Enhancing Control of Highly Pathogenic Avian Influenza in Developing Countries through Compensation: Issues and Good Practice Enhancing Control of Highly Pathogenic Avian Influenza in Developing Countries through Compensation Issues and Good Practice © 2006 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. The findings, interpretations, and conclusions expressed herein are those of the authors and do not necessarily reflect the views of the Board of Executive Directors of the World Bank, the governments they represent, or of any other organization. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of the World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Rights and Permissions The material in this work is copyrighted. Copying and/or transmitting portions or all of this work without permission may be a violation of applicable law. The World Bank encourages dissemination of its work and will normally grant permission promptly. For permission to photocopy or reprint any part of this work, please send a request with omplete information to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA, telephone 978-750-8400, fax 978-750-4470, www.copyright.com. All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, World Bank, 1818 H Street NW, Washington, DC 20433, USA, fax 202-522-2422, e-mail pubrights@worldbank.org. Cover Artwork: Detail from a painting by Gerardo Bravo Garcia Avian Flu Series, 2006. Oil & Gold Leaf on Canvas. Courtesy of the World Bank Art Program. Cover Design: Bill Pragluski, Critical Stages. iii Contents Introduction 1 Objectives of the Report and of Compensation 1 Background and Rationale for Assessing Good Practices 3 Context in Which Compensation Occurs 4 Approach to Ascertaining Good Practice 4 Deciding Who to Compensate: The Beneficiaries 7 Who Has Been Compensated in Control of Animal Diseases Other Than HPAI? 7 Who Is Involved in Poultry Production? 7 Ensuring That Eligible People Are Included 9 Establishing Losses 13 Direct, Consequential, and Indirect Losses 13 Types of Losses That Compensation Schemes Typically Take into Account 14 Losses Covered by Private Insurance or Public-Private Partnership in Developed Countries 16 Setting the Level and Timeliness of Compensation 19 Introduction 19 Approaches for Setting Compensation Rates 19 Who Defines the “Market” Price? 20 Choice of Price Baseline to Establish Compensation Rates 21 Establishing the Compensation Rate 23 Timeliness of the Payment 25 Promoting Awareness, Communications, and Capacity Building 27 Background 27 Experience to Date 28 Preparing the Communication Plan 29 ACkNOWLEDGMENTS v ACRONyMS AND ABBREvIATIONS vii ExECUTIvE SUMMARy ix 1 2 3 4 5 Enhancing Control of Highly Pathogenic Avian Influenza in Developing Countries through Compensationiv Preparing the Messages 29 Mechanisms and Channels of Communication on Compensation 29 Consistency of Message and Policy 30 Costs of the Awareness Campaign 30 Use of Nonpublic Sector Human Resources 30 Organizing Payment and Accountability 33 Introduction 33 Insights from Country Case Studies 34 Critical Success Factors 34 Who Pays? 34 Assessing Financial Needs 35 Payment Mechanisms and Timeliness 36 Certification 37 Use of Local Governments and Social Accountability 37 Core Components of Compensation 38 Good Practice Recommendations for Rapid Disbursements 38 Continuous Monitoring and Oversight 39 Where Are We and Where Do We Go from Here? 41 Compensation as an International Public Good 41 Characteristics of Good Practice in Compensation Schemes Designed to Help Control Disease Spread in an Emergency 42 Changes in the Structural Context under Which Compensation Is Occurring 44 The Future of Compensation in the Context of the Future of the Poultry Sector 45 Appendices 1. Examples of Compensation Schemes for Culling under Animal Disease Outbreaks in Selected Countries 49 2. Four Main Types of Poultry Production Systems 61 3. Avian Influenza Projects Approved under the GPAI 63 4. Sample Poultry Culling Record Owner’s Certificate 77 References 79 Endnotes 85 List of Figures Figure 2.1: Poultry System Continuum in 2006 with Country Examples 10 Figure 4.1: Farm-Gate Prices for Finished Broiler Chickens and Day-Old Chicks (DOC) from October 2005 to September 2006 in Egypt 22 Figure 4.2: Differences between Number of Commercial and Native Chickens in the Philippines 23 Figure 4.3: Differences in Prices between Commercial and Native Chickens in the Philippines 24 Figure 6.1: A Suggested Animal Health Fund Mechanism 38 6 7 v Acknowledgments This report was drafted by an interdisciplinary and interagency task team led by Christopher Delgado of the Agriculture and Ru- ral Development Department of the World Bank, involving staff and consultants from the World Bank, the Food and Agriculture Organization (FAO), and the International Food Policy Research Institute (IFPRI). The core drafting team consisted of Christopher Delgado, Patricia Mckenzie, and Cornelis de Haan (World Bank); Anni McLeod and Ana Riviere-Cinnamond (FAO); and Clare Nar- rod (IFPRI). Additional contributions were received from: Jimmy Smith, Samjhana Thapa, and Melissa Williams (World Bank); Jim Hancock, Dan vadnjal, and Jonathon Rushton (FAO); and Marites Tiongco and Devesh Roy (IFPRI). Funding was provided by the Operations Policy and Country Services and the Agricultural and Rural Development Departments of the World Bank, and by FAO, IFPRI, and the World Animal Health Organization (OIE). OIE kindly provided a venue for and supported a review meeting to discuss an earlier draft at its headquarters in Paris. Cordial thanks are ex- tended to the participants of that meeting and multiple meetings with colleagues on the World Bank’s Avian Influenza Regional Focal Points team in Washington. very helpful written reviews of an ear- lier draft were provided by Mario Bravo, Christiane Bruschke, Alain Dehove, Jacqueline Dubow, Simeon Ehui, Eric Marin, Piers Merrick, Stephen Mink, Laurent Msellati and klaus von Grebmer. Overall assistance with funding and guidance was received from Henning Steinfeld and Joseph Domenech (FAO); Bernard vallat (OIE); Max- imo Torero (IFPRI); and Sushma Ganguly, Olga Jonas, and Francois Le Gall (World Bank). Although every attempt was made to incor- porate fully all suggestions received, the core drafting team accepts responsibility for the views expressed in the report and resolution of multiple valuable suggestions for revision. vii Acronyms and Abbreviations AHF animal health fund APHIS Animal and Plant Health Inspection Service AoA Agreement on Agriculture BSE bovine spongiform encephalopathy CBPP Contagious Bovine Pleuropneumonia CSF Classical Swine Fever CTAs community thematic associations DOC day-old chicks EC European Commission ELGA Greek Agricultural Insurance Organisation EU European Union FAO Food and Agriculture Organization FMD Foot and Mouth Disease GDP gross domestic product GPAI Global Program for Avian Influenza HPAI highly pathogenic avian influenza IBRD International Bank for Reconstruction and Development IDA International Development Association IFPRI International Food Policy Research Institute LFAs less favored areas LPAI low pathogenic avian influenza NGO nongovernmental organization OECD Organisation for Economic Co-operation and Development OIE World Animal Health Organization PCR polymerase chain reaction PvS performance, vision, and strategy PvS performance, vision, and strategy SMEs small and medium enterprises SOEs state-owned enterprises TB tuberculosis TSk Animal Disease Fund Uk United kingdom WTO World Trade Organization ix Highly pathogenic avian influenza (HPAI) under current conditions poses a major risk to human and animal health. Efforts to contain the disease are therefore in national and global interest. As the most widely practiced control methods for poultry involve culling birds that are infected or in regions immediately around infected animals, the most common practice to ensure the cooperation of owners of birds is to compensate them for the culling of their animals to achieve this public goal. Early identification of HPAI and the imme- diate culling of diseased or suspected animals are critical elements of reducing the risk of the disease spreading. The international com- munity and national governments have responded to this challenge by establishing funding mechanisms to enable compensation to as- sist in this strategy. Payment of compensation to farmers whose animals are being culled enhances producer cooperation through better motivation to comply with the disease reporting and culling requirements of dis- ease control packages. It reduces the time lag between an outbreak and containment actions, and hence diminishes the overall cost of control. To the extent that it reduces the virus load, it also reduces the risk of the virus mutating to becoming transmissible from hu- man to human. Enhancing early reporting and complete culling of diseased or suspected birds is thus the first objective of compensa- tion schemes. A second objective can be to reimburse losses of pri- vate citizens who have complied with a disease control process for the public good. This is compatible with the first objective. While the imperative of disease containment drives compensa- tion schemes, the reality of the severe impact of culling on very poor people cannot be ignored. However, a compensation scheme cannot cover all livelihoods losses caused by livestock disease control and it cannot replace social safety nets. This requires other measures, out- side the scope of this paper. The report seeks to provide guidelines on good practice for payment of compensation as part of HPAI stamping-out strate- gies. It is meant for national and international managers and proj- ect staff involved in containing HPAI. It responds to a request of the Senior Officials Meeting on Avian and Human Influenza held in vienna, June 6–7, 2006, and the result of the work of a multidis- ciplinary team from the World Bank, FAO, and IFPRI. The report is based on review of the well-established literature of compen- sation practices in the developed world, staff interviews, experi- ence, and newly emerging gray literature (project documents, mis- Executive Summary Enhancing Control of Highly Pathogenic Avian Influenza in Developing Countries through Compensationx sion reports, and so forth) on compensation in the developing world, and specific field visits to Egypt, Indonesia, and vietnam. Preparedness Is Key An effective and efficient compensation scheme will compensate the appropriate beneficiaries for the ap- propriate losses and at an appropriate level, with only a short interval between culling and payment of the compensation. This will only be possible if a number of elements are already in place before an outbreak. There needs to be appropriate legislation for the control of animal disease in force that spells out clearly the rights and responsibilities of govern- ment, livestock sector and marketing personnel, and farmers in animal disease control. There needs to be widespread awareness of the dangers of the disease and how to mitigate them. Funds have to be readily available and the procedures and sequenc- ing of actions to be followed for compensation need to be agreed in advance. Preparations for the imple- mentation of expedient and transparent payment schemes need to be in place. Procedures and sequencing of compensation require knowing who to compensate, when, how much, and how, and all the stakeholders have to be aware of and have faith in the system. Widespread knowledge in advance of what the stakes are (in- cluding poultry holdings) and identification of the stakeholders are key elements in improving the gov- ernance of the use of compensation resources, which is especially difficult in emergency situations. Because preparedness is essential to using cull- ing and compensation effectively and efficiently for disease control, countries need to make a host of arrangements without necessarily having national precedents to guide the way. The present document tries to illustrate key lessons from countries such as Thailand and vietnam (and others) that have learned by doing and incorporated many of the les- sons in revised strategies. Even with guidelines from elsewhere, national avian influenza committees will still need to negotiate specific arrangements with national stakeholders in a way that fits local condi- tions, and this takes time and effort. Countries faced with outbreaks before they have their contingency plans in place will need to adopt the most basic measures. Even so, the same issues of who to compensate, when, for what, how, and how much still apply. However, the need to move quick- ly for disease control will force many of the normal oversight tasks to a later date and is likely to make governance issues even more difficult. Finally, it will be difficult to delink compensation practices from both changing needs for effective dis- ease control and the issue of equitable production systems change as the disease becomes endemic. This aspect is also introduced briefly in the conclud- ing chapter. Identification of Beneficiaries As a general rule, the beneficiaries of compensa- tion are the owners of the animals. Other supply chain participants, such as feed suppliers and mar- ket operators, may also incur losses when livestock production and sales are disrupted by disease, but they have not normally benefited from compensa- tion schemes. The type of production system sig- nificantly shapes feasible identification procedures. Large, highly biosecure poultry farms (the so-called sector 1 and 2 under FAO/OIE nomenclature) have generally good inventory records and culling is well controlled. Farmer documents are then a basis for compensation. Under conditions of contract farming in these systems, ownership of the birds decides the benefi- ciary. If the contractor is the owner, he/she would be compensated, and takes the responsibility for re- imbursing the integrator. If the integrator owns the bird, he/she will receive the compensation. In a few cases, arrangements have been made to pay the con- tractor for lost income on a wage per day basis, with funds subtracted from the integrator share before payment. The issue of how to incorporate contract growers into compensation process remains a prob- lem that many countries are only now beginning to look into. More attention needs to be addressed to this issue lest it becomes a loophole limiting effec- tive control of the disease. Identification of the beneficiaries for small en- terprises and back yard systems (the FAO/OIE de- nominated sectors 3 and 4) is more complex, as re- cords are normally not available, and factors such as differential ownership by gender come into play. Sur- veys as part of the preparedness planning (not after the disease emerges), including the identification of ownership patterns, broad awareness of the existence of compensation and payment as an integral part of the stamping-out process, are then key factors to en- sure a broad participation of the sectors 3 and 4. Type of Losses to Be Compensated Normally, compensation covers only the so-called direct losses, which include the value of the animals, Executive Summary xi and sometimes also (in richer countries) the costs re- lated to the disposal of dead animals and cleaning and disinfection. So-called farm-level consequen- tial losses, due to business interruption, movement control, and price effects are not compensated, al- though in many developed countries private insur- ance schemes exist for such losses. Dead animals before culling are often not compensated, however there may be a rationale to do so at least partially where either dead animals have market value (and thus there is the danger they will be sold) or disease control teams cannot respond within 72 hours of dis- ease reporting by the farm in question. In all cases the accurate computation of losses is greatly aided by having adequate farm-level records of poultry holdings, and it will be important to promote such a database prior to disease outbreak. Finally, the lion’s share of actual economic losses to the countries in question may be indirect: lost feed sales, diminished tourism, absenteeism at work, and so forth. These losses are never covered by public compensation schemes. In principle, they could be insurable under private sector contracts outside the livestock sector if risks are well known, but they rarely are. Setting Compensation Rates Compensation rates are variously set on the basis of (a) market value; (b) budget availability; and (c) production costs. Setting the cost on the basis of market value, wherever possible, is the preferred policy, as basing the cost on budget availability of- ten leads to underpayment, and hence poor compli- ance with the culling operation, and production cost would favor inefficiencies, and is more complex to establish. Experience that emerges from the review in establishing compensation rates based on market values shows: Compensation rates as a percentage of a refer- ence market price should be set before the dis- ease emerges, as part of an overall preparedness plan, using average preoutbreak market prices at the farm gate, computed with due regard for seasonality and the transport costs from the local community relative to the reference market. For special category birds (rare breeds, indigenous poultry, fighting cocks, grandparent stock, other bird types), where market prices are not readily available, consultation with the stakeholders is required to set realistic levels. Uniformity of rates across the country and for different classes of birds improves the implemen- tation efficiency of the program, and should be pursued in situations with good control. How- ever, in situations of poor movement control, • • differentiation by type of bird (layer, broiler) and age/weight of the group might be needed to fit compensation as close as possible with prevailing market prices. An interesting intermediate solu- tion might be to pay not on the basis of numbers but on the basis of the total weight of the flock. Compensation rates should be no less than 50 percent of the reference market value of suspect- ed birds at the farm gate, and no more than 100 percent. The rationale for the preferred range of 75–90 percent of the reference price and multiple considerations for being closer to one or the other limit are discussed in the report. Rates should be considerably lower for diseased birds and even less, but positive, for dead birds, to provide posi- tive incentives for early and complete reporting. Careful attention needs to be paid to bird move- ments during compensation to ensure that an incentive is not being created for the influx of healthy birds to disease zones or diseased birds to disease-free zones. In dealing with small farmers in developing countries, compensation should be paid within 24 hours of culling by cash (or possibly voucher where handling cash presents a security threat and credible local formal financial institutions such as rural post offices are available); any delay is likely to have a significant effect on reporting. Establishing Awareness Experience from on-going campaigns highlights the absolute necessity of communication on dis- ease control and compensation, which when done properly may run from 10 percent to 20 percent of the total package cost. The package should contain components of consultation with the beneficiaries, advocacy, and information, using multiple media and channels. The specific messages on compen- sation should explain to affected farmers the need for mandatory culling in cases of suspicion of avi- an influenza as a necessary measure to protect the health of the entire human population. They should contain the principles, procedures and grid of com- pensation levels, precise information on the exact amounts, and payment procedures. Messages and media should be prepared ahead of time with inputs from both technicians and communication special- ists. They should also be consistent over time, since frequent policy and message changes undermine the credibility of the campaign. Private sector opera- tors, such as para-veterinarians, can play a critical role in awareness raising and overall support to the campaign, and their input on retainer fees should be more encouraged than is currently the case. • • Enhancing Control of Highly Pathogenic Avian Influenza in Developing Countries through Compensationxii Payment Systems To promote early notification of suspected out- breaks, compensation for culled birds must be paid promptly following the birds’ destruction. Critical el- ements from an appropriate payment system follow. Rapid access to adequate funding for immediate deployment as needs arise is essential. Sources typically are government’s own funds from the National Treasury, farmer’s contributions, and those of donor partners. National budgets need contingency funds of at least 3–5 percent of total budget to facilitate a rapid central contribution in the event of an outbreak; alternative contingen- cy planning will be necessary where this is not available. The share of compensation payments in total animal disease control expenditures under out- breaks ranged from 0–45 percent in the cases studied, with a central tendency of about 35 per- cent. Holding large sums as contingency reserves to allow a rapid response engenders a consider- able cost. For compensation planning purposes, the upper range of foreseen culling during a se- vere outbreak should be capped at 10 percent of the national flock. Many outbreaks are controlled with culling of less than 1 percent of the national flock. Once the share of infected and closely asso- ciated birds exceeds 5 percent of the total nation- al flock, vaccination typically starts substituting for culling and compensation. These percent- ages, multiplied by the size of the national flock and again by 75 percent of the average farm- gate poultry price, provides a rough estimate of the range of funds that need to be accessible for compensation payments per se on short notice. Countries that are important poultry exporters and wish to avoid vaccination (such as Thailand under its 2004 outbreak) should plan at the 10 percent (high) limit, countries with little in the way of poultry exports and a large percentage of smallholder poultry producers at 5 percent, and countries with little trade concern, a high degree of biosecurity, and a creditworthy public finance system at 1 percent. The system should be simple enough to be used in difficult field situations and should make use of existing institutions (for example, line minis- tries, veterinary services, financial institutions). It is important to clarify responsibilities in advance, make provincial cross-agency coordination ar- rangements, and establish local contingency funding. If no system is in place when the disease emerges, the focus will need to shift to a greater reliance on ex post independent scrutiny to avoid inordinate delays in paying compensation. Eligibility databases and emergency payment (see above) procedures should be prepared as • • • • part of the emergency part preparedness plans; where lacking, they will both need to be set up when the disease emerges, posing considerable difficulties. The veterinary services (assessing the need and reliability of the culling), the Ministry of Finance (payment), civil authorities (security), and com- munity leadership (transparency) should all be directly involved in the payment process. For sectors 1 and 2, bank transfers are the most adequate instrument; cash payments are the pre- ferred method for those farms of sectors 3 and 4 without banking access. vouchers are often less credible for immediate motivation of rural house- holds, but may work where they can be integrat- ed with a dense local network of trusted financial institutions, such as rural post offices. To the extent possible, maximum use should be made of local banking entities, producer’s orga- nizations, veterinary services, and nongovern- mental organizations (NGOs). Their fiduciary assessment should be part of the preparedness planning. The Way Forward While over time the international public good ar- gument regarding the risk of human-to-human transmission of HPAI might diminish, transmis- sion between animal populations of different countries will continue to be a main reason for in- ternational funding of disease control in develop- ing countries. Moreover, in the likely event of the disease becoming endemic within certain coun- tries, this will have major effect on the poor, and interventions under those conditions therefore deserve international support from an equity per- spective. Stricter disease control requirements will have a major effect on the structure of the industry, with implications still to be clearly identified for the future viability of the sectors 3 and 4. None- theless, compensation is likely to remain neces- sary for many years to come to promote the early eradication of outbreaks and to avoid the spread of transmissible animal diseases. Under such conditions, compensation will: Become part of modified stamping-out strategies, with probably a lower priority to culling. Clear principles of how stamping-out strategies should evolve, and how compensation fits into such evolving strategies are needed. Have to become more dependent on the coun- tries proven political will to improve the key institutions for animal health, in particular for early alerts and independent disease reporting. The OIE tool for Performance, vision and Strat- • • • • • Executive Summary xiii egy (PvS) is a useful instrument to assess govern- ment capabilities. Be restricted to sectors 3 and 4, and be funded from a mixture of national and international pub- lic funds, the latter in particular for the poorer countries. • Be funded for the large commercial sectors through private initiatives, probably as a mix between mandatory levies and voluntary insur- ance; in many cases the public sector needs to work with the private sector to find equitable ways to develop these systems. • � Introduction1 Compensation as used in this paper relates to indemnification of private actors for losses incurred as a result of public action under- taken to promote the public good, such as in the case of payments to farmers for culled birds. Compensation is typically used as part of a stamping-out strategy in emergency situations or in support of inter- ventions for eradication after successful reduction of incidence. Compensation is fundamental to control strategies for the H5N1 strain of highly pathogenic avian influenza (HPAI for short), as it provides an incentive to the producer to report suspected disease outbreaks and comply with culling actions. It therefore reduces the time lag between an outbreak and containment actions and hence the overall control cost, and to the extent that it diminishes the vi- rus load, reduces risk of the virus mutating to become transmissible from human to human. However, the implementation issues are extremely tricky, as poorly designed schemes can in fact hasten the spread of the disease. Compensation can also mollify the effects of critical social and political shocks where large numbers of farms are decapitalized through culling of their stock. However, as will be dis- cussed, mixing greater social objectives with compensation for dis- ease control typically lessens the effectiveness of the latter without bringing to bear what is required for the former. Implicitly this report treats compensation as an international pub- lic good. This stems from the possibility of catastrophic international spillovers from HPAI outbreaks affecting human health. Compen- sation remains a key tool in support of stamping-out strategies for animal disease control even when human health is not at stake (as with Foot and Mouth Disease), but does not involve the same degree of priority, nor the implication that funding dedicated to the creation of international public goods should assist in meeting needs. Objectives of the Report and of Compensation Objectives of the Report The main objectives of the report are to discuss the purposes of com- pensation, review experience, link compensation practices to success in culling strategies, analyze how inappropriate compensation pack- Enhancing Control of Highly Pathogenic Avian Influenza in Developing Countries through Compensation� ages (such as in overcompensation) can create per- verse incentives, illustrate pitfalls and uncertainties, and develop recommendations based on observed good practice. The desired outcome is to facilitate changes in compensation practices that improve disease control through more prompt reporting of disease and bet- ter compliance with culling initiatives. The primary target audience consists of decision makers in the public sector of developing countries, their devel- opment partners in international agencies, bilateral donors and international NGOs, and public/private coalitions, producer organizations, and NGOs in developing countries. The paper will review issues and formulate rec- ommendations on the following items of good practice: Deciding who to compensate; Types of losses to be compensated; Setting the level and timeliness of compensation; Promoting awareness, communication, and ca- pacity building; Organizing payment while promoting account- ability; and Shifting compensation strategies as the disease becomes endemic. Purposes and Scope of Compensation Schemes Compensation schemes have been implemented for other diseases of livestock, including highly contagious animal diseases such as such as Foot and Mouth Disease (FMD), Classical Swine Fever (CSF) and Contagious Bovine Pleuropneumonia (CBPP) and zoonotic diseases such as rabies, bovine spongiform encephalopathy (BSE), and brucello- sis. Appendix 1 lists examples of different sorts of compensation schemes that have been applied for animal disease outbreaks. HPAI is a recent entrant to the scene, and is now included in compensation schemes in a growing list of countries, shown in the appendix. Compensation processes are most fully documented in industrialized countries and those with highly developed commercial livestock indus- tries such as Brazil and Chile. Historically, governments have entered into ani- mal health compensation programs for a variety of reasons. These range from the productivity cost of letting a disease spread throughout the country, to avoid restrictions on exports as in the case of OIE- notifiable diseases such as FMD and CSF, to emer- gency measures to protect human health (such as in the case of BSE or most recently HPAI). The first • • • • • • documented compensation schemes were applied to owners of cattle slaughtered due to infection of cattle plague (Rinderpest) in Britain and Ireland un- der the Cattle Disease Prevention Act of 1866, and then in 1884 in the United States (Wiser 1987).1 Com- pensation schemes today continue to be designed primarily as part of a disease control strategy, to encourage farmers to report outbreaks or potential disease problems, and to discourage them from sell- ing animals rather than presenting them for culling (FAO e-consultation 2006). Animal diseases that warrant compensation sys- tems typically have several dimensions: (a) the type of disease is highly contagious, may be zoonotic, and potentially engenders serious economic impact; (b) the type of animal is typically of importance to commercial farming (cattle, small ruminants, swine, or poultry), or closely related to these animals; (c) the source of funds involves a mix of public and private cost-sharing with the mix depending on the diversity of production systems (from large inte- grated farmers to backyard smallholders); and (d) the disease status of the country in question, with compensation being less prevalent if the disease is endemic. HPAI in most developing countries fits all of these conditions to varying degrees, but above all has the potential to inflict devastating damage on global public health if the virus mutates (or more specifically is reassorted) to become transmissible among humans. The human health dangers alone are enough to justify using international develop- ment funds to speed control through culling and compensation schemes, as the international “public bad” of transmission across borders of a mutated vi- rus dangerous to humans is clear to all. It is tempting to mix the issue of compensation for disease control purposes with a host of issues beyond disease control. It can be argued that there is a moral obligation of states to compensate for private property destroyed in the public good, where the destroyed property is a legally held and productive asset, especially in the case of very poor people. Furthermore, poultry is often a sig- nificant share of the assets of the poor in developing countries, in particular for women and rural people with few collateral assets with which to obtain fi- nance for rebuilding their source of livelihood. In Muslim countries located in the humid tropics, poultry is the main source of animal protein, rais- ing particularly thorny problems for achieving rural acceptance of stamping-out policies in coun- tries such as Indonesia and Malaysia. Poultry also Introduction � remains one of the best options for future pro-poor agricultural growth (poultry and pork consump- tion have been growing at 5 percent plus per capita per annum in developing countries over the last 25 years compared to less than 0.5 percent for cereals) (Delgado et al. 1999). Yet compensation for disease control needs to be kept conceptually and operationally separate from compensation for damages or asset rebuilding of the poor. Both are vital issues, but they require dif- ferent conceptual approaches, different operational measures and time frames, and have separate con- stituencies (and thus funding sources). As will be discussed below, the losses from animal disease out- breaks go far beyond lost sales of animals by farm- ers and cannot all be realistically addressed by pro- grams designed to speed up reporting of outbreaks and facilitate compliance with culling. Furthermore, as the disease becomes endemic, outbreaks go from being new natural catastrophes to more predictable occurrences. The latter may require different measures for containment, in- cluding other methods for disease control such as vaccination, and practically-speaking local HPAI outbreaks may be harder to distinguish from those of other poultry diseases that are endemic, such as Newcastle Disease. Containment when the disease is endemic will surely require greater cost sharing by producers and consumers. Finally, recapitalizing the rural poor after repeated outbreaks is a complex operation involving credit, extension, institutional changes, other measures to promote biosecurity, and a host of other items beyond the scope of rapid- response compensation schemes. Background and Rationale for Assessing Good Practices Following the logic above, compensation is a vital tool to reduce risks to animal and human health through increasing the width, depth, and speed of producer compliance with reporting and culling or- ders. It also helps reduce both the direct economic costs (such as falls in exports) and the typically much larger indirect costs (such as lost economic activity and the decline of tourism) of disease outbreaks.2 A further complication in developing countries is that livestock death and disease are considered to be some of the main factors contributing to poverty (World Bank and DFID 1999). How compensation is managed will greatly affect whether the poor will benefit on an equal basis with large-scale producers. The institutional framework for compensation will also help determine whether small-scale produc- ers remain in the industry; if they cannot remain in smallholder poultry because of stricter biosecurity measures needed to control the disease, the format of compensation (cash versus targeted credits, for example) can help promote the move to alternative livelihoods. The Global Program for Avian Influenza (GPAI) allows for the use of up to US$500 million in Inter- national Bank for Reconstruction and Development (IBRD) loans or International Development Associa- tion (IDA) credits or grants for new projects, or for restructuring existing projects, to undertake imme- diate actions to control avian influenza. A share of the funds is expected to be allocated to compensate for economic losses as a result of culling measures to contain HPAI. In addition, client country gov- ernments and World Bank operational staff have requested guidance on these issues. More than 25 countries are expected to receive financing under the GPAI by December 2006, and the presentation of a satisfactory compensation scheme is included as a condition of disbursement in the GPAI funding for several countries. Guidelines for good practices are needed for predicting funding needs for compensation during emergency culling, both to eliminate unnecessary costs from overestimating needs and to avoid loss of credibility from not having sufficient resources at hand for adequate or timely payments. Presently, the numerous schemes being followed for HPAI compensation exhibit significant variation across and even within countries, along with highly vari- able outcomes in terms of disease containment. Approaches should be developed to allocate fund- ing by geographic zones within countries based on forecasted needs using transparent criteria. Sources of funding need to be identified, and any differen- tiation between uses of funds from different sources explored (such as differences in use of multicountry regional, national host government, local, private sector, and donor partner funds). Management ar- rangements, accounting procedures, and disburse- ment arrangements need to be decided. The Senior Officials Meeting on Avian and Hu- man Influenza held in Vienna, June 6–7, 2006, iden- tified the need for guidelines based on best practices for compensation of producers of culled birds un- der developing country conditions as a top prior- ity in the fight against avian influenza. The World Bank offered to take the lead in preparing a report on good practice to be useful to its own staff, client countries, and partner agencies with regard to the design of compensation schemes. Enhancing Control of Highly Pathogenic Avian Influenza in Developing Countries through Compensation� Context in Which Compensation Occurs Animal Production Systems and Degree of Development Compensation has been given in some developed countries since the 19th century for culling in the context of a variety of contagious animal diseases, and also in the context of disaster relief in a vari- ety of developing countries. While this report will draw on these experiences, the focus here is solely on compensation for culling of birds associated with avian influenza control programs, with particular reference to developing countries. Using the FAO system of classification of pro- duction systems (FAO/OIE 2005), poultry farms in developing countries fall into one of four sectors. Sector 1 is an industrial and integrated system that produces commercial breeds of poultry using tech- nologies similar to those found in Organisation for Economic Co-operation and Development (OECD) countries, with adaptations where necessary for differences in temperature and humidity, and with relatively high levels of biosecurity.3 Sector 2 is an industrial but not integrated system that produces commercial breeds of poultry at levels of scale com- parable to production units found within sector 1, but typically at lower (but still significant) levels of biosecurity. Sector 3 is neither integrated nor industrial, typi- cally with low levels of biosecurity in developing countries, producing commercial breeds of poultry, at relatively small scales compared to most well-es- tablished commercial poultry operations.4 Sector 4 is small scale or backyard, produces native breeds of poultry and has little or nothing in the way of bi- osecurity measures. Sector 4 still accounts for a large share of poultry production in most developing countries, but most countries have all four sectors represented simultaneously. Appendix 2 provides a detailed description of the sectors. Countries that have a large proportion of sector 1 and 2 farmers are more likely to be those involved in export trade. The Legislative Environment Culling involves the destruction of private property for the public good, a process that should never be arbitrary. Compensation schemes implicitly rec- ognize the obligation of the public entity to make good on the loss to the private entity affected, even if the motive is to provide an incentive for rapid and peaceful compliance. The need for appropriate leg- islation for the control of animal diseases is key to the success of culling and compensation strategies. The legislation needs to include the specification of the right and conditions for governments to seize private property, including the obligation of gov- ernments to compensate when they take property. It is critical that such legislation is in place before an outbreak. This need is not limited to HPAI control, yet clear legislation on this is often lacking in devel- oping countries. The World Animal Health Organization (OIE) recommends that destruction of diseased animals and products and adequate disposal and disinfec- tion should be clearly spelled out in legislation as a responsibility of the central government. This re- mains the case even when the actual tasks are car- ried out by others. Furthermore, implementation of the legislation is crucial: the articles should clearly specify who decides on culling, who pays, when, and how payments are made. There needs to be a clear chain of command, especially as many differ- ent kinds of actors (public and private) are involved and speed and coordination are imperative. Variation in the Strength of Public Adminis- tration, Animal Health, and Rural Financial Systems Institutional environments for animal disease con- trol and compensation payments also vary signifi- cantly across countries, including across developing countries. Key elements are the technical and insti- tutional strength of public administration, animal health services, and rural financial systems and their capacity to enforce other control measures, in particular movement control when the disease emerges. These range from the well established (Australia, European Union [EU], and Hong Kong, for example), to the emerging countries (Thailand and Vietnam, for example), to still weak institution- al environments (most other developing countries). Approach to Ascertaining Good Practice The report is the product of a multidisciplinary, multiagency activity based on review of existing literature and interviews with selected developing country stakeholders, staff, or consultants from in- ternational agencies that have been involved with the implementation of compensation schemes for the control of HPAI and with conditional cash trans- Introduction � fers to large numbers of poor people. Interviews of experts from OECD countries who have experi- ence with compensation for other diseases, such as Newcastle Disease, FMD, and CSF were also used. Explicit attention was paid to necessary precondi- tions for the success of developed country strate- gies, such as strong veterinary services, availability of diagnostic labs, and so forth. Field visits in sup- port of the present paper were made to Indonesia, Egypt, and Vietnam. Finally, the study team fed into the terms-of-reference of an FAO-managed “E-con- sultation” of 200 experts around the world on avian influenza compensation, which fed directly into the present report. Judgements about “good practice” are often sub- jective. The team adopted the following working indicators of success and lack of success in compen- sation practices. Indicators of Success The spread of the disease is reduced compared to what happened in countries and regions where compensation was not used, or used late or in- consistently. Those who are entitled to be paid under an- nounced rules are in fact paid. Compliance with reporting and culling is in- creased relative to estimated numbers of dis- eased and suspected (that is, still healthy animals within the immediate neighborhood). Livelihoods’ distress directly related to the dis- ease control process is significantly reduced. Measures for reporting, culling, and compensat- ing are linked, enshrined in law, and preferably agreed on by key stakeholders. There is clear and accurate communication about entitlement. Indicators of Lack of Success The disease spreads. Persons who were entitled to compensation were not paid. The compensation process added to inequities. There was a high level of noncompliance. The operating procedures for reporting, culling, and compensating were delinked and poorly de- scribed or poorly understood. There was inconsistency in regulations or enforce- ment across small geographical or administrative areas and between different size producers. • • • • • • • • • • • • Conclusions and Recommendations from Chapter � The primary objective for establishing a compensation scheme is to promote effective disease control. To achieve this objec- tive, a scheme must operate efficiently and quickly and be well understood by those on whom its successful operation depends. It must also fit into the context of the whole disease control plan. A second objective may be to reimburse losses of private citi- zens who have complied with a disease control process for the public good. This is compatible with the first objective. Compensation schemes associated with mandatory culling have a long history in animal disease control, but are relative- ly new in the case of poultry, and particularly H5N1 HPAI. The potential danger to global public health from HPAI clear- ly make its control a priority global public good. Many of the lessons learned from compensation schemes involving HPAI control will be of great use in future outbreaks of other dis- eases. Successful compensation practices need to fit the production systems, animal health service and laboratory availability, ad- ministrative constraints, and the enforcement capabilities and rural financial systems context of the countries in which they operate. The most critical need is to enact clear legislation, pri- or to any outbreak, spelling out the rights and responsibilities of government, service agents such as private veterinarians and market agents, and farmers in the event of an outbreak or to prevent the threat of an outbreak. A clear chain of com- mand is necessary for successful disease control, including the implementation of compensation. Compensation schemes cannot cover all livelihoods’ losses caused by livestock disease control and they cannot replace all social safety nets, or administration becomes impossible. The limits to compensation must be identified so that addi- tional private and public measures can be put into place to reduce loss of livelihoods. While the objective of disease control drives compensation schemes, the reality of the severe impact of culling on very poor people cannot be ignored in developing countries. If smallholders are not effectively brought into compensation schemes, it is likely that they will be reluctant to report out- breaks, with negative consequences for all sectors. Nor is it re- alistic to believe that small backyard operations will disappear in response to prohibitions; experience shows that it is more likely that they will go underground and escape any effective veterinary control. It will be difficult to delink compensation practices from the issue of equitable systems change as the disease becomes endemic in specific developing countries, and thus compensation practices will likely need to change over time, as will disease control strategies. � This chapter examines the potential beneficiaries of compensation in support of disease containment poultry sectors that vary greatly across and within countries. Where the poultry sector is very di- verse, different approaches may be needed for different production systems within a country. An attempt will be made throughout to differentiate approaches between smallholder and backyard pro- ducers, on the one hand, and large-scale integrated and commercial producers on the other, despite the fact that the literature is much larger on compensation of the latter than of the former. Who Has Been Compensated in Control of Animal Diseases Other Than HPAI? Drawing on the pre-H5N1 precedents, primarily in developed countries, the beneficiaries in compensation schemes have been the owners of the animals, who have been paid for the direct loss of their livestock through culling and in some cases for various di- rect additional costs that were incurred during stamping out. There are many other kinds of losses associated with animal disease out- breaks, and these are discussed in Chapter 3. Who Is Involved in Poultry Production? Types of Production Systems As discussed in Chapter 1, developing countries often produce poultry in a variety of distinct production systems with different levels of biosecurity and institutional support, as described in Ap- pendix 2. Each type of sector presents particular challenges in de- signing compensation programs. Sectors 1 and 2 can be grouped together in designing compensa- tion processes. Production units in these systems tend to be large scale, with well-organized biosecurity systems and quality control processes. A well-run sector 1 or 2 farm in theory has a good chance of seeing suspect cases early. However it is recognized that even when there are sentinel cases, the general awareness of the disease may still be low, resulting in delay in identifying susceptible cases. Deciding Who to Compensate: The Beneficiaries 2 Enhancing Control of Highly Pathogenic Avian Influenza in Developing Countries through Compensation� Culling operations are on a large scale and the farm staff can be trained to take part in destruction of carcases and disinfection of premises, so that some of the direct costs of culling may initially be borne by the farmer. Large-scale commercial farms usu- ally keep good records on numbers of birds, cost of production, and sales. These usually can be verified after the fact, if necessary, with input suppliers. It is fairly straightforward under these conditions to calculate compensation numbers and rates based on culling. Export trade usually takes place from sector 1. Difficulties arise when contract growers are part of the system, since they may not technically be the owners of birds yet they may have made invest- ments in production and may also suffer income loss from culling or during downtime. In Brazil and Thailand, for example, 70 to 80 percent of birds are found in sectors 1 and 2 (Delgado, Narrod, and Tion- gco 2003; Narrod and Pray 2001). Many (most, in ex- port zones) are contract farmers following biosecure practices and are well monitored by veterinarians employed by the integrator company. Broiler pro- duction in particular tends to be highly integrated, while layer production is somewhat less integrated. Sectors 3 and 4 also exist in these countries, but tend to account for relatively small and shrinking shares of production. Sector 3 farms are commercially oriented, but tend to be less biosecure. They often do not have consistent records on birds kept, production costs, or sales. Since the genetics and feeds used have more in common with sectors 1 and 2 than with sector 4, the birds being compensated have more in common with the commercial product than in the case of the backyard sector.5 Further, sector 3 farmers in developing countries may sell into more than one market, with differenc- es in prices, and may change their pattern of sales by season, in a way that is predictable but not doc- umented. They may also sell through middlemen. The close contact nature of humans to birds in these systems that are integrated with family households carries the higher possibility of farm families being exposed to the virus. It is also very easy for birds from a small-scale flocks to be sold, moved, or eaten in advance of a culling operation and be very dif- ficult to trace. The last point is even more applicable to sector 4. Sector 4 flocks are kept with minimal inputs. In most places, birds roam freely around the farmstead or village, in others they are kept in family court- yards enclosed by walls, but in either situation they eat household scraps and what they can find in the environment. Birds and eggs are primarily for con- sumption in the households that produce them or sale to neighbours or in local markets so the owner can easily be identified. The owner is likely to be a woman, while the head of household may not be a woman, and compensation processes need to be designed with the actual owner in mind. Record keeping if any is minimal in sector 4. Birds may be of indigenous breeds with a different mar- ket value to those in sector 3, and there are likely to be limited records of holdings. Sector 4 farms are the most likely among farm types to be excluded (or at any rate absent from) from official culling schemes. Small-scale urban poultry keepers are included in this sector and may be the hardest of all to locate. A field assessment of Egypt carried out as part of the present work suggested that more than one poultry producer in ten falls into this group. Because of con- cerns about human health, urban smallholder flocks are highly likely to be targeted in a restructuring plan and a number of countries have banned this type of operation. Such bans often result in back- yard poultry production going underground. In emerging countries that have experienced HPAI, such as Indonesia and Vietnam, sector 4 may still account for approximately one-third of all birds, but sector 3—which previously was expanding and increasingly prevalent—has shrunk dramatically.6 Poultry Systems as a Continuum More than any other livestock system, poultry sys- tems lend themselves to scaling up, and when they do, they may also change geographical location to be closer to final markets in large towns (Delgado, Narrod, and Tiongco 2003; Costales et al. 2003; Poapongsakorn et al. 2003). Industrialization may be encouraged by governments as a means of sat- isfying growing domestic demand, developing an export market, or developing a sector structure that lends itself more easily to disease control. There continue to be large numbers of producers keeping birds on a small scale during the transition phase. Where domestic demand is growing quickly, sec- tors 1 and 2 often cannot grow fast enough to satisfy it, and supply from sector 3, which responds very quickly to market changes, may be preferred to im- porting. Countries can be placed on a continuum accord- ing to the state of their poultry market, from those that are concentrating the sectors very quickly to those where change is very slow or hardly happen- ing at all. As suggested in Figure 2.1, the developed countries are at one end of the spectrum, with Bra- Deciding Who to Compensate: The Beneficiaries � zil, Malaysia, and Thailand steadily converging on them in terms of productivity, and Cambodia, Bo- livia, and Burkina Faso are at the other end. Indo- nesia, Egypt, Nigeria, and Vietnam are among those that fell in between, and may in fact face the great- est challenge in designing compensation schemes, because they require equal but different efforts to be made for the large commercial systems and the smallholder systems. Contract Farming A problem in who to compensate arises in the case of contract farming, which accounts for the lion’s share of poultry production in many developed countries, such as the United States, and is increas- ingly important in many developing countries. In this vertically coordinated system, management of the poultry is separated from ownership, and both parties have sunk costs in the poultry flock; the inte- grators provide capital and the contractors provide labor, land, and farm facilities. In countries with tested legal systems for enforcing contracts and large sector 1 and 2 producers, such as the United States, the prevalent practice is to pay the own- ers of the birds—the integrator—and to encourage them to compensate the contract growers (Ott and Bergmeier 2005). In developing countries, there is a greater risk that contracting farmers will never see any part of the payment, particularly if the govern- ment is relying on the integrator to disburse some of the payments to growers. A suggested practice here is to split compensation paid to the two par- ties, with the contractor share being based on days of labor input and the sum of compensation per bird between integrator and contractor being the same as for independent farmers. Other Agents with Incomes Directly at Risk from Culling In the industrialized countries, with production coming mainly from sectors 1 and 2, market chains tend to be fairly simple and integrated. Distinct chains can be found for different types of birds or products. The contractual relationships between units are strong, with units owned by the same par- ent company or bound by legal supply contracts. In rare cases where production comes from sector 4, most marketing occurs within very short chains, often from the farmer directly to the consumer. In developing countries, other than for sector 1, relationships along supply chains are rarely based on written contracts, although in some cases they may involve long-term verbal agreements reinforced by family or ethnic ties. As cities have grown, sup- ply chains have lengthened and widened and have become more complex and more anonymous com- pared to village-level marketing (Delgado, Narrod, and Tiongco 2003). Accordingly, the number of criti- cal points for movement and disease control within supply chains have multiplied, as have the number of agents who stand to lose if poultry is culled any- where along the chain. The complexity of supply chains in sector 4 of developing countries and the relative ease with which producers in this sector can sell diseased ani- mals into these chains illustrates the importance of disease control containing adequate compensation schemes for sector 4 farmers to control disease at the source, despite the greater difficulty in implement- ing such schemes in sector 4 compared to the other sectors. As to supply chain agents other than primary producers, it is typically not feasible to include these agents in compensation schemes, even if their losses are directly related to state-mandated sei- zure of livestock products. Literature searches re- veal no cases of compensation paid to any market chain participants other than poultry owners, with the exception of one isolated case of live markets in Hong Kong. In that case, the government developed a compensation and financial package to aid all em- ployees of live market trades that were affected by the mass cull and temporary suspension of business in May 2001 (Legislative Council Secretariat 2002). The Hong Kong approach is a solution applicable only in the rare case where the number of persons involved is small and easily located in one place, and the resources involved are minor compared to the overall resources of the government involved. However if market chains are clearly the source of infection or spreading the disease, compensation in some form might lessen the risk of them continu- ing to be a source of infection/spreading the disease and perhaps this should looked at more closely in a more in-depth study than this one. Ensuring That Eligible People Are Included Deciding Who Is Eligible Unless HPAI compensation schemes depart radical- ly from those for other livestock diseases and from HPAI schemes to date, or some other very unusual Enhancing Control of Highly Pathogenic Avian Influenza in Developing Countries through Compensation�0 considerations hold (as in Hong Kong, above), those eligible for compensation should be confined to the persons making the management and sales deci- sions for the birds or eggs involved in the stamping- out processes. In most cases, these are the owners of the birds, with a few exceptions discussed below. Who decides eligibility for compensation depends primarily on the structure of the sector. Where sec- tors 1 and 2 are strong and there is previous experi- ence in designing compensation schemes, the voice of industry tends to be strong. This is particularly the case where the large-scale private sector (sectors 1 and 2) contribute to the funding of animal health services, as discussed further in Chapter 5. Small- scale producers are seldom consulted to any great degree unless they are contracted farmers with a strong union. If the scheme is being designed in an emergency, the process is likely to be more top down for lack of feasible alternatives. Over time, there is value in developing a consultative process around disease control methods, including compensation, because a scheme that is designed to take account of the priorities of different actors is more likely to encourage compliance with disease control. In countries with highly developed poultry sec- tors, the main challenge may be to bring together the actors in market chains that may normally be quite separate, to discuss openly with those who may be competitors how to organize and fund compensation. Both public and private actors are likely to be part of the decision making. Where systems are dominated by sectors 3 and 4, the challenge is to find organizations that really represent the scattered and highly diverse pro- ducers. Otherwise, smallholders are unlikely to be represented in the decision-making process. Identifying and Locating Eligible People As well as determining who will be eligible for compensation, it is important to be able to locate them and to communicate with them their rights and responsibilities. • • Deciding Who to Compensate: The Beneficiaries �� Sector 1 and 2 producers are likely to be part of official records as registered companies. Where no database exists on these producers, it is a priority to construct one. Sector 3 producers may not be registered, although countries whose poultry sec- tors are becoming more concentrated are likely to introduce registration of flocks above a certain size. Sector 4 producers are rarely if ever registered and it is unrealistic to assume that they can be. There are reports from at least two low-income countries that the very process of sending field agents from village to village to register producers during an outbreak helped spread the disease. If compensation is linked to culling, then regis- tration of those eligible can be linked to the culling process. When a culling team goes to a farm, details of the farm, the bird owners, and the type and num- ber of birds can be recorded, as well as any other information about costs that may be compensated. Chapters 3 and 5 explore the process in more detail. On sector 1 and 2 farms, as well as physical in- spection, farm records will be accessed and an of- ficial evaluator assigned. On sector 3 and 4 farms, where there may be no records, a physical count of the birds culled is necessary. Since successful cull- ing and compensation rely on trust, the registration process needs to be robust and appropriate to the local situation. Common practice is to involve three persons representing three authorities— veterinary, financial, and civil—in registration. The relationship between public and private vet- erinary service provision in the context of HPAI can change for the commercial sectors in the event of a serious disease outbreak such as HPAI. While the large-scale commercial sector may continue to rely on its own veterinary staff, the public sector veteri- narians retain a fundamental oversight role of all sectors for disease control purposes, and under most legislation their judgement prevails in the event of disagreement. Important issues in identifying people eligible for compensation processes are: Ensuring that people take part in culling schemes and are registered when they do. Determining who owns the bird in the cases of: contract farmers, a bird culled at market, or where the head of household is not the same person as the bird owner (husband paid for wife’s poultry, and so forth). • • Identifying those who have knowingly contrib- uted to disease spread and should not be eligible for compensation. Poultry owners who are typically excluded from compensation despite suffering direct losses during stamping out are: Those who are not covered by official culling schemes. Reports from several countries have provided information about small flocks being killed before official culling, birds being sold rather than culled because of lack of trust in the compensation processes, or birds dying of dis- ease before culling teams arrived. Those who lose business through movement con- trol although their birds are not culled.7 This may arise if ring vaccination is used together with limited culling, so that birds stay alive but cannot be sold at the normal time because of restricted movement. Producers who are outside of official disease con- trol areas but suffer from falling demand, falling prices, or difficulties in securing inputs. Encouraging Farmers to Report Early Usually farmers are reluctant to report the first case of HPAI, either because they do not understand the issues or for fear of social and economic implica- tions. In at least one case, paying a bonus above the set compensation rate has been suggested as an in- centive to get farmers to report as soon as they sus- pect a case in a non-disease area.8 The most critical incentive to get sectors 3 and 4 to report is to pro- vide them compensation within a 24–48 hour period from culling. This however is not always possible as discussed further in Chapter 4. Mechanisms of Solidarity among Livestock Owners In many countries, livestock owners, organizations, cooperatives, groupements sanitaires, and so forth, have been found to be effective ways to mobilize everyone in a locality, including small-scale produc- ers, to work together to contain disease outbreaks. Such collective action at the producer level greatly simplifies the public sector tasks and typically im- proves governance. It seems likely that they will work well where this is an option to assist in the implementation of compensation schemes. • • • • Enhancing Control of Highly Pathogenic Avian Influenza in Developing Countries through Compensation�� Conclusions and Recommendations from Chapter � In compensation schemes for animal diseases, beneficiaries are normally live- stock owners who have directly been affected by culling schemes. Well-designed schemes try to reward for compliance with culling and encourage early reporting of suspect cases while exerting penalties for behavior that will lead to disease spread. Others in the livestock market chains, such as input suppliers, are not normally compensated, and neither are sectors other than the livestock sector that suffer consequential losses from animal diseases. Potential beneficiaries of compensation schemes for HPAI include owners of poul- try and eggs in four sectors, defined as 1 (industrial), 2 (commercial with good biosecurity), 3 (commercial within limited biosecurity, often small scale), and 4 (backyard scavenging). Different procedures and funding sources are likely to be appropriate for sectors 1 and 2 and sectors 3 and 4. Often producers are fearful of being the first farms to be found infected with HPAI, thus there is reluctance to bring birds in for diagnosis; schemes should con- sider providing a bonus to producers providing evidence of the sentinel case in a region. In countries where good practices can be identified, producers tend to be primar- ily sector 1 and sector 2, working closely with government and often focused on international trade. Under conditions of contract farming, the beneficiary is the owner of the birds. If the integrator legally owns the bird, he/she will receive the compensation, al- though often arrangements are made to pay the contractor for lost income based on days worked and these payments are subtracted from payments to integra- tors. A well-designed scheme requires clarity in advance as to whom the intended ben- eficiaries are, and this is dictated by the needs for effective disease control. Where sectors 3 and 4 are large, this requires ways to identify beneficiaries, including those not officially registered as poultry keepers. Compliance is most widespread where there is a consultation process to ensure that the needs of all types of beneficiaries are considered when planning the process, including a grievance mechanism for those whose voices are not being heard. �� This chapter examines the literature on compensation to assess which types of losses have been incurred as a result of animal diseases and what subset of these are typically included in existing compensation schemes. It also examines how losses are determined and recorded. Following Meuwissen et al. (1999), losses associated with disease out- breaks may be classified into direct losses and consequential losses (van Asseldonk et al. 2003 and 2006; Meuwissen et al. 1999). To these categories we add “indirect” losses outside the farm sector. Direct, Consequential, and Indirect Losses Direct losses comprise the value to farmers of the animals culled under depopulation and welfare control measures, and the disposal and disinfection costs to farmers associated with the outbreak. The direct cost to the public sector of organizing culling is not counted as an item to compensate farmers for, and thus is not included in this discussion. Consequential Losses Consequential losses arise at the farm level as a result of culling but are additional to direct losses. They fall into one or more of the follow- ing categories. Business interruption: Occurs because farm buildings become wholly or partly unused due to stamping-out and welfare slaugh- ter or breeding prohibition, and stay empty until restriction zones are lifted. These costs are sometime referred to as “down time.” Losses related to established restriction zones: Farms in restric- tion zones face long periods in which animals and manure can- not be transported off the farm. These periods are characterized by animal welfare problems, extra feeding costs, and emergency measures for housing of pigs and storage of manure. Such losses will widely vary across farms and are difficult to assess. Additional animal repopulation costs: These losses include extra farm level costs of restocking beyond the cost of the stock itself. Losses from emergency vaccination: Where countries are engaged in poultry trade, meat from vaccinated animals may be ineligible • • • • Establishing Losses 3 Enhancing Control of Highly Pathogenic Avian Influenza in Developing Countries through Compensation�� for export to high price destinations and become redirected to lower priced domestic or regional markets. Price effects: Livestock epidemics can have a se- vere impact on meat prices. The impact depends on aspects such as the size of the epidemic (duration, size of restricted area), reactions of other countries (closure of borders, increased production) and whether vaccination is applied (which generally leads to long periods of export limitations). Indirect Losses These losses are incurred outside the on-farm sector, as in the case of lost feed sales, lost tourism, and lower economic activity in rural areas. Research shows that such losses are typically significant multiples of direct plus consequential on-farm losses (Thompson et al. 2002; World Bank 2006d). In 1999–2000, an outbreak of H7N1 HPAI in Italy resulted in US$122 million in compensation for destroyed birds, it was estimated that indirect costs exceed US$400 million for a total cost of over US$512 million (Halvorson et al. 2003). Types of Losses That Compensation Schemes Typically Take into Account Direct farm-level losses are typically partly com- pensated by governments (whether these are local, national, or regional agencies, such as the European Commission (EC), is discussed in Chapter 6). The types of losses that have always been compensated are for the animals that have been culled (includ- ing breeding stock, eggs, and semen). Schemes rare- ly cover the full value of rare or precious animals such as grandparent stock or fighting cocks. Some countries in their compensation scheme include direct farm level costs associated with culling such as cleaning, disinfection, and disposal, but most developing countries do not. Consequential losses are almost always completely borne by the farmers themselves if not insured privately. Indirect losses are never compensated directly as part of an animal health program, but may be covered in part by insur- ance in countries with developed insurance markets. Direct Losses Compensated in the EC In Europe, compensation of direct losses is partly based on EC directives for OIE List A diseases.9 Direct losses in these cases include the value of the animal destroyed, welfare control measures, and—in this case from the developed world—cer- • tain organizational aspects such as farm-level costs associated with monitoring of farms in restriction zones. Compensation of direct losses is either based on a predetermined animal value or actual market value at the time of culling. Generally, 100 percent of direct losses such as animal value of producers are compensated. Other eligible direct costs may be (i) refunded from national budgets, (ii) cofinanced through public-private financing schemes where farmers pay a compulsory levy, or (iii) incurred by the producer (Riviere-Cinnamond 2004, 2006). Animals culled by the authorities can be directly recorded on a list made at the time of culling by the culling team or by an accompanying recording team. The number, type, age, and so forth, of animals are recorded according to predetermined criteria. The categories need to be agreed on in advance, the re- cords signed by more than one reputable signatory (including from the government veterinary team), as well as the animal owner, and copies of records kept in more than one place. The list may also be on the basis of records from premises culled and existing animals records, how- ever this is likely to apply only to large sector 1 and 2 operations that are more likely to be registered and have reliable records on animal numbers, types, ages, and possibly weights, as well as on other ani- mal products such as eggs. The income appraisal ap- proach used in the outbreak of low pathogenic avi- an influenza (LPAI) in Virginia implicitly took into account the cost of inputs. The approach involved calculating the profit of meat bird production and then adding total meat cost of production (Ott 2006). Alternatively, losses can be based on the record of the total weight of animals when culled. However, in Laos this only applied to very small flocks. In Scotland, the level of compensation for ani- mals slaughtered is to some extent dependent on the nature of the disease and the disease status of the animals. Under some compensation schemes (for example for brucellosis and BSE for some suspects and beef animals), the level of compensation may be less than the animals’ actual market value, while un- der other schemes, such as bovine tuberculosis (TB), reliance on valuations of individual animals before slaughter has sometimes resulted in compensation payments that significantly exceed the market value of the animals. The current system, requiring on-site valuation of cattle also affects the efficiency of dis- ease control measures. Delays in removing infected animals may be caused by the need to arrange and agree upon individual valuations. Increased effi- ciency and the speed of settling compensation will help to reduce the spread of the disease. It will also Establishing Losses �� reduce the risk of animals dying before a valuation can be carried out, in which case farmers receive no compensation (Executive Environment and Rural Affairs Department 2004). However, on-site valu- ation is mainly applied for large livestock, not for small animals such as poultry. Lessons from Vietnam’s Experience There were differences across provinces in records used for culling during the 2003/04 outbreak in Viet- nam. Ha Tay province had a census of all animals in the province, and could cross-check the number of animals each farmer claims to have culled. The pro- cess was made public so that neighboring farmers could corroborate these numbers. In An Giang prov- ince, where poultry production concentrates mainly on duck production, there were problems in deter- mining the number of animals actually culled. Of- ten data provided by farmers did not match the data collected by the authorities. Authorities claimed that farmers mis-stated numbers and categories so as to obtain higher levels of compensation. As a result, recommendations arose to change the approach to supervision of culling, to include the fol- lowing members: head of the village, the veterinari- an covering the village, a representative of the wom- en’s union (and other existing unions), and farmers whose animals were being culled. It was suggested that the data gathered at village level should then be sent to the district, and consequently to the province authorities (Riviere-Cinnamond 2005). Losses due to unproductive herds and flocks tend not to be compensated; however they are some- times compensated under the process of “welfare culling.” This is not an infrequent problem. When a movement ban has been imposed for a long time and farmers are unable to continue feeding animals, they may be officially culled and compensation paid. This in fact was observed in parts of Vietnam. Disposal and Disinfection Costs The costs associated with the disposal of culled in- fected flocks and farm disinfection can be large. In some countries, such as the United States, the gov- ernment does the culling and disposal of the animals to ensure it is done and done right (Ott 2006). Thus these costs are not borne by the farmer. In Gaza, the Palestinian Authorities contracted out to the farm- ers the culling of their own animals. How this was monitored is not clear. Some compensation schemes in richer countries, where the government does not do these tasks di- rectly, cover farmers’ expenses for the cleansing and disinfection of buildings and disposal of carcasses. In Canada, owners of animals ordered destroyed are awarded compensation for disposal costs including transportation, slaughter, cleaning and disinfection, labor, and equipment.10 Interestingly, the central government of Vietnam has allocated some US$0.84 million to state-owned enterprises to cover veteri- nary services (77 percent) and disinfection (23 per- cent), but individual farmers are not compensated (Riviere-Cinnamond 2005). There is some question as to whether compen- sation for disinfection cost should be part of the compensation scheme, particularly in countries where there are limited funds. Government does not necessarily want to get in the habit of compen- sating for normal biosecurity measures. Cleaning is part of the normal biosecurity poultry management process and deep cleaning of the litter tends to also take place after one or more cycles, depending on the local practices. In any event, compensation for disposal and disinfection should not be at the cost of reducing the rate paid to owners of the birds Sick and Dead Animals In developing countries, animals are killed fairly fast when they are sick in sectors 3 and 4, and ei- ther eaten or sent to the market. In developed coun- tries, it is generally considered good practice to compensate sick birds less than healthy birds on the grounds that sick birds are worth less in the market and that a lower payment also provides an incentive for early culling. This is more difficult under sector 3 and 4 conditions in developing countries. Once birds begin to exhibit symptoms of HPAI, they are dead within 72 hours, and official response teams may not visit remote farms within that time frame. Most schemes globally do not cover losses from animals dead by the disease before official culling. This is justified sometimes on the belief that dead birds have no market value. It is also designed to create a strong incentive for early reporting and dis- courage transportation of infected carcasses among zones. This is especially necessary in the absence of good record keeping in sectors 3 and 4, since there is little documentary proof that diseased animals actually originated on the farm in question. Fur- thermore, poultry losses from other diseases such as Newcastle’s Disease are as high as 30 percent in sec- tors 3 and 4 of many developing countries, and it is costly to ascertain the cause of death after the fact. The assumption that dead birds have no market value does not stand up in sector 4 of many develop- Enhancing Control of Highly Pathogenic Avian Influenza in Developing Countries through Compensation�6 ing countries, where dead and dying animals are of- ten sold. Furthermore, there is a credible likelihood that some birds will be dead by the time that official culling teams arrive at the farm, even if the disease was reported at first appearance of symptoms. Because disease control is the overarching objec- tive, it is better when dealing with sectors 3 and 4 in developing countries to compensate partially for dead animals, but less than for healthy ones and sick ones. Losses Covered by Private Insurance or Public-Private Part- nership in Developed Countries As mentioned above, compensation schemes rarely if ever cover consequential losses. However, farm- ers in richer countries have typically mitigated their consequential livestock disease risks through a va- riety of public and private mechanisms that may eventually be useful for sector 1 and 2 operations in developing countries. Options used in Europe, for example, are: (i) private insurance schemes, (ii) free public disaster assistance, (iii) and public-private partnerships (Riviere-Cinnamond 2004). Private Insurance Private insurance schemes in Europe exist for cer- tain types of consequential losses from livestock disease, as in the United Kingdom, the Netherlands, and Germany. In Italy, private insurance schemes are exclusively for dairy production and sheep, and participation is very limited (less than 5 percent). In the Netherlands, farmer participation is less than 10 percent, and in Germany, participation levels vary depending on the type of animal (for dairy cows less than 50 percent, cattle 30 percent, sows 42 percent, and hogs 23 percent). All private insurance policies exclude direct losses, which are separately cov- ered by the government with partial EC support. Losses due to business interruption and in re- striction zones are in some cases covered. These are compensated through set daily rates for the duration of the period of restrictions. Losses as a result of movement prohibition are covered by private insurance schemes in Germany, but not in the Netherlands. Public Disaster Assistance Free public disaster assistance exists in Finland and France. Although funds are public, the scheme is administered by private insurers on a no-risk com- mission basis. There is therefore no risk transfer to private insurers. In France, consequential losses due to FMD are covered in a declared outbreak, but the funds for compensation are taken from a fund accumulated in prior years from farmers’ contribu- tions (€0.33 per livestock unit per year), with an un- derlying guarantee by the central government. Public-Private Partnerships Finally, public-private partnerships for the com- pensation of consequential losses exist in Denmark, Finland (to some extent), and Spain (through a structure named Agroseguro cofinanced by the gov- ernment). Under this type of partnership, the gov- ernment may act as an insurer or reinsurer. Under these schemes, the risk of losing financial resources is transferred to the private insurers. For example, in Denmark, the government pays 20 percent above the value of the animals culled to cover consequen- tial losses. In Spain, the government covers conse- quential losses up to a maximum of 41 percent of the value of culled animals in the case of cattle and 32 percent in the case of sheep and goats. Establishing Losses �� Conclusions and Recommendations from Chapter � Losses stemming from outbreaks of animal disease fall into three categories: direct (birds, eggs, sometimes the cost of disposal and disinfection), consequential (other on- farm costs or lost farm income due to the outbreak), and indirect (all other losses, off farm). Compensation typically has been limited to direct losses, although private and pub- lic-private schemes exist to defray farm risk of consequential losses in the developed countries. The types of direct losses that have always been compensated are for the animals that have been culled (including breeding stock, eggs, and semen). Some countries also include the direct cost associated with culling such as cleaning, disinfection, and disposal, though strictly this is not a cost eligible for compensation. Care should be taken that such costs are not mixed in with normal biosecurity practic- es. There is occasional compensation associated with welfare killings, but the amount tends to be at different levels. In addition, if there are cases where production is well recorded as in sectors 1 and 2, there may be a good case to compensate for dead birds associated with a disease. If covered at all, consequential losses tend to be covered under animal insurance schemes. Compensation for consequential losses can ideally be based on actual losses incurred or a flat rate based on the estimated loss over the period of business interrup- tion. The options are (i) private insurance schemes, (ii) free public disaster assistance, (iii) and public-private partnerships. Indirect losses occur outside the livestock sector as a result of disease outbreak. They include diverse items such as backward linkages to input suppliers, lost tourism, and lower demand for local products in income-depressed zones. Generally these losses are several times greater than direct and consequential losses combined, but they are never compensated as part of disease control strategies. Though it is not the scope of this paper, more work is needed on understanding these alternative mechanisms for sharing risk from consequential and indirect losses from animal diseases that are not normally covered by compensation schemes. �� Introduction This chapter deals with determining compensation rates and when it should be paid. Setting the right price and paying it at the right time are of critical importance. Prices need to be high enough to encourage farmers to engage in early and complete reporting of the disease. On the other hand, they need to be low enough to avoid encouraging farmers from still disease-free areas to present their animals to be culled, or others to move potentially sick birds across zones in hopes of receiving compensation in excess of prevailing market prices. Both of the latter behaviors have been observed in the field. Compensation prices are derived from applying a mandated percentage rate to a price series denominated in currency units per item compensated; the latter series may be collected from actual markets, adjusted down to the farm gate, and averaged over preset periods, to provide flexibility. Getting compensation prices right is tricky in an environment where normal market prices typically vary greatly across space and season, as they do in the best of times in most developing countries. The key issue is to provide an incentive for timely reporting in one’s own zone. Unfortunately, there is little existing work confirming empirically that high compensation prices are positively associated with high reporting, even if this is what we would predict. There is even less rigorous empirical evidence link- ing high compensation prices to effective disease control, although there is now a body of case studies suggesting that paying too little is linked to disease spread. Approaches for Setting Compensation Rates There are three main ways in practice in which countries have set compensation prices. The Ad Hoc Approach Based on Resource Availability Several of the developing countries afflicted with outbreaks of HPAI investigated for this report appear to have based compensation per bird on an estimate of total funds available for compensation—typi- cally limited due to budgetary constraints—divided by an estimate Setting the Level and Timeliness of Compensation 4 Enhancing Control of Highly Pathogenic Avian Influenza in Developing Countries through Compensation�0 of the total number of birds to be culled. Sometimes this led to very low percentages of the market value of birds being compensated. The Cost-of-Production Approach The cost-of-production approach has been applied in some countries, and determines the price of a product by summing the unit costs of the resources that went into making it. The cost can be composed of the cost of any of the factors of production includ- ing labor, capital, feed, breeding stock, or technology. Countries deciding to base the compensation price on the costs of production decide so because either they do not have a market price for the ani- mal (that is, there is only layer production hence the animal is not traded at the end of its productive life; this is the case of Kosovo) or the countries are not infected and have the time to establish those rates and categories during “peace time” (this is the case of Mauritania and Senegal, for example). It is rec- ommended, however, that this practice should be limited to those rare cases where no market prices are available. The compensation rate obtained from the cost-of- production approach at times can be higher than the market price. The method is subject to measurement errors, possibly some of them deliberate. More im- portantly, by its nature, it rewards inefficient pro- ducers or categories of producers. Finally, it tends to reward record keeping, which introduces a bias toward large commercial operations that is inconsis- tent with the principal of providing an incentive for disease control in sectors 3 and 4. The Market Value Approach In principle, the market value approach sets the compensation price at the level the good in question will fetch in the market place, adjusted back to the point at which the decision to sell is made (in this case, the “farm gate”).11 Usually rates are set based on an average set of prices received during the last several proceeding months, and thus prior to the dip in prices usually observed during an outbreak. The present report advocates a market-based ap- proach. Moreover, considerable difficulties exist in matching prices to goods, regions, and time periods, as will be discussed below. Who Defines the “Market” Price? When the compensation is paid by the public sector, government institutions normally define the com- pensation levels. Private industry in countries with large sectors 1 and 2 are invariably active in lobby- ing in this regard, both in terms of direct interest and—often just as important—to promote disease control essential to regaining export certification. National poultry associations often are involved, as in the case in Gaza and Mauritania. In Australia, Canada, and the United Kingdom, rates are often set on the basis of technical considerations supplied by independent certified evaluators, using govern- ment guidelines. In developing countries, public sector institutions generally define the rate of com- pensation. Two issues that are especially important are discussed in the following sections. The Level of Decision Making Simplicity is a key requirement for a compensation system, as it reduces transaction costs and the pos- sibility of rent-seeking behavior of farmers. The es- tablishment of uniform procedures to determine the level of the rate is generally recommended. Uniform levels of compensation for the entire country are also often recommended for the same reason. However, if there are major differences be- tween different regions in poultry market prices, some geographical differentiation might be needed to avoid transport of healthy birds from lower-priced regions into the disease area where compensation is paid. Indonesia for instance has proposed introduc- tion of a variable rate to take account of geographi- cal price differences in view of concerns over the possible movement of birds. The regional differen- tiation concern is even more true when regions in one country border on another country with little effective border control, as often happens in remote areas of developing countries. A border province in one country may have market prices that are much closer to the border region in the adjacent country than to other regions of its own country. Imposing a uniform national price would then encourage cross- border smuggling of both healthy and disease stock for culling. The Degree of Involvement in the Stakeholders Public awareness and broad-based ownership are important aspects of a successful compensation scheme. These can be achieved through an involve- ment of all stakeholders in the setting of the price, as is the case in Australia, for example. Where the com- pensation is funded by the private sector (levies), such as in the Netherlands, companies also have a Setting the Level of Timeliness and Compensation �� major say in deciding the level, although the govern- ment has introduced caps. There are few examples of strong private-sector stakeholder involvement in setting the price in developing countries, although attempts have been made, such as in an East Asian country. The independence of the evaluators is critical, in particular in the absence of good marketing informa- tion. The assessed values of animals rose threefold after the outbreak of the FMD in the early 2000s in the United Kingdom (NAO 2002). With no function- ing animal markets, the government lacked a clear frame of reference to assess or influence the valua- tions against which compensation was paid, evalu- ators were appointed in collaboration with parties whose stock was being culled, and there is at least anecdotal evidence that valuations were signifi- cantly too high. However, local poultry associations should still be consulted to ensure effectiveness and transparency in the price-setting process. Importance of Remaining Flexible to Chang- ing Market Situations Setting prices without paying attention to what is happening to market prices may lead to the move- ment of poultry from one area to another to reap the benefits of higher compensation practices. Market conditions can also change radically over time. To prevent perverse incentives (either inappropriate presentations for culling when the rate is too high or not reporting when it is too low) it is important to blend consistency with flexibility when setting rates. One way to do this is to set rates in terms of a percentage of an easily verifiable market price that is adjusted periodically (like an index interest rate in the adjustable mortgage market of developed countries). Where such a series does not exist (most countries in the world have a retail poultry price se- ries for the capital city), it will be necessary to set up a mechanism for keeping such a series as soon as possible (discussed further below). The principle then should be to derive an appropriate adjustment fact downward from the urban retail price series (if that is the one kept) to the farm-gate level in each province or major region. Choice of Price Baseline to Establish Compensation Rates The Animal Epidemic Act of Thailand provides farmers a compensation of 75 percent of the farm- gate value of animals that are destroyed. A 100 per- cent compensation was provided during the early epidemic (January–May 2004) because the epidemic was widespread and devastating to Thai farmers. In the second wave (July–Dec 2004), compensation was reduced to 75 percent. At this time, compensa- tion per bird was (in U.S. dollars) US$0.38–US$0.65, depending on the type of poultry (US$0.38 for quail; US$1.13 for broiler; US$2.00 for meat duck; US$2.25 for backyard chicken; US$3.50 for layer chicken, lay- er duck, or goose; US$7.25 for turkey; and US$65.00 for ostrich) (Tiensin et al. 2004). Generally, compen- sation rates are set for the class of animals that has the largest representation at market, that is, full- grown broilers and layers. Price Information by Sector and Seasonality Developed or middle-income countries tend to have well-established monitoring systems of the fluctua- tions of prices over time, and products are fairly uniform. However, products vary more in develop- ing countries across production sectors, and there is much less price information for the poorer sectors. For instance in Indonesia, sectors 1 and 2 have solid pr