Managing Food Price Risks and Instability in an Environment of Market Liberalization Agriculture and Rural Development Department 34499 REPORT NO. 32727-GLB Managing Food Price Risks and Instability in an Environment of Market Liberalization Agriculture and Rural Development Department THE WORLD BANK AGRICULTURE AND RURAL DEVELOPMENT DEPARTMENT © 2005 The International Bank for Reconstruction and Development / The World Bank 1818 H Street, NW Washington, DC 20433 Telephone 202-473-1000 Internet www.worldbank.org/rural E-mail ard@worldbank.org All rights reserved. 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 accu- racy 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 accept- ance of such boundaries. The material in this publication is copyrighted. Copying and/or transmitting portions or all of this work without permission may be a violation of applicable law. The International Bank for Reconstruction and Development/ The World Bank encourages dissemination of its work and will normally grant 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, Dan- vers, 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. Contents ACKNOWLEDGMENTS vii A C R O N Y M S A N D A B B R E V I AT I O N S ix E X E C U T I V E S U M M A RY xi 1 Introduction 1 Objectives of the Study 2 Outline of the Report 3 2 Toward a Typology for Food Policy Analysis 5 A Macrotypology of Countries 5 Diversity in Domestic Food Consumption 6 Dependence on Trade and Access to Global Markets 7 Capacity to Meet Food Import Requirements on Commercial Terms 8 Variability in Domestic Food Production 9 Combining Criteria into a Country Typology 10 A Microtypology of Households 13 Rural Household Participation in Markets 13 A Reduced Role for Food Staples 15 Main Messages for the Design of Food Security Policies 15 3 The Nature and Extent of Food Price Instability and Risk 17 Variability in World Grain Prices 17 Price Variability in Domestic Food Markets 18 Sources of Domestic Price Instability 19 Will Prices Become More Unstable? 20 Main Messages on the Nature and Extent of Food Price Variability 21 4 The Costs of Food Price Instability and Risk 25 Economic Efficiency Costs 25 Effects on Income Distribution and Household Food Security 26 Macroeconomic Externalities 27 Main Messages on the Costs of Food Price Instability 28 iii iv Contents 5 Food Market Reforms: Experiences and Lessons 29 State-Led Food Marketing Systems: Sometimes Effective, Always Costly 29 Market Liberalization: Uneven Implementation, Uneven Results 30 Growing Controversy over Food Market Liberalization 30 Three Cases of Food Market Liberalization 31 Southern Africa: Should Governments Continue to Intervene in Maize Markets? 31 Liberalization of Rice Markets in Bangladesh: An Antidote to the Floods of 1998 32 Mexico’s Maize Market Reforms: Good Intentions but Mixed Results 33 Lessons from Reform Experiences in Asia, Africa, and Mexico 33 6 Policy Options for Making the Transition to Private Markets 35 Guiding Principles for Public Policy Design 35 Creating Space for Private Markets to Operate 37 Predictable Implementation of a Well-Defined Food Security Strategy 37 Subsidy Reform to Level the Playing Field for the Private Sector 37 Remove Remaining Restrictions on Grain Movements and Imports 38 Introducing Flexibility into Pricing Policies 39 Tendering to the Private Sector 39 Policies for Direct Support to Market Development in the Medium Term 40 Improving Crop Forecasting and Market Information Systems 40 Liberalizing Regional Trade 40 Building Private-Sector Capacity 41 7 Specific Policy Options for Managing Price Instability and Risk 45 Market-Based Risk Management Instruments 45 Financial Markets 46 Warehouse Receipts 46 Futures and Options Contracts 47 Index-Based Weather Insurance 51 Commodity-Linked Finance 52 Assessing the Potential of Market-Based Risk Management Instruments 53 The Advantages of a Market-Based Approach 53 Challenges to Implementing a Market-Based Approach 53 Main Messages on Market-Based Approaches 54 Counter-Cyclical Safety Nets to Protect the Poor and Vulnerable 54 Variable Tariffs to Manage World Price Shocks 56 An Arm’s-Length Food Reserve to Manage Domestic Shocks? 57 Contents v 8 Managing the Policy Dialogue 59 9 Conclusions and Recommendations 63 How Big Is the Problem, and Who Suffers Most? 64 What Do We Learn from Policy Reforms? 65 Moving Forward: Broader Policy Options 65 Specific Policy Options for Managing Price Instability and Risk 66 The Importance of Local Circumstances 68 Entry Points for the World Bank 68 Appendix 1. List of Workshop Papers 71 Appendix 2. Detailed Data for the Macrotypology of Countries 73 NOTES 75 REFERENCES 77 Figures 2.1 Diversity of Staple Consumption, 2002 7 2.2 Trade Status of Dominant Staple Commodities, 1994–2003 8 2.3 Import Share of Staple Food Utilization, 1994–2003 8 2.4 Value of Average Annual Cereal Imports as a Share of Foreign Exchange Reserves, 1994–2003 9 2.5 Cuddy–Della Valle Index of Production Variability for Staples, 1995–2004 10 3.1 Coefficient of Variation of Wheat Producer Prices, 1971–2002 19 3.2 Coefficient of Variation of Maize Producer Prices, 1971–2002 19 3.3 Wholesale Maize Prices (US$/ton) in Addis Ababa Relative to Import and Export Parity Prices, Based on U.S. White Maize Traded via Djibouti, 1998–2003 20 3.4 Wholesale Maize Prices (US$/kg) in Southern Africa, 1994–2003 20 3.5 Trends in World Grain Stocks, 1989–2004 (million metric tons) 22 Tables 2.1 Countries, Income Status, and Dominant Staples Based on Calorie Intake 6 2.2 Typology of Countries According to Exposure to Global Price Shocks 11 2.3 Typology of Countries According to Exposure to Domestic Price Shocks 12 2.4 Distribution of Farm Households According to Participation in the Staple Grain Market, Africa (% of total) 14 2.5 Distribution of Rural Households in Indonesia, Vietnam, and Mexico According to Participation in the Staple Grain Market (% of total) 14 2.6 Who Is Affected by Food Price Instability, and How? 15 2.7 Role of White Maize in Farm Incomes and Consumer Expenditures, Eastern and Southern Africa 16 2.8 Household Food Budget Shares (%) in Rural and Urban Areas 16 vi Contents 3.1 Variability in World Prices of Major Grains, 1971–2003 18 3.2 Index of Variability in Monthly White Maize Prices in Southern Africa, January 1994 to August 2003 21 3.3 Disaggregation of Variance Components in Producer Prices for Maize, Selected African Countries (%) 22 7.1 Market-Based Risk Management Instruments and Their Potential Users 46 7.2 Major Global Futures and Options Exchanges for Food Staples 48 A.1 Country-Specific Variables Used to Develop the Country Typology 73 A.2 Commodity-Specific Variables Used to Develop the Country Typology 74 Boxes 2.1 Reliance on Food Aid Can Intensify Food Price Shocks 9 2.2 A Growing White Maize Market Challenges the Rationale for Stockholding 10 2.3 Major Changes in Asia’s Food Policy Environment 13 3.1 The Imperfect Transmission of World Prices to Domestic Markets 21 3.2 Global Climate Change and Unstable Food Production 22 6.1 Direct Payments to Producers 38 6.2 Tendering of Grain Procurement for Public Distribution in Ethiopia 39 6.3 Market Information for Emergency Responses 41 6.4 Variability and Covariance of Maize Production in Africa, 1995–2004 42 6.5 Farmer-Managed Cereal Banks in Western Kenya 43 6.6 Public-Private Partnerships—The Uganda Grain Traders Model 43 7.1 The Zambian Warehouse Receipt Program 47 7.2 Example of Futures Hedging 48 7.3 The Potential for Forward Contracts 50 7.4 A Proposal for Weather Insurance in Malawi 51 7.5 Emergency Assistance Funds from International Financial Institutions 55 7.6 NEPAD’s Sobering Findings on Strategic Reserves 57 8.1 Managing the Dialogue on Wheat Policy in Pakistan 60 8.2 Key Elements of a Poverty and Social Impact Analysis 60 Acknowledgments This report was prepared by Derek Byerlee, formerly Agricultural and Rural Development Department of the World Bank, and Bob Myers and Thom Jayne, Department of Agricultural Economics, Michigan State University. Much of the report is based on a set of pa- pers prepared for an international workshop in Washington, D.C., February 28 to March 1, 2005 (see appendix 1). Special thanks are due to authors who prepared and presented these papers: Alberto Valdes, Alexander Sarris, Antonio Nucifora, Beatriz Avalos-Sartorio, Bob Myers, Chris Gilbert, Don Mitchell, Eleni Gabre-Madhin, Ghada Shields, Hector Ibarra, Jean-Charles Le Vallee, Jonathon Coulter, Jonathon Kydd, Julie Dana, Ousmane Badiane, Peter Timmer, Ralph Cummings, Jr., Rasheed Faruqee, Thom Jayne, Trina Haque, and William Foster. Coauthors, discussants, and other participants in the workshop contributed to a better understanding of the issues. This report was developed in collaboration with the U.K. Dep- artment for International Development (DfID), which provided fi- nancial support for the workshop. Neil MacPherson and Tim Foy provided valuable professional support, and Kate Nutt and Catherine Loake managed the workshop logistics very efficiently. Peer review comments from Donald Larson, Henry Gordon, and Michael Morris, and comments from Raph Cummings, Jr., and Jonathon Coulter suggested many improvements to the report. Sushma Ganguly, supervising sector manager, provided wise guid- ance and encouragement throughout the project. Jock Anderson (pro- fessional guidance), Fleurdeliza Canlas (administrative support), Kelly Cassaday (editing), and Aparna Nemana (research assistance) made important contributions. vii Acronyms and Abbreviations CLVI Cuddy–Della Valle Index CV Coefficient of variation FAO Food and Agriculture Organization GDP Gross domestic product NEPAD New Partnership for Africa’s Development PSIA Poverty and Social Impact Analysis WFP World Food Programme WTO World Trade Organization ix Executive Summary Food price risks and instability are perennial issues that have dogged food policy debates for decades. Their persistence is understandable, given the continued importance of food staples as a wage good, their high share of national income and expenditures in low-income coun- tries, and political sensitivities to sharp changes in food prices. Since the 1990s these issues have taken on new urgency in the con- text of market liberalization. The controversy over price instability and its social and political costs has arguably been the Achilles’ heel of food market reform programs, programs that have progressed very slowly in many countries, especially with regard to public food marketing agencies. In several cases, reforms have been reversed. Some are reluctant to liberalize food markets because of fears about the potential impacts on food price instability, or out of the convic- tion that food prices have become more unstable in countries that have liberalized. Others contend that “halfway� reforms create the worst of all possible worlds, where the private sector is encouraged to operate in an environment in which governments continue to in- tervene in discretionary and unpredictable ways that make prices even less stable. Over the years commodity price stabilization and risk manage- ment have received considerable attention from researchers and pol- icymakers in industrial and developing- country contexts. This new study was motivated by the need to revisit the problem of food price instability and risk in low-income countries and to investigate the benefits and costs of alternative policy responses. In particular the study aimed to provide guidance on how to make the transition from state-dominated markets to private markets in ways that do not ex- pose producers and consumers to the risk of unacceptable price spikes and collapses. Five key questions are addressed: 1. What are the sources and magnitudes of food price shocks? 2. What are the magnitudes (actual and potential) of the eco- nomic and social costs stemming from food price instability in low-income countries? 3. What is the status of food market reforms, and what can be learned from the experience to date? 4. How can countries sequence reforms in ways that promote effi- cient market development and protect the interests of the poor? 5. What are appropriate policy responses to food price instability and risk in a liberalized market environment? xi xii Managing Food Price Risks and Instability in an Environment of Market Liberalization This report marshals the “best thinking� globally to are, to a lesser extent, other landlocked African outline a framework for analyzing policy responses— countries such as Ethiopia and some Sahelian coun- conventional and emerging. The report also reviews tries. Food production in these countries is highly experiences with policy reform across low-income variable, and their capacity to operate on world countries of Asia and Africa, including relevant markets is limited by high transport costs and for- experience from other regions (particularly Latin eign exchange constraints. America, where most countries have implemented The first conclusion—obvious but too often extensive reforms). The report draws extensively overlooked—is that food policy decisions and mar- on contributions from academics and practitioners ket reforms are highly specific to their context. who shared their knowledge at an international More attention needs to be paid to a country’s par- workshop on this theme held in Washington, D.C., ticular stage of development, food consumption from February 28 to March 1, 2005 (see appendix 1 patterns, agroclimatic factors, geographical situa- for a list of papers that are available at http:// tion, and institutional setup in designing appropri- www.passlivelihoods.org.uk/default.asp?project_id ate food policies. =240&nc=4921.) Finally, this report draws on the A country typology hides considerable hetero- broader knowledge base within the World Bank geneity within countries between rural and urban and the wider development community. areas, regions, and households, but generally the consumption patterns of urban households, even THE SIZE OF THE PROBLEM poor households, have become more diversified over time, giving them more flexibility to handle Country context defines the problem and its magni- sharp spikes in the price of the dominant food sta- tude. Food policy decisions must be tailored to the ple. In rural areas, the empirical finding that emerges individual circumstances of each country, but as a consistently in most parts of the developing world starting point for identifying an appropriate policy is that a majority of households are net food buy- response, countries can be grouped roughly accord- ers, while a relatively small minority of wealthier ing to common needs and risks. A simple framework households are grain sellers. The poor, who are over- for classification focuses on low-income countries whelmingly net food purchasers, suffer dispropor- and regions in which food consumption is domi- tionately from high food prices. Among producers, nated by one staple: rice in Asia and Madagascar, the impacts of low food prices are at least partially wheat in Pakistan and in the Middle East and North offset by the tendency for prices and output to be Africa, white maize in eastern and southern Africa, negatively correlated. and millet and sorghum in Sahelian countries of This leads to a second major conclusion: Food pol- West Africa. These are the countries and regions icy should generally emphasize the impacts of un- where the poor are most exposed to sharp move- stable food prices on consumers—rural and urban, ments in the prices of food staples, especially spikes and especially the poorest and most vulnerable— in the prices paid by consumers. These countries more than impacts on producers. were further classified according to their potential exposure to price shocks from domestic climatic events and to shocks generated by instability in HOW SIGNIFICANT ARE world grain markets. FOOD PRICE SHOCKS? Based on this classification, rice and wheat im- porters, especially in the least developed countries At the global level, variability in world grain prices such as Madagascar, Bangladesh, and the Republic remains significant, with coefficients of variation of Yemen, are most exposed to world price shocks. around trend of 20 to 30 percent for rice, wheat, and Many other Asian and middle-income countries are white maize. Although there is no evidence that exposed as well, but their greatly improved infra- variability has increased—indeed, prices were most structure and foreign exchange reserves place them unstable in the 1970s—there is concern that changes in a much better position to handle such shocks in world markets, especially reductions in the stocks than three decades ago, when many public food held by major producers (China, the United States, marketing agencies were established. Landlocked and the European Union) and rapid growth in de- countries in southern Africa that depend on maize mand in Asia, may provoke higher and less stable are most exposed to domestic sources of shocks, as prices in the future. Executive Summary xiii The evidence is limited on the magnitude and not large in most cases. The most persuasive cases frequency of price instability in domestic food mar- for the negative effects of high food prices can be kets, actual and potential. In general, producer prices made for effects on household food security and for wheat and maize in importing countries have nutrition and on macroeconomic performance. These been more stable than international prices, reflect- costs could be significant in certain situations—for ing transactions costs of transmitting international example, in the poorest countries with poor in- prices into domestic markets, as well as continuing frastructure, weak capacity to import, dependence policy interventions in many countries that insu- on a single dominant staple, and susceptibility to late domestic markets from world prices. There is drought—all characteristics of several landlocked no convincing evidence to date that domestic food countries in Africa. price instability has increased over time in the sam- ple of countries reviewed. Domestic price instability tends to be highest in LESSONS AND EXPERIENCES two groups of countries. Latin American countries, FROM POLICY REFORMS where macroeconomic shocks, especially sharp ex- The record of food market reforms in low-income change rate devaluations, have resulted in highly and even many middle-income countries is mixed at unstable prices, comprise the first group. African best. Some countries, such as India, have maintained countries, especially landlocked countries where their parastatal systems more or less intact, but the wedge between export and import prices is high mounting costs have made most of these systems because of high transport costs and poor market unsustainable. Other countries, such as Bangladesh, infrastructure, comprise the second group. The high Mali, and Mozambique, have introduced and sus- import-export parity wedge, combined with high tained significant reforms that have enabled them domestic production variability, increases the impact to weather a major natural disaster at a much lower of domestic shocks, especially drought, on prices. A cost than in the past, with tolerable levels of price contributing factor, particularly in southern Africa, is instability. Notably, these countries have exploited the uncertainty created by unpredictable govern- trade opportunities, especially regional trade, as ment interventions in food markets and imports. the main mechanism for stabilizing domestic grain Under a full market liberalization scenario, food prices. price shocks, whether from global or domestic But what about the many countries stuck halfway sources, are potentially significant in many situa- in the reform process, hovering between old paras- tions. For example, in Ethiopia the price wedge tatal models and private, market-led approaches? In between import and export parity has allowed this situation, discretionary interventions to meet maize prices in Addis Ababa to fluctuate from about an emergency (or even just a declaration of the in- US$50 to nearly US$250 per ton in recent years, and tention to intervene) have been especially destruc- probably more in more remote regions. Likewise, tive to incentives for private-sector participation. countries depending on rice imports have faced Other important lessons have been learned from world export prices falling from US$340 per ton in the varied experience with market reforms. Many 1996 to a low of US$170 per ton in 2001, and re- countries paid insufficient attention to designing bounding to more than US$300 per ton in 2005. an orderly sequence of reforms that systematically increased the role of the private sector and built confidence in a market-based approach. Nor was WHAT ARE THE COSTS sufficient attention given to political economy con- OF PRICE INSTABILITY? siderations (such as vested interests that maneuver The costs of unstable food prices can include the to maintain the status quo) and to designing a re- loss of economic efficiency, detrimental impacts on form program that takes account of these realities. the welfare of the poor (including undernutrition and reduced survival), and negative macroeconomic MOVING FORWARD: externalities that retard economic growth. There is little consensus and generally weak evidence on BROADER POLICY OPTIONS the magnitude of these costs. The effects of unsta- Policies are chosen within a set of constraints ble food prices on economic efficiency are probably formed by the political system and by limitations xiv Managing Food Price Risks and Instability in an Environment of Market Liberalization on availability of public funds. These constraints production and market information systems, force governments to make explicit tradeoffs in storage, and market networks are often the allocating public expenditures, and it is impera- first priorities for improving food market tive that these tradeoffs are made in ways that performance. enhance the long-run performance, growth, and sta- 3. Liberalize trade, especially by promoting re- bility of the food sector and the economy as a whole. gional trade, for one of the most effective This study highlights a number of policy options “quick wins� for reducing food price volatility for moving forward, recognizing that it is especially in small and medium-size countries. Liberal- difficult to make generic recommendations for such ization of trade shifts a country’s exposure a country-specific and complex topic. One general away from domestic shocks and toward glo- recommendation is that food policy decisions, rather bal price shocks, but global shocks are usually than focusing on price stabilization options per se, lessened if trade with neighboring countries is should take a holistic approach to food security in encouraged. Regional trade requires action on which long-run productivity growth and market a number of fronts, including long-run invest- development constitute the first priority. This leads ments in infrastructure, but the development to four specific recommendations: of (a) consistent rule-based policies to lift dis- cretionary export bans and import restric- 1. Address problems of food price instability tions, (b) smooth border-clearing procedures, and food insecurity in a holistic framework and (c) harmonized regulations, such as phy- that includes: tosanitary rules, would go a long way toward • Measures to improve overall productivity creating the incentives for private traders to of food staples, especially investments in engage in regional trade. research and development and irrigation 4. Sequence market reforms in a consistent man- • Measures to reduce the severity of domes- ner that creates space for the private sector to tic shocks caused by climatic events (such operate. “Big bang� approaches to market re- as measures to promote irrigation or crop form have rarely worked in practice. For mar- diversification) kets, including regional markets, to develop • Measures to improve the overall efficiency of over the long run, consistent progress must be markets, including investments in transport made in opening space for the private sector. and communication infrastructure, storage, More analytical work and policy dialogue will information systems, market regulations, provide a better basis for designing a logical, and institutional arrangements that improve sequential program of reforms. Finally, gov- coordination along the market chain ernments should implement the agreed pro- • Measures to mitigate the impacts of shocks, gram in a predictable and consistent manner. including market-based measures (such as A generic sequence that would gradually in- forward pricing and weather insurance) as crease the role of the private sector includes: well as countercyclical safety nets. • Eliminating blanket subsidies and revising The corollary of this recommendation is that remaining subsidies in ways that level the direct public interventions in food markets to playing field for the private sector and tar- manage food price risk should be a last resort. get the poor 2. Reallocate resources from short-run, “fire- • Removing remaining restrictions on grain fighting� interventions to manage food prices, movement within a country and reducing to investment in long-run market and private- restrictions on grain imports and exports sector development, including incentive frame- • Moving away from fixed procurement and works, market institutions, and infrastructure release prices toward seasonally adjusted consistent with item 1 above. Nonetheless, prices and price bands even investments in market development must • Tendering remaining public procurement, be sequenced in ways that confer measurable imports, and even storage to the private gains in the short to medium term. Public- sector, using a highly transparent process to private partnerships (for example, through increase efficiency, reduce rent-seeking, and farmer and trader associations) to develop build private-sector capacity. Executive Summary xv SPECIFIC POLICY OPTIONS FOR kets. These alternatives are already available where the basis risk is low, which appears to MANAGING PRICE INSTABILITY be the case for wheat and white maize for AND RISK many countries, using U.S. and South African Within an overall public policy strategy for food futures markets. systems that emphasizes the transition to private • Weather-indexed insurance for use by farm- markets and long-run market development, there ers, safety net programs, and (potentially) are roles for the public sector in enhancing price sta- consumers. While not designed specifically bility and managing food sector risks. Two of these for price risk management, weather-indexed will be a standard part of the toolkit of most food insurance can mitigate the impacts of price security strategies: piloting and facilitating the adop- spikes or climatic shocks. Successfully piloted tion of various market-based risk management at the farm level in India and Mexico, weather instruments, and countercyclical safety nets. Two insurance can be used more widely where others may have a role in certain situations and, weather indices are good proxies for crop when accompanied by specific safeguards to ensure losses, and especially if domestic insurers can “arm’s-length,� rule-based management: variable reinsure on global markets. tariffs and strategic reserves. The public sector should support the development of a basic enabling environment for these instru- ments by conducting the analytical work and build- Market-Based Risk Management Instruments ing the capacity to pilot and scale up programs Several risk management instruments show con- that promote the development of financial systems, siderable promise in managing food price risks, in- communication and information systems, regula- cluding facilitation of private storage (warehouse tions, and an appropriate business climate. receipt systems), futures and options markets, and Some recent discussions have also noted the po- weather-indexed insurance. These alternatives are tential for the public sector to use market-based in- rarely used in low-income countries, partly because struments to reduce exposure to risks from its own the public sector dominates food markets and partly operations in food markets. Yet direct trading of because the enabling conditions are lacking, such futures, options, or insurance contracts by govern- as access to finance, information systems, commu- ments or public food agencies should be approached nication systems, market regulations, and capacity. with extreme caution. Large government futures or The major focus of the public sector should be to options positions are not recommended for two rea- create an environment that facilitates the private sons. First, even if the public sector is successful in sector’s adoption of these instruments, especially using these instruments, the public sector is likely to in the following ways: undermine incentives for the private sector to use them. Second, given the poor record of public-sector • Warehouse receipts, for use initially by larger- interventions in food markets, there is little reason to scale farmers, processors, and traders, and believe that the public sector’s use of market-based over the longer term by the small-scale sector. risk management instruments would be immune to Warehouse receipts have much potential to re- the same inefficiencies and rent-seeking forces that duce risks from seasonal price fluctuations, have plagued conventional public food agency op- develop finance markets, encourage invest- erations. ment in storage, and eventually (when widely If governments do choose to become involved in adopted) to reduce both seasonal and interan- direct procurement to manage a small strategic food nual price fluctuations. They cannot be im- reserve, market-based risk management strategies plemented if an appropriate regulatory and may have a potential role in these operations. In business environment is lacking, however. such cases, options have distinct advantages over • Futures and options using existing global mar- futures—first, because of their role as price insur- kets, for use mainly by large-scale traders and ance, and second, because purchasing options re- processors and strong intermediaries, such as quires only a single, up-front premium, whereas well-developed farmer or trader associations, futures can entail continuing margin calls if prices to reduce exposure to risks from global mar- move unfavorably. Even when using options, an xvi Managing Food Price Risks and Instability in an Environment of Market Liberalization effective hedging strategy requires considerable in- must be approved by the World Trade Organization vestments in analytical capacity and a long-run com- (WTO), and indeed a preferable outcome would be mitment, otherwise hedging could add to risk rather for the triggers and monitoring of their implementa- than reduce it. The misuse of futures and options tion to be subject to WTO oversight to maintain max- may expose governments to even greater fiscal risks imum transparency. and rent-seeking than conventional public-sector Technically, variable tariffs could also be used to operations in food markets, unless special manage- reduce risks from price spikes in global markets, but ment safeguards are in place. tariffs must be high enough initially that they can be lowered when world prices rise sharply. Given that high tariffs on food grains are generally undesirable Countercyclical Safety Nets for both efficiency and equity reasons (most poor A second major priority for interventions to man- households, including rural households, are net food age risks is to support the development of counter- purchasers), variable tariffs are unlikely to be useful cyclical safety nets in ways that are market friendly. for managing world price spikes. Countercyclical safety nets, which kick in when high food prices or low production threaten house- hold food security, are an integral part of any pro- Strategic Reserves gram to manage food price risks. Food aid and Many countries still maintain publicly owned re- food-for-work programs remain the most impor- serves to reduce food price instability. In a liberalized tant safety nets in many countries. In the past, how- market economy, the primary reason to maintain ever, untimely imports and sales of food aid, along such reserves should be a targeted food distribu- with poor targeting, often undermined market de- tion scheme (if there is one), although in a few cases velopment. Food aid and other safety net programs reserves can be maintained to cope with emergen- can support long-run market development by: cies, especially in landlocked countries with poor infrastructure. In some cases, reserves may be large • Converting from food to cash transfers where enough to influence domestic market prices, and food markets already function reasonably well judicious use of these reserves may help reduce the • Scaling up local and regional procurement of impact of domestic shocks on food prices, espe- food aid, perhaps including the maintenance of cially where there is a large wedge between import a small and well-managed emergency reserve, but ensuring that the timing of food aid pro- and export parity prices. Critical safeguards must curement does not aggravate price instability be in place, however, to ensure that operations of • Incorporating rainfall insurance into safety food reserve agencies do not destabilize markets. net programs to enhance their ability to trig- These safeguards include arm’s-length, “central ger timely and better-targeted responses to a bank�–type autonomy, highly professional man- drought agement and analytical capacity, strict rule-based • Better targeting of food aid through im- market operations to meet a narrowly defined ob- proved information systems and the use of jective, and tendering of operations, including stor- self-targeting approaches, including “infe- age, to the private sector. rior� grains • Integrating safety nets with market develop- ment activities, such as the use of food aid to THE IMPORTANCE OF construct local market infrastructure. LOCAL CIRCUMSTANCES Returning to the country typology discussed earlier, it is clear that food policy design and approaches to Variable Tariffs managing food sector risks will vary widely, de- Under certain circumstances, variable tariffs can be pending on each country’s context. The overall pri- used to manage downside price risks to producers orities on productivity enhancement and market from exposure to global markets. To be effective, development are fairly generic; they apply in many variable tariffs should be triggered by well-defined contexts. However, quite different strategies will rules to reduce political capture and be highly trans- emerge across countries and regions when moving parent in their operation. Technically, their use also to sequenced reforms, creating space for the private Executive Summary xvii sector, and addressing specific priorities for manag- market reform will be more effective if it seeks wide ing market risks. The Asian countries, in particular, stakeholder dialogue and pays special attention to still have a considerable reform agenda to open transitional and sequencing arrangements that space for the private sector. Likewise, the opportu- mitigate the negative effects of policy changes on nity to apply various market-based risk instruments particular groups. The use of Poverty and Social depends significantly on the extent that a country is Impact Analyses (PSIAs) to ensure wide buy-in and exposed to domestic versus global shocks. ownership in this delicate reform process is a step in the right direction and should be scaled up. ENTRY POINTS FOR THE WORLD BANK Pilot and Evaluate New Market-Based Instruments Food market reform and food security remain crit- ical areas for Bank engagement. Interest in these The recent move by the Bank’s commodity-based issues is burgeoning in many countries, including risk management group to analyze the applicabil- those which have not yet embarked seriously on re- ity of market-based risk management instruments forms and those which seem stuck halfway through for food staples is providing encouraging results the process. The Bank needs to revamp its analyt- and should be scaled up. However, this work should ical work in this critical area, paying particular focus on analytical support and capacity building attention to the following points. to facilitate adoption of these instruments by the private sector and to promote the emergence of necessary institutions and intermediaries. Extreme Manage the Policy Dialogue Better caution should be used in promoting use of these Too often, the Bank’s analytical work has proposed approaches by public food marketing or strategic broad recommendations on market reforms but reserve agencies. paid little attention to how those reforms should be sequenced. The “big bang� approaches generally Support Activities at the Regional have not worked, and part of the challenge in mov- and Global Levels ing forward is to be alert for opportunities to move toward second- and even third-best options rather This report has highlighted the potential for regional than waiting for the opportunity for full reform. trade as a mechanism to stabilize prices within a Good analytical work will have to be combined with region, and this prospect raises a huge agenda for much more time- and resource-intensive policy analytical work and policy dialogue to reduce pol- dialogue that is attuned to political realities (for icy and institutional barriers to trade in nearly all example, vested interests). Advice on food grain regions. 1 Introduction Food price risks and instability are issues that have dogged food pol- icy debates for decades, with good reason. Unstable prices for im- portant food staples, such as maize, rice, and wheat, can have acute economic, social, and political consequences (Timmer 1995). Highly unstable prices can lead to inef�cient agricultural production deci- sions, especially when markets for credit and risk are poorly devel- oped.1 The human costs of food price shocks can be disastrous for the poor, because food staples often constitute a large share of poor farm- ers’ incomes and poor consumers’ expenditures. Food price instabil- ity is a frequent forerunner of macroeconomic shocks and political turmoil, which discourage long-run investment and curtail growth. Food prices can become extremely unstable and risky as a result of climatic events, world price fluctuations, an inelastic supply-and- demand response in domestic markets, and high transportation costs. In many low-income countries, the potential for food price risks is fur- ther increased by weak market infrastructure, a poorly developed pri- vate sector, and incomplete or poorly functioning �nancial and risk markets. A growing concern is that these long-acknowledged sources of instability are being aggravated by less familiar forces. Commodity stocks can buffer price instability, but current world stocks for grains are at historically low levels, and even relatively small swings in ex- ports or imports from large countries such as China could send major shock waves through world grain markets (Mitchell and Le Vallee 2005). Climatic cycles and global climate change may increase devel- oping countries’ exposure to droughts, floods, and other extreme cli- matic events that heighten the risk of severe fluctuations in food production.2 Until the 1980s, the traditional policy response to food price insta- bility in developing countries was direct government intervention. Governments orchestrated the purchase and sale of food, controlled food prices, and restricted internal and external trade, usually through grain marketing parastatals. While these interventions may have re- duced price instability and risk, in many cases they also imposed major economic costs (Schiff and Valdes 1992). Aside from the high costs that are often observed when public institutions take on mar- keting functions, direct government intervention is frequently sus- ceptible to rent-seeking and inequitable distribution of bene�ts.3 Over time, such interventions have led to changes in domestic price levels (which often fall below border prices), high treasury costs, and large income transfers (often from the poor to the wealthy; see Jayne and Rukuni 1993). 1 2 Managing Food Price Risks and Instability in an Environment of Market Liberalization By the 1980s, direct intervention was viewed crashes in critical food markets, with unacceptable widely as a major impediment to the growth and de- human, economic, and political costs. velopment of the food sector in developing coun- Commodity price stabilization and risk manage- tries.4 Donors and many governments began to ment have received considerable attention from re- promote the reform of food marketing and price poli- searchers and policymakers over the years, in both cies as a central component of structural adjustment industrial and developing- country contexts. This programs. The success of these market reforms in new study of the food price instability and risk prob- providing positive price incentives to farmers within lem in low-income countries investigates the bene- tolerable bounds of price instability, and the extent to �ts and costs of alternative policy responses and, which they have actually opened markets to the pri- more particularly, provides guidance on how to vate sector, have been the subjects of considerable de- make the transition from state-dominated markets bate.5 Even so, it is clear that many countries have to private markets in ways that do not expose pro- implemented food market reforms only partially, ducers and consumers to the risk of unacceptable and that deeper reforms are seriously constrained by price spikes and collapses. concerns over increased food price instability and Five important questions are addressed: risk. A number of countries have reversed the reform process and re-established quasi-governmental pro- 1. What are the sources and magnitudes of food grams for procuring, storing, and importing food.6 price shocks? 2. What are the magnitudes (actual and poten- tial) of the economic and social costs stemming OBJECTIVES OF THE STUDY from food price instability in low-income The search for appropriate policy responses to food countries? price risk and instability has again become a major 3. What is the status of food market reforms, and and contentious issue. There is no clear consensus what can be learned from the experience to regarding how best to deal with problems of food date? price risk and instability, especially in low-income 4. How can countries sequence reforms in ways countries and especially in the context of continued that promote ef�cient market development market reform. and protect the interests of the poor? For the past several decades, the World Bank’s po- 5. What are appropriate policy responses to food sition on food marketing policy has rested on three price instability and risk in a liberalized mar- planks (Meerman 1997): (1) liberalize food markets ket environment? and reduce direct government purchasing and sell- ing; (2) encourage the development of private-sector This report marshals the “best thinking� globally marketing services and innovation by investing to outline a framework for analyzing policy in public goods, such as marketing infrastructure, responses—conventional and emerging—as well as market information, and grades and standards experiences with policy reform. The report reviews systems; and (3) put greater reliance on interna- experiences with policy reform across low-income tional and regional trade, rather than government- countries of Asia and Africa, including relevant managed buffer stocks, to even out local imbalances experience from other regions (particularly Latin in supply and demand. America, where most countries have implemented Some countries embraced most elements of this extensive reforms). The report draws extensively on approach with success; examples include Mali, contributions from academics and practitioners Mozambique, Bangladesh, Vietnam, and many who shared their knowledge at an international countries in Latin America. Governments of most workshop on this subject,7 and on the broader other countries, however, especially in eastern and knowledge base within the World Bank and the southern Africa, South Asia, the Middle East, and wider development community. North Africa, still intervene heavily in food mar- Two additional points will clarify the scope of kets. Part of the problem is that the World Bank and the study. First, the terms “food price instability� others have been ineffective in providing guidance and “food price risk� are both used in this report. to countries on how to manage the transition from Food price instability refers to any abrupt change in public- to private-market operations. It is often ar- price, irrespective of whether the change is pre- gued that the complete liberalization of markets dictable. But price fluctuations can arise from un- will expose countries to high risks of price spikes or predictable shocks as well as from predictable Introduction 3 trends or seasonal patterns. Risk is associated only price risks and instability. From the outset, it is rec- with the unpredictable shocks. Many of the costs ognized that because the problem is highly speci�c stemming from volatile food prices, especially on to a given context, it may be helpful to examine the the producer side, are associated with risk rather problem through an analytical framework that at- than instability per se. However, predictable but tempts to take major differences in local context into extreme price movements can also have signifi- account. A typology of countries and also of house- cant costs in terms of consumer welfare and macro- holds is constructed, based on secondary data, to as- economic instability. sess potential exposure to different sources of price Second, the management of food price risk and risk and instability. Information on the nature instability cannot be separated from the wider issue and extent of food price instability and risk in of food market reforms, and this report necessarily low-income countries, in terms of both global price addresses both themes. Price instability, real or per- and domestic production shocks, is also provided, ceived, is a critical influence on the pace of reforms along with available information on the costs of and the extent of market liberalization. As demon- food price risk and instability. strated later in this report, food price instability is to Next, the report focuses on past and prospective some extent itself a manifestation of the way reforms policy responses to the problem. Lessons for future are implemented. For these reasons, the issues and policy dialogues are distilled from a review of ex- policies discussed in this report are in many ways periences with food market reform. Speci�c policy broader and more complex than might be expected options are described for managing the transition in a study of food price instability and price stabi- to private markets, including market-based instru- lization policy. ments for managing risk, “quick wins� to foster private market development, and the role of safety nets. The final sections of the report focus on OUTLINE OF THE REPORT ways of advancing the policy dialogue in this po- This report begins by marshaling information for litically sensitive area and present conclusions understanding and analyzing the problem of food and recommendations. 2 Toward a Typology for Food Policy Analysis The sources, size, and consequences of food price risk and instabil- ity vary substantially across and within countries. These elements depend on a country’s speci�c situation as well as on the local and household characteristics within that country. Similarly, the appro- priate policy response to food price risk and instability will vary across and within countries because of differences in geography, patterns of food production and consumption, and institutional ca- pacity to implement alternative policies. What might be appropriate for rice in Indonesia may not work for maize in Ethiopia, and vice versa. This chapter provides a framework for identifying populations that face price risks from different sources, mainly from world price instability and domestic supply shocks. It develops a “macrotypol- ogy� of low-income countries based on secondary data that indicate the likely degree of a country’s exposure to domestic weather shocks and global price shocks. The second part of the chapter develops a ty- pology of households that indicates likely differences in the way that food price instability and risk will affect different types of house- holds. The discussion throughout the chapter recognizes that this framework is only a starting point, and that country and household situations change over time, sometimes quite rapidly. A MACROTYPOLOGY OF COUNTRIES Several criteria are used to classify countries in terms of their expo- sure to global and domestic sources of food price shocks. Income level is the �rst criterion: Low-income countries are most likely to be affected by price shocks because of the high share of food staples in national income, and because they have less means to cope with shocks. Twenty-�ve low-income countries (income status was based on the World Bank Atlas method of classi�cation)8 with a population of more than 10 million were selected for analysis. For comparison, four lower-middle-income and two upper-middle-income countries were included. The dominant food staple within each country was identi�ed from Food and Agricultural Organization (FAO) food balance sheets for 2002 (table 2.1). In some countries, no single food staple dominated, so two important food staples were included. Secondary data sources, mainly from FAOSTAT, were used to further classify countries ac- cording to several criteria. 5 6 Managing Food Price Risks and Instability in an Environment of Market Liberalization Table 2.1 Countries, Income Status, and Dominant Staples Based on Calorie Intake Country Country Code Dominant Staple Income Status Dominant Staple, 2002 Bangladesh Ban Rice L 1,631 Burkina Faso Buf Millet/Sorghum L 1,265 Cambodia Cam Rice L 1,420 Cameroon Cmr Maize L 418 Cassava L 279 Chile Chi Wheat UM 909 Cote d’Ivoire CIv Rice L 577 Egypt, Arab Rep. Egt Wheat LM 1,093 Maize LM 594 Ethiopia Eth Maize L 396 Wheat L 302 Ghana Gha Cassava L 635 Maize L 364 India Ind Rice L 827 Wheat L 498 Indonesia Ina Rice LM 1,465 Kenya Ken Maize L 714 Madagascar Mas Rice L 973 Malawi Mlw Maize L 1,126 Mali Mal Millet/Sorghum L 781 Mexico Mex Maize UM 1,057 Morocco Mor Wheat LM 1,343 Mozambique Moz Cassava L 719 Maize L 534 Nepal Nep Rice L 941 Niger Ngr Millet/Sorghum L 1,424 Nigeria Nia Millet/Sorghum L 646 Pakistan Pak Wheat L 999 Senegal Sen Rice L 746 South Africa S.Af Maize LM 936 Sudan Sud Millet/Sorghum L 737 Tanzania Tan Maize L 657 Uganda Uga Plantains L 447 Cassava L 307 Maize L 260 Vietnam Vtm Rice L 1,662 Yemen, Rep. Yem Wheat L 912 Zambia Zam Maize L 1,088 Zimbabwe Zim Maize L 705 Source: Authors Note: L = low income, UM = upper middle income, LM = lower middle income. Diversity in Domestic Food Consumption Upward price spikes can severely jeopardize the welfare of low-income consumers. The extent to which consumption is concentrated Of course, much depends on the degree to which on one staple food commodity is probably the sin- one commodity can be substituted for another. gle most important variable influencing vulnera- From a policy perspective, fluctuations in supplies bility as well as political sensitivity to unstable food and prices in one market can be partially absorbed prices. When consumption is highly concentrated by other markets, to the extent that consumer de- on one staple, the implication is that the staple mand is flexible enough to shift to substitute foods makes up a large share of consumer expenditures. when the price of one food staple rises. In most Toward a Typology for Food Policy Analysis 7 cases, there is scope for substitution between the se- Figure 2.1 Diversity of Staple Consumption, 2002 lected commodities, but it is far from perfect. For example, in Africa millet and sorghum are very 1.0 close substitutes and have been grouped into one commodity for this analysis. However, the domi- 0.8 Vtm Diversity (HH Index) nance of rice in Indonesian and Bangladeshi con- sumption patterns and the very limited potential to Cam Pak 0.6 Chl Mex shift to other staples restrict the potential for using Ngr Zam trade in other staples as a strategy to moderate vari- 0.4 Sud ability in rice prices. Even within commodities, Moz strong preferences often emerge for particular va- 0.2 Uga rieties or grain types, such as the preference for Nia Eth Cmr Nia Nia white maize over yellow maize, with the result that 0.0 Rice Wheat Maize Millet/ Cassava/ commodity-speci�c data can hide cases of imper- Sorghum Plantain fect within-commodity substitution. The dominance of one commodity is calculated Note: Country abbreviations are listed in table 2.1. Source: Authors. using the Her�ndahl-Hirschman Index, which is commonly used to measure the market share or, in this case, the share Si of calories from different starchy food staples in each country. The index, HHI = Σ(Si)2, is 1 if only one staple is consumed, and climatic events will dominate the sources of price in- zero for an in�nitely diverse consumption basket. stability. This predominance is especially marked in The index is summarized in �gure 2.1, and indi- landlocked countries or in large countries with very vidual data are given in appendix 1. poor infrastructure, where there is a wide wedge be- The countries that tend to concentrate most on tween import and export parity prices (Byerlee and one staple are the rice economies of Southeast Asia, Morris 1993). In landlocked Zambia, for example, but some countries are also highly dependent on export and import parity prices for white maize can wheat (Pakistan, Morocco, Yemen, and Chile), differ by US$150 per ton—more than the normal CIF maize (Mexico and the countries of southern Africa), port price of maize. Likewise in another landlocked and millet/sorghum (Burkina Faso, Mali, and country, Ethiopia, the elasticity of price transmission Niger). Countries in which cassava is a major staple, between global cereal prices and prices in the capi- and coastal countries of West and Central Africa, tal, Addis Ababa, is estimated to be 0.8, but it falls to generally have the most diverse food consumption less than 0.2 for many of the more remote producing baskets, with diversity indices of 0.25 or less. regions (Nicita 2005). These differences in tradability were captured in three variables: (1) the consistency of trade sta- Dependence on Trade and tus, measured by the number of years in the past Access to Global Markets 10 years that a country has imported or exported; From the perspective of food price policy, one of (2) the dependence on imports as a percentage of the most important characteristics of a country is utilization; and (3) coastal versus landlocked sta- whether it consistently imports, exports, or fluctu- tus.9 The results by commodity and country (�g- ates between importing and exporting its major food ures 2.2 and 2.3) reveal very strong differences in staple. If a country is a consistent food importer tradable status. and its markets function reasonably well, domestic Rice is universally traded, and all the countries in prices should move in line with import parity prices. which rice is a staple food are exposed to global price Similarly, if a country is a consistent exporter, then shocks, mostly as regular importers or, in the case of domestic prices should move in line with export par- India and Vietnam, as exporters. Wheat is also ity prices. In both cases, domestic price instability highly traded (in this case all countries are net im- will be determined largely by global price shocks. porters). Countries in West Africa where rice is im- But when a country fluctuates periodically between portant, and in the Middle East and North Africa import and export status or a commodity is not trad- where wheat is important, also import a relatively able (for example, cassava), domestic shocks from high share of their consumption. 8 Managing Food Price Risks and Instability in an Environment of Market Liberalization of surplus production and high transport costs can Figure 2.2 Trade Status of Dominant Staple cause supply gluts and depressed food prices. Commodities, 1994–2003 Finally, it should be noted that the middle-income countries included in the sample are relatively 15 more dependent on trade to supply staple food Net imports (no. of years from 94–03) All other Chi, Egt, Cmr, Egt, Ngr needs (often one-quarter or more of total needs). 10 countries Mex, Moz Yem Sud, Eth 5 Nia Capacity to Meet Food Import Requirements Ind S.Af Nia on Commercial Terms 0 For a country that must consistently import its pri- –5 mary food staple, the impact of either domestic or global price shocks depends in part on its capacity –10 Ind, Vtm Gha to import additional food from world markets. This capacity was measured by the value of food im- –15 ports as a share of foreign exchange reserves (fig- Rice Wheat Maize Millet/ Cassava/ Sorghum Plantain ure 2.4).10 Nearly all of the countries with the lowest capacity to import are in Africa, although Note: Country abbreviations are listed in table 2.1. Bangladesh and Yemen would use one-quarter or Source: Authors. more of their foreign exchange reserves just to meet their average annual cereal imports. Based on these criteria, Ethiopia, Malawi, Sudan, and Yemen are At the other extreme, millet/sorghum and cas- highly exposed to price shocks. sava are rarely of�cially traded, and some African Given their weak capacity to import food com- countries engage very little in trading maize, either mercially, many low-income countries depend on because they are self-suf�cient in most years or be- food aid to meet food supply gaps. This is espe- cause high transport costs impede trade, especially cially so in Africa, where several countries depend in landlocked countries such as Ethiopia. For maize on food aid for 10–15 percent of consumption. in Africa, the trend over the past decade has been Although food aid reduces demands on foreign ex- for countries such as Zambia, Malawi, and Kenya change and may help overcome domestic food to become importers with increasing regularity, shortfalls, dependence on food aid may weaken a although in years of good rainfall, the combination country’s ability to manage world price shocks, and the management of local food aid procurement and release may exacerbate domestic price insta- Figure 2.3 Import Share of Staple Food Utilization, bility (box 2.1). 1994–2003 A �nal issue in assessing vulnerability in expo- sure to trade is the availability of the required 100 grains on world markets and the size of a country’s Yem imports in relation to world market volumes. The 80 Sen widespread intervention in domestic rice markets 60 in Asia from the 1960s was founded on the belief Percent Egt Egt that world rice markets were too thin to rely on 40 imports to manage domestic shocks. However, CIv 20 Mex Rashid, Cummings, and Gulati (2005) show that rice markets have become more robust over time 0 Ind and that all but a few of the largest rice-consuming S.Af Vtm countries could participate in rice markets without –20 Rice Wheat Maize Millet/ Cassava/ influencing prices. A similar situation has occurred Sorghum Plantain in Africa for white maize. Almost all maize traded Note: Country abbreviations are listed in table 2.1. outside the region was yellow maize, but this situ- Source: Authors. ation has changed in recent years (box 2.2). For nearly all of the countries in this sample, trade in Toward a Typology for Food Policy Analysis 9 staple food commodities is not large enough to in- Figure 2.4 Value of Average Annual Cereal Imports as fluence world market conditions, so this consider- a Share of Foreign Exchange Reserves, ation is not discussed further, although it might be 1994–2003 important in a few cases, such as rice in Indonesia and white maize in South Africa. 120 Percent of cereal imports 100 Sud Variability in Domestic Food Production to reserves (94–03) When domestic events, such as bad weather, cause 80 Nig prices for nontradable commodities to soar, the 60 magnitude of these shocks will be closely related to Sen the variability of domestic production. Many fac- 40 Ban Yem tors are responsible for variability in production, Mor Mex but two dominate. The �rst is the size of a country. 20 Pak Egt Uga Chi Ind Mal Larger countries typically have more diverse re- 0 gional climatic conditions that reduce risks for the Asia Middle East/ Sub-saharan Latin N. Africa Africa America country as a whole. The second is drought. The magnitude of drought shocks depends on the level Note: Country abbreviations are listed in table 2.1. Source: Authors. of rainfall and on whether irrigation is available. Production variability around trend was sum- marized for each country/commodity combination using the Cuddy–Della Valle Index (CLVI): Singh and Byerlee (1990) show a signi�cant nega- tive relationship across countries between the CV of CLVI = CV (1 − R 2 ) , 0.5 national wheat yields and both (1) the percentage wheat area under irrigation and (2) the total area of where CV is the unadjusted coef�cient of variation wheat cultivated in a country. By contrast the CLVI over 1994–2003 and R2 is the coef�cient of determi- for African maize, rice, wheat, and millet/sorghum nation for the log-linear time trend regression over generally exceeds 15 percent. It surpasses 20 per- the same period (Cuddy and Della Valle 1978). cent in southern Africa, partly because of the re- Because Asian countries are large and a major gion’s dependence on rain-fed rather than irrigated share of food crop production occurs under irriga- agriculture rainfall, and partly because country tion, the magnitude of production variability is gen- sizes are small and regional trade potential is not erally low, on the order of 2–7 percent (�gure 2.5). well developed.11 Box 2.1 Reliance on Food Aid Can Intensify Food Price Shocks Several African countries are relatively dependent on respect to world grain prices may therefore be consid- food aid (appendix 2). This dependence poses two erably higher than one for countries that depend on dilemmas in managing food price shocks. First, inter- food aid [Taylor and Byerlee (1991)]. Second, the un- national supplies of food aid are negatively correlated timely release of food aid into local markets or—in with world grain prices. In other words, when world countries where food aid is procured locally—the un- prices increase sharply, the availability of food aid de- timely procurement of grain can contribute to price in- creases. The burden on foreign exchange increases, stability in local markets. For example, in Ethiopia, because countries may have to increase commercial more food aid is procured just before harvest when food imports to offset food aid shortfalls just when the prices are highest, further increasing seasonal price price of commercial imports is high. The elasticity of swings (World Bank, Forthcoming). foreign exchange requirements for grain imports with Source: Taylor and Byerlee (1991); Barrett and Maxwell (2005); World Bank (Forthcoming). 10 Managing Food Price Risks and Instability in an Environment of Market Liberalization Box 2.2 A Growing White Maize Market Challenges the Rationale For Stockholding Until recently, the world market for white maize was large white maize supply response in the United States thinly traded, so small absolute changes in import de- to export to Mexico. These developments have miti- mand in southern Africa had the potential to influence gated the potential for white maize prices and supplies world prices. The rationale for some level of stockhold- to become tight when southern Africa experiences a ing is more compelling in such cases. As a result of the drought and have thus reduced the rationale for keep- North American Free Trade Agreement (NAFTA), how- ing large government stockpiles of white maize to sta- ever, the white maize market has become much more bilize supplies [Tschirley and others (2004)]. heavily traded. Since 1997, NAFTA has induced a Estimated World Exports of White Maize (000 Tons) 1976–1980 1981–1985 1986–1990 1991–1997 1998–2003 Average Average Average Average Average South Africa 621 888 972 1132 621 United States 156 126 272 505 1,254 World total 1,197 1,579 1,930 Na Na Source: Tschirley and others (2004). Combining Criteria into a Country Typology countries that appear to be especially vulnerable to domestic production shocks. When the criteria discussed previously are com- In each table, the �rst-level division (see the �rst bined and countries are grouped according to their column) is based on the dietary concentration of median levels for major variables, two classi�ca- consumption of starchy food staples.12 The idea is tions emerge. The �rst (table 2.2) identi�es coun- that higher dietary concentration leaves consumers tries that appear to be especially vulnerable to more exposed to price shocks in dominant com- global price shocks. The second (table 2.3) identi�es modities, regardless of the source of the shocks. The second-level division (see the �rst row) is based on the country’s trade status with respect to the com- Figure 2.5 Cuddy–Della Valle Index of Production modity, as measured by the consistency of imports Variability for Staples, 1995–2004 (or exports). In this case, the median value is 10, so countries are divided into those that have imported 50 (or exported) in each of the last 10 years and those CV of production (1995–2004) Mor that have not. Many of the latter have been im- 40 Zim porters in most years, so the share of consumption Sud provided by imports is reported in parentheses 30 Zam within the body of the table. Countries are further Sud classi�ed into coastal and landlocked as an addi- 20 CIv tional measure of potential tradability and trans- Mal Sen Uga portation costs. 10 Table 2.2 groups countries according to their ex- Ind Egt Mex Nia Nia posure to global markets and, hence, world price 0 Rice Wheat Maize Millet/ Cassava/ shocks. A further division in this table relates to Sorghum Plantain countries’ capacity to import, measured by food sta- Note: Country abbreviations are listed in table 2.1. ple imports as a share of foreign exchange reserves. Source: Authors. Country-commodity combinations in the top left corner of the table are therefore the most exposed to Toward a Typology for Food Policy Analysis 11 Table 2.2 Typology of Countries According to Exposure to Global Price Shocks Exposure to Global Markets (net importer in all 10 years) Ratio of Diversity of Cereal Imports to Yes No Food Staple Foreign Consumption Reserves Coastal Landlocked Coastal Landlocked Higher import Yemen–Wh (93.0) Niger–M/S Kenya–Mz (9.5) Zambia–Mz (10.5) to reserves ratio Bangladesh–Ri (2.5) (0.47) Malawi–Mz(6.9) (≥ median 19%) Madagascar–Ri (4.9) Zimbabwe–Mz(–8.7) High staple concentration Lower import to Vietnam–Ri (–12.5) Pakistan–Wh (6.5) (HHI >= 0.43) reserves ratio Morocco*–Wh (41.8) (< median). Cambodia–Ri (1.5) Chile*–Wh (24.5) Mexico*–Mz (20.1) Indonesia*–Ri (3.7) Higher import Senegal–Ri (75.8) Sudan–M/S (–1.8) to reserves ratio Côte d’Ivoire–Ri(32.0) Ethiopia–Wh Ghana–Mz (–0.1) Ethiopia*–Mz (0.7) (≥ median 19%) Cameroon–Mz (0.8) (39.2) Cameroon–Ca(–0.0) Burkina Faso–M/S Ghana–Ca (–0.2) (–0.1) Low staple concentration Lower import to India–Ri (–2.4) Nigeria–S/M (–0.1) (HHI >= 0.43) reserves ratio Mozambique–Mz Nepal–Ri India–Wh (–1.8) Uganda–Ca (0.0) (= 0.43) Low variability Bangladesh–Ri (2.5) Pakistan–Wh (6.5) in staple Cambodia–Ri (1.5) Tanzania–Mz (1.8) production Indonesia*–Ri (3.7) Ghana–Mz (–0.1) (< 8.9%) Madagascar–Ri (4.0) South Africa*–Mz Mexico*–Mz (20.1) (–10. 8) Vietnam–Ri (–12.5) Sudan–M/S (–1.8) High variability Mozambique–Mz Nigeria–M/S (–0.1) Mali–M/S (0.3) in staple (16.4) India–Wh (–1.8) Burkina Faso–M/S (–0.1) production Senegal–Ri (75.8) Cameroon–Ca(–0.0) Uganda–Ca (0.0) (≥ median 8.9%) Egypt*–Mz (38.7) Mozambique–Ca(0.0) Ethiopia–Mz (0.7) Low staple Côte d’Ivoire–Ri(32.0) concentration (HHI < 0.43) Low variability Cameroon–Mz (0.8) Nepal–Ri Uganda–Pl (0.0) in staple Egypt*–Wh (47.4) (1.3) production India–Ri (–2.4) Ethiopia–Wh (< 8.9%) Ghana–Ca (–0.2) (39.2) Source: Authors. Note: CLVI is a measure of production variability (see section 2.1.4). HHI is the Her�ndahl-Hirschman Index = ∑(Si )2,where Si is the share of calories from starchy food staple i. Staple foods are Ca = cassava, Mz = maize, M/S = millet/sorghum, Pl = plantain, Ri = rice, and Wh = wheat. Numbers in parentheses indicate the percentage of utilization imported for dominant staple, 1995–2004. Countries in italics have food aid greater than 50 percent of cereal imports, 1999–2004. Countries marked with an asterisk (*) are classi�ed as middle income, based on the World Bank Atlas method (see footnote 9). Shaded countries are most exposed to global shocks. domestic price shocks; the subgroupings reflect the ability in production and low participation in trade, extent of domestic production variability. The right although the staple is usually less dominant. side of the table displays the country-commodity Although there are many other ways to group combinations that are most exposed to domestic countries, some instructive patterns emerge from shocks because these countries participate less in the classi�cation used here. In summary, the rice- trade. The country-commodity combinations in the and wheat-consuming countries of Asia have the top right corner are most vulnerable, notably the most stable production and depend on global mar- African countries that depend mostly on maize and kets for only a small share of consumption. They are have highly variable production. All of these coun- exposed to global shocks but should have the capac- tries import in most years, however, and trade is an ity to manage them, because only a small change in increasingly viable option. Another grouping in- domestic production or consumption is needed to cludes millet/sorghum in Niger, Mali, and Burkina clear markets in response to a price shock. On the Faso and maize in several other African countries other hand, African producers and consumers of where production variability is high. The countries maize and millet/sorghum are relatively exposed to in this group are largely self-suf�cient, however, and domestic production shocks caused by high pro- prices could swing widely between export and im- duction variability. In many cases, limited partici- port parity. This group is characterized by high vari- pation in trade will magnify the impacts of these Toward a Typology for Food Policy Analysis 13 shocks, although some countries will not suffer as the urban population, because rural households are much because local diets are more diversi�ed. Still equated with farm households and farmers are other countries, notably the southern African coun- equated with the production and sale of food. In tries and Kenya, depend heavily on maize but par- fact, rural households participate in grain markets ticipate extensively in trade. Their capacity to import in widely varying ways, and the overwhelming is quite limited, however, owing to constraints on evidence is that now the majority of the rural pop- export earnings and foreign exchange, and food aid ulation, and especially poor households, are net is used to �ll the gap. These countries are especially purchasers of grain. Small-scale farm households vulnerable to shocks that affect the region as a whole, generally fall into one of four categories with re- because they have to pay higher import prices when spect to their participation in major staple grain severe drought leads to regional shortfalls. Finally, it markets (tables 2.4 and 2.5). is important to note that these classi�cations can Households that sell staple grains account for change quite rapidly, as seen over the past 30 years roughly 20–35 percent (40 percent in Vietnam) of in Asia, where constraints on developing ef�cient the rural farm population. In Eastern and southern food markets and food imports have been dramati- Africa, two subgroups fall within this category: cally reduced (box 2.3). • A very small group (about 2–4 percent of the total rural farm population) of relatively A MICROTYPOLOGY wealthy smallholder farmers with 5–15 hect- OF HOUSEHOLDS ares of land, who sell between 5 and 50 tons of Important differences and trends at the household grain per farm annually and account for half of level affect the nature and costs of price instability marketed output at the microlevel. Household surveys provide im- • A larger group of farm households (20–30 per- portant data on emerging “empirical regularities� cent of the total rural farm population) selling surrounding food production, consumption, and much smaller quantities of grain (between 0.5 trade in low-income countries. and 5 tons per farm), who tend to be slightly better off than households that buy grain (see Rural Household Participation in Markets below). A widespread misconception is that high grain Households that buy staple grains generally make prices bene�t the rural population at the expense of up 60–70 percent of the rural population (their Box 2.3 Major Changes in Asia’s Food Policy Environment Over the past 30 years, food markets in Asia have be- ments in infrastructure, and adoption of improved come more integrated and ef�cient, reflecting dra- crop production technology in major Asian countries. matic changes on several fronts. The table below Production variability has also declined signi�cantly provides evidence of changes in foreign exchange in response to these changes [Naylor, Falcon, and earnings and the ability to import cereals, improve- Zavaleta (1977); Singh and Byerlee (1990)]. Percentage of Percentage of Cereal Imports Percentage Rice Area Wheat Area Import Capacity as Percentage Paved Roads of Arable Planted to Planted to Year Index of Foreign Reserves (000 km) Area Irrigated Modern Varieties Modern Varieties 1970 25 56 78 27 32 27 2000 143 5 293 45 79 95 Note: Countries include Bangladesh, India, Indonesia, Pakistan, the Philippines, and Vietnam. The import capacity index measures the ability of current foreign exchange reserves to import a given quantity of cereals. Source: Rashid, Cummings, and Gulati (2005). 14 Managing Food Price Risks and Instability in an Environment of Market Liberalization cent of the rural population. This group comprises Table 2.4 Distribution of Farm Households According relatively wealthy households that sell grain and to Participation in the Staple Grain Market, Africa (% of total) buy back lesser amounts of processed meal, as well as relatively poor households that make distress Household Category Ethiopia sales of grain after harvest only to buy grain later in with Respect Zambia Mozambique Kenya (maize the season (typically less than 5 percent of the total). to Main Staple Grain (maize) (maize) (maize) and teff) Households that neither buy nor sell staple grains make up a small proportion of the rural popula- Sellers only 19 13 11 13 tion in countries where one staple crop dominates. Buyers only 33 51 58 60 Buy and sell However, in countries where cassava is the main sta- (net buyers) 3 7 13 ple in speci�c and usually very remote regions, such 12a as northern Zambia and parts of Mozambique, a siz- Buy and sell (net sellers) 5 16 12 able fraction of the rural population is autarkic with Neither buy nor respect to the primary national food crop (maize). sell 39 24 8 2 According to their status as sellers and pur- All 100 100 100 100 chasers of food grains, different types of households Top 50% of total will be impacted quite differently by price instabil- sales 3 3 2 2 ity and may be quite sensitive to whether it is sea- Bottom 50% of sonal or interannual (table 2.6). total sales 21 10 9 11 Several implications can be drawn from this ty- a The Mozambique data do not allow net buyers and net sellers to be pology of rural households: distinguish among households who both buy and sell 1. Staple grain sales are highly concentrated Source: Jayne, Tembo, and Nijhoff 2005, based on data from Zambia: Central Statistical Of�ce Supplement to the Post Harvest Survey, among a relatively small proportion of the 2000/01 marketing year; Kenya: Tegemeo Institute Household Survey, rural farm population. These households will Egerton University, 1999/2000 season; Mozambique: Trabalho do be hurt most when prices of food staples col- Inquerto Agricola (TIA), Ministry of Agriculture and Rural lapse, but they are also in a much better posi- Development (MADER), and National Institute of Statistics (INE), 2001/02 season; and Ethiopia: Central Statistical Authority, Govern- tion to weather such shocks, because their ment of Ethiopia, Food Security Survey, 1995/96 season. income and asset levels are much higher than the national average.13 Also, because produc- tion and sales tend to be relatively high in years when prices are low (and vice versa), the insta- number is higher in drought years and lower in bility in grain sellers’ revenue over time is good years). They are generally poorer and have likely to be lower than instability in prices. smaller farm sizes and asset holdings than the me- 2. The majority and the poorest segment of farm dian rural household. households are grain purchasers and are most Households that both buy and sell staple grains vulnerable to rising prices (which also reduce within the same year generally make up 5–15 per- the cash available for purchasing farm inputs). Because these households rarely sell the main staple crop, even in a good year, they do not Table 2.5 Distribution of Rural Households in Indonesia, Vietnam, and Mexico According to Participation face output price risk, and downward swings in the Staple Grain Market (% of total) in grain prices have little effect on production incentives. Indonesia Vietnam Mexico 3. While an inverse relationship between prices (rice) (rice) (maize) and quantities marketed may dampen the ef- fects of price shocks on farm incomes, this is Producers who are net sellers 29 43 25 not true for farm households that are (a) chronic Producers who are net buyers 10 41 45 food purchasers in normal and poor rainfall Nonproducers 61 16 30 years or (b) self-suf�cient in normal years but Total 100 100 100 transitory food buyers in bad years. For both Source: Indonesia: Dawe and Timmer 2005; Vietnam: Ryan 1999; groups of households, higher grain prices exac- and Mexico: Avalos-Sartorio (personal communication). erbate the weather shock, because grain must be purchased for consumption. Toward a Typology for Food Policy Analysis 15 Table 2.6 Who Is Affected by Food Price Instability, and How? Relative Importance of Inter-annual Type of Household Price Instability Problem or Intra-annual Price Variability 1. Poor consumers High prices reduce real incomes, especially Peaks in both in years of low harvest 2. Net de�cit producers As in 1, but high prices also discourage Peaks in both investment in high-value crops 3. Net de�cit sellers As in 2, but low prices immediately after Intra-annual troughs harvest reduce real income 4. Surplus producers Price collapse at bumper harvest reduces real Both, particularly intra-annual troughs incomes and depresses incentives for investment in intensi�cation Source: Poulton et al. 2005 4. Attempts by food-purchasing farm house- cash revenues from agriculture or consumer expen- holds to diversify into higher-valued crops de- ditures (table 2.7). Gradually shrinking landhold- pend on their con�dence in being able to ings over the past decades have caused farmers to procure food at tolerable prices. Price instabil- shift toward crops that provide greater caloric value ity can act as a disincentive to diversify to crop- per unit of land. The elimination of pan-territorial ping patterns that raise farmers’ incomes but pricing on maize in some countries has also diversi- increase their risks in food markets (Fafchamps �ed cultivation patterns. In Zambia, the value of cas- 1992). sava and sweet potato production is now roughly 5. Food policies that alter mean price levels over 75–85 percent of the value of maize production time (for example, relative to border prices) can (Govereh, Jayne, and Shaffer, forthcoming). have unanticipated income distributional ef- The rapid rise in wheat and rice consumption in fects that run counter to explicit poverty re- urban areas of Africa has moderated the effects of duction goals. To the extent that the poor are variability in coarse grain prices in many of those net purchasers of staples, they are directly hurt areas. Data from recent surveys in urban Kenya in- by policies that raise the prices of these com- dicate that expenditures on wheat now exceed those modities. The bene�ts of food policies that on white maize, the traditional staple. The share of raise food prices are captured predominantly expenditure on maize is less than 10 percent in by the small minority of wealthier households urban areas (Traub and Jayne 2004; Muyanga and that sell most of the staple grains. This situation others 2005). More diverse diets, particularly in is evident for maize in Kenya and rice in urban areas, make it easier for households to stabi- Indonesia and Madagascar, where substantial lize their food expenditures through substitution. import tariffs are in place. Low-income groups remain susceptible to sharp in- creases in cereal prices, however, because staple food still accounts for 30–40 percent of consumer ex- A Reduced Role for Food Staples penditures (tables 2.7 and 2.8). Household data also reveal important trends over time. In particular, the dominance of food staples as MAIN MESSAGES FOR THE DESIGN a share of producer incomes and consumer expen- ditures is shrinking, so extreme movements in the OF FOOD SECURITY POLICIES prices of the major food staple now have much All food security policies—but especially policies di- smaller effects. For example, in eastern and southern rected at managing food price risks and instability— Africa, the dominant staple—white maize—now must be designed in accordance with the macro- generally accounts for 10 percent or less of producer and microlevel conditions prevailing in a particular 16 Managing Food Price Risks and Instability in an Environment of Market Liberalization Table 2.7 The Role of White Maize in Farm Incomes and Consumer Expenditures, Eastern and Southern Africa Cash Revenue from Maize as Expenditure on Maize Products Value of Maize Output as Percentage of Cash Revenue as Percentage of Urban Country Percentage of Total Farm Output from Agriculture Household Income Kenya 19 10 1 to 10 South Africaa na na 1 to 7 Mozambique 27b 6 Na Zambia 45 b 16 15 Note: na = information not available. a Study sample pertains to urban and peri-urban region outside Umtata, Eastern Cape Province, 2004. b Does not include fruits, vegetables, or livestock products in total farm output. Source: Kenya: Muyanga et al. 2005; Tegemeo Institute Rural Household Survey, 1999/00 crop season; South Africa: Traub and Jayne 2004; Mozambique: Ministry of Agriculture and Rural Development (MADER), in collaboration with the National Institute of Statistics (INE), Trabalho do Inquerito Agricola (TIA), 2002; Zambia: Post Harvest Survey, 1999/00, and Supple- mental Survey to the PHS, 2000; Ministry of Agriculture and Central Statistical Of�ce, Government of Zambia; and del Ninno, Dorosh, and Subbarao 2005. country. The typology of countries constructed in that the majority of poor farmers and other rural this section reveals very large differences in poten- households depend increasingly on the market to tial exposure to global price shocks and domestic meet their staple food needs in most years. production shocks. In general, the Asian economies Finally, food policy must take account of rapidly face the least risk, although their dependence on one changing production and consumption patterns. commodity (rice) may exacerbate risks to some ex- Many parameters of food security policy that tent. The potential for risks from domestic shocks reigned in the 1970s and 1980s, especially food is far higher in the maize- and millet/sorghum- self-suf�ciency and price stabilization, are outdated. dominated economies of Africa, where the capac- Indeed, in terms of household welfare, producer ity to manage these risks is constrained by levels of revenue from the sale of a major staple food is often foreign reserves and export earnings. less than 20 percent in many parts of Asia and At the microlevel, food policy must distinguish Africa. Similarly, consumers’ expenditures on a between myth and reality. The notion of smallholder major staple food now occupy a surprisingly low subsistence farmers who produce a surplus for the share of total household expenditures in many low- market in good years is largely a myth. The reality is income countries. Table 2.8. Household Food Budget Shares (%) in Rural and Urban Areas Zambia, 1998 Ethiopia, 2000 Bangladesh, 2000 Maize All Cereals Food Share Maize All Cereals Food Share Rice All Grains Rural (bottom 30%) 27 44 77 13 43 85 53a 55a Rural (average) 21 33 74 10 34 71 42 45 Urban (bottom 30%) 28 34 69 4 33 65 43a 46a Urban (total) 15 22 59 2 20 46 29 33 a Bottom 20%. Source: del Ninno, Dorosh, and Subbarao 2005 3 The Nature and Extent of Food Price Instability and Risk How great is the problem of food price instability in low-income coun- tries, and why does food price instability vary across countries? Although government policy has a critical role in reducing or adding to food price instability, particularly as it relates to whether market lib- eralization policies have improved or worsened the variability in food prices, detailed discussion of this issue is deferred to a later discussion. This chapter examines the magnitude of two major exogenous sources of food price instability—world price shocks and domestic supply shocks (primarily induced by climate or natural disasters)—and how they are transmitted to domestic food prices. It also briefly examines whether price shocks are likely to become more severe or frequent in the future. VARIABILITY IN WORLD GRAIN PRICES As noted in chapter 2, the three major grains traded internationally for food use are rice, wheat, and white maize. Prices of these grains tend to move quite closely together, indicating that to some extent they are substitutes in world markets (although this is somewhat less true for white maize). In general (a) world prices of all the major grains have been declining in real terms, (b) there is no statistical evidence that these prices are becoming more variable over time, and (c) in the case of rice, world prices appear to have become more stable over time (table 3.1).14 There is no evidence of any recent increase in world price variabil- ity for grains, but the absolute levels of variability are high. The coef- �cients of variation (CVs) are 33 percent for rice, 29 percent for wheat, and 23 percent for yellow maize (probably higher for white maize).15 If fully transmitted to domestic markets in low-income countries, such high levels of price variability could pose problems for farmers and poor consumers. Another important dimension of food price variability is the per- sistence of world price shocks. If shocks dissipate quickly, the effects of price instability are less pronounced, because over time producers and consumers come to understand the temporary nature of the in- stability and adjust their behavior accordingly. But if shocks are per- sistent and move prices away from their long-run trend levels for an extended period, then instability has more lasting effects that require greater (and more painful) adjustments. Sarris (1998, 2000) �nds rela- tively low persistence in world cereal prices, and these results are sup- 17 18 Managing Food Price Risks and Instability in an Environment of Market Liberalization fact, in a few instances, such as wheat in India, vari- Table 3.1 Variability in World Prices of Major Grains, ability in producer prices has declined signi�cantly. 1971-2003 Cereal price variability does appear to vary con- siderably across countries, however. In the case of Yellow Rice Wheat Maize wheat, five of seven importers have CVs that are substantially lower than the CV of world prices Coef�cient of variation (CV) (29 percent). In India, where prices have been the 1971–2003 (%) 33 29 23 most stable, wheat marketing policies insulate do- Regression t-statistic for: mestic prices from world prices (�gure 3.1). The two Price trend –7.5 –7.2 –9.7 5-year moving standard deviation –7.5 –4.6 –4.5 exceptions occur in Bolivia and Mexico, where CVs 5-year moving CV –4.5 0.0 –0.2 surpass 75 percent. Much of this instability proba- bly results from macroeconomic shocks, especially Note: All regressions are linear trends over the period 1971–2003. sharp devaluations of exchange rates and high do- The CV is the standard deviation of detrended price divided by mean of actual (not detrended) price. mestic inflation, rather than from the fundamental Source: Hazell, Shields, and Shields 2005. characteristics of Latin American grain markets or grain marketing policies. In the case of maize, the majority of importing countries have a CV that is at or below the CV of world prices for yellow maize (23 percent), with the major exceptions again occurring in the Latin ported by Leon and Soto (1997). It appears that price American countries of Bolivia, Brazil, and Mexico, shocks for rice do persist for lengthy periods of 9–18 where CVs surpass 65 percent (�gure 3.2). years, however, implying that at least some major Among African maize producers that are more grain prices can cycle away from their long-run intermittent traders, about half have CVs above trend for extended periods (Cashin, Liang, and 20 percent, whereas the other half fall below 20 per- McDermott 1999). cent (�gure 3.2). Maize producer price variability in most African countries remains comparable to vari- PRICE VARIABILITY IN DOMESTIC ability in world maize prices.17 The trend in price variability is signi�cant in only a few countries, no- FOOD MARKETS tably Tanzania and Namibia, and always negative Domestic food prices can differ substantially from (see Hazell, Shields, and Shields 2005). Overall this world prices because of transport and transaction sample of African countries offers little evidence that costs and because of the insulating effect of trade domestic price variability is high relative to world and domestic grain marketing policies. Relatively prices or that instability is increasing over time. little information is available on price variability in Some of these African countries nevertheless have domestic markets of major food staples at the indi- higher domestic producer price variability than vidual country level. An exception is Hazell, Shields, others, and data for particular countries, regions, and Shields (2005), who analyze producer prices for and commodities reveal serious price instability in wheat and maize in developing countries that con- Africa. From 1996 to 2003, when world maize prices sistently import at least one of these major staples.16 were relatively stable, the wholesale price of maize in These countries are expected to experience instabil- Addis Ababa varied from just about US$50 per ton to ity in domestic food prices resulting from world nearly US$250 per ton (�gure 3.3). Maize is effec- price shocks, especially if agricultural markets have tively a nontradable commodity in Ethiopia, and �g- been liberalized and trade restrictions are minimal. ure 3.3 shows how domestic maize prices have Hazell, Shields, and Shields (2005) also analyze vari- cycled between export and import parity over this ability in producer prices for maize in 12 African period. The high cost of transporting grain in and out countries that have weak infrastructure and more in- of Ethiopia creates a wedge of about US$150 per ton termittent patterns of trade. between import and export parity. Like world prices, producer prices in most coun- There is also little information on grain price in- tries studied by Hazell, Shields, and Shields (2005) stability faced by consumers in low-income coun- have generally trended down over time, and the tries. Data from Malawi, Mozambique, and Zambia study found no evidence of increased variability. In indicate that consumers in southern Africa have ex- The Nature and Extent of Food Price Instability and Risk 19 perienced highly variable retail prices of white Figure 3.1 Coefficient of Variation of Wheat Producer maize in the past decade (�gure 3.4). This variation Prices, 1971–2002 tends to be higher in inland and more remote areas, again partly because of the large wedge between import and export parity prices. At inland loca- Bolivia tions in southern Africa, coef�cients of variation of Mexico 50 percent or above are common for consumer Egypt prices of maize (table 3.2). Mozambique, with the Morocco most liberalized markets, has the lowest price vari- Syria ability in the capital city of Maputo, but variability India was much higher in remote areas of the country. Kenya Overall, these results for Africa indicate con- siderable cross-country variation in the magni- 0 10 20 30 40 50 60 70 80 90 tude of domestic producer and consumer price Percent instability, with more variable prices generally oc- Source: Hazell, Shields, and Shields 2005. curring in countries that trade intermittently and have a large wedge between import and export parity (that is, in countries where transportation and marketing costs are high; see the country Figure 3.2 Coefficient of Variation of Maize Producer typology in chapter 2). Prices, 1971–2002 SOURCES OF DOMESTIC Bolivia Brazil PRICE INSTABILITY Honduras To what extent is variability in domestic prices Mexico caused by world price shocks versus shocks to do- Morocco mestic production? For many reasons, changes in Syria world prices are transmitted imperfectly to domes- tic markets. Some of these reasons are related to African maize producers Botswana transaction costs in markets and others to policy in- Ethiopia terventions (box 3.1). Hazell, Shields, and Shields Kenya (2005) use a variance decomposition approach to ex- Malawi amine how domestic prices of wheat- and maize- Mali importing countries are influenced by world prices Namibia and exchange rates. They �nd that these variables Niger explain a very small share of domestic price vari- Nigeria ability in their sample of countries, which suggests Rwanda that most variation in domestic prices arises from Tanzania domestic factors, such as production shocks. Uganda For the 12 African countries where infrastructure Zambia is generally poorly developed, Hazell, Shields, and Shields (2005) use a similar decomposition approach 0 20 40 60 80 100 to estimate the contribution of both world prices and Percent domestic production variability to domestic pro- Source: Hazell, Shields, and Shields 2005. ducer price variability. Surprisingly, variability in world prices accounted for at least 25 percent of do- mestic maize price fluctuations in only three coun- tries—Malawi (post 1991 only), Mozambique, and Zambia (table 3.3). Instability in domestic maize pro- duction accounted for more than 25 percent of pro- ducer price variability in �ve countries—Botswana, 20 Managing Food Price Risks and Instability in an Environment of Market Liberalization Ethiopia, Mali, Niger, and Nigeria. The approach Figure 3.3 Wholesale Maize Prices (US$/ton) in Addis used to undertake the decomposition is crude and Ababa Relative to Import and Export Parity ignores lag effects. Nevertheless, the results provide Prices, Based on U.S. White Maize Traded support to the argument that domestic production via Djibouti, 1998–2003 instability is an important source of price variability in countries that are relatively self-suf�cient and 400 where there is a signi�cant wedge between import 350 and export prices. 300 WILL PRICES BECOME Import parity (Addis) 250 MORE UNSTABLE? US$ / ton Some observers believe that the long, steady de- 200 cline in world prices is ending and that a period of 150 more unstable prices is looming. The two factors Addis Ababa wholesale cited most often to support this view are the decline 100 in global grain stocks and the prospects that global climate change will increase the uncertainty in agri- 50 Export parity cultural production. Because global grain stocks can buffer fluctua- 0 tions in supply and demand, they can help stabilize 1996 1997 1998 1999 2000 2001 2002 2003 prices. Mitchell and Le Vallee (2005) note important Source: del Ninno, Dorosh, and Subbarao (2005). changes in the stockholding policies of the three largest grain-producing countries and regions: China, the European Union, and the United States. Together these producers account for 69 percent of Figure 3.4 Wholesale Maize Prices (US$/kg) in world grain stocks. Over the past �ve years, global Southern Africa, 1994–2003 grain stocks have declined by almost 50 percent, with stocks in China declining the most (�gure 3.5). 0.6 China—now a consistent net importer of rice—is Lusaka, Zambia forecast to become the world’s largest importer of 0.5 Maputo, Mozambique wheat (Mitchell and Le Vallee 2005). The European Lilongwe, Malawi Randfontein, South Africa Union began to reform its grain policy starting in Nominal US$ per kg 0.4 1992 and reduced its stocks. The United States changed its policy even earlier, in 1985, and also 0.3 subsequently reduced its grain stocks. The com- bined impact of policy changes in these countries 0.2 has been to reduce their grain stocks from 31 per- cent of total use in 1999 to 18 percent in 2003—the 0.1 lowest level since the mid-1970s. As stocks decline, prices may become more vul- 0.0 nerable to sharp upward swings if there are cli- 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 matic shocks or rapid shifts in global demand. Source: Mozambique and Zambia: wholesale maize prices from Ministry Furthermore, if China and other Asian countries of Agriculture; Malawi: retail maize prices from Ministry of Agriculture; South Africa: wholesale white maize price data from SAFEX, Randfontein. maintain their current rate of economic growth, food exports and imports from Asia are expected to be- come considerably larger and more volatile, leading to higher and more unstable world food prices. If this speculation is correct, food-de�cit countries may have to deal with both increased food price variability and possibly higher prices in the future. The Nature and Extent of Food Price Instability and Risk 21 On the domestic side, there is concern that production variability may increase, particularly Table 3.2 Index of Variability in Monthly White Maize Prices in Southern Africa, because of global climate change. Global climate January 1994 to August 2003 patterns have long-run cycles that are inherently noisy, and it is dif�cult to draw de�nitive relation- Coef�cient Coef�cient ships between climate patterns and crop yields. Country/Location of Variation Country/Location of Variation There is growing evidence, however, that extreme climatic events are becoming more numerous and Mozambique: South Africa: are leading to wider fluctuations in crop yields in Beira 45 Randfontein 34 some particularly vulnerable areas of the world Maputo 26 (for example, see Hulme 1996). If this is indeed the Nampula 58 case, food production variability will increase in Malawi: Zambia: low-income countries that experience more fre- Karonga 159 Chipata 55 quent extreme climatic events, such as the counties Lilongwe 71 Choma 58 of southern Africa (box 3.2). Nkhata Bay 62 Lusaka 59 Rumphi 44 Ndola 42 MAIN MESSAGES ON THE Note: SAFEX prices are for April 1996 to August 2003; Randfontein for March 1996 to January 2003. NATURE AND EXTENT OF Source: Mozambique and Zambia: wholesale maize price data from FOOD PRICE VARIABILITY Ministry of Agriculture; Malawi: retail maize price data from Ministry of Agriculture; South Africa: white maize wholesale from SAFEX, and Despite the perception that the liberalization of grain white maize selling price from Randfontein Maize Board. markets has made food prices less stable, there is no evidence that price variability has been increasing over time, either in global or domestic markets. But it is unrealistic to assume that policymakers will es- gions and if global climate change further destabi- cape the recurrent challenge of dealing with food lizes food production. price instability. Instability appears to have dimin- Most domestic markets, even in middle-income ished little in many countries where food prices are countries that consistently import grain, appear to quite variable, and in fact it may increase if grain be integrated only weakly into global markets. stockpiling policies change in major producing re- Various transaction costs along supply chains im- Box 3.1 The Imperfect Transmission of World Prices to Domestic Markets Many studies have found that movements in world ports themselves can have an important effect on IUVs prices are transmitted very imperfectly into local mar- expressed in local currency. For example, a sudden in- kets. The transformation from world price to producer crease in food imports could lead to a worsening bal- price begins with the average annual import price paid ance of trade, causing the currency to devalue and by a country, the import unit value (IUV). The IUV need making imports more expensive in local currency. not closely follow the average annual world price of a The mapping of IUV in local currency to the average commodity because of differences in quality and the producer price can be affected by government interven- country’s seasonal distribution of imports, the use of tion in the form of import taxes and attempts at price forward price contracts, and the particular mix of world stabilization. Marketing institutions also matter, espe- market locations used. Food aid shipments of some sta- cially the size and temporal behavior of marketing and ples at concessionary rates may also affect the average processing margins retained by all the relevant market IUVs paid for those staples. In some low-income coun- intermediaries, and private sector decisions about stor- tries with limited foreign exchange earnings, food im- age, international trade, and regional shipments. Source: Hazell, Shields, and Shields (2005). 22 Managing Food Price Risks and Instability in an Environment of Market Liberalization Table 3.3 Disaggregation of Variance Components in Producer Prices for Maize, Selected African Countries (%) Variance of Variance of Covariance Country Period of Analysis Border Prices Domestic Production and Unexplained Total Botswana 1974–1995 1 49 50 100 Ethiopia 1971–2002 1 22 77 100 (2) (34) (64) Malawi 1980–2002 8 1 9 100 (43) (4) (53) Mali 1988–2002 2 28 70 100 (1) (32) (67) Mozambique 1985–2002 25 2 73 100 (2) (1) (97) Namibia 1980–2002 3 11 86 100 (1) (1) (98) Niger 1971–1995 5 57 38 100 Nigeria 1971–2002 12 26 62 100 (7) (16) (77) Rwanda 1971–2002 5 1 6 100 (10) (4) (86) Tanzania 1971–1995 19 15 66 100 Uganda 1980–1995 1 10 89 100 Zambia 1985–1995 37 5 58 100 Note: Figures in parentheses are for post-reform periods beginning 1991. Source: Hazell, Shields, and Shields 2005. Figure 3.5 Trends in World Grain Stocks, 1989–2004 Box 3.2 Global Climate Change and Unstable Food Production 600 500 Global climate change may affect variability in food production by changing the inci- Million metric tons Total 400 dence of pest losses, altering the length of the growing cycle, and increasing the fre- 300 quency and severity of weather events such Non-China as droughts and cyclones. Studies by FAO 200 suggest that rain-fed agriculture in the devel- 100 oping world is especially subject to greater stresses from climate change (FAO 2005). 0 Persistent and more frequent severe weather 98/99 99/00 00/01 01/02 02/03 03/04 04/05 changes (such as drought) are likely to make Source: Mitchell and Le Vallee (2005). food supplies more unstable in some re- gions, leading to greater price instability on a local and global scale. Climate models suggest that recently observed variability in rainfall in southern Africa is related closely to long-term warming of the Indian Ocean, and the incidence of drought in southern Africa is projected to increase. Source: Revkin (2005). The Nature and Extent of Food Price Instability and Risk 23 pede the direct transmission of world price changes high transportation costs often drive a large wedge into local markets, and most developing country between export and import parity prices. These cir- governments still intervene in food markets to pro- cumstances explain at least part of the higher vari- tect producers and consumers from world price ability in domestic food prices observed in Africa, swings. but untimely and unpredictable policy interventions Finally, production variability for food grains have also contributed to this variability, as will be such as maize is higher in African countries, where seen in chapter 5. 4 The Costs of Food Price Instability and Risk There is general agreement that highly volatile food prices can impose signi�cant costs on a society, particularly in low-income countries that import food and where many poor households spend a large propor- tion of their income buying food staples. But the naïve notion that volatility in food prices is somehow “bad� and stability “good� masks the complex and often poorly understood issues surrounding the costs of food price instability. The potential costs of food price instability and risk can be broken down into three major categories: (1) economic inefficiency costs; (2) negative effects on income distribution and household food secu- rity; and (3) macroeconomic externalities. ECONOMIC INEFFICIENCY COSTS In market-based economies, price changes are important because they help redirect resources to more ef�cient uses and encourage ef�cient changes in consumption bundles following economic shocks. Yet there is a concern that “too much� price instability is costly. Food price instability certainly can impose costs on food market partici- pants, but considerable costs can also be incurred by not adjusting to the underlying shocks that cause prices to change. So how much price instability is “too much,� and what are the resulting costs? To answer this question, it is important to understand why “exces- sive� food price instability in low-income countries causes the misallo- cation of resources and results in economic inef�ciency. Several explanations have been put forward (see, for example, Dawe 2001; Timmer 2002). Briefly, food price risk is thought to lead producers, con- sumers, and traders to engage in such risk-reducing strategies as di- versi�cation into lower value but more stable products, minimal use of purchased inputs, and lack of trade in remote locations. Price risk may also reduce investment and make farmers and traders reluctant to use new technologies. While these are rational responses to risk, resources might be used more effectively if risks could be pooled or distributed more ef�ciently. Each of these responses works to distort resource al- location, investment, and consumption patterns away from their most ef�cient levels, with a resulting decline in the productive potential of the economy. The economic foundation for the link between price instability and economic inefficiency lies with the theory of incomplete mar- kets (Newberry and Stiglitz 1981). Without a complete set of credit, 25 26 Managing Food Price Risks and Instability in an Environment of Market Liberalization insurance, and forward markets, risks cannot be and what about countries that typically are either pooled and shared efficiently, which distorts pro- nearly self-suf�cient in a staple or consistently im- duction and consumption decisions. There is little port it? In these cases, domestic food grain prices doubt that most poor households in low-income may vary more than world prices (because of countries have only very limited and informal op- switching between import and export parity), and portunities to obtain credit, insure risks, and sell food expenditures may take up a much higher pro- commodities forward (Morduch 1995; Townsend portion of total consumer expenditures. Gabre- 1995). It has been argued that in such circum- Madhin, Barrett, and Dorosh (2003) argue that food stances the efficiency costs of food price instabil- price instability can be much more costly for coun- ity are high, and price determination should not tries facing such circumstances. It is possible that be left completely to market forces (see, for exam- economic ef�ciency losses from food price instabil- ple, Poulton and others 2005). ity are higher than the 0–2 percent range typically The key empirical issue surrounding food price found for export crops, provided countries have instability is the size of the efficiency losses in par- food consumption concentrated in a few staple ticular countries and situations. Newberry and crops, are landlocked or otherwise experience high Stiglitz (1981), Scandizzo, Hazell, and Anderson transportation and transaction costs, and spend a (1984), and others provided early estimates of the large part of their foreign exchange reserves and ef�ciency costs of price instability for a range of consumer incomes on food (see the country typol- countries and commodities, and their work has ogy in chapter 2). Despite the logic of these argu- been followed by many other quantitative estimates ments, surprisingly little empirical evidence is and simulation studies investigating the magnitude available to evaluate the size of these effects in par- of ef�ciency costs.18 Most of these studies have con- ticular countries and situations. cluded that the ef�ciency costs from agricultural commodity price instability are generally quite small when measured as a proportion of incomes (around EFFECTS ON INCOME 0–2 percent). DISTRIBUTION AND There are four main reasons for these generally HOUSEHOLD FOOD SECURITY small estimates of economic ef�ciency costs. First, The poor are the most vulnerable to food price in- prices and yields are typically negatively correlated, stability and risk. Compared to households with so price instability does not necessarily lead to large more assets, poor households have fewer options for fluctuations in producer incomes. Second, relative diversifying their productive activities and must welfare effects on consumers are small unless con- spend a larger proportion of their income on food. sumers spend a large proportion of their income on The human cost of adverse price shocks can be dev- the commodity in question and are highly averse to astating as poverty, malnutrition, and even famine risk. Third, the welfare effects on producers and con- deepen and become more prevalent. sumers generally move in opposite directions: If Most research on the distributional effects of sharp price changes have a negative effect on pro- commodity price instability and price stabilization ducer welfare, they will often have a positive effect schemes has focused on the effects on producers on consumer welfare. Fourth, most smallholder versus consumers, �nding that the relative burden farming households are producers as well as con- borne by producers versus consumers depends on sumers of food, and (as shown in chapter 2) more such factors as the strength of consumer and pro- often than not they are net buyers of food staples ducer risk aversion, the shape and elasticity of sup- rather than net sellers. Price instability will have lit- ply and demand curves, the nature and source of tle effect on the welfare of households that are not the instability, and the proportion of producer in- participating in the market in a major way. comes and consumer expenditures devoted to food An important point for this discussion, however, commodities (Just and others 1978). Unfortunately is that the vast majority of quantitative estimates of the distributional effects of food price instability on economic ef�ciency losses from commodity price in- consuming versus producing households cannot be stability have focused on export crops that typically generalized and need to be evaluated on a case-by- comprise a very small proportion of domestic con- case basis. sumer expenditure, such as cocoa, coffee, cotton, However, a couple of key points should be kept jute, rubber, and wool. But what about food crops, in mind. First, many farming households in low- The Costs of Food Price Instability and Risk 27 income countries both produce and consume food, MACROECONOMIC and if they are neither major sellers nor major buy- ers, then food price instability will be of little con- EXTERNALITIES sequence to them. Second, larger-scale farmers It has been argued that food price instability can who have a signi�cant marketable surplus will be impose negative externalities on the general econ- affected as producers, while smaller-scale farmers omy, particularly when a food staple is a wage who are net purchasers of food (the most common good or represents a large proportion of a country’s case—see chapter 2) will be affected more like con- gross domestic product (GDP) (Bidarkota and sumers (Finkelshtain and Chalfant 1997). Crucini 2000; Dawe 2001; Dawe and Timmer 2005). Second, the distributional effects discussed so Food price shocks can ripple through the economy far assume implicitly that all households are food and cause major macroeconomic instability, from secure—in other words, their incomes are suf�cient business cycle fluctuations to exchange rate move- to ensure adequate food intake for survival and ments, changes in the general price level, and po- reasonable health, even at relatively high food litical upheaval. Macroeconomic instability may in prices. Consumption choices then come down to turn retard economic growth (Choudhury 1995; trading off preferences for food against preferences Ramey and Ramey 1995; Deaton and Miller 1996). for other types of consumption goods. But for food- Various mechanisms have been suggested to ex- insecure households, whose survival and nutri- plain how food price instability reduces investment tional status are threatened under adverse price or levels and economic growth rates in low-income production outcomes, the costs of food price insta- countries (Dawe and Timmer 2005).19 bility may be much higher than indicated by tradi- How large are these negative growth effects? tional welfare analyses of economic ef�ciency and Timmer (2002) argues that rice price stabilization distribution. These higher costs stem from two added one-half to one full percentage point to GDP main sources: growth in the Indonesian economy in the 1970s. This is a huge number. Myers (2005) shows that a 1. A reduced probability of survival. Using a sustained growth reduction of this magnitude cor- value-of-life approach, Myers (2005) calculates responds to an estimated welfare effect of approxi- that if high food prices reduce the probability mately 5–11 percent of economy-wide income per of survival by 1 percent, the resulting welfare year. Clearly, growth effects of this magnitude will loss would be valued at approximately 5 per- dominate the traditional welfare estimates of the cent of household income. In very poor house- costs of commodity price instability, but the exis- holds where high food prices may signi�cantly tence and magnitude of these growth effects remain reduce the probability of survival, the value-of- controversial. Other researchers have found no sta- life losses dominate the traditional welfare tistically signi�cant link between commodity price costs estimated for food price instability. fluctuations and the rate of economic growth (see 2. A reduced capacity to work. Even if price MacBean 1966; Deaton 1992; and Kannapiran 2000). shocks do not reduce survival probabilities, A �nal caveat is that these macroeconomic “ex- they may decrease nutritional status to the ternalities� are not externalities in the strict welfare point where labor productivity is severely sense, and it is not clear that macroeconomic fluc- curtailed. Traditional welfare analyses of food tuations should be interpreted as indicators of price instability take no account of these po- “market failure.� To be sure, one role of successful tential labor productivity losses. governments is to provide a stable macroeconomic Given the paucity of detailed case studies estimat- environment within which investment is encour- ing the size of these effects in particular market sit- aged and economic growth can occur. But it is not uations, it is very dif�cult to provide a complete clear that food price stabilization schemes are the evaluation of these “food insecurity� costs of food most appropriate way to achieve these goals, even price instability. Even so, they would appear to be when food products represent a major share of potentially large in many low-income countries. For GDP. Macroeconomic stability might best be example, Ersado, Alderman, and Alwang (2003) achieved using traditional macroeconomic policies and others have shown that drought can perma- such as monetary policy, �scal policy, and trade nently reduce human nutrition levels and hence and exchange rate management, rather than food labor productivity. price stabilization schemes. 28 Managing Food Price Risks and Instability in an Environment of Market Liberalization MAIN MESSAGES ON THE COSTS big these additional welfare costs really are for dif- ferent countries in different situations. OF FOOD PRICE INSTABILITY The general message, therefore, is that it is likely The empirical evidence from traditional welfare that the overall costs of food price instability will analyses suggests that economic inef�ciency costs be quite high, at least for some low-income coun- from food price instability are likely to be quite tries in certain situations, but the evidence on just small, even in low-income countries subject to in- how high remains surprisingly sparse. Of course, complete and underdeveloped markets. This is es- even when the costs are high, price stabilization pecially true if the country is diversi�ed in its schemes may not necessarily improve welfare. production and consumption of food crops. Price stabilization schemes have costs of their own, Nevertheless the traditional welfare costs of food and they may not counteract all of the costs of price price instability may be considerably higher in other instability. situations: when countries are specialized in pro- Are there other, more effective ways to overcome duction or consumption of one or two major food the negative effects of food price instability? For ex- staples; when production variability is high and ample, if the major cost involves malnutrition or links with global markets are poor; and when large death for poor households when food prices soar, numbers of consuming households spend a large targeted subsidy and income support programs for proportion of their income on food. the poor may be more appropriate. If the major cost Add the possible effects of food price instability involves negative macroeconomic externalities, ap- on survival and nutrition in low-income house- propriate monetary and �scal policies may be more holds, and on macroeconomic growth rates, and the effective than food price stabilization schemes. The costs of food price instability may also climb signi�- remainder of this report discusses how to evaluate cantly. Unfortunately there is considerable contro- appropriate ways of dealing with food price insta- versy and very little empirical evidence on just how bility and risk, and their consequences. 5 Food Market Reforms: Experiences and Lessons Views on how public policy can best address problems of food price instability and risk in low-income countries have followed broader trends in development thinking about the role of the state in food markets. This chapter provides a brief overview of the evolution of policy interventions in food markets and the lessons that can be dis- tilled from these experiences for designing and implementing future policies. Three case studies of market reform are presented later in this chapter. STATE-LED FOOD MARKETING SYSTEMS: SOMETIMES EFFECTIVE, ALWAYS COSTLY Throughout the 1950s, 1960s, and 1970s, most governments engaged directly in food marketing through parastatal marketing boards or corporations that usually purchased, stored, and sold crops; con- trolled prices; and restricted private trade within and across borders. The prevailing view at the time was that private food markets were very underdeveloped and could not be relied on to bring about de- sired outcomes. In particular, the potential for highly unstable prices was viewed as signi�cant and costly. Price stability for staple foods was seen as necessary to foster the adoption of improved technology, promote macroeconomic stability, and support the incomes of the poorest households (Jones 1994; Meerman 1997; Cummings, Rashid, and Gulati 2005). In a number of cases, these interventions stabilized food prices and signi�cantly improved poor households’ access to food for a decade or more, as seen in Asia with the green revolution and in a small number of African countries (Kenya and Malawi during the 1970s; Zimbabwe and Zambia during the 1980s). This position has been bolstered by re- cent analyses indicating that price stabilization in certain Asian coun- tries did achieve signi�cant welfare gains and growth effects for the economy (Dawe 2001; Timmer 2002; Dawe and Timmer 2005). Despite this good news, it is now well known that state-led poli- cies incurred high and increasingly unsustainable costs. State-led models required heavy subsidies to offset the high costs of supplying food to urban areas. Most price policy interventions kept food prices low to bene�t urban consumers. In many cases, producer prices were supported while consumers were subsidized, adding to the �scal costs of interventions (Byerlee and Sain 1986). Because parastatal op- erations were susceptible to rent seeking, politically powerful groups 29 30 Managing Food Price Risks and Instability in an Environment of Market Liberalization were able to appropriate many of the bene�ts.20 In oly, swiftly lost a major part of their market to infor- all of the African and most of the Latin American mal operators of small hammer mills, whose num- cases, the treasury costs associated with these poli- bers rapidly expanded. The increased availability cies eventually led them to become �scally unsus- of hammer-milled maize meal at 60–75 percent of tainable (Jayne and Jones 1997). the cost of meal from large commercial mills par- tially or fully offset the adverse effect of eliminating consumer subsidies. Household surveys indicate MARKET LIBERALIZATION: that low-income consumers in particular shifted UNEVEN IMPLEMENTATION, quickly to hammer-milled meal and bene�ted from UNEVEN RESULTS the reforms.21 In Asia, Bangladesh successfully liberalized much The liberalization of food markets was a central el- of its food marketing system. This liberalization, ement of the structural adjustment programs that together with the opening of private trade to India began in the 1980s. As the costs of government and other neighbors, has been an important factor intervention became clear—in terms of subsidies in maintaining stable prices at much lower costs, as well as distorted incentives—donors and many even in the wake of a major natural disaster in 1998 governments pushed for market liberalization. (see the case study on Bangladesh below). Policy-based lending in the 1980s called for the pro- motion of marketing systems driven by the private sector, with little or no direct government inter- GROWING CONTROVERSY OVER vention in markets, and supported by standard public good investments in infrastructure. These FOOD MARKET LIBERALIZATION programs also called for liberalizing trade rather Controversy over the reform programs and their im- than maintaining food stocks as a means of balanc- pacts has increased. Up to the mid-1990s, there was ing supply and demand. a notable convergence among economists and the The implementation of food market reforms has development community regarding the merits of varied widely across countries. Dependent on struc- market reforms and the need to continue this course tural adjustment loans, African and Latin American (van de Walle 2001). From about 1995, this consen- countries began to implement many aspects of the sus has weakened considerably. Dissatisfaction with reforms in the 1980s, but some argue that these re- the limited progress in redressing poverty and farm forms have been partial (see below). In Asia and in productivity problems, especially in Africa, has the Middle East and North Africa, the record has often been attributed to food market reforms and been more mixed. Some countries, such as Vietnam wider macroeconomic adjustments (for example, and Bangladesh, implemented wide-ranging re- see Poulton and others 2005). The role of food aid forms, while others, such as India and Morocco, has generally increased (Barrett and Maxwell 2005), maintained their parastatal food marketing sys- and agricultural growth and poverty reduction tems largely intact, along with their increasing �s- remain elusive, especially in the food crop sector cal burdens (Cummings, Rashid, and Gulati 2005). (Byerlee, Diao, and Jackson 2005). This dissatis- Liberalization produced some clear wins. Fiscal faction has sparked a critical re-examination of the subsidies and marketing margins have fallen in assumptions underlying support for market liber- many cases, although not as comprehensively or by alization policies. as much as anticipated. Experiences from eastern Some argue that the reforms have not worked and southern Africa indicate that the adverse ef- because the private sector has not responded. One fects of reducing or eliminating consumer food school of thought maintains that many of the per- subsidies were sometimes counteracted by other ceived economic benefits of liberalization have aspects of policy reform that encouraged new en- failed to materialize, or that they have not been as trants and greater competition in the food system. great as anticipated (Reardon and others 1999; Sachs For example, in the early 1990s, private marketing 2001; Kherallah and others 2002). The private sector of grain in urban areas was liberalized and subsidies has been slow to emerge in many cases, and private on commercially produced maize meal were elimi- agencies for trade, storage, and input �nancing have nated in Zambia, Kenya, and Zimbabwe. Large- not expanded to the extent expected (Dorward and scale millers, who had enjoyed a protected oligop- others 2004). It is also widely perceived that market Food Market Reforms: Experiences and Lessons 31 reforms exposed farmers and consumers, especially forces but rather the impact of a mixed policy envi- those in remote areas, to greater price volatility ronment in which legalized private trade operates (Gabre-Madhin, 2001). Price instability and its social uneasily within a context of continued strong gov- and political costs have arguably been the Achilles’ ernment intervention in food markets (Bird, Booth, heel of many food market reform programs. and Pratt 2003; Avalos-Sartorio 2005; Jayne, Tembo, These analyses have emphasized market failures, and Nijhoff 2005). This school of thought also high- transaction costs, and coordination failures, leading lights the role of government intervention in chan- to arguments to reconsider some form of direct gov- neling resources to key political constituencies in ernment intervention in markets beyond the stan- the new multiparty political systems that emerged dard public goods–type investments (Reardon and in the 1990s. Indeed, the uneven manner in which others 1999; Dorward and others 2004). This view reforms have been implemented reflects the influ- suggests that the traditional dichotomy between ence of these elements of the political economy. private and public goods is unlikely to provide an Both sides of this controversy have merit, as seen effective or useful guide to policymakers, nor is it in the case studies that follow. It is naïve to expect likely to provide a suf�cient basis for understand- the private sector to step in immediately after the ing the appropriate role of government in food mar- withdrawal of public marketing agencies, espe- keting (Joffe and Jones 2004). cially where transaction costs and risks are high. A second school of thought contends that re- Yet continuing and unpredictable public interven- forms have not gone far enough. Another interpre- tion in food markets has aggravated the risks and tation of the experience in many countries is that undermined the incentives for private investment market reforms were implemented only partially, in market development. The challenge lies in mak- sometimes quickly reversed, and often accompanied ing an orderly transition to an improved marketing by other legal changes that undermined the intent of system. The case studies in the next section describe the reforms.22 Private trade was indeed legalized but three experiences with implementing food market seldom supported, and in many countries old paras- reform. They point to some of the bene�ts as well tatals (or their new incarnations) continue to operate as the lessons from market liberalization. as major actors in markets (Meerman 1997). In some cases, successor agencies were formed with new names, ostensibly signifying a more truncated gov- THREE CASES OF FOOD ernment role con�ned to holding strategic reserves, MARKET LIBERALIZATION but in practice their activities were wide ranging and discretionary. Discretionary market operations by The experiences summarized in the following brief government without transparent and consistently case studies range from relatively negative in enforced decision rules have exacerbated the coor- southern Africa to more positive in Mexico and dination problems, risks, and transaction costs Bangladesh, which made signi�cant progress in faced by private traders in developing food mar- stabilizing markets through private trade and other kets (Brunetti, Kisunko, and Weder 1997), which in means (although Mexico continues to incur high turn have increased price instability (Jayne, Tembo, �scal costs). The lessons emerging from these cases and Nijhoff 2005). Incentives for private storage and others are summarized at the end of the chapter. and investment in storage facilities are often de- pressed by anticipated government actions to re- Southern Africa: Should Governments duce the magnitude of intra-seasonal price rises, Continue to Intervene in Maize Markets? creating a vicious cycle that makes continued gov- ernment action more necessary. Public investment Malawi and Zambia have had a long history of to improve the functioning of markets has also re- heavy government involvement in maize markets mained low, partly because public budgets for agri- (Byerlee and Eicher 1997). Both countries intro- culture continue to be channeled disproportionately duced reforms to allow more private-sector partic- to subsidies (Jayne and others 2002; Cummings, ipation in food markets, but they still maintained Rashid, and Gulati 2005). parastatal food agencies. Their experience in man- Because of these factors, empirical assessments aging the 2002 drought shows how government in- of food market performance since the 1990s in many terventions worsened the impacts of that shock on cases reflect not the impact of unfettered market food prices (Tschirley and others 2004; Rubey 2005). 32 Managing Food Price Risks and Instability in an Environment of Market Liberalization In Malawi, the government was presented with and other trade, regularly exporting from the north a food balance sheet in May 2002 that forecast a and importing from South Africa to the south. deficit of 430,000 tons for the 2002–3 season. The Largely for this reason, maize prices in Mozambique government acted promptly by importing 250,000 remained relatively stable during the 2001–2 crisis, tons of maize entirely through public channels well below levels in Zambia and Malawi (Tschirley (the National Food Reserve Agency) and arranged and others 2004). for 150,000 tons of food aid, for a total formal inflow of over 400,000 tons, nearly covering the forecast Liberalization of Rice Markets in Bangladesh: de�cit. Unfortunately these decisions did not con- An Antidote to the Floods of 1998 sider the large informal flows of white maize from Mozambique into southern Malawi—an estimated Rice dominates food consumption in Bangladesh, 150,000–250,000 tons—which left the country with providing 87 percent of starchy calorie intake. For a large maize surplus (Whiteside 2003). many years, a public marketing agency purchased In March 2003 the government, facing a good rice at �xed prices and distributed it through subsi- harvest and the prospect of storing maize for more dized sales channels, but in the early 1990s a num- than a year, decided to sell some of its accumulated ber of reforms were introduced. The general ration stocks, depressing market prices to very low levels shops for subsidized rice were closed. The govern- (less than two-thirds the levels in Zambia and south- ment initiated targeted rice distribution programs ern Mozambique). This decision undermined incen- for the poor and other populations at risk, and the tives to farmers and private traders and ran up large government liberalized the rice trade. The private �scal costs. The government’s actions reinforced sector began to import substantial amounts of rice the impression, developed through previous expe- from India following poor harvests in Bangladesh. rience, that future shortages might not necessarily This system was put to the test following the severe provide pro�t opportunities for private importers. floods of 1998, which caused enormous crop losses, In Zambia, the government had frustrated pri- especially to the aman (monsoon season) rice crop. vate importers during past food shortages by send- The government encouraged private rice imports ing confusing signals to markets. Similar confusion through favorable policies such as removal of the was evident during the 2001–2 food crisis, when rice tariff and expedited clearance of rice imports. the government announced its intention to import Other factors contributed to a positive trade envi- 200,000 tons of maize grain to cover a national de�cit ronment, including ample rice stocks in India, con- and to sell that grain at below-market prices directly siderable foreign exchange reserves in Bangladesh, to a small number of selected large-scale millers. and the depreciation of the Indian rupee in the early Given this announcement, potential private im- to mid-1990s. Private-sector rice imports exceeded porters stayed out of the market. When, facing a re- 200,000 tons per month for seven months in late source constraint, the government instead imported 1988 and early 1999 (Dorosh 2001), compared to only 130,000 tons very late in the season, maize 34,000 tons in the entire 1996–97 year, and roughly prices rose steeply (Nijhoff and others 2003). Because 80,000 tons per month in the 1997–98 year. If the the grain was channeled to large-scale millers, con- private sector had not been involved during the sumers had to buy re�ned meal at a high price rather 1998 crisis, rice prices would have risen by an esti- than purchasing the less expensive grain and milling mated 40–60 percent (del Ninno and Dorosh 2003). it in a local hammer mill. The poor suffered from reduced rice consumption In Mozambique, by contrast, private trade plays and illness because of the floods, but the liberalized a more prominent role on a regular basis and the policies prevented rice prices from rising above im- government simply stays out of the import busi- port parity and forestalled a much worse deterio- ness. Southern Mozambique contains the nation’s ration in calorie consumption levels. largest urban population and is perpetually food The public sector is not entirely disengaged from de�cient. The center of the country is typically but rice markets in Bangladesh. The government dis- not always in surplus, whereas the north produces tributes rice to the poor and to other populations at a surplus every year. In response to this production risk during emergencies. The government contin- pattern and to the long distances and high costs of ues to hold a strategic reserve and tries to set a floor transporting maize from the north to the south, price for producers. Public procurement and distri- Mozambique has maintained open borders to maize bution have fallen to about 3 percent of national rice Food Market Reforms: Experiences and Lessons 33 utilization, however, which is considerably lower velopment of private storage operations and dis- than in most other Asian countries (the comparable courage forward price markets. Finally, most of the �gure for India is 25 percent) (Rashid, Cummings, bene�ts of indirect market interventions have gone and Gulati 2005). to a minority of large commercial producers rather than to subsistence farmers. Although Mexico has implemented some important aspects of reform, Mexico’s Maize Market Reforms: elements of the reform that were designed to be Good Intentions but Mixed Results23 politically acceptable to important stakeholders still Maize is the dominant staple of Mexico, occupy- hinder the development of a pro-poor and fully ing about half of the cultivated land and grown competitive maize marketing system. by nearly three-fourths of the country’s farmers. Traditionally, the government food marketing parastatal, CONASUPO, subsidized maize pro- LESSONS FROM REFORM duction, established import quotas, acted as a major EXPERIENCES IN ASIA, AFRICA, importer, and set a national price for producers. AND MEXICO CONASUPO also fixed and subsidized consumer prices. This brief overview and case studies highlight some CONASUPO was judged to be inefficient. Its of important lessons emerging from reform experi- operations were perceived to have exacerbated in- ences over the past decade or more. equities in income distribution for several reasons: Price stabilization, as part of a much larger pack- age to support agriculture, has contributed to eco- • Most small-scale producers are net buyers of nomic development and stability, especially in Asia. grain (see chapter 2), and the bene�ts of pan- Some argue that nearly all examples of sustained territorial pricing were highly skewed toward agricultural development in recent decades involved bigger farms and owners of irrigated land major direct state intervention in setting prices and • The pricing system encouraged production in distributing food staples (Dawe and Timmer 2005; areas involving high marketing and transport Poulton and others 2005). Cummings, Rashid, and costs and discouraged private traders Gulati (2005) offer a more nuanced summation of • Consumer subsidies were disproportionately the Asian experience, arguing that price stabiliza- focused on urban areas and did not reach most tion did contribute to economic growth and macro- consumers in rural areas where many of the economic stability, but that such interventions are poor live unlikely to be cost-effective in most developing • The high costs of subsidies were eventually countries today. They emphasize the importance unsustainable of government commitment on many fronts to Avalos-Sartorio (2005) describes how a series of tran- achieve success in agricultural development, in sitional liberalization policies prior to the implemen- general terms embodied in “improved institutions, tation of the North American Free Trade Agreement incentives, and investments.� in 1994 drastically reduced CONASUPO’s interven- A commitment to stabilizing food prices through tions. A new state agricultural marketing agency, state marketing operations and stockholding is ASERCA, was established to facilitate private mar- likely to impose high costs on the public treasury. ket development. Unlike CONASUPO, ASERCA The government subsidy bills in Asia “are stagger- does not directly buy and store agricultural com- ing in all countries that continue to have signi�cant modities. Instead it supports target incomes to pro- parastatal presence� (Cummings, Rashid, and Gulati ducers and subsidizes farmers’ and traders’ use of 2005). In Pakistan’s Punjab, wheat subsidies exceed market-based instruments, including hedging with expenditures by the Department of Agriculture. options. Under its PROCAMPO program, Mexico In some countries of eastern and southern Africa, introduced to farmers direct income transfers, which marketing board losses on maize trading and stock- are decoupled from producer prices. The new poli- holding have sometimes amounted to 4 or 5 percent cies have been largely successful in encouraging of GDP (Jayne, Tembo, and Nijhoff 2005). The high private markets. Overall �scal outlays remain high, �scal cost of Mexico’s parastatal (CONASUPO) although less distortionary, and the level and tim- gave strong impetus to reform in the 1990s (Avalos- ing of marketing subsidies may hamper further de- Sartorio 2005). The opportunity cost of these �scal 34 Managing Food Price Risks and Instability in an Environment of Market Liberalization outlays, in terms of the contribution to agricultural ators and sometimes exacerbating price instabil- growth and poverty reduction that might have been ity. There is a rationale for maintaining some role achieved if these resources had been invested in for the public sector in markets, especially during core public goods, is likely to have been very high. the transition to more liberalized markets, but dis- Food price stabilization policies are almost in- cretionary and unpredictable government actions evitably captured by special interests and become often pose greater threats to long-run market de- difficult to reverse. Stabilization programs that velopment than price instability in unregulated were modest in scope and designed to protect the markets (Coulter 2005). Systems in which private more vulnerable segments of the farm community trade coexists with continued direct government have generally evolved into very large, costly pro- operations generally have not performed well when grams that mostly bene�t a small percentage of the government operations are highly uncertain and population, often larger-scale farmers and proces- discretionary (Jayne, Tembo, and Nijhoff 2005). sors, who have thwarted efforts to introduce re- Conditions change, and food price policies must forms (Avalos-Sartorio 2005; Cummings, Rashid, be continuously updated. Because production and and Gulati 2005; Jayne, Tembo, and Nijhoff 2005). consumption patterns and broader economic con- In most cases, donors and analysts have underesti- ditions change over time, the rationale for particu- mated how political economy factors may influence lar programs requires periodic reassessment. Most the implementation of food market reforms. Where arguments for cereal price stabilization, for exam- it is dif�cult to dismantle parastatals, a second-best ple, are founded on the notions that (1) one or two solution is to implement measures to reduce �scal major food grains account for the lion’s share of costs, enhance ef�ciency, and (most important) en- producer incomes and consumer food expendi- sure that they do the least harm to the long-term pri- tures and (2) there is limited substitution between vate development of the food system (Cummings, commodities. Yet the capacity of producers, con- Rashid, and Gulati 2005). sumers, treasuries, and foreign exchange reserves Partial liberalization, with continuing govern- to handle price shocks has increased tremendously ment intervention in food markets, has often un- (see chapter 2). Despite these changes, “the policy dermined the transition from a publicly controlled environment in agriculture is still hung up with marketing system to a market-oriented one, by the priorities of the 1960s and 1970s� (Cummings, creating a high-risk environment for private oper- Rashid, and Gulati 2005). 6 Policy Options for Making the Transition to Private Markets A central theme in this report is that the major priority of public pol- icy for food systems should be long-run market development rather than short-run, “�re-�ghting� responses to price instability. Almost all governments are concerned with three broad food policy objectives: increased food productivity to promote long-run income growth; the long-run development of markets to enhance ef�cient resource allo- cation and exchange; and protection of the interests of the poor and vulnerable from transitory crises. Effective policy options for mak- ing the transition to private markets must seek a balance among these three broad food policy objectives, and limited public budgets mean that trade-offs will be needed. These issues are at the heart of this chapter, which begins by set- ting out eight broad principles to guide public policy interventions for managing food price risks and instability. These principles are meant to complement the far more detailed analysis that must take place in most countries, given that the appropriateness of a speci�c policy intervention and details of its design are heavily conditioned by country-speci�c factors such as those presented in chapter 2. The initial policy and institutional framework in a given country will also weigh heavily in policy design. Over time, investments in market development can reduce the economic and political costs of price instability, especially if viable risk markets can be instituted. This process is necessarily a long-term one, but this chapter presents a few potential “quick wins� that can help in the medium term. Policies speci�cally designed to manage market risks are discussed in chapter 7. GUIDING PRINCIPLES FOR PUBLIC POLICY DESIGN The management of food price risks and instability should be viewed as part of a more holistic strategy to develop food marketing systems that foster broad-based economic growth, poverty reduction, and food security. A holistic approach is necessary to avoid instituting systems that narrowly target price stabilization but leave insuf�cient public resources for broader improvements in marketing ef�ciency and small- holder productivity. Research has consistently shown high payoffs to sustained public goods investments in areas such as research and development and physical infrastructure (Antle 1983; Oehmke and Crawford 1996; Alston and others 2000; Evenson 2001). Food security 35 36 Managing Food Price Risks and Instability in an Environment of Market Liberalization strategies should refocus their emphasis from price more effective if it reflects wide stakeholder dia- stabilization per se to the sustainable promotion logue, is attuned to political realities, and pays of productivity growth, market development, and special attention to transitional and sequencing poverty reduction. A central component of such a arrangements that mitigate the negative effects of strategy is to reduce costs and risks throughout the policy changes on particular groups (van de Walle food system, thereby making food prices more sta- 2001; Bird, Booth, and Pratt 2003). ble or improving producers’ and consumers’ ability Policy should be consistent, predictable, and to manage unstable food prices. Other components transparent, with open, rather than discretionary, of this strategy include investments in irrigation, “rules of the game.� Market interventions, if ap- technologies that improve drought tolerance, ex- propriate, should be based on a well-de�ned set of tension services, and sustainable means of coordi- a priori rules and not left to the discretion of paras- nating credit, input, and output markets so that tatal agencies or government policymakers (Barro smallholders will face fewer risks, manage them and Gordon 1983; Just 1985; Myers 1992a; Innes better, and absorb their impacts better. 2003). These rules, together with transparency in The corollary to the �rst principle is that because government actions, should be maintained over the public resources are always scarce, public expendi- long term to provide a predictable policy environ- tures should balance the need for long-run invest- ment in which markets can grow. ments in sustainable market development and The costs of market failures should be balanced productivity growth with short-run policy instru- carefully against the costs of government failures. ments for managing food price risks and instabil- Policy interventions have often been designed to ity. Governments often feel pressure to prioritize compensate for the divergence between outcomes public expenditures toward policies and programs that are expected in conditions of perfect compe- that offer short-term payoffs but may do little to pro- tition and outcomes that occur in the real world. mote long-term growth. For example, many African However, such market failure is not sufficient jus- countries are caught in a vicious cycle of frequent tification for government intervention. The sus- food crises that demand an emergency response that ceptibility of many government interventions to draws resources away from long-term investment inefficiency and rent seeking often creates a wide in broader rural development. Notwithstanding gap between the potential and actual bene�ts of the urgency of such crises, the ability to get on a policies. These costs of government failure must long-run growth trajectory requires increased pub- be assessed against the costs of the market failure lic investments, including investments in market the intervention aims to correct. development, research, and infrastructure. Donors, especially food aid donors, should do Short-run interventions should promote rather no harm. Donor interventions have the potential to than undermine long-run market development. support long-run market development as well as to An important role for food policy is to meet short- address short-term needs, such as safety nets. Past term food policy objectives in ways that ensure donor interventions, especially the provision of food that these interventions do not undermine long- aid, have sometimes weakened market develop- term market development. Many interventions ment and sometimes, through the untimely release that stabilize prices in the short run, such as export or local procurement of food aid, aggravated price bans, sudden changes in import tariff rates, and instability. Donors have frequently provided pol- subsidies to offset high import prices, can weaken icy advice that lacks coherence across different op- market incentives and hinder the development of erations or lacks consistency over time (Bird, Booth, market-oriented risk management institutions. and Pratt 2003.) Finally, grain market policies in Policy design should consider not only economic donor countries themselves, especially trade poli- and equity dimensions but also the political econ- cies, shift domestic price instability to global mar- omy of the reform process. The success of alternative kets, with negative impacts on food-importing policies may depend on the capacity and objectives countries (Anderson and Tyers 1993). of governments and the nature of their support base. Policy should foster the use of market-based Conceptually appealing options may not work in instruments for managing risks in ways that shift practice, owing to powerful vested interests that can risks away from the most vulnerable food market influence policy implementation and outcomes. participants. Policy should play an effective role in Advice on reforming food grain markets will be encouraging the emergence of intermediaries that Policy Options for Making the Transition to Private Markets 37 facilitate the use of market-based risk instruments price instability must be evaluated against a num- that will bene�t the whole market chain, without ber of subsidiary policy objectives: destroying incentives for the private sector to par- 1. Defining a “tolerable� level of price vari- ticipate in risk markets. ability; it is much easier and cheaper to re- move extreme prices than to stabilize prices CREATING SPACE FOR PRIVATE completely 2. De�ning the respective roles of domestic pro- MARKETS TO OPERATE duction, imports, and reserves in stabilizing In those countries—and there are many—where food prices and supplies, and the role envi- governments still dominate grain markets, the pri- sioned for the private sector and government ority should be to identify those functions that can in carrying out each of these functions remain under government control and those for 3. Minimizing distortions to long-run market which an orderly transition to private markets can equilibrium prices, de�ned in terms of border be mapped out. Although marketing systems may parity prices (Byerlee and Morris 1993; Timmer be liberalized in the sense that private traders are 1995)25 free to buy and sell, the behavior of government 4. Utilizing scarce public resources in activities agencies and policies may leave little space for pri- that offer high returns in the form of pro-poor vate traders to operate (Jayne and others 2002; growth and market development Coulter 2005). How then can policies supporting 5. Protecting the interests of the poorest and most direct government intervention be restructured to vulnerable and ensuring they are the major yield some of their bene�ts but control costs, mini- bene�ciaries of any remaining subsidies di- mize rent seeking, and provide an expanded role rected at household food security for private-sector participation? The exact sequence, 6. Minimizing the risks of unexpected �scal im- of course, will depend on country-speci�c initial pacts of remaining interventions conditions, but a plausible sequence for a country Crafting a food security strategy within these some- in which a parastatal still dominates food market- times conflicting objectives and constraints is a ing is outlined below. demanding task. Transcending all of these points is the need to Predictable Implementation of a provide clear, well-publicized, and consistent rules Well-De�ned Food Security Strategy to which the government will adhere in achieving these objectives. Attaining greater predictability of Food market reforms are often stalled or muddled government policy—for example, knowing under by the lack of clear objectives for government in- which conditions a given parastatal will intervene tervention in food markets and a well-articulated in the market, and how—will provide more space food security strategy. Many (perhaps most) coun- for the private sector to undertake socially useful tries still de�ne food security in terms of national roles within a managed food system. In most coun- food self-suf�ciency that are ill adapted to the cur- tries, the movement from discretionary interven- rent era of liberalized markets and trade. Food secu- tions to rule-based interventions is the �rst essential rity has dimensions of access to food as well as the step toward providing an enabling environment means to acquire food, and it must be assessed at for private market development. the household rather than the national level.24 This wider de�nition of food security requires a multi- faceted food security strategy, involving measures Subsidy Reform to Level the Playing Field to reduce food prices by enhancing productivity or for the Private Sector trade, improve food production in food-insecure The next order of business should be to eliminate farm households without ready access to markets, untargeted subsidies that alter market prices for all increase the incomes of the poor, and implement producers and consumers. Removing blanket sub- targeted measures to meet the immediate food and sidies is a prerequisite for leveling the playing �eld nutritional needs of the seriously undernourished. for the private sector. This action frees scarce re- Once this broader de�nition of food security is sources for investment in long-run market devel- accepted, interventions in food markets to reduce opment (for example, in infrastructure). Targeted 38 Managing Food Price Risks and Instability in an Environment of Market Liberalization subsidies to protect the poorest and most vulnera- markets in South Africa in the late 1980s and early ble households will in most cases remain crucial, 1990s. Providing second payments in a smallholder but care should be taken to design them in ways farm context may present some dif�culties in ensur- that minimize their effects on the level and volatil- ing that traders (who buy from farmers and sell to ity of market prices. the crop-marketing authority) pass along the second Subsidies result from the way that producer payment to farmers, but this problem is not insur- and consumer prices are set. Setting fixed pan- mountable when the second payment is announced seasonal and pan-territorial prices at planting time publicly. exposes marketing agencies to the risk of substan- Target any remaining subsidies to the poorest con- tial losses and distorts private incentives. To avoid sumers and producers. On the producer side, one strat- these problems in the transition to liberalized mar- egy for targeting subsidies and removing market kets, a number of intermediate steps are possible distortions is through decoupled direct payments and often feasible. to producers (box 6.1). On the consumer side, sev- Reduce the �scal costs of producer price support. An eral of targeting mechanisms are available, includ- alternative to �xed producer support prices is a ing food-for-work programs, the use of “inferior� pooled pricing system, in which a marketing agency or “self-targeted� commodities for free distribu- announces at planting time an initial highly con- tion, and other types of countercyclical safety nets servative price that will be paid to farmers, but the (discussed later). full price paid to farmers is not determined until after the crop is harvested and sold. In this way, Remove Remaining Restrictions on farmers know the minimum price at planting, but Grain Movements and Imports the marketing agency has the flexibility to alter the �nal price after it has determined domestic and ex- Private markets cannot function effectively when port sales revenue and marketing costs. The export the movement of grain is restricted within a coun- parity price (in a grain-importing country) or the try. Such restrictions not only undermine market variable costs of inputs are examples of conservative development but, in decentralized political systems, options for determining the initial payment. If im- greatly heighten tensions between regions (as in plemented in a disciplined, nondiscretionary man- Pakistan; see box 8.1). Likewise levies and taxes ner by an ef�ciently managed marketing agency, on the movement of grain between districts, as in pooled pricing has the virtue of minimizing the risk Zambia and Tanzania, constrain the development of losses for the public marketing agency (and hence and integration of markets across regions, actions the need for government subsidization) while pro- essential for reducing the effects of production vari- viding a minimum support price for producers. A ability on prices. One of the easiest “quick wins� to pooled price system was instituted with some suc- stimulate market development in many countries cess as a transitional step toward liberalized maize would be to lift such restrictions and taxes perma- Box 6.1 Direct Payments to Producers By supporting the incomes of the poor directly, in support programs are feasible, tariffs can be reduced ways that do not distort market incentives, it is possi- and other subsidies can be phased out relatively ble to ease the pain of market reforms. “Decoupled� quickly. Experiences in Mexico and Turkey show that support programs provide direct cash assistance to this approach can be practical for relatively advanced farmers without raising food prices or encouraging countries (World Bank 2005a), but most low-income recipients to remain in unproductive activities. They countries cannot afford direct payments or lack the offer �xed and guaranteed payments to farmers, inde- institutions (especially a land registration system) to pendent of the quantity produced, usually per hectare implement them effectively. In these cases, reforms up to a maximum area. In countries where decoupled may need to be introduced more gradually. Source: World Bank (2005a). Policy Options for Making the Transition to Private Markets 39 nently, enshrining the freedom to move grain across then free to move prices within this band, but ex- district or state borders into law (or even into the treme price movements are eliminated by the floor constitution!). and ceiling prices. Clearly, this option should in- A logical next step is to liberalize trade on imports crease the incentives for private-sector participation and exports, allowing the private sector to enter this while reducing the costs of the program (Buccola market (see the �nal section of this chapter). Licenses and Sukume 1988; Pinckney and Valdes 1988). In to import or export should also be replaced by the practice, grain procurement and release to enforce registration of grain shipments to serve as an infor- price bands is dif�cult to manage, and success stories mation base for all market participants. are scarce. Price bands in food-importing countries are somewhat easier to implement through variable tariffs (see discussion in chapter 7). Introducing Flexibility into Pricing Policies Countries should move away from pan-territorial Tendering to the Private Sector and pan-seasonal prices. Fixing prices over time and space provides disincentives to private traders If relaxation of �xed prices is accepted, a public mar- to store grain or move grain from surplus to de�cit keting agency can authorize procurement, sales, areas. A simple means of alleviating this problem is imports, and even storage based on a competitive to set procurement and release prices from public tender system, perhaps using the system to protect stocks based on seasonally adjusted storage costs a price band (floor and ceiling prices). This strategy (with the lowest price after harvest). This strategy essentially requires the marketing agency to partic- provides incentives to the private sector to partici- ipate in the market along with everybody else, al- pate in storage activities. In Pakistan such a system though the scale of their participation may lead to of seasonally “cascading prices� met with an en- the exercise of market power. As with direct gov- couraging response from the private sector.26 It is ernment operations in the market, the tendering of conceivable that prices can be �xed by region ac- grain purchases to the private sector requires good cording to transport costs and surplus–de�cit status, knowledge of supply and demand conditions, in- but that would be dif�cult to implement in practice. cluding accurate crop production estimates, esti- More information on grain markets is required, mates of marketed quantities being sold off the and it is politically awkward to pay lower prices to farm, and price information systems to provide farmers in more remote regions with higher trans- timely feedback on the effects of market operations. port costs (Coulter 2005). This system, if properly implemented, may help A further re�nement is to move from �xed prices develop capacity in the private sector. An exam- to setting floor and ceiling prices, which the gov- ple of tendering for procurement is the local grain ernment or marketing agency defends through pur- purchases by Ethiopia’s Food Security Program chases or sales onto the market. Market forces are (box 6.2), although setting a support price is not an Box 6.2 Tendering of Grain Procurement for Public Distribution in Ethiopia Under the Food Security Program in Ethiopia, local needs. Several evaluations indicate that under appro- trading firms are invited to bid on purchasing set priate conditions and operating modalities, local pur- quantities of grain in designated surplus areas and chase programs have the potential to stabilize prices delivering them to designated deficit regions, where and encourage investment by local private traders. they are distributed as food aid or sold in markets. These programs also have the potential to destabilize The program, which is designed to complement rather prices, however, if the tendering agency miscalculates than substitute for market activities, is scaled up or supply and demand conditions and issues tenders for down depending on harvest estimates and food aid too much grain. Source: Amha and others (1996). 40 Managing Food Price Risks and Instability in an Environment of Market Liberalization objective. On the consumer side, Malawi is design- A major priority in many countries is develop- ing a system to tender maize sales in remote regions ing improved crop forecasting and supply esti- through the private sector. mates, which can help private and public marketing actors avoid exacerbating market instability through poorly informed trade and stock release decisions POLICIES FOR DIRECT SUPPORT (box 6.3). Food supply estimates must be devel- TO MARKET DEVELOPMENT oped within the context of overall food balance sheets that include substitute “food security� crops. Early IN THE MEDIUM TERM warning systems in drought prone areas have been Coulter (2005), Gabre-Madhin (2005), and others developed in most African countries in recent years provide a broad overview of the many entry points to guide emergency responses, and some, such as for public policy for market development, centered the systems in Mali and Ethiopia, seem to work rea- on the “three Is�—incentives, institutions, and in- sonably well. frastructural investments. The other major priority is market information systems that are commercially oriented but at least • Incentives largely relate to policy reform, to partially publicly financed. Most public systems level the playing �eld and create space for the do little more than collect market prices and report private sector as discussed above. them, too often late and inconsistently.27 In some • Institutional arrangements include contract en- cases (for example, in Kenya and, very recently, forcement, market networks and coordinated Malawi), the tendency has been to bypass public value chains, grades and standards, and mar- systems in favor of private systems, which are seen ket information services. as potentially more client oriented and sustainable. • Infrastructure includes public investment in Yet the public good nature of basic market infor- roads, communications, and critical market- mation means that fully private systems will not ing infrastructure. be pro�table for the foreseeable future, and public- Consistent implementation of these “three Is� is the sector involvement is most likely needed. At the key to successful management of food price risks same time, these information services should have and instability over the long term. A full discussion the financial and managerial autonomy to gener- of these policies is beyond the scope of this report, ate revenue, seek additional outside funding, and but this section will focus on three interventions manage these funds. The objective is to supply in- that can often make a difference in the short to creasingly relevant information to private traders medium term: (1) improving forecasting and infor- while providing policymakers with information, mation systems, (2) liberalizing regional trade, and analysis, and perspectives to make well-informed (3) building private-sector capacity. emergency response and market development decisions.28 Improving Crop Forecasting and Market Information Systems Liberalizing Regional Trade In many countries, estimates of food production Liberalizing border trade can be an important part are unreliable. Public agencies and private traders of overall policy liberalization in the food sector, often over- or underestimate import needs, or with considerable potential to stabilize food prices they make poor decisions on storage. For exam- and reduce food prices to consumers. Trade liber- ple, Zambia no longer can accurately estimate alization protects domestic food markets against maize production from the large-scale farm sector, domestic shocks by allowing more food to be im- as responses to its annual production question- ported in times of shortage and exported in times naires are low. Likewise, food balance sheets and of plenty. Historically, many countries have chosen import requirements are often determined without to take the opposite approach, which is to tax im- reference to informal cross-border trade or local ports and discourage exports to keep domestic “food security crops� such as cassava, which in- markets isolated from international shocks. flates of�cial import requirements and exacerbates If production variability is not highly correlated food price uncertainty and volatility (see the case of among countries in a region, and if those countries Malawi in chapter 5). are well integrated through regional trade, it may Policy Options for Making the Transition to Private Markets 41 Box 6.3 Market Information for Emergency Responses Market information systems must serve governments share data must be combined with improved and donors during a crisis. During the onset of a crisis, and more comprehensive food balance sheets to timely price information is needed to assess the degree account for substitution effects and avoid biases to which supplies in more accessible areas are reach- that lead to overestimates of food import and ing more remote areas through markets. During the food aid needs in crises. crisis response, these data are also needed to deter- • Incidence of different coping mechanisms used mine whether food aid is reaching intended bene�cia- by households, classi�ed by their likely order of ries and not depressing markets. Finally, these systems appearance during a crisis compared to some need to track price trends for food staples and the baseline. Simple questionnaires can collect this assets, especially livestock, that tend to be liquidated information, along with the levels of a range of during crises. Plummeting livestock-to-staple price household assets that might be liquidated over ratios are a classic indicator of mounting vulnerability the course of a crisis. as increasing numbers of households sell livestock to • Household income shares from different purchase staple foods. sources and an assessment of the likely impact Better information must be complemented by better of a crisis on the level of income from each analysis. In times of crisis, a number of different types of source can be very useful in determining the information and analysis can improve policy responses: balance between food aid, cash transfers, and • Information on household budget shares and market responses. cross-price elasticities of demand among staples, • Simple models to predict likely internal and broken down by income level. Consumer budget regional informal trade flows. Source: Tschirley and others (2004). be possible to cancel the effects of small country cially better ports and road connections. In the size on production variability. Production variabil- short to medium term, however, policy and insti- ity across an entire region is almost always consid- tutional changes can facilitate regional trade by erably lower than it is for individual countries reducing or eliminating tariffs and by reducing (Badiane and Resnick 2005). Even in Southern cross-border trade barriers, both regulatory (for Africa, where maize production variability is highly example, phytosanitary standards) and bureau- correlated in South Africa, Zambia, and Zimbabwe, cratic (for example, border crossing documenta- production patterns in Mozambique and Malawi tion). Even unilateral liberalization by a country are largely uncorrelated with the rest of the region, can greatly facilitate the use of trade to reduce sup- which bodes well for regional trade to help stabilize ply and price instability. Regional trade will work price levels (box 6.4). Badiane and Resnick (2005) most effectively, however, when countries in a re- conclude that regional trade in Africa is a win-win gion liberalize domestic markets and harmonize option in terms of both ef�ciency and price stability. remaining food policy interventions (banning ex- Despite its potential to stabilize food supplies port restrictions, for example)—a much more daunt- and prices, regional trade occupies only a small ing challenge. share of the total grain trade in most regions, with some exceptions. Bangladesh is one country that Building Private-Sector Capacity has pursued regional trade liberalization with con- siderable success (Dorosh 2001) (see chapter 5); Poor coordination along the value chain often re- another is Mozambique, which enjoys the most duces the ef�ciency of food markets and leads to stable prices in southern Africa (Tschirley and underinvestment in market development, because others 2004). complementary investments are not made by the Ef�cient regional trade certainly depends on the various players in the supply chain (Dorward and long-run development of key infrastructure, espe- Kydd 2002). In such situations, joint public-private 42 Managing Food Price Risks and Instability in an Environment of Market Liberalization Box 6.4 Variability and Covariance of Maize Production in Africa, 1995–2004 Production variability in small countries may be re- generally lower for the region than for individual duced through regional trade, depending on the co- countries. The covariance of production is still very variance of production among countries within a high among several countries of southern Africa, region. For this sample of countries and commodi- but it is relatively low among others, highlighting ties, grouped by region, the table on the left indicates the potential for regional trade to at least partially that variability in food production (as measured by stabilize individual countries’ food supplies and the coefficient of variation of maize production) is prices. Source: Authors. National and Regional Production Correlation Matrix for Maize Production in Variability, Eastern and Southern Africa Southern Africa, 1995–2004 CV of Maize South Country/Region Production, Years Zimbabwe Zambia Malawi Africa Mozambique Ethiopia 12.6 Zimbabwe 1.00 0.42 0.12 0.44 −0.31 Kenya 8.9 Zambia 1.00 0.09 0.38 0.11 Tanzania 11.2 Malawi 1.00 0.17 0.24 Uganda 8.2 South Africa 1.00 0.41 East Africa 5.8 Mozambique 1.00 Malawi 21.6 Mozambique 4.9 Source: Calculated from FAOSTAT. South Africa 20.3 Zambia 30.6 Zimbabwe 40.9 Southern Africa 17.9 Source: FAOSTAT. action can reduce the transaction costs and risks of farm, community, or national level (as with maize private investment in critical services. In smallholder storage in Uganda; see box 6.6). agriculture, for example, public-private action may These approaches require a supportive general include support to build sustainable commercial business environment. They also require a rela- relationships along the value chain, business devel- tively sophisticated capacity within government to opment services, and speci�c mechanisms for re- (1) develop a strategic vision as the basis for coor- ducing the risks facing private investors (such as dination efforts and (2) design, monitor, and eval- co-�nancing critical investments). uate effective interventions. Such investments aim at building particular in- A high priority in many systems is to build pri- stitutional arrangements that can be maintained by vate storage capacity so the private sector can the private sector without public support after the play a more effective stabilizing role in the food initial phase. They may include strengthening links systems of developing countries. The experience between smallholders and markets by building net- with group or collective storage initiatives by small- works; facilitating contractual arrangements; and scale farmers has been mixed (Coulter 2005). More providing matching grants for coordination along success has been achieved by schemes that focus on the marketing chain, such as the farmer-owned ce- increasing storage capacity at the individual farm, real banks in Kenya (box 6.5); and co-�nancing crit- firm, or family level. These schemes, which pro- ical market infrastructure, such as storage at the vide technology, materials, and sometimes credit, Policy Options for Making the Transition to Private Markets 43 Box 6.5 Farmer-Managed Cereal Banks in Western Kenya The Rockefeller Foundation has supported the estab- tablished, with more than 2,000 farmer-shareholders. lishment of community-owned cereal banks managed Linked to a commercially owned market information by associations of maize producers in western Kenya. system, the banks have obtained maize prices surpass- These banks essentially allow farmers to bulk their ing those paid by local traders by 50 percent or more. sales (�rst at the village level and then across villages), A local nongovernmental organization facilitated for- establish uniform quality standards, and negotiate mation of the associations. While initial experiences sales with large purchasers, especially millers. The have been very positive, it remains to be seen if simi- cereal banks have also established retail outlets in lar institutional innovations can be scaled up rapidly the villages. More than 20 such banks have been es- and effectively. Source: A. Adesina, personal communication. essentially subsidize investments in private storage (Brunetti, Kisunko, and Weder 1997). Efforts to capacity. build private-sector capacity are unlikely to go very The government’s role in building the private far until incentives and space are provided for the sector’s capacity pertains just as much to what it private sector to operate. By realigning government should not do as to what it should do. Survey evi- policies and programs so that they do not overly dence from private traders and potential investors constrain private-sector incentives, many countries in Africa during the 1990s showed that fear of pol- can achieve “quick wins� that will improve how icy reversal was a major impediment to investment markets function and reduce price instability. Box 6.6 Public-Private Partnerships—The Uganda Grain Traders Model Maize traditionally has been a nonstaple crop in Enter the Uganda Grain Traders. This consortium Uganda, grown by smallholders under low levels of of 16 private grain traders was formed in 2001 in inputs and yielding little. The grain market was not well partnership with the government to exploit Uganda’s developed, and operational reserves for storage were emerging export market for grain (maize and grain lacking. Exports through ad-hoc border trade with legumes), coordinate export development, and im- Kenya and purchases by the World Food Programme prove the quality of exports. The public sector con- (WFP) were highly variable (WFP purchases of maize tributed by financing construction of the storage ranged from a low of 8,500 tons in 1997 to a high of facilities. The consortium operates the facilities and 98,500 tons in 2003). This variability, combined with buys and sells grain (which includes selling to export variable demand from the Kenyan market, led to highly markets and the WFP). The better-organized market unstable prices. Three major price collapses have and expanded storage facilities have developed occurred since 1988, with prices fluctuating between Uganda’s grain exports and helped stabilize domestic US$50 and more than US$300 per ton. prices. Source: Magnay (2004); Poulton and others (2005). 7 Speci�c Policy Options for Managing Price Instability and Risk Chapter 6 discussed broad policy options for making the transition to private markets and promoting market development, options that would go a long way toward reducing price instability and risk. This chapter describes four policy options that focus speci�cally on man- aging price instability and risk. The �rst two—(1) piloting and facili- tating the adoption of market-based risk management instruments and (2) providing countercyclical safety nets for the poor—are con- sistent with creating space for private markets and transitioning to a market-based system. The second two—(3) variable tariffs and (4) stra- tegic reserves—are more interventionist policies. They must be applied with great care, if at all, and be accompanied by speci�c safeguards to ensure “arm’s-length,� rule-based management. Market-based risk management instruments and countercyclical safety nets might best be viewed as long-run investments that require institutional innovation and support and that eventually can be fully consistent with long-run market development. Variable tariffs and strategic reserves might best be viewed as short-run measures for achieving speci�c, immediate food security objectives, which ulti- mately may conflict with an extended transition to a market-based system. MARKET-BASED RISK MANAGEMENT INSTRUMENTS A market-based risk management instrument is any freely exchanged �nancial contract that allows parties on one or both sides of the ex- change to reduce their exposure to risk or alleviate its consequences. A simple example is a bank loan, which can smooth variable income flows and allow consumption to remain relatively stable over time. A more complex example is the purchase of a weather derivative, which pays off when rainfall falls outside an objectively de�ned and mea- sured normal range. Many types of market-based instruments either are being used or could be used to manage food system risks in developing countries. Similarly, many participants in the food system potentially could benefit from using these instruments, ranging from individuals, households, and firms engaged in producing, storing, processing, and trading food commodities to public marketing agencies partic- ipating in and regulating food markets. Table 7.1 summarizes the major types of market-based risk management instruments and sug- 45 46 Managing Food Price Risks and Instability in an Environment of Market Liberalization Table 7.1 Market-based Risk Management Instruments and Their Potential Users Risk Management Instrument Warehouse Futures Weather-index Commodity-linked Potential User Credit Markets Receipts and Options Insurance Finance Small-scale farmer High potential High potential Low potential Moderate potential Low potential Small-scale trader or processor High potential High potential Low potential Low potential Low potential Larger-scale farmer High potential High potential Moderate potential High potential Low potential Larger-scale trader or processor High potential High potential High potential Low potential Moderate potential Consuming households High potential Low potential Low potential Low potential Low potential Public food/strategic reserve agency High potential Moderate potential Moderate potential Moderate potential Moderate potential Source: Authors. gests the degree to which different potential users Warehouse Receipts might �nd them useful. The major instruments— A warehouse receipt system allows participants to credit markets, warehouse receipts, futures and op- deposit a stated quantity of a specified quality of tions contracts, index-based weather insurance, a commodity into a warehouse, where it can be and commodity-linked �nance—are discussed in the sections that follow. pooled with other grain of similar quality. A receipt is issued to the owner as evidence of location and ownership. The receipt is a negotiable instrument Financial Markets that can be sold or used as collateral for a loan, Credit markets are coping mechanisms: They do backed by the claim to the commodity held in the not reduce risks per se, but help individuals or warehouse. �rms mitigate the consequences of negative shocks Warehouse receipts facilitate risk management after they have occurred. Access to credit markets in three main ways (Lacroix and Varangis 1996; facilitates borrowing to maintain consumption lev- Coulter and Onumah 2002; Coulter 2005). First, els when incomes fall; makes critical investments they give participants better access to formal credit possible; and also reduces or delays distress sales markets by providing reliable, veri�able collateral of assets, which are often detrimental to long-run for loans to mitigate the consequences of shocks. productivity and growth (Rozensweig and Wolpin Second, warehouse receipt systems facilitate pri- 1993; Morduch 1995; Townsend 1995). vate storage, giving farmers the flexibility to mar- Reliable access to credit at reasonable terms has ket their crop at different times of the year rather high potential to benefit all participants in the than strictly at harvest, when prices are usually the food sector (table 7.1). Financial markets provide lowest. This diversi�cation of sales across time the foundation for a market-based approach to helps to manage risk and, when widely adopted, risk management. Without available and effective can also help reduce seasonal price variability (Lai, financial markets, it is difficult to see how many Myers, and Hanson 2003). Third, a well-structured of the more sophisticated instruments discussed and reliable warehouse receipt system generally later will succeed in managing food sector risks, makes food marketing more ef�cient by acting as a except perhaps for the largest firms and public clearinghouse that enforces ownership claims agencies that can access international credit mar- and can be an impartial third party that guarantees kets. Policy approaches to facilitating development performance on contracts. of rural financial markets are discussed in detail Warehouse receipts are already widely used in in World Bank (2005b) and are not addressed fur- grain marketing systems around the world to pro- ther here. vide secure collateral for credit and as an instru- Speci�c Policy Options for Managing Price Instability and Risk 47 ment for delivering traded commodities. Public Instead, traders offset their commitment by taking food agencies and food relief agencies may also out an opposite position in the same contract (in participate in and use warehouse receipt systems. other words, buying contracts that were previ- South Africa has developed a substantial ware- ously sold and selling contracts previously bought). housing industry for agriculture, but similar ser- Because prices fluctuate between the time the ini- vices are in very short supply elsewhere in eastern tial position is taken out and the time it is closed and southern Africa. The only comparable systems out, holders of the contracts make pro�ts or losses. in this region are the warehouse receipt systems for By taking out futures positions whose returns are grain in Zambia (box 7.1) and coffee in Tanzania, negatively correlated with profits from produc- and a few localized pilot schemes for grain in tion, trading, or processing operations, the cash Uganda and Kenya. If models like those in Zambia position becomes hedged, and overall portfolio could grow and be replicated, they would add sig- risk is reduced. Box 7.2 provides a simple example. nificantly to private storage capacity for small- Options are different from futures in that they holder farmers and also improve the efficiency, give the option buyer the right, but not the obliga- transparency, and competitiveness of grain mar- tion, to buy (a call option) or sell (a put option) the keting systems. underlying asset (usually a futures contract in the These systems and other means of improving case of commodity options) at a strike price speci- private storage capacity and access to credit are �ed in the option contract. The option can be exer- unlikely to provide immediate relief for problems cised at a speci�ed maturity date (and sometimes caused by short-run price instability and food inse- before, at the discretion of the buyer). Trade in op- curity. Instead they should be viewed as long-run tions can be used to put a floor under losses but investments in institutional capacity building. still allow individuals and firms to participate in Several conditions must be in place to ensure that gains when prices move in their favor. In this way, they can succeed and that a range of stakeholders options operate a lot like price insurance, because participate: (1) an effective system of grades and a premium (the price of the option) is paid up standards; (2) suf�cient trust, integrity, and quality front in order to reduce risk by guaranteeing a min- control to ensure that there is essentially no default imum return. risk in using them; and (3) regulatory procedures One of the major difficulties in using futures and oversight to ensure the integrity of the system. and options to manage food system risks in low- income countries is that there are few relevant markets. Table 7.2 lists some of the world’s major Futures and Options Contracts food commodity futures and options exchanges. Commodity futures contracts are commitments Almost all of the high-volume markets are located to make or take delivery of a specific quantity of in industrial countries. Their contract speci�cations a specified quality of a commodity at a particular are designed to meet the needs of industrial country location and time in the future. In most well-func- producers, traders, and processors. A major excep- tioning futures markets, only a small percentage of tion is SAFEX in South Africa, which provides re- contracts are satis�ed by actual product deliveries. gional futures markets for wheat, white maize, and Box 7.1 The Zambian Warehouse Receipts Program The Zambian warehouse receipt program, launched in most of which was collaterally �nanced, and total 2000, is regulated by the Zambian Agricultural warehouse capacity registered for 2005–6 is 120,000 Commodities Agency Ltd., a nongovernmental agency tons. Commercial farmers constituted most of the early owned by stakeholders. To date, the program involves adopters of the service, but smallholders accounted four certified warehouse operators and four banks. In for 9 percent in 2004–5, and their participation ap- 2004–5, farmers deposited 65,500 tons of maize, pears to be growing. Source: Coulter (2005). 48 Managing Food Price Risks and Instability in an Environment of Market Liberalization Box 7.2 Example of Futures Hedging Suppose a trader buys 100 tonnes of white maize at But futures prices have also fallen, and so the 500 rand [R] per tonne with the intention of holding it, futures price for September delivery a month later transporting it, and �nally reselling it to an urban- is now R600 per tonne. The trader buys the futures based processor. The trader does not yet have a sell contract back at this price and makes R18 × R100 = price and so is exposed to the risk of price declines. R1,800 on the futures trade (minus brokerage com- The trader sells one futures contract (equivalent to 100 missions). Hence, losses on the physical trade were tonnes) for September delivery at a price of R618 per offset by gains on the futures trade and overall tonne. A month later the trader has the maize trans- portfolio risk is reduced. ported and ready to sell but the prices have fallen and If the prices had risen over the month instead of the price received from the processor is only R480 per fallen, then extra pro�ts on the physical trade would tonne. The trader has lost R20 × R100 = R2,000 on the have been offset by losses on the futures trade and, physical trade. again, overall portfolio risk would be reduced. Source: Authors. yellow maize. SAFEX contracts have been growing In the short run, existing global markets may be steadily in liquidity since the market was estab- useful for managing food price risks, depending on lished in 1995. basis risk—the extent to which local grain prices Some developing countries, such as India and are correlated with futures prices quoted on global China, are moving to establish local futures and op- futures exchanges. If these prices move together tions exchanges, although there are severe obsta- closely, then the potential for managing price risks cles to developing national futures exchanges in will be high, but if they are only loosely correlated, low-income countries, such as weak marketing in- basis risk will be high, and futures and options frastructure and lack of liquidity. Investing in the hedging will not be effective at reducing price risks. development of local exchanges should therefore at The degree of basis risk will differ by commodity best be viewed as a very long-run response to prob- and location and needs to be evaluated empirically lems of food price instability. on a case-by-case basis. However, unlike coffee, Table 7.2 Major Global Futures and Options Exchanges for Food Staples Location Market Main Food Crop Contracts USA, Chicago Chicago Board of Trade Yellow maize, wheat, rice USA, Kansas City Kansas City Board of Trade Wheat USA, Minneapolis Minneapolis Grain Exchange Yellow maize, wheat Canada, Winnipeg Winnipeg Commodity Exchange Wheat Europe Euronext White and yellow maize, wheat, potatoes Argentina Rosario Futures Exchange Maize, wheat Brazil Bolsa de Mercadorias & Futuros Maize Japan, Tokyo Tokyo Grain Exchange Maize, rice China, Dalian Dalian Commodity Exchange Maize, rice, beans China, Zhengzhou Zhengzhou Commodity Exchange Wheat India National Commodity and Derivatives Exchange Maize, rice South Africa, Pretoria South African Futures Exchange White and yellow maize, wheat Source: Authors. Speci�c Policy Options for Managing Price Instability and Risk 49 cocoa, and to some extent sugar, which have glob- 2005; and Sarris, Conforti, and Prakash 2005). In ally integrated markets (in other words, low basis this case, countries are essentially hedging their risk), food grain markets tend to be more localized export revenues or import bills, presumably to and insulated from one another because of trans- enhance macroeconomic stability and fiscal out- port costs, quality differences, and trade restrictions lays. But when a public agency does the hedging, (chapter 3). it is not always clear how the bene�ts of hedging will Some case studies have examined the basis risk be passed back to the producers, traders, processors, and hedging potential for particular food crops in and consumers that make up the food system. If the particular countries. Faruqee, Coleman, and Scott public agency is directly involved in procurement (1997) evaluated wheat imports in Pakistan and (it buys and imports or exports the grain itself), found good hedging potential using U.S. wheat then the gains or losses from hedging can be passed and futures and options contracts. This �nding has back along the supply chain by altering domestic been supported by an analysis of hedging aggre- prices bid or offered by the agency. gate wheat and maize imports in several develop- One means of providing the necessary coordina- ing countries using Chicago Board of Trade wheat tion without direct procurement is for a public and maize futures and options (Sarris, Conforti, agency to act as an intermediary. An example is the and Prakash 2005). Dana, Gilbert, and Shim (2005) Agricultural Product Options Program of ASERCA evaluate the potential for Malawi and Zambia to in Mexico (see Avalos-Sartorio 2005). Grain pro- hedge maize imports using SAFEX in South Africa, ducers purchase a subsidized option premium from concluding that hedging could be an effective risk the program in return for a guaranteed minimum management strategy. These studies suggest that price at harvest. ASERCA then constructs pooled basis risk is low enough that existing global futures hedges for all of the participating producers using and options markets may provide effective hedging Chicago Board of Trade options contracts. ASERCA potential for food imports into low-income coun- never takes control of the grain, which is marketed tries, at least in some important cases. through private channels, but it passes the bene�ts, Where hedging potential exists, a key question and part of the costs, of option hedging back to is who will do the hedging. Among the potential farmers. In this way ASERCA essentially acts as an users (table 7.1), small-scale farmers and traders intermediary to facilitate price risk management by would generally �nd the costs of individual partic- producers. Although farmers have made little use ipation prohibitive. Trading on global futures and of this program to date (primarily because other options markets requires considerable resources, government programs have reduced or destroyed including access to credit, use of foreign exchange, the incentive to participate—see Avalos-Sartorio good market intelligence, reliable and speedy com- 2005), it does show how creative public policy can munications, and the analytical capacity to con- facilitate a market-based approach to risk manage- struct risk-minimizing portfolios. Furthermore, the ment even when transaction costs, information volume speci�cations on most global futures and costs, and diseconomies of scale are signi�cant ob- options contracts are too high to be of use to small- stacles to small-scale farmers’ participation. A sim- scale operations. Even in industrial countries where ilar program operates in Brazil. the exchanges are located, farmers make little di- Intermediation can also occur without direct rect use of futures and options markets. government involvement. For example, large-scale Larger-scale traders and processors (and even traders, processing �rms, supermarket chains, co- large-scale farmers) have higher potential to use operatives, or farmer organizations can offer �xed futures and options because they have better access or floor price contracts to smaller-scale producers, to the required resources and their scale of opera- traders, and processors (box 7.3). The intermedi- tions can accommodate the quantities speci�ed in aries pool the risks and hedge them using global the contracts. A fairly large and sophisticated oper- futures and options markets. This is exactly what ation is required to trade directly in these markets. happens in many industrial countries. In the United The most commonly suggested strategy for low- States, for example, individual farmers (particu- income countries to use global food futures and op- larly those operating on a smaller scale) make very tions markets is for a public agency that controls or little direct use of futures and options markets, but regulates imports to do the hedging (as in Faruqee, grain elevators (that is, traders) offer these farmers Coleman, and Scott 1997; Dana, Gilbert, and Shim cash contracts that have forward �xed or floor 50 Managing Food Price Risks and Instability in an Environment of Market Liberalization Box 7.3 The Potential of Forward Contracts Forward commodity contracting can be an important is available for future delivery. Finally, the existence of market-based risk management instrument. Forward futures markets greatly promotes forward contracting contracts are agreements to exchange a speci�ed and substantially lower costs, by allowing buyers to quality and amount of a commodity at a future date offer small-scale, individualized forward contracts to and location, but at a �xed price agreed upon today. producers. Buyers can then pool the resulting risks Relative to futures, forward contracts can be written and manage them on futures markets. for any amount and so have more flexibility to be used Although forwards are potentially more flexible by small-scale operators. Any delivery location and and useful than futures contracts for small-scale farm- product grade can be speci�ed in the contract so there ers and traders, futures contracts are low-cost, highly is generally no basis risk. liquid, and easily transferable �nancial instruments There are many ways that forward contracting that do not incur default risk. In many developing could be facilitated. Formal commodity exchanges fa- countries the inability to enforce forward contracts, cilitate both cash and forward commodity buying and especially for staple food crops, means that default selling, by providing implicit or explicit guarantees on risks are too high to support viable forward markets contract performance, encouraging networks where without some form of guarantee on performance. trust relationships can be built up to a level that will Hence futures can transfer relatively large risks much support forward exchange, and by imposing costs on more easily and at much lower cost than forward those who default (e.g., exclusion from further partici- contracts, although as noted above, they are in prac- pation in the exchange). Warehouse receipts can also tice complementary, when traders offer forward con- play an important role in facilitating forward contract- tracts to producers, backed by operation in futures ing by acting as a guarantee that the speci�ed product markets. prices embodied in them. For example, the elevator work in countries that are consistent importers can offer farmers a forward contract that prices the (exporters) and where import (export) requirements grain at planting but does not require delivery until are known well in advance. For example, if a coun- harvest. Alternatively, the elevator can offer a con- try that expected to import maize actually produces tract at planting that requires the farmer to deliver enough maize to export, then hedging the expected at harvest and guarantees a minimum price but import requirement before the harvest is known allows the farmer to receive a higher price if prices could lead to unexpected and possibly large losses. rise over the growing season. The elevator is able to Of course, uncertainty about the right quantity to offer these contracts because it pools the resulting hedge is a problem that also plagues individual risks across a large number of farmers and then farmers and �rms, but they probably have better hedges the aggregate risk on futures and options knowledge of their production situation—and can markets. This strategy allows elevators to be com- respond more quickly to changes in that situation— petitive and attract business, while enabling farm- than a centralized government agency hedging ers (indirectly) and elevators (directly) to manage aggregate imports or exports. their price risk through futures and options trading. Because the use of futures and options markets The choice between direct government procure- by the public and private sector are unlikely to co- ment and hedging versus a decentralized approach, exist easily, governments must choose between in which trade is undertaken by the private sector centralized control of procurement and hedging and hedging is encouraged via intermediation activities or a decentralized approach that encour- (by �rms, strong farmer organizations, or by pub- ages more private-sector participation. The latter lic agencies), is an important choice. If procurement approach has significant advantages and is more and hedging are undertaken directly by a govern- consistent with the long-run emergence and devel- ment agency, the incentives for private individuals opment of market-based institutions. Extensive, and �rms to participate will be reduced signi�- decentralized use of futures and options contracts cantly. Furthermore, this approach will really only will not emerge rapidly or spontaneously, however. Speci�c Policy Options for Managing Price Instability and Risk 51 Growth will require public investments in educa- weather stations throughout a country. For exam- tion and capacity building, as well as institutional ple, a farmer may pay a premium for an insurance innovations that facilitate indirect use of these in- contract that pays US$25 for every 1 millimeter that struments by smaller-scale farmers and traders. the observed rainfall index falls below its critical One �nal point about futures and options hedg- level of 500 millimeters per year, up to a maximum ing is that even when relevant markets are avail- of US$5,000 (in other words, there are no extra pay- able, they allow risk reduction only over the short ments if rainfall drops below 300 millimeters per run and generally are not useful for hedging annual year). If observed rainfall is below the threshold income fluctuations over long periods (Gardner level, leading to low yields, the farmer receives a 1989; Lence and Hayenga 2001). This limitation in payment that can at least partially compensate for the degree of risk reduction has the bene�t of forc- reduced crop production. ing market participants to continue to be respon- Index-based weather derivatives are quite com- sive to longer-run changes in prices, which is mon in industrial countries, where contracts focus desirable from an economic ef�ciency perspective. primarily on heating-degree and cooling-degree days in major cities, and are used by �rms whose returns depend heavily on the weather, such as elec- Index-Based Weather Insurance tricity companies. Weather insurance is less com- Index-based weather insurance is a class of �nan- mon in developing countries, but a private market cial derivatives written against deviations from for rainfall insurance is emerging in India, and sev- threshold rainfall or temperature indices that are eral other schemes have been piloted or investi- constructed from weather data measured at secure gated (box 7.4). Box 7.4 A Proposal for Weather Insurance in Malawi A proposal for weather insurance in Malawi has two the farmer has to bear before the insurance pay- components (see Ibarra and others 2005): a microlevel outs begin) insurance product that can be sold to individual farm- • Require that farmers have access to credit so ers and a macrolevel product that the government can they can afford the premium, and that insurers use to obtain emergency funds to meet food security are willing to offer the product at premium levels commitments in times of drought. that remain attractive to farmers. The microlevel product would: The macrolevel product would: • Focus on the important maize-producing region • Focus on countrywide maize production surrounding Lilongwe • Construct a rainfall index that is correlated with • Construct a rainfall index that is highly corre- the average maize yield in Malawi, based on lated with maize yield outcomes in the region, rainfall data collected at weather stations based on rainfall data collected from the throughout the country Lilongwe airport • Estimate the �nancial burden facing the govern- • Estimate the extent of �nancial loss per unit area ment food reserve agency in times of yield stress that is associated with changes in the index (for (for example, to �nance food imports or costly example, a 1-millimeter reduction in the rainfall social safety net policies) index below a “normal� trigger level might cause, • Structure an insurance product that pays out on average, a maize yield reduction of 10 kilo- according to the agency’s need for funds as the grams per hectare, valued at 15 kwacha (MK) per country-wide rainfall index declines kilogram, which gives an overall payout of • Require that the exact nature of the agency’s �- MK150 per millimeter of the index per hectare) nancial burden is speci�ed, and that an insurer is • Set the trigger level that determines the de- willing to offer the product at premium levels ductible on the insurance (the amount of risk that remain attractive to agency participation. Source: Ibarra and others 2005. 52 Managing Food Price Risks and Instability in an Environment of Market Liberalization It should be clear that weather insurance is not time. The price of insurance could rise because focused directly on managing price risks, at least insurers will require a risk premium that com- for the microlevel product used by farmers. In fact, pensates them for taking on this catastrophic risk, when producers are receiving payouts on their and if this premium is high enough, it will destroy rainfall insurance, yields should generally be low the incentives for insurers to participate (Duncan and prices higher (although incomes will be low and Myers 2000). The risk premium could be kept because of reduced yields). In this way the insurance lower by reinsuring part of the risk on global insur- acts more like an income safety net for producers ance markets, if opportunities are available. and less like price insurance. However, in principle In summary, while index-based weather insur- there is no reason to restrict rainfall insurance to ance may not be attractive to all food sector par- producers. Consuming households might also ben- ticipants in all situations, these contracts do have e�t from purchasing rainfall insurance if it pro- considerable potential for managing risks and vides income when local food prices soar owing to providing a safety net in times of climatic stress. low rainfall and low local yields. The only real re- Farmers, both small- and large-scale, are the obvi- quirements for this alternative to be feasible is that ous potential users, but others, including traders the premium must be attractive to consuming house- and even consuming households, may potentially holds given the risks they face, and they must be benefit. Public agencies may also have potential able to pay the premium in advance. Weather in- demand for these insurance products, but an ob- surance could also be used to manage the food aid jective measure of an agency’s liability under unfa- requirements of donor agencies, as proposed in vorable weather outcomes is required. There is also Ethiopia (Morris 2005). a danger that rent-seeking will eat into the insur- Governments and government agencies could ance payouts when they occur if the agency is not also use index-based weather derivatives to insure credibly committed to using the funds for their their liabilities in times of climatic crisis (box 7.2), intended purpose. but this strategy would be subject to severe rent- Similar to the case of futures and options, the seeking problems without a credible commitment growth and development of index-based weather to use the insurance payouts for their intended pur- insurance will require public investment in devel- pose (Myers 1992a; Innes 2003). oping not only the insurance products but the insti- The advantage of index-based weather insur- tutions that are needed to support viable insurance ance is that it is based on objective measures of markets. This is another example of long-term insti- readily observable events that cannot be influenced tution and capacity building that is consistent with by human behavior. Such schemes avoid the moral long-run market development. hazard and adverse selection problems that plague traditional agricultural insurance schemes, which Commodity-Linked Finance are based on individual farm yields. They also have low transaction costs and can be scaled down to A problem with most rural credit products is that payout levels that might interest relatively poor in- there may be little connection between the in- dividual households. come flows of borrowers and the service flow re- The weakness of the index-based approach is quirements of the debt. In other words, farmers that the returns to individual farmers or traders (or may be required to make large loan payments the food prices paid by individual consumers) may precisely when their current incomes are low. One not be strongly correlated with the weather index potential means of overcoming this problem is and hence with the insurance payout. For example, with commodity-linked finance. While there are if a farmer fails to receive a payout when yields are many different types of commodity-linked �nance, low, then the insurance will not provide effective commodity-linked bonds are a prominent exam- risk management. This issue is similar to the basis ple (Priovolos and Duncan 1991). These bonds risk issue for futures and options trading, and it can link the principal, and possibly interest payments, destroy the incentive to insure. Another compli- to future realization of a speci�ed set of commod- cation is that enormous demand for index-based ity prices, so that when commodity prices are high, insurance products will expose the insurer to cat- debt service obligations are also high, but the bond astrophic risk. If the insured event occurs widely, issuer has the income to service the debt (and vice many payouts will have to be made at the same versa). In this way, commodity-linked finance Speci�c Policy Options for Managing Price Instability and Risk 53 helps hedge price risk and smooth consumption when risks and the degree of risk aversion are high streams. (Anderson 2001). Although commodity-linked bonds (and other From a policy perspective, a market-based ap- forms of commodity-linked �nance) are interesting proach to risk management should not require the in principle, they have several limitations for man- large, persistent budgetary outlays that historically aging food price risks in low-income countries. have been a feature of price stabilization schemes. Often the institutions and market infrastructure for Even if public agencies are trading futures and supporting these financial products are lacking. options, the trading profits and losses should ap- In industrial countries, commodity-linked �nance proximately cancel each other in the long run, if is used only by large firms that can accommodate the futures and options markets operate ef�ciently. the accompanying high transaction costs. A major It is important to note, however, that there could be problem is that there may be strong incentives to large trading losses in the short run, which would issue the bonds, but often there are no strong incen- presumably be offset by gains in physical trading tives to buy them, other than for speculative pur- operations or be passed back to others if the agency poses. Because buyers require a significant risk is operating as an intermediary. premium before they are willing to hold these Perhaps the most important advantage of market- bonds, the interest rates can be quite high. For the based risk management instruments is that in gen- same reason, these bonds tend to be very illiquid. eral they facilitate and enhance the private sector’s The only viable way for commodity-linked �nance role in the food system rather than displace it. The to offer real risk management alternatives for in- use of market-based risk management can improve dividual farmers and households appears to be price discovery, enhance market efficiency, and through some kind of public or private intermedi- improve price transparency and information dis- ary that issues the bonds on a wide scale and then semination throughout the marketing channel. packages the resulting financial instruments into These secondary bene�ts occur most commonly products that might be accessible and of use to with organized commodity exchanges. For futures individual farmers and households. Commodity- and options to work effectively, there must be an linked �nance would appear to hold more promise open, highly transparent system of exchange that for managing the macroeconomic risks associated facilitates information dissemination. These mar- with import-export fluctuations and the external kets also generate incentives to collect market debt positions of governments rather than the intelligence and information (because futures and individual risk portfolios of small-scale produc- options exchanges provide a forum for making ers and households (O’Hara 1984; Myers and trading pro�ts based on superior information) and, Thompson 1989). in so doing, help to disseminate this information to other market participants through the price system. Finally, an important social bene�t of futures and ASSESSING THE POTENTIAL options markets is that they facilitate the collection OF MARKET-BASED RISK of time-series data on market prices that can be MANAGEMENT INSTRUMENTS used for evaluating market performance over time. The Advantages of a Market-Based Approach Challenges to Implementing a Relying on a market-based approach to managing Market-Based Approach food system risks has a number of distinct ad- vantages (Anderson 2001; Larson, Anderson, and For several reasons, low-income countries rarely Varangis 2004). Participation is generally volun- use market-based instruments to manage food sec- tary, so people participate only at a level that suits tor risks. When local shortages occur, it may be their particular situation. In contrast, participa- dif�cult to enforce contracts for food staples. The tion in traditional price stabilization schemes is small size of the farming and trading enterprises that compulsory: Everybody is subject to the stabilized serve the traditional food sector in these countries, prices. Furthermore, the welfare gains to individu- and the poorly developed �nancial markets, limit als and �rms using market-based risk management the liquidity required for successful trading. Few strategies are sometimes substantial, particularly low-income countries have the market intelligence 54 Managing Food Price Risks and Instability in an Environment of Market Liberalization systems, grades and standards systems, commu- Main Messages on Market-Based Approaches nication systems, storage and marketing infra- Market-based risk management instruments have structure, and experience and education to use some clear advantages for managing food price these markets effectively. Basis risk is another major risks in low-income countries, in ef�cient ways that impediment to both futures and options trading allow voluntary participation. The evidence sug- and index-based weather insurance. gests that hedging potential is considerable in some Somewhat ironically, continuing government cases, even when restricted to existing global futures intervention in food markets may be one of the most and options markets. However, effective develop- serious impediments to innovation and to the de- ment and use of such markets is clearly not going velopment of risk management markets for the food to occur without active public policy support. sectors in many countries. Interventionist policies Although there are many barriers to participation, reduce or destroy the incentive to participate in especially for small-scale producers, traders, and market-based risk management mechanisms, be- processors, the public sector can play an important cause there is no incentive to manage risk when role in reducing these barriers and facilitating the prices are effectively stabilized through policy, and use of market-based risk management instruments. because such policies tend to disconnect local prices An option that could be adopted very quickly from world prices, which reduces the hedging is the direct trading of market-based risk manage- potential of the global markets. Furthermore, if ment instruments by public food marketing agen- government interventions are discretionary and cies to hedge government liabilities. Yet this is a dif�cult to predict, they simply add another layer risky venture for the public sector. Such trading of risk that individuals and �rms may �nd dif�cult requires not only considerable information and to hedge using available market-based risk man- analytical capacity, but it is subject to the same agement instruments. problems of inef�ciency and rent seeking that have In a liberalized market environment, however, plagued direct public intervention in food markets governments can play an important role in facili- in the past, especially when there is no credible tating and expanding the use of market-based risk commitment regarding how the gains will be spent management instruments. This role includes in- (and the losses �nanced). A preferred strategy is to vesting in: encourage the private sector to use these markets • Basic market infrastructure, such as systems by making long-run investments in the standard for transport, communications, grades and public goods that create an enabling environment standards, and market information (see sec- for �nance and risk markets, including grades and tion 6.3). Lacking these basic investments, standards, credit market development, communica- more sophisticated risk management instru- tion systems, market intelligence systems, regula- ments are unlikely to succeed. tions, and support for local or regional commodity • Institutions that support the development of exchanges and insurance products. There may also rural �nance markets, expand the availability be a role for policy support of market intermediaries of credit, and encourage and facilitate private that provide access to risk management markets grain storage. for small-scale operations, particularly when these • Analytical capacity, technical support, and ed- markets are just beginning to develop. Perhaps most ucation to facilitate use of global futures and important, governments can provide a predictable options markets by large-scale domestic pro- policy environment that does not destroy the in- ducers, traders, and processors. centives for private individuals and �rms to trade • The development and support of intermedi- market-based risk management instruments. ary institutions that can pool and repackage the risks facing small-scale producers, traders, COUNTERCYCLICAL SAFETY NETS and processors and then hedge the pooled TO PROTECT THE POOR risks using global futures, options, and insur- ance markets. AND VULNERABLE • The development of objectively measured Because of the perceived risks involved in eliminat- weather indices that can provide a foundation ing consumer subsidies or experiencing large price for index-based weather insurance. spikes under liberalization, safety nets are generally Speci�c Policy Options for Managing Price Instability and Risk 55 required to protect the poorest and most vulnerable means to protect the livelihoods of the rural poor groups of the population. The role of safety nets when shortfalls in food availability are accompa- within the context of a market-based approach to nied by shortfalls in purchasing power (Sen 1981). managing food price risk and instability may be Successful countercyclical programs require that very important in providing reassurance that the additional resources kick in on a timely basis, that poorest segments of the population will be pro- they are well targeted to the most vulnerable, and tected in the transition away from state-dominated that they support and do not undermine market food distribution. Governments have been toppled development. Current public safety net programs by food riots often enough for the threat of such an for the chronically poor can provide a scaffold to occurrence to jeopardize attempts at reform. scale up implementation and coverage when needs Most public safety net programs in low-income are relatively high (Alderman and Haque 2005). countries are “pro-cyclic� in the sense that public Effective design of countercyclical safety nets also resources are most available when they are least requires that they can be scaled back after a crisis. needed. As mentioned, resources to procure grain The timely availability of funds is likely to be the for food aid programs are most plentiful in good most critical element for a countercyclical safety net harvest seasons and least available in drought years program to function well. Weather-based triggers (Hicks and Wodon 2001). This means that in times for activating budgets may be one way to shorten of stress most developing countries will be less able the time between identifying the need for transfer to target income or food to the poor without assis- programs and implementing them. As discussed tance from external donors. below, when food aid is anticipated but expected In countercyclical safety net programs, budgets to arrive later than desired, the use of emergency feasibly can be scaled up in times of need and scaled reserves (physical or financial) can help provide back when the need passes (Alderman and Haque timely assistance. 2005). Countercyclical safety net programs are likely Domestically resourced programs are subject to to be very important for governments seeking to the “Samaritan’s dilemma,� which is that govern- eliminate or reduce the costs of price stabilization ments may not devote as much attention or re- policies. They provide the means to transfer cash or sources to developing effective national programs if in-kind income to households and communities, international donors and humanitarian organiza- and they serve as insurance against variable income tions are perceived to be ready to assist if a food cri- streams in response to climatic and other shocks. sis occurs (Gurenko and Lester 2004; Alderman and The incorporation of an insurance function into the Haque 2005). Indeed international �nancial institu- design of safety net programs recognizes that in- tions already provide a number of such programs, jecting food into markets is often an inadequate many on concessional terms (box 7.5). Attention Box 7.5 Emergency Assistance Funds from International Financial Institutions Effective implementation of countercyclical nets re- prove the temporary nature of the price shock quires timely access to financial resources in kind or and the absence of concessionality. in cash in the event of a severe shock, such as a • The IMF’s Poverty Reduction and Growth Facility, world price spike or a widespread natural disaster. which has more of a grant element but is re- International �nancial institutions have set up a variety stricted to low-income countries. of mechanisms. These include: • Support under the Emergency Assistance and • The International Monetary Fund’s (IMF’s) Natural Disasters window of the IMF, which has Compensatory Financing Facility, which is espe- a concessional element for eligible members of cially designed to facilitate food imports during the Poverty Reduction and Growth Facility. sharp changes in global prices. This facility was • Emergency Recovery Loans of the World Bank, not widely used because of the requirement to which include quick disbursing elements. Source: Alderman and Haque (2005). 56 Managing Food Price Risks and Instability in an Environment of Market Liberalization to moral hazard problems of this type must be ad- VARIABLE TARIFFS TO MANAGE dressed in the implementation and management of effective countercyclical safety net programs. WORLD PRICE SHOCKS Safety net programs must be capable of target- Variable tariffs can be used as a short-run policy in ing benefits effectively to the poor and excluding food-importing countries to insulate domestic food others, for two reasons. First, to the extent that markets from large world price shocks. The chal- those with effective demand for food receive free lenge with such policies is to manage the tariff level food through safety net programs, markets will in a way that allows domestic prices to track world experience depressed demand, with the potential prices in the long run and that maintains the private to exacerbate price instability and undermine long- sector’s incentive to participate in international run market development. Second, because the trade. The historical tendency to manage variable resources available for safety net programs are tariffs in a very discretionary way makes private- limited, such programs must be able to prevent sector planning dif�cult and opens the programs to benefits from spreading to those who are rela- capture by vested interests, especially large-scale tively less poor. farmers. If variable tariffs are used, therefore, rates In certain instances, the ability to cost-effectively should be set according to well-speci�ed rules rather target safety net programs can be enhanced through than discretion. the distribution of “inferior� or “self-targeted� com- Variable tariffs work best for imposing a floor modities, which are important in the diets of poor price in food-importing countries because the tariff people and are also consumed disproportionately can be raised in the event of an extreme drop in world prices. Foster and Valdes (2005) suggest that by them (Jayne and others 1999; Gutner 2002). the floor price should be based on the cost of pro- Examples of self-targeted goods include unre�ned duction in the most ef�cient exporting country, to wheat flour in countries such as Pakistan and the minimize risks of encouraging inef�cient domestic Arab Republic of Egypt, and yellow maize in eastern production. Other countries, such as Chile, have and southern Africa. Moreover, by supporting the used a fixed departure from a moving average operation of marketing channels for self-targeted border price as the trigger. Unless the tariff is al- commodities, low-income consumers can rely on ready high, variable tariffs do not address effects the market to a greater extent for their food needs, of price spikes on consumers, and because high thereby reducing the magnitude of resources re- tariffs on food grains are sources both of ineffi- quired for targeting the needy through adminis- ciency and higher inequality (the poor are penal- tered food assistance programs. ized), they are not usually a desirable option. Nor Finally, safety net programs must be able to are variable tariffs appropriate for price extremes support market development in the long run while generated by domestic shocks in countries that op- protecting the poor in the short run. Public works erate in wide bands between import and export programs, such as food-for-work programs, are an parity. Furthermore, under current World Trade important type of safety net with the potential to Organization (WTO) rules, the scope for variable help build markets through the creation of roads tariffs is limited to the bound tariff (the tariff level and other forms of market infrastructure. These declared to the WTO), although proposals are being programs are self-targeted to the extent that the discussed to allow variable tariffs as a safeguard to wage rate is set below the market wage, although developing countries that import food.29 Finally, if some research has shown that even relatively well- countries are to liberalize and encourage regional off households may bene�t from enrolling mem- trade, variable tariffs have to be agreed at the re- bers in public works programs if the opportunity gional level, as implemented in the Andean zone. cost of their time is low (Barrett and Clay 2003). In sum, variable tariffs have some scope to pro- Some of these programs can also be combined tect producers from extremely low prices in food- with efforts to build market demand, such as the importing countries, but they require very open “input for asset� program in Malawi, which pays and transparent rules that would preferably be for the construction of roads and other public monitored by the WTO to prevent abuse and polit- works, partly with vouchers to purchase fertilizer, ical patronage (Foster and Valdes 2005). They or the Ethiopian safety net program, which is switch- should be used only for a very small number of ing from food-for-work to cash payments, partly to “strategic commodities� that have well-de�ned in- support domestic market demand for grains. ternational reference prices. Finally, it is clear that Speci�c Policy Options for Managing Price Instability and Risk 57 variable tariffs are of limited value for protecting from local grain production, they involve public against price spikes, a goal that is often the main procurement. A tendering system is the most effec- concern of food-importing countries. tive approach, but the size of such an operating re- serve is generally too small to influence national markets and therefore unlikely to be a feasible tool AN ARM’S-LENGTH FOOD RESERVE for stabilizing prices. In some countries, such as TO MANAGE DOMESTIC SHOCKS? Ethiopia and Uganda, food aid operations have The last and most dif�cult step for countries un- reached a level that they can influence markets dergoing market liberalization and privatization is either negatively (for example, through the untimely how to deal with public grain reserves. Countries release of food aid into a surplus market or the un- maintain such reserves for three major reasons timely local procurement of food aid) or positively (NEPAD 2004): (for example, when strategic procurement helps develop markets or stabilize prices) (Coulter 2005). 1. Emergency reserves for a major natural disas- In practice, social objectives could be combined with ter, such as a severe drought, especially in procurement by, for example, requiring that tenders eastern and southern Africa, usually linked to be supplied from more remote, poorer regions that food aid donations have surplus grain but thin markets. Ef�ciently run 2. Food security reserves for servicing both public procurement could provide a much-needed emergency relief and a public distribution stimulus for competition and demand in such mar- system (mainly in Asia) for the chronically kets. It is important, however, to recognize the trade- poor, again often supported in part through offs between ef�ciency and social objectives. food aid donations On a larger scale, many countries in Africa, after 3. Buffer stocks, now often known as strategic the closure of public food marketing agencies, still reserves, aimed at smoothing prices, but also attempt to operate a buffer stock to support prices serving as emergency relief and supporting in years when harvests are good and to dampen public distribution systems, if they exist price rises in years when they are poor, or even to Clearly the first two objectives, which operate ride out extreme prices in world markets. Of course, largely on the consumer side of the market, are not these same reserves also serve during crises and focused on stabilizing prices per se, although they for public food distribution systems. Despite their do target food security for vulnerable consumers. appeal, the record of such operations is not encour- In principle, a small and well-managed stock could aging (box 7.6), and consumers often face greater provide “degrees of freedom� in responding to a instability in food prices and availability when crisis until commercial imports and food aid can such strategic reserves are used, as seen in Malawi arrive. To the extent that these reserves are supplied (see chapter 5).30 Box 7.6 NEPAD’s Sobering Findings on Strategic Reserves A comprehensive review by NEPAD (2004) captures crop by falling prices and lack of markets. the record of food reserve agencies as follows: This has led them to reduce plantings with In southern Africa, continued attempts to use subsequent adverse impact on the overall strategic grain reserves to help stabilize cereal production and grain availability situation in prices for both producers and consumers have following years. At the same time, consumers undermined market incentives for private have also faced greater instability in grain traders to perform normal arbitrage func- markets, with respect to both physical quanti- tions that could otherwise have satisfied ties available and price. In most cases, there- governments’ food security objectives in most fore, experience with strategic grain reserves years. As a consequence, small farmers have in this part of Africa up to now has been less often been penalized for producing a surplus than satisfactory. Source: NEPAD (2004:34). 58 Managing Food Price Risks and Instability in an Environment of Market Liberalization The case for strategic reserves is strongest in land- • Central-bank type autonomy, with complete locked countries that are close to self-suf�ciency in a independence from political processes, and major staple, and where reliance on trade to equal- with clear and well-de�ned objectives ize supply and demand can potentially lead to • Highly professional management, with a good large price swings (from export to import parity). information system and analytical capacity But even here, timely access to �nancial resources • Flexibility to hold the combination of grain is needed for reserves to operate effectively, and and �nancial reserves that minimizes costs any grain reserve must be combined with a �nan- within acceptable levels of risk cial reserve (usually in foreign currency). In coastal • Clear and open rules for market intervention countries, the �nancial reserve should be all that is and transparency in its interventions needed (Poulton and others 2005). For example, • Access to a fund or �nancial markets, to pro- Senegal depends solely on a dedicated �nancial re- vide flexibility to respond in an emergency serve for drought emergencies (NEPAD 2004). A These fairly strict requirements have proven very professionally managed reserve could also take out dif�cult to implement. Whether the development of insurance or hedge to reduce �nancial exposure. an arm’s-length, professionally managed reserve Conceivably some of the problems with reserves can be achieved in practice is unclear and would could be surmounted by setting up an arm’s-length, vary by country and region. Such a reserve is also professionally managed reserve along the following costly, occupying resources that have signi�cant lines: opportunity costs. 8 Managing the Policy Dialogue There is little doubt that the slow progress in reforming food mar- kets reflects weaknesses in how the World Bank, other donors, and reformers within countries have managed the policy dialogue. Broad recommendations on privatization and public-sector with- drawal have been less successful in the case of food markets relative to export crop markets, since both producer and consumer interests are at stake, involving the welfare of a large cross-section of the pop- ulation. In a few cases a “big bang� approach was implemented, as when Mexico committed to a free trade agreement with the United States and Canada, but even there, reforms in maize marketing re- mained partial after Mexico built safety nets and “sweeteners� into the reform program on the consumer and producer sides (Avalos- Sartorio 2005). The large number of stakeholders and the political sensitivity of food market reforms suggest that a time-intensive, resource-consuming process of policy dialogue is essential to design a phased sequence of reforms that is broadly acceptable. Vested interests in public paras- tatals as well as private agencies often stall reforms. Politicians are ex- tremely cautious in implementing food market reforms if they perceive that they will lead to more volatile food prices. Two somewhat different approaches to managing the policy dia- logue are evident in Pakistan and Malawi. In Pakistan, after more than two decades of donor-commissioned studies and recommenda- tions to liberalize wheat markets, the policy dialogue has advanced considerably over the past year, facilitated by a resident external ad- viser to the Federal Minister for Food and Agriculture (box 8.1). The key to success has been a series of structured meetings among key policymakers in each province and at the federal level, followed by a national meeting (Faruqee 2005). Broad consensus has been estab- lished for moving forward and giving a much larger role to the pri- vate sector. Reforms will still be partial, however, and key details remain to be resolved regarding the management of a strategic re- serve and the maintenance of a minimum support price. In Malawi, the policy dialogue on maize market reforms has been facilitated by a Poverty and Social Impact Analysis (PSIA), imple- mented over the past two years (box 8.2). As in Pakistan, in Malawi the new approach followed a continuing stalemate between donors and the government of Malawi on reforming the food marketing parastatal. Participatory workshops were a central part of the process to break the stalemate, although a much wider range of stakeholders 59 60 Managing Food Price Risks and Instability in an Environment of Market Liberalization Box 8.1 Managing the Dialogue on Wheat Policy in Pakistan International �nancial institutions have long but unsuc- An intensive dialogue, initiated in 2004 by the advi- cessfully pressured Pakistan to reform wheat markets, sor to the federal minister for food and agriculture, in- often by attaching conditions to policy loans. Although volved all relevant public decision makers in ministries the ration shop system of distributing wheat to con- and parastatals at the federal and provincial levels. sumers was disbanded in the 1980s, vested interests They focused on: within the government and its parastatals, as well as in • Clarifying the objectives of the current policy, es- private businesses (especially wheat millers), stalled pecially the conceptual distinction between the more extensive reform. A further complication was minimum support price and the price that was that wheat-surplus provinces, notably the Punjab, had set to procure stocks for the strategic reserve different interests from de�cit provinces, such as North • Distinguishing the stocks needed for the strategic West Frontier Province. Adding to the pressure was reserve from those needed to manage the public the fact that wheat subsidies in the largest province, food distribution program Punjab, exceeded the sum of all other expenditures • Targeting remaining consumer subsidies on the on the agricultural sector, crowding out desperately poorest needed investments in roads, irrigation, and research • Moving toward a price band and away from and development. Another major �scal issue was the �xed producer and consumer prices high variability in subsidies from year to year, depend- • Agreeing to ban restrictions on moving wheat ing on the harvest and import prices. across provinces In 2002, the federal government introduced limited • Developing a strong crop forecasting and infor- reforms, notably a seasonally adjusted wheat procure- mation system for public and private decision ment and release price, which provided needed space makers for the private sector to operate. In 2004, however, the • Phased downsizing of public agencies involved government of Punjab, unable to procure its targets at in wheat markets. the announced price, once again restricted the move- While many details are being worked out and the ment of wheat out of the province, sparking a sharp policy has yet to be implemented, the dialogue is confrontation with de�cit provinces. promising. Source: Faruqee (2005). Box 8.2 Key Elements of a Poverty and Social Impact Analysis The Poverty and Social Impact Analysis (PSIA) was • Uses local expertise as much as possible, again introduced by donors and international financing to foster ownership agencies as a central element of the national dialogue • Employs a diverse toolbox that mixes quantitative on sensitive policy reforms. The analytical process and qualitative methods gives special attention to participatory processes of • Uses transparent assumptions and processes policy discussion as well as analytical studies to • Addresses risks to policy implementation. identify the winners and losers from prospective pol- For more details on the PSIA approach and studies icy reforms. In practice, a good PSIA: that have already been completed, see • Is undertaken early, to inform the way that reforms www.worldbank.org/psia. are designed • Consults a wide range of stakeholders, to foster broad ownership Source: Nucifora and Lisulo (2005). Managing the Policy Dialogue 61 was involved in Malawi than in Pakistan (Nucifora The main messages from these ongoing experi- and Lisulo 2005). The Malawi case has also focused ences, as well as others (Faruqee 2005; Nucifora and much more on analyzing who will win or lose from Lisulo 2005), can be summarized as follows. the reform. The dialogue resulted in broad agree- • Much more time and resources must be in- ment that the private sector should handle maize vested in the policy dialogue over sensitive marketing in urban areas and in rural areas with reforms such as those involving food market- good infrastructure but that public intervention is ing parastatals. It is especially important to needed over the medium term to serve producers build trust among key stakeholders. and especially consumers in more remote areas. • Both standard economic analytical tools and Details on how to support this “social market� are participatory approaches are important in still being formulated. Another issue is that Malawi building the evidence for the dialogue. has a separate food reserve agency (aside from the • Particular attention should be given to se- food marketing parastatal), whose performance quencing reforms over time, and “second- has been widely questioned but was not analyzed best� outcomes will be the norm. Phasing in the PSIA (see Bird, Booth, and Pratt 2003 and reforms to score quick wins is one way to build chapter 5). support for further reforms. Of course, participatory stakeholder approaches • Donors should provide a consistent and co- to policy reform do not guarantee that a transpar- herent message on the reform process (Bird, ent discussion of stakeholder interests will ensue Booth, and Pratt 2003). While donors and ex- or that consensus will be achieved on the way for- ternal experts can support valuable analytical ward. Stakeholders have asymmetric information studies, local champions and local expertise and knowledge about the likely effects of reforms. are essential for success. Knowledgeable stakeholders who stand to lose • It is important to put in place a system to from reforms can attempt to manipulate the pub- monitor and evaluate implementation of the lic discussion and turn opinion against reform, reforms, paying special attention to monitor- which occurred during maize market reforms in ing price movements. Kenya in the 1990s and in Zimbabwe after the mid- • Reforms are best implemented in a good har- 1990s. Improvements in policy and the use of pub- vest year, but the acid test for sustainable re- lic expenditures cannot be ensured through such a forms is their performance following a bad process. harvest or global price shock. 9 Conclusions and Recommendations Food price risk and instability are perennial issues that have dogged food policy debates for decades. Their persistence is understandable, given the continued importance of food staples as a wage good, their high share of national income and expenditures in low-income coun- tries, and political sensitivities to sharp changes in food prices. Since the 1990s, these issues have taken on new urgency in the context of market liberalization, suggesting that the standard policy responses are inadequate. The reform of food markets, particularly public food marketing agencies, has been very slow in many countries. In several cases, re- forms have been reversed. The relationship between food market reforms and food price instability is especially controversial. Some are reluctant to liberalize food markets because of fears about the potential impacts on food price instability, or out of the conviction that food prices have become more unstable in countries that have liberalized. Others contend that “halfway� reforms create the worst of all possible worlds, in which the private sector is encouraged to operate in an environment where governments continue to inter- vene in discretionary and unpredictable ways that make prices even less stable. Over the years, commodity price stabilization and risk manage- ment have received considerable attention from researchers and pol- icymakers in industrial- and developing- country settings. This study was motivated by the need to revisit the problem of food price insta- bility and risk in low-income countries and investigate the bene�ts and costs of alternative policy responses. In particular, the study aimed to provide guidance on how to make the transition from state- dominated markets to private markets in ways that do not expose producers and consumers to the risk of unacceptable price spikes and collapses. It has addressed �ve key questions: 1. What are the sources and magnitudes of food price shocks? 2. What are the magnitudes (actual and potential) of the economic and social costs stemming from food price instability in low-in- come countries? 3. What is the status of food market reforms, and what can be learned from the experience to date? 4. How can countries sequence reforms in ways that promote ef�- cient market development and protect the interests of the poor? 5. What are appropriate policy responses to food price instability and risk in a liberalized market environment? 63 64 Managing Food Price Risks and Instability in an Environment of Market Liberalization HOW BIG IS THE PROBLEM, poor households, have become more diversi�ed over time, giving them more flexibility to handle sharp AND WHO SUFFERS MOST? spikes in the price of the dominant food staple. In Country context de�nes the problem. Food policy rural areas, the empirical finding that emerges decisions must be tailored to the individual circum- consistently in most parts of the developing world stances of each country but, as a starting point for is that a majority of households are net food buy- identifying an appropriate policy response, coun- ers, while a relatively small minority of wealthier tries can be grouped roughly according to common households are grain sellers. The poor, who are needs and risks. A simple framework for classi�ca- overwhelmingly net food purchasers, suffer dis- tion focuses on low-income countries and regions in proportionately from high food prices. Among pro- which food consumption is dominated by one sta- ducers, the impacts of low food prices are at least ple: rice in Asia and Madagascar, wheat in Pakistan partially offset by negatively correlated production and the Middle East and North Africa, white maize variability. in eastern and southern Africa, and millet/sorghum This analysis leads to a second major conclusion: in Sahelian countries of West Africa. These are the Food policy should generally emphasize the im- countries and regions where the poor are most ex- pacts of unstable food prices on consumers—rural posed to sharp movements in the prices of food and urban, and especially the poorest and most staples, especially spikes in the prices paid by con- vulnerable—rather than on producers. sumers. These countries were further classi�ed ac- How significant are food price shocks? At the cording to their potential exposure to price shocks global level, variability in world grain prices remains from domestic climatic events and to shocks gen- significant, with coefficients of variation around erated by world grain markets. trend of 20 to 30 percent for rice, wheat, and white Based on this classi�cation, rice and wheat im- maize. Although there is no evidence that variabil- porters, especially the least developed (examples ity has increased (indeed, prices were most unsta- include Madagascar, Bangladesh, and Yemen) are ble in the 1970s), there is concern that changes in most exposed to world price shocks. Many other world markets, especially reductions in the stocks Asian and middle-income countries are exposed as held by major producers (China, the United States, well, but their improved infrastructure and foreign and the European Union) and rapid growth in de- exchange reserves have placed them in a much mand in Asia, may provoke higher and more un- better position to handle such shocks than three stable prices in the future. decades ago, when many public food marketing The evidence on the magnitude and frequency agencies were established. Landlocked countries of price instability in domestic food markets, actual in southern Africa that depend on maize and, to a and potential, is limited. In general, producer prices lesser extent, other landlocked African countries for wheat and maize in importing countries have (such as Ethiopia and some Sahelian countries) are been more stable than international prices, reflect- most exposed to domestic sources of shocks. Food ing dif�culties in transmitting international prices production in these countries is highly variable, into domestic markets, as well as continuing policy and their capacity to operate on world markets is interventions in many countries that insulate domes- limited by high transport costs and foreign exchange tic markets from world prices. There is no convincing constraints. evidence to date that domestic food price instabil- The �rst conclusion—obvious but too often over- ity has increased over time in the sample of coun- looked—is that food policy decisions and market tries reviewed. reforms are highly specific to their context. More Domestic price instability tends to be highest in attention needs to be paid to a country’s particular two groups of countries. The �rst group comprises stage of development, food consumption patterns, Latin American countries where macroeconomic agroclimatic factors, geographical situation, and in- shocks, especially sharp exchange rate devaluations, stitutional setup in designing appropriate food have resulted in highly unstable prices in a number policies. of cases. The second group is African countries, A country typology hides considerable hetero- especially landlocked countries where the wedge geneity within countries between rural and urban between export and import prices is high because areas, regions, and households. But generally the of high transport costs and poor market infrastruc- consumption patterns of urban households, even ture. The high import-export parity wedge, com- Conclusions and Recommendations 65 bined with high domestic production variability, have exploited trade opportunities, especially re- increases the impact of domestic shocks, especially gional trade, as the main mechanism for stabilizing drought, on prices. A contributing factor, particu- domestic grain prices. larly in Southern Africa, is the uncertainty created But what about the many countries that are by unpredictable government interventions in food stuck halfway in the reform process, hovering be- markets and imports. tween old parastatal models and private, market- Under a full market liberalization scenario, led approaches? In this situation, discretionary food price shocks, whether from global or domes- interventions to meet an emergency (or even just a tic sources, are potentially signi�cant in many situ- declaration of the intention to intervene) have been ations. For example, in Ethiopia the price wedge especially destructive to incentives for private-sec- between import and export parity has allowed tor participation. maize prices to fluctuate from about US$50 to nearly Other important lessons have been learned from US$250 per ton in recent years in Addis Ababa, and the varied experience with market reforms. Many probably more in remoter regions. Likewise, coun- countries paid insuf�cient attention to designing tries depending on rice imports have faced world an orderly sequence of reforms that systematically export prices falling from US$340 per ton in 1996 to increased the role of the private sector and built a low of US$170 per ton in 2001, and rebounding con�dence in a market-based approach. Nor was to more than US$300 per ton in 2005. suf�cient attention given to political economy con- What are the costs of price instability? The costs siderations (such as vested interests that maneuver of unstable food prices can include the loss of eco- to maintain the status quo) and to designing a re- nomic ef�ciency, detrimental impacts on the welfare form program that takes account of these realities. of the poor (including undernutrition and reduced survival), and negative macroeconomic externalities that retard economic growth. There is little consen- MOVING FORWARD: BROADER sus and generally weak evidence on the magnitude POLICY OPTIONS of these costs. The effects of unstable food prices on economic ef�ciency are probably not large in most Policies are chosen within a set of constraints formed cases. The most persuasive cases for the negative by the political system and by limitations on avail- effects of high food prices can be made for effects ability of public funds. These constraints force gov- on (1) household food security and nutrition and ernments to make explicit tradeoffs in allocating (2) macroeconomic performance. These costs could public expenditures, and it is imperative that these be signi�cant in certain situations—for example, in tradeoffs are made in ways that enhance the long- the poorest countries with poor infrastructure, weak run performance, growth, and stability of the food capacity to import, dependence on a single domi- sector and the economy as a whole. nant staple, and susceptibility to drought—all char- This review highlights a number of policy op- acteristics of several landlocked countries in Africa. tions for moving forward, recognizing that it is es- pecially dif�cult to make generic recommendations for such a country-speci�c and complex topic. One WHAT DO WE LEARN FROM general recommendation is that food policy deci- POLICY REFORMS? sions, rather than focusing on price stabilization options per se, should take a holistic approach to The record of food market reforms in low-income food security in which long-run productivity growth and even many middle-income countries is mixed and market development constitute the �rst prior- at best. Some countries, such as India, have main- ity. This leads to four speci�c recommendations, tained their old parastatal systems more or less in- summarized below. tact, but mounting costs have made most of these systems unsustainable. Other countries, such as 1. Problems of food price instability and food Bangladesh, Mali, and Mozambique, have intro- insecurity need to be addressed in a holistic duced and sustained signi�cant reforms that en- framework that includes: abled them to weather a major natural disaster at a • Measures to improve overall productivity much lower cost than in the past and with tolerable of food staples, especially investments in levels of price instability. Notably, these countries research and development and irrigation 66 Managing Food Price Risks and Instability in an Environment of Market Liberalization • Measures to reduce the severity of domes- 4. Sequence market reforms in a consistent man- tic shocks caused by climatic events (such ner that creates space for the private sector to as measures to promote irrigation or crop operate. “Big bang� approaches to market re- diversi�cation) form have rarely worked in practice. For mar- • Measures to improve the overall ef�ciency kets to develop over the long run, including of markets, including investments in trans- regional markets, consistent progress must be port and communication infrastructure, made in opening space for the private sector. storage, information systems, market regu- More analytical work and policy dialogue will lations, and institutional arrangements that provide a better basis for designing a logical, improve coordination along the market sequential program of reforms. Finally, gov- chain ernments need to implement the agreed pro- • Measures to mitigate the impacts of shocks, gram in a predictable and consistent manner. including market-based measures (such A generic sequence that would gradually in- as forward pricing and weather insurance) crease the role of the private sector includes: as well as countercyclical safety nets. • Eliminating blanket subsidies and revising The corollary of this recommendation is that remaining subsidies in ways that level the direct public interventions in food markets to playing �eld for the private sector and tar- manage food price risk should be a last resort get the poor. (see below). • Removing remaining restrictions on grain 2. Resources should be reallocated from short- movement within a country and reducing run, “�re-�ghting� interventions to manage restrictions on grain imports and exports. food prices to investment in long-run market • Moving away from �xed procurement and and private-sector development, including in- release prices toward seasonally adjusted centive frameworks, market institutions, and prices and price bands. infrastructure in line with the recommenda- • Tendering remaining public procurement, tion above. Nonetheless, even investments in imports, and even storage to the private sector, using a highly transparent process market development must be sequenced in to increase ef�ciency, reduce rent-seeking, ways that confer measurable gains in the short and build private-sector capacity. to medium term. Public-private partnerships (for example, through farmer and trader asso- ciations) to develop production and market SPECIFIC POLICY OPTIONS FOR information systems, storage, and market net- works are often the �rst priorities for improv- MANAGING PRICE INSTABILITY ing food sector performance. AND RISK 3. Liberalization of trade, especially the pro- Within an overall public policy strategy for food motion of regional trade, is one of the most systems that emphasizes the transition to private effective “quick wins� for reducing food price markets and long-run market development, there volatility in small and medium-sized coun- are roles for the public sector in enhancing price tries. Liberalization of trade shifts a country’s stability and managing food sector risks. Two of exposure away from domestic shocks and to- these will be a standard part of the toolkit of most ward global price shocks, but global shocks food security strategies: (1) piloting and facilitating are usually lessened if trade with neighbor- the adoption of various market-based risk man- ing countries is encouraged. Regional trade agement instruments and (2) countercyclical safety requires action on a number of fronts, includ- nets. Two others may have a role in certain situa- ing long-run investments in infrastructure, tions and when accompanied by speci�c safeguards but the development of (a) consistent rule- to ensure “arm’s-length,� rule-based management: based policies to lift discretionary export bans (3) variable tariffs and (4) strategic reserves. and import restrictions, (b) smooth border- Market-based risk management instruments. Several clearing procedures, and (c) harmonized reg- risk management instruments show considerable ulations, such as phytosanitary rules, would promise in managing food price risks, including go a long way toward creating the incentives facilitation of private storage (warehouse receipt for private traders to engage in regional trade. systems), futures and options markets, and weather- Conclusions and Recommendations 67 indexed insurance. These alternatives are rarely instruments to reduce exposure to risks from its used in low-income countries, partly because the own operations in food markets. Yet direct trad- public sector dominates food markets and partly ing of futures, options, or insurance contracts by because the enabling conditions are lacking, such governments or public food agencies should be as access to �nance, information systems, commu- approached with extreme caution. Large govern- nication systems, market regulations, and capacity. ment futures or options positions are not recom- The major focus of the public sector should be to mended for two reasons. First, even if the public create an environment that facilitates the private sector is successful in using these instruments, the sector’s adoption of these instruments, especially in public sector is likely to undermine incentives for the following ways. the private sector to use them. Second, given the poor record of public-sector interventions in food • Warehouse receipts, for use initially by larger- markets, there is little reason to believe that the pub- scale farmers, processors, and traders, and lic sector’s use of market-based risk management over the longer term by the small-scale sector. instruments would be immune to the same inef�- Warehouse receipts have much potential to ciencies and rent-seeking forces that have plagued reduce risks from seasonal price fluctuations, conventional public food agency operations. develop �nance markets, encourage invest- If governments do choose to become involved in ment in storage, and eventually (when widely direct procurement to manage a small strategic food adopted) to reduce both seasonal and inter- reserve (see below), market-based risk management annual price fluctuations. They cannot be im- strategies may have a potential role in these oper- plemented if an appropriate regulatory and ations. In such cases, options have distinct advan- business environment is lacking, however. tages over futures—first, because of their role as • Futures and options using existing global mar- price insurance, and second, because purchasing kets, for use mainly by large-scale traders and options requires only a single, up-front premium, processors and strong intermediaries, such as whereas futures can entail continuing margin calls if well-developed farmer or trader associations, prices move unfavorably. Even when using options, to reduce exposure to risks from global mar- an effective hedging strategy requires considerable kets. These alternatives are already available investments in analytical capacity and a long-run where the basis risk is low, which appears to be commitment, otherwise hedging could add to risk the case for wheat and white maize for many rather than reduce it. The misuse of futures and op- countries, using U.S. and South African futures tions may expose governments to even greater �scal markets. risks and rent-seeking than conventional public- • Weather-indexed insurance for use by farmers, sector operations in food markets, unless special safety net programs (see below), and (poten- management safeguards are in place (see below). tially) consumers. While not designed for price Countercyclical safety nets. A second major prior- risk management per se, weather-indexed ity for interventions to manage risks is to support insurance can mitigate the impacts of price the development of countercyclical safety nets in spikes or climatic shocks. Successfully piloted ways that are market friendly. Countercyclical at the farm level in India and Mexico, weather safety nets, which kick in when high food prices or insurance can be used more widely where low production threaten household food security, weather indices are good proxies for crop are an integral part of any program to manage food losses, and especially if domestic insurers can price risks. Food aid and food-for-work programs reinsure on global markets. remain the most important safety nets in many countries. In the past, however, untimely imports The public sector should support the development and sales of food aid, along with poor targeting, of a basic enabling environment by conducting the often undermined market development. Food aid analytical work and building the capacity to pilot and other safety net programs can support long- and scale up programs that promote the develop- run market development by: ment of �nancial systems, communication and in- formation systems, regulations, and an appropriate • Converting from food to cash transfers where business climate. food markets already function reasonably well Some recent discussions have also noted the • Scaling up local and regional procurement of potential for the public sector to use market-based food aid, perhaps including the maintenance 68 Managing Food Price Risks and Instability in an Environment of Market Liberalization of a small and well-managed emergency re- tral bank�–type autonomy, (2) highly professional serve, but ensuring that the timing of food management and analytical capacity, (3) strict rule- aid procurement does not aggravate price based market operations to meet a narrowly de�ned instability objective, and (4) tendering of operations, includ- • Incorporating rainfall insurance into safety ing storage, to the private sector. net programs to enhance their ability to trig- ger timely and better-targeted responses to a drought THE IMPORTANCE OF LOCAL • Better targeting of food aid through improved CIRCUMSTANCES information systems and the use of self- Returning to the country typology discussed earlier, targeting approaches, including “inferior� it is clear that food policy design and approaches grains to managing food sector risks will vary widely, • Integrating safety nets with market develop- depending on each country’s context. The overall ment activities, such as the use of food aid to priorities on productivity enhancement and mar- construct local market infrastructure. ket development are fairly generic; they apply in Variable tariffs. Under certain circumstances, vari- many contexts. However, quite different strategies able tariffs can be used to manage downside price will emerge across countries and regions in moving risks to producers from exposure to global markets. to sequenced reforms, creating space for the private To be effective, they should be triggered by well- sector, and addressing speci�c priorities for man- defined rules to reduce political capture and be aging market risks. The Asian countries, in particu- highly transparent in their operation. Technically, lar, still have a considerable reform agenda to open their use also must be approved by the WTO, and space for the private sector. Likewise the opportu- indeed a preferable outcome would be for the trig- nity to apply various market-based risk instruments gers and monitoring of their implementation to be depends a lot on the extent that a country is exposed subject to WTO oversight to maintain maximum to domestic versus global shocks. transparency. Technically, variable tariffs could also be used to reduce risks from price spikes in global markets, but ENTRY POINTS FOR tariffs must be high enough initially that they can be THE WORLD BANK lowered when world prices rise sharply. Given that Food market reform and food security remain crit- high tariffs on food grains are generally undesirable ical areas for Bank engagement. Interest in these for both ef�ciency and equity reasons (most poor issues is burgeoning in many countries, including households, including rural households, are net those which have not yet embarked seriously on re- food purchasers), variable tariffs are unlikely to be forms and those which seem stuck halfway through useful for managing world price spikes. the process. The Bank needs to revamp its analytical Strategic reserves. Many countries still maintain work in this critical area, paying particular attention publicly owned reserves to reduce food price in- to the following points. stability. In a liberalized market economy, the pri- Manage the policy dialogue better. Too often, the mary reason to maintain such reserves should be a Bank’s analytical work has proposed broad recom- targeted food distribution scheme (if there is one), mendations on market reforms but paid little atten- although in a few cases reserves can be maintained tion to how those reforms should be sequenced. The to cope with emergencies (especially in landlocked “big bang� approaches generally have not worked, countries with poor infrastructure). In some cases, and part of the challenge in moving forward is to be reserves may be large enough to influence domes- alert for opportunities to move toward second- and tic market prices, and judicious use of these reserves even third-best options rather than waiting for the may help reduce the impact of domestic shocks on opportunity for full reform. Good analytical work food prices, especially where there is a large wedge must be combined with much more time- and re- between import and export parity prices. Critical source-intensive policy dialogue that is attuned to safeguards must be in place to ensure that opera- political realities (for example, vested interests). tions of food reserve agencies do not destabilize Advice on food grain market reform will be more markets, however, including (1) arm’s-length, “cen- effective if it seeks wide stakeholder dialogue and Conclusions and Recommendations 69 pays special attention to transitional and sequencing private sector and promote the emergence of nec- arrangements that mitigate the negative effects of essary institutions and intermediaries. Extreme policy changes on particular groups (van de Walle caution should be used in promoting use of these 2001; Bird, Booth, and Pratt 2003). The use of PSIAs approaches by public food marketing or strategic to ensure wide buy-in and ownership in this deli- reserve agencies. cate reform process is a step in the right direction Support activities at the regional and global level. and needs to be scaled up. The Bank’s analytical work can play an important Pilot and evaluate new market-based instruments. role in informing global actors in food markets on The recent move by the Bank’s commodity-based the use of safeguard measures such as variable risk management group to analyze the applicabil- tariffs by developing countries. More importantly, ity of market-based risk management instruments this report has highlighted the potential for regional for food staples is providing encouraging results trade as a mechanism to stabilize prices within a re- and should be scaled up. However, this work should gion, which raises a huge agenda for analytical work focus on analytical support and capacity building and policy dialogue to reduce policy and institu- to facilitate adoption of these instruments by the tional barriers to trade in nearly all regions. Appendix 1 List of Workshop Papers Alderman, H., and T. Haque. 2005. “Countercyclical Safety Nets Jayne, T. S., G. Tembo, and J. Nijhoff. 2005. “Experiences of Food for the Poor and Vulnerable.� Market Reform and Price Stabilization in Eastern and South- Avalos-Sartorio, B. 2005. “What Can We Learn From Past Price ern Africa.� Stabilization Policies and Market Reform in Mexico?� Mitchell, D., and J. Le Vallee. 2005. “International Food Price Badiane, O., and D. Resnick. 2005. “Regional Trade Liberalization Variability: The Implications of Recent Policy Changes.� and Domestic Food Market Stabilization in African Countries.� Myers, R. 2005. “Costs of Food Price Instability in Low-Income Coulter, J. 2005. “Making the Transition to a Market-Based Grain Countries.� Marketing System.� Nucifora, A., and A. Lisulo. 2005. “Lessons in Managing Policy Cummings, R. W., Jr., S. Rashid, and A. Gulati. 2005. “Grain Price Dialogue in Malawi: Reforming the Agricultural Development Stabilization Experiences in Asia: What Have We Learned?� and Marketing Corporation.� Dana, J., C. L. Gilbert, and E. Shim. 2005. “Hedging Grain Price Poulton, C., J. Kydd, S. Wiggins, and A. Dorward. 2005. “State Risk in the SADC: Case Studies of Malawi and Zambia.� Intervention for Food Price Stabilisation in Africa: Can it Dawe, D., and Timmer, C. P. 2005. “Managing Food Price Work?� Instability in Asia: A Macro Food Security Perspective.� Sarris, A., P. Conforti, and A. Prakash. 2005. “The Use of Or- Faruqee, R. 2005. “Reforming Wheat Policy in Pakistan.� ganized Commodity Markets to Manage Food Import Price Foster, W., and A. Valdes. 2005. “The Merits of a Special Safeguard, Instability and Risk.� Price Floor Mechanism under Doha for Developing Countries.� Papers presented to the international workshop “Managing Food Gabre-Madhin, E. 2005. “Getting Markets Right.� Price Instability in Low-Income Countries,� February 28 to Hazell, P., G. Shields, and D. Shields. 2005. “The Nature and March 1, 2005, Washington, DC. All papers are available at Extent of Domestic Sources of Food Price Instability and Risk.� http://www.passlivelihoods.org.uk/default.asp?project_id Ibarra, H., U. Hess, J. Syroka, and A. Nucifora. 2005. “Use of =240&nc=4921. Weather Insurance Markets for Managing Food Supply Risk: Malawi Case Study.� 71 Appendix 2 Detailed Data for the Macro-Typology of Countries This appendix provides empirical background on the basis for grouping 25 developing countries according to their vulnerability to global and domestic food price shocks (see tables 2.2 and 2.3). Table A.1 Country-speci�c Variables Used to Development the Country Typology Average Annual Diversity Index Food Aid as a Cereal Imports of Food Share of Cereal to Foreign Consumption, Utilization, Reserves, Country (Si )2 (2002) 1999–2003 1994–2003 Bangladesh 0.77 1.5 29.7 Burkina Faso 0.55 1.5 25.6 Cambodia 0.81 1.0 4.6 Cameroon 0.18 0.4 53.8 Chile 0.54 0 2.2 Côte d’Ivoire 0.23 0.8 26.1 Egypt, Arab Repub. of 0.36 0.1 11.8 Ethiopia 0.22 13.0 36.2 Ghana 0.21 3.0 23.1 India 0.40 0.1 0.3 Indonesia 0.54 0.5 7.9 Kenya 0.43 9.1 26.9 Madagascar 0.47 1.1 21.2 Malawi 0.48 6.8 36.1 Mali 0.35 1.2 8.1 Mexico 0.58 0 6.6 Morocco 0.51 0.8 12.7 Mozambique 0.33 10.3 16.6 Nepal 0.33 0.6 2.2 Niger 0.80 1.5 65.6 Nigeria 0.25 0.0 9.6 Pakistan 0.61 0.6 11.0 Senegal 0.34 1.4 63.8 South Africa 0.42 0.3 8.9 Sudan 0.52 5.7 98.6 Tanzania 0.30 3.1 15.5 Uganda 0.22 6.6 8.6 Vietnam 0.82 0.2 0 Yemen, Repub. of 0.49 4.7 19.0 Zambia 0.56 9.5 26.7 Zimbabwe 0.51 9.9 30.6 73 74 Managing Food Price Risks and Instability in an Environment of Market Liberalization Table A.2 Commodity-speci�c Variables Used to Develop the Country Typology Cuddy-Della Valle Number of Years Net Imports (Exports) of Net Imports of Food Index of in 10 as Net Dominant Staple as Staple as a Percentage Production Country Dominant Staple Importer (Exporter) Percentage of Utilization of Total Exports Variability Bangladesh Rice 10 2.5 3.1 5.0 Burkina Faso Millet/Sorghum 3 –0.1 –0.0 12.0 Cambodia Rice 10 1.5 0.6 5.0 Cameroon Cassava 2 –0.0 0 2.0 Maize 10 0.9 0.0 6.0 Chile Wheat 10 24.5 0.4 11.1 Côte d’Ivoire Rice 10 32.0 2.4 19.1 Egypt, Arab Repub. of Wheat 10 47.4 3.2 2.2 Maize 10 44.6 5.5 8.9 Ethiopia Maize 9 0.7 0.3 12.6 Wheat 10 39.2 15.7 8.5 Ghana Cassava –10 –0.2 –0.0 4.3 Maize 5 –0.1 0.0 11.1 India Rice –10 –2.4 –1.7 7.0 Wheat 3 –1.8 –0.2 5.4 Indonesia Rice 10 3.7 0.9 1.6 Kenya Maize 7 9.5 1.6 8.9 Madagascar Rice 10 4.0 2.7 2.7 Malawi Maize 9 6.9 6.5 21.6 Mali Millet/Sorghum 3 0.3 –0.1 18.7 Mexico Maize 10 20.1 0.5 3.7 Morocco Wheat 10 41.8 3.9 46.3 Mozambique Cassava 1 0.0 0.0 4.9 Maize 10 16.4 6.4 11.1 Nepal Rice 10 1.3 0.9 2.9 Niger Millet/Sorghum 10 0.5 0.3 14.2 Nigeria Millet/Sorghum 1 –0.1 –0.0 3.0 Pakistan Wheat 7 6.6 2.1 5.5 Senegal Rice 10 75.8 9.1 16.7 South Africa Maize 1 –10.8 –0.4 20.3 Sudan Millet/Sorghum 3 –1.8 –0.4 24.6 Tanzania Maize 7 1.8 1.2 11.2 Uganda Cassava 2 0 –0.2 13.7 Maize 5 –0.2 0 8.2 Plantains 0 0 0 1.9 Vietnam Rice –10 –12.5 –5.4 2.3 Yemen, Repub. of Wheat 10 93.0 7.5 10.6 Zambia Maize 8 10.5 2.4 30.6 Zimbabwe Maize 3 –8.7 –0.1 40.9 Notes next group of households, which sells the remaining half of CHAPTER 1 all maize. 1. See, for example, Newberry and Stiglitz (1981); Myers (1988, 1992b); and Williams and Wright (1991). 2. See Tobey, Reilly, and Kane (1992); Reilly and Hohmann CHAPTER 3 (1993); Antle (1995); Winters and others (1998); and Mendelsohn and Dinar (1999). 14. These results can be attributed to the fact that the standard 3. See Bates (1981); Toye (1992); Sahn, Dorosh, and Younger deviation of fluctuations around trend is declining at about (1997); Jayne and others (2002); McPherson (2002); and the same pace as the mean, or even faster than the mean in Garcia Garcia (2004). the case of rice (see table 3.1). 4. See World Bank (1994, 2000); Sahn, Dorosh, and Younger 15. These statistics reflect the variability of prices in the 1970s. (1997); Kherallah and others (2002); and Dorward and others For the period 1981–2003, the respective CVs for rice, wheat, (2004). and yellow maize are 25, 22, and 19 percent. 5. See Barrett (1997, 1999); Reardon and others (1999); World 16. These price series are based on FAO data and reflect national Bank (2000); Sachs (2001); Jayne and others (2002); and average prices. In some cases, they may be of�cial producer Dorward and others (2004). prices, although in practice many producers receive prices 6. In eastern and southern Africa, for example, Zambia, Kenya, that are signi�cantly higher or lower than of�cial prices. For Zimbabwe, and Malawi have either retained or reinstituted example, Pakistan does not collect farm harvest price data, parastatal marketing boards to stabilize prices, hold buffer although informal observation suggests that farm harvest stocks, and achieve income transfer objectives (Jayne, Tembo, prices in some years depart sharply from the minimum sup- and Nijhoff 2005). Likewise, after initial steps toward liberal- port price. ization, Pakistan once again banned the movement of wheat 17. White maize prices are likely to be at least slightly more across provinces, and the public sector competed aggres- variable than yellow maize prices. sively with the private sector to procure wheat in 2004. 7. Papers from the workshop, held in Washington, D.C., from February 28 to March 1, 2005, are listed in appendix 1 and are available at http://www.passlivelihoods.org.uk/ CHAPTER 4 default.asp?project_id=240&nc=4921. 18. For example, Myers (1988); Braverman and others (1990); Islam and Thomas (1994); and Finkelshtain and Chalfant (1997). CHAPTER 2 19. In economies with incomplete markets, and where food pro- 8. Low-income countries are defined as having a GNP per duction represents a major share of GDP, food price instabil- capita of US$765 or less; lower-middle-income counties ity can cause savings to be held in less liquid form, thereby as US$766–3,305; and upper-middle-income countries as reducing the funds available for capital investment. Second, US$3,036–9,385 (World Bank 2004). All low-income countries uncertainty about future price levels can increase the value with a population of more than 10 million were included, ex- of waiting to invest, which in turn leads to a reduction in the cept those with serious internal conflicts that would distort rate of investment. Third, in the presence of food price in- production and trade trends. Morocco, Egypt, Indonesia, and stability, producers and investors may have a dif�cult time South Africa were the lower-middle-income countries, and separating “permanent� from “temporary� shifts in invest- Mexico and Chile the upper-middle-income countries, in- ment returns, and this dif�culty may lead to an inef�cient cluded in the sample. allocation of investment funds. 9. These data capture formal trade, although of course there is often considerable cross-border informal trade. 10. Results for an alternative measure, the percentage of total CHAPTER 5 export revenues spent on imports of the food staple, are re- ported in appendix 1. 20. See Bates (1981); Mosley, Harrigan, and Toye (1991); Bates 11. Erratic policies have also probably encouraged variability in and Krueger (1993); and Van de Walle (2001). maize production in these countries (see chapter 5). 21. See Government of Zambia (1995); Rubey (1995); Jayne and 12. Concentration of consumption is measured with the others (1999); and Balat and Porto (2005). Liberalization also Her�ndahl-Hirschman Index (see notes to table 2.2). bene�ted urban consumers in Tanzania, but for somewhat 13. In Zambia these households enjoy income levels and asset different reasons. Here, the volume of subsidized grain distri- holdings that are three to four times greater than those of the buted through the state marketing system was insuf�cient 75 76 Managing Food Price Risks and Instability in an Environment of Market Liberalization to meet demand. As subsidies became increasingly irrele- 26. However, the next year, the government undermined this vant to most consumers, who lacked access to the rationed reform by aggressively intervening in markets and restrict- grain, they increasingly depended on parallel markets, even ing grain movement (box 8.1). before markets were officially liberalized (Bryceson 1993). 27. A notable exception is Mali, where price information fol- Liberalization reduced food prices by removing some of the lowing the 2004–5 drought has been used extensively to policy-related barriers on private trade. guide cereal import decisions by the government and pri- 22. See van de Walle (2001); Jayne and others (2002); McPherson vate sector. See Statz (2005). (2002); Bird, Booth, and Pratt (2003); and Avalos-Sartorio 28. Market information systems for countries within a region (2005). also must be linked through ef�cient means of communica- 23. Summarized from Avalos-Sartorio (2005). tion so that information available in one country is immedi- ately available in all countries of the region. CHAPTER 6 24. In some of the least-developed countries, the lack of foreign CHAPTER 7 exchange remains a potentially critical national constraint 29. For a full discussion of variable levies and tariffs within on using food imports to meet shortfalls from a severe nat- WTO rules, see Foster and Valdes (2005). ural disaster. 30. Even seasonal price movements can become more extreme 25. For importing and exporting countries, standard measures when reserves operate. Mozambique, with no food reserve of border parity can be used. However, for countries that are and no restrictions on maize trade, shows a typical seasonal approximately self-suf�cient, market equilibrium prices may price rise for maize at retail of about 50 percent in its de�cit fluctuate within a wide band between import and export southern region. Malawi, on the other hand, frequently holds parity, and more complex supply-demand analysis is needed a large reserve and intervenes in other ways in the market, to compute equilibrium prices [see Byerlee and Morris (1993); but shows the highest seasonal price movement, averaging Mullen and others (2004)]. 90 percent over the last decade [Tschirley and others (2004)]. References Alderman, H., and T. Haque. 2005. “Countercyclical Safety Bates, R. 1981. Markets and States in Tropical Africa: The Political Nets for the Poor and Vulnerable.� Paper presented to the Basis of Agricultural Policies. Berkeley: University of California workshop “Managing Food Price Instability in Low-Income Press. Countries,� February 28 to March 1, 2005, Washington, DC. Bates, R., and A. Krueger, eds. 1993. Political and Economic Inter- Alston, J. M., C. Chan-Kang, M. C. Marra, P. G. Pardey, and T. J. actions in Economic Policy Reform: Evidence from Eight Countries. Wyatt. 2000. A Meta-analysis of Rates of Return to Agricultural Oxford: Basil Blackwell. R&D. IFPRI Research Report No. 113. International Food Bidarkota, P., and M. J. Crucini. 2000. “Commodity Prices and the Policy Research Institute, Washington, DC. Terms of Trade.� Review of International Economics 8(4): 647–66. Amha, W., J. Stepanek, T. Jayne, and A. Negassa. 1996. “Meeting Bird, K., D. Booth, and N. Pratt. 2003. Food Security Crisis in Food Aid and Price Stabilization Objectives through Local Southern Africa: The Political Background to Policy Failure. Forum Grain Purchase: A Review of the 1996 Experience.� Working for Food Security in Southern Africa, London: Overseas Dev- Paper 7. Grain Market Research Project, Addis Ababa, Ethiopia. elopment Institute. http://www.aec.msu.edu/agecon/fs2/ethiopia/wp7.pdf. Braverman, A., R. Kanbur, A. S. P. Brandao, J. Hammer, M. de Anderson, J. R. 2001. “Risk Management in Rural Development: Rezende Lopes, and A. Tan. 1990. “The Costs and Bene�ts of A Review.� Rural Development Strategy Background Paper 7. Agricultural Price Stabilization in Brazil.� WPS 564, World World Bank, Washington DC. Bank, Agriculture and Rural Development Department, Anderson, K., and R. Tyers. 1993. “More on Welfare Gains to Washington, DC. Developing Countries from Liberalizing World Food Trade.� Brunetti, A., G. Kisunko, and B. Weder. 1997. “Institutional Journal of Agricultural Economics 44(2): 189–204. Obstacles to Doing Business: Region-by-Region Results from Antle, J. M. 1983. “Infrastructure and Aggregate Agricultural a Worldwide Survey of the Private Sector.� Policy Research Productivity: International Evidence.� Economic Development Working Paper 1759. World Bank, Washington, DC. and Cultural Change 31(2): 609–20. Bryceson, D. F. 1993. “Urban Bias Revisited: Staple Food Pricing ———. 1995. “Climate Change and Agriculture in Developing in Tanzania.� In C. Hewitt de Alcantara, ed., Real Markets: Social Countries.� American Journal of Agricultural Economics 77(3): and Political Issues of Food Policy Reform. London: Frank Cass. 741–6. Buccola, S., and C. Sukume. 1988. “Optimal Grain Pricing and Avalos-Sartorio, B. 2005. “What Can We Learn From Past Price Storage Policy in Controlled Agricultural Economies: Appli- Stabilization Policies and Market Reform in Mexico?� Paper cation to Zimbabwe.� World Development 16(3): 361–71. presented to the workshop “Managing Food Price Instability Byerlee, D., X. Diao, and C. Jackson. 2005. Agriculture, Rural in Low-Income Countries,� February 28 to March 1, 2005, Development and Pro-poor Growth: Country Experiences in the Washington, DC. Post-Reform Era. Washington, DC: World Bank. Badiane, O., and D. Resnick. 2005. “Regional Trade Liberalization Byerlee, D., and C. K. Eicher, eds. 1997. The Emerging Maize-based and Domestic Food Market Stabilization in African Coun- Revolution in Africa: The Role of Technologies, Institutions and tries.� Paper presented to the workshop “Managing Food Policies. Boulder: Lynne Rienner. Price Instability in Low-Income Countries,� February 28 to Byerlee, D., and M. Morris. 1993. “Calculating Levels of Pro- March 1, 2005, Washington, DC. tection: Is It Always Appropriate to Use World Reference Balat, J., and G. Porto. 2005. “Globalization and Complementary Prices Based on Current Trading Patterns?� World Development Policies: Poverty Impacts in Rural Zambia.� World Bank, 21(5): 805–15. Development Research Group, Washington, DC. Byerlee, D., and G. Sain. 1986. “Food Pricing Policy in Developing Barrett, C. 1997. “Liberalization and Food Price Distributions: Countries: Bias Against Agriculture or for Urban Consumers?� ARCH-M Evidence From Madagascar.� Food Policy 22(2): American Journal of Agricultural Economics, 68(4): 961–9. 155–73. Cashin, P., H. Liang, and J. C. McDermott. 1999. “How Persistent ———. 1999. “The Effects of Real Exchange Rate Depreciation are Shocks to World Commodity Prices?� Working Paper on Stochastic Producer Prices in Low-Income Agriculture.� 99/80. International Monetary Fund, Washington, DC. Agricultural Economics 20(3): 215–30. Choudhury, M. A. 1995. “An Empirical Analysis of Global Barrett, C., and D. Clay. 2003. “Self-Targeting Accuracy in the Commodity Price Fluctuations with Special Reference to Presence of Imperfect Factor Markets: Evidence from Food- Developing Countries.� In O. Suliman and M. El-Baghdadi, for-Work in Ethiopia.� Journal of Development Studies 39(5): eds., Global Commodity Price Stabilization: Implications for 152–80. World Trade and Development. Westport, Conn. and London: Barrett, C., and D. Maxwell. 2005. Food Aid After Fifty Years: Greenwood, Quorum Books, 97–123. Recasting Its Role. London: Routledge. Coulter, J. 2005. “Making the Transition to a Market-Based Grain Barro, R. J., and D. B. Gordon. 1983. “Rules, Discretion, and Marketing System.� Paper presented to the workshop “Man- Reputation in a Model of Monetary Policy.� Journal of Monetary aging Food Price Instability in Low-Income Countries,� Economics 12(1): 101–22. February 28 to March 1, 2005, Washington, DC. 77 78 Managing Food Price Risks and Instability in an Environment of Market Liberalization Coulter, J., and G. Onumah. 2002. “The Role of Warehouse in Low-Income Countries,� February 28 to March 1, 2005, Receipt Systems in Enhanced Commodity Marketing and Washington, DC. Rural Livelihoods in Africa.� Food Policy 27: 319–37. Faruqee, R., J. R. Coleman, and T. Scott. 1997. “Managing Price Cuddy, J., and P. A. Della Valle. 1978. Measuring the Instability Risk in the Pakistan Wheat Market.� The World Bank Economic of Time Series Data. Oxford Bulletin of Economics and Statistics Review 11(2): 263–92. 40(1): 79–84. FAS-USDA (Foreign Agriculture Service of the U.S. Department Cummings, R. W., Jr., S. Rashid, and A. Gulati. 2005. “Grain Price of Agriculture). 2004. Grain: World Markets and Trade. Foreign Stabilization Experiences in Asia: What Have We Learned?� Agricultural Service Circular Series, FG 05-04, May. http:// Paper presented to the workshop “Managing Food Price In- www.fas.usda.gov/grain/circular/2004/05- stability in Low-Income Countries,� February 28 to March 1, 04/graintoc.htm. 2005, Washington, DC. Finkelshtain, I., and J. A. Chalfant. 1997. “Commodity Price Dana, J., C. L. Gilbert, and E. Shim. 2005. “Hedging Grain Price Stabilization in a Peasant Economy.� American Journal of Agri- Risk in the SADC: Case Studies of Malawi and Zambia.� cultural Economics 79(4): 1208–17. Paper presented to the workshop “Managing Food Price In- Foster, W., and A. Valdes. 2005. “The Merits of a Special stability in Low-Income Countries,� February 28 to March 1, Safeguard, Price Floor Mechanism under Doha for Developing 2005, Washington, DC. Countries.� Paper presented to the workshop “Managing Dawe, D. 2001. “How Far Down the Path to Free Trade? The Food Price Instability in Low-Income Countries,� February 28 Importance of Rice Price Stabilization in Developing Asia.� to March 1, 2005, Washington, DC. Food Policy 26(2): 163–75. Gabre-Madhin, E. 2001. Market Institutions, Transaction Costs, and Dawe, D., and Timmer, C. P. 2005. “Managing Food Price Social Capital in the Ethiopian Grain Market. IFPRI Research Instability in Asia: A Macro Food Security Perspective.� Report 124. International Food Policy Research Institute, Paper presented to the workshop “Managing Food Price In- Washington, DC. stability in Low-Income Countries,� February 28 to March 1, Gabre-Madhin, E. 2005. “Getting Markets Right.� Paper pre- 2005, Washington, DC. sented to the workshop “Managing Food Price Instability Deaton, A. S. 1992. “Commodity Prices, Stabilization, and Growth in Low-Income Countries,� February 28 to March 1, 2005, in Africa.� Research Program for Development Studies Dis- Washington, DC. cussion Paper 166. Princeton University, N.J. Gabre-Madhin, E. Z., C. B. Barrett, and P. A. Dorosh. 2003. Deaton, A. S., and R. Miller. 1996. “International Commodity Technological Change and Price Effects in Agriculture : Conceptual Prices, Macroeconomic Performance and Politics in Sub- and Comparative Perspectives. Discussion Paper, International Saharan Africa.� Journal of African Economies 5(3): 99–191. Food Policy Research Institute (IFPRI), Washington, DC. del Ninno, C., and P. Dorosh. 2003. “Public Policy, Markets, and Garcia Garcia, J. 2004. “BULOG Interventions in Commodity Household Coping Strategies in Bangladesh: Avoiding a Food Markets, Income Transfers, and Rent-Seeking: 1994/ Security Crisis Following the 1998 Floods.� World Development 95–1997/98.� Draft paper, World Bank, Washington, DC. 31(7): 1221–38. Gardner, B. 1989. “Rollover Hedging and Missing Long-Term del Ninno, C., P. Dorosh, and K. Subbarao. 2005. Food Aid and Futures Markets.� American Journal of Agricultural Economics Food Security in the Short and Long Run: Country Experience 71(3): 311–18. from Asia and Sub-Saharan Africa. Draft paper, World Bank, Govereh, J., T. Jayne, and J. Shaffer. 2005. Trends in Zambian Washington, DC. Agriculture, 1990–2003. Working Paper 14. Food Security Dorosh, P. 2001. “Trade Liberalization and National Food Research Project, Lusaka, Zambia. Security: Rice Trade between Bangladesh and India.� World Government of Zambia. 1995. A Review of Maize Marketing Development 29(4): 673–89. Liberalization During 1994: The Transition Programme 1995/96. Dorward, A., and J. Kydd, 2002. “The Malawi 2000 Food Crisis: Lusaka: Food Security Division, Ministry of Agriculture, The Rural Development Challenge.� Presented at “Malawi Food and Fisheries, Lusaka. after Banda: Perspectives in a Regional African Context.� Gurenko, E., and R. Lester. 2004. “Rapid Onset Natural Disasters: Centre for Commonwealth Studies, University of Stirling. The Role of Financing in Effective Risk Management.� World Dorward, A., J. Kydd, J. Morrisson, and I. Urey. 2004. “A Policy Bank Policy Research Working Paper 3278, Washington, DC. Agenda for Pro-Poor Agricultural Growth.� World Development Gutner, T., 2002. “The Political Economy of Food Subsidy 32(1): 73–89. Reform: The Case of Egypt.� Food Policy 27(5): 455–76. Duncan, J., and R. J. Myers. 2000. “Crop Insurance under Hazell, P., G. Shields, and D. Shields. 2005. “The Nature and Catastrophic Risk.� American Journal of Agricultural Economics Extent of Domestic Sources of Food Price Instability and 82(4): 842–55. Risk.� Paper presented to the workshop “Managing Food Ersado, L., H. Alderman, and J. Alwang. 2003. “Changes in Price Instability in Low-Income Countries,� February 28 to Consumption and Saving Behavior before and after Economic March 1, 2005, Washington, DC. Shocks: Evidence from Zimbabwe.� Economic Development and Hicks, N., and Q. Wodon. 2001. “Social Protection for the Poor Cultural Change 52(1): 187–215. in Latin America.� CEPAL Review 73: 93–113. Evenson, R. E. 2001. “Economic Impacts of Agricultural Re- Hulme, J., ed. 1996. Climate Change and Southern Africa. An Ex- search and Extension.� In B. L. Gardner and G. C. Rausser, ploration of Some Potential Impacts and Implications in the SADC eds., Handbook of Agricultural Economics. Amsterdam: North Region. Norwich: Climatic Research Unit, School of Environ- Holland/Elsevier. mental Sciences, University of East Anglia. Fafchamps, M. 1992. “Cash Crop Production, Food Price Ibarra, H., U. Hess, J. Syroka, and A. Nucifora. 2005. “Use of Volatility, and Rural Market Integration in the Third World.� Weather Insurance Markets for Managing Food Supply Risk: American Journal of Agricultural Economics 74(1): 90–9. Malawi Case Study.� Paper presented to the workshop FAO (Food and Agriculture Organization). 2005. “Background “Managing Food Price Instability in Low-Income Countries,� Document.� Prepared for the Special Event on Impact of February 28 to March 1, 2005, Washington, DC. Climate Change, Pests and Diseases on Food Security and Innes, R. D. 2003. “Crop Insurance in a Political Economy: An Poverty Reduction, 31st Session of the Committee on World Alternative Perspective on Agricultural Policy.� American Food Security, May 23 to 26, 2005, Rome. Journal of Agricultural Economics 85(2): 318–35. Faruqee, R. 2005. “Reforming Wheat Policy in Pakistan.� Paper Islam, N., and S. Thomas. 1994. “Food Price Stabilization Policies presented to the workshop “Managing Food Price Instability in Developing Countries: Analytical Issues and Empirical References 79 Evidence.� In A. Sorkin and I. Sirageldin, eds., Research in Meerman, J. P. 1997. “Reforming Agriculture: The World Bank Human Capital and Development. Volume 8: Nutrition, Food Goes to Market.� Operations Evaluation Study, World Bank, Policy and Development, Greenwich, Conn. and London: JAI Washington, DC. Press: 57–130. Mendelsohn, R., and A. Dinar. 1999. “Climate Change, Agri- Jayne, T. S., J. Govereh, A. Mwanaumo, J. Nyoro, and A. Chapoto. culture, and Developing Countries: Does Adaptation Matter?� 2002. “False Promise or False Premise: The Experience of World Bank Research Observer 14(2): 277–93. Food and Input Market Reform in Eastern and Southern Mitchell, D., and J. Le Vallee. 2005. “International Food Price Africa.� World Development 30(11): 1967–86. Variability: The Implications of Recent Policy Changes.� Jayne, T. S., and S. Jones. 1997. “Food Marketing and Pricing Paper presented to the workshop “Managing Food Price Policy in Eastern and Southern Africa: A Survey.� World Instability in Low-Income Countries,� February 28 to March Development 25(9): 1505–27. 1, 2005, Washington, DC. Jayne, T. S., M. Mukumbu, M. Chisvo, D. Tschirley, B. Zulu, M. Morduch, J. 1995. “Income Smoothing and Consumption Weber, R. Johansson, P. Santos, and D. Soroko. 1999. “Successes Smoothing.� Journal of Economic Perspectives 9(3): 103–14. and Challenges of Food Market Reform: Experiences from Morris, J. 2005. “Can Insurance Break Ethiopia’s Vicious Cycle Kenya, Mozambique, Zambia, and Zimbabwe.� International of Hunger?� Financial Times, May 10. Development Working Paper 72, East Lansing, Michigan State Mosley, P., J. Harrigan, and J. Toye. 1991. Aid and Power: The World University. Bank and Policy Based Lending in the 1980s. London: Routledge. Jayne, T. S., and M. Rukuni. 1993. “Distributional Effects of Mullen, K., D. Sun, D. Orden, and A. Gulati. 2004. “Producer Maize Self-Suf�ciency in Zimbabwe: Implications for Pricing Support Estimates (PSEs) for Agriculture in Developing and Trade Policy.� Food Policy 18(4): 334–41. Countries: Measurement Issues and Illustrations from India Jayne, T. S., G. Tembo, and J. Nijhoff. 2005. “Experiences of Food and China.� MTID Discussion Paper 74, International Food Market Reform and Price Stabilization in Eastern and Southern Policy Research Institute, Washington, DC. Africa.� Paper presented to the workshop “Managing Food Myers, R. J. 1988. “The Value of Ideal Contingency Markets in Price Instability in Low-Income Countries,� February 28 to Agriculture.� American Journal of Agricultural Economics 70(2): March 1, 2005, Washington, DC. 255–67. Joffe, S., and S. Jones. 2004. Stimulating Private Investment and ———. 1992a. “Intervention Bias in Agricultural Policy.� Agri- Market Development for Agriculture: New Approaches and cultural Economics 7(3–4): 209–24. Experience. Oxford, U.K.: Oxford Policy and Management. ———. 1992b. “Incomplete Markets and Commodity-linked Jones, S. P. 1994. Privatization and Policy Reform: Agricultural Finance in Developing Countries.� World Bank Research Marketing in Africa. Oxford, U.K.: Food Studies Group, Observer 7(1): 79–94. University of Oxford. ———. 2005. “Costs of Food Price Instability in Low-Income Just, R. E. 1985. Automatic Adjustment Rules in Commodity Countries.� Paper presented to the workshop “Managing Programs. In B. L. Gardner, ed., U.S. Agricultural Policy: The Food Price Instability in Low-Income Countries,� February 1985 Farm Legislation. Washington, DC: American Enterprise 28 to March 1, 2005, Washington, DC. Institute for Public Policy Research. Myers, R. J., and S. R. Thompson. 1989. “Optimal Portfolios of Just R., E. Lutz, A. Schmitz, and S. Turnovsky. 1978. “The External Debt in Developing Countries: The Potential Role of Distribution of Welfare Gains from Price Stabilization: An Commodity-Linked Bonds.� American Journal of Agricultural International Perspective.� Journal of International Economics Economics 71(2): 517–22. 8(4): 551–63. Muyanga, M., T. S. Jayne, G. Argwings-Kodhek, and J. Ariga. Kannapiran, C. 2000. “Commodity Price Stabilization: Macro- 2005. “Staple Food Consumption Patterns in Urban Kenya: economic Impacts and Policy Options.� Agricultural Econo- Trends and Policy Implications.� Working Paper, Tegemeo mics 23(1): 17–30. Institute of Agricultural Policy and Development, Egerton Kherallah, M., C. Delgado, E. Gabre-Madhin, N. Minot, and M. University, Kenya. Available online at www.tegemeo.org. Johnson. 2002. Reforming Agricultural Markets in Africa. Naylor, R., W. Falcon, and E. Zavaleta. 1997. “Variability and Baltimore: Johns Hopkins University Press. Growth in Grain Yields 1950–1994: Does the Record Point to Lacroix, R., and P. Varangis. 1996. “Using Warehouse Receipts in Greater Instability?� Population and Development Review 23(1): Developing and Transition Economies.� Finance and Develop- 41–58. ment, September: 36–9. NEPAD (New Partnership for Africa’s Development). 2004. Lai, J-Y, R. J. Myers, and S. D. Hanson. 2003. “Optimal On-Farm “Study to Explore Further Options for Food-Security Reserve Grain Storage by Risk-Averse Farmers.� Journal of Agricul- Systems in Africa.� Draft paper, NEPAD, Pretoria, South Africa. tural and Resource Economics 28(3): 558–79. Newberry, D., and J. Stiglitz. 1981. The Theory of Commodity Price Larson, D. F., J. R. Anderson, and P. Varangis. 2004. Policies Stabilization: A Study in the Economics of Risk. Oxford: Clarendon Regarding Risk in Agricultural Markets. Draft paper, World Press. Bank, Washington, DC. Nicita, A. 2005. “Doha Implications for Poverty Reduction in Lence, S. H., and M. L. Hayenga. 2001. “On the Pitfalls of Multi- Ethiopia.� Draft paper, Development Economics Department, year Rollover Hedges: The Case of Hedge-to-Arrive Con- World Bank, Washington, DC. tracts.� American Journal of Agricultural Economics 83(1): 107–19. Nijhoff, J. J., T. S. Jayne, B. Mwiinga, and J. Shaffer. 2003. Leon, J., and R. Soto. 1997. “Structural Breaks and Long-Run “Markets Need Predictable Government Actions to Function Trends in Commodity Prices.� Journal of International Develop- Effectively: The Case of Importing Maize in Times of De�cit.� ment 9(3): 347–66. Zambia FSRP Policy Synthesis 6, Michigan State University, MacBean, A. I. 1966. Export Instability and Economic Development. East Lansing. Cambridge, Mass.: Harvard University Press. Nucifora, A., and A. Lisulo. 2005. “Lessons in Managing Policy Magnay, J. 2004. “The Ugandan Maize Industry.� Presented at Dialogue in Malawi: Reforming the Agricultural Development the NEPAD/IGAD Regional Conference on Agricultural and Marketing Corporation.� Paper presented to the work- Successes in the Greater Horn of Africa, Nairobi, Kenya. shop “Managing Food Price Instability in Low-Income Coun- McPherson, M., ed. 2002. Restarting and Sustaining Growth and tries,� February 28 to March 1, 2005, Washington, DC. Development in Africa. Equity and Growth through Economic Oehmke, J. F., and E. W. Crawford. 1996. “The Impact of Research, Harvard Institute for International Development, Agricultural Technology in Sub-Saharan Africa.� Journal of Cambridge. African Economies 5(2): 271–92. 80 Managing Food Price Risks and Instability in an Environment of Market Liberalization O’Hara, M. 1984. “Commodity Bonds and Consumption Risks.� Schiff, M., and A. Valdes. 1992. The Plundering of Agriculture in Journal of Finance 39(3): 193–206. Developing Countries. Washington, DC: World Bank. Pinckney, T. C., and A. Valdes. 1988. “Short-Run Supply Manage- Sen, A. 1981. Poverty and Famines: An Essay on Entitlement and ment and Food Security: Results from Pakistan and Kenya.� Deprivation. Oxford: Clarendon Press. World Development 16(9): 1025–34. Singh, A. J., and D. Byerlee. 1990. “Relative Variability in Wheat Poulton, C., J. Kydd, S. Wiggins, and A. Dorward. 2005. “State Yields Across Countries and Over Time.� Journal of Agri- Intervention for Food Price Stabilisation in Africa: Can it cultural Economics 41(1): 21–32. Work?� Paper presented to the workshop “Managing Food Staatz, J. 2005. “Building Long-Term Food Security while Price Instability in Low-Income Countries,� February 28 to Managing Food Crises: Insights from Mali,� Presentation March 1, 2005, Washington, DC. to USAID/AFR, July 28. www.aec.msu.edu/agecon/fs2/ Priovolos, T., and R. Duncan, eds. 1991. Commodity Risk Manage- mali_fd_strtgy/index.htm. ment and Finance. Oxford: Oxford University Press. Taylor, D., and D. Byerlee. 1991. Food Aid and Food Security: A Ramey, G., and V. A. Ramey. 1995. “Cross Country Evidence on Cautionary Note. Canadian Journal of Agricultural Economics the Link Between Volatility and Growth.� American Economic 39: 163–75. Review 85(5): 1138–51. Timmer, C. P. 1995. “Getting Agriculture Moving: Do Markets Rashid, S., R. Cummings, and A. Gulati. 2005. “Grain Marketing Provide the Right Signals?� Food Policy 20(5): 455–72. Parastatals in Asia: Why Do They Have to Change Now?� ———. 2002. “Agriculture and Economic Development.� In Discussion Paper 80, International Food Policy Research Gardener, B. and G. Rausser, eds., Handbook of Agricultural Institute, Washington, DC. Economics, Vol. 2. Amsterdam: Elsevier Science. 1487–1546. Reardon, T., C. Barrett, V. Kelly, and K. Savadogo. 1999. “Policy Tobey, J., J. Reilly, and S. Kane. 1992. “Economic Implications of Reforms and Sustainable Agricultural Intensification in Global Climate Change for World Agriculture.� Journal of Africa.� Development Policy Review 17(4): 375–95. Agricultural and Resource Economics 17(1): 195–204. Reilly, J., and N. Hohmann. 1993. “Climate Change and Agri- Townsend, R. M. 1995. “Consumption Insurance: An Evaluation culture: The Role of International Trade.� American Economic of Risk-Bearing Systems in Low-Income Economies.� Journal Review 83(2): 306–12. of Economic Perspectives 9(3): 83–102. Revkin, A. C. 2005. “Ocean Warmth Tied to African Drought,� Toye, J. 1992. “Interest Group Politics and the Implementation of New York Times, May 24. Adjustment in Sub-Saharan Africa.� Journal of International Rozenzweig, M. R., and K. I. Wolpin. 1993. “Credit Market Development 4(2): 183–98. Constraints, Consumption Smoothing, and the Accumulation Traub, L. N., and T. S. Jayne. 2004. “The Effects of Market of Durable Production Assets in Low-Income Countries: Reform on Maize Marketing Margins in South Africa.� MSU Investments in Bullocks in India.� Journal of Political Economy International Development Working Paper 83, Michigan 101(2): 223–44. State University, East Lansing. Rubey, L. 1995. Maize Market Reform in Zimbabwe: Linkages be- Tschirley, D., J. Nijhoff, P. Arlindo, B. Mwinga, M. Weber, and T. tween Consumer Preferences, Small Enterprise Development Jayne. 2004. “Anticipating and Responding to Drought Emer- and Alternative Market Channels, Ph.D. diss., Michigan State gencies in Southern Africa: Lessons from the 2002–2003 Ex- University, East Lansing. perience.� Prepared for the New Partnership for Africa’s ———. 2005. Malawi’s Food Crisis: Causes and Solutions. Re- Development (NEPAD) Conference on Successes in African port for U.S. Agency for International Development (USAID), Agriculture, November 22–25. 2004, Nairobi, Kenya (www. Lilongwe, Malawi. aec.msu.edu/agecon/maizemarket/index.htm). Ryan, J. G. 1999. “Assessing the Impact of Rice Policy Changes van de Walle, N. 2001. African Economies and the Politics of in Vietnam and the Contribution of Policy Research.� Impact Permanent Crisis, 1979–1999. Cambridge: Cambridge Univ- Assessment Discussion Paper 8. International Food Policy ersity Press. Research Institute, Washington, DC. Whiteside, M. 2003. “Enhancing the Role of Informal Maize Sachs, J. 2001. “New Approaches to International Donor Assis- Imports in Malawi Food Security.� Consultancy report for tance. Presentation to International Development Research the Department for International Development (DFID), Center, Ottawa, Canada.� ECAPAPA Electronic Newsletter London. 4.37 (www.asareca.org/ecapapa). Williams, J. C., and B. D. Wright. 1991. Storage and Commodity Sahn, D., P. Dorosh, and S. Younger. 1997. Structural Adjustment Markets. Cambridge: Cambridge University Press. Reconsidered: Economic Policy and Poverty in Africa. Cambridge: Winters, P., R. Murgai, E. Sadoulet, A. de Janvry, and G. Frisvold. Cambridge University Press. 1998. “Economic and Welfare Impacts of Climate Change on Sarris, A. H. 1998. “The Evolving Nature of International Price Developing Countries.� Environmental and Resource Economics Instability in Cereal Markets.� FAO Commodity and Trade 12(1): 1–24. Division, Report 4, Commodity Policy and Projections Ser- World Bank. 1994. Adjustment in Africa: Reforms, Results and the vice of FAO, Rome. Road Ahead. New York: Oxford University Press. ———. 2000. “World Cereal Price Instability and a Market-Based ———. 2000. Can Africa Claim the Twenty-First Century? Instrument for LDC Food Import Risk Management.� Food Washington, DC: World Bank. Policy 25(2): 189–209. ———. 2004. World Development Report 2005: A Better Investment Sarris, A., P. Conforti, and A. Prakash. 2005. “The Use of Climate for Everyone. Washington, DC: World Bank. Organized Commodity Markets to Manage Food Import ———. 2005a. Agricultural Growth and the Poor: An Agenda for Price Instability and Risk.� Paper presented to the workshop Development. Directions in Development Report, World Bank, “Managing Food Price Instability in Low-Income Countries,� Washington, DC. February 28 to March 1, 2005, Washington, DC. ———. 2005b. Meeting Development Challenges: Renewed Approaches Scandizzo, P. L., P. B. R. Hazell, and J. R. Anderson. 1984. Risky to Promoting Rural Finance. Washington, DC: World Bank. Agricultural Markets: Price Forecasting and the Need for Inter- ———. Forthcoming. Ethiopia: Policies for Pro-Poor Agricultural vention Policies. Boulder: Westview Press. Growth. Washington, DC: World Bank.