without which the land law will have little impact on the ground. Note finally that while this is not apparent in Table 3.4, the importance of accessing labor is also an acute concern. The availability of labor in rural areas seems to be decreasing. Field work documented a so-called "rural exodus," primarily of youth escaping rural areas in search of further education and better life opportunities in urban towns and cities. In addition to the thousands of uneducated youth it adds to the informal and unemployed labor market in Bissau, this exodus leaves rural areas without the productive capacity required to farm the lands and ensure sustainable rural livelihoods. Again, however, households with large proportions of elderly and female members are at a particular disadvantage. Conclusion This chapter has provided results from a recent small-scale household survey on changes in citizens' well-being over time, their trust in various local and national authorities, sources of conflicts at the local level, and the ways with which citizens deal with these conflicts. There is a clear perception among the population that its well-being has deteriorated as a result of the 1998 conflict as well as an increase in violence and lack of security even after the conflict, with no clear sign of improvement. The population has little trust in national institutions, such as the army, the police, the judicial system, and the central government. The results also suggest that local conflicts often emerge because of the competition for scarce productive resources, but poorer households deal with these conflicts in a different way than wealthier households. References Acemoglu, Daron. 2003. "The Form of Property Rights: Oligarchic Versus Democratic Societies." NBER Working Paper No. 10037. ------. 2006. "Modeling Inefficient Institutions." Forthcoming in R. Blundell, W. Newey, and T. Persson, eds., Advances in Economic Theory and Econometrics: Proceedings of 2005 World Congress of the Econometric Society. Acemoglu, D., and J. A. Robinson. 2006a. Economic Origins of Dictatorship and Democracy. Cambridge University Press. ------. 2006b. "Economic Backwardness in Political Perspective." American Political Science Review 100: 115­31. Acemoglu, D., S. Johnson, and J. A. Robinson. 2002. "Reversal of Fortune: Geography and Institutions in the Making of the Modern World Income Distribution." Quarterly Journal of Economics 117 (November):1231­94. Fosu, A. K. 2001. "Political Instability and Economic Growth in Developing Economies: Some Specification Empirics." Economics Letters 70 (February): 289­94. Economist Intelligence Unit (EIU). 2005. Country Report­Guinea-Bissau, April 2005. London: EIU. Forrest, J. B. 2003. Lineages of State Fragility: Rural Civil Society in Guinea-Bissau. Oxford: James Currey. Conflict, Livelihoods, and Poverty in Guinea-Bissau 41 Government of Guinea-Bissau. 2002. "Programme National de Bonne Gouvernance." Bissau. Kaufmann, D., A. Kraay, and M. Mastruzzi. 2005. "Government Matters IV: Governance Indicators for 1996­2004." Policy Research Working Paper series no. WPS 3630, World Bank, Washington, D.C. Lourenço-Lindell, I. 2002. Walking the Tight Rope: Informal Livelihoods and Social Networks in a West African City. Stockholm: Stockholm University. Mamdani, M. 1996. Citizen and Subject: Contemporary Africa and the Legacy of Late Colonialism. Princeton: Princeton University Press. Mendy, P. K. 1994. Colonialismo portugues em África: a tradiçãao de resistencia na Guiné- Bissau (1879­1959). Bissau: Instituto Nacional de Estudos e Pesquisa. Menezes, G. 2005. "Inauguration of Guinea-Bissau's President to Go Ahead." Broadcast report on Voice of America, September 29, 2005. Pélissier, R. 1989. História da Guiné: Portugueses e Africanos na Senegâmbia (1841­1936). Lisbon: Editorial Estampa. Rodrick, D. 1999. "Where Did All the Growth Go? External Shocks, Social Conflict, and Growth Collapses." Journal of Economic Growth 4:385­412. Rudebeck, L. 1997. "To Seek Happiness: Development in a West African Village in the Era of Democratization." Review of African Political Economy 71:75­86. Rudebeck, L. 2001. "On Democracy's Sustainability: Transition in Guinea-Bissau." SIDA Studies, No. 4, Gothenburg: SIDA. United Nations Development Programme. 2004. Human Development Report 2004: Cul- tural Liberty in Today's Diverse World. New York: UNDP. World Bank. 2006. "Guinea-Bissau: Integrated Poverty and Social Assessment." World Bank Report No. 34553-GW. CHAPTER 4 Poverty and Its Determinants in Guinea-Bissau Edward G. E. Creppy and Quentin Wodon20 Given the poor economic performance of Guinea-Bissau over the last few years, including a severe recession toward the end of 2002, poverty is likely to be high and to have risen in recent years even compared to its high postconflict level. The first objective of this chapter is to estimate the share of the population in poverty in 2002, predict how it may have evolved since then, and assess the levels of growth that will be required to reduce poverty measures in the future. The chap- ter also provides a brief poverty profile and an analysis of the determinants of poverty using the 2002 nationally representative survey, which was recently made available for analysis. Geographic location, demographic structure (both household size and headship), employment (both in terms of sector and type), education, and migration all have potentially large effects on the consumption level and thereby poverty of households. P olitical instability, a lack of investment, and the impact of the 1998 conflict on GDP per capita have led to low levels of per capita GDP and high rates of poverty in Guinea-Bissau (as discussed in Chapter 2 of this volume; see also World Bank, 2006). As will be documented in this chapter, close to two-thirds of the population live with levels of consumption per equivalent adult below a purchasing power parity adjusted at $2 per day. One in three persons in poverty today can be said to be poor because of the 1998 conflict, as argued in the second chapter of this book. Yet, there is renewed hope that the economic and political situation will improve. As the country implements its Poverty Reduction Strategy and as it benefits again from important support from donors, it will be important to search for opportunities for growth and broader policies that will lead to poverty reduction. The objective of this chapter is twofold. In the first section, we provide estimates of poverty in 2002 and how the share of the population in poverty may have evolved since then. 20. The authors are grateful to Momar Sylla for providing the Guinea-Bissau 2002 survey data. 43 44 World Bank Working Paper Simulations for the potential impact of future growth rates are also provided, showing that high growth rates will be needed for many years, if only to return to the poverty levels that would have been observed without the conflict. Then, evidence is provided on the main income sources of the population by quintile of per capita consumption and by geographic area, showing that the poor depend in large part on primary sector activities, which include cashew nuts and fishing. Because these are also sectors with potential for growth, it is argued that there is hope for implementing growth policies for the poor in the country through the promotion of these sectors. In the second section, we provide a brief poverty profile and an analysis of the determi- nants of poverty using the 2002 nationally representative survey which was recently made available for analysis and three different measures of well-being on the basis of consumption, income, and wealth, respectively. Geographic location, demographic structure (both house- hold size and headship), employment (both in terms of sector and type), education, and migration all have potentially large effects on the consumption level and thereby on the poverty of households (Lourenço-Lindell, 2002). Poverty Estimates and Income Sources Estimates of poverty in Guinea-Bissau are based on the 2002 ILAP survey conducted by the National Statistical Institute (the survey was actually implemented between March 2001 and April 2002). The survey's sample size is 3,216 households, of which 672 are located in Bissau and 2,544 in the rest of the country. The estimates of poverty that are provided follow the methodology used for the PRSP. Both the data used for the analysis and the methodological choices are somewhat nonstandard, so that the poverty estimates must be considered as indicative only. In the absence of better data, it is difficult to improve on the estimations in a significant way. Yet at the same time the broad conclusions that emerge from the analysis in this paper seem plausible enough to warrant attention. Consider first the household consumption aggregate. This aggregate is based on a lim- ited expenditure module with data obtained according to the recall method. Some of the weaknesses of the data are apparent in Table 4.1. Across quintiles, about 40 percent of household expenditures are allocated to the purchase of food, with another 12 percent allocated to autoconsumption of food items. What is surprising is that these food consumption shares do not decrease with the overall consumption level (per capita) by household. Food consumption as a share of total consumption, in fact, increases when total consumption increases, which is rarely observed in other countries. Also, autoconsumption for energy-related expenditures appears very high, between 20 percent and 28 percent of total expenditures in the first four quintiles. It could be that the unreliable supply of electricity requires that households look for alternative sources of energy, which may indeed be costly. However, the magnitude of the estimates is very surprising, and may again suggest weak- nesses in the underlying data set. The methodology for estimating poverty is also somewhat nonstandard, again because of weaknesses in the data (a limited questionnaire on food consumption, among others). Instead of computing a poverty line based on the cost of basic needs method as is typically done, we follow the PRSP as well as previous work by Sylla (2004) in assessing the value of the inter- national $1 and $2 poverty line in local currency units. It turns out that the extreme poverty Conflict, Livelihoods, and Poverty in Guinea-Bissau 45 line corresponding to the $1 threshold is CFAF 108,000 per person per year. The poverty line corresponding to the $2 threshold is therefore CFAF 216,000. Again, following the method used for the preparation of the PRSP, the indicator of well-being on which the poverty measures are based is the consumption level per equivalent adult. Household members above 15 years of age are considered as adults, while younger members count for one half of an adult. Household-based poverty measures suggest a head count of poverty of 58.9 percent and a head count of extreme poverty of 16.8 percent. When population weights are taken into account (that is, when we count individuals instead of households, thereby taking into account household size), the incidence of poverty is estimated at 65.7 percent, while that of extreme poverty is 21.6 percent (see Table 4.2). In what follows, we will rely mostly on pop- ulation-based estimates. There are noticeable differences in the poverty measures by region. First, as expected, the share of the population in poverty is higher in rural areas (70.3 percent) than in urban areas (52.6 percent). Next, according to the survey, some 79.1 percent of the population lives in rural areas (versus 20.9 percent in urban areas). This would mean that out of the 65.7 percent of the population in poverty, 52.1 percent live in rural areas. Said differently, 79.3 (= 52.1/65.7) percent of the poor would be living in rural areas. For the extreme poor, the proportion would be even larger, since out of 21.6 percent of the popula- tion in extreme poverty, 19.0 percent live in rural areas (that is, 88 percent of the extreme poor are living in rural areas according to the survey's weights). Table 4.1 Expenditure Shares by Quintile, 2002 (in percentage of total consumption) Type/Quintiles Poorest 2 3 4 Wealthiest Food 36.24 42.08 41.64 42.55 43.00 Food (Autoconsumption) 13.20 11.62 12.91 12.95 11.83 Energy 4.19 4.09 3.99 3.91 3.42 Energy (Autoconsumption) 28.08 23.91 22.86 20.78 11.27 Education 1.30 1.27 1.26 1.35 1.49 Sanitation 4.43 3.78 3.57 3.65 4.05 Sanitation (Autoconsumption) 0.14 0.19 0.17 0.30 0.25 Health 2.37 2.81 2.75 2.55 3.30 Clothing 5.39 5.83 5.40 5.52 5.97 Accommodation 1.36 1.69 1.92 2.68 7.69 Ceremonies 2.82 2.30 3.07 3.36 5.91 Transfers 0.33 0.25 0.33 0.30 1.55 Other 0.15 0.16 0.13 0.10 0.27 Total 100.00 100.00 100.00 100.00 100.00 Source: Authors' estimates using 2002 ILAP. There are, however, two reasons that suggest that the above estimates of the share of the extreme poor living in rural areas are overestimated. First, one could argue that part of the reason why we observe such a high concentration of poverty in rural areas is because of the fact that we use a single poverty line for poverty measurement. To the extent that the cost of living is higher in urban areas than in rural areas, we may underestimate poverty in 46 World Bank Working Paper Table 4.2 Population-based Poverty Measures by Region and Areas, 2002 Poverty (US$1) Extreme Poverty (US$2) P0 P1 P2 P0 P1 P2 Bissau 52.6 17.5 7.6 9.8 2.0 0.6 Bafata 73.2 30.5 16.0 27.2 7.5 3.1 Gabu 67.0 25.7 12.8 20.4 5.6 2.5 Cacheu 64.7 27.9 15.2 28.4 7.8 3.1 Oio 80.4 34.8 18.6 35.3 9.0 3.6 Biombo/Bolama 63.9 22.4 10.0 14.5 2.8 0.8 Quinara/Tombali 70.3 27.4 13.9 23.1 6.5 2.5 Rural areas 70.3 28.5 14.7 25.7 6.8 2.7 Urban areas 52.6 17.5 7.6 9.8 2.0 0.6 Total 65.7 25.7 12.9 21.6 5.5 2.2 Source: Authors' estimates using 2002 ILAP. urban areas and overestimate it in rural areas. Second, some estimates indicate that only 66 percent of Guinea-Bissau's population may live in rural areas,21 instead of 79 percent as suggested by the survey. These differences may be because of definitional issues and popula- tion dynamics and are difficult to describe given the lack of actualized data and the conflict that the country experienced in 1998. Yet overall, even if we were using different poverty lines and rural population shares, we would probably still find a much higher number of the poor in rural as compared to urban areas, as observed in other West African countries where bet- ter data have made such estimations easier. Beyond the urban/rural comparison, Table 4.2 suggests the existence of large differences in poverty measures according to Guinea-Bissau's main regions. The coastal towns of Bissau and to a lower extent Biombo/Bolama and Cacheu have the lowest levels of poverty, whereas the interior areas of Bafata and especially Oio have much higher levels of poverty. Yet in com- parison to some other Sub-Saharan African countries, and especially given the fact that we are using a common poverty line for the country as a whole, the differences in poverty between regions are in fact somewhat limited. That is, poverty seems to be widespread every- where in the country. It is also worth noting that the annual consumption per capita recorded in the survey is CFAF 164,061 and the consumption per equivalent adult is CFAF 211,927. Since poverty is estimated using expenditures per equivalent adult, the value of CFAF 211,927 is the reference figure, and it is below the level of the poverty line, at CFAF 216,000. In countries such as Guinea-Bissau where the average consumption level is below the poverty line, it is clear that, broadly speaking, growth policies are key for poverty reduction, and more important than redistribution policies. The poverty measures in Guinea-Bissau are high, in part, because of the choice of the poverty line (which is to some extent arbitrary), but mostly because of structural weaknesses in the economy and the impact of the 1998 conflict. As argued in the first chapter of the book, 21. The last population census was conducted in 1991. Current estimates are based on census data and electoral cadastre. The figures are from the U.S. Census Bureau in 2004. Conflict, Livelihoods, and Poverty in Guinea-Bissau 47 up to one in three households in poverty today can be said to be poor because of the large negative impact of the conflict on GDP per capita. But even after the conflict, the situation does not seem to have improved markedly. When asked in 2001­02 whether they were bet- ter off today as compared to one year ago, some 60 percent of households responded in the negative, and the same holds for their perceptions regarding changes in well-being at the community level. There are few differences between households according to their position in the distribution of consumption in these perceptions. This means that the better off have the same negative perceptions as the poor as to changes in well-being between, say, the year 2000 and the period corresponding to the implementation of the survey. At the time of the survey, close to half of the population declared not being able to meet their food needs, with a higher such proportion in the bottom quintiles of the distribution, as would be expected (see Figure 4.1). Figure 4.1 Problem Satisfying the Food Needs of Households by Quintile, 2002 Wealthiest 4th Often 3rd Sometimes 2nd Never/Rarely Poorest National 0 20 40 60 80 100 Source: Authors' estimates using 2002 ILAP. How may the poverty measures have changed since 2001­02? To answer this question, we must rely on a number of assumptions in the absence of better data to carry the simula- tions. First, we will assume that growth in consumption per adult as measured in household surveys follows GDP per capita growth closely as measured in the national accounts. Second, we will assume that growth does not change the relative prices of goods, nor that it has any effect on inequality (at the international level, there is basically no correlation between growth and changes in inequality). Under such assumptions, we may use the household level data from the 2002 ILAP survey to simulate the impact of changes in GDP between 2002 and 2005 on poverty, simply by scaling up the observed consumption levels of households in 2001­02 by the observed level of growth in real terms of per capita GDP between 2002 and 2005, and computing again the poverty measures using the same poverty line. Real GDP contracted by 7.2 percent by the end of 2002, due in part to bad weather that led to a 15 percent fall in the production of cashew nuts and weak agricultural performance, and also probably in part to the continued suspension of donor-funded policy lending repre- senting about 7 percent of GDP. In 2003, while agriculture grew by about 5 percent, real GDP grew by only 0.6 percent because of a contraction in other sectors and poor fiscal management (including a decision to increase the salaries of the military by more than ten-fold). Growth resumed somewhat at 2.2 percent in 2004 thanks to enhanced fiscal management and 48 World Bank Working Paper increased donor support. But even that is not enough to reduce poverty in any meaningful way, as population growth rates, which probably had remained almost constant at 2.4 percent from 1970 to 2000, are now estimated for the period 2000­04 at 2.0 percent per year (U.S. Census Bureau 2005). Growth estimates for 2005 are not available, but it is unlikely that sub- stantial progress have been made. In fact, a deterioration (or at least no improvement) in liv- ing standards is suggested in the results from the small scale 2005 IPSA survey. In this 2005 IPSA survey, 32 percent of the sampled households estimated that they were better off at the time of the survey than one year prior (which would correspond to 2004), versus 54 percent who said that they were worse off, and 15 percent who declared no change in well-being. Overall, given a rate of population growth of at least 2 percent per year, there must have been a substantial decrease in GDP per capita from the early months of 2002 (the time at which the household survey was completed) until 2005. According to data from the World Development Indicator (WDI) database (World Bank 2004) and our own calculations, the decrease in per capita GDP may be of the order of 10.9 percent since 2002. Then, under the assumption mentioned above regarding the absence of relative price shifts, changes in inequality, and the link between GDP per capita and consumption per equivalent adult, the share of the population in poverty may have increased from 65.7 percent in early 2002 to 72.3 percent by the end of 2005. What would it take to reduce poverty in the future? The same method as that used above for the projections of poverty to 2005 can be used to provide estimates of future poverty as a function of growth in per capita GDP. Table 4.3 provides the number of years that would be required with various annual GDP growth rates to achieve cumulative per capita growth rates of respectively 25 percent, 50 percent, 75 percent, and 100 percent. For simplicity, a single population growth rate was used of 2.5 percent per year, which is below the current level, but takes into account the fact that population growth will decline over time (clearly, in the long run, for some of the large number of years required in the table, the rate of population growth would decrease much more, but the exercise is provided just for giving an idea of what could be achieved in the next 10 years or so, by the 2015 deadline for reaching the Millennium Development Goals in 2015). The results suggest that if the economy were to grow at 4 per- cent per year, it would take 11 years to achieve a cumulative growth rate of 25 percent, which would in turn reduce the share of the population in poverty from the estimated level of 72.3 percent in 2005 to 58.8 percent by 2020. To give another example, assuming an annual growth rate of 8 percent per year of the economy, the cumulative growth rate would reach close to 75 percent by about 2015 (as shown in the Table 4.3, the number of years required is 9.6). This would in turn lead to a share of the population in poverty of about 36.4 percent. While these simulations are nothing but a number crunching exercise based on very strong assumptions, they do provide an idea of the magnitude of the challenges ahead. The message from Table 4.3 is somewhat disheartening, as even with much higher growth rates than were observed in the recent past, Guinea-Bissau is not likely to reduce its level of poverty well below the level that would prevail today if the conflict had not taken place (this level is estimated at 43 percent for the headcount index in Chapter 2). On the other hand, if growth were achieved in sectors in which the poor are engaged, progress toward poverty reduction could be more rapid. Table 4.4, which is also derived from the 2002 ILAP survey, provides data on the main income sources of households by quintiles of the population ranked according to consumption per capita. As will be discussed in the following section, the data on income sources are far from perfect. Yet the data clearly show, Conflict, Livelihoods, and Poverty in Guinea-Bissau 49 Table 4.3 Number of Years Required to Achieve Cumulative Growth and Poverty Targets Number of Years Required to Reach Target Share of the Population 4% 5% 6% 7% 8% in Poverty (%) Growth Growth Growth Growth Growth Baseline estimate in 2005 72.3 -- -- -- -- -- Per capita GDP growth 25% 58.8 11.3 7.5 5.7 4.6 3.8 Per capita GDP growth 50% 47.5 20.5 13.7 10.3 8.3 7.0 Per capita GDP growth 75% 36.4 28.3 18.9 14.3 11.5 9.6 Doubling of per capita GDP 27.9 35.0 23.4 17.7 14.2 11.9 Source: Authors' estimates using 2002 ILAP. as expected, that agricultural products and to some extent income from fishing (in the sec- ond quintile) are very important for the poor. These are precisely areas where poverty could be decreased if adequate policies were to be implemented. In Table 4.4, a similar analysis is provided to assess differences in income sources between regions. It is useful to note that all regions are equal in terms of population size. Accord- ing to the United Nations (2001), Bissau, which includes the capital city, concentrates about 25 percent of the population, followed by Oio (14.6 percent), Bafata (13.8 percent), Cacheu (13.8 percent), Quinara (4 percent), and Boloma-Bijagos (2.3 percent). Table 4.4 suggests that the share of total revenue obtained from private and public sector salaries as well as from nonagricultural products is significantly higher in Bissau than in any other region. In Table 4.5, it is evident that the regions of Bafata, Quinara/Tombali, and Oio derive the highest shares of income from agricultural products, including cashew nut production.22 As for fish products, they are most important in the areas of Cacheu and Ohio, again two of the poorest regions in the country (Oio is the poorest region and Cacheu the fourth poorest). Poverty Profile and Determinants of Poverty It is standard practice to provide a poverty profile in a paper on poverty. Such a profile is a set of statistics giving the probability of being poor according to various characteristics, such as the level of education of the household head or the area in which a household lives (Coudouel and others 2002). Given that a poverty profile is already available in Sylla (2004), we provide here only a few results. In terms of demographic variables, household size is cor- related with poverty, with larger households (many of whom live in rural areas) being poorer. The probability of being poor also increases with the age of the household head. In the sur- vey, there are 2,713 male-headed and 503 female-headed households. The level of con- sumption per equivalent adult is higher for female-headed households than male-headed 22. Even though all these regions are also rice producers, rice farming in Bafata is more modernized than in the other regions. Whereas the other regions use mainly traditional and labor intensive methods, Bafata is a beneficiary of a Chinese pilot program, which uses a combination of animals and tractors for rice cultivation. Hence the 52 percent share of household income from agricultural products in Bafata is mainly from the sale of rice. 50 World Bank Working Paper Table 4.4 Revenue Shares by Quintile of Per Capita Consumption, 2002 Revenue Shares by Sources, Quintiles Poorest 2 3 4 Wealthiest Local drinks 3.67 3.72 2.65 2.90 1.57 Livestock 2.88 2.04 4.12 2.25 1.34 Animal products 0.71 0.19 0.11 0.09 0.17 Fish products 2.57 6.84 3.38 3.31 2.73 Firewood 2.18 0.58 0.95 0.42 0.50 Salary--agric. sector 2.86 1.74 1.29 1.26 0.47 Salary--private sector 8.06 17.14 21.19 19.42 29.28 Salary--public sector 8.00 12.82 19.97 13.92 13.06 Income from equipment rentals 13.07 7.99 9.33 9.60 7.62 Pensions 0.66 1.10 0.75 0.44 1.93 Transfers 5.86 7.67 6.74 13.84 12.49 Nonagriculture 7.31 6.09 5.49 7.89 16.12 Agricultural products 42.18 32.07 24.04 24.67 12.72 Total 100.00 100.00 100.00 100.00 100.00 Source: Authors' estimation from 2002 ILAP survey. households. Similarly, the probability of being poor at the household level is 61 percent for male-headed households versus 51 percent for female-headed households. One reason for this finding is that women have been active in the informal sector as small traders, a sector that may have been less affected by adjustment policies and the conflict than the male- dominated public sector. Also, an increasingly larger number of women in rural areas are Table 4.5 Revenue Shares by Type of Activity and by Region, 2002 Biombo and Quinara and Sources Bissau Bafata Gabu Cacheu Oio Bolama Tombali Local drinks 1.34 1.32 0.29 6.32 2.23 7.08 3.75 Livestock 0.08 7.04 9.93 0.91 8.68 1.62 1.81 Animal products 0.00 0.45 1.15 0.47 0.26 0.27 0.10 Fish products 0.61 0.61 0.18 8.86 6.32 17.59 5.72 Firewood 0.58 0.14 0.28 0.46 3.05 0.52 0.39 Salary--agric. sector 0.66 0.35 1.83 2.18 1.54 1.44 2.29 Salary--private sector 33.15 9.58 21.59 5.19 9.15 4.64 15.48 Salary--public sector 20.92 2.78 4.52 4.78 5.05 5.84 6.27 Rentals 6.40 6.56 3.63 26.73 4.42 13.03 7.70 Pensions 2.00 0.20 0.45 0.02 0.56 0.48 0.39 Transfers 13.03 13.81 9.69 6.85 6.94 4.93 6.80 Nonagriculture 14.77 4.67 15.89 4.12 4.27 7.76 4.01 Agricultural products 6.47 52.49 30.57 33.10 47.53 34.81 45.30 Total 100.00 100.00 100.00 100.00 100.00 100.00 100.00 Source: Authors' estimation from 2002 ILAP survey. Conflict, Livelihoods, and Poverty in Guinea-Bissau 51 involved in the processing of cashew nuts. According to the ministry of agriculture, the plant- ing and processing of cashew provides work for 82 percent of the rural workforce, 49 percent of which are women. Table 4.6 provides statistics on poverty and the distribution of the poor according to quintiles for a few employment and education variables. Consider first the distribution of the poor as compared to population shares (first six columns in table 4.6). The first column in Table 4.6 provides the population shares in the various categories. For example, 53.96 per- cent of the population belongs to households whose head works in the agriculture sector. Yet as expected the proportion is much higher, at 72.81 percent, for the share of population in the first quintile of consumption per equivalent adult that belongs to households with a head engaged in agriculture (in the wealthiest quintile, the proportion of the population in agricultural households is lower than the national average, at 42.05 percent). The table also gives data on the percentage of the population with a head in various cat- egories (last column in Table 4.6). The share of individuals in poverty is highest, at 69.6, per- cent among individuals who belong to a household whose head is in agriculture. Perhaps surprisingly, the poverty rate among households with a head employed in public adminis- tration is high at 56.6 percent, but this is related to the fact that a large numbers of household heads in public administration are actually working in low-skill jobs, for example as drivers, technicians, cleaning personnel, etc. Individuals in households with a head working in the private formal sector have a lower incidence of poverty, whereas those who are self-employed (private, informal) have a high incidence of poverty. Finally, there is also a correlation between the level of educational attainment of the household head and the probability for the household to be poor. The proportion of individuals living in households where the head has no education at all is highest among the bottom quintiles of the distribution of con- sumption, as expected. Regarding the headcount index of poverty, it is at 32 percent in house- holds where the head has a tertiary education, a fairly high value but a rate still well below that of 71 percent among households where the head has no education at all. Poverty profiles are informative, but they also have limits. The main drawback is that they cannot be used to assess the determinants of poverty. For example, the fact that some household group is poor (say, agricultural workers) may be due in large part to other char- acteristics of this group (say, the educational level of the group's members). In order to pro- vide more insights into the determinants or correlates of poverty, we provide in Table 4.7 the results from a regression analysis of the determinants of well-being with three dependent variables: the logarithm of the per capita consumption of households (we could have used consumption per equivalent adults and this would probably not have made a large differ- ence, except for the demographic variables); the logarithm of the income per capita of house- holds; and the logarithm of an assets index that aims to capture the wealth of households. Separate regressions are provided for the full sample and for rural and urban areas separately. The main independent variables that are used to explain the levels of the dependent variables include: (1) family structure variables and their square (number of infants, children, and adults), (2) the characteristics of the household head (the gender of the head, the age group the head belongs to, the marital status of the head, the migration status of the head, the head's level of education, his/her employment status, his/her sector of activity, and whether he/she works in the public or private sector); (3) the region where the household is located; and (4) household access to basic services such as potable water, schools, public transportation, food markets, and health facilities. 52 World Bank Working Paper Table 4.6 Repartition of Poverty by Group and Headcount Index for Selected Characteristics, 2002 Repartition of the Poor by Category (Column Sum is 100%) Headcount Poorest 2nd 3rd 4th Wealthiest Index of All Quintile Quintile Quintile Quintile Quintile Poverty Employment sector Agriculture 53.96 72.81 57.36 54.19 52.71 42.05 69.60 Industry 5.66 1.52 7.63 6.19 4.52 7.29 59.30 Construction 6.45 6.77 6.32 7.27 5.38 6.66 65.80 Transport 1.67 1.01 0.63 1.75 1.85 2.49 52.90 Commerce 10.55 5.98 9.33 9.37 11.11 14.31 53.60 Services 7.13 2.94 7.71 6.76 8.31 8.36 61.80 Education 0.73 1.06 0.44 0.85 0.43 0.90 65.10 Health 0.61 0.36 0.42 0.06 0.93 1.01 49.20 Public 13.25 7.54 10.16 13.55 14.76 16.92 56.60 administration Total (%) 100.00 100.00 100.00 100.00 100.00 100.00 -- Employment type Public 14.90 9.99 12.15 13.87 16.16 19.14 57.90 Private formal 9.20 3.16 7.60 10.50 8.75 13.03 51.70 Private informal 75.90 86.85 80.25 75.63 75.09 67.83 67.30 Total (%) 100.00 100.00 100.00 100.00 100.00 100.00 -- Educational No level 58.24 69.17 63.27 60.72 55.94 48.44 71.00 Primary 24.42 23.45 25.81 24.37 24.79 23.76 65.00 Secondary 14.75 6.27 10.17 13.86 16.43 22.07 49.00 Superior 2.60 1.10 0.75 1.05 2.84 5.73 32.00 Total (%) 100.00 100.00 100.00 100.00 100.00 100.00 -- Source: Authors' estimates using 2002 ILAP. Yet before presenting the results of the regressions, a word of caution on the data is war- ranted. As shown in Figure 4.2, there are large differences between the average consumption and income per capita recorded in the survey. Income per capita is clearly underestimated, in part because we do not have good data on production for autoconsumption. Hence the estimates of the impact of household characteristics on consumption per capita are proba- bly better than those obtained for income per capita. Also, we find relatively few differences in assets between households, suggesting that those regressions probably also have limita- tions. Still, despite the weaknesses in the raw data, it remains useful to estimate regressions for all three measures of well-being to see if the conclusions regarding the determinants of well-being are robust across measures. We now turn to the results in Table 4.7. Consider first the impact of demographic char- acteristics: an increase in the number of infants, children, or adults in a household is likely to cause a decrease in per capita consumption. To some extent, this is an automatic result since Conflict, Livelihoods, and Poverty in Guinea-Bissau 53 Figure 4.2 Mean Values of Well-Being Indicators by Consumption Quintiles, 2002 (in CFAF) 400000 350000 300000 Income per capita Consumption per capita Wealth index 250000 200000 150000 100000 50000 0 1 2 3 4 5 Source: Authors' estimates using 2002 ILAP. the indicator of well-being depends directly on household size. Note that the negative impact of having more adults is not statistically significant for per capita consumption in urban areas, perhaps because of better employment opportunities there so that adult members could work to meet their consumption needs. Note also that the impacts on income per capita and on assets are much weaker. In fact, a unit increase in the number of adults is expected to have a positive effect on asset accumulation (this variable is not normalized by household size). In female-headed households in rural areas, consumption per capita and the accumula- tion of assets is likely to be higher than in households headed by men, all other things being equal. For example, female-headed households in rural areas have average consumption lev- els that are 20 percent higher than households headed by male counterparts. A similar find- ing is observed for the accumulation of assets, albeit to a smaller degree. This result is a clear indication that female-headed households are not necessarily worse off than households headed by males, as was already alluded to in previous sections. As for the age of the house- hold head, in most cases the results are not statistically significant. Finally, in rural areas, con- sumption levels are higher for heads who are single or married (monogamous or polygamous) as compared to heads who are divorced or widowed. In urban areas however, only house- holds headed by those who are single are likely to benefit from higher consumption levels, as compared to the other groups. The relationships between these variables and income per capita as well as assets are weaker. Another finding from Table 4.7 is that education is key to increase consumption, income, and asset accumulation. In rural areas, consumption increases by 12 percent for heads with secondary education and by 35 percent for heads with tertiary education, as compared to heads with no education at all. In urban areas the effect is even stronger at 54 World Bank Working Paper Table 4.7 Impact of Household Characteristics on Welfare, 2002 Consumption Per Capita Income Per Capita Wealth (assets index) All Urban Rural All Urban Rural All Urban Rural Family Structure Number of infants 0.12 0.18 0.08 0.13 N.S. 0.15 -0.024 -0.06 N.S Number of infants sq. 0.01 0.02 NS NS NS 0.02 NS NS NS Number of children 0.13 0.20 0.10 0.19 -0.19 0.19 NS NS NS Number of children sq. 0.01 0.02 0.01 0.02 NS 0.02 NS NS NS Number of adults 0.12 NS 0.16 NS NS NS 0.024 0.05 NS Number of adults sq. 0.01 NS 0.01 NS NS NS NS NS NS Household Head Head below 35 NS NS 0.08 NS NS NS 0.04 NS 0.05 Head between 35 and 55 NS NS NS NS NS NS 0.04 NS 0.05 Female head 0.13 NS 0.19 NS NS NS .08 NS 0.07 Marital Status of Head Single 0.26 0.35 0.19 NS NS NS NS NS NS Married (monogamous) 0.01 NS 0.15 NS NS NS NS NS 0.07 Married (polygamous) 0.13 NS 0.16 NS NS NS NS NS NS Temporary migration Migrated (at least once) 0.10 0.16 0.10 NS NS NS 0.05 NS 0.05 Education of the Head Primary education NS NS NS NS NS NS 0.06 0.08 0.04 Secondary education 0.17 0.24 0.12 NS NS NS 0.19 0.21 0.16 Tertiary education 0.51 0.53 0.35 0.55 0.70 NS 0.48 0.48 0.43 Employment (Head) Construction/Industry NS NS NS NS NS NS 0.09 NS 0.09 Transport/Commerce NS NS NS 0.60 0.46 0.47 0.21 0.15 0.24 Education/Health/Services NS NS -0.17 0.54 NS 0.48 0.17 NS 0.17 Administration NS NS NS 0.32 NS NS 0.10 NS NS Employment (Head) Private sector NS NS NS NS 0.42 NS NS NS NS Business owner NS NS NS NS NS NS NS NS NS Access to Services Potable water (<15) -0.16 NS NS NS NS NS NS NS 0.10 Food market (<15) -0.11 NS NS 0.51 -0.50 0.50 -0.05 NS NS Public transport (<15) 0.12 NS 0.12 NS NS NS 0.06 0.39 0.06 Primary school (<15) NS NS NS 0.25 0.72 NS NS 0.22 0.05 Secondary school (<15) NS NS NS NS NS NS 0.09 NS 0.15 Health facility (<15) 0.16 0.25 0.12 NS NS 0.24 NS NS NS (continued) Conflict, Livelihoods, and Poverty in Guinea-Bissau 55 Table 4.7 Impact of Household Characteristics on Welfare, 2002 (Continued) Consumption Per Capita Income Per Capita Wealth (assets index) All Urban Rural All Urban Rural All Urban Rural Regional Dummies Bissau 0.35 0.79 0.43 Bafata 0.24 NS 0.13 Gabu 0.27 -0.34 0.13 Cacheu 0.15 0.29 0.05 Biobo/Bolama 0.32 0.40 NS Quinara/Tombali 0.18 NS 0.06 Source: Authors' estimates. Note: NS = not significant. Displayed coefficients are significant with a 10% level of confidence. Underlined coefficients are significant with a 5% level of confidence. Bold coefficients are significant with a 1% level of confidence. Omitted categories: Age Group of Household Head (aged 55 or over); Gender of Head (Male); Marital Status of Head (Divorced/Separated/Widowed); Migration Status of Head (Never Migrated); Education of Head (Never Attended); Sector of Employment of Head (Agriculture); Type of Employment (Government Employee); Access to Basic Services (30 minutes or more); Regional Dummies (Oio). 24 percent for secondary school educated heads and 53 percent for heads with tertiary edu- cation. The impact of education on the accumulation of assets is also high in both urban and rural areas for assets. In urban areas for example, asset holdings increase by 8 percent, 21 percent, and 48 percent respectively for heads with primary, secondary, and tertiary education, whereas, in rural areas, the gains are 4 percent, 16 percent, and 43 percent, respectively. The impact of education on income appears to be limited to urban areas where incomes for households with heads with tertiary education are 70 percent higher than incomes for heads with no education at all. In contrast, the impact of the sector and type of employment is much weaker, and even surprisingly so. There are only a few statistically significant coefficients in the regressions while in many other countries more differences are observed between differ- ent types of occupations. The most notable exception to the lack of statistical significance of most coefficients is the fact that households in rural areas with heads employed in transport/ commerce and education/health fields benefit from a 47 percent gain in income per capita as compared to households with heads working in the agricultural sector. In urban areas, the only sector that seems to have a large positive impact on income is transport/commerce. The results also indicate that urban household heads who work in the private sector, are likely to increase income per capita by 42 percent compared to their counterparts who work in government, but the impacts are not found to be statistically significant for consump- tion and assets. Another set of variables in the regression, which clearly matters for standards of living, relates to geographic location to health and education facilities. In general, physical capital endowments of households are important determinants of poverty outcomes. Low educa- tion has a pervasive effect on poverty because in addition to its impact on income, it has sig- nificant spillover effects on other socioeconomic factors, such as the health status of children, 56 World Bank Working Paper reproductive behavior, infant and child mortality, and employment.23 In the case of Guinea- Bissau, living nearby a school or a health facility, and means of public transportation tend to be positively associated to consumption, income, and assets. The role of these endowments increased as the economic recession of the 1980s became more severe. In contrast, those households who live near food markets seem to be poorer, because these tend to take place in poorer neighborhoods. What is more important is the fact that, controlling for all the characteristics discussed above, households who live in Bis- sau, Bafata, Gabu, Cacheu, Biombo/Bolama, and Quinara/Tombali are better of than house- holds living in the reference category of Oio, the poorest region in the country. For example, living in Bissau is associated with a 79 percent increase in income compared to residing in Oio. More generally, the gains observed for Bissau explain much of the migration patterns observed in the country, and the fact that Bissau itself has relatively high levels of poverty probably in part because of migration. Finally, temporary migration for more than a month in the last 12 months also leads to gains in consumption, income, and assets. This is important given that during periods of har- vests, some people, particularly younger women and youth, seasonally migrate towards the rural areas of the country to offer their labor to farmers, particularly cashew producers (an activity referred to locally as pirmi). During this period, they set temporary residence in the country's interior, and live and work with relatives or friends. Others simply work alone, offering their services to small cashew producers. Remuneration is either in cash or in-kind, such as in cashews, and is frequently traded for rice for consumption. Many of the women also extract cashew juice, which is later transported to the cities and sold as wine, providing a major source of income during the cashew season (April to July). Temporary migration therefore serves an important livelihood purpose in rural areas. In addition, those with kin or family in urban areas with sufficient means often send their children to live with their rel- atives in the towns and cities to enable them to continue with their studies. Some ethnic groups also engage in seasonal migration from one rural area to another, either in pursuit of wage-labor, for trading and commerce purposes, to return to their "native lands" to engage in rice production, or to lead their livestock through transhumance toward more fertile lands. Others have emigrated from the Casamance region of Senegal and have settled indefinitely in the northern areas of Guinea-Bissau. Others again, particularly from the northern and eastern areas of Guinea-Bissau, have permanently emigrated into neighbor- ing countries or to Europe (primarily Portugal and France) to join the ever-increasing num- bers of the Guineaspora--Guineans who have left Guinea-Bissau but who remain intimately connected with their homeland in a variety of ways, including through remittances to fam- ily and friends. Conclusion This chapter has provided poverty measures for Guinea-Bissau as a whole and for the main regions, as well as elements for a poverty profile. Overall, Guinea-Bissau presents determi- nants of poverty similar to other Sub-Saharan countries: the predominance of poverty in 23. This issue is well documented in Creppy (2003). Conflict, Livelihoods, and Poverty in Guinea-Bissau 57 rural areas; and the sensitivity of poverty to key household characteristics such as size, level of education, health conditions, and migration status. Some 65 percent of the population of the country may have been living in poverty in 2002, and this percentage may have risen to 72.3 percent since then. High rates of growth sustained for many years will be needed in the future to simply return to the levels of poverty that could have been observed if the 1998 con- flict and subsequent political instability had not taken place. Yet if growth were promoted in the sectors where the poor are active, including in cashew production and fishing, faster gains could be achieved. This chapter has also provided a basic poverty profile and an analysis of the determinants of well-being such as consumption, income, and assets of households. It was found that apart from geographic location, a wide range of variables affect the level of well-being of house- holds. Higher household size is correlated with lower levels of consumption and income per capita. Households headed by women enjoy a slightly higher level of well-being. Education has an important and large positive impact on standards of living, but the impact of employ- ment is less clear. Even if households involved in agriculture are poorer than other house- holds, the differences in poverty levels between categories of employment are lower than expected. Temporary migration also seems to improve standards of living. References Coudouel, A., J. Hentschel, and Q. Wodon. 2002. "Poverty Measurement and Analysis." In J. Klugman, ed., A Sourcebook for Poverty Reduction Strategies. Washington, D.C.: World Bank. Creppy, E. G. 2003. "The Growth Effects of the Socio-Economic Status of Women: A Panel Data Approach." American University. Economist Intelligence Unit (EIU). 2005. Country Report--Guinea-Bissau, April 2005. London: EIU. Lourenço-Lindell, I. 2002. Walking the Tight Rope: Informal Livelihoods and Social Networks in a West African City. Stockholm: Stockholm University. Sylla, M. B. 2004. Evaluation de la pauvreté en Guinée Bissau (2001­2002). Bissau. United Nations. 2001. Estudo Qualitativo Sobre o Perfil da Pobreza na Guiné-Bissau: Persepções e Soluções Locais. World Bank. 2004. World Development Indicators. Washington, D.C. ------. 2006. "Guinea-Bissau: Integrated Poverty and Social Assessment." World Bank Report no. 34553-GW. CHAPTER 5 Livelihoods in Guinea-Bissau Estanislao Gacitua-Mario, Hakon Nordang, and Quentin Wodon This chapter provides an analysis of the modes of livelihood of Guinea-Bissau's population. The country is poor and highly dependent on agriculture and extractive activities such as fishing. Yet beyond such general statements, it is useful to describe in qualitative terms the main activities that take place both in rural areas at the regional level, and in urban areas through mostly informal employment. This is done on the basis of fieldwork carried out in the country in early 2005. In addition, the chapter provides a brief quantitative empirical analysis of the main income sources of households, and a more detailed analysis of two key sectors of the economy that could form the basis of a poverty reduction strategy: cashew nuts and fishing. T here is no doubt that most of the population in Guinea-Bissau is poor. Standards of living have been historically low, and the conflict that took place in 1998 had a very great negative impact on GDP per capita and thereby on poverty, as discussed in Chapter 2. As documented in the fourth chapter, 65.7 percent of the population was poor in 2002 and 21.6 percent were the extreme poor (Sylla 2004). These estimates of poverty are based on the 2002 ILAP survey conducted by the National Statistical Institute between March 2001 and April 2002, and a poverty line and extreme poverty line corresponding to $2 per person per day in purchasing parity terms. It is feasible to dramatically reduce poverty in Guinea-Bissau. The country has a largely untapped potential for agriculture. It is endowed with fertile soils, abundant water, and a favorable climate. It is the sixth-largest producer of cashew nuts (6 percent of the world cashew production) after India, Vietnam, Brazil, the Ivory Coast, and Tanzania. Its weather is also favorable to the production of other cereals and a large variety of fruits. Agricultural production rests mainly on cashew nuts, rice and subsistence farming. In addition to some 90,000 traditional farm households in tabancas, there are 2,000 ponteiros that have access to important land concessions. The cashew nuts sector represents a growth opportunity that could directly benefit the poor, but a further expansion in the sector should be planned in such a way as not to increase food security risks because of a corresponding reduction in rice production. 59 60 World Bank Working Paper The country also has large and valuable fish resources. Yet despite the vast fishery potential in Guinea-Bissau, only 10 percent of the fish caught annually are caught by domestic artisanal fishers. Indeed, fishing in Guinea-Bissau has traditionally been a mar- ginal activity to complement agricultural activities. Fishing is, on average, practiced some 4­6 days a month, over 8­10 months, for a total of less than 50 days a year. According to the FAO Fishery Country Profile (2001; see also FAO and World Bank. 2002), in 1996 fish- ing was the main activity for only 18 percent of those engaged in fishing, while agriculture was the main activity for 78 percent. Still, the fishery sector provides about half of govern- ment's budgetary revenues though fishing license payments. Because many fishermen are poor, improving the performance of this sector should lead to poverty reduction. Despite high rates of poverty throughout the country, it is also important to realize that all households are not equally poor. How well any given household is doing depends essentially on its sources of income and modes of livelihood. Building on previous work by Lourenço-Lindell (2002) and diagnostic reports on informal livelihoods published by INEP (INEP 1996, 2001; Cardoso and Imbali 1993), the main objective of this chapter is to analyze the sources of livelihoods in the country, and document how these sources dif- fer between regions and to some extent between ethnic groups. For example, as mentioned above, the country is dependent on agriculture and extractive activities (mainly fishing). Yet beyond such general statements, in order to inform poverty reduction strategies, it is necessary to describe in more detail the main activities that take place in rural areas at the regional level. Similarly, it is useful to describe the strategies, mostly informal, that are used by the poor to make a living in urban areas. Finally, an important contribution of the chapter is to document some of the salient features of the cashew nuts and fishing sectors so as to identify opportunities for growth, and to assess whether such growth would likely be "pro-poor"--that is, whether the poor would benefit more proportionately from this growth than other groups of the population. The chapter is structured as follows. In the first section, we provide qualitative insights into the strategies used by households for their livelihood in rural areas. This is done by location, with distinctions being made between the northern regions, the east, the south, and the island areas. Next, in the second section, we provide a similar analysis of livelihoods for urban areas, with a focus on Bissau where an overwhelming majority of the urban pop- ulation lives. In the third section, after presenting simple statistics derived from a recent small-scale household survey in order to identify the key sources of income of households, we turn to a more detailed discussion of two key sectors for the poor: cashew nuts and fish- ing. A conclusion follows. Livelihoods in Rural Areas In Guinea-Bissau, rural livelihoods are characterized in large part by their dependence on the cashew and rice sectors, as well as subsistence agriculture. Rice remains the most widely grown staple crop in the country, although in the east and amongst the Fula and Mandinga, corn is the preferred staple crop. Crop diversification is the norm--dictated by the necessity to reduce risks in order to ensure food security. Cashews (including production and sale of cashew wine) have played a particularly important role for many households, providing an important source of income--as well as consumption if traded for rice--during parts Conflict, Livelihoods, and Poverty in Guinea-Bissau 61 of the dry season (April to July). Other sources of consumption and income come from the cultivation of peanuts, cassava, beans, and so forth. Horticulture (tomatoes, onions, and so forth), fruit production (mangos and bananas in particular), and exploration of forest products (wood, coal, straw, extraction of palm oil and wine, and so forth) are also pursued, partly as a complement to rice and corn for consumption purposes, but increasingly as a source of income as well. Some households breed chickens and hens, goats, and other small animals. While fishing has traditionally been considered a marginal activity undertaken during times of poor harvests and during the "hunger season," it is an important coping strategy for those communities based along the coast and rivers. In the sector of Caió, in northern Guinea-Bissau, the main ethnic group is the Manjacos, who mainly live off agriculture, particularly cashew production. In addition to the Manjacos, the sector is also populated by migrating Fulas, Mandingas, Pepel, and Mauritanians. Among the male Fulas and the Mauritanians, the principal activity is small trade and com- merce, while the Pepel that mainly occupy the coastal areas primarily engage in seasonal fishing during the dry season, many return to the region of Biombo during the rainy season to engage in agricultural activities. In terms of the gendered division of labor, men tend to be more involved in the clearing and planting of agricultural lands, while women--in addition to tending to the domestic chores--are more involved in the transportation and sale of crops. Women from the non-Islamic ethnic groups, much like in urban areas, are also more actively engaged in the sale and resale of foodstuffs. In the east, where the Fula and Mandinga ethnic groups make up the majority of the population, the men tend to work on corn production in the highlands, while women tend to cultivate rice in the lowlands (bolanhas). While cotton was for many years the main cash-crop in the east, supported by large investments in the area by the parastatal SAGB (the Cotton Society of Guinea-Bissau), in recent years cotton plantations have been abandoned here also in favor of cashew cultivation. Households in the east also rely on livestock breeding (mainly cattle), though this is mainly used to ensure livelihood security rather than for consumption. Livestock serve as a particularly important source of collateral, which is used to invest in agricultural production, and in some cases also for the emigration of household members. For those households that engage in the latter, remittances also become an important source of income. The East, moreover, has for centuries been the center of the main trading route in Guinea-Bissau. The lumos there, especially in the city of Gabú, attract mobile trades from all of the countries in the subregion (Guinea-Bissau, Senegal, Gambia, and Guinea-Conakry, amongst others). The lumos, however, are most active during the dry season. The east, and to some extent parts of the north, suffer from drought during the dry season, with negative consequences for food security in these regions. According to respon- dents in the focus groups of the areas surveyed, the droughts have become severe in recent years with diminishing and faltering rains. In addition to the impacts on food security, this has also led to increasing tensions and, at times, local conflicts over access to water and naturally irrigated lands, particularly between cattle breeders and agriculturalists. In search for watering holes for their cattle, many livestock breeders engage in transhumance across the eastern regions toward the north during the worst parts of the drought, and in the process often trespass on the lands of agriculturalists. Households in the fertile south, despite the lack of access to services and infrastructure described earlier, are well endowed with natural resources, including fish, and therefore, in most cases, manage to avoid the food security dilemmas of the east. Indeed, the south 62 World Bank Working Paper is where a majority of the locally produced rice and food in Guinea-Bissau is cultivated. Lowland (bolanha) rice cultivation in paddy fields is the primary activity for most house- holds, followed by the production of fruits--pineapples, bananas, and oranges--as well as explorations of forest products (extraction of palm oil and wine) and horticulture. Fishing, while rarely a primary activity for households, is a strategy often used to supplement con- sumption and incomes, which becomes particularly important in times of poor rice harvests and during the dry season. Women typically engage in the sale of fish and production of wine from honey. Among the Balante, women typically also engage in the production and sale of congutu, a snack made from rice and peanuts. Given the poor road and transport conditions in the south, however, households and communities have a hard time marketing and selling their produce. Indeed, as a result of the relative isolation of many communities, the local economy is mainly based on barter- ing and exchange, with rice serving as a common currency. Unable to access markets, the communities are rarely compensated well for their produce. The constraints in accessing markets, and transforming and conserving their products, each year lead to the rotting and wastage of tons of products, particularly fruits and horticultural produce. Due to the geo- graphical isolation, especially of these communities, many of the households aim for agri- cultural self-sufficiency. Due to the lack of electricity in most of Tabancas, transformation and conservation of the produce is scarce, making it difficult for most households to engage in much consumption-smoothing behavior. In recent years, many of the households in the south have also been affected by the slow advance of saltwater into the rice fields in the lowlands (bolanhas). Lacking access to the credit needed to repair the dams in the paddy fields, many households have been forced to relocate to middle-lying areas. Other households cultivating in the highlands, have also-- because of the increasing deforestation in the area and recent forest-protection initiatives-- willingly or unwillingly been displaced toward the mid-lying areas. Some complain that they are currently unable to sustain their livelihoods in these areas given the different farming techniques and crops required to develop the areas. In the islands, unlike on the mainland, fishing serves as a primary livelihood strategy. Nevertheless, households still pursue a mix of livelihood strategies, combining fishing with agricultural activities and the exploration of forest products. Fishing, extraction of palm oil and wine, and clearing and planting agricultural fields, are generally the main activities pursued by men. Women sell the fish, and produce horticultural products for consumption and cash. Young males also take part in fishing and extraction, as well as the sculpting of wood statues for the tourism industry, but in peri-urban areas may also occasionally provide casual labor for tourists and hotels in the area. For the most part, the tourism industry and commerce on the islands are dominated by foreigners, the former by the French, and the latter by mobile traders from neighboring countries. The latter also work in the artisanal fishing industry. Note that while Guinea-Bissau is endowed with an abundance of fish and seafood, it currently lacks a regulatory framework and mechanisms monitor to supervise this valuable resource. As a result, there is a danger that the country--and particularly the communities in the archipelagos who rely on fishing--will suffer from overuse and fishing by foreign industrial fleets and artisanal fishers from the neighboring countries. The focus groups conducted in the villages surveyed on the islands reported that people's livelihood strategies have already been negatively affected by this. Conflict, Livelihoods, and Poverty in Guinea-Bissau 63 Overall, perhaps the main constraints for the development of people's livelihood strategies were--in addition to the low level of access to basic services--the difficulty in marketing goods and the lack of transformation and conservation opportunities in the case of fishing. As a result, people have few incentives to work beyond what they need to feed themselves. The government has recently made some investments in developing the arti- sanal fishing industry on the islands (including by establishing a vocational school in the town of Bubaque), but so far these investments have had little impact on the livelihoods of the islanders. Livelihoods in Urban Areas Over the last 15 years there has been a strong growth of the urban population (5.1 percent per year). This rapid urbanization has been fueled in the recent past by a lack of employment opportunities in rural areas and the displacement caused by the war. Yet it also represents a long-term trend. From 1975 (at independence) to 2002, the urban population grew from 16 percent to 44 percent (FAO-CP 2004). Rural-urban migration has sped up over the last decade, concentrating the population in four main areas: Bissau-Biombo, Cacheu, Bafata- Bamdadinca, and Buba. The rapid urban growth has dramatically increased urban density and pressure over an almost inexistent urban infrastructure and public services (FAO-CP 2004 and UNDP 2001). Today most urban households are slum dwellers (World Bank 2001).24 Data from the small household survey implemented in 2005 for this study suggests that 70 percent of households had one or more members who migrated for less than a year, while 30 percent have someone that migrated on a permanent basis. Most short-term migrations (45 percent) are to Bissau, followed by other rural communities (28 percent) and other cities (13 percent). Permanent migration targeted other countries (35 percent), followed by Bissau (29 percent), other communities (27 percent), and other cities (9 percent).25 Even though migration is likely to lead to higher incomes for migrants (as suggested in Chapter 4), both temporary and permanent migration has led to a large increase in the number of urban dwellers, many of whom are poor, live in slums, and organize their livelihoods around informal activities since the private sector--concentrated in agriculture, trade, and construction sectors--is rather small. The lack of a vibrant private sector is itself due in large part to political uncertainty, inadequate infrastructure, a small industrial base, lack of appropriate financial services, and limited local technical skills. All of these factors are contributing to high rates of poverty. Thus, even though poverty on an aggregate level remains lower in urban than in rural areas, increasingly in Guinea-Bissau poverty and vulnerability are also becoming urban issues, with many urban households for parts of the year living on only one meal each day-- a phenomenon referred to in Guinea-Bissau as um tiro--the "one shot." Moreover, there are indications that many urban dwellers may be more vulnerable to particular shocks than 24. SDStats, the Social Development Database, drawing on data from UN Habit (2003). 25. Prior studies by INEP suggest that there are some differences in the migratory behavior among ethnic groups. For example, Mandigas and Fulas would be more likely to migrate to foreign countries (Senegal, Gambia, and France), while Manjacos and Mandingas would tend to move more to the regional capital or Bissau. 64 World Bank Working Paper rural populations, because they cannot survive in subsistence agriculture. The country has experienced an extensive process of informalization, particularly when it comes to the livelihood strategies available to the urban poor. According to results from the 2005 IPSA survey, more than 90 percent of the urban families in urban areas live essentially from activities in the informal sector (fazer bida, "making a living"), or of a mix of informal and formal (mainly public sector) activities. Despite higher incomes than in rural areas, many urban households lack assets (includ- ing access to kinship and social networks, and to land), which constrains their ability to cope with unexpected stress and uncertainty. There are multiple causes contributing to this situation, including high levels of unemployment, low wages, non-payment of salaries, and poor access to basic social services and infrastructures. In fact, independent of their place of residence (rural, urban, or peri-urban), most families cannot survive with one income source, making the most common feature of livelihoods the diversification of strategies to make ends meet. A string of activities by various household members is thus required. The survival of families is guaranteed by a combination of several income-generating activities (formal and informal). Trade and commerce, most of which is informal, play a dominant role among income- earning activities in urban areas. Women have historically played an important role as traders. Indeed, women's role as traders goes back to precolonial times, and, for Bissau, have been particularly pivotal in ensuring the regular supply of fresh foods. Small garden agriculture practiced by women is a very important activity at the urban level (existing estimates by UNDP suggest that more than 2,000 plots were identified in the green belt of the city of Bissau). The principal products are vegetables that are sold in local markets for internal consumption. Among urban households, some 30 percent practice small garden agriculture, which during the dry season covers 70 to 80 percent of the demand for fresh vegetables (up to 60 percent in the rainy season). Trading women are often called bideiras, and many of them come from the coastal ethnic groups (Manjaco, Mancanha, Pepel). They mainly trade in foodstuffs, such as locally produced vegetables, fruits, alcoholic drinks (such as cashew wine), roasted groundnuts, and fish. As already mentioned, many also cultivate small horticultural plots for con- sumption and income. In Bissau, where the land traditionally belongs to the Pepel ethnic group, this often occurs on land that is either borrowed or rented during the dry season, when the land is not being used for rice cultivation. As many of the local products are seasonal, however, sellers often need to complement these activities by engaging in other strategies when these products are out of season, such as the collection and sale of firewood and coal. Many women also engage in the provision of personal services, including doing domestic work for others, dressmaking, and weaving. The participation of women in market activities intensified following the process of market liberalization that was initiated in the mid-1980s. This may have resulted from an attempt to compensate for households' declining incomes throughout this period as well as women's increasing expenses as the state retreated from the delivery of social services (Lourenço-Lindell 2002). As found in the urban focus groups carried out for this study, women, have in some cases become the main breadwinners in the household--which accord- ing to some men have made women the de facto heads of the household (INEP 2005). This is also to some extent supported by the 2002 household survey data, which suggest that female heads of household tend to be better off than male heads of household. According Conflict, Livelihoods, and Poverty in Guinea-Bissau 65 to INEP, this is increasingly leading to a profound transformation of the notion of the family, particularly in the urban areas of Guinea-Bissau. This inversion of gender roles, they suggest, is on the one hand leading to an increasing independence and autonomy for some women within the household, but on the other hand has led to situations of instability within the family, sometimes resulting in separation and divorce, putting a further strain on households' capacity to cope. This latter point corresponds with the findings from the 2002 Light Household Survey (INEC 2002) that will be discussed in the next section, and that suggests that divorced or widowed heads of household are generally worse off. Nevertheless, while potentially better off than men, women are often more vulnerable given that they are the ones that are responsible for most family expenses, such as clothing, education, and health expenses. For urban men, casual work (surne) is increasingly the main activity and livelihood strategy as formal wage opportunities become limited. Most urban men indicate that they are "on call"--waiting for temporary jobs, typically in construction (including bricklaying), loading work (often at the dock), and transportation (taxis, private buses). While the par- ticular activities pursued depend on the location (neighborhood), ethnic group, and skills of the individuals, most of the men end up unemployed and peddling in the streets. The situation has worsened over the last decade, as the number of informal wage-seekers and operators in urban areas has been swelled by an exodus of youth from rural areas in search of a better life. Even children are increasingly turning to the informal sector to assist house- holds in making ends meet. Cardoso and Imbali (1993) document a rise in the level of street children in Bissau. Some, mainly men, have jobs in the formal sector, mostly as teachers or health workers. Other low-wage public sector employees include members of the military or police. A major- ity of these, however, aim to complement low and irregular payment of salaries by also turning to the informal sector, and often juggle multiple activities to ensure their livelihoods. Some men, particularly those drawn from the eastern ethnic groups (Fula, Mandinga, and so forth) also play a historically important role as traders, reflecting long indigenous trade traditions in Guinea-Bissau and the broader subregion. These workers, who go by the name of djilas, are mobile traders who tend to concentrate in the trade of imported foodstuffs and rice, and who in recent years have become dominant in Bissau. In urban settings, these men have recently also been joined by women also drawn from the eastern groups. Together with the bideiras, the djilas are part of "a marketing system [. . .] in which traders are linked by shifting and flexible ties into a decentralized and unbounded network that connects markets located in different areas" (Lourenço-Lindell 2002:76). These markets include lumos, so-called weekly and rotating markets spread around the country. Some of these traders may be considered formal in that they are registered and hold trade licenses, but because of their constant geographical mobility and the relatively small size of their enterprises, most of them easily escape registration and the various taxes and regulations of the state.26 26. While some of these do pay a daily fee to the municipal council, and are therefore supposedly "registered" and, hence, "formal" traders, according to a survey conducted by Lourenço-Lindell (2002: 74), most of the ones that rely on these strategies for their livelihoods do not. In any case, for all intents and purposes, even those that do pay the daily fee were involved in activities that could otherwise be described as "informal"--activities where participants lacked a written work contract, if employed, and which offered highly irregular incomes. 66 World Bank Working Paper In recent years, particularly after the war, there has been an increasingly visible presence in Bissau (especially in the large Bandim market) of merchants and traders from neigh- boring countries and others in the subregion (particularly Senegal, Mali, Guinea-Conakry, Mauritania, and Nigeria), though their presence is also noted in other regions primarily along the ancient trade routes through the eastern provinces of the country. While some of these traders have been able to build prosperous trade careers, often through engaging in vertical business networks headed by import/export firms, a large share of those households that rely on trading and commerce as a primary livelihood strategy seem confined today in low- income activities. Many indeed live from hand to mouth, and are barely surviving. In peri-urban areas outside of the Bissau metropolitan area, as well as in primary and secondary towns in the regions (where the regional and sector state administrations are based), livelihoods are further intertwined with rural/agricultural activities. Indeed, in the peri-urban town of Bubaque, on the island of the same name, most households engage in fishing and exploitation of forest resources (palm oil and wine). Public employment is scarce, and only a small minority engages in wage-labor in the small tourism industry on the island. Hence, the main difference between the livelihood strategies of poor households in peri-urban areas and rural households would be that some of these households are getting part of their income from household members working in the formal wage sector. Never- theless, the overwhelming majority of the peri-urban population lives off agriculture as in any Tabanca in the rural areas. What characterizes both those households that live primarily off agricultural activities and those that rely on income flows from public work or services, is that neither make enough from any one activity to guarantee their survival. Therefore, most households are forced to pursue several income and consumption streams from several different household members to ensure their livelihoods throughout the year. Income Sources, Cashew Nuts, and Fishing Income Sources of Households So far, we have focused on an analysis of modes of livelihood in rural and urban areas today. In the future, high and sustained growth rates will be needed in order to reduce poverty. The key for the poor to benefit from future growth will be that this growth takes place in large part in the areas where they are most involved, which includes cashew nut production and fishing. We now turn to these two sectors to analyze their strengths and weaknesses. However, before doing so, it is useful to provide empirical data on key sources of income for the poor to show that, indeed, the poor would stand to benefit from growth in these two sectors. Table 5.1 is derived from the 2005 small-scale IPSA survey that was carried out for this study. The table provides data on the main sources of income per pop- ulation tercile. Each of the three terciles represents a third of the households interviewed in the survey; we used terciles instead of quintiles because of the limited sample size. Because the data does not include a consumption module, we used an assets index instead to rank households by level of well-being. Thus the first tercile can be interpreted as rep- resenting the households who have the lowest level of wealth in the survey, while the third tercile accounts for the wealthiest households. Clearly, and as expected, agriculture-related activities are the main source of income for a majority of the population, especially in rural areas, and in the bottom third of the Conflict, Livelihoods, and Poverty in Guinea-Bissau 67 population. For three out of four households belonging to the first tercile, agriculture is the main activity, and for another 19 percent, it is the second-most-important activity. The share of households declaring that fishing is their first activity is much lower. However, when looking at secondary activities, fishing is clearly important, especially among the poorer segments of the population. In contrast, wages and trade activities are more preva- lent among wealthier households. Thus, since the poor are mostly involved in agriculture (including cashew nut production) and in fishing activities, they stand to benefit substantially from growth in these sectors, provided that mechanisms ensure that the poor get their fair share of the revenues generated by these sectors (for example, producers should receive a fair share of the export price for cashew nuts). In what follows, we look at key characteristics of the cashew nut and fishing sectors in order to shed light on opportunities and risks in both sectors. Given that a separate analysis of cashew policies is available in Chapter 6, we only provide here a brief discussion of key risks in the sector because of its interaction with rice production. Our discussion of the fishing sector is more extensive, since it is not covered elsewhere in this book. The Cashew Sector It is argued in Chapter 6 that the cashew sector is a key potential source of growth for Guinea-Bissau, and that this could yield significant benefits for the poor. A recent study requested under the World Bank's Private Sector Rehabilitation and Development Project and conducted by Jaeger and Lynn (2004) provides a very good overview of the sector. The study notes that with production rising to 83,000 tons, cashew nuts now represent more than 90 percent of the country's exports. However, the Guinea-Bissau cashew sector has so far failed to move beyond commodity production. There is neither a positioning in the market nor any added value to the crop. If the country were able to move up the value chain in production by processing cashew nuts instead of simply exporting raw cashews, this could generate significant growth, with potentially large benefits and less vulnerability for the poor (see Table 5.1). At the same time, the monoculture of raw cashews also poses a threat to long-term development and food security because of the negative impact that cashew production has had on the rice sector. Although rice is the main staple food in the country, its farming has sharply declined in favor of cashews over the past three decades. As a result the country relied heavily on rice imports to meet the domestic demand for cereal. The gap filled by rice imports represents about 25 percent of the trade balance. Bringing rice production up to pre-independence levels would require the introduction of high-yield rice seeds, and the rehabilitation of rural roads. The rice deficit has been filled by barter transactions of rice and raw cashew nuts starting in 1984. In recent years more cash has been entering the value chain of cashew production, but the cashew-rice nexus and barter system still persists. This is particularly dominant during the early stages of the cashew season, which is the most acute part of the dry season. Today between 50 percent and 70 percent of all imported rice is likely to be used for barter- ing purposes. While there are upsides to the bartering system, in that it helps address the primary needs--access to rice--of households particularly in the early parts of the cashew harvest season, it often means that money is not entering the value chain at this level, and partly as a result often small farmers do not get to benefit from the potential value of the 68 World Bank Working Paper Table 5.1 Main Sources of Income for Households by Tercile, 2005 Tercile Area 1 (poorest) 2 3 (richest) Urban Rural Total First source of income (%) Agriculture 73.38 61.94 24.11 29.01 68.5 55.06 Livestock 0.00 0.00 0.89 0.76 0.00 0.26 Small animals 1.44 0.00 0.00 0.00 0.79 0.52 Fishing 3.6 1.49 2.68 3.05 2.36 2.6 Extraction activities 4.32 3.73 0.89 1.53 3.94 3.12 Wages 2.16 7.46 25.89 24.43 3.94 10.91 Seasonal work 2.88 6.72 4.46 3.82 5.12 4.68 Formal employment 1.44 2.99 3.57 4.58 1.57 2.6 Petty trade 2.88 5.22 13.39 13.74 3.15 6.75 Trade 1.44 2.24 8.04 5.34 2.76 3.64 Leasing (aluguel) 0.00 0.00 1.79 1.53 0.00 0.52 Remittances 0.00 0.75 0.89 0.76 0.39 0.52 Spiritual services (Murundade) 2.16 1.49 0.00 0.76 1.57 1.30 Others 4.32 5.97 13.39 10.69 5.91 7.53 All sources of income 100 100 100 100 100 100 Second source of income (%) Agriculture 19.10 16.25 25.93 23.33 18.75 20.40 Livestock 7.87 5.00 3.70 4.44 6.25 5.60 Small animals 1.12 3.75 1.23 0.00 3.13 2.00 Fishing 20.22 7.50 4.94 2.22 16.25 11.20 Extraction activities 12.36 13.75 2.47 5.56 11.88 9.60 Wages 3.37 3.75 8.64 8.89 3.13 5.20 Seasonal work 7.87 8.75 2.47 4.44 7.50 6.40 Formal employment 3.37 3.75 4.94 5.56 3.13 4.00 Petty trade 16.85 17.50 28.4 26.67 17.50 20.80 Trade 3.37 7.50 2.47 2.22 5.63 4.40 Alugel 2.25 0.00 0.00 1.11 0.63 0.80 Pension 0.00 0.00 3.70 3.33 0.00 1.20 Remittances 0.00 6.25 3.70 2.22 3.75 3.20 Transport work 1.12 0.00 0.00 0.00 0.63 0.40 Spiritual services (murundade) 1.12 1.25 0.00 1.11 0.63 0.80 Others 0.00 5.00 7.41 8.89 1.25 4.00 All income sources 100 100 100 100 100 100 Source: Authors' estimation from 2005 IPSA survey. Conflict, Livelihoods, and Poverty in Guinea-Bissau 69 cashew that they produce. There is evidence that in the early stages of the cashew season (March­May), a larger share of the exchanges at the farm gate is through bartering. Despite a fixed price for cashews (250 CFAF/kg) in principle, producers during the early stages of the harvest frequently accept terms of trade at the ratio of 1:1 (in terms of bags of rice/cashew) or if cash is used at prices as low as 100 CFAF/kg. From June onwards, more cash is entering the cashew trade, and at times prices as high as 350 CFAF/kg can be found. It is likely that poorer producers benefit from such price variations. There have also been developments in the value chain itself with a higher role played by Indian firms. Some historical background may help to explain these changes. During the colonial era, Guinea-Bissau used to be at an aggregate level self-sufficient in terms of rice production, and was toward the end a net exporter of rice. The production of rice has steadily declined since then, with potentially great implications for food security. Small farmers have shifted or diversified their livelihoods by cultivating products that could be sold for cash or traded for imported rice. In the early 1980s, responding to a continuous fall in world prices for groundnuts, a new cash crop--cashew nuts--began to be exported by the state trade enterprises that were later joined by private operators. Export volumes increased rapidly. A majority of cashews are today supplied by small producers, who exchange them with traders and intermediaries for imported rice (early in the cashew season) or cash (later in the cashew season). These traders and intermediaries are connected to import/export firms who import rice to be bartered at the farm gate and export raw cashew nuts to other countries, particularly India, for processing. These firms tend to benefit greatly from the substantial differences in the value of the two products on international markets. Initially, the shift away from rice production toward cashew cultivation provided an opportunity for rice-producing populations to reduce food insecurity. The switch also reduced the overall labor burden. The move toward flexible exchange rates, as well as further liberalization of the economy, brought about favorable export conditions and a multi- plication of actors in the market. Yet although these growing exports may have positively impacted macroeconomic indicators, the cashew rice economy has discouraged rice pro- duction in the country and has acted to accentuate dependency on food imports. The close connection between the two commodities has led to a disinterest in importing rice outside the cashew season. The population, particularly the urban population, which tends to be dependent on rice imports, has had to put up with sharp fluctuations in the availability of rice and rice prices. In addition, interviews conducted for this study suggest that while con- ditions generally improved for all involved in the cashew trade, the civil war in 1998 and the instability that followed worsened conditions in rural areas. Many of the initial actors in the market, particularly traders and import/export firms, were forced to drop out of the market. As a result a relatively monopsonistic market ensued, where the conditions for cashew producers for trading rice and cashew seem to have worsened. The situation has recently become more complex through the enhanced role played in this cashew-rice value chain by Indian cashew processing companies. Especially since the war, the banking and financial sector in Guinea-Bissau has collapsed, with the result that most private businesses have had to look outside of the country for access to capital and funds (more established firms, however, may still get access to capital). Very often the only actors willing to provide funds to these import/export firms have been the Indian processing companies. Being both semi-monopsonistic buyers of cashews as well as the "lenders of 70 World Bank Working Paper last resort," these Indian firms are able to set very strict financing terms. Hard data are difficult to come by, but according to Badji, former minister of agriculture (interview of December 14, 2004), Indian firms get to buy cashew at prices well below the world market price. He estimated that producers in the past may have received about 20­30 percent of the full export value of cashew at the farm gate, with 20­25 percent going to intermediaries, and 40­60 percent going to import/export companies. Of this latter figure, Badji estimated that Indian companies received about 70 percent of the import/export companies' share, in return for providing financing. To compensate for this, he claimed, the actors at the lower end of the value chain have become increasingly squeezed, to the detriment of small producers. As discussed in Chapter 6, and as argued in more detail by Jaeger and Lynn (2004), Guinea-Bissau could potentially have a sizeable processing industry that would create thousands of jobs and earn a great deal of value added that is currently lost by not process- ing raw cashew. The development of the cashew processing industry depends on the ability to attract investment, most of which will have to come from outside the country. Yet it will also be important to avoid the risk of an excessive dependency of rural and urban house- holds on cashew production (the latter through rice-cashew bartering). In addition to the very real risk relating to fluctuations in the price of cashews on the world market, people's well-being and livelihoods are also highly dependent on environmental and climactic variables related to the production of cashews. Recent news of a fungal disease having infected cashews grown in Guinea-Bissau, for example, could prove to have a devastating impact on the livelihoods of small farmers. Moreover, soil fertility is declining, partly as a result of the monocropping of cashews, which also reduces returns to small farmers over the long run. The Fishing Sector The fishing sector in Guinea-Bissau is underdeveloped. Despite this, fishing serves as an important livelihood and particularly a coping strategy for most coastal/river villages. Moreover, according to the ministry of fisheries, revenues collected from (offshore) fishing licenses to a range of large, mainly industrialized countries (European Union, China, Italy, Japan) currently contributes about half, if not more, of government revenues (existing data refer to estimations of 40 percent of government revenues, but in the last three years actual contributions have been higher). The fishing sector, hence, provides a good entry point to address the interplay between government policy at the macro level (for example, what role is played by different actors--private sector, license-holders, including donors, other gov- ernment ministries, and so forth), as well as implications for livelihoods at the microlevel. Given the heavy reliance on offshore fishing licenses, this discussion also introduces ques- tions of transparency of government revenue flows (where does the money go?), as well as mechanisms of financial and public accountability that are of great importance in countries that are highly dependent on natural resources. While it is beyond the scope of the present chapter to discuss all these issues, we can nevertheless outline the key economic opportunities that fishing could represent for poverty reduction and growth. Foreign fishers, operating both artisanal and industrial fishing vessels, have historically caught most of the fish in Guinea-Bissau, and continue to do so. Although open to the sea, Guinea-Bissau is essentially an agricultural country, in which traditional farmers in coastal Conflict, Livelihoods, and Poverty in Guinea-Bissau 71 areas merely supplement incomes and consumption, especially in the dry season and difficult years, by fishing for subsistence. There is little tradition or culture of involvement in marine fisheries as exists in Senegal or Ghana. While artisanal fishing in Guinea-Bissau, by being limited to areas close to the coast, the islands, and the rivers, should in theory face little direct competition from industrial and foreign fishing, in practice foreign artisanal and industrial fishermen fish within the zones reserved for local artisanal fisheries. Artisanal fisheries are concentrated along the coast, islands, and rivers, and the mullets and ethmaloses found here constitute roughly 80 percent of the artisanal catch. A majority of this catch is taken during two periods of the year, from April to May (also coinciding with the dry/hungry season) and from October to December. Currently about 3,500 artisanal fishers are reported to be operating in the country, of which at least half are from Senegal and Guinea-Conakry. Their number may have been almost halved since the war, when it is thought that many of the foreign artisanal fishers abandoned Guinea-Bissau. Yet most of the artisanal catch is still taken by foreign artisanal fishers and likely transported to neighboring countries. Only about 10 percent or so is reported to be taken by domestic fishers. This small proportion is attributed to the fact that fishing in Guinea-Bissau has tra- ditionally been a marginal activity--mainly to complement agricultural activities, or as a coping strategy--practiced some 4­6 days a month, over 8­10 months, for a total of less than 50 days a year. According to data from the FAO, in 1996 fishing was the main activ- ity of approximately 18 percent of the total numbers of fishers while agriculture occupied 78 percent. These fishers typically use small, traditional canoes, and hand lines or small gillnets. A second reason for the small share of the artisanal catch accounted for by nationals is that declining catches in Senegal, Guinea-Conakry, and Ghana have caused a substantial migration of fishers from these countries into Guinea-Bissau waters. The net benefits for Guinea-Bissau from this migration are likely to be modest and even probably negative-- these fishermen sell most of their catch back to their own countries, and pay nominal license fees or fish illegally--while the costs, in terms of additional fishing efforts on already heavily exploited stocks, may well be substantial. As a result, although foreign fishermen have been fishing in local waters for decades, reports of clashes between local and foreign artisanal fishers appear to be increasing. While domestic artisanal fisheries are small, the share of added gross production value is far higher in artisanal fisheries (75­90 percent) than in industrial fisheries (35­50 percent), and local processing and marketing create additional value. Artisanal fisheries also support service industries (boat building, processing, and transport). Nevertheless, the local market for fish in Guinea-Bissau is not well developed yet and prices are generally low; virtually no export-oriented processing and marketing facilities exist that could offer higher prices, as exist in Senegal. Local demand for high-value fish is limited as purchasing power is declining. Few incentives exist for foreign fishers to land fish locally. Despite these constraints, artisanal fishing holds promise and the government is giv- ing high priority to the development of local artisanal fisheries, emphasizing the employ- ment creating potential and the potential of local value added that could be generated by domestic small-scale fisheries. The IUCN has supported a highly successful project based on comanagement in Buba (local microprojects, credit systems, low-cost bioecological and socioeconomic monitoring system). The World Bank/GEF is also launching a project to 72 World Bank Working Paper improve the protection of local marine parks and development of artisanal fisheries. The African Development Bank has also approved a project that includes a component in support of small-scale fisheries. As for industrial fishing, the fleet is authorized to fish anywhere in the country's Exclusive Economic Zone (EEZ) at least 12 miles off the coast; artisanal fisheries have exclusive rights within the 12-mile zone. In 2000, 60 percent of the legal fishing fleet constituted chartered vessels, 17.5 percent from the European Union, and 7.5 percent from China. Since then, the chartered fleet has declined dramatically from 85 in 2000, to 57 in 2001, to 4 in 2002, as license fees have increased four-fold (to 140,000 CFA/GRT). The number of licensed vessels has increased recently however, and given the almost complete lack of surveillance and enforcement, it is doubtful that all formerly licensed chartered vessels have left the country's waters. Some 80 percent of the so-called local industrial fleet (apparently many foreign com- panies have made deals with local companies to fish under the Guinea-Bissau flag) and for- eign licensed vessels land their catches in Dakar; a much smaller percentage in Abidjan and Las Palmas. Fish landings of foreign industrial vessels in Bissau are still rare, and mostly limited to those of Chinese vessels. In fact, demand for and prices of fish in Guinea-Bissau are low in comparison to Senegal in part because of a lack of fish processing capacity and because of price policies that depress prices for some industrially caught species in the local market. In addition, poor sector services, red tape and corruption, and the absence of a standard quality control system that satisfies major importer markets undermine efforts to bring fishing onshore in Guinea-Bissau. As already mentioned, surveillance and monitoring systems are lacking. While most licensing agreements put upper limits on the total tonnage of the vessels that are allowed to operate under the agreements, they do not in any way define how catches can be effectively monitored and transshipments controlled. Guinea-Bissau thus currently has in effect no way to determine the volume and value of the fish caught in its waters, and has no economic basis to assess the fair value of fishing licenses. A new licensing agreement with the European Union under a partnership approach includes more emphasis on mechanisms for moni- toring and surveillance, but the EU remains the only license holder that has agreed to help with monitoring efforts. Overall, the majority of the current domestic returns from the sector are derived from the sale of fishing licenses and compensation payments from foreign agreements, particularly that with the European Union. While exports of frozen fish and shrimp were once a signif- icant part of the economy (roughly 17 percent of total exports in 1987), they now constitute less than 1 percent of total exports. In 2002, estimated revenue from these compensation payments and fishing licenses amounted to about $19.7 million, but the potential of the sector for the economy, and for the poor who would participate in the expansion of artisanal fisheries, is much larger. As for the sector's employment level, it is difficult to assess because many fishermen fish only part time, often not more than a couple of days a month. While over 7,000 people are officially registered as industrial fishers and deckhands, less than 400 actually work at any one time. One estimate of sector employment suggests that some 15,000 people are involved in the sector: 2,100 fishers and some 13,000 people in ancillary services and industries. This figure, however, is probably too high and may include employment that no longer exists in local processing plants and service facilities that have shut down. Conflict, Livelihoods, and Poverty in Guinea-Bissau 73 Overall, the fisheries sector can have a potentially large impact on future growth and poverty reduction in Guinea-Bissau. To maximize this potential, efforts are needed in three areas. First, given the high importance of fishing to public finances, enhanced governance of the sector--including through improved monitoring, control and supervision of fishing activities, but also extending to enhanced transparency and accountability of the collection and use of fishing revenues--may be required to ensure the productive use of fishing resources in increasing growth and combating poverty. Second, the artisanal fishing sector needs to be further developed to enhance the returns to fishing for coastal and river com- munities. Onshore fishing resources remain largely underexploited. However, partly because of cultural reasons, but more importantly as a result of low incentives, coastal communities mainly pursue fishing to complement other livelihood strategies, or as a coping strategy during times of droughts and bad harvests. Finally, to enhance the domestic benefits of industrial fishing, there is a need to improve the investment climate to bring industrial activities onshore. This would help the poor indirectly through increased economic growth, but also directly through enhanced employment opportunities for the substantial potential pool of capable people that could work in the sector. Conclusion In order to reduce poverty, what matters is growth in the sectors where the poor make a living. Hence it must be recognized that the livelihood strategies of the rural poor differ across regions and are influenced by the social characteristics of households and their members. The main objective of this chapter is to present a qualitative analysis of livelihoods in Guinea-Bissau. The analysis suggests that markers of social diversity in Guinea-Bissau-- particularly in the form of ethnicity, gender, religion, and age--seem to interact in structured ways with the livelihood options and strategies chosen by different households. The differ- ences are key to understanding economic and productive activities, as well as to the poten- tial success of policy interventions aimed at enhancing the livelihood bases of the majority of Guineans. Trading and commercial activities, for example, tend to be conducted by males from the ethnic groups typical in the eastern provinces of the country, mainly Fula and Mandinga ethnic groups. Women from the coastal groups, such as the Manjaco, Mancanha, and Pepel ethnic groups, have also historically been important traders, particularly of food- stuffs. There is also some division as regards agricultural cultivation, with the coastal groups--and particularly the Balante--cultivating rice mainly in lowlands, paddy fields, whereas the Fula and Mandinga tend to specialize more in the cultivation of corn in the highlands. While these relationships between identity and livelihoods seem to be true at a general level, there are always spaces and options for reinterpretation, particularly in more ethnically and religiously mixed areas, such as urban and peri-urban spaces, as well as in times of stress and coping, which have been typical in recent years. Across the country, living conditions are especially difficult for rural households, who face a range of constraints in improving their livelihoods. Agricultural activities are mainly nonmechanized and rely in most areas on manual labor only (the use of animals for traction and plowing is more common only in the eastern part of the country). Many households, furthermore, lack basic tools and implements, such as plows and rice husking tools. Access 74 World Bank Working Paper to markets, credit, and inputs (such as seeds and fertilizers) is low across the country, some- thing which is exacerbated by the devastated transport and energy infrastructure, the limited activities of the private sector particularly in rural and remote areas, as well as a retreating and reduced state with limited capacity and resources to provide agricultural extension services to the rural poor. Given many of these constraints, it is not surprising that house- holds participating in the focus groups that were implemented for this study reported the low value of local products and the inherent difficulties related to reaping the benefits of their production. The lack of certainty that results from this leads to low and suboptimal investments, in addition, of course, to subsistence-level harvests and consumption. While a lot of these constraints and obstacles could be overcome with a minimal level of investment, a lot of these investments presume a present and capable state, as well as a functioning and vibrant private sector. In the current context, however, the vacuum left by the state, par- ticularly in rural regions, has left the provision of most basic services to traditional and informal authorities and organizations. Guinea-Bissau has many natural resources with a lot of potential, particularly in agri- culture and farming (including cashews), forest exploration, and fishing. Indeed, the primary sector constitutes the backbone of the economy, providing more than half of GDP, account- ing for about 80 percent of the country's labor force, and more than 90 percent of current exports (mainly cashew). On the fiscal side, revenues from fish licenses alone ensure about half of the government's annual budget. Yet for the primary sector to ensure food safety and create new jobs, policymakers would need to adopt a coherent development strategy in the context of PRSP, which would aim at rehabilitating and encouraging rice production, diversifying agricultural production, promoting the processing of raw cashew into exportable cashew kernels, and developing the fishing sector. Such a strategy would also help in urban areas, which is important at this moment as the government is readying to implement major civil service and security sector reforms. These reforms would revive the private sector and contribute to create new jobs to absorb part of the labor force to be discharged from the public payroll. References Cardoso and Imbali. 1993. As questões institucionais e o programa estrutural na Guiné- Bissau. In INEP, Os efeitos sócio-económicos do programa de ajustamento estrutural na Guiné-Bissau. Bissau. INEP. Food and Agriculture Organization (FAO). 2001. Fishery Country Profile--Guinea-Bissau (2001). Rome: FAO. Available at: http://www.fao.org/fi/fcp/en/GNB/profile.htm. ------. 2004. Programme national d'investissement a moyen terme (PNIMT). Nouveau Partenariat pour le Développement de l'Afrique (NEPAD), Programme detaille pour le developpement de l'agriculture Africaine (PDDAA) République de Guinée-Bissau TCP/Gbs/2907. Food and Agriculture Organization and World Bank. 2002. Diagnosis and Emergency Restructuring Plan to Enhance the Domestic Benefits of the Fishing Sector and its Integration in the World Economy. Assistance to Fisheries Sectors in West Africa, Joint FAO/World Bank Fisheries Sector Memorandum. INEC. 2002. Evaluation de la pauvreté en Guinée Bissau (2001­2002). Conflict, Livelihoods, and Poverty in Guinea-Bissau 75 Instituto Nacional de Estudios e Pesquisa (INEP). 1996. Subsídio para a política de region- alização da SNV. Análise Regional: Tombali e Quinara. ------. 2001. Estudo qualitativo sobre o papel; da pobreza na Guiné-Bissau (Soluções e percepção local). INEP/PNUD Bissau. ------. 2005. "Analise social e de pobreza integrada--estudio qualitativo." Background report for the Poverty Assessment (IPSA), Bissau. Jaeger, P., and S. Lynn. 2004. "Guinea-Bissau Cashew Sector Development Study." Private Sector Development Project of Guinea-Bissau. Lourenço-Lindell, I. 2002. Walking the Tight Rope: Informal Livelihoods and Social Networks in a West African City. Stockholm: Stockholm University. Sylla, M. B. 2004. Evaluation de la pauvrete en Guinée Bissau (2001­2002). Guinea-Bissau. United Nations Development Programme. 2001. Relatorio nacional de desenvolvimento humano 2001. Segurança Humana. PNUD Guinea-Bissau / PNG. World Bank. 2001. "Economic Rehabilitation and Recovery Credit (ERRC) Memorandum of the President." World Bank Report no. P7373. CHAPTER 6 Cashew Production, Taxation, and Poverty in Guinea-Bissau Boubacar-Sid Barry, Edward G. E. Creppy, and Quentin Wodon Agriculture is the engine of Guinea-Bissau's economy. The sector relies mainly on cashew nuts, rice, and the subsistence production of food crops. Cashews represent 90 percent of the country's exports and the principal source of income in rural areas. Unfortunately, cumbersome administrative arrangements, weak legal systems, and an absence of credit often lead to high transaction costs for cashew buyers and exporters, which help decrease the farm-gate price of the raw nuts. This chapter provides a review of the cashew sector in Guinea- Bissau, as well as estimates of the likely impact of changes in farm-gate prices and export taxes on poverty among cashew producers and in the country as a whole. The chapter also notes that over the last three decades, the production of rice has significantly decreased in favor of cashew farming. This situation represents a threat to food security. For the rural sector to ensure food security and create new jobs, policymakers would need to adopt a coherent agrarian development strategy in the context of the PRSP, which would aim at rehabili- tating and encouraging rice production, and also promoting the processing of raw cashews into exportable cashew kernels, in order to generate more value added in the cashew sector. A griculture has been for a long time the backbone of Guinea-Bissau's economy as the country is endowed with fertile soils, abundant water, and a favorable climate. The sector represents more than half of the GDP, employs four-fifths of the labor force, and contributes to more than 90 percent of the country's export earnings through cashew nut exports. Over the past three decades, the production of cashew nuts has increased substantially. Guinea-Bissau is now the sixth-largest producer of cashew nuts (6 percent of the world cashew production) after India, Vietnam, Brazil, Ivory Coast, and Tanzania. While the cashew sector has had large positive macroeconomic impacts, and while it is the main source of livelihood of many among the rural poor, it suffers from a number 77 78 World Bank Working Paper of weaknesses. First, as in the other economic sectors, public investments in agriculture have stalled in recent years due the reallocation of public resources to defense and security programs. The efficiency of the cashew sector has probably been affected by this lack of investment, if only in terms of transport costs. Similarly, private investments suffered from the lack of adequate financing mechanisms, which is one of the reasons why Guinea-Bissau produces and exports raw cashews instead of moving up the value chain. This situation places the sector as well as cashew producers in a vulnerable condition vis-à-vis price fluc- tuations on the international markets, and the absence of a coherent rice-cashew strategy represents a major threat to rice production and food security because the increase in cashew production has been obtained at the detriment of rice production (there is in fact a whole system of barter of cashew nuts for rice). On a regulatory and administrative level, the cost of doing business in the cashew sector also remains relatively high, as documented by Jaeger and Lynn (2004).27 Instability and weak governance remain major disincentives for much-needed domestic and foreign investments in the sector. This situation is further complicated by changing regulations, unclear and cumbersome administrative requirements for existing and new businesses, and dysfunctional legal enforcement as well as inadequate utilities and seaport systems. In addition, the current export tax on cashew nuts (6 percent) is high relative to other export taxes (typically at 2 percent), thereby having a negative effect on rural incomes. The objective of this chapter is to provide a brief review of key issues in the cashew sec- tor, as well as estimates of the potential impact on poverty of selected cashew policies. It is found that an increase in farm-gate prices, which could be achieved through a reduction in transaction and financing costs, could help to reduce poverty. The same would likely be true for a reduction in export taxes. In contrast, an increase in export taxes on raw cashew exports in order to promote the creation of processing facilities could, at least in the short run, affect farmers negatively. While all these results are not sufficient to serve as a basis for policy rec- ommendtions, they provide important insights into the economy of Guinea-Bissau. Guinea-Bissau's Cashew Sector: Brief Review Livelihood of the Poor and Farm-gate Price Guinea-Bissau is endowed with fertile soils, abundant water, and a favorable climate. Agri- cultural production includes cashew nuts, rice and other cereals, fruits, fishing, livestock, and forestry products. Exports are, however, concentrated on cashew nuts, with only min- imal additional revenues from fish and seafood, fruits, palm kernels, and timber. Started as a marginal activity with an export volume of about 2,500 metric tons in the 1970s, exports of cashew nuts have increased steadily during the past three decades, reaching over 80,000 metric tons by 2004. Today, the country is the sixth-largest producer of cashew nuts (6 percent of the world production) after India, Vietnam, Brazil, the Ivory Coast, and Tan- zania. Episodes of good cashew performance are often associated with higher per capita incomes and better economic performance, as illustrated in Figure 6.1, suggesting a key role for the sector in the economy. 27. For further reference see Government of Guinea-Bissau (2004, 2005a, and 2005b) and Franca (1994 and 1995). Conflict, Livelihoods, and Poverty in Guinea-Bissau 79 Cashew planting was first established in the northern regions of Ohio and Biombo, and expanded thereafter rapidly in the east and the south. According to estimates by ANAG (quoted by Jaeger and Lynn 2004), about 5 percent of the country's land or 175,000 hectares are devoted to cashew production, and the land allocated to cashew nut production con- tinues to grow today at a rate of about 4 percent per year. Due to the fact that many trees planted in the recent past are bearing fruit, exports may very well reach up to 140,000 tons by 2010. Even more conservative estimates would lead to a level of exports above 100,000 tons per year by 2010. Small farmers, most of whom are engaged in cashew production, com- pose approximately 90,000 households with an average farm size per household of less than 3 hectares. Larger farmers or ponteiros, with land right concessions assigned by the govern- ment, own some 2,200 properties covering 27 percent of the country's arable land, with each property ranging from 20 to 3,000 hectares. Figure 6.1 Real GDP and Cashew Production Volume, 1999­2003 (change in %) 10.0 80 70 5.0 60 0.0 50 ­5.0 40 ­10.0 30 20 ­15.0 10 ­20.0 0 ­25.0 ­10 ­30.0 ­20 Real GDP Growth Rate (left Growth in Cashew Production (right axis) axis) Source: Authors using IMF and FAO data. As shown in Figure 6.2, an apparently positive feature of the cashew market in Guinea- Bissau is the fact that farmers get a fairly high share of the export price, about 70 percent according to Jaeger and Lynn (2004). This suggests that, within Guinea-Bissau, farmers may get their fair share. However, cashews are provided to exporters by traders and inter- mediaries who buy the crops from farmers, which are then exported to India for processing. Due to the weakness of the banking sector in Guinea-Bissau, Indian firms often provide financing for exporters, and this is done on strict financing terms, thereby putting downward pressures on the prices paid to exporters and ultimately the farmers. As cited in Chapter 5, Badji, a former minister of agriculture interviewed in December 2004, estimated that farm- 80 World Bank Working Paper ers may overall obtain only 20­30 percent of the full export value of cashews; with another 20­25 percent going to intermediaries, and a much larger 40­60 percent going to import and export firms. Specifically, in return for providing financing, 70 percent of the share allocated to exporters and importers may very well benefit Indian importers rather than exporters from Guinea-Bissau. The estimates provided by Badji are much less favorable than those provided by Jaeger and Lynn (2004) regarding the transaction costs in the market within Guinea-Bissau, and they also put the price received by the country's exporters in broader perspective. Said differently, farmers in the end may well benefit from only a fairly limited share of the actual value of their production on the world market. Still, working in the sector is beneficial for agricultural workers who are hired to col- lect the crops and perform other work. Brushing, forest clearing, and planting are the pri- mary activities required for establishing a cashew plantation. These activities are generally carried out by unskilled laborers and can take 30 days of labor at a cost of 60,000 CFAF per hectare ($110, estimates provided by Jaeger and Lynn 2004). Small farmers typically col- lect their own production, while larger planters hire workers for the job. Jaeger and Lynn (2004) suggest that about CFAF125 ($0.25) per kilogram is paid as labor costs during col- lection (50 percent of the average farm-gate price per kg). If a worker gathers 35 kg/day of nuts, he or she will earn CFAF 8,750 every four days, or 2,188 per day (about $4.50), which is an attractive rural wage rate. Figure 6.2 Breakout of Cashew Export Costs 60 50 40 $mn 30 US 20 10 0 Farmer Price Collection & Exporter costs Port charges & Customs Taxes Export evacuation stevedoring Inspection and Certificates Source: Jaeger and Lynn 2004, based on exporter interviews and government figures. The more fundamental issue is that Guinea-Bissau essentially produces and exports raw cashews, while the sector would generate much more value added if cashew nuts were processed and conditioned before being exported. Another constraint facing the sector is the absence of research and development, and vulgarization activities. Cashew trees are grown Conflict, Livelihoods, and Poverty in Guinea-Bissau 81 naturally without using advanced scientific methods, while in other countries research programs are carried out in order to improve the size of the nuts and their capacity to resist deadly diseases such as the Fungi Anthracnose and Oidium Anacardii, which have already affected plantations in Mozambique. So far, Guinea-Bissau has been fortunate as these infections have not spread in the country, but risks remain. The sector also remains vulnerable to price shocks and a decline in export prices over time (see Figure 6.3), with a decline over time of the price of the commodity and adverse price fluctuation on the world markets translating into some volatility for cashew producers. In addition, there is also volatility in cashew prices within a given year. For example, the farm gate price was between 60­70 percent of the FOB value of the export price in 2004. Jaeger and Lynn (2004) note that the price doubled during the first weeks of the season, fell back as the export buying price moved higher, and then dropped again because of increased speculations on the side of the Indian exporters. Figure 6.3 Trend in Cashew Prices, 1979­2003 1600 1400 1200 1000 Trend 800 $/ton 600 Actual 400 200 0 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 Source: Authors based on FAO data. In addition, as already mentioned above when discussing the prices received by exporters, microfinance is too weak not only to finance the whole value-added chain but also to finance other improvements in the sector. There is only one commercial bank in Guinea-Bissau and nearly all banking activities are concentrated in the capital city, Bissau. Commercial loans are limited to a few small enterprises in the form of short-term agricul- tural working capital (crédit de campagne). Access to credit is very limited for the majority of the population because of lack of collateral. Improving access to microfinance for small holders could facilitate the diversification of the sector and help create jobs. Regulatory Framework, Transaction Costs, and Taxation28 Guinea-Bissau could develop a sizeable cashew processing industry that could create thou- sands of jobs and earn a great deal of value added. However, the development of such an 28. This section is based in large part based on Jaeger and Lynn (2004). 82 World Bank Working Paper industry requires both domestic and foreign investment. A study by Jaeger and Lynn (2004) suggests that obstacles to such investments include a lack of adequate institutional and regulatory arrangements, as well as high domestic and international transportation costs due to poor rural road infrastructure and the small number of freight companies, and, to a lower extent, taxation. Consider first the broad institutional and regulatory setting. Political instability and weak governance have created major disincentives to domestic and foreign investments necessary for the development of the sector. There are numerous risks associated with changing regulations, unclear and cumbersome administrative requirements for existing and new businesses, dysfunctional legal enforcement, and inadequate utilities and seaport systems. Addressing these issues would require improving the investment climate (simpli- fying and clarifying relevant laws and taxes, streamlining investment and export-related administrative procedures, rehabilitating the physical and administrative infrastructure of the port, and so forth). In regard to the cashew sector specifically, laws concerning cashew nut trading have changed often in recent history, including two changes in 2000 and another in 2001. The laws are unclear, and the traders believe that enforcement may not always correspond to what the laws say. For example, it is currently not clear whether foreign enterprises are per- mitted to buy raw cashew nuts as intermediaries or exporters. Not only does this reduce the competition for the nuts but the uncertainty is an unambiguous danger signal to investors. When enforcement is believed to be contrary to the law, there must be recourse through the courts, which requires an effective judicial system. In addition, the absence of coordination between producers and other stakeholders, as well as the lack of supervision on the government side also contribute to persistent weaknesses in the sector. Currently, there is no private organization that encourages and sustains collaborative actions between all parties involved in the cashew chain. Stakeholders in the cashew sector need to organize themselves under a structure that they create to promote their mutual interests. On the government side, a small office dedicated to policies related to cashews could also be set up. Another issue is the fact that transportation costs are relatively high, especially for exports out of the country. The country does not have a railway system, and possesses only 10 per- cent of paved roads in a network of 4,400 km. The road network in rural areas is generally unpaved and in bad condition. The country has a large river transportation network yet to be developed. Beyond the poor infrastructure within the country, the cost of freight to India (a major importer of Guinea-Bissau's cashew nuts for processing) is significantly higher in Guinea-Bissau than in other West African ports. The cost of a 20-foot container shipped from Bissau to India was quoted as $1,400 in June 2004 (rising to $1,700 by late June), compared with $600 from Abidjan. The higher costs are in part because of shipping line charges for calling at Bissau ports and the absence of competition in Guinea-Bissau's freight market. Finally, taxation of the sector, although reduced over time, may still lead to some negative trickle-down for farmers. Exports of cashew nuts are currently taxed at the rate of 6 percent (since 1989 this rate has been gradually reduced from 34 percent). The reduction was intended to boost raw cashew export. However, the authorities may decide in the future to promote cashew processing by raising rates in order to discourage raw cashew exports. Although the overall net effect of such policy may be positive in the long run, it could have a strong negative effect on rural incomes in the short term (as kernel processing will take years to fully develop). If 80,000 tons are exported at $650 per ton, a 10 percent Conflict, Livelihoods, and Poverty in Guinea-Bissau 83 tax rise could equate to $5 million, a big part of which might in the end have to be borne by farmers. Since farmers benefit only from part of the export price, a 10 percent tax increase at FOB level would be equivalent to a drop in revenues of at least 14 percent at the farmer's level under full pass-through. Thus, in the absence of a processing capacity, which would lead to higher margins, a tax increase could penalize farmers and worsen poverty. Impact of the Cashew Boom on Rice Production Rice production represents a fundamental livelihood strategy for most of the rural popula- tion in Guinea-Bissau, and is recognized as a priority in the Agrarian Development Policy Charter. Rice accounts for 62 percent of the national cereal production and 75 percent of actual cereal consumption equivalent to 130,000 tons per year or 130 kg per capita annu- ally. Yet, domestic paddy rice production constitutes 77,000 tons of milled rice only. This means that the country has faced a chronic rice production shortage of 45,000­60,000 tons in recent years. The gap is filled by rice imports, which have represented about 25 percent, or $13.4 million, of the trade account deficit in recent years. This is in sharp contrast to the production levels attained before independence, during the 1950s and 1960s, a time at which production used to generate a surplus for export to neighboring countries in addition to urban centers. In fact rice production had increased by more than 10 percent per year even during the 1980s, largely because of improved economic incentives. But production stalled following the cashew boom of the 1990s, due mainly to increased reallocation of agricultural land and labor force to cashews (Figure 6.4). Figure 6.4 Cashew Nut Production, 1979­2003 (in metric tons) 90000 80000 70000 60000 50000 40000 30000 20000 10000 0 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 Source: Authors based on FAO data. There are several reasons for the growth of the cashew sector to the detriment of rice production. First, farmers are naturally inclined to engage in commercial crop pro- duction of more guaranteed value. Second, cashew production also allows farmers to 84 World Bank Working Paper reduce their production cost, especially in terms of labor inputs that are higher for rice than cashews. Indeed, virtually no labor is required until nut harvesting after planting. Thus new areas are increasingly converted into cashew plantations, especially in the eastern and southern part of the country. Most of such newly established plantations have yet to mature for fruit bearing, hence, as mentioned earlier, higher levels of pro- duction are expected in the future. Third, another disincentive for rice production is its resale value. The local rice variety is not viewed favorably enough to give incentives to engage surplus production beyond what would be necessary for subsistence needs for a family. The strategy of replacing a subsistence crop with a cash crop involves a major food security risk. Farmers need to survive within what the rice harvest permits as it comes to a close in September­October in the small valleys, or October­November in mangroves. They have to wait for the cashew nuts to ripen for harvest in March/April through June when barter transaction with rice is possible. Farmers' subsistence rice stock often runs out before such transaction is possible, leading to food security risks. This implies that cashew producers exchange part of their production to obtain rice. These barter transactions amount to CFAF 15 billion, or about $26 million based on local market value of raw cashew nuts. Necessary investment to attain self-sufficiency in rice production would be less than one-third of the costs paid in the barter system. The rehabilitation and promotion of rice production is seen as essential for ensuring food security and sustainable poverty reduction. Bringing rice production to pre-independence levels would require the introduction of high-yield rice seeds, and the rehabilitation of rural roads. Introducing new species of high-yield rice compatible with the country's cli- mate is one option. For example, the super ratooning rice brand developed in southeast China's Fujian Province can reach more than 16 tons per hectare per year. The species is able to bear fruit twice a year, and exceeded the Chinese national acceptable criteria for super hybrid rice. The high-yield rice may bring a higher yield and alleviate the heavy labor otherwise required of farmers in Guinea-Bissau. If all actual land devoted to rice production were cultivated using the type of rice described above, the country's total pro- duction of rice is likely to exceed 2.2 million tons per year in the long run. This would allow covering domestic needs and still potentially generate more than 2 million tons for exports. Achieving such a production capacity would of course require that farmers have access to credit for buying fertilizers and other modern agrarian equipment for sowing, maintaining and harvesting crops. It would also require good rural road networks in order to ship goods to local and regional markets. The first step toward planning such a change would be to prepare a coherent agricultural development strategy in the context of the PRSP, and reinforce rice production pilot phases currently being implemented in the eastern regions of the country. Distributional Issues in Cashew Production, Pricing, and Taxation Basic Statistics from the 2005 IPSA Perceptions Survey Household level data on cashew production are scarce. The nationally representative 2002 ILAP survey (INEC 2002, Sylla 2004) was to have a module on income sources, including income from various crops, but information by crop is not actually available in the survey Conflict, Livelihoods, and Poverty in Guinea-Bissau 85 data files. Therefore, the only source of information available to conduct some distributional analysis is a small-scale survey conducted in 2005 for the IPSA report on Guinea-Bissau by the World Bank (World Bank 2006). The survey was conducted in both urban and rural areas, with about 400 households participating. Apart from general information on income sources, the 2005 IPSA survey includes specific questions on cashew sales and barter. Summary statistics on the results are provided in Table 6.1. Slightly more than a third of the households (37.4 percent) have positive sales, with a higher proportion in the bottom tercile (46.5 percent) than in the top tercile (26.1 percent). The proportion of households with positive cashew sales reaches close to half among those household who declare having land, with smaller differences by tercile. While poorer households are more likely to produce and sell cashews, the total value of the sales is higher among wealthier households when the sample is restricted to those who sell cashews. However, the differences are not very large in the sample, ranging from an average sale value of about CFAF 121,000 in the bottom tercile to 176,000 in the top tercile. This lack of difference may be because of the small scale of the survey, in which large producers (ponteiros) are not likely to be well represented. Said differently, the sample is likely to represent the situation of relatively small producers only, rather than the situation of all producers in the country. Another important result is the fact that exchange by barter (exchange of cashews for rice) is very common, especially among the poorest tercile, where they represent 55.9 per- cent of all transactions at the end of the season (and 73 percent of all transactions year-round). In the top quintile, barter accounts for 39.3 percent of sales at the end of the season (59.4 percent year-round). This is probably because wealthier households have other means to purchase rice, and prefer to sell their production in cash rather than in kind. Table 6.1 Cashew Sales by Tercile, 2005 Tercile (%) Area (%) 1 2 3 Urban Rural Total Positive sale 46.5 39.4 26.1 28.9 42.8 37.4 Positive sale among those with land 50.4 47.0 43.6 50.6 46.4 47.6 Sales value when positive (CFAF) 120,924 147,540 176,189 170,292 131,332 143,020 Time of sale Beginning of season 1.3 4.3 3.1 2.5 3.1 2.9 Mid-point of season 3.1 2.6 1.5 2.2 2.6 2.5 End of season 22.7 23.9 36.0 29.3 25.6 26.9 Time of barter Beginning of season 5.3 10.6 12.2 11.8 7.7 9.2 Mid-point of season 11.8 7.1 7.9 7.1 10.0 9.0 End of season 55.9 51.5 39.3 47.1 51.0 49.6 Source: Authors from the 2005 Qualitative Survey. 86 World Bank Working Paper Impact of Cashew Prices and Taxes on the Poor Although the above data are fairly limited, they can nevertheless be used to perform indicative simulations regarding the impact of cashew policies on poverty and inequality if we are willing to make a number of assumptions and to combine these data with results obtained from the nationally representative 2002 ILAP survey. Two types of simulations are performed. First, we can assess the impact on poverty of an increase in the prices paid to producers, thanks, for example, to a higher price obtained by exporters or through a reduction in transaction and transport costs. Similarly, we can assess how a reduction or an increase in the tax to be paid on cashew exports would affect the poor. Consider first the impact of changes in the farm-gate prices of cashews. Empirical evi- dence worldwide suggests that improved access to markets through improved business environments; better institutional, regulatory, and tax arrangements; and adequate rural infrastructure may have a strong potential for reducing poverty. The specific discussion for Guinea-Bissau in the previous section suggested that although farmers may receive a relatively large share of the export price, the export price itself is low due in part to the unavailability of financing in the country. As exporters must rely on financing provided by Indian firms on strict terms, farmers ultimately suffer from low farm-gate prices. The simulations in Table 6.2 provide a very rough idea of how poverty might evolve among cashew producers if these producers were to receive a higher price for their crop. The first line in the table provides the value of household cashew production. For simplicity, we will assume that the cashew producing households from the first tercile of Table 6.2 in the 2005 IPSA survey are representative of the extreme poor who are cashew producers, while households in the first two terciles are representative of cashew producers who are either poor or extreme poor. Thus, the average value of cashew production among the extreme poor is estimated at CFAF 120,924 (from Table 6.1), while for the poor, the corresponding value is the straight average of the value of the production for the two first terciles in the 2005 IPSA survey, namely CFAF 134,232. For poverty analysis, we need to transform these estimates into estimates per equivalent adult, (each household member above 15 years of age counts for one equivalent adult, while household members below that age count for only one half of an equiv- alent adult, following the poverty measurement methodology used in the country). The esti- mated numbers of equivalent adults in Table 6.2 are from the 2002 ILAP survey. Dividing household cashew production value by the number of equivalent adults generates cashew production value per equivalent adult, which can then be compared to the levels of total con- sumption per equivalent adult in the 2002 ILAP survey in order to assess how consumption per equivalent adult might evolve with an increase in revenues from cashew nuts (assuming all the increase in revenues is used for consumption purposes by beneficiary households). The results suggest that a 15 percent increase in farm-gate prices for cashew nuts could result in an increase in consumption for the extreme poor of 9.5 percent, and 3.3 percent for the poor. The results for the extreme poor are likely to be overestimated because the estimated share of total per capita consumption accounted for by cashew nuts is large. On the other hand, the very poor who are not cashew producers themselves but work on farms as day laborers could benefit from trickle-down effects which are not accounted for here. Next, taking into account the measures of poverty and extreme poverty obtained in the 2002 ILAP survey, and estimates of the elasticity of poverty reduction to an increase in consumption (which are preliminary at this stage), we find that among cashew producers, extreme poverty could be reduced by 3.14 percentage points and poverty by 1.81 percentage point under a 15 percent Conflict, Livelihoods, and Poverty in Guinea-Bissau 87 Table 6.2 Impact of Change in Cashew Prices on Poverty Impact on Impact on Assumptions poverty extreme poverty Value of household cashew production (CFAF) 120,924 134,232 Household size (number of equivalent adults) 7.6 6.6 Per capita value of cashew production (CFAF) 15,911 20,338 Mean consumption/equivalent adult among the poor (CFAF) 25,152 91,895 Gain in consumption with 15% rise in cashew price (CFAF) 2,387 3,051 Headcount index of poverty/extreme poverty (%) 21.6 65.7 Elasticity of poverty reduction to consumption growth -1.53 -0.83 Change in consumption with 15% rise in cashew price (%) 9.5% 3.3% Change in poverty among cashew producers (%) -14.5% -2.8% Percentage point change in poverty among cashew producers -3.14 -1.81 Share of households producing cashew nuts (%) 46.5% 43.0% Impact on poverty/extreme poverty among all households -1.46 -0.78 Source: Authors from 2002 ILAP and 2005 IPSA surveys. increase in cashew prices. These are substantial impacts that result from the important role that cashew revenues play among those households who produce and sell cashews. Finally, in order to assess the potential poverty impact on poverty in the country as a whole, one must take into account the fact that not all the poor or extreme poor are cashew producers. Even though only about half of the households are likely to be involved in cashew produc- tion and sales according to the 2005 IPSA survey, the net impact on national poverty remains fairly large, with anticipated reductions in the share of the poor or extreme poor of 0.78 and 1.46 percentage points respectively. Note again that the estimates in Table 6.2 do not take into account potential derived impacts through wage increases for cashew farm workers. What about the impact on poverty of taxing cashew exports? An increase in the export tax on unprocessed cashew nuts of about 10 percent could potentially (with full pass- through to farmers) generate a reduction in their farm-gate price of close to 15 percent as discussed in previously (this assumes that farmers get 70 percent of the export price).29 In other words, this would have equal, but reverse, effects on poverty as the estimates in Table 6.2 (poverty could increase by the amounts estimated in Table 6.2). Alternatively, if the 6 percent export tax were eliminated, and if we were again to assumer full pass-through to farmers, farmers could benefit from an increase in consumption of 8.6 percent, which would have poverty impacts of about half the estimates presented in Table 6.2. The issue of the impact on producers of changes in export taxes implies a difficult dilemma, as an increase in taxes to encourage exporters to move up the value chain toward cashew process- ing may have short-term detrimental impacts on the poor. Yet in the medium- to long-term, it could be hoped that the positive effect on income of cashew processing on producers would more than offset the adverse short-term impact. If it were decided to increase export taxes, a gradual increase in the tax rate could help to reduce short-run effects. 29. See Diop and others (2005) for a similar case study on Rwanda. 88 World Bank Working Paper Conclusion Cashews represent 90 percent of the country's exports and the principal source of income in rural areas. Unfortunately, cumbersome administrative arrangements, weak legal systems, and an absence of credit have led to high transaction costs which decrease the farm-gate price of the raw nuts. This chapter has provided a review of the cashew sector in Guinea- Bissau, as well as estimates of the likely impact of changes in farm-gate prices and export taxes on poverty among cashew producers and in the country as a whole. The chapter has noted that over the last three decades, the production of rice has decreased in favor of cashew farming. According to some, this situation may represent a threat to food security. More generally, for the rural sector to ensure food security and create new jobs, policymakers would need to adopt a coherent agrarian development strategy in the context of the PRSP, which could among others aim at rehabilitating and encouraging rice production, and also promoting the processing of raw cashew into exportable cashew kernels in order to generate more value added in the cashew sector. References Diop, N., P. Brenton, and Y. Asarkaya. 2005. "Trade Costs, Export Development, and Poverty in Rwanda." World Bank Policy Research Working Paper no. WPS 3784. França, M. 1994. "Cashew of Guinea-Bissau: Potential and Constraints." Trade and Invest- ment Promotion Support (TIPS), Report No. 36E. ------. 1995. "Agro-negócio do arroz na Guiné-Bissau." Trade and Investment Promotion Project (TIPS)--USAID. Government of Guinea-Bissau. 2004. "Synthesis of Potentiality and Opportunities of Business and Investment in Guinea-Bissau." Ministry of Trade, Guinea-Bissau. ------. 2005a. "Programa integrado de industrializacao acelarada e producao de energia electrica: Castanha de Caju." Ministry of Trade, Guinea-Bissau. ------. 2005b. "Guinea-Bissau Cashew Sector Development Study: Part II Action Plan." Ministry of Finance, Private Sector Rehabilitation and Development Project, Guinea- Bissau. INEC. 2002. Evaluation de la pauvreté en Guinée Bissau (2001­2002). Jaeger, P., and S. Lynn. 2004. "Guinea-Bissau Cashew Sector Development Study." Private Sector Development Project of Guinea-Bissau. Lourenço-Lindell, I. 2002. Walking the Tight Rope: Informal Livelihoods and Social Networks in a West African City. Stockholm: Stockholm University. Sylla, M. B. 2004. Evaluation de la pauvrete en Guinée Bissau (2001­2002). Bissau. World Bank. 2006. "Guinea-Bissau: Integrated Poverty and Social Assessment (IPSA)." Report no. 34553-GW. Eco-Audit Environmental Benefits Statement The World Bank is committed to preserving Endangered Forests and natural resources. We print World Bank Working Papers and Country Studies on 100 percent postconsumer recy- cled paper, processed chlorine free. The World Bank has formally agreed to follow the rec- ommended standards for paper usage set by Green Press Initiative--a nonprofit program supporting publishers in using fiber that is not sourced from Endangered Forests. For more information, visit www.greenpressinitiative.org. In 2006, the printing of these books on recycled paper saved the following: Trees* Solid Waste Water Net Greenhouse Gases Total Energy 203 9,544 73,944 17,498 141 mil. * Pounds Gallons Pounds CO2 Equivalent BTUs 40" in height and 6-8" in diameter Conflict, Livelihoods, and Poverty in Guinea-Bissau is part of the World Bank Working Paper series. These papers are pub- lished to communicate the results of the Bank's ongoing research and to stimulate public discussion. Conflict and political instability have weakened Guinea- Bissau's productive infrastructure considerably during the past three decades. This situation contributes to an increase in the degree of vulnerability of the population, especially in rural areas where most economic activities continue to take place. As growth has been weak, poverty levels remain high. This book provides a collection of papers on conflict, liveli- hoods, and poverty in Guinea-Bissau based on both the nationally representative 2002 household survey and a small scale survey with both quantitative and qualitative compo- nents implemented in 2004. The chapters deal with growth and poverty, institutions and social networks, the determi- nants of poverty, the means of livelihoods of the population, and finally cashew production and taxation. World Bank Working Papers are available individually or on standing order. These are also available online through the World Bank e-Library (www.worldbank.org/elibrary). ISBN 0-8213-7021-9 THE WORLD BANK 1818 H Street, NW Washington, DC 20433 USA Telephone: 202 473-1000 Internet: www.worldbank.org E-mail: feedback@worldbank.org