AGRICULTURE GLOBAL PRACTICE DISCUSSION PAPER 10 AGRICULTURAL SECTOR RISK ASSESSMENT: METHODOLOGICAL GUIDANCE FOR PRACTITIONERS WORLD BANK GROUP REPORT NUMBER 100320-GLB JANUARY 2016 AGRICULTURE GLOBAL PRACTICE DISCUSSION PAPER 10 AGRICULTURAL SECTOR RISK ASSESSMENT: METHODOLOGICAL GUIDANCE FOR PRACTITIONERS © 2016 World Bank Group 1818 H Street NW Washington, DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org Email: feedback@worldbank.org January 2016 This volume is a product of the staff of the International Bank for Reconstruction and Development/The World Bank. The findings, interpretations, and conclusions expressed in this paper do not necessarily reflect the views of the Executive Directors of The World Bank or the governments 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 concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Rights and Permissions 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 permission to reproduce portions of the work promptly. 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, telephone 978-750-8400, fax 978-750-4470, http://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, e-mail pubrights@worldbank.org. Cover photos: Upper left: Kz rain desert—Mariusz Kluzniak. Upper right: Harvesting irrigated fields—Curt Carnemark/World Bank. Lower left: Woman plant—Yosef Hadar/World Bank. Lower right: Drought in Kenya—Neil Palmer/CIAT. CONTENTS Acronyms and Abbreviations vii Acknowledgments ix Executive Summary xi Chapter One: Introduction 1 Overview 1 What Is an ASRA? 2 Why Bother about Agricultural Risk? 2 Why an ASRA? 3 Volatility as the “New Normal” 4 Climate Change and Agricultural Risk 4 Resilience of Agricultural Systems and Climate-Smart Agriculture 5 Chapter Two: Agricultural Risk Management Conceptual Framework 7 The Conceptual Framework for ARM 7 Risk Management Strategy 11 Chapter Three: Operational Approach: How to Conduct an ASRA 13 What Is the Operational Approach? 13 The Sequential Flow Process 15 Work Plan for the ASRA 15 What Analysis and When? 15 Chapter Four: Step 1: Background Research and Risk Identification 17 Sources of Information 18 Understanding the Agricultural Sector 18 Constructing an Event Time Line 19 Selection of Commodities 20 Identification of Key Risks 21 Assessment of the Risk Management Landscape 21 Chapter Five: Step 2: Risk Quantification 23 Rationale 23 How to Quantify Losses Associated with Risk 24 Scenario with Poor-Quality Data 26 Scenario with No Data 26 Frequency and Intensity 27 A Note on Livestock 28 Quantifying Market Risks—Price Volatility 30 Limitations 31 Agricultural Sector Risk Assessment: Methodological Guidance for Practitioners iii Chapter Six: Step 3: Fieldwork 33 Purpose 33 Fieldwork Activities 33 The ASRA Team 34 Fieldwork Deliverables 35 Chapter Seven: Step 4: Risk Prioritization 37 Variables for Risk Prioritization 37 Prioritization Matrix 37 From Commodity Risk to Sector Risk 38 Chapter Eight: Step 5: Prioritization of Risk Management Solutions 41 Long List of Solutions 41 Risk Solutions Filters 42 Short List of Solutions 43 Chapter Nine: Understanding Vulnerability to Risk 45 Stakeholders’ Risk Profile Assessment 45 Incorporating the Vulnerability Assessment in the ASRA 46 Chapter Ten: Toward a Risk Management Strategy 51 Actions and Stakeholders 51 Key elements of the Risk Solutions Process 52 Risk management Intervention Areas 53 Stocktaking and Gap Analysis 53 Steps to Prepare the Risk Management Strategy and Action Plan 55 Contents of the Risk Management Strategy and Action Plan 56 Chapter Eleven: Final Considerations 59 The Operational Experience 59 The Importance of an ASRA 60 One Size Does Not Fit All 60 Challenges 61 A Last Word 62 References 63 Appendix A: Glossary of Terms 65 Appendix B: Agricultural Sector Risk Assessments Conducted by the World Bank 69 Appendix C: Data Requirements for Agricultural Sector Risk Assessments 71 Appendix D: Guidelines for Focus Group Discussions with Farmers 73 Appendix E: Example of Rainfall Analysis and Crop Production in Ghana 83 Appendix F: Guidelines for Consultants 91 Appendix G: Sample Terms of Reference for Consultants 95 Appendix H: Example of a Commodity Risk Profile in Ghana 99 Appendix I: Example of a Regional Risk Profile in Upper West Region of Ghana 101 iv Agriculture Global Practice Discussion Paper Appendix J: Table of Contents from an ASRA in Kenya 103 Appendix K: Prioritization of Risk Management Measures in Niger—a Filtering Approach 105 Appendix L: Examples of Agricultural Risk Management Solutions 109 BOXES Box 2.1: Distinguishing between Risks, Constraints, and Trends 9 Box 5.1: Alternative Approach to Calculation of Production Risk Losses in Mozambique’s ASRA 27 Box 6.1: Production Risk: A Bottom-Up Approach to Fieldwork 34 Box 6.2: Outputs of the ASRA Fieldwork 35 Box 8.1: Final Prioritization of Risk Solutions in Tanzania’s ASRA 43 Box 9.1: Taking into Account the Social Dimensions of Agricultural Risk in Brazil 49 Box 11.1: The Influence of ASRAs on World Bank Operations 60 FIGURES Figure ES.1: The Sequential ASRA Flow Process xiii Figure 1.1: Contribution of ASRA to Climate Change Adaptation 6 Figure 2.1: Key Components of Agricultural Risk Management 8 Figure 2.2: The Conceptual Framework for ARM 8 Figure 2.3: Risk Management Layers 11 Figure 3.1: The Sequential ASRA Flow Process 15 Figure 3.2: Calendar of Activities for an ASRA in Tanzania 16 Figure 3.3: Analytical Components at Commodity and Aggregate Levels 16 Figure 4.1: Time Line of Major Shocks to Agricultural Production in Senegal, 1980–2012 19 Figure 4.2: Events Time Line of Risks to Cocoa Production in Ghana 20 Figure 5.1: Paddy Rice Yield Losses in Tanzania (Tons per Ha), 1981–2010 25 Figure 5.2: Expected Average Losses for Adverse Crop Production Events in Niger 28 Figure 6.1: Field Team Responsibilities for an ASRA in Tanzania 35 Figure 7.1: Three Key Variables for Risk Prioritization 38 Figure 7.2: Reprioritization of Risks from Commodity to Sector Level 38 Figure 8.1: Risk Solutions Prioritization Process 41 Figure 9.1: Mapping Perils According to Vulnerability in Ghana’s Cocoa Sector 48 Figure B9.1.1: Risk Incidence Based on Likelihood, Expected Impact, and Relative Economic Relevance of Subsectors 49 Figure B9.1.2: Risk Incidence Based on Likelihood, Expected Impact, and Relative Economic Relevance of Subsectors Weighted by the Number of Families in the Family Agriculture Sector 50 Figure 10.1: Solutions Assessment Steps 53 Figure E.1: Weather Station Distribution with Region Centroids 84 Figure E.2: Monthly Rainfall Patterns by Region 85 Agricultural Sector Risk Assessment: Methodological Guidance for Practitioners v Figure E.3: Correlation Matrix Plot 89 Figure E.4: Factor Loadings Plot 90 Figure E.5: Mean Factor Scores by Region 90 TABLES Table 4.1: Relative Importance of Commodities by Value in Tanzania 20 Table 4.2: Key Identified Risks in Mozambique’s Agricultural Sector 21 Table 5.1: Step-by-Step Methodology for Quantification of Losses Using Yield Data 24 Table 5.2: Step-by-Step Methodology for Risk Quantification of Losses Using Production Data 24 Table 5.3: Paddy Rice in Tanzania: Losses for Years in Which Yield Dropped Below the Threshold 25 Table 5.4: Disaggregated Losses by Commodity in Paraguay 26 Table 5.5: Disaggregated Losses by Region in Paraguay 26 Table 5.6: Assessment of Frequency and Intensity of Risk in Mozambique’s Cotton Supply Chain 28 Table 5.7: Illustration of Risk Prioritization in Livestock for Rwanda 30 Table 7.1: Risk Classification Example 38 Table 7.2: Risk Prioritization Matrix for Cash and Export Crops in Tanzania 39 Table 7.3: Risk Prioritization Matrix for All Crops in Malawi 39 Table 8.1: Long List of Solutions from Tanzania’s ASRA 42 Table 8.2: Decision Filters for Prioritization of Risk Solutions 43 Table 8.3: Illustration of Decision Filters for Prioritization of Risk Solutions 43 Table 9.1: Vulnerability Matrix for an ASRA in Paraiba, Brazil 47 Table 10.1: Examples of Risk Management Measures 52 Table 10.2: Illustration of Project Stocktaking 53 Table 10.3: Gap Analysis Conducted for Tanzania’s ASRA 54 Table 10.4: Example of an Action Plan Structure 56 Table B.1: ASRA Country and Assessment Year 69 Table E.1: Standardized Cumulative Rainfall 86 Table E.2: Impact of Rainfall Parameters on Crop Yield 88 Table E.3: Principal Components Analysis: Three Eigenvalues and Proportion of Variance Explained 89 Table E.4: Correlation of Components 89 Table F.1: Template for Commodity Supply Chain Profile 92 Table K.1: Relative Benefits of Risk Management Measures 106 Table K.2: Decision Filters and Intervention Classification 107 vi Agriculture Global Practice Discussion Paper ACRONYMS AND ABBREVIATIONS ARM Agricultural risk management IFAD International Fund for Agriculture Development ARMT Agriculture Risk Management Team MT Metric ton ASRA Agricultural sector risk assessment NGO Nongovernmental organization CSA Climate-smart agriculture NRM Natural resource management FAO Food and Agriculture Organization (of the UN) OLS Ordinary least squares FAOSTAT Food and Agriculture Organization Corporate SD Standard deviation Statistical Database TA Technical assistance FMD Foot-and-mouth disease TTL Task team leader GDP Gross domestic product USAID U.S. Agency for International Development Agricultural Sector Risk Assessment: Methodological Guidance for Practitioners vii ACKNOWLEDGMENTS This report was developed by a team led by Vikas Choudhary, Senior Economist, and Carlos Arce, Consultant from the Agricultural Risk Management Team (ARMT) at the World Bank. The content was developed with contributions from Stephen P. D’Alessandro, Åsa Giertz, Kilara Suit, Traci J. Johnson, Tobías Baedeker, and Jorge Caballero. This work would not have been possible without the numerous contributions from all the agricultural specialists who have participated in conducting Agricultural Risk Assessments during the last four years, of whom there are too many to list. The team would also like to extend its special appreciation to the stakeholders of agricultural supply chains in Africa, Latin America, and Eastern Europe who actively engaged in the development of sector risk assessments in those regions, bringing a great sense of realism to this experience. Their active participation obliged the team to be realistic and practical. The team would like to thank peer reviewers for their honest and useful comments when revising earlier versions of this work. They are Willem G. Janssen, Holger A. Kray, and Diego Arias, from the World Bank. The team also thanks Lena Heron, Gloria Kessler, and Krista Jacobs from the U.S. Agency for International Development (USAID) for their support and suggestions for refining the document. The leadership of Marc Sadler, Mark Cackler, and Preeti Ahuja was instrumental in delivering the various risk assessments and developing these methodological guide- lines. Generous funding for this activity was provided by the United States Agency for Inter- national Development and a Multi-Donor Trust Fund on Agricultural Risk Manage- ment financed by the Ministry of Foreign Affairs of the Government of the Netherlands and the State Secretariat for Economic Affairs (SECO) of the Government of Switzer- land. Their encouragement and financial support allowed the gathering of significant experience that serves as the basis for this work. Agricultural Sector Risk Assessment: Methodological Guidance for Practitioners ix EXECUTIVE SUMMARY In the agricultural sector, risks are inherent and ubiquitous, posing potentially serious consequences for stakeholders and consumers. Risks disrupt supply chains, causing extensive financial and economic losses. Agricultural risks are also the principal cause of transient food insecurity, creating a poverty trap for millions of households across the developing world that enforces a vicious cycle of shock and recovery. Climate change is exacerbating this cycle by shifting the frequency and intensity of weather- related risks and increasing uncertainty. Effective agricultural risk management (ARM) is crucial to increasing economic growth, improving food security, and reducing poverty. Although levels of risk vary within and between countries, lower-income and highly agriculture-dependent coun- tries are more vulnerable to agriculture-related risks. In these countries, there is an urgent need to better assess risks, understand the interconnections between different types of risk, and improve agricultural risk management strategies. AGRICULTURAL SECTOR RISK ASSESSMENT: AN INTEGRATED APPROACH The World Bank’s conceptual framework for ARM looks at risk from an integrated perspective. The framework views agricultural sector risk assessment (ASRA) as a tool to help decision makers understand risk exposure and to provide the basis for develop- ing appropriate risk management solutions. ASRA is a consultative, time-bound pro- cess that provides an orderly way to analyze, identify, and prioritize risks. The process facilitates risk management policy and program design by providing a framework and tools for decision making. The ASRA’s target audience includes country-level stake- holders involved in selected agricultural commodity systems, development agency decision makers, and developing country policy makers. This report provides methodological guidance for conducting an ASRA. The guid- ance is based on the World Bank’s Agriculture Risk Management Team’s experience assisting governments in conducting sector and key commodity supply chain risk assessments in more than 20 countries. Agricultural Sector Risk Assessment: Methodological Guidance for Practitioners xi strategy facilitates expedient resource mobilization when PRACTICAL AND risk events occur. For example, ex ante planning can assist CUSTOMIZABLE TOOLS FOR farmers in getting back into production more quickly by PRACTITIONERS using previously identified delivery channels. Finally, by The document contains practical guidance on the com- better managing risk, poor rural households can avoid plexities of developing a holistic ASRA and provides a depleting their assets or increasing their vulnerability in stepwise approach to conducting ASRAs, incorporat- the future, particularly women in disadvantaged positions ing the lessons learned from assisting countries in main- in many countries. streaming risk management practices into mid-term development strategies and investments. Over time, the ASRA can add value in a variety of contexts, methodologies and tools have been refined, but the focus including (i) broader agricultural sector analyses and remains on approaching risks and risk management development strategy processes; (ii) constraint and through an integrated perspective. opportunity analyses undertaken in the identification and formulation of development projects that focus on The guidelines in this report are primarily designed for value chain integration, agricultural commercialization, use by development practitioners working in the agricul- rural finance, export promotion, landscape planning, tural sector in developing countries. The framework can and so on; (iii) planning, implementation, and monitor- be adapted to the particular circumstances of developing ing of sector reform programs, including those involv- economies. In ARMT’s experience, each specific assess- ing shifts in the commercial, regulatory, and other roles ment needs to be adapted to account for a country’s of governments in agriculture; (iv) investment apprais- unique combination of risks, institutional arrangements, als by private and development finance institutions or as and fiscal constraints. part of strategic assessments of agricultural lending portfolios’ quality and risk exposure; and (v) as an entry point for the introduction of climate-smart agriculture REAPING THE BENEFITS (CSA) aiming at strengthening resilience in agricultural OF ASRA systems. The ASRA process is a rigorous, inclusive approach to identifying potential solutions to mitigate, transfer, and ASRA is devised to support broader sector strat- cope with agricultural risks. It is also an entry point for egy formulation efforts and the identification and for- mainstreaming ARM solutions in national development mulation of proposals for investment, capacity building, strategies, policies, and programs. By bringing together a and policy and regulatory reform in strategically impor- broad group of stakeholders from the public, private, and tant agricultural supply chains or within the sector as a nongovernmental sectors, an ASRA can facilitate knowl- whole. edge exchange between siloed academic disciplines and ministries. The process also facilitates priority setting and The outcome of an ASRA is only a first step in the alignment of policies and programs to support country risk management process. Real changes that incor- development objectives. porate comprehensive risk management into planning and budgeting will likely be a gradual process. Ultimately, The benefits of a more holistic and targeted risk integration of risk management practices into develop- management strategy are many. Adopting a risk ment strategies will strengthen resilience in the longer management strategy can help mitigate risks before they term, thus reducing vulnerabilities among agricultural occur. It can also lessen the impact on the government’s sector stakeholders and increasing the success of agricul- fiscal balance, particularly at catastrophic levels, by tem- tural investment strategies. pering the need for costly and often poorly targeted post shock humanitarian relief to affected communities. One size does not fit all. Owing to the diversity of Additionally, a well-planned ex ante risk management risks, approaches for managing them, and country xii Agriculture Global Practice Discussion Paper contexts, a singular blueprint for a risk management road- This sequential flow process has proved to be map is not feasible. This document presents many exam- very practical in conducting ASRAs to date, but it ples of risk management strategies that can be pursued by should be adapted to the particular circum- various groups of stakeholders. stances of the country where it is applied and should respond to the assessment’s specific objectives. The ASRA IMPLEMENTATION step-by-step activities shown in this diagram are sequen- tial; the findings of each step inform and serve as the basis This report presents detailed operational guid- for the next. The immediate outcome of the process is a ance on how to apply the conceptual framework short list of prioritized risks. for conducting an ASRA. Rather than presenting a methodological blueprint, this document outlines a prac- Based on the risk prioritization, a short list of tical approach that can be adapted to the particular cir- potential risk management solutions can be cumstances of developing economies. The ASRA may identified. This outcome is the basis for designing a risk be tailored to take into account the sector’s structure and management strategy that in turn has various steps, end- institutional dynamics, including factors such as supply ing in an action plan for implementation. The short- chains, institutions, fiscal constraints, and stakeholders’ listed solutions can be evaluated using decision filters, capacity to deal with the effects of realized risks. Likewise, such as feasibility and affordability, either in a consultative risk management solutions identified in a particular coun- stakeholder setting or individually. This report contains a try will not necessarily fit the context of another country, list of risk mitigation, transfer, and coping solutions iden- however similar they may appear. tified in previous ARMT country engagements as well as a list of potential decision filters. These lists are not The ASRA process is a dynamic one that requires careful exhaustive, but provide a starting point for discussion. planning given the participation of a variety of stakehold- ers and the degree of analytical skills and experience This report presents a framework for the risk desired for the assessment team. Figure ES.1 summarizes identification process, providing comprehensive the step-by-step ASRA process. instructions for methodical data collection and targeted analysis. The authors discuss methods for quantifying risks under different data quality scenarios, approaches for overcoming common data limitations, and FIGURE ES.1. THE SEQUENTIAL ASRA FLOW alternative methodologies for quantifying risk and esti- PROCESS mating losses when data are incomplete. Examples of data visualization tools are included, which are useful for understanding the potential impact of different risks and for communicating findings with country governments and other stakeholders. Risk Assessment / Risk Monitoring Reevaluation TOWARD A RISK MANAGEMENT STRATEGY Solutions Effective risk management typically requires a Implementation Assessment combination of measures, some designed to remove underlying constraints and others designed to address risk directly. The report discusses different approaches for how Operationalizing & the risk solutions identified in the ASRA can be developed Mainstreaming Risk management into action plans incorporated into stakeholders’ strate- gies to reduce agricultural risks. Resource availability will often determine what is possible. Whatever the case, Agricultural Sector Risk Assessment: Methodological Guidance for Practitioners xiii integrated risk management programs are more effective key variables are often very limited, which nega- than stand-alone programs. tively affects the ability to conduct a thorough risk assessment. When governments assume a holistic approach » Subjective bias. Despite all efforts to bring objectivity and improve their ability to assess risks and their and give clear analytic indicators, a broad ASRA interlinkages, they can more effectively help farm- is essentially a subjective tool and this is one of the ers manage their risks by providing targeted informa- more challenging aspects, especially during risk tion and training where they are most needed. Translating prioritization. the integrated risk management strategy into concrete » Risk appetite. Poorer producers have fewer buffers action requires the use of several kinds of implementation and therefore tend to have a low risk appetite. instruments. This report contains tested examples of risk Many of these households manage risks by avoid- management measures and highlights integrated approaches ing productivity-enhancing inputs, preferring yield that target risks at both the micro- and macrolevels. stability over higher-yield inputs that may increase volatility. According to the International Fund ASRA CHALLENGES AND for Agriculture Development (IFAD 2011), “Risk LIMITATIONS avoidance strategies thus have high opportunity costs: some studies estimate that average farm Significant practical and methodological challenges limit incomes could be 10 to 20 percent higher in the the assessment process and its application. The report dis- absence of risk.” cusses approaches for overcoming these challenges; how- » Attribution. It is rare that one event leads to losses ever, one good practice is to be aware of these issues and such that all losses can be attributed to it. In many to be consistent and explicit about the associated assump- cases, multiple risk events occur in a given year, tions. The key challenges include the following: which makes it difficult to attribute proportions to » Historical bias. Assessments conducted to date have different risks. been structured to analyze risk from a historical » Client expectations. An ASRA is an intermediate pro- perspective. This type of assessment provides lim- cess for managing risk, not a final product. Govern- ited opportunity for analyzing new risks that might ment political will and commitment are necessary occur in the future. to translate the suggested solutions into concrete » Data limitations. Accessibility and reliability of government strategies and budgeting. good-quality, disaggregated, time series data on xiv Agriculture Global Practice Discussion Paper CHAPTER ONE INTRODUCTION OVERVIEW This document provides methodological guidance for conducting an ASRA. It is pri- marily designed for use by development practitioners working in the agricultural sec- tor in developing countries. Besides presenting the conceptual underpinnings of agricultural risk management, the document offers practical guidance on the com- plexities of developing a holistic ASRA upon which an integrated risk management strategy can be designed. Rather than presenting a methodological blueprint, this document outlines a practical approach that can be adapted to the particular circum- stances of developing economies. The conceptual framework and step-by-step illustrations presented in this methodo- logical guidance are based on nearly a decade of experience of the World Bank’s Agriculture Risk Management Team in assisting developing economies in conducting sector and key commodity supply chain risk assessments. Whereas other institutions and academic reports offer insights into the development of tools to assess agricultural risks, the guidance herein provides a practical, stepwise approach to conducting ASRAs, incorporating the lessons learned from assisting countries in incorporating risk management practices into mid-term development strategies. This approach was developed and refined over time based on experience, but each specific assessment needs to be adapted to account for a country’s unique combination of risks, institu- tional arrangements, and fiscal constraints. The World Bank’s conceptual framework for ARM looks at risk from an integrated perspective. The framework views risk assessment as a decision support tool for designing strategies to manage risks in agriculture. The ASRA’s primary objective is to assist developing countries and stakeholders to better comprehend the complexity of ARM and to develop appropriate solutions to mitigate, transfer, and cope with agricultural risks. Despite notable progress in this field, much remains to be done to mainstream ARM into development planning and investment decisions. The mainstreaming of ARM and Agricultural Sector Risk Assessment: Methodological Guidance for Practitioners 1 subsequent development of resilient and sustainable agri- The ASRA is devised as a consultative and time-bound cultural systems has been hindered by four main constraints: process to be carried out over an estimated 12-month 1. Lack of an integrated operational approach to period. The assessment draws upon available data and ARM that is embedded in country development qualitative and quantitative information collected through and investment planning; stakeholder interviews and dialogue. This information 2. Lack of expertise and capacity in developing covers input supply from farm production, assembly, pro- countries in this field; cessing, and logistics through to the final consumer. A set 3. Lack of interaction and knowledge exchange by of guidelines is included to facilitate the identification and stakeholders and practitioners, resulting from a characterization of different risks (appendix C) and to previously siloed approach to products, strategies, structure stakeholder exercises (appendix D). and risks; and 4. A tendency to see a constraints-based approach as The rest of this introductory chapter addresses the con- the only strategy to inform investment planning. text and pressing needs for assessing risk in agriculture. Chapter 2 offers a holistic conceptual framework for the The expanding range of experiences across several coun- design of risk management strategies that can be applied tries and agricultural commodities has extended and and adapted according to the specific circumstances of refined the understanding of agricultural risks, their each country. Chapter 3 describes the operational impacts, their transmission along supply chains, and the approach for conducting risk assessments for the agricul- efficacy of different strategies to manage them. These tural sector. Chapters 4 through 8 detail a practical step- guidelines try to capture the lessons learned as a contribu- by-step process to assess risk. Chapter 9 introduces the tion to the process of mainstreaming risk management issue of vulnerability, whereas chapter 10 discusses the practices into policy and investment decisions. Although design of risk management strategies and the mainstream- the literature addressing issues related to risk uses defini- ing of risk management into development-oriented invest- tions of terms that conform to the particular conceptual ments. Chapter 11 offers as final considerations the need and regulatory requirements of their users and intended to pay special attention to operational challenges. purposes, risk can be looked at from various angles and disciplines, and the terminology can be confusing. Appen- dix A contains a glossary of terms used in this document WHY BOTHER ABOUT as they apply to ASRAs. AGRICULTURAL RISK? Agricultural risk is the principal cause of transient food WHAT IS AN ASRA? insecurity and disruption to agricultural supply chains. It An ASRA is simply an orderly process to analyze, identify, creates a poverty trap and causes volatile economic growth and prioritize risk, which serves as the basis for the design in many developing economies. Crop failure is often the of risk management strategies.1 The follow-up of incor- biggest shock faced by rural households and perhaps also poration of the ASRA findings into government policies the biggest poverty trap. Agricultural risks have a pro- and plans is not addressed in this document because every found impact on poverty because they undermine rural country has its own decision-making process influenced entrepreneurs’ (particularly farmers’) possibilities to accu- by the political economy, institutional strength and proce- mulate assets, invest in and develop businesses, and gain dures, and fiscal constraints. access to health and education services. The increasing prevalence of “shock-recovery-shock” cycles vastly reduces many countries’ ability to plan sustainable devel- 1 These methodological guidelines refer to the risk assessment only. The sec- opment paths. This is especially the case in agriculture- ond key step is to develop a full-fledged risk management strategy, which is not dependent countries such as those in Sub-Saharan Africa. addressed in detail in this document. The way in which countries have incor- porated the findings into detailed strategies and budgets has depended on the unique circumstances of each country, given their political economy, institu- Risks are inherent, ubiquitous, and varied in the agricul- tional capacity, and fiscal constraints. tural sector, posing potentially serious consequences for all 2 Agriculture Global Practice Discussion Paper stakeholders. This stems from a range of factors including the vagaries of weather, the unpredictable nature of bio- WHY AN ASRA? logical processes, the pronounced seasonality of produc- An ASRA identifies key vulnerabilities and areas requir- tion and market cycles, the geographic separation of ing priority attention in investment, capacity building, production from end users, and the unique and uncertain and policy reform. The primary objective is to help deci- political economy of the food and agricultural sectors, sion makers understand the risk exposure of agricultural both domestically and internationally (Jaffee, Siegel, and sector stakeholders and to identify risk management strat- Andrews 2008). egies for prioritized risks. It provides a systemwide approach to identify risks, risk exposure, the severity of Agricultural risk is the principal source of agricultural potential losses, and options for risk management by the supply chain disruptions and commodity chain failures. private and public sectors. Ultimately, the ARM frame- Such disruptions have negative impacts that can last sev- work is designed to inform and facilitate integration of a eral years. As a result, many governments introduce policy stronger risk management focus into sector policy plan- measures to protect their population, such as export bans. ning and development. These policies can trigger a ripple effect that adversely The ASRA’s target audience includes country-level stake- affects neighboring countries (possibly reducing grain sup- holders involved in selected agricultural commodity sys- plies and triggering conflict). Ultimately, these protection- tems, development agency decision makers, and ist policies may have adverse consequences for the country developing country policy makers. The ASRA is devised implementing them. to support broader sector strategy formulation efforts and the identification and formulation of proposals for invest- Although levels of risk vary from one country to the next, ment, capacity building, and policy and regulatory reform lower-income and highly agriculture-dependent countries in strategically important agricultural supply chains or are more vulnerable to agriculture-related risks. A press- within the sector as a whole. ing need exists for them to better assess risks, understand where they are most vulnerable, and reinforce their risk An ASRA can add value in a variety of contexts, including (i) management strategies. Highlighting the types and orders broader agricultural sector analyses and development of magnitude of potential risks can improve planning and strategy processes; (ii) constraint and opportunity analyses investments in prioritized strategies critical for mitigating undertaken in the identification and formulation of devel- identified risks, promoting risk transfer mechanisms, and opment projects that focus on value chain integration, making ex ante provisions for coping with realized risks. agricultural commercialization, rural finance, export pro- motion, landscape planning, and so on; (iii) planning, The benefits of a more holistic and targeted risk manage- implementation, and monitoring of sector reform pro- ment strategy are many. Adopting a risk management grams, including those involving shifts in the commercial, strategy can help mitigate risks before they occur. It can regulatory, and other roles of governments in agriculture; also lessen the impact on the government’s fiscal balance, (iv) investment appraisals by private and development particularly at catastrophic levels, by tempering the need finance institutions or as part of strategic assessments of for costly and often poorly targeted post shock humanitar- agricultural lending portfolios’ quality and risk exposure; ian relief to affected communities. Additionally, a well- and (v) as an entry point for the introduction of CSA planned ex ante risk management strategy facilitates aimed at strengthening resiliency in agricultural systems. expedient resource mobilization when risk events occur. For example, ex ante planning can assist farmers in return- The outcome of an ASRA is only a first step. Real changes ing to production more quickly by using previously identi- that incorporate comprehensive risk management into fied delivery channels. Finally, better risk management planning and budgeting will likely be a gradual process. helps poor rural households, women, and other vulnera- Ultimately, integration of risk management practices into ble groups avoid depleting their assets or increasing development strategies will strengthen resilience in the indebtedness in response to shocks. longer term, thus reducing vulnerabilities among Agricultural Sector Risk Assessment: Methodological Guidance for Practitioners 3 agricultural sector stakeholders and increasing the success attracted the attention of policy makers and the donor of agricultural investment strategies. community by elevating the need to invest in agricultural systems and implement risk management systems to ASRAs can complement other types of risk assessments, reduce volatility and protect investments. The develop- including (i) household or area-based risk assessments, typ- ment community now recognizes the need to bridge the ically focused on the vulnerability of different types of development-relief divide and to shift the focus from sim- households (that is, women in disadvantaged positions), the ply increasing productivity and food aid effectiveness application of (typically) informal risk sharing and coping toward a holistic and integrated risk management or resil- mechanisms, and the need and scope for supplementary ience-building approach. Food security concerns, often social protection measures; (ii) hazard vulnerability assess- driven by underlying risks, are currently at the top of ments intended to justify the role and design of agricul- many countries’ political agendas and discussions on agri- tural insurance instruments for specific segments and levels cultural risks are central to the Group of Twenty (G-20) of stakeholders exposed to production risks; and (iii) finan- and other multilateral and sovereign policy deliberations. cial risk assessments focused on the possible budgetary and other macroeconomic impacts of major shocks. The majority of the world’s projected 9 billion people by 2050 will be located in developing countries and will be disproportionately affected by climate change. To feed VOLATILITY AS THE “NEW this growing population, policy needs to shift from design- NORMAL” ing stand-alone risk products to adopting holistic and inte- grated risk management approaches with an emphasis on During the past 20 years, new macroeconomic forces have increasing sustainability and resiliency. The ARM frame- changed the shape of many agricultural risks, not only for work must move beyond the individual farmer and seek to food crops but also for agricultural supply chains more build capacity in the design of integrated risk manage- broadly. In light of long-term trends such as globalization ment strategies for farmers, supply chain stakeholders, and climate change, managing agricultural risk is ever more and the agricultural sector as a whole. These strategies crucial to shortening the shock-recovery-shock cycle. Cli- need to be incorporated into government plans to rein- mate change is increasing vulnerability and contributing to force the process of managing the “new normal” of a protracted crises, whereas globalization is connecting the world with higher volatility in the agricultural sector. world as never before through trade, financial markets, and politics. The global economy has become a more compli- In the face of multiple risks, a critical consideration is cated space whereby, for example, shifts in U.S. grain pro- the resilience of primary producers, agribusiness enti- duction can affect Niger, and economic growth in China ties, institutions, and other supply chain stakeholders for can increase global demand for high-protein agricultural collective action, coordination, and public-private coop- products. Some analysts observe that the changing funda- eration. One cannot understand the competitiveness and future mentals of demand for and supply of agricultural com- potential of a sector without understanding players’ ability to modities have led to the creation of thin commodity markets anticipate and respond to shocks. An agricultural sector develop- that are intrinsically more vulnerable to adverse events. ment strategy that focuses solely on productivity and removal of The foregoing trends are likely to aggravate the current lev- constraints is incomplete if it ignores risk and risk management els of volatility and lead to further uncertainty. considerations. The 2007–08 global food price crisis was an inflection point for attention to global food security. Following the crisis, interest has renewed in agriculture as a key driver of CLIMATE CHANGE AND development and poverty reduction for many developing AGRICULTURAL RISK economies. Furthermore, shocks including recent The long-term changing climate is a significant source of droughts in Australia, Russia, and the Horn of Africa; risk to agricultural and food systems, beyond the short- floods in Pakistan; and several food safety failures have term weather risks addressed in this methodological 4 Agriculture Global Practice Discussion Paper guidance. Scientists predict that climate change will alter shocks that disrupt their livelihoods and encircle them into climate conditions, with increased weather and climate recurrent poverty traps. Mitigation of agricultural risk is variability expected for most locations. Climate projec- increasingly focused on reducing the exposure and increas- tions suggest that impacts will include shifts in average ing the resilience of production systems and livelihoods to growing conditions, increased rain and temperature vari- adverse impacts.2 The ASRA therefore largely seeks to ability, and greater uncertainty in predicting future cli- understand which characteristics make a country, agricul- mate and weather conditions. More concretely, these tural sector, or segment of producers resilient and to iden- impacts will translate into an overall warming trend, an tify the specific measures and processes that strengthen increasingly erratic distribution of precipitation, more fre- resilience. Resilience can be strengthened in many differ- quent and more extreme weather events, and spatial shifts ent ways and at different levels through political, economic, in pests and disease outbreaks. Climate change thus sociological, and technological interventions. For example, requires adjusting both to new average climatic conditions drought can be countered by building irrigation systems, and preparing for more volatile weather with more fre- improving water-harvesting techniques, and using agro- quent and intense extreme events in most locations. The ecological technologies such as conservation farming, and remaining uncertainty over future long-term climate by breeding new drought-resistant and drought-tolerant change will lead to more short-term weather volatility crops and livestock. Open trade policies to facilitate trans- overall. Combined, these effects form the “new normal” border access to food can also strengthen resilience. to which all stakeholders need to adapt. One key strategy for mainstreaming resilience into farm- The possibility of the appearance of new types of extreme ing systems is the concept and practice of CSA. CSA is an events will pose further challenges to the agricultural sec- approach for transforming and reorienting agricultural tor. For example, it is predicted that cropping systems will systems to support food security under the new realities of no longer be viable in many locations. In Africa, for climate change. It aims to achieve three simultaneous out- instance, under a range of scenarios projected to 2050, 35 comes: (i) increased productivity; (ii) enhanced resilience; million farmers across 3 percent of the continent’s land and (iii) reduced emissions. A wide range of practices and area are anticipated to switch from mixed crop-livestock approaches can increase the “climate-smartness” of pro- systems to livestock only (Jones and Thornton 2008). duction, from agro-forestry to rangeland management to climate and weather information services. Although climate change is expected to produce both winners and losers overall, losses will far outweigh the ASRA promises to serve as a methodological guidance to gains, and the poor will be disproportionately affected identify and justify strategies that strengthen the resilience because of their dependence on agriculture and a lower of agricultural systems aimed at adapting agriculture to capacity to adapt (World Bank 2008). Yet stakeholders climate change conditions (figure 1.1). The process can be must develop the capacity to identify thresholds triggering met by reconciling the short-term objectives of risk man- potential new hazards and anticipate which novel extreme agement with the medium- to longer-term objectives of events may arise in order to help prepare farmers as well climate change adaptation. Early incorporation of CSA as national and regional systems in dealing with the risks into farming system practices also contributes to the associated with the new normal. objectives of increasing productivity, managing volatility, and boosting overall farming system resilience. An ASRA RESILIENCE OF can serve as a key entry point for operationalizing CSA strategies by providing the analytical and operational tools to strengthen resilience AGRICULTURAL SYSTEMS in agricultural systems. AND CLIMATE-SMART AGRICULTURE 2 Resilience can be defined as “the ability to withstand, recover from, and reorga- The vast majority of producers in developing countries are nize in response to crisis so that all members of society may develop or maintain smallholders who are particularly vulnerable to external the ability to thrive” (World Bank 2012). Agricultural Sector Risk Assessment: Methodological Guidance for Practitioners 5 FIGURE 1.1. CONTRIBUTION OF ASRA TO CLIMATE CHANGE ADAPTATION What ARM can contribute to meeting the climate challenge: 1 A proven tool for building resilience to climate and weather Climate volatility Change 2 A key entry point for the operationalization of climate-smart agriculture where resilience is first priority Implications of climate change for ARM: Agriculture 3 Increasing risks = increasing importance of ARM Risk 4 Need to adapt frameworks and approaches: Management a. Incorporation of climate projections b. Decision making under uncertainty & capacity building to meet the unknown 6 Agriculture Global Practice Discussion Paper CHAPTER TWO AGRICULTURAL RISK MANAGEMENT CONCEPTUAL FRAMEWORK This conceptual framework for ARM looks at risk from a holistic perspective and views the ASRA process as a decision support tool for designing integrated ARM strat- egies. The primary objective is to help decision makers understand the risk exposure of agricultural sector stakeholders and to provide the basis for developing appropriate solutions to mitigate, transfer, and cope with agricultural risks. An expanding range of experience across several countries and agricultural commod- ities has extended and refined the ARMT’s understanding of agricultural risks, their impacts, and their transmission across the sector as well as the efficacy of different strategies to manage risks. Distilling the experience and lessons learned in more than 20 countries, the World Bank developed the conceptual framework presented herein for assessing agricultural risks with an emphasis on practical application. This methodology for assessing agricultural risks is based on a decade of lessons learned by the World Bank’s ARMT, which has conducted numerous sector risk assess- ments in Niger, Honduras, Belize, Grenada, Paraguay, Ghana, Rwanda, Malawi, Kenya, Senegal, Tanzania, and Mozambique. Likewise, valuable lessons were learned from risk assessments for specific agricultural supply chains in Ghana (cocoa), Kenya (horticulture), Mozambique (cotton), Vietnam (coffee), Guyana (rice), Haiti (coffee), and Uganda (dairy and coffee), among others. (See appendix B for a complete list of risk assessments completed by the World Bank.) THE CONCEPTUAL FRAMEWORK FOR ARM A holistic perspective needs to (i) engage all stakeholders who are affected by agricul- tural risks and who take actions to manage them; (ii) consider the full range of risks; (iii) analyze different strategies to manage risk; and (iv) understand the different steps in the risk management process (as summarized in figure 2.1). An ASRA typically involves prioritizing the major risks affecting various stakeholders. The risk prioritiza- tion subsequently informs identification of specific risk management interventions and strategies to be implemented in the short to medium term. Agricultural Sector Risk Assessment: Methodological Guidance for Practitioners 7 FIGURE 2.1. KEY COMPONENTS OF AGRICULTURAL RISK MANAGEMENT TASKS RISKS STAKEHOLDERS STRATEGIES INTERVENTIONS • Identify • Production • Producers • Mitigation • Policy reform • Assess • Market • Comercial sector • Transfer • Investments • Manage • Enabling • Public sector • Coping • Technical environment assistance As these components are not linear, they are better illus- RISKS VERSUS CONSTRAINTS trated in a more dynamic fashion, as presented in AND TRENDS figure 2.2, wherein the cycle of tasks (Identification, Conventional analyses of agricultural value chains typi- Assessment, and Management) interacts with the other cally emphasize efficiency and productivity, often through components and flows around the needs of various stake- identification and removal of “critical constraints” or bot- holders at the center of the process. tlenecks. This static focus ignores volatility in both pro- duction and markets and its effect on supply chain Risk management is a process that needs to be reevalu- performance; it also does not address the incidence, allo- ated periodically, as underlying conditions (for example, cation, or implications of risk or how incentives to add risk dynamics, levels of vulnerability) change over time. value among different actors (that is, smallholders and Conducting periodic risk assessments can provide valua- small and medium enterprises [SMEs]) might explicitly ble snapshots that assist in the design of new instruments alter risk patterns. Whereas the definition of risk implies and refinement of existing interventions. the presence of uncertainty and the probability of losses (for example, from a pest outbreak), the definition of con- The following sections briefly describe the key compo- straint implies a certain factor known to cause suboptimal nents of the ARM framework, which make up the basis performance in agriculture (for example, poor soil) for addressing the operational rationale behind an ASRA, (see box 2.1). as explained in chapters 3 through 8. TYPES OF RISK Risks faced by agricultural stakeholders can be classified FIGURE 2.2. THE CONCEPTUAL primarily into three categories: production, market, FRAMEWORK FOR ARM and enabling environment risks. Depending on the Instruments market integration of any particular supply chain and its Investments Technical assistance context, each type of risk can be present, dominant, or Policy absent. Each can also affect unique segments of the sup- ply chain or the entire chain. 1. Production risks: A large number of risks Stakeholders affect the volume of production of agricultural Producers commodities. These include nonextreme weather Commercial sector Risks Strategies events (too little rainfall, too much rainfall, hail, Public sector Production Mitigate frost, low temperature) and less frequent but cata- Market Transfer strophic weather events (severe floods, droughts, Enabling environment Cope hurricanes, cyclones). Outbreaks of pests and diseases can adversely affect yields, as can dam- age by animals, fire, and wind. Human-induced 8 Agriculture Global Practice Discussion Paper BOX 2.1. DISTINGUISHING BETWEEN RISKS, constantly and are rarely associated with only CONSTRAINTS, AND TRENDS one specific geographic location.3 Aspects of market risk may directly affect individual actors It is important to differentiate risks from constraints in a supply chain and may affect producers in a and trends. single community and producer group in differ- Risks are uncertain events that have the probability to ent ways. cause losses. The element of uncertainty is present. As a 3. Enabling environment risks: Sudden changes symptom, yield volatility might be caused by a drought or in the given scenarios in which business takes pests or a disease outbreak. place such as unexpected changes in government Constraints are conditions that lead to suboptimal per- or business regulations, the macroeconomic envi- formance. For example, low yield (symptom) might be ronment, political risks, conflict, trade restrictions, caused by lack of access to inputs or poor technology. The logistics, and corruption are all major enabling element of certainty is present. environment risks that can lead to financial losses Trends are longer-term or “chronic” patterns (reversible for agricultural stakeholders. These risks have or irreversible) that provide context. For example, declin- systemic impacts on decision making, productiv- ing yield (symptom) might be caused by structural changes ity, and market options. Because incentives can in agriculture or changes in climatic patterns (for example, change (including the distribution of rewards and desertification). risks in the supply chain), these risks can result in changes in yield quantity and quality and can even lead to disruptions in the flow of goods, services, problems such as theft, fraud, and arson are also information, and finance. notable. These risks are usually manifested at the farm level but often affect the entire supply chain CATEGORIES OF STAKEHOLDERS via risk transmission. Production risks are mostly Stakeholders affected by agricultural risks can be classified associated with yield reductions but can also affect into the following three major categories. A fourth cate- product quality (especially hail and wind damage gory, consumers, is also affected by agricultural risks, but and high humidity and excess rain leading to pests addressing demand-side issues directly is outside the scope and diseases) and disrupt the flow of goods and of this assessment. services. 1. Producers: Operators of marginal, small, and 2. Market risks: Market risks are related to medium-size farms make up the vast majority issues that affect price, quality, availability, and of agricultural producers in developing coun- access to necessary products and services. Prices tries. They are most often severely exposed to for inputs and outputs can be highly volatile, risks and their impacts. Their capacity to man- particularly in commodity markets where both age risk is limited and they are often decapital- local and global supply and demand conditions ized by catastrophic events. Because of this, they are constantly changing. Other market risks are the leading recipients of direct and tempo- include exchange rate and interest rate volatility rary government support. Larger-scale opera- and counterparty and default risks. These risks tors in agriculture run more technically intensive usually materialize at the market level but have and integrated farming systems and have higher backward linkages to the farm, thereby affecting capacity to manage production risks but are many different stakeholders. Directly related to highly exposed to market risks. price risks are risks associated with quality. Qual- ity is affected by availability of affordable inputs, 3 Smallholder farmers typically face a systemic market risk in that their most delivered and applied in a timely fashion, and accessible (localized) markets may be characterized by lack of access to informa- by decisions about production, postharvest, and tion, poor transport and storage facilities, and low numbers of regularly active processing practices. Market-related risks vary buyers. Agricultural Sector Risk Assessment: Methodological Guidance for Practitioners 9 2. Commercial sector stakeholders: Commer- losses. They are particularly useful for risks that cial stakeholders, including agribusinesses, trad- occur with relatively high frequency but with ers, wholesalers and retailers, financial institutions, lower impact intensity. Risk mitigation options are and input providers, have a vested interest in bet- numerous and varied. Examples include adop- ter ARM. This segment of participants is exposed tion of improved agronomic practices such as soil to the various risks sourced at any point in the sup- drainage and mulching, conservation farming, ply chain, but their roles and positions afford them and the use of short-duration and disease- and a much higher capacity to protect their interests, stress-resistant cultivars; irrigation and flood con- manage risks, and smooth income over time. trol infrastructure; soil and water conservation 3. Public sector: Public sector institutions, para- measures; changes in cropping patterns; crop and statals, government, and donors are important livestock diversification; income diversification; stakeholders that are fiscally exposed to agricul- improved early warning systems; and modern tural risks and are often the main actors support- information and decision support systems. ing risk management activities, and who also take 2. Risk transfer (ex ante): As not all effects of into account impacts on consumers. At times, pol- realized risks can be mitigated, risk transfer tools icy decisions made by these stakeholders may also and mechanisms transfer the potential financial be a source of risk. consequences of particular risks from one party to a willing third party, usually for a fee or premium. These mechanisms usually trigger compensation INTEGRATED RISK MANAGEMENT in the case of a risk-generated loss (for example, LAYERS purchasing insurance, reinsurance, financial The vulnerability of individual stakeholders and the agri- hedging tools). Although insurance and hedging cultural sector as a whole depends on the nature of the are well-known forms of risk transfer, in devel- risks (that is, their correlation, frequency, timing, and oping countries the use of informal risk transfer severity) and the effectiveness of the risk management within families and communities is also extremely instruments in use. It is unrealistic to suppose that all risks important. can be managed, as one solution or product cannot serve 3. Risk coping (ex post): Some risks cannot be as a “silver bullet” for all risks in all circumstances. Indeed, mitigated or transferred, so risk coping strategies the existing literature discusses a complex variety of strat- are needed to help stakeholders better absorb and egies. Rather than reviewing all possible risk management recover from their impacts. These instruments strategies, the strategic framework used in these guidelines improve the affected population’s resilience to presents a simplified approach for risk management strat- withstand and cope with events through ex ante egy development for illustrative and practical purposes. preparation to sustain production and livelihoods The conceptual framework can be adapted to be as com- following an event. Examples include some form plex or as simple as needed, given a country’s circum- of compensation (cash or in-kind), social safety net stances. programs, buffer funds, savings, strategic reserves, Following the assessment of risks and analysis of stake- and livelihood recovery programs (for example, holder vulnerability, risk management strategies can be government assistance to farmers, debt restruc- proposed. A practical way to identify solutions is by classi- turing, contingent financing). Such interventions fying possible risk management strategies into three cate- are often financially beneficial and the ability to gories: mitigation, risk transfer, and coping. The appropriate quickly respond to events often reduces losses. set of strategies depends in part on participants’ capacity to effectively use them. Figure 2.3 illustrates these risk management strategies in 1. Risk mitigation (ex ante): Risk mitigation the context of increasing layers of risk depending on the strategies are actions taken prior to a risk event probability of occurrence (frequency) and the intensity or to reduce the likelihood of risk or the severity of potential to cause losses (severity). 10 Agriculture Global Practice Discussion Paper FIGURE 2.3. RISK MANAGEMENT LAYERS OUTPUT AND PRIORITY MEASURES Probability LAYER 3 very low frequency, The end product (output) of the ASRA process is a set of very high losses LAYER 2 actions that require special attention to reduce vulnerabil- Risk mitigation low frequency, medium losses + risk transfer ity to shocks associated with the key priority risks. More- + risk coping LAYER 1 Risk mitigation over, resources are scarce and stakeholders need to see an high frequency, + risk transfer explicit risk prioritization and strategy identification pro- low losses cess whereby the proposed interventions provide clear Risk mitigation returns in terms of agricultural growth, poverty reduc- tion, food security, or other agricultural policy objectives. The measures can then be incorporated into government Severity plans and budgets. Risk management strategies are operationalized by instru- The combination of different activities selected to man- ments that can be planned, budgeted, and implemented age risk ultimately depends largely on the findings of the for. These instruments fall into three main categories: ASRA, the characteristics of the identified risks, various 1. Policy reform: Improved risk management actors’ existing capacity to manage risk, and the fiscal often entails policy reforms (for example, legal constraints to implementing an integrated strategy. This or regulatory reforms to improve access to agri- framework can be applied to prioritize risks and interven- cultural inputs; changes in information policy to tions in a country with many risk management practices make agricultural information easily accessible to already in place, whereby stakeholders identify priorities all; changes in government policy related to price and gaps in their current risk management strategies to formation, government procurement, or strategic enable them to adapt to a changing risk landscape. grain reserves). 2. Agricultural investment: Although policy reforms mainly require political will, other risk RISK MANAGEMENT mitigation measures can be costly. Examples are STRATEGY financial investments in irrigation infrastructure, The findings of an ASRA ultimately need to be incorpo- research into drought- and disease-resistant rated into an actual plan of action with a sequencing of and pest-tolerant cultivars, soil and water con- activities and a budget financed over the medium term. In servation, weather infrastructure, or updated this sense, the process of risk assessment has to be devel- agricultural services (for example, agricultural oped with clearly designated leaders who not only partici- extension systems or disease surveillance sys- pate in the process but also find it useful to incorporate tems). Some of these measures may already be risk management strategies into medium-term planning. part of a government program, with the ASRA Typical stakeholders interested in the outcomes of the simply calling for additional investments to ASRA process are the Ministry of Finance, the Ministry strengthen capacity in those areas more vulner- of Agriculture, and other organizations related to the able to external shocks. agricultural sector. The public sector is the most appropri- 3. Technical assistance (TA): TA is geared ate leader given the key role that public expenditures toward building local stakeholders’ capacity (for (investments and current expenses) play in ARM. The jus- example, training in price risk management; fea- tification for public sector interventions is further sup- sibility studies for various instruments; flood risk ported by the public good nature of risk management, modeling work; development of early warning particularly in countries where the vast majority of pro- systems). Recent developments in information ducers are small to medium-size farmers in vulnerable systems addressing agricultural risks can be easily positions. transferred to public and private institutions that Agricultural Sector Risk Assessment: Methodological Guidance for Practitioners 11 can adapt the instruments to a country’s specific Stakeholders’ underlying objectives and their specific conditions. motivations for participating in the focal dialogue vary, cutting across commercial, personal, political, economic development, and even humanitarian concerns and con- SHARED RESPONSIBILITIES siderations. Some goals, perspectives, and expectations The ASRA methodology described in subsequent chap- may be shared; others may not. It should not be assumed ters calls for early and active participation of public and that a common set of goals exists at the start of the private sector representatives. Risk management is not the assessment process. Indeed, a key objective of the pro- sole responsibility of the government or a group of par- cess is to build a greater degree of common understand- ticipants in the sector. Moreover, it is not the sole responsi- ing as well as a commitment toward shared goals. The bility of the Ministry of Agriculture because risks can result should be a set of practical risk management originate at any point along a commodity supply chain and measures that stakeholders agree will result in a more can be transmitted to other participants in agribusiness, resilient sector. Early consultations and involvement of the financial sector, or industry. In turn, risks can affect government representatives (especially from the Minis- stakeholders in different ways. For example, food proces- try of Finance) and interested donor agencies will sors and consumers could perceive sudden price hikes enhance ownership of the process and facilitate incorpo- negatively, but producers could view them as good news. ration of the recommendations into government plans. All participants thus have a stake in ensuring that their risk Stakeholders are also more willing to adopt new instru- perception, capacity to manage, and vested interests are ments and strategies if they are involved in the process. taken into account and balanced. This is of utmost impor- Once stakeholders understand and accept the value tance in agriculture, where the political economy of food added of the ASRA process, the proposed strategies to production usually plays a key role in policy making. A manage risk will be more easily incorporated into their well-informed ASRA with broad participation by key agri- working plans. cultural sector players can go a long way in the final design of acceptable and practical risk management measures, Agricultural stakeholders will have incorporated risk with results that can be monitored over time. management activities within their own area of influence to protect their financial interest against shocks. It is there- COLLABORATION AND JOINT fore of crucial importance for the ASRA to (i) understand OWNERSHIP very well the institutional and socioeconomic context of In facilitating the dialogue and the identification of risks every group of stakeholders; (ii) identify and assess current and priority measures, the ASRA takes into account the risk management practices and their strengths and limita- divergent goals and motivations of different stakeholders. tions; (iii) assess public and private sector capacity to man- In so doing, it advances individual stakeholders’ acceptance age risk; and (iv) maximize consensus among stakeholders of the analysis, recommendations, and other outcomes that the ASRA’s recommendations will strengthen current regardless of their motivations. risk management practices. 12 Agriculture Global Practice Discussion Paper CHAPTER THREE OPERATIONAL APPROACH: HOW TO CONDUCT AN ASRA Whereas the previous chapters explained the context, justification, and conceptual framework for conducting an ASRA, the remainder of the document focuses on the operational steps to apply the framework in practice. As mentioned earlier, rather than presenting a methodological blueprint, this document outlines a practical approach that can be adapted to the particular circumstances of developing economies. The ASRA may be tailored to take into account the sector’s structure and institutional dynamics, including factors such as supply chains, institutions, fiscal constraints, and individuals’ capacity to deal with the effects of realized risks. Likewise, risk manage- ment solutions identified in a particular country will not necessarily fit the context of another country, however similar they may appear. WHAT IS THE OPERATIONAL APPROACH? A great deal of literature exists on the subject of risk management. No less extensive are variations in the understanding and use of terminology. For the purpose of these guidelines, the ASRA is considered the process of identifying and prioritizing the major risks that explain agricultural gross domestic product (GDP) volatility. Once those risks are identified and assessed, the following issue is determining how stake- holders manage those risks. This process involves identifying optimal risk management solutions to be incorporated into sector risk management strategies led by the public sector. Risk analysis using a holistic framework can be very complex and can involve sev- eral stakeholder groups. Stakeholder participation is encouraged through a process of facilitation. Given the complexities of agricultural production, processing, and commercialization, there is no shortage of risks and potential solutions to manage them. Moreover, effective risk management generally requires close cooperation Agricultural Sector Risk Assessment: Methodological Guidance for Practitioners 13 between various sector actors, each with their own par- analyze the performance of agricultural sup- ticular vested interests, in complex political economy ply chains usually contain valuable information scenarios. The facilitation process requires risk assess- regarding efficiency and productivity constraints, ment teams to assume a neutral position and play the which is also helpful in identifying the underly- role of “honest broker” among parties, with the explicit ing risks. mandate to answer the following questions: 3. Stakeholder interviews. This important technique 1. What are the major agricultural risks? assesses various stakeholders’ capacity to manage 2. What are the optimal solutions to manage the key risks, their responses to shocks, and their degree risks? of vulnerability in the presence of risk. Interviews 3. What are the gaps in current risk management can identify the causality of risk and risk trans- strategies and plans? mission along supply chains, as different actors 4. What is the action plan to strengthen resilience to perceive and react differently to risk. Addition- shocks in agriculture? ally, interviews with key informants who have extensive experience in analyzing the agricultural The risk assessment team establishes the methodology sector and have gained a reputation among stake- and facilitates discussion among various stakeholders to holders are a valuable way to build a realistic risk answer those questions. The use of quantitative data is storyline. critical to inform and bring a level of objectivity to the 4. Qualitative analysis. As not all risks can be assessed process. using quantitative techniques, the use of qualita- tive analysis needs to be introduced. This is partic- The ASRA approach relies primarily on analysis of time ularly true when comparing the importance and series data, supported by other qualitative techniques for ranking of nonmeasurable risks such as enabling assessing risk. Techniques used in various ASRAs include environment risks. Team members participating the following: in the ASRA use their best judgment during the 1. Time series data analysis. This technique is pre- prioritization process. ferred for assessing production risks. It involves 5. Workshops and focus group discussions. Group discus- statistical analysis of annual performance indi- sions with stakeholders at different levels and cators for major agricultural commodities, stages of the ASRA process are particularly help- namely, changes in area, yield, and production. ful for securing feedback and validating findings. Most ministries of agriculture keep historical Additionally, group discussions help to neutralize records of these indicators. Ideally, it is optimal those with potential vested interests who might try to obtain data disaggregated by region to isolate to direct attention to particular risks or solutions and analyze regions with major volatility and that suit their convenience. production losses. In the absence of time series data, production analysis can rely on yearly Experience shows that all of these techniques are needed reports written by government agencies, com- to some degree at different moments during the ASRA modity boards, producers’ associations, and the process. The process involves a blend of science and art, private sector. whereby quantitative analysis is combined with other ana- 2. Secondary data analysis. An early literature review lytical methods and fact-finding techniques to facilitate of peer-reviewed publications by academic insti- the prioritization process. Any degree of subjectivity can tutions, social scientists, and agriculture special- be put into context through discussions with stakeholders ists and an Internet search of media reports can as well as group discussions within the ASRA. It is impor- help build the storyline related to the occurrence tant that underlying assumptions related to subjective of risks, risk exposure, and vulnerability of vari- value judgments are clearly articulated so that their merits ous communities. Recent analytical reports that can be openly debated. 14 Agriculture Global Practice Discussion Paper current interventions and make the agricultural sector THE SEQUENTIAL FLOW more resilient to shocks. PROCESS The ASRA process proposed herein is a dynamic one that requires careful planning given the participation of a vari- WORK PLAN FOR THE ASRA ety of stakeholders and the degree of analytical skills and As part of the planning process, it has proved to be practi- experience desired for the assessment team. Figure 3.1 cal to elaborate a calendar of activities shared with coun- summarizes the step-by-step ASRA process. terparts to make them aware of when and for what purpose their participation will be needed. The more This sequential flow process has proved to be very practi- engaged the local counterparts, the greater the degree of cal in conducting ASRAs to date, but it should be adapted buy-in for the methodology and ultimately of the results. to the particular circumstances of the country where it is Figure 3.2 illustrates a calendar of activities used in an applied and should respond to the assessment’s specific ASRA in Tanzania. This calendar is not a blueprint but objectives. The step-by-step activities shown in this dia- can be used as a reference. gram are sequential; the findings of each step inform and serve as the basis for the next. In this illustration, the risk team listed four detailed steps that were jointly monitored between the assessment team and the The ASRA’s outcome is a short list of prioritized government of Tanzania. Note that the calendar has a third risks and corresponding solutions. This outcome is step called “Solution Assessment” that involves developing the basis for designing a risk management strategy that in the details on how to address the major risks prioritized ear- turn has various steps, ending in an action plan for lier in the ASRA. It is the responsibility of government implementation. The action plan details the interventions authorities and their partners to incorporate the action plan that stakeholders need to implement to complement in government strategies and development projects, as well as to monitor and evaluate implementation. FIGURE 3.1. THE SEQUENTIAL ASRA FLOW Chapters 4 and 5 provide step-by-step guidance for assess- PROCESS ment teams to conduct a risk assessment and a solutions assessment. Chapter 10 provides less standardized guid- ance for incorporating action plans into government strat- egies and programs given the unique nature of institutional decision making in each country. Risk Assessment / Risk Monitoring Reevaluation WHAT ANALYSIS AND WHEN? Figure 3.3 shows the various pieces of analysis included in Solutions the ASRA. The process starts with sectorwide analysis Implementation Assessment during the background research phase (step 1). The analy- sis is then disaggregated at the supply chain commodity level to allow enough granularity for the risk prioritization Operationalizing & process (step 4). Finally, the various commodity analyses Mainstreaming Risk management are aggregated to conclude a risk prioritization and solu- tion assessment that is representative of the entire agricul- tural sector (step 5). Agricultural Sector Risk Assessment: Methodological Guidance for Practitioners 15 FIGURE 3.2. CALENDAR OF ACTIVITIES FOR AN ASRA IN TANZANIA Tanzania: Risk Management Review Workplan 2012 2013 Nov Dec Jan Feb Mar Apr May June July Aug Sept Oct Nov Dec Step I: Inception Mission (1 week in-country mission) Activity 1.1: Meeting the government counterpart and explaining the objective Activity 1.2: Meeting other relevant stakeholders and soliciting their buy-in Activity 1.3: Identifying local partner institution and consultants Activity 1.4: Preliminary data collection Step II: Risk Assessment (7 week preparation, 3 weeks in-country mission, 6 weeks write-up) Activity 2.1: Background data collection Activity 2.2: Analysis of background data Activity 2.3: Team selection and contracting Activity 2.4: Risk assessment mission planning Activity 2.5: Risk assessment mission Activity 2.6: Risk assessment wrap-up workshop and sharing preliminary findings Activity 2.7: Report writing Activity 2.8: Sharing preliminary report with clients and relevant stakeholders Step III: Solution Assessment (2 week in-country mission, 4 week preparation and 8 week write-up) Activity 3.1: Background data collection Activity 3.2: Collection and classification of risk management intervention inventory in the country Activity 3.3: Analysis of efficacy and relevance of different RM interventions in the country Activity 3.4: Team selection and contracting (solutions experts) Activity 3.5: Solution assessment mission planning Activity 3.6: Solution assessment mission Activity 3.7: Risk Management Road Map workshop (2 days) Activity 3.8: Development of risk management implementation plan Activity 3.9: Report writing Activity 3.10: Sharing solution assessment and RM implementation plan with the clients Step IV: Dissemination Activity 4.1: Printing of Agriculture Sector Risk Assessment document Activity 4.2: Printing of Solution Assessment document Activity 4.3: Disseminating the document with the government & partner institutions Activity 4.4: Disseminating the document online (FARMD, USAID website, ARMT website, etc.) Activity 4.5: Disseminating the findings in relevant forums/meetings FIGURE 3.3. ANALYTICAL COMPONENTS AT COMMODITY AND AGGREGATE LEVELS DISSAGGREGATED COMMODITY LEVEL Commodity 1 Commodity 2 AGGREGATED Commodity 3 AGGREGATED SECTOR LEVEL SECTOR LEVEL Commodity 4 Commodity 5 Commodity 6 Commodity 7 Commodity 8 ANALYSIS ANALYSIS ANALYSIS • Sector • Risk identification • Risk prioritization • Commodities • Risk quantification • Long list of solutions • Policies & • Timeline • Solutions filtering strategies • Short list of • Risk prioritization • Risk Landscape solutions • Capacity to manage • Risk Identification • List of Solutions • Timeline 16 Agriculture Global Practice Discussion Paper CHAPTER FOUR STEP 1: BACKGROUND RESEARCH AND RISK IDENTIFICATION The first step of the ASRA process is to gather background information on the agricultural sector, collect relevant data for risk analysis, and conduct a desk-level anal- ysis before doing fieldwork. The objectives of this step are the following: » Understand the context by analyzing current and past agricultural policies and strategies. » Understand the dynamics of the political economy, when possible. » Identify the structure of the agricultural sector and the role it plays within the broader economy (for example, size of sector, share in trade, labor in agricul- ture, rural poverty). » Analyze trends in time series data for the various commodities related to area, yield, and production. » Quantify and analyze variations in yield by commodity and region, noting years that show large drops in yields. » Analyze time series rainfall data to determine trends and years of extreme volatility. » Match historical drops in yield with realized risks, when possible. » Identify major risks by crop, region, and stakeholder group. » Analyze levels of vulnerability among various stakeholder groups. » Analyze trends and variations in time series data of prices for various agricul- tural commodities and inputs. » Analyze trends and levels of volatility in the foreign exchange rate and interest rates. » Identify and predict likely impacts from climate change related to projected changes in weather and temperature. By conducting this background analysis before the fieldwork, the team can identify in advance the sector’s risk landscape and its disaggregation by commodity and by region. This exercise allows the team to become familiar with the country context, the broader economy, and more particularly, the agricultural sector. The findings of the background Agricultural Sector Risk Assessment: Methodological Guidance for Practitioners 17 research help guide the team as it proceeds with risk quan- tification, explained in detail in chapter 5. The qualitative UNDERSTANDING THE analysis is conducted as part of the fieldwork, explained in AGRICULTURAL SECTOR detail in chapter 6. To situate risks in their context, the first stage is to form a better understanding of the agricultural sector, including SOURCES OF INFORMATION its role in the economy, the importance of key agricul- Sources of information are numerous and vary from one tural commodities, the organizational structure of associ- country to the next. Government agencies (Ministry of ated supply chains, key stakeholders and their functions, Finance, Ministry of Agriculture, and other agriculture- performance over time, future outlook, and the enabling related agencies, including agricultural commodity environment in which they are embedded. It has proved boards) can provide a great deal of the relevant data and to be practical to select as a proxy a mixed bag of agricul- documents needed for the background research. Addi- tural commodities that together represent at least tional sources of information include the following: (i) 80 percent of agricultural GDP. This provides the assess- websites of various international agencies involved in ment team with a handful of commodities that are easier analyzing the agricultural sector, with a focus on macro- to handle and that are also representative of the sector in or sectorwide policies and constraints, value chain analy- terms of value. Finally, it is important to understand the sis, regional or territorial development issues, or specific agricultural sector’s relative importance in the economy selected topics (for example, agricultural insurance, (that is, as a share of GDP, share of trade, contribution to grain reserves, food security, commodity exchanges); (ii) employment, levels of poverty and food insecurity in universities and similar analytical institutions that pub- rural areas). lish the findings of research conducted on the agricul- tural sector, in particular the Consultative Group for Preliminary analysis of agricultural supply chains often International Agricultural Research (CGIAR); (iii) entails a comprehensive review of existing material reports or analytical documents from commodity asso- (reports of agricultural sector reviews, studies, research ciations; (iv) the World Bank, which has provided useful theses, major news items, trade reports, and so on). Issues information since 2009 when it changed its communica- to be thoroughly researched include the following: tion policy toward full disclosure, making available to the » Demand conditions. Identification of major markets public analyses at country, sector, and commodity-spe- of the commodity in question (export, regional, cific levels, as well as on agriculture policy–related issues. and domestic), end use of the commodity, char- Similarly, regional development banks such as the Inter- acteristics of the commodity, quality attributes, American Development Bank (IADB), Asian Develop- demand and supply dynamics, and current and ment Bank, and African Development Bank, are emerging market trends. excellent sources of analytical information; (v) the Food » Supply chain structures. The current structure of and Agriculture Organization Corporate Statistical supply chains, which includes the flow of goods, Database (FAOSTAT), which has been closing the gap information, and finance, and the degree of con- between its time series datasets and those published by centration of different operators (in terms of vol- ministries of agriculture. Additionally, the Food and ume or value). Agriculture Organization (FAO) of the UN actively ana- » Stakeholders and their functions. A brief description of lyzes a wide range of agricultural issues at all levels; and the various actors in the agricultural system (those (vi) the U.S. Department of Agriculture’s Economic directly affecting the flow of goods, as well as those Research Service (USDA/ERS) does interesting analyti- providing support services) and their functions. cal agricultural sector work in developing economies. » Future outlook. Near-term projections (demand pro- Lastly, Internet search engines are an invaluable way to jections, supply forecasts, and commentary on the access a wealth of information. Appendix C provides a country’s commodity outlook). list of background data that facilitate or significantly » Governance and coordination mechanisms. The domi- contribute to the assessment of risks. nant governance structures and coordination 18 Agriculture Global Practice Discussion Paper mechanisms and how they promote or inhibit the A simple graph of agricultural GDP growth over time can flow of information, risk sharing, and risk manage- help identify years in which agriculture was shocked by ment across the sector. the realization of some type of event or risk. Those events » Enabling environment. The broader global enabling can also be identified by the literature review conducted environment, as well as national and regional as part of the initial background research. The supporting enabling environments (for example, national and information may appear dispersed and come from various regional trade policy, level of trade cooperation sources. Figure 4.1 illustrates a completed event time line and integration). These play a key role in shaping associated with changes in agriculture growth rates in the sector, so it is useful to identify their implica- Senegal. tions for risks in the sector. Once the available information has been compiled and organized into a time line that captures production shocks CONSTRUCTING AN over a period of 25 to 30 years, it will need to be verified EVENT TIME LINE during the field visits. The description of the nature, fre- Developing an event time line of major past risk events quency, and impact of realized risks at the aggregate sec- helps establish correlation and, it is hoped, causation, tor level paints a broad picture that needs disaggregation between events and volatility. It assists in providing early because sector aggregation masks volatility at the com- indications of the major risks and their relative signifi- modity and regional levels. The subsequent tasks are to cance for sector growth. Time lines can also be useful in disaggregate the information to reveal that volatility and developing some initial hypotheses around principal identify the major risks posed to particular commodities risks facing the agricultural sector. These hypotheses can in specific regions, and to determine who is primarily then be field-tested for rejection, further refinement, or affected. Drawing such a time line for each specific com- acceptance. Creation of such a time line greatly facili- modity helps identify the risks associated with specific tates an ASRA, but extensive time series data and infor- commodities as well as their frequency over time. For mation on all major events and their impacts on example, from figure 4.2 it can be hypothesized that vari- commodities’ performance are often required and can ous risks have produced severe losses to cocoa production be difficult to obtain. in Ghana. The team can go to the field to test whether FIGURE 4.1. TIME LINE OF MAJOR SHOCKS TO AGRICULTURAL PRODUCTION IN SENEGAL, 1980–2012 180 Crop production index Food production index Livestock production index 160 140 120 100 Late/erratic 80 Locusts, rainfall; 2004 locusts, Erratic Locusts, 2011 60 Locusts, Late rains, rainfall; 1988 1992 regional Severe birds, 40 Severe droughts, drought; 2007 Severe 1996–98 cold rains; drought; drought, locusts, 20 locusts, 1980 2002 1983–84 0 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 Source: World Bank 2014d. Agricultural Sector Risk Assessment: Methodological Guidance for Practitioners 19 FIGURE 4.2. EVENTS TIME LINE OF RISKS TO COCOA PRODUCTION IN GHANA GHANA 1983 ‘‘Big’’ Bushfire 1993 –1994 2001– 2002 Loss of 60,000 HA Swollen Shoot Outbreak Cross-border Smuggling Loss of 22,620 MT 10 million trees cleared Loss of 50,000 MT US$35.6 million cost of US$20.2 million US$56.1 million 1980 1990 2000 2010 2001– 2002 Blackpod Outbreak Loss of 30,000 MT US$35 million 1996 – 1997 Blackpod Outbreak 2009 – 2010 Loss of 32,000 MT Cross-border Smuggling Loss of US$46.7 million Loss of 50,000 MT US$135.1 million these early hypotheses are correct and obtain a firsthand TABLE 4.1. RELATIVE IMPORTANCE OF assessment of stakeholders’ current capacity to manage COMMODITIES BY VALUE IN similar risks. TANZANIA Tanzania Relative % of AG. SELECTION OF Commodity Importance GDP COMMODITIES Livestock 18 A developing economy’s agricultural sector is often Cash crops Export value based on production of a wide range of agricultural Tobaco l 2.1 products, making it difficult to select the commodities of Cotton 2 1.8 focus for analysis. In such cases, filtering criteria can be Cashew nuts 3 1.5 useful. Teams can rank commodities in terms of their Coffee 4 1.2 relative importance to export earnings or by their use for Tea 5 0.5 food security and domestic consumption (see table 4.1 Food crops Food security for an illustration from Tanzania). A simple rule of Maiz 1 31.1 thumb used in prior assessments is to choose a group of Paddy rice 2 12.5 commodities whose output represents approximately 80 Bananas 3 8 percent of agricultural GDP to arrive at a representative Millet/sorghum 4 5.4 Cassava 5 5.2 group of commodities. Some governments choose to Vegetables 6 4 include emerging crops for which current production Sweet potatoes 7 3.1 may be limited but for which surging market demand Other 5.6 demonstrates a growing trend. Other commodity selec- tion factors include area and number of farms or pro- Total 100 ducers involved in the production of each commodity or Source: World Bank 2013b. 20 Agriculture Global Practice Discussion Paper contribution to rural employment. This information can As the agricultural sector is highly vulnerable to rain- assist in selecting the commodities to be analyzed. fall variability, a detailed quantitative analysis of the Regardless of the group of commodities chosen, the impact of rainfall at a disaggregated level is useful. country counterparts, as end users of the findings, need Appendix E provides an illustrative example of a to endorse the selection. quantitative analysis of rainfall and its impact on crop production in Ghana. IDENTIFICATION OF KEY RISKS ASSESSMENT OF THE Provided enough depth has been devoted to the literature RISK MANAGEMENT review, the team should have the information to construct a long list of risks that have been mentioned as sources of LANDSCAPE volatility. This serves as the basis for the process of risk Another component of the background research is identi- identification. In some cases, enabling environment risks fication of current risk management programs and related are prevalent because of the nature of the political econ- interventions by the public and private sectors. Sometimes omy in the sector. In other cases, weather risks might be those programs were not necessarily designed with the significant, especially for agricultural supply chains pro- objective of risk management, but addressing risk is often- ducing food crops and livestock in drought-prone coun- times implicit in their activities. Understanding the exist- tries. Market risks tend to be more predominant in ing risk management landscape lays the groundwork for export-oriented agricultural supply chains focused on the team to begin identifying gaps. Once the literature production of commodities with highly volatile interna- review is concluded, a gap analysis can be completed dur- tional markets (that is, cocoa, cotton, coffee). ing field visits. The initial list of risks that arises from the preliminary For example, the ASRA team in Mozambique found three literature review and data analysis will be subject to fur- intervention categories aligned with the National Invest- ther analysis based on their historic frequency of occur- ment Plan for the Agrarian Sector (PNISA) and its Strate- rence and their capacity to cause major losses (see chapters gic Plan (PEDSA); many of them were being implemented 5 and 7). Table 4.2 illustrates the long list of risks identi- with positive impacts, albeit at a smaller, regionally local- fied in an ASRA conducted in Mozambique. ized level. The team placed greater emphasis on TABLE 4.2. KEY IDENTIFIED RISKS IN MOZAMBIQUE’S AGRICULTURAL SECTOR Production Risks Market Risks Enabling Environment Risks Drought (late onset of rain, early cessation of rain, Domestic price volatility Political unrest and violent conflict irregular rainfall, cumulative rainfall) Flood International price volatility Infrastructure disruption Cyclone Exchange rate risk Implementation risk Locusts Input price volatility (for example, fertilizer, diesel) Wild animals Counterparty risk Irregular rainfall Grain-eating birds Alluvial and soil erosion Heat and excessive temperature Wildfire Source: World Bank 2013a. Agricultural Sector Risk Assessment: Methodological Guidance for Practitioners 21 recommending the scale-up of those measures to the The initial risk assessment requires understanding the national level to meaningfully affect the whole agricultural landscape of these interventions, assessing their relative sector. The same situation arose for the ASRA team in Par- efficacy, understanding principal challenges to success and aguay, whereby foot-and-mouth disease (FMD) was identi- scale, and identifying leverage points and necessary inter- fied as a priority risk for small-scale livestock herders. The ventions to increase access to a wide majority of agricul- government was already implementing programs to control tural sector stakeholders. Assessing solutions to help FMD outbreaks through the National Service of Animal prioritize specific interventions, scaling-up priority pro- Quality and Health (SENACSA), but the ASRA revealed grams, and putting in place a risk management imple- serious gaps. Strengthening the program became one of mentation plan are the next steps in the process of building the government’s risk solutions as a follow-up to the ASRA. resilience to external shocks in agriculture. 22 Agriculture Global Practice Discussion Paper CHAPTER FIVE STEP 2: RISK QUANTIFICATION The task of quantifying losses caused by different risk events is a key step in the ASRA process, but it is not easy. Financial losses attributed to agricultural risks are caused by a variety of shocks related to production, market, or enabling environment factors. The impacts of some risks can be complicated to quantify in monetary terms, whereas others cannot be quantified at all. Assigning proxy values can provide an estimate of the magnitude of financial losses to allow comparison of risks, paving the way for the prioritization of major risks. Various approaches exist to quantifying risk. The World Bank’s ARMT relies primar- ily on time series analysis to assess production risks; variations from the historical mean to assess market risks; and qualitative analysis to assess enabling environment risks. Chapter 5 focuses on how to quantify financial losses using time series analysis when a complete set of time series data is available. The chapter also discusses methods for quantifying losses under poor data scenarios. RATIONALE Quantifying losses associated with a risk event enables comparison or risk ranking as an intermediate step in the risk prioritization process. Loss quantification provides an order of magnitude of the indicative losses in terms of agricultural GDP, which helps to justify investment in risk management solutions that reduce the impact of external shocks and strengthen resilience. In fact, losses are a key piece of information for any cost-benefit analysis for investing in agricultural resilience. Ideally, if data availability permits, the quantification of losses for each commodity under study should be done as part of the deskwork phase before the field visits. The findings are very useful for determining the magnitude of loss, its causes, the stake- holders affected, and stakeholders’ capacity to manage. Calculating loss estimates before the field visit makes the fieldwork much more efficient and helps facilitate dis- cussions with stakeholders. Agricultural Sector Risk Assessment: Methodological Guidance for Practitioners 23 losses using yield data and production data is shown in HOW TO QUANTIFY LOSSES tables 5.1 and 5.2, respectively. ASSOCIATED WITH RISK Several methods can be used to quantify risk, ranging Alternatively, if time series data of yields cannot be from calculating the simple variation from a mean value obtained or are not reliable, the calculations can be made to using sophisticated statistical software applications as is using production data either from the country or from commonly done in the insurance industry. The focus FAOSTAT. Using production data instead of yield data herein is on conducting a simple time series analysis of means that the loss calculation also includes changes in risk events to arrive at the monetary value of losses caused area, not just yield (production = yield times area). It is by production risks. Other chapters address in detail how important to use the same methodology for all commodi- to reconcile the quantitative assessment of production ties so that the quantification of losses uses the same yard- losses with the qualitative assessment used to rank the sig- stick to enable risk prioritization. nificance of losses associated with market and enabling environment risks. Figure 5.1 provides an illustration of a loss calculation using rice yields in Tanzania. It is worth noting the following: For production risks, calculations are made individually 1. The blue line represents the actual historical for each commodity under study. They are then aggre- records of rice yields in tons per hectare from the gated to quantify losses for the whole agricultural sector. 1981–82 agricultural cycle until 2009–10, repre- The step-by-step methodology to quantify production senting 29 years of data. TABLE 5.1. STEP-BY-STEP METHODOLOGY TABLE 5.2. STEP-BY-STEP METHODOLOGY FOR QUANTIFICATION OF FOR RISK QUANTIFICATION OF LOSSES USING YIELD DATA LOSSES USING PRODUCTION Calculate Indicative Value of Yield Losses DATA for Individual Crops Calculate Indicative Value of Production Losses for Individual Crops 1 Obtain time series data for yields of a commodity from in-country data or FAOSTAT. 1 Obtain time series for total crop production from 2 Derive linear trend for yields using ordinary least in-country data or FAOSTAT. squares (OLS) and associated predictive values. 2 Obtain corresponding nominal price data. 3 Calculate standard deviation (SD) for linear trend. 3 Calculate value of crop production in constant prices, 4 Derive threshold: Trend value minus (0.33 multiplied for either a given year or an average for a given period by SD). of years; this can be done using the most recent 5 Calculate loss below threshold = Trend value minus in-country prices or the prices for the corresponding yield value for each year where yields are lower than period from FAOSTAT data. threshold. 4 Derive linear trend for value of crop production using 6 Calculate total output loss: Yield loss multiplied by OLS and associated predicted values total area harvested. 5 Calculate SD for linear trend. 7 Get yearly prices of commodity and convert them 6 Derive threshold: Trend value minus (0.33 multiplied from nominal to constant value. by SD). 8 Calculate value loss of output: Total output loss 7 Calculate loss below threshold: total crop prod value multiplied by price (at constant value). minus threshold value. 9 Convert to US$ using annual exchange rates. 8 Convert to US$ using annual exchange rates. 10 Calculate annual average losses: Sum of value loss 9 Calculate annual average losses: Sum of value loss for for years below threshold divided by total number of years below threshold divided by total number of years years in series. in series. 24 Agriculture Global Practice Discussion Paper FIGURE 5.1. PADDY RICE YIELD LOSSES TABLE 5.3. PADDY RICE IN TANZANIA: IN TANZANIA (TONS PER HA), LOSSES FOR YEARS IN WHICH 1981–2010 YIELD DROPPED BELOW THE 3.0 Yield (x) Trend 0.3 trend THRESHOLD 2.5 Years Loss (MT) Loss (US$) 1982/83 65,762 44,465,202 2.0 1983/84 64,303 43,478,588 1.5 1990/91 161,252 109,031,104 1991/92 94,242 63,722,253 1.0 1996/97 184,113 124,488,936 1997/98 255,477 172,741,869 0.5 2000/01 239,066 161,645,055 0.0 2004/05 191,871 129,734,130 2007/08 250,504 169,379,182 19 /82 19 /84 19 /86 19 /88 19 /90 19 /92 19 /94 19 /96 19 /98 20 /00 20 /02 20 /04 20 /06 20 /08 0 /1 81 83 85 87 89 91 93 95 97 99 01 03 05 07 09 2008/09 176,671 119,456,593 19 Note: Ha = Hectare. Annual average 58,044 39,246,307 As percentage of Agricultural GDP: 0.7 2. The red line represents the linear trend of yield over Note: MT = Metric ton. time. This linear trend can be thought of as the yield that farmers would “expect” in the absence of volatility. In other words, investment decisions for Table 5.3 shows the estimated losses associated with paddy growing rice are expected to be made based on a rice production in Tanzania using time series data. normal year, represented by the trend line. 3. The green line represents the threshold. The These calculations are done for each commodity under threshold represents deviations from the trend (the study, taking care to use the same methodology for each “expected” yield) that would normally occur in calculation. Once the calculations for all commodities the course of doing business. In other words, yield are done, they are aggregated to estimate losses for the drops down to the threshold line are considered sector as a whole. Table 5.4 illustrates the aggregation losses that are part of “doing business.” This is not of losses across commodities for an ASRA conducted in desirable, but is not considered extraordinary. In Paraguay. this case, it is calculated as one-third of a standard deviation, an arbitrary rule arrived at by trial and Similarly, when data availability allows, the calculations error. Other ways to establish the threshold (either can be disaggregated by province or region to provide proportions or other measures) do not provide more granularity in the findings. Table 5.5 illustrates results as consistent as those obtained from using regionally disaggregated results for losses found in Para- the current threshold. The important issue is that guay’s ASRA. the threshold captures major drops in yield or pro- duction that are worth taking into account for the Production shocks can affect prices in the short term, purpose of risk identification and quantification. especially for goods with limited trade, so using average 4. Any deviations in yield that drop lower than the annual prices to demonstrate the value of the losses may threshold are considered losses incurred because be misleading. It is therefore important to determine what of a “risk” event. In this case, in 10 years out of 29 type of prices to use in the calculation: yield dropped below the threshold. » Constant prices: show the value associated with 5. Consequently, losses in monetary value need to be production losses only, enabling a comparison of calculated for those years only. the volume lost from year to year. Agricultural Sector Risk Assessment: Methodological Guidance for Practitioners 25 TABLE 5.4. DISAGGREGATED LOSSES BY TABLE 5.5. DISAGGREGATED LOSSES BY COMMODITY IN PARAGUAY REGION IN PARAGUAY Commodity Local Currency (₲) US$ Province Local Currency (₲) US$ Garlic (1,540,068,785) (355,428) Concepcion (24,136,471,088) (5,570,383) Cotton (38,294,643,861) (8,837,905) San Pedro (112,826,815,210) (26,038,960) Irrigated rice (20,372,821,741) (4,701,782) Cordillera (17,821,730,343) (4,113,023) Rice (4,497,983,960) (1,038,076) Guaira (27,770,086,845) (6,408,975) Sweet potato (6,588,120) (1,520) Caaguazu (146,778,337,805) (33,874,530) Sugarcane (28,833,977,166) (6,654,507) Caazapa (64,400,703,955) (14,862,844) Onion (1,856,184,503) (428,383) Itapua (211,046,236,386) (48,706,724) Frutilla (1,590,928,493) (367,166) Misiones (21,647,858,182) (4,996,044) Locote (3,124,627,686) (721,123) Paraguari (25,204,096,945) (5,816,778) Maize (229,559,269,995) (52,979,291) Alto Parana (312,719,057,093) (72,171,488) Cassava (124,232,296,459) (28,671,197) Central (9,122,516,360) (2,105,358) Groundnuts (20,861,192,474) (4,814,492) Ñeembucu (5,171,201,082) (1,193,446) Potatoes (343,895,779) (79,367) Amambay (39,059,203,341) (9,014,356) Beans (9,264,417,036) (2,138,107) Canindeyu (143,261,178,695) (33,062,815) Sesame (17,441,343,466) (4,025,235) Pte. Hayes (3,060,072,506) (706,225) Soybean (616,409,635,438) (142,259,320) Alto Paraguay (2,246,771,475) (518,526) Tomato (16,556,806,939) (3,821,096) Boqueron (14,691,763,924) (3,390,668) Wheat (23,589,511,795) (5,444,152) Total (1,180,964,101,235) (272,551,143) Carrot (248,496,520) (57,350) Banana (14,596,653,216) (3,368,718) Pineapple (3,057,428,271) (705,615) series production data using a variety of sources. No one Pomelo (4,685,329,530) (1,081,313) single way exists to reconstruct missing data, as it will Total (1,180,964,101,235) (272,551,143) always be based upon assumptions that need to be thor- Note: ₲ = Paraguayan guaranties. oughly tested with key informants with long experience in the agricultural sector. The methodology used for recon- structing time series production data for Mozambique is » Real prices: show joint production-price shocks, shown in box 5.1 for illustrative purposes. In Mozam- as relative price changes are captured. For some bique, disruptions in data occurred because of the long countries, ASRAs have shown that joint produc- civil war. Other countries experience problems in captur- tion-price shocks have a larger impact on the sec- ing time series data on agriculture because of budget tor than production shocks alone. However, this problems or complexities in capturing and processing method may also show that prices compensate for information. In such cases, the risk assessment team needs some of the impacts of risks on production. to creatively reconstruct data using available sources and Both methods are useful for understanding the relative thoughtful assumptions. importance of different risks. SCENARIO WITH NO DATA SCENARIO WITH POOR- If the risk assessment team finds itself in a situation where QUALITY DATA the quality and quantity of data are still limited even with Oftentimes the risk assessment team will find that the time various sources, it will need to rely more heavily on find- series data needed for production risk identification and ings from interviews with stakeholders and key infor- quantification are of poor quality or have too many miss- mants. Thus, the fieldwork plays an even more crucial ing observations. The alternative is to reconstruct the time role, because the actual identification of risk events, the 26 Agriculture Global Practice Discussion Paper BOX 5.1. ALTERNATIVE APPROACH TO the magnitude of losses incurred and their frequency, CALCULATION OF PRODUCTION apart from the usual questions regarding stakeholders’ capacity to manage and their vulnerability. Testing the RISK LOSSES IN MOZAMBIQUE’S findings in workshops with stakeholders and focus group ASRA discussions in every commodity supply chain under In Mozambique, available data on actual losses caused by study is imperative. adverse events are neither particularly accurate nor consis- tent within individual data sources. To facilitate compari- son and ranking of the losses attributable to various events, FREQUENCY AND INTENSITY different data sources were combined to generate a more or The frequency of risk events and their capacity to pro- less consistent time series. The calculations were as follows: duce losses (intensity) are variables that need to be esti- 1. Gross production values were derived from FAOSTAT mated once the team has identified the major risks to the data. agricultural sector. These variables and the vulnerability 2. The share of production from crops and agriculture to the impact of risk (expressed in this approach as the that were studied represented approximately 60 percent capacity to manage risk) will serve as the basis for the risk of gross production value, crops included those that prioritization, explained in detail later in the guidelines. were fairly representative of the sector structure. However, at this stage it is worth noting that these param- 3. The amount of land planted in hectares for arable and per- manent crops per year was derived from FAOSTAT data. eters need to be estimated for each commodity under con- 4. The estimated value per hectare of land planted with sideration. Table 5.6 illustrates how observed risks to arable and permanent crops was calculated by dividing cotton in Mozambique were assessed by frequency (Prob- the figure for the share of production from crops and ability of Event—categorized as highly probable, proba- agriculture using constant 2004–06 U.S. dollars by the ble, and occasional) and intensity (Potential Severity of estimated amount of land planted for arable and per- Impact—with impacts ranging from negligible to cata- manent crops. 5. The number of hectares lost per risk event was taken strophic). from analysis of annual reports from SETSAN (food security situation reports), the PES (the Annual Balance The risks located in the upper right-hand, darker-shaded of Economic and Social Plan and Annual Economic corner represent the greatest risk for this particular com- and Social Plans), Famine Early Warning Systems Net- modity. To obtain the broader picture for the sector as a work (FEWS NET) reports, Global Information and whole, this exercise would be replicated for all commodi- Early Warning System (GIEWS)/FAO reports, and data ties under study, using the findings from the quantitative downloaded from the Early Warning System. 6. The estimated loss was calculated by multiplying the assessment and those from interviews with stakeholders estimated value per hectare by the hectares lost in a (more on this in later chapters). given year caused by risk events. The event time line discussed in the section “Constructing Source: World Bank 2013a. an Event Time Line” in chapter 4 is a useful tool for risk prioritization because it provides evidence about the intensity and frequency of major events that caused agri- cultural GDP to drop. frequency of their occurrence, and their intensity will be derived primarily from interviews. Bubble graphs are a good illustrative tool to use to com- pare and contrast different production risks relative to This approach has many limitations, the main one being intensity (as indicated by losses) and frequency, where that people generally have short memories and more data are available. Figure 5.2 provides a visual represen- easily remember recent events. Consequently, the extent tation of frequency and intensity using bubbles to denote and depth of field interviews need to be greater than in the relative significance of various risks in Niger. In this scenarios with more data availability. In this case, the case, drought is the biggest risk for Niger’s agricultural team needs to pay particular attention to get a sense of sector. Agricultural Sector Risk Assessment: Methodological Guidance for Practitioners 27 TABLE 5.6. ASSESSMENT OF FREQUENCY AND INTENSITY OF RISK IN MOZAMBIQUE’S COTTON SUPPLY CHAIN Potential Severity of Impact Probability of Event Negligible Moderate Considerable Critical Catastrophic Highly Sudden change Farmers’ credit default International Crop substitution. probable in orders for (chemicals). cotton price chemicals. volatility. Probable Port delays. Weather Ginners’ credit default. Pests (aphids, (droughts, floods, Transaction tax default worms, and so and so on). and payment delays by on). ginners. Occasional Carrying large Fire. Loss of soil fertility. Exchange rate quantities of Unreliability of risk. cash. transportation Sample testing market. delays. Source: Cotton supply chain risk assessment for Mozambique. “Mozambique Cotton Supply Chain Rapid Risk Assessment”: http://siteresources.worldbank.org/ INTCOMRISMAN/Resources/MZ_CottonRiskReport_FINAL(Nov2010).pdf. FIGURE 5.2. EXPECTED AVERAGE also because they represent the accumulation of savings LOSSES FOR ADVERSE CROP and capitalization of productive activities. PRODUCTION EVENTS IN NIGER 90 Assessing risks to livestock production is a complex techni- 80 cal challenge, and no unique method exists for conducting such a risk assessment. Therefore, it is suggested that an Indicative loss $US m 70 60 Drought experienced expert on livestock is incorporated as part of 50 the risk assessment team, and that local expertise is sought 40 out and relied on as much as possible. 30 20 Price Locusts In general, the following steps could be followed to con- 10 0 Floods duct a risk assessment for livestock: 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 1. Identification of the various supply chains. Livestock pro- Frequency duction and marketing could be organized along Source: World Bank 2014a. different and particular supply chains. Examples are livestock pastoralists, intensive livestock pro- duction, silvo-pastoralists systems, intensive dairy A NOTE ON LIVESTOCK production, artisanal cheese channels, and so on. Livestock activities are very important for a vast propor- The exposure and vulnerability to risks might be tion of rural households. Shocks that affect livestock pro- very different in many countries. Understanding duction can be devastating in many developing countries, how these supply chains work, what stakeholders particularly in Sub-Saharan Africa where entire commu- are involved and who they are, and their turnover nities depend on pastoral activities for their livelihoods. is the starting point for a risk evaluation. For other households, livestock are valued not just for 2. Identification of threatening events (hazards and shocks). their products in terms of meat and dairy products but The objective is to have a qualitative understanding 28 Agriculture Global Practice Discussion Paper of the risks (historically present and poten- To answer these questions, qualitative and quantitative tial) for each identified supply chain, noting the analysis will be needed. Stakeholders’ risk perceptions following: could be obtained by interviews, secondary literature, a. Types of risks. What are the major risks for each focus group discussions, surveys and questionnaires, dis- supply chain associated with losses that are cussions with stakeholders, and expert opinions. recorded or remembered by stakeholders? b. Historical events. What shocks (production risks, The outcome of a livestock risk assessment would in prin- market risks, enabling environment risks) have ciple include the following: affected the supply chains during the past 20 » Estimation of potential losses per level of hazard years? and stakeholders for each supply chain c. Severity and frequency. How frequently are stake- » Impact on investment, income, debt, employment, holders exposed to shocks and how severe are and export earnings due to risks they? How long do the shocks last? » Analysis of vulnerability for groups of stakeholders d. Degree of dispersion. Were all supply chains and comparison under different types of risks and affected equally? What stakeholders were geographic location affected most? What regions are more prone to losses? There are many ways how to organize and analyze the e. Characteristics of affected stakeholders. What are the information gathered via the steps described above. But socioeconomic profiles of the most affected the common variables of the analysis for each identified stakeholders? risk are intensity to cause losses, frequency of occurrence, 3. Assessing losses. For each of the value chains in and capacity to manage. Analyses of those variables are the livestock subsector, it is practical to catalogue the inputs for risk prioritization in the livestock subsector. assets and income levels. The following questions Using a prioritization table similar to that shown for crops can serve as guidance: can capture the results of these analyses. Table 5.7 pre- a. What is the average herd per household in sents an illustration of a risk assessment for livestock in each supply chain? Rwanda. A practical advantage of expressing the results b. What is the estimated income from livestock in this table is that the prioritization of livestock risks can production? be consolidated with those for crops to complete the sec- c. How are income levels affected by different tor dimension. shocks? d. How is production affected per hazard and An additional challenge for the risk assessment team is how long does it take to recover to the pre- how to assess risks that have never occurred in the coun- event position? try’s livestock subsector but have the potential to cause 4. Capacity to manage. This step aims at identifying and severe losses if they do happen. For instance, an outbreak assessing the capacity to manage risks for each of FMD in any country in Central America would have group of stakeholders active in the supply chain. severe consequences given the weight that beef exports Relevant questions include the following: play in the economy and the fact that these countries are a. Before the event. Are households applying mitiga- FMD-free (without vaccinations), which allows them to tion practices that would lower the impact of export to any country without restrictions. The risk is ever negative shocks? How widespread is the use of present. To address this challenge, a livestock expert esti- good practices? mated the impact of such an outbreak in Honduras using b. During and after the event. How do households losses from outbreaks in other countries. For the prioriti- cope with losses—selling assets, borrowing, zation of solutions in Honduras, the team relied on insti- selling labor, migrating, lowering consump- tutional assessments done by the Regional Sanitary tion? What are the most common coping Organization (OIRSA) that revealed weaknesses in pre- strategies? vention. Agricultural Sector Risk Assessment: Methodological Guidance for Practitioners 29 TABLE 5.7. ILLUSTRATION OF RISK PRIORITIZATION IN LIVESTOCK FOR RWANDA Impact and Probability of Event Negligible Moderate Considerable Critical Highly Probable • Milk contamination • Disease (1 in 3) outbreak(s) Probable • Drought (1 in 5) • MCC (milk collection center) power cuts Occasional • Glut (price risk) • Drug and livestock (1 in 10) inputs contamination and adulteration risk Remote • Aflatoxins (1 in 20) • Maize production shortages Source: World Bank 2014c. For domestically traded commodities, price vola- QUANTIFYING MARKET tility is usually intimately related to supply conditions. RISKS—PRICE VOLATILITY Price volatility is not necessarily autonomous, but the Market risks are generally related to issues that affect effect or symptom of realized production risks. Analyzing price, quality, availability, and access to necessary prod- time series data on production and correlating price spikes ucts and services. Of these, price risks are typically vola- with corresponding realized production risks is a practical tile, particularly in commodity markets where both local way to confirm this assumption. Managing price risk of and global supply and demand conditions constantly this nature is handled by public policy, a topic addressed change. Price uncertainty has a direct impact on decision in chapter 8. making related to the selection of crops and enterprises and investments with the intent of maximizing profit. The methodology for quantifying losses associated with price volatility could potentially mirror that used for Quantifying price risk could become a complex exercise quantifying production losses. In the former case, it with no straightforward way to assess volatility. For the risk applies to losses to producers when price drops below a prioritization process, the team needs to distinguish given threshold. Though it can be interesting to quan- between internationally traded export commodities (that tify the magnitude of the price effect, the results should is, cocoa, cotton, coffee, soybean) and those that are mostly be interpreted carefully as (i) price volatility, whether a traded domestically (that is, food crops). For interna- spike or drop, can differentially affect different stake- tionally traded export commodities, provided the holders (farmers, supply chain actors, consumers, gov- country is not a price giver but a price taker, domestic ernment) and financial losses are difficult to quantify stakeholders are generally exposed to price volatility. The because each episode of volatility is associated with volatility of these commodities has been thoroughly stud- some stakeholders losing and others gaining; (ii) com- ied in the academic and business community. As the com- plexities arise in isolating price movements from supply plexities of those calculations can go beyond what is responses; and (iii) absorption of price shocks is strongly reasonably needed for the ASRA, the team can assume linked to market structure (that is, storage facilities, that if a country relies heavily on export earnings from commercialization, players). Because of these complex export commodities, it is exposed to international price factors, ASRAs generally rely on qualitative measures to volatility. The focus should thus be geared toward identify- evaluate price volatility. ing who is exposed and who is mostly affected within the sector, as well as those stakeholders’ capacity to manage Similarly, enabling environment risks by their nature do volatility. not lend themselves to empirical observation and direct 30 Agriculture Global Practice Discussion Paper attribution. Qualitative measures are thus usually deployed and the vulnerability of specific stakeholders (as discussed to compare and contrast the effects of enabling environ- in the following chapters). ment risks with those of other agricultural risks. Finally, although many production, market, and enabling LIMITATIONS environment risk shocks have longer-term consequences Quantifying losses is often difficult because of the scar- and losses, for simplicity’s sake, it is helpful to restrict the city of time series data on the occurrence of events and calculation only to the immediate direct losses. Identifying the associated losses. Assessment teams must rely on the net multiplier effects in the economy requires general proxies or qualitative rather than quantitative measures equilibrium models that go beyond the scope of the to assess impacts and compare risks. For instance, in ASRA. countries with a single crop that accounts for a great portion of total exports and GDP (for example, soybeans By the time the desk study is complete, the team should in Paraguay or cocoa in Ghana), the reduction in have produced (i) a time line of major shock events that exported volume may be a good proxy for roughly esti- have caused volatility in the country’s agricultural GDP; mating losses associated with events such as droughts or (ii) a long list of past risk events with a relatively good floods. approximation of their frequency; (iii) a quantification of losses for each commodity because of those risks in terms Whereas the methods and examples provided above cover of yield losses (volume) and value losses (monetary value); production risks (drought, flood, pests and diseases, and so (iv) an assessment of the importance of price volatility of on), it is more difficult to quantify losses related to market export commodities in terms of the magnitude of shock and enabling environment risks. Income losses derived to the sector; and (v) an assessment of major current inter- from price drops are very difficult to isolate from varia- ventions by the public and private sectors addressing risk. tions in supplied volumes. However, it is possible to ana- The findings of the deskwork serve as the basis for formu- lyze price volatility and transmission and then estimate lating preliminary hypotheses regarding risk prioritization the domestic supply chain’s exposure to external shocks and the potential solutions to identified risks. Agricultural Sector Risk Assessment: Methodological Guidance for Practitioners 31 CHAPTER SIX STEP 3: FIELDWORK For illustrative purposes, one can imagine deskwork as a “top-down” approach, which means that the starting points are the sector statistics, programs, and assessment of value chains risks and commodities. The fieldwork, by contrast, can be considered the “bottom-up” side of the exercise, where the team seeks to have direct one-to-one dis- cussions about risks with key stakeholders of major agricultural supply chains. PURPOSE The purpose of fieldwork is not just to corroborate the findings of the desk assessment, but to help the team identify the story line. The team seeks to (i) corroborate the time line of events; (ii) test the causality of risks identified; (iii) test whether the losses can be validated against what happened in reality; and (iv) assess stakeholders’ capacity to manage risks. Box 6.1 shows an example of the line of enquiry that needs to be dis- cussed by team members as part of the background research and prior to arriving in country for the ASRA fieldwork. FIELDWORK ACTIVITIES Fieldwork consists of a limited number of activities, but these require remaining in the field for at least two weeks to conduct interviews with key stakeholders of main supply chains. Organizing the timing and logistics of the visit is particularly important when the assessment is done in a large country with difficult access to remote areas and lim- ited overnight accommodations. It is advised to plan the trip logistics well in advance, including details of appointments, transportation, and interviews. Having a consultant with local knowledge who can assist with making arrangements prior to the arrival of the risk assessment team has proven to be an indispensable element. The main activities typically covered during the fieldwork include the following: (i) Data mining. This is a valuable opportunity to fill gaps in the information needed. Some pieces of information exist only in hard copy and only by visiting institu- tions can the team access them. For example, commodity boards’ annual reports have valuable information about supply chain dynamics reported on an annual basis, including causes for drops in volume and logistics and market issues. Agricultural Sector Risk Assessment: Methodological Guidance for Practitioners 33 BOX 6.1. PRODUCTION RISK: A BOTTOM-UP homogenous groups (that is, small farmers pro- APPROACH TO FIELDWORK ducing cotton). Focus groups facilitate identifica- tion of the group’s capacity to manage risks and What is the story line? Line of enquiry: discover how they cope with risk as individuals and » Identify the causes of losses and their attribution as a community. An early gender analysis by sup- (single or multiple causes). ply chain will show where it will be more critical » Assess how losses affected various members of the to interview women versus men. Risks may have supply chain. differential impacts on men and women, depend- » Find out how shock was absorbed. ing on their level of participation in certain supply » Establish how different stakeholders managed risks. » Determine whether losses were evenly distributed by chains. Similarly, gendered differences in access to area or by farmer groups. resources and markets may influence the strategies » Identify differences by sex, age, or size of producers, men and women use to manage risks. Appendix processors, and so on. D describes the methodology for conducting focus » Corroborate the frequency of those events. group interviews with farmers. » Establish the persistence of the impact of events (iv) Validation workshop. At the end of the fieldwork, the (short term or medium term). team holds a workshop with key stakeholder rep- » Get a sense of long-term threats to supply chain. » Determine whether any stakeholders went out of resentatives to discuss the findings and facilitates business. a group exercise to prioritize risk management » Understand how small, medium, large farms man- solutions. The group exercise consists of filtering aged in terms of mitigation, transfer, and coping a long list of potential solutions to the identified mechanisms. risks into a prioritized short list to reduce volatility » Obtain evidence from stakeholders: reports, publica- and therefore losses in the agricultural sector. tions, evaluations, and so on (about losses and fre- quency of events). » Test whether magnitude of losses estimated from the THE ASRA TEAM top-down approach are correct. Based on the World Bank’s experience conducting various » Assess capacity to manage risks by existing institu- tions managing risks (formal and informal). ASRAs, a team of between three and five members can » Analyze government’s current strategies to respond easily conduct an ASRA. The actual number of team to shocks. members depends on the size of the country and the trans- » Elicit stakeholders’ perceptions of risk priorities. port facilities to cover interviews with relevant supply chains » Identify stakeholders’ suggested solutions. actors at various levels. Team members should preferably be seasoned agricultural specialists with broad agricultural sector experience (agricultural sector review specialist, agri- (ii) Interviews. The team interviews at least one rep- cultural supply chain specialist, and so on) rather than resentative of each level of value chain actor for experts or specialists in narrow fields (animal health, inte- each supply chain under assessment. This means a grated pest management, and so on) as an unbiased view is high volume of interviews, because the ASRA will critical for risk prioritization and identification of solutions. cover farmers, processors, input suppliers, govern- Appendix G provides an example of terms of reference for ment agencies, financial intermediaries, service risk assessment experts. Experts in narrow fields will be providers, traders, and exporters. Some value needed at the solutions assessment phase to assist the gov- chains have a regional concentration and it is pos- ernment in targeting the most appropriate risk solutions. sible to meet most stakeholders by travelling to the region. Most government agency representatives It is practical to prepare a package of information with can be met in the national or provincial capital some guidelines for the team prior to its engagement in cities. fieldwork. A simple PowerPoint presentation discussed in a (iii) Focus groups and gender analysis. Group discus- videoconference or over Skype can clarify the scope of the sions are particularly useful for addressing large, field mission and the expected deliverables from each team 34 Agriculture Global Practice Discussion Paper member. Appendix F contains an example of guidelines FIGURE 6.1. FIELD TEAM RESPONSIBILITIES discussed with team members prior to a field assessment. FOR AN ASRA IN TANZANIA Prior to starting the fieldwork, the team leader should assign responsibilities for specific commodities and regions to each team member based on spatial considerations. A team of four specialists can plan four different routes to meet the relevant stakeholders related to a commodity. Supply chains are typically located in spatial clusters and team members should be able to cover the interviews by following a well-planed route. Other commodities, such as many food crops (maize, cassava), are virtually every- where. Depending on country circumstances, the team will agree on how to share responsibilities to cover all commodities and their stakeholders. For instance, analyzing risk to livestock activities can take three dimensions: (i) farmers’ focus groups can be used to evaluate systems in which livestock (ruminants) are per- manently a part of a rural household’s production system and food security; (ii) for more extensive cattle production systems, particularly common in Sub-Saharan Africa, the team will need to secure appointments with groups of pastoralists potentially gathering at cattle markets around towns; and (iii) for more intensive cattle production sys- field to assess their findings, organize them into standard- tems where participants are active in some type of related ized formats, and discuss them as a group. The aggregated supply chain (milk, cheese, industrial processing, and so sector risk perspective is shaped in these team delibera- on), the team will need to interview stakeholders individu- tions. At the end of the field mission, the team presents its ally, as with any other commodity supply chain. findings at a stakeholder workshop to (i) obtain feedback on the risk prioritization and vulnerability exercise; and The simple map in figure 6.1 shows the responsibilities for (ii) engage stakeholders in prioritizing risk solutions (the three subteams that conducted interviews for key agricul- details of this process are addressed in chapter 7). Box 6.2 tural crops in Tanzania. Each subteam covered a particu- shows the expected outputs of the field assessment. lar area with specific responsibilities to interview supply chain participants for specific commodities. Each team member planned a well-prepared itinerary route and BOX 6.2. OUTPUTS OF THE ASRA made appointments for interviews in advance. FIELDWORK FIELDWORK DELIVERABLES » Risk identification matrix » Quantification of risks The field mission is the culmination of the ASRA. Each » Detailed analysis of individual risks, interrelated team member has only a partial view of the risk process risks, and associated commodities and regions because he or she has focused on analyzing specific com- » Relative impacts of risk modities in selected territories. The next step is thus to » Vulnerability ranking and assessment integrate the findings through a process of aggregation to » Long list of potential solutions » Stakeholder list of filter mechanisms to produce a achieve a sector perspective. It is imperative that team short list of solutions for a solutions assessment members allow enough time after coming back from the Agricultural Sector Risk Assessment: Methodological Guidance for Practitioners 35 CHAPTER SEVEN STEP 4: RISK PRIORITIZATION As a result of the background research and fieldwork, the team has already under- taken a prioritization exercise for each commodity that shows the ranking of each identified risk in terms of (i) the frequency in which risk events occur; (ii) the severity of the impact (intensity); and (iii) stakeholders’ capacity to manage the identified risk events. VARIABLES FOR RISK PRIORITIZATION Categorizing risk in terms of frequency and intensity can be confusing without the use of consistent definitions or guidelines shared with the team. Again, no magic recipes or standardized procedures exist for establishing these categories, but prior discussion with team members helps to harmonize an approach. As a general rule of thumb based on experience, table 7.1 provides some guidance on how to categorize risks based on probability of event (frequency) and severity of impact (intensity or capacity to produce losses). Apart from the frequency of events and the severity of impact, a third variable will ultimately determine the ranking of risk: stakeholders’ capacity to manage the identi- fied risk (see figure 7.1). For example, if the team finds that stakeholders already have high capacity to manage a particular identified risk, this particular risk will not be ranked among the most important. As mentioned earlier, the capacity to manage each identified risk is assessed mainly during stakeholder interviews, but also by using sec- ondary information collected during the background work prior to the field visit. Chapter 9 addresses the concept of vulnerability and capacity to manage risk in greater depth. PRIORITIZATION MATRIX The use of a prioritization matrix helps enormously in categorizing risks in terms of frequency and intensity. If team members use this table in a standardized manner, results can be shared and discussed among stakeholders and team members in an easy- to-understand fashion, and corrections can be made accordingly and quickly. (Recall Agricultural Sector Risk Assessment: Methodological Guidance for Practitioners 37 TABLE 7.1. RISK CLASSIFICATION EXAMPLE table 5.6, which showed an example of a risk prioritiza- Probability Severity tion matrix for cotton in Mozambique.) of Event Indicator of Impact Indicator Highly Within Catastrophic More than FROM COMMODITY RISK probable 2-year 50% losses TO SECTOR RISK interval Translating individual commodity risk prioritization into Probable Within Critical Between 5-year 30% and an aggregated sector risk prioritization is a collective exer- interval 50% cise, because prior to this stage team members have Occasional Within Considerable Between assessed risks for individual commodities, but a broader 10-year 15% and sector perspective has not yet been established. After com- interval 30% pleting an individual prioritization matrix for each com- Remote Within Moderate Between modity, the team proceeds to reclassify those risks in terms 20-year 5% and of frequency, severity of impact, and capacity to manage interval 15% from a sector perspective. This involves choosing only the Improbable Within Negligible Less than risks located in the upper right-hand corner of each com- 40-year 5% modity’s risk prioritization matrix and relocating them in interval a single aggregate risk prioritization matrix. This process involves not just familiarization with the risk assessment done for each commodity, but a change in per- FIGURE 7.1. THREE KEY VARIABLES FOR spective from the commodity supply chain to a holistic RISK PRIORITIZATION agricultural sector perspective. For instance, the risk of “erratic rainfall” could have a potentially severe impact Frequency of Capacity to for stakeholders participating in the bean supply chain, Severity of impact events manage but might not have such a high impact on the sector as a whole. Figure 7.2 illustrates the process of risk aggrega- tion and reprioritization. FIGURE 7.2. REPRIORITIZATION OF RISKS FROM COMMODITY TO SECTOR LEVEL COMMODITIES SECTOR PRIORITIZATION COMMODITY ZED RISKS ONLY ! PRIORITIZED 38 Agriculture Global Practice Discussion Paper TABLE 7.2. RISK PRIORITIZATION MATRIX FOR CASH AND EXPORT CROPS IN TANZANIA Impact Likelihood Moderate Considerable Critical Catastrophic Highly probable Sesame flea beetle infestation (insects) (S). Diseases (for example, CBD, Cotton price Fungal diseases (for example, powdery CWD, CLR) (CO). volatility (Cot). mildew) (C). Insect/pests (cotton bull Counterparty risk (famers) and side- worm, and so on) (Cot). selling (Cot and Tob). Drought (Cot). Probable Pests (for example, thrips) (Co). Erratic rainfall (CO). Regulatory risk. Price volatility (CO). Price volatility (unstable Counter party (Ginners) and world prices) (C). international buyers (Cot). Excess rainfall (Tob) . Occasional Occurrence of severe drought (S). Source: World Bank 2013b. Note: (S) = Sesame, (C) = Cashew nuts, (Cot) = Cotton, (Tob) = Tobacco, (CO) = Coffee, (CBD) = Coffee berry disease, (CWD) = Coffee wilt disease, (CLR) = Coffee leaf rust. Tables 7.2 and 7.3 show the prioritization matrices com- potential to cause damages or their frequency of occur- pleted for Tanzania (cash and export crops) and Malawi rence is low, or both. (all crops), respectively. The dark areas (upper right-hand corner) represent the most significant risks because of Note that existing losses can be a useful proxy for capacity their potential to cause the greatest losses and the high to manage risks. On the one hand, if existing losses are frequency of their occurrence. The lighter dark boxes low (low impact) despite frequent occurrence of risk, then represent the second level of importance, whereas the the capacity to manage risk is supposedly high and vulner- clear boxes represent identified risks that either have low ability low. On the other hand, if existing losses are TABLE 7.3. RISK PRIORITIZATION MATRIX FOR ALL CROPS IN MALAWI Impact Likelihood Moderate Considerable Critical Catastrophic Highly probable Hailstorms. Pests and diseases (food Drought events, including (1 in 3 years) Untimely distribution of inputs and export crops). – False start of, or (cotton). Price volatility and shorter than normal, Theft (sugarcane, tea, food crops). uncertainty (tobacco, rainy season; Damage from wild animals. tea, cotton, sugar). – Extended dry spells; Power outage (sugarcane, tea). Unpredictable – Higher-than-average Exchange rate (risk mainly for regulatory environment temperatures. smallholders). for traders. Probable Side-selling (cotton). Unpredictable maize (1 in 5 years) Excess of rainfall increasing harvesting market interventions and processing cost (tea, sugar). causing price volatilities in Floods (food crops). the maize market (recent). Occasional Export shipments (1 in 10 years) rejected (tobacco). Source: World Bank 2014b. Agricultural Sector Risk Assessment: Methodological Guidance for Practitioners 39 substantial (high impact), even for low frequency events, addressed to reduce losses in the agricultural sector. The then vulnerability is assumed to be high. highest priority risks are those located in the boxes in the upper right corner of the aggregated risk prioritization The final outcome of the aggregated risk prioritization matrix. The next step is to identify solutions for those process is a list of key priority risks that explain the prioritized risks. causes of agricultural GDP volatility and need to be 40 Agriculture Global Practice Discussion Paper CHAPTER EIGHT STEP 5: PRIORITIZATION OF RISK MANAGEMENT SOLUTIONS In this step the team and stakeholders identify a short list of solutions that they agree will make up the optimal strategy for managing risk in the agricultural sector. As defined in chapter 2, risk management strategies (solutions) fall into three categories: risk mitigation, risk transfer, and risk coping. The long list of solutions, categorized into these three types of strategies, provides the starting point for this step (figure 8.1). LONG LIST OF SOLUTIONS If the line of enquiry was followed during the field visit, the team has identified a list of potential solutions that are meant to contribute to the management of identified risk. Once the sector risk prioritization matrix is complete (recall table 7.3), the team analyzes the long list of potential solutions for the prioritized risks in the matrix and divides those potential solutions into the three categories of risk strategies. The initial long list is often not exhaustive and might not provide concrete next steps, but it does provide a menu of options from which to choose. Table 8.1 shows the long list pre- pared for an ASRA in Tanzania. In conducting Paraguay’s ASRA, the team came up with more detailed long list of solu- tions organized by subsectors, because a huge gap exists between family agriculture and commercial agriculture, such that Paraguay is considered to have a dual agricultural system. The team adapted the methodology to devote special attention to family agri- culture, the subsector with the vast majority of poor small producers, for whom the government identified separate policies, including separate risk management strategies. FIGURE 8.1. RISK SOLUTIONS PRIORITIZATION PROCESS Long list of Prioritization Short list of solutions filters solutions Agricultural Sector Risk Assessment: Methodological Guidance for Practitioners 41 TABLE 8.1. LONG LIST OF SOLUTIONS FROM TANZANIA’S ASRA Risk Mitigation Transfer Coping Drought Drought-tolerant varieties Insurance Food reserves Water harvesting and irrigation Food imports Improving early warning systems Social safety net programs Reforestation and afforestation Risk financing Land and water management Agronomic practices for on-farm drought management Crop diversification Price volatility Managing food stocks Hedging Imports Reform maize trade policy Trade policies Increased domestic processing Social safety net programs Improved quality to access stable niche markets Improved market information systems and transparency Contract farming Improved storage facilities Infrastructure development Foster competition in markets Diseases Scale-up disease-tolerant varieties Quarantine measures On-farm agronomic practices On-farm agronomic practices Early warning systems Integrated pest management Integrated pest management (IPM) Quarantines measures Improved phytosanitary laboratory systems Improved extension services Pests On-farm agronomic practices Quarantine measures Early warning systems On-farm agronomic practices Integrated pest management Integrated pest management Quarantines measures Improved phytosanitary laboratory systems Improved extension services Regulatory risks Improved efficacy of commodity councils Promote proactive rather than reactive policies Develop long-term commodities policies Improved transparency in policy decision making Source: World Bank 2013b. Thus, Paraguay’s ASRA presented a set of solutions for team alone or in a workshop involving stakeholders. It is each main subsector (that is, commercial agriculture, fam- preferable to involve stakeholders, not only to gain their ily agriculture, and livestock). Although the ASRA team support for the process and outcomes but also to take should generally follow the standardized risk assessment advantage of participants’ knowledge. approach presented herein, it can be adapted to each country’s particular characteristics and requirements. Using decision filters in a participatory environment to evaluate and prioritize the risk management proposals helps the team make rational choices in a context of qual- RISK SOLUTIONS FILTERS itative and subjective assessments. Tables 8.2 and 8.3 The screening process for identifying a short list of priori- show decision filters that have been used while prioritizing tized solutions through agreed filters can be done by the solutions for some countries. (See appendix K for a 42 Agriculture Global Practice Discussion Paper TABLE 8.2. DECISION FILTERS FOR decision filter used in Niger, Mozambique, and Ghana.) PRIORITIZATION OF RISK The filters in these tables are shown for illustrative pur- poses only and are not exhaustive. The team should SOLUTIONS choose filters depending on the circumstances or intro- Applicability to current Public sector: Is the proposed duce other filters that stakeholders find more appropriate. agricultural policy and solution in line with current programming or business and existing agricultural objectives policy and programs and SHORT LIST OF SOLUTIONS priorities, and so on? The outcome of the long-list prioritization exercise is a Private sector: Is the short list of solutions that will be further assessed by proposed solution in line with current and existing experts in those topics to ultimately inform the action plan business objectives? that will be part of the government’s policies and strate- Feasibility of implementation Is the proposed solution gies. Box 8.1 shows the short list of solutions developed “easy” to implement in the from the solutions prioritization process carried out in short to medium term? Tanzania’s ASRA. The next step is to recruit experts for Affordability of Is the proposed solution each solution to develop a detailed action plan for the implementation affordable to put into action short to medium term that will be incorporated into gov- and implement? ernment strategies and investments or will inform donors’ Scalability of Is the proposed solution formulation of development projects. implementation easy to scale-up and make available to an increased number of beneficiaries? Finally, the team combines the analysis from all steps and Long-term sustainability Is the proposed solution produces a report. Appendix J provides the table of con- sustainable in the long term? tents from an ASRA conducted in Kenya as an example. TABLE 8.3. ILLUSTRATION OF DECISION FILTERS FOR PRIORITIZATION OF RISK SOLUTIONS Applicability Feasibility Affordability Scalability Sustainability TOTAL SCORE (1-5) (1-5) (1-5) (1-5) (1-5) (Maximum 25) Solution 1 Solution ... Solution ... Solution ... Solution ... Solution ... Solution n Agricultural Sector Risk Assessment: Methodological Guidance for Practitioners 43 BOX 8.1. FINAL PRIORITIZATION OF RISK SOLUTIONS IN TANZANIA’S ASRA The short-list areas for deepening the risk solutions are, in sustainable way. This would create a better balance between brief, the following: the short-term food security goal and the long-term produc- tivity growth aim. Highly drought- (pest-) tolerant seeds. There are weak- nesses in the supply chains for delivering drought-tolerant Risk management strategies for key export crops seeds, disease-resistant seeds, and planting material and inef- with high price volatility (in principle, coffee and cot- ficiencies in seed markets that need to be addressed. In prin- ton). The way these supply chains are organized bears upon ciple, it involves food crops such as maize and rice and export which stakeholder is exposed to price risk. By analyzing the crops such as cotton and coffee. This would imply the need to physical and financial flows on current transaction arrange- effectively intervene in the short to medium term to make the ments for exports, a set of options on how to reduce exposure seeds supply chains work more effectively along the range of to risk can be explored. This would imply a need to deepen stakeholders involved, from breeders to seed producers and to institutional arrangements and clarify current roles of the farmers, as well as to clearly define the roles of the public and public and private sectors. private sectors in developing this market. Whereas there are already interventions of various temporal Good agricultural practices to address drought and and spatial natures in Tanzania on these short-list solutions, pests and diseases. Widespread, improved agricultural the key issue is to focus on identifying the gaps of current inter- risk mitigation practices can have a very significant impact ventions and designing a package of solutions that address in reducing risks derived from irregular or insufficient rain- the main underlying causes of risk. A risk management solu- fall as well as from diseases and pests. This implies a need tions assessment will be planned as a follow-up to current risk to strengthen the existing disconnected technology systems identification. The coming assessment will have the task of through effective coordination between research, extension, linking the risk management interventions to the Agricultural and training, including the effectiveness of information and Sector Development Program by developing concrete propos- communication outreach to farmers. als (policy solutions, investment solutions, and TA solutions) for better managing the risks identified in the short list. In Balanced maize trade policy. The export and import particular, the mission will identify the risk management gaps policy has to be predictable and stable and at the same time in existing interventions; and, propose a set of interventions allow for a transparent market. Policy predictability, market for incorporating them into the medium-term ASDP, which transparency, and fewer nontrade barriers would result in could be financed by the public sector and donors. greater incentives for farmers to invest in technology that increases productivity and reduces production volatility in a Source: World Bank 2013b. 44 Agriculture Global Practice Discussion Paper CHAPTER NINE UNDERSTANDING VULNERABILITY TO RISK As vulnerability is used to prioritize risks and to design and target risk management interventions, assessing vulnerability is an important part of the ASRA. The concept of vulnerability can be applied at different levels (that is, farmer, household, commu- nity) and over various areas (that is, farm, district, province, region), which can lead to confusion. For the purpose of these guidelines, vulnerability is defined as the capac- ity of stakeholders to manage agricultural risk and recover from exter- nal shocks. The capacity to manage risk is assessed during the background work prior to the field visit and during the field visit through stakeholder interviews. Using the methodology outlined in previous chapters provides the team with a good overview of the risks with the biggest impacts on the agricultural sector and the overall economy. Depending on the size of the sector, the magnitude of the impacts, and the variability of their incidence across regions, however, the assessment of aggregate losses can sometimes mask the distribution and scope of impacts on specific groups of stakeholders. The vulnerability assessment is therefore an important tool for more fully analyzing the distributional impacts of risks and developing more efficient risk man- agement mechanisms. STAKEHOLDERS’ RISK PROFILE ASSESSMENT Understanding risk profiles entails analyzing the roles of different stakeholders across the agricultural sector, including the three principal groups (producers, commercial sector stakeholders, and the public sector) but also subgroups, and understanding their risk management capacities. For example: » Who is involved in the value chain analyzed (different stakeholders, segments of population, gender, and so on)? » How does the identified risk affect different stakeholders (for example, farm- ers vs. processors vs. traders and exporters; consumers vs. producers; regional Agricultural Sector Risk Assessment: Methodological Guidance for Practitioners 45 differences; landowners vs. farm workers; men vs. women)? INCORPORATING » What are the current management practices? How THE VULNERABILITY do different stakeholders manage risks and are ASSESSMENT IN THE ASRA these instruments effective? Why or why not? What The vulnerability assessment’s findings can be used in the are the limitations of current practices? Why are ASRA to (i) prioritize risks; and (ii) propose risk some risks not being managed? management instruments. Figure 9.1 maps different perils » What is the management capacity of stakehold- in Ghana’s cocoa sector according to their impacts and ers and supporting institutions to manage a given stakeholders’ capacity to manage them. The figure clearly risk and how effective are the practices and strate- shows that black pod, smuggling, and virus diseases gies employed by various stakeholders to mitigate, (CSSVD) are the risks with the highest impacts and the cope, or share the risk with others? lowest stakeholder capacity to manage. Vulnerability variation between households can be seen as the function of three factors: sensitivity, adaptive capacity, TARGETING WOMEN and exposure. Sensitivity is the degree of impact of the In some cases, certain risks may have relatively low impacts initial shock. Sensitivity can be thought of as the elasticity on the sector or a geographic region, but relatively high of household welfare (that is, consumption levels) in impacts on a particularly vulnerable group of the popula- response to a shock. Adaptive capacity is the ability of tion. Targeting this group with risk management mecha- the household to access ex post coping strategies that help nisms may be a priority. For example, losses in Region A it return to preshock welfare levels. Exposure is the prob- may outweigh those of Region B in terms of monetary ability of a given shock materializing and affecting the value, but if the majority of the country’s food insecure household’s assets. lives in Region B and is dependent on crops prone to risk- related production losses, policy makers may want to pri- Whenever possible, the analysis of these three aspects oritize risk management interventions in Region B, should be done at the local level using available welfare particularly targeting the food insecure. Similarly, finding indicators. Except for poor smallholder farmers, most out that a particularly vulnerable part of the supply chain stakeholders along the commodity supply chain can dis- is predominantly run by women may, because of the cuss their vulnerability to risks during interviews. For poor country-specific cultural context, influence the design of smallholder farmers, who are often the most vulnerable, the relevant risk management policy. Box 9.1 discusses focus group discussions are an important technique to dis- Brazil’s ASRA, in which the vulnerability assessment cuss their risks, mitigation strategies, and coping mecha- influenced the risk management policy priorities. nisms after a disaster. In countries where the initial diagnosis shows that women A vulnerability matrix can map stakeholders and the way in certain communities or regions are in a vulnerable posi- they manage different agricultural risks. A vulnerability tion, analysis needs to be deepened with the auxiliary of table can provide an overview of all stakeholders, stake- secondary research or information that can identify vul- holders along individual supply chains, or stakeholders nerable households more precisely and target interven- associated with a particular commodity. For example, in tions designed to assist those households—even when at an ASRA for Mongolia, the vulnerability matrix lists the sector level those impacts appear marginal. Even small impacts of risk and risk management instruments for shocks to the livelihood of vulnerable groups can have a herders with different herd sizes. Table 9.1 shows the vul- catastrophic impact for large groups of households, nerability matrix used to map stakeholders throughout impacts that are usually masked at the broader cumulative the agricultural sector in the Brazilian state of Paraiba. sector level. 46 Agriculture Global Practice Discussion Paper TABLE 9.1. VULNERABILITY MATRIX FOR AN ASRA IN PARAIBA, BRAZIL Most Important Significance Current Capacity to Stakeholders Risks of Risk Manage Risk All farmers Unexpected irregular Crop and animal losses. Drought-tolerant varieties precipitation and severe Reduced quality of products: and variable crop cycles. drought events. fruits, sugarcane. Family farmers: greater exposure to food insecurity. Smallholders Insect pests (caterpillars), Pests represent an important Only a minority of the diseases (viruses, rots), risk to production if not growers adopts mitigation nematodes, mealy bug. properly controlled. practices, such as pest Animal diseases: exotic Elimination of infected control or resistant varieties. diseases (FMD, BSE, animals and losses of Deficiencies in the provision HPAI), prevalent diseases. production of meat, milk, of TA services and honey. insufficient supply chain Restriction to the marketing coordination prevent family of products. agriculture farmers to access Public health risk from appropriate technologies to zoonosis. control pests and diseases. FMD vaccination. Livestock Drought. Loss of animals, decreased Some financial aid by the producers Animal diseases: exotic production. government—government diseases (FMD, BSE, Quarantine and restricted programs such as Garantia HPAI); prevalent diseases movement and trade of Safra e Bolsa Estiagem. of economical and animals. FMD vaccination. public health importance Public health risk for the Federal and state control, (brucellosis, tuberculosis, transmission to humans. prevention, eradication, and classical swine fever). Food safety compromise contingency programs in caused by contamination of operation. meat and milk. Financial loses. Reduced exports. Sugarcane Drought. Less raw material available Plants accrue financial losses. processing plants to process. During severe droughts, sugar processors experience increase in production cost because plants operate at lower capacity. Consumers Foodborne diseases. Important food safety Federal and state meat and problems. milk control and inspection programs. Planned: State Food Security and Safety System. Government Drought. Social instability. Budget provisions for (national) Animal diseases: exotic Budget implications. risk coping programs and prevalent diseases. (drought support programs, Food safety. contention and emergency funds, compensation funds). Source: World Bank 2014a. Note: BSE = Bovine spongiform encephalopathy; HPAI = Highly pathogenic avian influenza. Agricultural Sector Risk Assessment: Methodological Guidance for Practitioners 47 FIGURE 9.1. MAPPING PERILS ACCORDING TO VULNERABILITY IN GHANA’S COCOA SECTOR Black pod (+) Anticipated losses from Risk Event Mirids/ capsids Smuggling CSSVD Cocoa price volatility (–) (+) Existing capacity to Manage Risk Other pests/diseases Exchange rate volatility Drought/ dry spell Regulatory Risks Counter-party risks Logistics breakdown Interest rate volatility Bushfires Loss of acreage Misappropriation Input price of Funds volatility (–) Source: World Bank 2011. 48 Agriculture Global Practice Discussion Paper BOX 9.1. TAKING INTO ACCOUNT THE SOCIAL DIMENSIONS OF AGRICULTURAL RISK IN BRAZIL In Paraiba, Brazil, sugarcane and fruit (especially grapefruit) risk according to the number of people affected led to differ- are the greatest determinants of agricultural output. Losses ent results: severe rainfall (estimated to occur 1 in 10 years), of these two commodities could be considered a priority in irregular rainfall (1 in 5), and pests and diseases (with lower the risk management strategy, as indicated in the graph below. expected impact than the other risks) were the main risks in Paraiba under these criteria for prioritization. This scenario The conclusion would be rather different if instead of pri- is one of much higher impact but of lower probability. The oritizing risk in terms of the commodities’ economic weight, solutions scenario involved strong actions directed toward the ASRA primarily accounted for the social dimensions of introducing innovative technology and improving agricultural agricultural risk. In Paraiba in 2006, family farmers num- practices. bered 148,047 out of a total 167,272 farmers. Prioritizing FIGURE B9.1.1. RISK INCIDENCE BASED ON LIKELIHOOD, EXPECTED IMPACT, AND RELATIVE ECONOMIC RELEVANCE OF SUBSECTORS 4.5 Severe drought rainfed sugar cane, family agriculture crops 4 Irregular rainfall (heavy rain followed by extended dry spells) (rainfed sugar cane, family Expected impact (1 low, 4 critical) 3.5 agriculture crops) Exotic diseases bananas and grapes Sugar and ethanol price cost uncertainty 3 Exotic diseases sugar cane Pesticide application without the needed technical knowledge 2.5 (fruit) Price volatility fruits Severe drought irrigated sugar cane Pests and diseases family 2 agriculture crops (maize, beans, cassava) 1.5 Irregular rainfall fruit Pests and diseases fruits 1 0% 5% 10% 15% 20% 25% 30% 35% 40% Probability Size of bubble determined by VBP of subsector Agricultural Sector Risk Assessment: Methodological Guidance for Practitioners 49 BOX 9.1. Continued FIGURE B9.1.2. RISK INCIDENCE BASED ON LIKELIHOOD, EXPECTED IMPACT, AND RELATIVE ECONOMIC RELEVANCE OF SUBSECTORS WEIGHTED BY THE NUMBER OF FAMILIES IN THE FAMILY AGRICULTURE SECTOR 4.5 Severe drought rainfed sugar cane, family agriculture crops 4 Irregular rainfall (heavy rain followed by extended dry spells) Expected impact (1 low, 4 critical) (rainfed sugar cane, family 3.5 agriculture crops) Exotic diseases bananas and grapes Sugar and ethanol price cost uncertainty 3 Exotic diseases sugar cane Pesticide application without the needed technical knowledge 2.5 (fruit) Price volatility fruits Severe drought irrigated sugar cane Pests and diseases family 2 agriculture crops (maize, beans, cassava) 1.5 Irregular rainfall fruit Pests and diseases fruits 1 0% 5% 10% 15% 20% 25% 30% 35% 40% Probability Size of bubble determined by VBP of subsector and number of farmers in the family agriculture sector Source: World Bank 2014a. 50 Agriculture Global Practice Discussion Paper CHAPTER TEN TOWARD A RISK MANAGEMENT STRATEGY Whereas the ASRA steps detailed in the preceding chapters can be standardized to some extent, the incorporation of risk solutions into government policies and plans is a highly country-specific process that cannot be standardized. Each country has its own decision-making process that takes place in a unique political economy. Chapter 10 discusses different approaches for how the risk solutions identified in the ASRA can be developed into action plans incorporated into stakeholders’ strategies to reduce agricultural risks. ACTIONS AND STAKEHOLDERS Effective risk management typically requires a combination of measures, some designed to remove underlying constraints and others designed to address risk directly. Resource availability will often determine what is possible. Whatever the case, inte- grated risk management programs are more effective than stand-alone programs. Translating the risk management strategy (whether mitigation, transfer, or coping) into concrete action requires the use of several kinds of implementation instruments. These instruments fall into three categories: agricultural investments, TA, and policy support. Some examples of risk management measures are highlighted in table 10.1. Risk management activities involve action by either individual or multiple stakehold- ers. Examples of activities that involve action by a single stakeholder include a farmer adopting agricultural practices to manage weather, pest, or disease risk or a trader using a back-to-back trading strategy or hedging to manage price risk. Other risk man- agement strategies necessitate collaboration between two or more actors in the supply chain. These include sharing market information to manage price risk or establishing better contractual relationships and information sharing to manage counterparty risk. The private sector can play an important role in managing the types of risk that can be handled through market tools such as insurance and futures markets. Futures contracts can be used to hedge price risk, for example. Besides private sector stakeholders, actions by public institutions, which potentially benefit many private actors, may be critical for successful risk management. These include strategies such as developing early warning Agricultural Sector Risk Assessment: Methodological Guidance for Practitioners 51 TABLE 10.1. EXAMPLES OF RISK early warning system is pursued at the government level. MANAGEMENT MEASURES Both actions seek to address the same risk, but are more effective when implemented together. Action Example Reduce the likelihood of Vaccination programs, event (mitigation) mass spraying programs, KEY ELEMENTS OF THE RISK good agricultural practices, physical infrastructure SOLUTIONS PROCESS (spillway, reservoir, dams, and A standardized risk solutions assessment is not feasible so on) because of the diversity of risks, approaches to man- Lead to adaptation Drought- and flood-resistant age them, supply chain contexts, and country contexts. (mitigation) varieties, changing crop Despite this diversity, some generic actions can be taken composition, agriculture while the government engages in the process of develop- diversification, and so on ing a risk management strategy and action plan. Reduce the losses (mitigation) Application of pesticides, insecticides, agricultural 1. Assistance of experts. Different from the practices, management ASRA team that identified the short list of solu- practices tions, a “new” team led by experts specialized in Reduce the spread of risk Quarantine programs, early each topic related to the short list of solutions can (mitigation-coping) warning and information be formed. dissemination, management 2. Discussion with stakeholders. Extensive approaches consultations and broader and more formal stake- Compensate the actor Agricultural insurance, holder discussions can be held in which the team (transfer-coping) buying futures and options shares the preliminary proposals with stakehold- for hedging, social safety net ers, including universities and specialized nongov- programs, and so on ernmental organizations (NGOs). 3. Evaluation of action options. The ASRA often results in a “laundry list” of possible activi- systems, developing and disseminating information about ties that could be undertaken to manage the iden- effective risk management, investing in agricultural exten- tified risks. During the risk solutions assessment, sion systems, conducting research and development on an exhaustive list of all possible actions to manage weather- and disease-resistant crop varieties, and investing the priority list should be developed. This should in infrastructure. When governments assume a holistic be followed by detailed evaluation of the selected approach and improve their ability to assess risks and their options and the current management practices interlinkages, they can more effectively help farmers man- and supporting programs and policies (in addi- age their risks by providing targeted information and train- tion to the stocktaking of projects and programs ing where they are most needed. carried out during the first phase). This exercise will help to consolidate the initial proposal and Risk management actions can be implemented at both bring forth operational issues, trade-offs, possible the micro and macrolevel. Microlevel actions are under- overlaps, and technical complexities. Along with taken by individual producers or a community, with the this, it is important to analyze underlying policy risk management decisions tailored to protect assets and or institutional bottlenecks that could hamper the improve resiliency. Macrolevel actions are implemented at achievement of desired objectives and to suggest a national level whereby ARM strategies are incorporated measures to overcome bottlenecks. into sectoral growth and investment and policy decisions. 4. Creation of a risk management strategy. To illustrate, mitigation strategies to minimize the effects On the basis of this evaluation, the next step is to of drought on crops may require producers to plant identify the strategic actions, objectives, and insti- drought-resistant cultivars, whereas development of an tutional implications. Rather than a stand-alone 52 Agriculture Global Practice Discussion Paper FIGURE 10.1. SOLUTIONS ASSESSMENT STEPS Focus on solutions that Stocktaking of current address main underlying Short list of solutions interventions and gap causes of risk and fill the analysis gaps strategy, this should be an integral part of the broader sector development strategy. STOCKTAKING AND GAP 5. Development of an implementation plan. ANALYSIS A detailed implementation plan, with clear mile- The gap analysis is based on a stocktaking of current proj- stones, time lines, sequencing of activities, and ects and programs that, intentionally or not, address the outcomes, is required to operationalize the strat- risk management issues identified in the short list of solu- egy. This implementation plan should have clear tions. The final result is a package of interventions action items for the public and private sector and designed to lower volatility, increase resilience in agricul- should be developed with broad consensus. ture, and protect vulnerable stakeholders. 6. Implementation. The government will lead implementation of the plan, which will require The stocktaking of projects and programs should incor- a coordinated effort from all stakeholders. The porate all the information needed to diagnose current process will involve generating and allocating interventions related to the short list of solutions. Table resources, allocating responsibilities for imple- 10.2 illustrates a useful way to summarize the findings. mentation, monitoring regularly to track progress, Table 10.3 illustrates the gaps analysis done for Tanza- and developing indicators to measure the sector’s nia’s ASRA to start narrowing specific solution areas that reduced vulnerability to risks. tackle the key risk issues. The table contains a number of projects and programs that were already addressing some RISK MANAGEMENT of the risk factors identified. The table indicates their con- INTERVENTION AREAS nection with the ASRA results and the potential gaps to Based on the ASRA results, this step identifies and evalu- be covered with specific risk management actions in addi- ates possible action options and then selects a strategy or tion to the existing ones. combination of strategies to reduce the risks and their negative effects on income and welfare. Generally follow- The solutions should be assessed by a team of specialized ing the risk prioritization, the solutions assessment will fol- experts to develop full proposals for the risk management low three steps (figure 10.1). actions that will make up the core of the risk management TABLE 10.2. ILLUSTRATION OF PROJECT STOCKTAKING Objective, Components Target and Key Financing Time Group Activities (US$) Name of Frame and Area (Indicate any Executing (Government, Project (Start- (Region, Specific Risk Unit and Main Cost by Grant, Loan, or Terminate) District, Management Institutional Components Donors, and Program or Pipeline and so on) Activity) Arrangements (US$) so on) Comments Agricultural Sector Risk Assessment: Methodological Guidance for Practitioners 53 TABLE 10.3. GAP ANALYSIS CONDUCTED FOR TANZANIA’S ASRA Statement on Additional Risk and Current Projects or Risk-Solving Activities Grading Solution Programs Perspective Gap? Proposed Drought— Irrigation systems, ASDP: A total of 353 Prospective sector Expansion of None. critical or land and water irrigation schemes were risk reduction impact coverage to considerable management. upgraded, rehabilitated, or is small because build up from and probable newly developed. nonmassive type of current projects’ or highly TAFSIP: Irrigation investment. experiences. probable risk. development, sustainable water and land use management. Feed the Future Program: Increase area under irrigation by 15.5% through the development of 7 smallholder irrigation schemes in Morogoro and Zanzibar. More extensive use There are available drought- Reduction of yield Planting Specific program of drought-resistant tolerant cultivars in Tanzania variability and crop materials and to be included cultivars. (for example, maize) and there losses but learning research results in set of specific is research under way (for required for optimal are available, proposals. example, coffee), but there is balance between risk information and low adoption. reduction and high promotion are productivity. missing. Pests and Good agricultural ASDP: Strengthening Existing research To build upRedesign current diseases— practices to address agricultural research and results need to be programs from the ASDP critical or drought and pest training. disseminated to and expand subprograms. considerable and diseases. Feed the Future Program: farmers. Specific geographically to and probable Agricultural support services knowledge is required cover the entire or highly and capacity building including to address pest and country or new, probable risk. research and development and disease prevention more specific financial services. and control. technology transfer program. Introduce disease- No comprehensive program Reduction of yield Progress in coffee, Specific program resistant cultivars in place. There are available variability and crop for instance, but no to be included and planting disease-resistant cultivars losses, depending on program covering in set of specific material. (coffee, and so on). the crop. most crops. proposals. Price Improved Agricultural Marketing Project-related No hedging Reforms needed volatility— understanding Systems Development interventions can instruments and to deal with moderate to of price risk Program: (a) Agricultural have good results in institutional price volatility critical and management, marketing policy terms of targeted framework and to deepen probable market information, development; (b) small stakeholders available at into institutional and highly and hedging. producers’ empowerment by but massive the producer arrangements and probable, Trade policies. building their entrepreneurial and sustainable and primary current roles of mostly cotton, Contract farming. and organizational capacity achievements require cooperative public and private cashew nuts, Improved storage. and improving their links to nationwide policies levels. sector. and coffee. Infrastructure markets; (c) introducing a and institutional development. warehouse receipt system. buildup. 54 Agriculture Global Practice Discussion Paper TABLE 10.3. Continued Statement on Additional Risk and Current Projects or Risk-Solving Activities Grading Solution Programs Perspective Gap? Proposed Improved efficacy Allowing the small farmers of commodity using the warehouses to councils. obtain loans for the period between harvest and sale; and (d) the development of rural marketing infrastructure, including storage facilities, marketplaces and roads. ASDP: marketing and private sector development. Improving overall sector policy, regulatory and legal framework. Maize short- Develop clear, USAID Tanzania’s Sera Difficult to have Policy framework Find adequate term policy long-term, efficient, Policy Project (launched impact from single still suffering policy equilibrium variability— and transparent in 2011), which focuses on projects if policy of great between the considerable commodities advancing policy reforms that framework is weak. variability and short-term food and probable and sectoral are critical to the agricultural discretionarily security goal and risk. development sector. (because of food the long-term policies. security goal) and productivity therefore provides growth aim. poor incentives to invest in production. Source: World Bank 2013b. Note: ASDP = Agricultural Sector Development Program; TAFSIP = Tanzania Agriculture and Food Security Investment Program. strategy. Ideally, a detailed and exhaustive cost-benefit plan to be incorporated in government policies. To analysis would inform the selection of the most appropri- accomplish the stated objective, the solutions team con- ate intervention options. But conducting a cost-benefit sults available literature, diagnoses the issues, and con- analysis of a multitude of options can be costly and time ducts interviews with various stakeholders related to the consuming. Alternatively, the use of prioritization filters risk solution under analysis. Most of the risk assessment that can be discussed in workshops with stakeholders can work is done in the field and requires extensive teamwork go a long way, not just in ownership of the results but also and close cooperation with stakeholders and government in arriving at a set of solutions on which the stakeholders at different levels. themselves agree. To prepare a risk management strategy and action plan, STEPS TO PREPARE THE RISK this phase of the fieldwork requires that the team do the following: MANAGEMENT STRATEGY 1. Review the key ARM issues identified during the AND ACTION PLAN ASRA with respect to the specific relevant areas The ASRA’s findings serve as the basis for identifying and of analysis (in the foregoing Tanzania example, designing an ARM strategy and corresponding action they would be utilization of drought-tolerant Agricultural Sector Risk Assessment: Methodological Guidance for Practitioners 55 seeds, appropriate agricultural practices, maize 6. Bundle those recommendations into elements of a trade, and cotton and coffee price volatility. This strategy, in terms of policy and regulatory frame- assessment would include an in-depth review of work issues, investments, capacity building and aspects such as: nature of the risk problem, fre- TA, and institutional reform, as appropriate. quency and actual or potential impact of risks, 7. Prepare a comprehensive plan of action, including longer-term sector trends, and fiscal and macro- the following information by strategic lines: activi- economic repercussions, as well as poverty and ties, responsible institutions, period of execution, food security implications. resources needed, and estimated cost. 2. Revise and complete the stocktaking of existing programs or policies that in some way address the CONTENTS OF THE RISK identified risk management issues contained in the ASRA. MANAGEMENT STRATEGY 3. Assess the related current regulatory and institu- AND ACTION PLAN tional framework and detect the main limitations The risk management strategy should contain the follow- that prevent local institutions (public and private) ing elements at a minimum: from performing effectively as vehicles for risk management instruments in the relevant areas. Risk management framework, including the fol- 4. For each relevant area, identify specific action lowing: options to address the risk management problem » Brief information about the concerned sector and over the short to medium term, considering the the risks in the specific area of analysis potential for market-driven solutions and govern- » Stocktaking of projects and programs ment interventions, while highlighting linkages » Discussion of the possible risk management action with the current public policy framework. Assess options to approach the identified risks and justifi- the positive aspects and gaps in existing public cation of the chosen option policies to mitigating, transferring, or absorbing the concerned risks. Description of the risk management strategy 5. Share findings and proposals with the team to with respect to the specific areas of analysis, including as identify a set of solutions that may be aggregated appropriate the following: in one or a few investment projects or program » Aim, scope, beneficiaries, main elements, institu- profiles. Work together in drafting the profiles of tional framework, connections with national poli- the proposed aggregated projects or programs and cies, and so on (end goal of the recommendations policies. or strategies, what or whom they are informing) TABLE 10.4. EXAMPLE OF AN ACTION PLAN STRUCTURE Strategic Responsible Period Results Line/Objective Actions Institution (Quarter/Year) Resources Cost (US$) Indicator A. A1 A2 A3 B. B1 B2 56 Agriculture Global Practice Discussion Paper » Anticipated risk management gains programs or through discussions with staff from projects » Concrete actions planned for the short and medium or the government. A second set of tables providing infor- term mation about the short-term activities by executing unit could be most useful for discussions regarding individual As the road map for the execution of the actions con- responsibilities and for allocating budget resources at the tained in the strategy, the action plan could be organized institutional level, as illustrated in table 10.4. as follows: a detailed presentation of the actions and their objectives with indications of the implementation period, The action plan can also contain complementary policy resources needed, and costs. This information may not be recommendations that may have been identified by the easily available but could be obtained by looking at similar ASRA team or are part of the government pipeline. Agricultural Sector Risk Assessment: Methodological Guidance for Practitioners 57 CHAPTER ELEVEN FINAL CONSIDERATIONS Agricultural risk management is an issue of development effectiveness, not just a mat- ter of kindness to farmers. The increasing presence of risk in the agricultural sector may not just exacerbate volatility, but may also become a major impediment to devel- opment, economic growth, and poverty reduction agendas. The nature and impor- tance of risk issues affecting economic growth and poverty reduction vary from country to country. So, too, do stark differences exist in a country’s capacity to manage agricultural risks. Disparities in risk vulnerability in the agricultural sector tend to be greater in lower-income countries, and, within countries, greater among poor than more affluent stakeholders along agricultural supply chains. Thus the need to tailor each ASRA to suit the particular realities of developing countries, as one size does not fit all. These guidelines should be considered a practical approach rather than a rigid “bulletproof ” methodology. THE OPERATIONAL EXPERIENCE The World Bank has gained notable experience in integrating agricultural risk issues into country work and lending over the last decade. Attention to agricultural risk issues in Country Assistance Strategies (CAS) has also increased during the same period. Likewise, ASRAs’ findings have informed various lending operations at the project level, either as stand-alone projects or as components of a larger intervention to increase sector productivity and resilience (box 11.1). Much work still needs to be done to establish an enabling environment that will foster country-led, country-specific strategies for mainstreaming agricultural management risk measures into agricultural policies and investments. The opportunities for improv- ing the development impact through risk mainstreaming include making development interventions more responsive to country risk conditions and commitments; making these interventions more strategic; and improving the alignment of policies, processes, and resources to support such interventions. Agricultural Sector Risk Assessment: Methodological Guidance for Practitioners 59 BOX 11.1. THE INFLUENCE OF ASRAS ON chains and identify key priority risk management WORLD BANK OPERATIONS actions that can inform (i) sector policy planning and development; and (ii) the country assistance The World Bank, in collaboration with Niger’s 3N Initia- program; tive, a national food security strategy led by the president’s » Develop and implement, as part of the country office, conducted an ASRA in 2012 to help prioritize risks and solutions to building resilience in Niger’s agricultural assistance program, priority policy and operational sector. The process resulted in the government of Niger interventions that respond to the ASRA; and developing the Plan d’action pour la gestion des risques agricoles » Monitor the implementation and results of these au Niger (PAGRA). The 3N Initiative considers the PAGRA policy and operational interventions. a critical tool for long-term planning in Niger, which suf- fers from frequent shocks and losses from agricultural risks. This 10-year action plan (2014–23) sets short-, medium-, THE IMPORTANCE and longer-term targets, with the overriding goal of OF AN ASRA strengthening the resilience of rural and semi-urban com- munities against the main agricultural risks. The World A key component of successful mainstreaming of risks Bank is supporting the government’s efforts to operational- into agricultural sector strategies is a sector-level risk ize PAGRA with a US$116 million Climate-Smart Agricul- assessment that identifies and prioritizes critical areas in ture Support Project. The government of Niger is working which risk-responsive actions are likely to enhance growth, toward implementing PAGRA, developing coordinating contribute to poverty reduction, and strengthen resilience structures, identifying good practices, planning for scale-up to external shocks. In line with this aim, the methods to of interventions, setting quantitative targets and identifying carry out an ASRA as established in this document need target groups, and sharing experiences at different levels of government. Niger’s experience operationalizing risk man- to be flexible and adapted to each country’s circumstances. agement could help inform other countries’ efforts toward The ASRA may, for example, be a stand-alone document, building resilience. a section of a country economic analysis, or part of agri- cultural development projects, programs, and strategies. In Kenya, the ASRA conducted by the World Bank resulted in a US$200 million lending operation on CSA and helped inform the government of Kenya’s Climate-Smart Agricul- The basic process is to diagnose the risk conditions that ture Program 2015-2030. inhibit growth, poverty reduction, and well-being in a particular country; based on this diagnosis, the risk- Paraguay’s ASRA informed the World Bank Country Part- nership Strategy (2015–18), which includes strengthening responsive development actions that would be strategic agricultural resilience and managing price volatility as one from the government’s point of view can be identified. of its three pillars. It is expected that a lending operation Where such actions involve donor assistance, risk consid- based on agricultural resilience will follow. erations can be integrated into development assistance The World Bank’s Rwanda Economic Update of February operations in interventions that the diagnosis suggests are 2015 focused specifically on agricultural sector risk assess- critical for managing risk and therefore poverty reduction ment to highlight the vulnerability of Rwanda’s agricul- or economic growth, and quality and outcomes can be tural sector. monitored. The World Bank has conducted ASRAs in 14 countries across four regions since 2013. These have contributed ONE SIZE DOES NOT FIT ALL to informed discussions, debates, and operations toward A singular blueprint for a risk management roadmap is building agricultural sector resilience in these countries. not feasible because of the diversity of risks, approaches to manage them, supply chain contexts, and country con- texts. A roadmap for managing the risk of a pest and dis- In some countries, it has proven effective to develop a ease outbreak could be quite different from one for basic process that involves working with stakeholders to managing counterparty risk. Nonetheless, despite the » Prepare an ASRA that analyzes the risk dimensions diversity of approaches, some generic steps can be of development across major agricultural supply adopted by all risk management approaches. 60 Agriculture Global Practice Discussion Paper This document, particularly chapters 7 through 9, pre- are not the biggest risks. An insurance expert will sents many examples of risk management strategies that be biased toward risk transfer solutions, whereas can be pursued by various groups of stakeholders. Similar an agriculture extension expert might be more to strategies pursued by the public and private sectors, inclined toward a risk mitigation solution. A good these strategies can and should be pursued in parallel and way to mitigate the bias is to use a process driven are complementary. At the micro level, risk management by data and empirical evidence and incorporate decisions can be uniquely tailored to protect individual different perspectives while prioritizing risks. producers’ or communities’ assets and improve their resil- » Risk appetite. It has been mentioned that poorer iency. At the macrolevel, ARM strategies can be incorpo- producers have fewer buffers and therefore tend to rated into sectoral growth and investment decisions. have a low risk appetite. Many of these households already engage in risk management strategies, CHALLENGES whether through diversification of crops or income, In conducting agricultural sector or supply chain risk but this may lead to a lower return on assets than assessments, some challenges do arise. These challenges, households would otherwise enjoy. According to primarily methodological, should be kept in mind while IFAD (2011), “Risk avoidance strategies thus have developing a risk assessment. They include the following: high opportunity costs: some studies estimate that » Historical bias. Assessments conducted to date have average farm incomes could be 10 to 20 percent been structured to analyze risk from a historical higher in the absence of risk.” perspective. This type of assessment provides lim- » Attribution. Attribution and proving the counterfac- ited opportunity for analyzing new risks that might tual have always been challenging and particularly occur in the future, whose frequency of occurrence complex in social sciences, unless clear-cut data are is zero (as yet unrealized risks). If these events available.4 It is rare that one event leads to losses were to occur in the future, they could lead to such that all losses can be attributed to it. In many catastrophic losses for the supply chain. The team cases, multiple risk events occur in a given year, should develop a mechanism to incorporate poten- which makes it difficult to attribute proportions to tial risks, and thus modify the framework described different risks. Furthermore, as many of these risk herein. events are interlinked through the causal chain, it » Data limitations. Accessibility and reliability of good- is difficult to isolate the losses from these events. quality, disaggregated, time series data on key For example, in Ghana in 2005/06, national pro- variables (production, cropped area, yield, prices, duction of maize dropped by 45 percent (350,000 major weather events, market events, policy deci- tons). Deeper analysis indicated that in the 2005/06 sions, and so on) are often very limited, which neg- season, some areas experienced droughts, there atively affects the ability to conduct a thorough risk was an attack of armyworm, cross-border armed assessment. conflict arose in the north, and bush fire destroyed » Subjective bias. Despite all efforts to bring objectivity large areas of savannah. Establishing the propor- and give clear analytic indicators, a broad ASRA tion of losses and attributing them to these four is essentially a subjective tool and this is one of the separate, but perhaps interlinked, risks would have more challenging aspects, especially during risk been quite challenging. prioritization. Depending on the experience of the » Client expectations. An ASRA is an intermediate pro- team leader and members, the tendency will be to cess for managing risk, not a final product. Govern- push specific agendas. An entomologist will tend ment political will and commitment are necessary to prioritize insect pests as the biggest risk, a plant to translate the suggested solutions into concrete pathologist might highlight diseases, a weather government strategies and budgeting. expert may push for flood and drought, and a com- modity price-hedging expert might tend to high- 4 Counterfactual means to prove the fact that the loss would not have happened if light price volatility, even in instances when these the event had not occurred. Agricultural Sector Risk Assessment: Methodological Guidance for Practitioners 61 These are some of the major challenges that limit the integration of a stronger risk management focus into sec- assessment process and its application. Nevertheless, tor policy planning and development. In the face of mul- ways exist to deal with these challenges. One good prac- tiple risks, the resilience of primary producers, tice is to be aware of and upfront in highlighting these agribusinesses, institutions, and other supply chain stake- issues and to be consistent and explicit about the associ- holders for collective action, coordination, and public- ated assumptions. private cooperation is a critical consideration. One cannot understand the current competitiveness and future poten- A LAST WORD tial of the agricultural sector in developing economies This methodological guidance hopefully provides a sys- without understanding stakeholders’ ability to anticipate temwide approach to identify risks, risk exposure, the and respond to shocks. An agricultural sector develop- severity of potential losses, and options for risk manage- ment strategy that focuses solely on productivity and ment for the private and public sectors. Ultimately, the removal of constraints is incomplete if it ignores risk and ARM framework is designed to inform and facilitate the risk management considerations. 62 Agriculture Global Practice Discussion Paper REFERENCES IFAD (International Fund for Agricultural Development). 2011. “Rural Poverty Report.” IFAD. Rome, Italy. Jaffee, S., P. Siegel, and C. Andrews. 2008. “Rapid Agricultural Supply Chain Risk Assessment: Conceptual Framework and Guidelines for Application.” Commod- ity Risk Management Group, Agriculture and Rural Development Department. World Bank, Washington, DC. Jones, P. G., and P. K. Thornton. 2008. “Croppers to Livestock Keepers: Livelihood Transitions to 2050 in Africa due to Climate Change.” Environmental Science & Pol- icy 12 (4): 427–37. WFP (World Food Program). 2009. Emergency Food Security Assessment Handbook. Rome, Italy: WFP. World Bank. 2008. Agriculture for Development. World Development Report 2008. Wash- ington, DC: World Bank. ———. 2011. “Cocoa Supply Chain Risks Assessment in Ghana.” World Bank, Washington, DC. ———. 2012. “Turn Down the Heat: Why a 4°C Warmer World Must Be Avoided.” World Bank, Washington, DC. ———. 2013a. “Mozambique Agricultural Sector Risk Assessment.” World Bank, Washington, DC. ———. 2013b. “Agricultural Sector Risk Assessment for Tanzania.” World Bank, Washington, DC. ———. 2013c. “Niger Agricultural Sector Risk Assessment.” World Bank, Washing- ton, DC. ———. 2014a. “Agricultural Sector Risk Assessment for the State of Paraiba, Brazil.” World Bank, Washington, DC. ———. 2014b. “Malawi Agricultural Sector Risk Assessment.” World Bank, Wash- ington, DC. ———. 2014c. “Rwanda Agricultural Sector Risk Assessment.” World Bank, Wash- ington, DC. ———. 2014d. “Senegal Agricultural Sector Risk Assessment.” World Bank, Wash- ington, DC. Agricultural Sector Risk Assessment: Methodological Guidance for Practitioners 63 APPENDIX A GLOSSARY OF TERMS Agricultural risk: Agricultural risk is a combination of the likelihood of the occur- rence of a hazardous event or exposure(s) and the severity of losses that can be caused by the event or exposure(s). The three main attributes of risk are event hazard, uncer- tainty, and losses. Climate change: Climate change refers to a statistically significant variation in either the mean state of the climate or its variability, persisting for an extended period (typically decades or longer). Climate change may be due to natural internal processes or external forces or to persistent anthropogenic changes in the composition of the atmosphere or in land use. In its Article 1, the Framework Convention on Climate Change (UNFCCC) defines climate change as a change of climate which is attributed directly or indirectly to human activity that alters the composition of the global atmos- phere and that is in addition to natural climate variability observed over comparable time periods. Climate: Climate in a narrow sense is usually defined as the average weather, or more rigorously, as the statistical description in terms of the mean and variability of relevant quantities over a period of time ranging from months to thousands or millions of years. The classic period is 30 years, as defined by the World Meteorological Organiza- tion (WMO). These quantities are most often surface variables such as temperature, precipitation, and wind. Climate in a wider sense is the state, including a statistical description, of the climate system. Constraints: These are conditions that lead to suboptimal performance. For example, low yields (symptom) might be caused by lack of access to inputs or poor technology. Enabling environment risk: Sudden changes in the given scenarios where busi- nesses take place that can lead to financial losses among agricultural stakeholders (for example, unexpected changes in government or business regulations, the mac- roeconomic environment, political risks, conflict, trade restrictions, logistics, and corruption). Agricultural Sector Risk Assessment: Methodological Guidance for Practitioners 65 Extreme weather event: An extreme weather event is Risk mitigation: This term describes activities designed an event that is rare within its statistical reference distribu- to reduce the likelihood of an adverse event or reduce the tion at a particular place. Definitions of “rare” vary, but severity of losses. It refers to activities leading to improved an extreme weather event would normally be as rare as or resilience to withstand external shocks, through ex ante rarer than the 10th or 90th percentile. By definition, the preparation to sustain production and livelihoods follow- characteristics of what is called extreme weather may ing an event. Risk mitigation options are numerous and vary from place to place. An extreme climate event is varied. Examples include crop and livestock diversifica- an average of a number of weather events over a tion, income diversification, irrigation, use of disease- and certain period of time, an average that is itself extreme stress-resistant cultivars, improved early warning systems, (for example, rainfall over a season). avoidance of risky practices, and adoption of improved agronomic practices such as soil drainage and mulching. Market risk: Market risk is related to issues that affect It is often confused with mitigation to climate change, price, quality, availability, and access to necessary prod- which refers to anthropogenic interventions to reduce the ucts and services. Prices for inputs and outputs can be sources or enhance the sinks of greenhouse gases. highly volatile, particularly in commodity markets in which both local and global supply and demand condi- Risk transfer: These are tools and mechanisms for tions constantly change. Other risks include exchange rate transfer of the potential financial consequences of par- and interest rate volatility and counterparty and default ticular risks from one party to another willing party, usu- risks. ally for a fee or premium. Although insurance and hedging are well-known forms of risk transfer, in developing coun- Production risk: This refers to risks that are present at tries the use of informal risk transfer within families and farm level and affect yields. A large number of risks affect communities is also extremely important. the volume of production of agricultural commodities. These include nonextreme weather events (too little rain- Risks: These are uncertain events that have the proba- fall, too much rainfall, hail, frost, low temperature, and so bility to cause losses. As a symptom, yield volatility might on). They also include less frequent but catastrophic be caused by a drought, pest outbreak, or disease out- weather events (severe floods, droughts, hurricanes, break. cyclones, and so on). Outbreaks of pests and diseases can also adversely affect yields, as can damage by animals, fire, Trends: These are longer-term or “chronic” patterns and wind. (reversible or irreversible) that provide context. Here, the declining yield symptom might be caused by structural Resilience: It is defined as the ability to withstand, changes in agriculture or changes in climatic patterns. recover from, and reorganize in response to crisis so that all members of society may develop or maintain the abil- Uncertainty: This is an expression of the degree to ity to thrive. which a value (for example, the future state of the climate system) is unknown. Uncertainty can result from lack of Risk coping: This term describes actions that help information or from disagreement about what is known or affected households and the government absorb loss. even knowable. It may have many types of sources, from They usually take the form of compensation (cash or in- quantifiable errors in the data to ambiguously defined kind), social protection programs, livelihood recovery pro- concepts or terminology, or uncertain projections of grams (for example, government assistance to farmers, human behavior. Uncertainty can therefore be repre- debt restructuring, contingent financing), safety net pro- sented by quantitative measures (for example, a range of grams, buffer funds, savings, strategic reserves). Such values calculated by various models) or by qualitative interventions are often financially beneficial, and the abil- statements (for example, reflecting the judgment of a team ity to respond quickly to events often reduces losses. of experts). 66 Agriculture Global Practice Discussion Paper Volatility: This is a measure of the dispersion of obser- Vulnerability: This is defined as the capacity of stake- vations from the trend value. Volatility can either be meas- holders to manage agricultural risk and recover from ured by using the standard deviation or the variance external shocks. between observations. Commonly, the higher the volatil- ity, the riskier the set of observations. The concept of Weather: This is the current atmospheric condition in a volatility risk has been consistently applied in the financial given place; it includes variables such as temperature, literature as the risk of a change of price of a portfolio as rainfall, wind, and humidity. a result of changes in the volatility of a risk factor. Agricultural Sector Risk Assessment: Methodological Guidance for Practitioners 67 APPENDIX B AGRICULTURAL SECTOR RISK ASSESSMENTS CONDUCTED BY THE WORLD BANK TABLE B.1. ASRA COUNTRY AND ASSESSMENT YEAR Country Assessment Year 1. Niger 2012 2. Ghana 2013 3. Tanzania 2013 4. Mozambique 2013 5. Paraguay 2014 6. Mongolia 2014 7. Kenya 2014 8. Senegal 2014 9. Malawi 2014 10. Rwanda 2014 11. Kazakhstan 2014 12. Tajikistan 2014 13. Kyrgyzstan 2014 14. Brazil (State of Paraiba) 2014 15. Brazil (State of Bahia) 2015 Agricultural Sector Risk Assessment: Methodological Guidance for Practitioners 69 APPENDIX C DATA REQUIREMENTS FOR AGRICULTURAL SECTOR RISK ASSESSMENTS Access to disaggregated, reliable, good-quality time series (preferably 20–30 years) data is a precondition for conducting an agricultural sector risk assessment. It is preferable that all statistics and data be provided at the disaggregated departmental or provincial level. 1. Agricultural structure of the country: • Composition of agricultural GDP, highlighting major commodities and percent share of different commodities (crop, livestock, fisheries, and so on) in agri- cultural GDP; • Time series data on acreage, production, yield of major commodities (crops and live- stock) that account for more than 80 percent of the country’s agricultural GDP; • Agricultural sector composition (by landholding size of the country); • Description of major production zones: Brief description of major agro-metrologi- cal zones, crop production and animal husbandry systems, land use patterns, description of irrigated versus rain-fed agriculture, and so on; • Agricultural census: Past four to five agricultural censuses; • Livestock census numbers by class of animal or latest census by region; • Rural household data: Number of crop (and livestock) households per region, average farm size, and so on; and • Cropping calendar for key crops grown in each region, including for each sea- son and major crop type, planting dates and harvest. 2. Crop production data: The following data are needed, broken down by crop season, summer and winter, and disaggregated by region or province. These data should be provided for those 7–10 agricultural commodities (including livestock and crops) that account for more than 80 percent of the country’s agricultural GDP: • Commodity production: Total volume of commodity produced (MT); • Commodity acreage: Total acreage under the commodity in question (sown and harvested areas); • Commodity yield: Provincial yield data, distinguished by the nature of the pro- duction system; Agricultural Sector Risk Assessment: Methodological Guidance for Practitioners 71 • Cropping calendar of the commodity for each re- 6. Livestock mortality statistics: It is hoped gion, including planting dates and harvest dates that the Ministry of Agriculture, Animal Health for each season and major crop type; and Department, or Veterinary Department records • Brief description of major agro-metrological mortality data for livestock by type of animal and zones, land use patterns, and irrigated versus by cause of loss in each province. If so, as many rain-fed agriculture, and so on, for the com- years as possible should be accessed. modity. 7. Meteorological data (time series data for 3. Market data: The following time series (30 past 30 years): Information about occurrence years) data are needed, disaggregated by depart- and impact of major natural hazards includ- ments or regions of the country for all major com- ing droughts, floods, hurricanes, locust upsurges, modities (7–10) that account for 80 percent of the major disease outbreak, and so on. Access to country’s agricultural GDP: monthly/decadal data on rainfall, temperature, • Retail prices; and other important weather variables. • FOB/export prices; 8. Government support to agriculture: Gov- • Farm-gate prices; ernment Disaster Relief Program? If yes: • Input prices: for fertilizer (diammonium phos- • Organization responsible for implementing; phate, urea, and so on), diesel, labor, major pes- • Events for which compensation is paid; ticides and insecticides, and so on; • Criteria for assessing losses and compensation • Value and volume of commodity exported and levels; imported; • Compensation payments by event (for example, • Value and volume of commodity processed; earthquake, hurricane, drought, flood) for the • Details on cost of production and margins, past 10 years: breaking even along the chain; Input price subsidies: If yes, details; • Currency exchange rates of major importing Output price support: minimum prices: If and exporting countries; and yes, details; • Interest rates. Other forms of government support to crop 4. Enabling environment and livestock producers (for example, taxa- • Time line of major political events/disturbanc- tion policy); and es that might have affected the commodity; Insurance subsidies: If yes, details. • Major policy changes/regulatory changes that 9. Trade data: Time series data (past 15–20 years) could have affected the commodity (description of of the change, as well as date of change); • Major import crops—volumes, price data (de- • Major changes in market regulations; livery prices/retail prices), country of origin, • Major disruptions in the supply chain and the and so on; cause and effect thereof; • Major export crops—volumes, price data (FOB • Time series inflation (general and food) data; and and farm-gate prices), major destinations; and • Agricultural GDP growth rate data. • Domestic crops—volume of interregional trade, 5. Commodity risk assessment and produc- farm-gate prices/retail prices. tion loss data: It is hoped that the Ministry of 10. Risk management intervention landscape: Agriculture or another ministry records commod- A number of ongoing government, development ity damage on an annual basis for major events in partner, and private sector-supported initiatives/ each province and cause of loss. Data for the past programs may exist in the area of ARM (mitiga- 15–20 years are very useful for this assessment. tion, transfer, or coping). Please provide a brief These data should also include major occurrences description of the interventions (nature of inter- of pest and disease outbreaks and their impacts on vention, coverage, budget, implementing agency, commodity production. and so on). 72 Agriculture Global Practice Discussion Paper APPENDIX D GUIDELINES FOR FOCUS GROUP DISCUSSIONS WITH FARMERS Basics. A focus group comprises people who have attributes in common and who are able to provide information and opinions about the topic or subject that is the focus of discussion. Focus groups may be homogeneous or heterogeneous depending on the purpose of the focus group meeting (for example, to identify a pattern, a homogeneous group is more appropriate, whereas to ascertain perceptions of risks affecting a whole supply chain, a heterogeneous group is more appropriate). Although the Ministry of Agriculture counterparts select the focus groups for an ASRA, the assessment team should » Be careful that no significant power differentials exist among group members, as this often results in influential people (often men, or better-off individuals) dominating the discussion. » Aim to cover a mixed sample of representative farming entities covering (i) small and (ii) medium farmers.5 Within this continuum, considerable differ- entiation may still exist within groups and should be accounted for. For example some small farmers (possibly in different regions) may have low input use (for example, limited inputs, family labor) versus high input use (for example, higher input use, hired labor). The level of commercialization may also vary among medium and large enterprises and regions in the country. » Keep the size of the focus group manageable. A focus group is most effective with 10–15 participants. All three field teams will conduct as many farmer focus group discussions as possible: on the major cash crop, on all food crops, and one more cash crop at the discretion of the field team member. 5 Large-scale farmers will be interviewed separately (key informant interviews). Agricultural Sector Risk Assessment: Methodological Guidance for Practitioners 73 Purpose. A focus group discussion is one tool (alongside Stage One: Preparing for a Focus Group Dis- literature review, secondary data collection, key informant cussion interviews, and so on) to help assess risk. Its purpose is to Prior to engaging in a focus group discussion, the field help the team: team needs to: » Identify risks (three principal types of agricultural » Memorize the key purpose of this discussion and risk: production, market, and enabling environ- key questions (please see guide and script below); ment risk); » Appoint a facilitator or moderator (international » Analyze and quantify risks (primary stakeholders; field team member) who will guide the content and losses; frequency; underlying causes; risk transmis- process of the discussion; sion); » Appoint a note taker (local field team member) to » Prioritize risks (frequency of occurrence and sever- record the focus group feedback (the facilitator or ity of impact); and moderator can take notes too, but should be careful » Identify current management practices (that is, sale that this does not negatively affect the flow of the of assets, borrowing, sale of labor, reduced con- discussion if he or she stops to write things down); sumption, government handouts, migration). » Develop a facilitator’s or moderator’s guide or script; The purposes of the focus group discussions are to assess farmers’ » Confirm whether a translator is needed (local perceptions of the risks they face in the supply chain and to examine field team member should be able to assume this how these risks and negative impacts could be managed more effec- responsibility); and tively. » Have sufficient stationery, including copies of all report formats (to be developed), notebooks, clip- ! Remember: board (for drawing, writing, or posting), tape, pens, Agricultural risk is defined as a combination of the likeli- pencils, markers, and sticky notes. hood of an occurrence of a hazardous event or exposure(s) Developing a facilitator’s or moderator’s guide and the severity of losses that can be caused by the event or script. The script is a guide for the facilitator to or exposure(s). The three main attributes of risk are event explain to participants the purpose of the group, review hazard, uncertainty, and losses. the focus group rules, and provide other important infor- mation. Suggested script and guide template for all field teams to A constraint is a bottleneck, a condition that hampers the follow: smooth functioning of the supply chain. The main char- acteristics of a constraint include condition, certainty, and Opening (10 minutes): suboptimal performance. “Hello. I will introduce myself and then I would like each of you to tell us your name, what you do, and whether you Agricultural risks ordinarily vary from season to season, whereas have participated in a similar discussion before. My name longer-term trends and conditions are more often attributable to con- is (. . .). [Have the focus group members introduce themselves.] straints. Today we would like to have a conversation with you about agricultural risks in your community. The World ! Remember: Bank has been asked by the government to assess the risks Agricultural risks can be artificial or natural, and can be affecting the agricultural sector. We have selected major idiosyncratic (affecting individual supply chain partici- food and cash crop supply chains across the country, and pants), covariate (systemic) within the chain (affecting just like we are here today with you, there are two other multiple supply chain participants), or covariate outside teams in other parts of Tanzania doing the same with the chain (affecting chain participants and the broader farmers in [list some commodities and some of the other regions]. economy). The focus here is on unexpected, adverse, sys- What we are trying to accomplish before we leave here temic events. today is to get a better understanding of the risks that you 74 Agriculture Global Practice Discussion Paper face in the supply chain, and to examine how these risks Stage Two: Conducting a Focus Group Discus- and negative impacts could be managed more effectively. sion We plan that this will take us approximately two hours. Because the aim is to understand what, who, how, where, Are there any questions?” [Respond to participant questions.] and when are the greatest expected losses, it might be worthwhile at the beginning of the discussion to have focus group “Before we continue, let’s agree on the rules of the discus- participants engage in a wealth breakdown: the group can be sion [Decide together on whether you want focus group members to asked to consider four (usually) wealth groups of equal participate by raising their hands to speak; write their ideas on a piece size, ranging from the poorest 25 percent to the richest of paper.] I would like to tell you that everything we discuss 25 percent (or small, medium, and large farmers). Visu- here will be kept confidential. We are not here to check on ally this is represented with piles of rocks, marbles, or licenses or for tax collection purposes. We will summarize crops (maize, rice, beans), or can be drawn on the earth the things you tell us and combine it with other focus or a clipboard. This differentiation may help at the end groups we are meeting across Tanzania. My job here of the process to more easily identify the significance of today is to make sure we discuss all of the issues we expected losses for different participants in the chain rela- planned to discuss, and my colleague here will help us by tive to their assets, livelihood and enterprise strategies, taking notes.” and performance outcomes. “Let’s begin.” (40 minutes) Proceed with the opening according to the script and start the [All discussions should cover five main areas: (i) livelihood profile; discussion with the agreed method of participation. (ii) risk exposure; (iii) risk transmission along the supply chain; (iv) risk management; and (v) gaps and opportunities to address risk; and All field teams should be prepared to discuss the five main areas be structured: first a question, then a probe. Please see Stage Two.] identified above as (i) livelihood profile; (ii) risk exposure; (iii) risk transmission along the supply chain; (iv) risk management; and Closure (10 minutes) (v) gaps and opportunities to address risk. The accompanying “Are there final questions? [Respond to questions.] Let me table includes indicative questions under each area. This reiterate what I think we heard the risks that you face in is not a questionnaire to be strictly followed and com- the supply chain are, and how these risks and negative pleted. It is meant to help the facilitator or moderator impacts could be managed more effectively [reiterate]. (international field team member) keep the discussion Thank you for participating in our focus group today. We organized. Questions asked should generally be open will be collecting feedback from all the focus group discus- ended to stimulate both individual contributions and sions undertaken across Tanzania, and together with group interaction. Questions should not elicit “yes or no” other analysis, will include it in our discussion with the answers. government.” Livelihood Profile Question: What are the main crops you produce for own food Probe: Percentage of the area of value, total area farmed and consumption? For cash? For export? owned (ha), ranked in order of importance. Question: Is livestock important for you? Probe: Do you consume livestock products that you produce? Do you derive income from livestock products? What percentage of income? Question: What percentage of your household income is from Probe: Try to find out what role agriculture plays in the crops, livestock, other? What is your average level of production household livelihood. and revenue in a given year? What has been the trend in recent years? Question: What inputs do you use? Where do you source inputs Probe: Fertilizer, seeds, pesticides. Check for reliability, quality, from? utilization. Agricultural Sector Risk Assessment: Methodological Guidance for Practitioners 75 Livelihood Profile Question: Is there any type of irrigation used? What types? Probe: Furrow, drip, overhead, and so on. Question: What factors motivate the planting of crops? What are Probe: Returns to assets vs. risk management, substitute, the alternatives? complementary goods. Question: What access do you have to local markets and traders? Probe: Formal vs. informal markets. What is the distance to the nearest trading center? Question: What access do you have to financial resources? Probe: Formal vs. informal finance (credit, trade finance, personal). Question: Are you a member of a cooperative or organization? What are the primary benefits of this relationship? Risk Exposure Question: In broad terms, what are the main sources of risk you Probe: Nature of risk (production, market, enabling face in: environment). Probe against specific risk factors having an • Sourcing inputs? effect on farm level (for example, weather, price, environment, • Production? labor standards, logistics, operational, trade policies). • Sales and marketing of goods? Question: What are the direct negative impacts that potentially Probe: Direct impacts of enabling environment (policy) risks arise from these risks? include competition from subsidized companies. Question: What crops are more vulnerable to risks? Probe: Probe against specific production risks (weather, SPS [sanitary and phytosanitary]), market risks (prices), and enabling environment. Question: What are the three main sources of risk that most Probe: Ranking of potential problem “areas.” concern you? Question: What are the key transaction points and types of interaction associated with risk and uncertainty? Question: Of the risks identified, what is their frequency? Probe: Often or seldom, seasonal, annual. Temporal impacts? Question: How would you describe the potential severity of Probe: Expected loss—minimal, low, medium, high, very high. impact and expected losses arising from major risks? Question: Overall, have underlying conditions in the supply Probe: Get a sense of long-term threats or opportunities to chain, and your position in particular, deteriorated or improved in supply chains. recent years? Have you kept any records to track this? Check for available records and request. Note: You may wish to probe specific risk aspects related to seasonality dimensions, contracting arrangements, direct impact of weather, and environmental factors. You are trying to corroborate frequency of risk events; identify the causes of losses; could be single cause, or multiple causes (percentage). Risk transmission Along The Supply Chain Question: Do you have regular input procurement arrangements? Probe: Timely provision of inputs, cost factors, logistics issues. With whom? How effective are existing input arrangements? Formality of arrangements. Question: Do you have fixed selling (contract) arrangements with Probe: Formality of contracting, length of trading processors or other intermediaries? How often are these negotiated? relationships. Logistics. Question: How are transport requirements met? How effective are Probe: Availability, affordability (petrol prices), dependability. transport facilities? Question: What feedback mechanisms or interactions (if any) exist Probe: Shared concerns related to environment, labor, food with traders and retailers? safety. Question: What spillover effects (that is, linked impacts) do farm- Probe: Impacts of production or supply shortfalls, labor level production problems have on the wider supply chain? Which constraints, unexpected events; are there perceptions of entities are most affected? equitably or inequitably shared risks within the chain? Note: You are trying to understand how losses affect various members of the supply chain and if losses are evenly distributed by area or by farmer group. 76 Agriculture Global Practice Discussion Paper Risk Management Question: What is done to address problems in advance of a risky Probe: Ex ante risk management strategies—investments in event? How long have these actions existed? infrastructure, technology, management practices, financial instruments, organizational arrangements. Question: What is done to address problems after a risky event? Probe: Ex post risk management strategies—reallocation or sale of assets (livestock), seek employment or migration, transfers, resort to savings, borrow from friends or family, and so on. Question: How effective have these actions been? What actions Probe: Ex ante and ex post risk management strategies have been most effective? Least effective? Why? Question: Who typically provides these actions? Probe: Self-made decisions, decisions by farmer organizations, formal vs. informal mechanisms. Question: What interventions have been supported by public sector Probe: Role of public sector and government agencies vs. market- “agents” (including donors and NGOs) to manage problems? based actions vs. community level. Ex ante vs. ex post risk. Question: How effective have public interventions been? Which Probe: Timing, targeting, delivery aspects. are more or less effective? Question: What has recent experience illustrated about farmer Probe: Ability to manage risk on own vs. need for external capacity to withstand major deviations, disruptions, and disasters “partners.” in the supply chain? Question: What information sources, if any, are used to predict or Probe: Early warning information, price tracking, local assess the potential frequency, magnitude, and severity of problems? knowledge. Question: How would you describe overall access to credit and Probe: Availability and affordability of credit and timely and insurance? What are the benefits and costs from credit and “fair” payment of insurance. insurance? Note: You are trying to understand the impact of events (short term, medium term); how losses are transmitted upstream and downstream along the supply chain; how different stakeholders manage these risks; how small, medium, and large farmers manage in terms of mitigation, transfer, or coping; how shock was absorbed; and whether any stakeholders went out of business. Gaps and Opportunities to Address Risk Question: What are the main lessons learned from experience in risk management? Question: What options could be explored to better manage risks Probe: Production, market, enabling environment. affecting farmers? Opportunities and constraints. Question: What are the perceived potential options for managing problems jointly with other supply chain entities? Question: What roles could others play? Probe: Community, public sector or government, private sector, NGOs, donors. ! Remember: nity to double-check major findings, and also identify pat- You are taking time out of focus group members’ busy terns and themes at the forefront. schedules. Keep it within the time limit you promised it would take! Stage Three: Reporting a Focus Group Discussion As soon as possible after ending a focus group discussion Ending a Focus Group Discussion (ideally, that same evening with the team), proceed with Proceed with the closure according to the script, and debrief with reporting the results of the focus group discussion, using focus group discussion participants immediately after each this format (copies of the report format will be provided to focus group when information is fresh. It is a good opportu- all field teams). Agricultural Sector Risk Assessment: Methodological Guidance for Practitioners 77 Focus Group Discussion No. __________ Region Name: ____________________ City/Village Name: ____________________ Livelihood Zone: ____________________ Date: __________ Team Member Names: ____________________ Livelihood profile Main Food Crop Production (Rank Small Medium Importance of Contribution to Diet) Farmers Farmers Comments 1 2 3 4 5 *Rank from 1: Most important to 5: Least important Main Cash Crop Production (Rank Small Medium Importance of Contribution to Income) Farmers Farmers Comments 1 2 3 4 5 Small Medium Livestock (Types and Size Held) Farmers Farmers Comments Narrative description of supply chain segment interviewed: Explain main characteristics of supply chain segment (participants, interactions) and general trend. 78 Agriculture Global Practice Discussion Paper Risk Exposure Comments (Impact: Small Medium Moderate, Considerable, Food Crop (Rank Exposure to Risk) Farmers Farmers Critical, Catastrophic) 1 2 3 4 5 *Rank from 1: Most exposed/Most vulnerable to risky event to 5: Least exposed/Least vulnerable to risky event Comments (Impact: Small Medium Moderate, Considerable, Food Crop (Rank Frequency of Risk) Farmers Farmers Critical, Catastrophic) 1 2 3 *Rank from 1:High Probability to 3: Low Probability Comments (Impact: Small Medium Moderate, Considerable, Cash Crop (Rank Exposure to Risk) Farmers Farmers Critical, Catastrophic) 1 2 3 4 5 *Rank from 1: Most exposed/Most vulnerable to risky event to 5: Least exposed/Least vulnerable to risky event Comments (Impact: Small Medium Moderate, Considerable, Cash Crop (Rank Frequency of Risk) Farmers Farmers Critical, Catastrophic) 1 2 3 *Rank from 1: High Probability to 3: Low Probability Agricultural Sector Risk Assessment: Methodological Guidance for Practitioners 79 Risks for food crops Frequency Impact Risks for Cash Crops Frequency Impact Note: It is hoped that this can lead to building a risk prioritization of the sector assessment. QUANTIFICATION TABLES PRODUCTION LOSSES MARKET LOSSES Narrative description: List the causes of losses; could be a single cause or multiple causes (percentage) Risk transmission along the supply chain Schematic representation of supply chain; indication of segments and transactions vulnerable to risk and direction of risk transmission. Current risk management practices Mitigation ( for Example, Coping Comments (It Would be Good Cash Crop Risk Event Drought-Tolerant Seed ( for Example, to Have these from Small and ( for Example, Drought, Output Price) Varieties) Transfer Borrowing) Medium Farmers) Comments (It Would be Good Food Crop Risk Event to Have these from Small and (for example, drought, output price) Mitigation Transfer Coping Medium Farmers) 80 Agriculture Global Practice Discussion Paper Suggested Gaps and Opportunities to Address Risk Opportunities for Public Opportunities for Opportunities for Cash Crop Risk Event Gaps Sector and Government Private Sector NGOs and Donors Opportunities for Public Opportunities for Opportunities for Food Crop Risk Event Gaps Sector and Government Private Sector NGOs and Donors Narrative description: A sense of prioritization? Agricultural Sector Risk Assessment: Methodological Guidance for Practitioners 81 APPENDIX E EXAMPLE OF RAINFALL ANALYSIS AND CROP PRODUCTION IN GHANA Whenever data and time allow, it is useful to conduct an analysis of the relationship between rainfall and agricultural yields for main crops and for various regions. An example of such analysis is hereby shown for Ghana. BACKGROUND An analysis of rainfall data provides useful information on the level and distribution of rainfall by region and the impact of various rainfall characteristics on crop yields. Ghana has 99 weather stations located throughout the country, although some regions have a higher density than do others. Analysis was based on daily rainfall data from 1981 to 2010. Figure E.1 shows the distribution of the weather stations (orange dots). As the weather stations do not have information on the region to which they belong, distance from the centroid of each region (i) was calculated for each station (j) using the Euclidean distance formula: (x ) +( y − y ) 2 2 Dist = i xj i j Where: Dist = Euclidean distance xi = longitude from region i’s centroid xj = longitude from station j yi = latitude from region i’s centroid yj = latitude from station j Each station was assigned to the region whose distance to the centroid was the small- est. Reference marks for the centroids of each region are indicated by the blue dots in figure E.1. Agricultural Sector Risk Assessment: Methodological Guidance for Practitioners 83 FIGURE E.1. WEATHER STATION DISTRIBUTION WITH REGION CENTROIDS 11 Vea Manga Bawku Binduri Garu Bolgatanga Zuarungu East UpperKugri Babile Upper West Walewale 10 Pong Tamale Nyankpala Northern Damongo 9 Bui Salaga Kpandae Mankango Krachi Nkwanta 8 Kintampo Latitude Prang Atebubu Brong Ahafo Berekum Akaa Ejura Volta Hohoe Bechem Mampong Ashanti 7 Barekese Effiduasi Agogo Forifori Kpandu Amedzofe Goaso Owabi Kwadaso Ashanti Konongo Tafo Kwahu Kpeve Mpraeso Bibiani Bekwai -Ashanti Nkawkaw Begoro Eastern Asesewa Boso Asante Krom Tsito Pankese Akrokerri College Bunso Farms Akosombo Ahunda Adaklu Kanayerebo Tafo Sefwi Wiawso Obuasi Kibi Huhunya Akokoaso Kpong Ohawu Kade Somanya Asutsuare Adidome Afife-Weta 6 Enchi Western Abesewa Gyaman Dunkwa Agric Kusi Asamankese Akropong Akwapim Pomadze Nsawam Aburi Aveyime Keta Afienya Assin Nyankumasi Pokoase Greater Accra Twifo Praso Atieku Breman AsikumaCentral Kwanyako Agona Swedru Bogoso Tarkwa Asuansi Winneba Apam Asebu Benso Cape Coast 5 Half Assini Esiama Pretsea Komenda Princes Town 4 –3.5 –3.0 –2.5 –2.0 –1.5 –1.0 –0.5 0.0 0.5 1.0 Longitude Source: World Bank 2011. within a region was used as the basis for analysis. To RAINFALL DISTRIBUTION determine whether a year was dry or humid, a standard- Most rain occurs during the summer months from June to ized cumulative rainfall variable was calculated for each September, followed by a dry winter from November to region, according to the formula: March. In the south, it is also common to find a dry period during August, usually referred to as the “dog days of summer” because of its relationship to the Dog Star of i = StdRain (∑ oct i mar ∫ Pre ec e i − mi ) Sirius in the Canis Major constellation. A unimodal rain- i si fall pattern is thus observed in the lower rainfall northern Where: regions and a bimodal pattern in the central and southern regions. Figure E.2 shows the monthly distribution of StdRain = standardized cumulative rainfall rainfall for each region. Pre = daily rainfall m = mean yearly rainfall As shown by these charts, most rainfall occurs in the s = standard deviation of yearly rainfall March–October period, with an average of more than i = year 100 mm per month, followed by a dry season from November to February. The ensuing analysis focuses on This variable makes it easier to discern drought and observed rainfall during the period March to October. excess rainfall events. Table E.1 shows the standard- ized cumulative rainfall by year and region, with red signifying a drought event and green an excess rainfall DROUGHT AND event. EXCESS RAINFALL Cumulative rainfall for all stations was calculated for the This analysis shows that drought typically affects numer- March–October period and the average of all stations ous regions simultaneously. 84 Agriculture Global Practice Discussion Paper FIGURE E.2. MONTHLY RAINFALL PATTERNS BY REGION 300 Zuarungu Station - Upper East Region 250 Babile Station - Upper West Region 200 Nyankpala Station - Northern Region 180 250 200 160 Cumulative rainfall Cumulative rainfall 140 Cumulative rainfall 200 150 120 150 100 100 80 100 60 50 40 50 20 0 3 10 43 99 138 173 280 189 58 5 2 2 6 25 64 106 134 184 233 195 65 8 5 2 8 32 76 109 135 156 168 179 71 6 2 0 0 0 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 250 Ejura Station - Brong Ahafo Region 250 Kwadaso Station - Ashanti Region 200 Begoro Station - Eastern Region 180 200 200 160 140 Cumulative rainfall Cumulative rainfall Cumulative rainfall 150 150 120 100 100 100 80 60 50 50 40 20 7 25 99 150 174 181 132 90 214 169 38 20 28 51 114 151 169 209 127 86 177 195 96 41 23 53 122 145 158 184 124 100 172 159 49 24 0 0 0 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 250 Akaa Station - Volta Region 250 Breman Asikuma Station - Central Region 250 Sefwi Wiawso Station - Western Region 200 200 200 Cumulative rainfall Cumulative rainfall Cumulative rainfall 150 150 150 100 100 100 50 50 50 8 28 91 126 151 190 177 193 175 151 49 24 19 63 119 135 164 214 118 72 136 180 114 48 23 40 127 128 193 213 125 72 155 208 74 28 0 0 0 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Source: World Bank 2011. Drought years: 1982, 1983, 1986, 1990, 1992, 1998, and it was more than adequate. The most humid year occurred 2005. During these years, rain was more than one stand- in 2007, affecting six regions—with rainfall in the Upper ard deviation below average in at least three regions. East region more than two standard deviations above the Drought was particularly severe and widespread in 1983, average. with nine regions affected—including several with cumu- lative rainfall more than two standard deviations below THE IMPACT OF RAINFALL the average. The most recent dry year was 2005, when the Eastern and Volta regions suffered from very low rains. ON CROP YIELD These data suggest that there is a 23 percent probability (7 RAINFALL PARAMETERS out of 30 years) that drought will occur in at least one Crops are sensitive to rainfall in different ways. Low region. cumulative rainfall is the main determinant of yield, but crops can also be affected by late onset of the rainy season Excess rainfall years: 1987, 1989, 1991, 1995, 1999, 2002, or early cessation of rains. Prolonged periods without rain 2007, and 2010. Rainfall was more than one standard can also reduce yields, as can excess rainfall. The follow- deviation above average during these years, meaning that ing variables were thus calculated for each weather station Agricultural Sector Risk Assessment: Methodological Guidance for Practitioners 85 86 TABLE E.1. STANDARDIZED CUMULATIVE RAINFALL Year UpEast UpWest North Brong A Ashanti Eastern Volta Central Western G Accra Dry Regs Exc Regs Conclusion 1981 –1.56 –0.75 0.01 –0.50 0.27 0.10 0.17 1.03 0.55 0.52 1 1 Neutral 1982 –0.03 –0.07 –0.78 –1.33 –1.73 –1.77 –1.83 –0.05 –0.98 0.62 4 0 Dry 1983 –0.81 –3.07 –1.55 –1.24 –1.99 –1.90 –2.09 –2.57 –1.58 –2.39 9 0 Dry 1984 –1.53 –1.91 –0.26 –0.17 1.78 0.77 0.95 0.50 1.72 0.84 2 2 Neutral 1985 –0.98 –0.09 0.28 0.57 0.85 0.74 –0.20 –0.14 –0.04 0.50 0 0 Neutral 1986 –0.42 0.69 –0.66 –0.26 0.73 –1.07 –0.88 –1.79 –1.31 –1.18 4 0 Dry 1987 –0.46 0.11 –0.27 0.75 0.83 1.38 0.49 1.40 2.15 0.86 0 3 Excess 1988 –0.20 –0.57 0.29 –0.35 –0.90 –0.06 0.47 –0.06 –0.12 0.40 0 0 Neutral 1989 1.39 0.12 2.23 1.35 0.25 0.40 1.30 0.80 0.92 0.80 0 4 Excess 1990 –1.22 –1.00 –0.86 –1.51 –1.02 –0.30 –1.38 0.06 –0.68 5 0 Dry 1991 0.58 –0.38 2.06 0.14 –0.26 1.85 1.03 1.62 –0.17 2.45 0 5 Excess 1992 0.57 –0.23 –1.20 –1.27 –0.86 –0.85 –0.97 –0.91 –1.05 –0.96 3 0 Dry 1993 –0.31 0.28 0.04 –0.26 –0.35 1.16 –0.09 –0.35 –0.31 –0.42 0 1 Neutral 1994 1.18 0.22 0.05 –1.51 –0.46 –0.04 –0.41 0.24 –1.64 –0.88 2 1 Neutral 1995 –0.72 1.19 0.31 0.37 0.48 1.71 1.76 0.77 0.03 0.47 0 3 Excess 1996 1.07 –0.81 0.60 0.01 –0.71 –0.46 –1.00 0.21 0.66 0.92 0 1 Neutral 1997 –0.18 0.62 –0.01 0.36 –0.42 –0.58 0.09 0.28 –0.31 0.91 0 0 Neutral 1998 0.45 –0.52 –2.08 –0.01 –0.51 –0.30 0.16 –1.01 –1.28 –1.17 4 0 Dry 1999 1.78 1.45 1.43 0.63 1.50 –0.09 0.74 0.52 –0.11 0.51 0 4 Excess 2000 –0.09 1.85 0.01 –0.07 –0.56 –0.35 0.71 –1.07 0.07 –1.53 2 1 Neutral 2001 0.10 –0.57 –0.94 –0.61 –0.34 –0.46 –1.39 –0.71 0.19 0.08 1 0 Neutral 2002 –0.68 0.36 –0.13 1.03 1.22 0.44 0.49 0.62 1.76 0.65 0 3 Excess 2003 0.45 0.45 0.05 –0.88 –0.35 0.00 0.58 0.08 –0.03 –0.31 0 0 Neutral 2004 –1.07 –0.24 0.93 –0.75 –0.19 –0.66 –0.30 –0.74 –0.98 1 0 Neutral 2005 –0.42 0.46 –0.72 0.02 –0.71 –1.97 –2.04 –1.07 –0.60 –0.99 3 0 Dry 2006 –0.19 1.01 –0.62 –0.44 0.73 –0.21 0.22 0.61 0.06 –0.09 0 1 Neutral 2007 2.62 –0.39 0.80 1.84 1.47 1.25 1.44 1.13 0.83 0 6 Excess 2008 –1.05 1.49 –0.25 1.39 0.90 1.02 0.85 0.98 0.54 1 3 Neutral 2009 1.00 0.10 0.50 –0.51 0.36 –0.12 0.46 –0.55 –1.17 –0.89 1 1 Neutral 2010 0.72 0.06 1.51 3.59 0.19 0.52 –0.04 0.98 1.16 0.56 0 3 Excess Dry Years 5 2 4 4 3 5 4 6 6 4 Exc Years 6 4 5 3 5 5 5 4 5 1 Source: World Bank 2011. Agriculture Global Practice Discussion Paper per year as the basis for closer analysis of the relationship which measures the proportion of the variability in yield between rainfall and crop yield: explained by each rainfall variable, is reported (in brack- 1. Cumulative rainfall (cumrain). The sum of rain- ets) to indicate the magnitude of this impact. The short fall from March to October, it measures the total time period for analysis and limited variability of some amount of rain that accumulates yearly from of the data limited the explanatory power of these regres- March to October. It is expressed in millimeters. sions, although some general trends are apparent. 2. Onset date (onset). The time of year in which the rainy season starts, defined as the first day of the The impact of individual rainfall parameters is most year with 20 mm or more of rain. It is measured apparent for the production of rice and groundnuts in as the number of days from the start of the year. Upper East region. Excess rainfall is the major risk, rather 3. Cessation date (cessation). The day the rainy sea- than drought, as shown by the negative signs for the son ends is defined as the day on which 90 percent impact of cumulative rainfall, number of rainfall days, of total rainfall period occurs. It is measured as and the maximum rainfall in any 10-day period. Ground- the number of days from the start of the year. nut yields respond positively to the earlier onset of rain in 4. Length (length). The length of the rainy season both Upper East and Upper West regions. is defined as the difference between the cessation date and the onset date, measured as number of The impact of rainfall in other regions is limited. Maize days. and yam yields are vulnerable to drought in BrongAhafo, 5. Rain days (events). The number of days in the yam yields are vulnerable to drought in Central, and period when rainfall was higher than 1 mm. maize yields are vulnerable to excess rainfall in Ashanti. 6. Dry spell (drysp). The longest number of consecu- This suggests that factors other than rainfall may be more tive days without rain. important determinants of yield in these regions. 7. Extreme excess rainfall (max10days). The yearly maximum amount of cumulative rainfall in any PRINCIPAL COMPONENT 10 consecutive days. ANALYSIS The influence of rainfall on yield was examined using As the variables used for analysis are all different attri- regional production data for maize, rice, millet, ground- butes of the same weather phenomenon (rainfall), some nuts, cassava, and yams for 1992–2009. will be correlated. Figure E.3 shows the correlation matrix for the first six variables.6 REGRESSION ANALYSIS Some variables are closely related, such as the length of The rainfall parameters described above were averaged the rainy season (durac) and the cessation date (ces90): the across the weather stations in each region and regressed higher the cessation date, the longer the rainfall period. on yield, as described below. These correlations can also be highly negative—such as the correlation between the length of the rainy season Yield = b0 + b1 cumrain (durac) and the onset date, because the later the rainy sea- Yield = b0 + b2 onset son starts, the shorter its duration. Yield = b0 + b3 events Yield = b0 + b4 cessation Because of this high correlation, principal component Yield = b0 + b5 length analysis was used to further analyze the impact of these Yield = b0 + b6 drysp parameters on crop yield. Table E.3 shows the first three Yield = b0 + b7 max10days eigenvalues and the corresponding proportion of vari- ance explained. Results are reported in Table E.2 for those crops and regions where the regression coefficient was statistically 6 The excess rainfall variable was not used in the principal component analysis significant (at 5 percent). The coefficient of determination, because it was introduced at a later stage. Agricultural Sector Risk Assessment: Methodological Guidance for Practitioners 87 TABLE E.2. IMPACT OF RAINFALL PARAMETERS ON CROP YIELD Upper Upper Parameter East West Northern BrongAhafo Ashanti Eastern Volta Central Western Cum rain Rice (−) Maize (+) (32%) (24%) GNuts (−) Yams (+) (35%) (23%) Onset of rain GNuts (+) GNuts (+) (27%) (39%) Number of Rice (−) rainfall days (26%) Cessation date Length of rainy season Dry spell Yams (−) (46%) Max days Rice (−) Maize (−) (42%) (26%) GNuts (−) (28%) The first component explains 52 percent of the total vari- Figure E.4 shows that the length of the rainy season ance accounted for by the six variables, and the second (durac), cessation date (ces90), and onset date (onset) are component explains an additional 25 percent of the highly correlated amongst themselves, and so constitute variance—so that the cumulative variance explained is the first factor together with the negatively correlated more than 77 percent. This reduces the dimensionality of onset date. This factor can be taken to represent the the original problem from six variables into two compo- length of the rainy season given that the length is high, nents with a reduction in variability of only 23 percent the onset date is low, and the cessation date is high as (100 percent to 77 percent). A third component would well. Hence when this factor is large, the rainy season is add another 10 percent of the variability explained. The very long. usual practice is to retain as many components as eigen- values are higher than one, which suggests retaining the Factor two consists of cumulative rainfall (cumrain), num- first two components. Table E.4 shows the correlation ber of rainy days (events), and length of the dry spell (factor loadings) of each component with each of the var- (drysp), which are positively correlated, together with dry iables in the two-factor solution (retaining the first two spell, which is negatively correlated. This means that components, meaningful loadings marked in red): when the dry spell is very long, cumulative rainfall and 88 Agriculture Global Practice Discussion Paper FIGURE E.3. CORRELATION MATRIX PLOT Source: Authors’ analysis. TABLE E.3. PRINCIPAL COMPONENTS TABLE E.4. CORRELATION OF COMPONENTS ANALYSIS: THREE EIGENVALUES Variable Factor 1 Factor 2 AND PROPORTION OF cumrain 0.114 0.893 VARIANCE EXPLAINED onset −0.746 −0.263 % Total events 0.228 0.895 Eigenvalue Variance Cumulative % ces90 0.861 0.053 durac 0.984 0.174 1 3.1297 52.16 52.16 drysp −0.113 −0.777 2 1.5165 25.27 77.44 3 0.6384 10.64 88.08 number of events will be low. This second factor repre- six variables into the two new variables (factors). These sents the intensity of rainfall during the year. scores are standardized so that the mean is equal to zero and the standard deviation is equal to one. Figure E.5 These two factors (or principal components) by definition shows the mean scores for each factor by region. are orthogonal, meaning that they are independent between themselves. They suggest that rainfall in Ghana The length of the rainy season is shorter than normal in has two main attributes: the length of the rainy season the Upper East and Upper West regions, as their mean (factor 1) and the intensity of rainfall (factor 2). factor scores are smaller than −1. By contrast, the Ashanti region has the highest mean score for factor one, meaning Based on this two-factor solution, it is possible to derive that the season is usually longer. For the intensity factor, factor scores, which are the transformation of the original the Volta and Western regions seem to have the most Agricultural Sector Risk Assessment: Methodological Guidance for Practitioners 89 FIGURE E.4. FACTOR LOADINGS PLOT Factor loadings, factor 1 vs. factor 2 Rotation: Varimax raw Extraction: Principal components 1.0 events cumrain 0.8 0.6 0.4 durac 0.2 ces90 Factor 2 0.0 –0.2 onset –0.4 –0.6 drysp –0.8 –1.0 –1.0 –0.8 –0.6 –0.4 –0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 Factor 1 FIGURE E.5. MEAN FACTOR SCORES BY » Cereal and groundnut yields are affected by adverse REGION patterns and events in both the length of the rainy 1 Regional Mean Factor Scores season and the intensity of rainfall. Moreover, yields can be adversely affected when the rainy sea- 0.5 0.6 0.8 0.8 son is either too long or too short. Both high and low intensity rainfall can reduce yields, but excess 0.4 0.3 0.4 0.4 0.2 0.3 0.3 0.1 0.2 0.1 0 –0.3 –0.3 0.0 rainfall appears to pose the highest risk. Together, Factor score –0.4 –0.5 –1.2 these results confirm the vulnerability of cereal –1.2 –1.8 Factor 1, Length Factor 2, Intensity and groundnut yields to a range of adverse rainfall –1 patterns and events in the lower rainfall zones in –1.5 which they predominate. » Cassava and yam yields are vulnerable to shorter –2 Upper Upper Northern Brong Ashanti Eastern Volta Central Western Greater rainfall seasons and lower intensity rainfall, East West Ahafo Accra although neither set of factors had a substantive impact. This may result from the higher and more reliable rainfall patterns in the transition and forest intense rainfall because their mean score is almost one zones where these crops predominate. standard deviation above the mean (0.8), whereas the Greater Accra region has the lowest intensity of rainfall. The impact of both sets of factors was higher in the drier BrongAhafo, Ashanti, Eastern, Volta, and Western regions savannah zones, particularly Upper East and Upper West, have similar rainfall conditions. as would be expected. 90 Agriculture Global Practice Discussion Paper APPENDIX F GUIDELINES FOR CONSULTANTS PRIOR TO MISSION In the ARMT’s experience, team members on missions who are unprepared spend time in stakeholder meetings trying to understand the structure of the sector and sup- ply chains instead of focusing on the risks to the sector. This is not the purpose of the mission! Consultants should therefore: » Read up on ARMT’s previous work and the methodology applied in risk assess- ments. » Read the preparatory work already conducted by ARMT on the country (in Dropbox). » Prepare background notes on relevant crops of livestock (see table F.1). During mission: Make sure you take notes! ARMT’s experience is that it is useful to start writing during the mission while thoughts are still fresh and to get a better sense of where the information gaps are before leaving the country. ARMT understands that this is not always feasible given the workload during missions, but you are strongly encouraged to think through what you have and what information you need to support the risk story that emerges for your commodities. It is so much more difficult to get the information once out of the country, and in ARMT’s experience, this significantly delays production of these reports. Additionally, you will need this information for your stakeholder presentation. Post mission (approximately two staff weeks depending on the number of com- modities): Please ensure timely submission of your inputs! The lead author depends on your work to do his or her job, and the ARMT will spend additional time on the report once it is received. After that, it is reviewed in the World Bank before it is sent to the client. Hence, the process is long and it is important to avoid delays. Guidelines on writing follow. Writing for ARMT: Because every country is different, the risk story will vary between countries and commodities, so no fixed format exists for ARMT’s reports. Nevertheless, certain standards are needed to ensure quality work. Additionally, in ARMT’s experience, substandard deliverables from team members (including poor Agricultural Sector Risk Assessment: Methodological Guidance for Practitioners 91 TABLE F.1. TEMPLATE FOR COMMODITY SUPPLY CHAIN PROFILE Production (including geography) Processing Markets, domestic (including price and so on) Markets, export Logistics Main stakeholders (private and public) writing) significantly delay completion of ARMT reports » ARMT’s audience is rather time constrained— and are both costly and time consuming to correct. Thus make sure the text is easy to follow. ARMT requests that any deliverables comply with the fol- » Allocate one or two paragraphs for each topic, lowing: unless it is absolutely necessary to expand » When you write your text, critically ask whether it beyond that. Make sure that the message of each would be clear to an audience that is not familiar paragraph is clear and that messages are not with the topic. repeated. 92 Agriculture Global Practice Discussion Paper » Use numbers to back up your arguments. ARMT’s » If policies are considered a risk, back up the audiences are aware of the agricultural risks in argument with a timetable with past regulatory their countries; ARMT’s value added is that it changes and time series figures that demonstrate quantifies those risks. the impacts of these regulatory changes in specific » When referring to fluctuations in prices, exchange months or years. rates, and so on, demonstrate them in figures and graphs. Agricultural Sector Risk Assessment: Methodological Guidance for Practitioners 93 APPENDIX G SAMPLE TERMS OF REFERENCE FOR CONSULTANTS BACKGROUND Risks are inherent, ubiquitous, and varied in agricultural systems, perhaps more so than in any other area of economic endeavor. They enforce poverty traps and pose serious consequences for all stakeholders. The apparent increase in the frequency of commodity price spikes and crop failures, added to concerns over climate change and food security, have increased global interest in ARM. The prevalence and complexity of the multiple risks facing agricultural systems and the failure to address them on an ex ante and integrated basis continue to leave countries and their agricultural supply chains at risk. The realization of these risks leads to a perpetual cycle of “shock- recovery-shock,” which endangers the sustainability of ongoing initiatives and remains a major impediment to the development of most agricultural sectors. With the objective of mainstreaming agricultural risk management into national sec- tor development strategic planning and promoting more resilient and sustainable agri- cultural systems, the World Bank’s Agriculture Risk Management Team provides TA that helps clients evaluate agricultural risks and put in place requisite systems for improved risk management. ARMT’s ARM framework looks comprehensively at the risks to the agricultural sector and comprises (i) risk assessment and prioritization; (ii) stakeholders’ assessments; (iii) analysis of risk management strategies; and (iv) screen- ing of potential risk management instruments. This activity has the potential to inform agricultural development strategies and invest- ment plans as well as the agricultural development programs of development partners or donors. By quantifying risk and providing an evidence-based prioritization of risk management interventions, it can facilitate objective policy and programming dia- logue and help focus design and delivery of TA activities. In the end, however, integra- tion of risk management practices into development strategies will reduce the short- and medium-term volatility in the agricultural sector and improve resilience in the longer term, thus reducing vulnerabilities among agricultural sector stakeholders Agricultural Sector Risk Assessment: Methodological Guidance for Practitioners 95 and increasing the potential success of agricultural invest- 1. More than 25 years of experience designing, ment strategies. implementing, and analyzing agricultural devel- opment programs in (specify region) CONSULTANT 1: 2. Broad-based experience in one or more relevant fields, including agricultural and rural investment, CROP SPECIALIST policy planning, sectoral analysis, food security, The consultant will act as the ASRA’s key focal point in agricultural marketing, or institutional develop- the field and will be expected to take the lead in planning, ment organizing, and facilitating the mission. In particular, par- 3. Experience with value chain methodology and ticipation in this assessment will consist of the following approaches a plus responsibilities: 4. Experience with rice, groundnuts, sorghum, or » Liaising directly with key governmental and non- other major crops relevant to the study governmental partners at all levels to increase 5. Relevant analytical skills, operational experience, awareness and support for the activity training, and credentials to perform the technical » Providing the team with mission-critical local per- tasks required for the risk assessment mission spective and sectoral context related to risk and risk 6. Fluency in spoken and written (specify language) management » Spearheading and coordinating information and data collection efforts before, during, and after the CONSULTANT 2: mission LIVESTOCK SPECIALIST » Taking the lead in identifying key stakeholders The consultant will provide technical expertise on the for informational interviews and focus groups, as livestock sector, related supply chains (meat, dairy, poul- appropriate, and developing mission itineraries try), and production systems. The consultant’s participa- and meeting schedules tion in this assessment will consist of the following: » Performing data analysis of crop production and estimated losses, and other types of analytical work Phase I: as determined by the task team leader (TTL) » Conducting research and compiling, organizing, » Managing activities of other local consultants to and analyzing primary and secondary data rel- ensure that all efforts are well coordinated and evant to risks in the subsector; appropriately focused, per the TTL’s direction » Identifying relevant stakeholders to participate in » Coordinating field visits and interviews and inde- consultations and assisting with the development pendently leading some portion of the consulta- of an appropriate mission agenda; and tions with stakeholders » Participating in team discussions in preparation for » Participating in team discussions at the end of fieldwork. the field visits to synthesize stakeholder input and Phase II: assisting with organizing a workshop (and identify- » Coordinating with other consultants and ASRA ing participants) to share the preliminary findings team members; with key counterparts » Facilitating interactions, including focus group dis- » Contributing independently to key sections of the cussions, with sector stakeholders to collect data; final report (in English) and performing any other » Analyzing production, market, and enabling envi- activities as instructed by the TTL ronment risks to the livestock sector, including meat, dairy, and poultry supply chains and risk SELECTION CRITERIA impacts on livelihoods; For providing assistance in designing and conducting » Assessing levels of vulnerability and risk manage- national agricultural sector risk review, the World Bank is ment capacity across pastoralist and agro-pastoral- recruiting a consultant with the following characteristics: ist livelihood systems; 96 Agriculture Global Practice Discussion Paper » Performing quantitative or qualitative analysis of frequency and severity of losses to the livestock CONSULTANT 3: subsector and related supply chains; VALUE CHAIN SPECIALIST » Participating in the quantification and prioritiza- The consultant has experience with value chain develop- tion of risks, using AMRT’s methodology; ment, postharvest, food safety, marketing, training and » Analyzing the existing risk management (mitiga- extension, and project management. The consultant’s tion, transfer, coping) landscape across the livestock participation in this assessment will consist of subsector at micro- (rural), meso- (market), and » Independently leading parts of the ASRA; macro- (institutional; policy) levels; » Assisting in identifying relevant supply chain actors » Recommending measures and interventions for to participate in stakeholder consultations; improved risk management using ARMT’s meth- » Organizing, planning, and managing independent odology; and meetings with a range of agricultural sector stake- » Participating in a stakeholders’ roundtable to vali- holders; date and share initial findings. » Collecting, organizing, analyzing, and sharing quantitative and qualitative data on agricultural Phase III: risks and risk management activities with the » Conducting follow up as necessary to address assessment team; knowledge and data gaps; » Identifying and reviewing existing resources on » Contributing key sections to the final report; and agricultural risks in the country; » Performing any other relevant technical task » Coordinating with other teams and consultants; required by the TTL. » Performing quantitative analysis of the frequency and severity of losses from agricultural risk; SELECTION CRITERIA » Analyzing existing and potential risk management For providing assistance in designing and conducting the measures (mitigation, transfer, coping); ASRA, the consultant should have » Prioritizing risks and risk management measures 1. A minimum of 10 years of experience working in using ARMT’s methodology; and studying animal health and husbandry, live- » Participating in a stakeholders’ roundtable to vali- stock management, feed and fodder management, date and share initial findings; processing technologies, livestock extension sys- » Authoring specific sections of the assessment tems, and related activities, preferably in (specify report and providing input to refine, finalize, and region); disseminate it; and 2. Minimum of a master’s degree in animal hus- » Performing any other relevant technical tasks bandry, livestock management, animal health, required by the TTL. veterinary services, dairy food technology, or a related field; SELECTION CRITERIA 3. Proven experience in conducting rapid assess- For providing assistance in designing and conducting the ments and in working effectively in multidisci- ASRA, the World Bank is recruiting a Consultant per the plinary teams; following: 4. Excellent communication, analytical, and report 1. More than 20 years of experience designing, writing skills, with fluency in spoken and written implementing, and analyzing agricultural devel- English and (specify language); opment programs in (specify region) 5. Experience with value chain methodology, 2. Broad-based experience in one or more relevant approaches, and concepts a plus; and fields, including agricultural and rural invest- 6. Relevant analytical skills, operational experience, ment, policy planning, sectoral analysis, food training, and credentials to perform the technical security, agricultural marketing, or institutional tasks required for the ASRA mission. development Agricultural Sector Risk Assessment: Methodological Guidance for Practitioners 97 3. Experience in one or more relevant subsectors, 5. Relevant analytical skills, operational experi- including staples, cash crops, horticulture, or live- ence, training, and credentials to perform the stock technical tasks required for the risk assessment 4. Experience with value chain methodology and mission approaches a plus 6. Written and oral fluency in local language a plus 98 Agriculture Global Practice Discussion Paper APPENDIX H EXAMPLE OF A COMMODITY RISK PROFILE IN GHANA Agricultural Sector Risk Assessment: Methodological Guidance for Practitioners 99 100 Agriculture Global Practice Discussion Paper APPENDIX I EXAMPLE OF A REGIONAL RISK PROFILE IN UPPER WEST REGION OF GHANA Agricultural Sector Risk Assessment: Methodological Guidance for Practitioners 101 102 Agriculture Global Practice Discussion Paper APPENDIX J TABLE OF CONTENTS FROM AN ASRA IN KENYA Acronyms and Abbreviations Acknowledgments Executive Summary Chapter One: Introduction Chapter Two: Agriculture Sector in Kenya Agroclimatic Conditions Rainfall Patterns and Trends Crop Production Systems Livestock Production Systems Variability in Crop Production Food Supply and Demand Agricultural Markets and Price Trends Livestock Production Food Security Constraints to Agricultural Growth Chapter Three: Agriculture Sector Risks Production Risks Pests and Diseases Market Risks Enabling Environment Risks Multiplicity of Risks Chapter Four: Adverse Impacts of Agricultural Risks Conceptual and Methodological Basis for Analysis Crop Production Risks Livestock Risks Chapter Five: Stakeholder Vulnerability Assessment General Trends in Vulnerability Livelihoods and Agroclimatic Conditions Poverty and Vulnerability Vulnerability Among Livelihood Groups Agricultural Sector Risk Assessment: Methodological Guidance for Practitioners 103 Vulnerability and Risk Management Case Study 7: Yusuf Khalif Abdi—Livestock Risk Management Capacity Herder, Garissa County Vulnerability in ASALs Case Study 8: Fresha Dairy—Milk Processors, Chapter Six: Risk Prioritization and Management Githunguri County Risk Prioritization Appendix C: Stakeholder Vulnerability Analysis Risk Management Measures General Trends in Vulnerability Illustrative Risk Management Measures Vulnerability, Livelihoods, and Agroclimatic Prioritization of Risk Management Measures Conditions Conclusion Poverty Status and Vulnerability References Vulnerable Groups Appendix A: Climate Change Impacts on Agriculture in Pastoralists Kenya Female-Headed Households (FHHs) Introduction Unskilled/Casual Wage Laborers Principal Findings Appendix D: Rainfall Analysis Climate Change and Severe Weather Events Appendix E: Weather and Yield Impact Analysis Methodologies Background Crop Predictions Summary and Key Findings Crops Resistant to Climate Change Weather Information Conclusions Annual Rainfall Distribution in Kenya Appendix B: Stakeholder Risk Profiles Drought and Excess Rainfall Analysis Case Study 1: Philip Mutua Mbai—Smallholder Rainfall—Yield Regressions Maize Farmer, Machakos County Appendix F: Crop Production Trends Case Study 2: Mrs. Maraba—Agro-input Dealer, Appendix G: Livestock Terms of Trade Analysis Eldoret Uasin Gishu County Appendix H: Options for Scaling Up Livestock Insurance Case Study 3: Leshamon Olekoonyo—Wheat in Kenya Farmer, Narok Appendix I: Results of Solutions Filtering Process Case Study 4: Marcel Wambua—Head of Food Crops Finance, Lesiolo Grain Handlers Limited Cash Crops Case Study 5: Michael Waigwa—Agricultural Livestock Underwriter, Cooperative Insurance Company Case Study 6: Wilson Murunya—Livestock Herder, Kajiado County 104 Agriculture Global Practice Discussion Paper APPENDIX K PRIORITIZATION OF RISK MANAGEMENT MEASURES IN NIGER—A FILTERING APPROACH Many risk management measures are complementary in nature and will contribute to improved risk management in the short, medium, and long term. In a resource-con- strained environment, however, decision makers are compelled to find the quickest, cheapest, and most effective measures among a myriad of possibilities. Ideally, a detailed, objective, and exhaustive cost-benefit analysis will help in selecting the most appropriate intervention options. But conducting a cost-benefit analysis of a multitude of options can be costly and time consuming. As such, using decision filters to evaluate and prioritize among a list of potential interventions can help in making rational resource allocation decisions in lieu of a detailed cost-benefit analysis. The Agriculture Risk Management Team conducted a filtering exercise in Niger to prioritize risk man- agement interventions using some simple criteria as filters (see table K.2). » Relative benefits: The most beneficial interventions were those that reduced the hazards, exposures, and losses associated with the various identified risks. Further, if the intervention could lead to additional benefits (for example, yield or efficiency improvements or cost reduction), it was rated highly. Table K.1, taken from the 2013 “Niger Agricultural Sector Risk Assessment” (World Bank 2013c), shows the “relative benefits” filter in actions. » Relative cost: Without detailed assessments, it may be difficult to estimate the cost of some interventions. ARMT’s experience is that the “relative cost” of interventions can be assessed more readily. For example, the cost involved in a large-scale irrigation project is usually much higher than the cost involved in setting up a system for seed distribution. » Scalability: Owing to a required prerequisite some interventions may ben- efit a small group of stakeholders, but other interventions may have much greater potential to be scaled-up and benefit a much larger group of stake- holders. Agricultural Sector Risk Assessment: Methodological Guidance for Practitioners 105 TABLE K.1. RELATIVE BENEFITS OF RISK MANAGEMENT MEASURES Yield or Addresses Type of Reduces the Reduces the Reduces the Compensates Productivity Multiple Intervention Hazard Exposure Losses after the Loss Improvements Risks Drought- No Yes Yes No Yes (in a drought, No tolerant seed not otherwise) varieties (M) Soil and water No Yes Yes No Yes Yes conservation/ NRM (M) Irrigation (M) No Yes Yes No Yes No Early detection No Yes Yes No Yes No and destruction of locust (M) Community- No Yes Yes No No No level food and fodder banks (M and C) Vaccination No Yes Yes No Yes No programs (M) Insurance (T) No No No Yes No Yes Shortening No Yes Yes No No Yes emergency response time (C) Strategic No Yes Yes No Yes No de-stocking (C) Contingent No Yes No Yes No Yes financing (C) Source: World Bank 2013c. Note: (M) = Mitigation; (T) = Transfer; (C) = Coping. NRM = Natural resource management. » Ease of implementation: The technical com- could have catastrophic long-term consequences plexity of an intervention and local stakeholders’ for the environment. Hence, it is important to capacity to implement it is another filter that can scrutinize the potential adverse impacts of a given be used in prioritization. Simpler interventions intervention on the environment. might be more readily accepted by stakeholders » Potential impact on poverty alleviation: and therefore easier to implement. Although some interventions directly contribute to » Return time: Some interventions have a long improved income and poverty alleviation, others gestation period, whereas others yield quick results. do so indirectly. Using this filter helps in identifying Risk management requires short-, medium-, and risk management interventions that might yield the long-term perspectives, but quick wins are often a largest poverty alleviation dividends. high priority for decision makers. » Adverse impact on the environment: Some Based on prioritization of risk and intervention measures, risk management interventions (for example, large- six interventions offered the greatest risk management scale spraying of chemicals for locust destruction) benefits in Niger: (i) drought-tolerant crop varieties; 106 Agriculture Global Practice Discussion Paper TABLE K.2. DECISION FILTERS AND INTERVENTION CLASSIFICATION Potential Adverse Impact on Type of Relative Ease of Return Impact on Poverty Intervention Scalability Cost Implementation Time Environment Alleviation Drought-tolerant High Medium Medium Short Low High or improved seed varieties (M) Soil and water High Medium Medium Medium Low High conservation and NRM (M) Irrigation (M) Low High Low Short-Med Moderate High Early detection High Medium High Short Moderate Low and destruction of locust (M) Community-level High Medium Medium Short Low High food and fodder banks (M and C) Vaccination High Medium Medium Medium Low High programs (M) Contingent High Low High Short Low Low financing (C) Shortening Medium Low Medium Short Low Low emergency response time (C) Strategic Low Medium Low Medium Low Low destocking (C) Insurance (T) Low Low Medium Medium Low Low Source: World Bank 2013c. Note: (M) = Mitigation; (T) = Transfer; (C) = Coping. NRM = Natural resource management. (ii) soil and water conversation and NRM interventions; (vi) livestock vaccination programs. Stakeholders could (iii) expansion of small-scale irrigation; (iv) support to choose other criteria as filters, but it is important to ensure community-level food and fodder banks; (v) continuous clarity, consistency, and objectivity while using them to support to early detection and destruction of locusts; and evaluate decision options. Agricultural Sector Risk Assessment: Methodological Guidance for Practitioners 107 APPENDIX L EXAMPLES OF AGRICULTURAL RISK MANAGEMENT SOLUTIONS Mitigation Transfer Coping Production Drought (crop) Sustainable land management Macro- (government) level Use of weather index for triggering (SLM) crop insurance early warning and response Conservation agriculture Farm-level crop insurance Food-for-work programs Natural resource management Weather-index insurance Cash-for-work programs Strengthen early warning systems Community risk pooling Promoting household and community savings Water harvesting Establishing and managing strategic grain reserves Irrigation expansion and Contingent financing and other development financial instruments for financing coping strategies Investment in small-scale irrigation Decentralized disaster contingent (dry season farming) to improve fund for rapid response to local nutrition and food diversity emergencies Development & dissemination of Cash payments drought-tolerant seed varietals Safety net interventions Migration for work Crop diversification Emergency food distribution to affected communities Improved water and soil Informal credit management Improved farming techniques Formal and informal borrowing (for example, intercropping, conservation tillage) Savings for consumption smoothing Sale of livestock and other assets Agricultural Sector Risk Assessment: Methodological Guidance for Practitioners 109 Mitigation Transfer Coping Drought (livestock) Interventions for improving Sovereign agricultural risk Facilitate early destocking of livestock feed and fodder financing livestock availability (pasture improvement and forage production) Community-level fodder and forage Public-private partnership in Livestock food and forage delivery banks livestock insurance Modifying transhumance patterns Emergency food aid More strategic destocking Cash- and food-for-work programs Health interventions (improving Fodder banks access to deworming medicines and salt licks) Invest in livestock sector Contingent financing and other infrastructure financial instruments for financing coping strategies Hay making and storage, irrigated Cash payments fodder production Livestock microfinance Migration for work Institutional reform Use of weather index for triggering early warning and response Reserve grazing pastures and Building water pans standing pasture Controlled livestock movement Exceptional livestock movements Conditional parks grazing, wildlife, and livestock coexistence Increase water conservation pans Intensification and strengthening of disease surveillance Community peacekeeping programs Strengthening community customary governance Capacity building (for farmers and local officers) Pest and disease Development and dissemination outbreaks (crop) of pest-and disease-resistant seed varietals Improved and sustained support for early detection and destruction of locusts Fumigation (storage) Promoting hermetically sealed sacks or silos (on farm) Pest and disease Increased vaccination campaigns Quarantine measures outbreaks (livestock) Improving veterinary services and Livestock vaccination access to services 110 Agriculture Global Practice Discussion Paper Mitigation Transfer Coping Excessive rainfall Improved land management Flood index insurance Public disaster relief payments and floods practices (for example, retain and store potential flood water) Micro- and small-scale irrigation Flood detection and early warning development Irrigation rehabilitation Social protection programs Terracing Windstorms Introduce vegetative barriers and crosswind ridges Development of stronge and resistant varietals and grafting techniques Climate-smart farming Bushfire Introduce and strengthen legislation governing bush burning Community education and environmental awareness Improved land management to prevent the invasion of grasslands by trees and shrubs Market Consumer price Community-level price stabilization Improving efficiency of emergency volatility grain reserve Improved use of existing market Shortening emergency response information for earlier response to time food price spikes Upgrade market information systems Lower barriers to trade Increase production Commodity price Commodity exchange Hedging volatility Use of forward contracts Forward contracts Strengthening management Warehouse receipts information systems (MIS) Interest rate volatility Exchange rate Hedging volatility Forward contracts Counterparty risk Strengthening regulatory and legal and default framework Use of contracts Logistical risks Improving the efficiency of private storage (lower losses, and so on) Marketing (road, storage) infrastructure development Agricultural Sector Risk Assessment: Methodological Guidance for Practitioners 111 Mitigation Transfer Coping Input price Lower barriers to trade volatility More transparent government support policy Strengthening import financing Use of forward contracts Infrastructure development Enabling Environment Regulatory risk Promote food safety and quality (domestic and standards int’l) Diversify market destinations and trading partners Erratic gov’t Development of long-term, intervention and transparent policy policy Institutional risk Strengthen land tenure systems and failure Political instability More inclusive, broad-based Social protection programs policy-making process, reinforcing traditional mechanisms and creating joint customary and formal mechanisms 112 Agriculture Global Practice Discussion Paper A gr i c ulture G lobal P ra c t i c e D i s c u s s i on P aper 1 0 W O R L D B A N K G R O U P R E P O R T N U M B E R 100320-GLB 1818 H Street, NW Washington, D.C. 20433 USA Telephone: 202-473-1000 Internet: www.worldbank.org/agriculture Twitter: @WBG_agriculture