Linking Farmers to Markets through Productive Alliances An Assessment of the World Bank Experience in Latin America November 2016 © 2016 International Bank for Reconstruction and Development / The World Bank 1818 H Street NW Washington DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org This work is a product of the staff of The World Bank with external contributions. The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors, or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Rights and Permissions The material in this work is subject to copyright. Because The World Bank encourages dissemination of its knowledge, this work may be reproduced, in whole or in part, for noncommercial purposes as long as full attribution to this work is given. Any queries on rights and licenses, including subsidiary rights, should be addressed to the Publishing and Knowledge Division, The World Bank Group, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2625; e-mail: pubrights@worldbank.org. Photographys from World Bank Projects and Shutterstock Concept and design: Jaime Sosa Linking Farmers to Markets through Productive Alliances An Assessment of the World Bank Experience in Latin America November 2016 TABLE OF CONTENTS Acknowledgements ix Abbreviations x Executive Summary 1 1. Introduction 5 2. The Productive Alliance Approach 9 3. Productive Alliances in Latin America and the Caribbean 13 4. Targeting Strategies: Geographic, Value Chains, Beneficiaries 21 5. Subproject Preparation and Selection 29 6. Subproject Financing 33 7. Results Monitoring 39 8. Achievements and Outcomes 41 9. Evaluation Strategies 49 10. Lessons Learned 55 11. Conclusions and Recommendations 59 Annex 1 Descriptions PA Projects in Latin America and the Caribbean 62 Annex 2 Key PA Project Characteristics 65 Annex 3 Targeting Strategies 69 Annex 4 Subproject Selection Process 71 Annex 5 Subproject Financing 74 Annex 6 PA Project Monitoring and Evaluation Systems 76 Annex 7 Results Frameworks 79 Annex 8 Productive Alliance Project Outcomes and Impact Evaluations 83 Annex 9 Theory of Change for Productive Alliance projects 93 Annex 10 Selected issues and possible solutions 95 References 98 LIST OF FIGURES Figure 2.1 Core Elements of a Productive Alliance 12 Figure 3.1 Geography of Productive Alliance Projects in Latin America 14 Figure 3.2 Stages of beneficiary producer organizations 17 Figure 3.3 End Market Levels 18 Figure 4.1 Distribution of rural population and smallholder producers 24 Figure 5.1 Steps of the Subproject Selection Process 31 Figure 8.1 Main areas of PA project impact 43 LIST OF TABLES Table 3.1 Conditional versus unconditional Productive Alliance projects 15 Table 3.2 Areas of emphasis 16 Table 3.3 Formality of Commercial Agreements 18 Table 4.1 Geographical Coverage of Productive Alliance Projects 22 Table 4.2 Targeting of Value Chains 23 Table 4.3 Eligibility criteria for Beneficiary Producers 25 Table 8.1 Financial Rates of Return 46 LIST OF BOXES In Practice Box 1: Eligibility Criteria for Targeting Buyers 28 In Practice Box 2: Preparation and Selection Criteria for Subproject Business Plans 32 In Practice Box 3: Institutional Arrangements of PA Subproject Selection Process 32 In Practice Box 4: Leveraging Subproject Funding through broader Alliances (Colombia) 36 In Practice Box 5: Upfront Engagement of the Commercial Financial Sector (Honduras) 37 In Practice Box 6: Revolving Funds as an Alternative Financing Mechanism (Colombia) 38 In Practice Box 7 Productive Alliances as a Tool for Effective Social Inclusion (Colombia) 44 In Practice Box 8: Dissolution versus Failure of a Productive Alliance (Colombia) 46 In Practice Box 9: Factors of Sustainability (Bolivia and Colombia) 48 In Practice Box 10: Measuring Successful Productive Alliances (Brazil Pernambuco) 52 In Practice Box 11: Building Partnerships for Obtaining Evidence 53 In Practice Box 12: Learning from Different Evaluation Methodologies (DIME) 54 LIST OF ANNEX FIGURES Figure A6.1 M&E System Example 1: Colombia PAAP 78 Figure A6.2 M&E System Example 2: Mexico SPSB 78 Figure A9.1 Theory of Change Logic 93 LIST OF ANNEX TABLES Table A2.1 Productive Alliance Projects in Latin America and the Caribbean 65 Table A2.2 PA Stand-alone versus PA Component 66 Table A2.3 Target Areas and Direct Beneficiaries 67 Table A3.1 Producer Eligibility Criteria for Subproject Proposals 69 Table A3.2 Producer Organization Requirements 70 Table A4.1 Institutional arrangements subproject selection 71 Table A5.1 Financial Support to Producer Organizations 74 Table A5.2 Financed Support and Cost Categories 75 Table A6.1 Subproject Monitoring 76 Table A7.1 Project Development Objectives and Indicators 79 Table A8.1 Scope of Projects 83 Table A8.2 Investment Support 83 Table A8.3 Key Outcomes and Impacts of Completed PA Projects 84 Table A9.2 Prototypical Theory of Change for Productive Alliance Projects 94 Linking Farmers to Markets through Productive Alliances ACKNOWLEDGEMENTS The preparation of this report has been led by Alexandra Christina Horst (Junior Professional Officer, Agriculture Global Practice) in close collaboration with Maria del Mar Polo (Agricultural Economist, Food and Agriculture Organization) and with guidance from Laurent Msellati (Practice Manager, Agriculture Global Practice), Michael Morris (Lead Agriculture Economist, Agriculture Global Practice) and David Tuchschneider (Senior Rural Development Specialist, Agriculture Global Practice). It has been peer-reviewed by Martien van Nieuwkoop (Practice Manager, Agriculture Global Practice), Panos Varangis (Head, Finance and Markets), Loraine Ronchi (Lead Economist, Trade and Competitiveness Global Practice), and Mark Lundy (Senior Researcher, International Center for Tropical Agriculture). The report was edited by Anna Roumani (Consultant, Agriculture Global Practice). The report has also benefited greatly from detailed comments, discussions and expert consultations within the World Bank Agriculture Global Practice, specifically from Fatima Amazonas (Senior Rural Development Specialist), Diego Arias (Senior Agriculture Economist), Marie-Helene Collion (former Lead Agricultural Specialist), Edward Bresnyan (Senior Agriculture Economist), Pierre Olivier Colleye (Senior Microfinance Specialist), Luz Diaz (Senior Agribusiness Specialist), Svetlana Edmeades (Senior Agriculture Economist), Barbara Farinelli (Agriculture Economist), Marianne Grosclaude (Lead Agriculture Economist), Steven Jaffee (Lead Agriculture Economist), Willem Janssen (Lead Agriculture Economist), Roy Parizat (Senior Economist), Norman Piccioni (Senior Rural Development Specialist), Anna Roumani (Consultant), Eli Weiss (Senior Rural Development Specialist), and Pierre Werbrouck (former Lead Agriculture Economist). Furthermore, comments and contributions have been received from colleagues within the World Bank and other development and research institutions: Luis Dias (Economist, FAO), Dino Francescutti (Senior Economist, FAO), Frank Hollinger (Economist, FAO) Barbara Jordan Vicente (Consultant Junior Economist, FAO), and Astrid Zwager (Research Analyst, World Bank Development Impact Evaluation Unit). Special recognition is extended to the Project Management Units of all Productive Alliance projects analyzed in this assessment for providing crucial inputs and documentation to this report. ix ABBREVIATIONS AF Additional Financing ALIADOS Sierra Rural Development Project (Peru) CCRB Coordinating Unit for Biological Corridors and Resources CDD Community-Driven Development CfP Call for Proposals CIAT International Center for Tropical Agriculture CLARs Local Resource Allocation Committees COMRURAL Rural Competitiveness Project CONABIO National Commission for Knowledge and Use of Biodiversity COSUDE Swiss Agency for Development and Cooperation COTESIP Technical Council for Sustainable Production Systems EDOs Enterprise Development Officers IRR Internal Rate of Return FPS National Fund for Social and Productive Investments GDP Gross Domestic Product GEF Global Environmental Facility IADB Inter-American Development Bank IBRD International Bank for Reconstruction and Development ICC Investment Coordination Committee ICR Implementation Completion and Results Report IDA International Development Association IFI Intermediate Financial Institution IRR Internal Rate of Return JSIF Jamaica Social Investment Fund M&E Monitoring and Evaluation MADR Ministry of Agriculture and Rural Development (Colombia) MIDA Ministry of Agricultural Development (Panama) MINAG Ministry of Agriculture (Peru) MINECO Ministry of Economy (Guatemala) MRC Management Review Committee MTR Mid-Term Review NGO Non-Governmental Organization x Linking Farmers to Markets through Productive Alliances NIC National Inter-sectoral Committee NPAS National Protected Areas System OGA Local Management Organization OGR Regional Management Organization PA Productive Alliances PAAP Productive Alliance Support Program (Colombia) PAR Rural Alliances Project (Bolivia) PDER Project to Support a Rural Economic Development Program (Guatemala) PDO Project Development Objective PCU Project Coordination Unit PO Producer Organization(s) POA Annual Operational Plan PPAR Project Performance Assessment Report PRORURAL Rural Productivity Project (Panama and Pernambuco, Brazil) REDI Rural Economic Development Initiative (Jamaica) REEC Rural Economic Evaluation Committees RIC Regional Inter-sectoral Committee(s) RPU Regional Project Units SDEs Entrepreneurial Services Providers SEGEPLAN Presidential Secretariat for Planning and Programming SENA National Apprenticeship Service (Colombia) SIGG Georeferenced Management Information System SIGIEP Project Monitoring System - Sistema Integral de Gestión de Información y Evaluación del Proyecto (Mexico) SNIP National Public Investment System (Peru) TA Technical Assistance xi EXECUTIVE SUMMARY i.  The World Bank Agriculture Global towards the producers’ needs: productive and hence revenues. In addition, Practice has identified “linking farmers investments, technical assistance, and producers also value the opportunity to markets” as a major development business development. These core to obtain technical assistance, improve challenge in recognition of the obstacles inputs are financed through public their negotiating power, and receive faced by smallholder producers as they grants provided by the project, which are payment promptly from the buyer(s). seek to compete in rapidly expanding matched by the beneficiary producers vi.  The basic concept of the modern agricultural value chains. As and in some cases also by the buyer(s). Productive Alliance approach is simple, laid out in its Agriculture Action Plan iv.  The financial support provided to and it has proved sufficiently flexible (2013-2015), the Agriculture Global Productive Alliance subprojects usually to adjust to a wide range of market Practice is working to improve the comes in the form of matching grants, realities and policy objectives. By competitiveness and entrepreneurship which are justified by the positive design, the PA approach resolves multiple of smallholder producers in developing externalities that are generated by the constraints in a simultaneous and tailored countries by strengthening value chains. subproject and the fact that commercial fashion by providing integrated solutions ii.  One approach to address this finance is often not available to rural that are adapted to local conditions. It challenge has been to promote smallholders. The levels of grant support has shown its flexibility across PA projects, Productive Alliances (PA) that and the arrangements for co-financing having demonstrated the ability to strengthen the linkages between among the alliance partners have adapt to differences in policy priorities, producers, buyers and the public varied considerably across PA projects; market opportunities, and countries’ sector within agriculture value- experience suggests that satisfactory economic conditions. In favorable chains. The PA approach provides outcomes are far more likely when a enabling environments, PA projects holistic solutions to address market minimum of 30% co-financing of the total benefit from complementary support imperfections that inhibit smallholder subproject costs is required of beneficiary systems from public and private sector producers’ socio-economic progress. producers. While many PA projects agents, encouraging a harmonization have required only in-kind contributions of public and private services in CONCEPT from producers, cash contributions line with the overall PA project’s encourage greater ownership of the objectives of improving smallholder iii.  A Productive Alliance involves supported subproject, ensure enhanced production and market integration. three core agents: a group of risk-sharing, and require greater smallholder producers, one or more buyers, and the public sector. These commitment on the part of producers. IMPLEMENTATION three agents are connected through v.  The design of the Productive EXPERIENCE a business proposition, or “business Alliance approach encourages vii.  The Productive Alliance approach plan”, which describes the capital and the development of two types of was introduced during the early 2000s services needs of the producers and productive alliances: (i) a horizontal in Latin America and the Caribbean proposes improvements that would alliance among the producers and (LAC). Since then, the World Bank has allow them to upgrade their production most importantly (ii) a vertical alliance provided more than US$1 billion in capacities and skills to strengthen between the producers and the financing to support 21 projects with their linkage with the market, i.e. the buyer(s). Major motivations identified by over 3,500 subprojects in ten countries buyer(s). The implementation of such both producers and buyers for joining across the LAC region. Projects a business plan through a subproject a vertical alliance have been increased promoting PA have also been introduced is typically supported through three stability in prices, assured sales, as well in countries in Sub-Saharan Africa and core inputs and/or activities directed as improvements in product quality East Asia. This widespread adoption is 1 Linking Farmers to Markets through Productive Alliances based on increasing evidence suggesting well as in the levels of product quality commercial finance, but these efforts - subject to further, intensified formal requirements they have pursued. Finally, have rarely been successful. A limited evaluation - that the PA -approach can PA projects have varied in the degree number of PA projects have successfully lead to increases in productivity, market of formality they have demanded with engaged the commercial financial sector integration, production, sales volume, respect to the commercial agreements from early on by making them part of value-addition, prices, and income of struck between producers and buyers; the alliance and subproject co-financing smallholder farmers, while generating these have ranged from formal written structure. However, most have not been on-farm and non-farm employment, contracts to more informal agreements. able to realize their intentions. The improving the quality of jobs, as well reasons for this are manifold and include x.  Based on differences in project as the inclusion of vulnerable groups. low coverage of financial services in objectives and expected outcomes, rural areas, lack of collateral from the viii.  The subproject preparation Productive Alliance projects in LAC smallholder producers, regulatory issues and selection process follows a have adopted a variety of targeting that prevent financial institutions from common procedure across Productive strategies regarding geography, making loans to groups of producers Alliance projects: First, a public agricultural value chains, and (and not individuals), among others. information campaign is carried out beneficiaries. For instance, most PA to raise awareness among potential projects have focused on specific beneficiaries. Subsequently, a call for priority areas in a given country based ACHIEVEMENTS initial subproject proposals is launched on socio-economic, demographic xii.  The generally positive to invite interested producers to submit and market criteria to ensure effective achievements of Productive Alliance initial subproject proposals. Next, the geographic targeting. PA projects have projects are assessed in this report in initial subproject proposals are screened financed subprojects with activities in terms of: (i) scope, (ii) social inclusion, and evaluated against predefined both agricultural production and rural (iii) socio-economic impacts, (iv) eligibility criteria. Viable proposals are non-farm economic activities, although efficiency, and (v) sustainability. then elaborated into more detailed the former have predominated. In terms Scope: Almost all Productive Alliance subproject business plans, often with of beneficiaries, PA projects generally projects in Latin America have the support of service providers. have targeted “transitional smallholder exceeded their appraisal targets for Finally, subproject proposals which producers” who lack well-established the number of alliance subprojects. meet (a minimum of) the established linkages to buyers and markets but have However, they have generally fallen criteria are approved for financing. the potential and willingness to increase short on the expected number of their productive and entrepreneurial ix.  Across Latin America, many beneficiary producers per producer capacities to engage in modern agri-food different variations have appeared organization. Another important markets. PA projects have established of the Productive Alliance approach, finding regarding beneficiaries is that transparent and technical criteria for distinguished in terms of their areas indirect beneficiaries of PA projects selecting beneficiary producers, to of emphasis, types of beneficiary are largely unaccounted for and have ensure credibility and a fair selection producers, end markets and formality been measured only in a few cases. process. With respect to buyers, targeting of commercial agreements between financially and commercially strong buyers Social inclusion: Productive producers and buyers. Generally, and ensuring their commitment to an Alliance projects have performed improving market integration and alliance with the producers are critical. well in including women and other competitiveness are the main areas Very few PA projects have established disadvantaged groups, such as of emphasis of Productive Alliance eligibility criteria for buyers, however, indigenous peoples or smallholder projects. Depending on the sector and which has been identified as an area for producers in post-conflict zones. the prevailing market structures, PA improvement for future PA interventions. Based on project evaluation results, projects have supported subprojects the PA approach appears to be an that work with producers at different xi.  To leverage Productive Alliance effective tool for social inclusion, as: stages of organization. Similarly, PA subproject financing, almost all (i) the participation of disadvantaged projects have differed in the types of Productive Alliance projects have groups in PA projects is high, end markets they have targeted, as aimed to enhance producers’ access to sometimes exceeding the proportion 2 of these groups in the overall and persist over the longer term, with based on a realistic assessment population; and (ii) the performance most alliances continuing to operate of the local/regional comparative of alliances involving these groups after project support has ceased. advantage in specific value chains. compared to alliances involving Sustainability: Productive Alliance »» Productive Alliance projects involving producers from non-disadvantaged projects have promoted longer- a multi-sector approach with different groups has been equally good term vertical alliances between institutions, components, and areas and in some cases even better. smallholder producers and buyers. of concentration are challenging to Socio-economic impacts: Productive Increasing evidence shows that a implement. First-time operations Alliance projects have generated significant portion of established with newly formed implementation significant positive impacts in alliances continue operating also agencies should be kept simple, use production, sales, income, and after subproject support ends. piloting and evaluation for future employment. Evidence from surveys However, as for development scale-up, and receive close supervision and evaluations of several projects operations in general, only few so that the learning opportunity is has shown that the PA approach has ex-post evaluations are done more optimized and the implementation led to increases in production volume, than two years after overall project agencies are capable of entering a productivity, access to improved inputs completion, leaving most assessments follow-on phase with confidence. and productive equipment, as well of sustainability limited to one or two »» Establishing assessment criteria for as to integration into new markets. years after subproject completion. Moreover, beneficiary producers the selection of potential buyers and regular monitoring of the producer- benefit from better product quality and LESSONS LEARNED diversification and hence increased buyer relationship can reduce the risk xiii.  Key lessons emerging from the of choosing an uncompetitive buyer or sales volume and prices. Specifically, Productive Alliance experience in promoting elite capture in imperfect increases in sales have ranged between Latin America include the following: markets. Measures to enhance the 20% and 60% and the average net income of beneficiary producers has continuity of a vertical alliance include »» Setting up a competitive subproject been around 30% higher compared improved identification of buyers selection process based on clearly to control groups. Moreover, some and their market competitiveness defined technical evaluation PA projects have been found to have and viability, brokerage services for criteria is crucial for establishing led to improvements in employment organizations that outgrow their credibility among stakeholders, and the generation of new job, but current alliance arrangements, and avoiding political interference, and the available evidence on jobs is an increased outreach to potential safeguarding the technical quality of based on relatively small samples and buyers outside the local spheres. varies strongly across value chains. selected subprojects. A competitive process ensures that subprojects »» Requiring cash contributions or Efficiency: Most Productive Alliance are approved based on the merit bank loans as co-financing from projects have generated satisfactory and viability of their business plan, producers can ensure a stronger average rates of return at the leaving few opportunities for financing buy-in. The provision of financial commonly assumed discount rate to be steered by government resources which are not in-kind of 12% and 10 year estimation officials or local authorities based creates a greater ownership of the period. Returns have often varied on non-technical criteria. success of the supported subproject greatly between products, however, and allows producer organizations to which underlines the importance of »» Identifying and analyzing promising strengthen their entrepreneurial skills accurately assessing the long-term value chains based on technical criteria, and commitment to the subproject. market potential of the products aligned with project objectives and to be financed under a PA project. market potential, are important to »» Building capacity of beneficiary Regarding sustainability, vertical ensure effectiveness and sustainability. producers over an extended period alliances often form between Such analysis should be done during while they grow and mature is smallholder producers and buyers project preparation and must be crucial for ensuring long-term 3 Linking Farmers to Markets through Productive Alliances success. Experience suggests that actors in the enabling environment »» Productive Alliance projects should accompanying beneficiary producer (e.g. financial, institutional, and incorporate an impact evaluation organizations – and buyers - over the educational). This will enable them strategy and dedicated budget course of their alliance subproject to tap into local knowledge and from the early design stage. Despite implementation period, rather than business opportunities and to better widespread application of the PA providing a one-off injection of integrate the alliance subprojects in approach and increasing evidence on resources, can build the capacity municipal development plans and results, limited large-scale rigorous needed to foster adaptation to related investments and services. and formal evidence exists so far specific and evolving business needs on its effectiveness. Because of this »» Productive Alliance projects should and ensure long-term survival. lack of evidence, several key design develop a systematic approach questions remain un-answered. »» More large-scale and representative linking beneficiary producers with the Hence, future PA projects should evidence is needed to substantiate commercial financial sector to leverage include relevant evaluation questions the results on socio-economic impacts financing. This could be done through in their results monitoring design and efficiency. Sound evaluation guarantees to stimulate the willingness and incorporate baseline and impact strategies and ensured budgeting of financial intuitions to co-finance evaluation data collection and for data collection and analysis can a significant part of the business analysis in their budget planning. bolster the evidence base supporting plan through a loan. Furthermore, use of the PA approach. Starting this PA projects could increasingly focus process at the project design stage on meeting other financing needs and rigorously following through of producers, such as payment increases the likelihood that a PA systems, insurance, or savings. project can demonstrate results and »» Productive Alliance projects should attribute them to the project. intensively promote the diversification »» Developing a plan for transitioning of buyers and markets. A more to post World Bank-financing diversified portfolio of offtake markets of Productive Alliance activities can serve as an instrument to increase ensures continuity without producers’ resilience to external disruptions for subprojects. This shocks and to protect producers from requires early definition of an exit exploitative behavior by buyers. strategy with transition planning »» Productive Alliance projects could beginning – at the latest - by the increasingly consider adopting time of the Mid-term Review and a multi-sector approach, where continuous monitoring thereafter. appropriate and desired by the client. In this manner, the generally socio- RECOMMENDATIONS economic development objectives of xiv.  Considering the achievements of the PA approach can be combined past and current Productive Alliance with high-level objectives such as projects and the lessons learned from environmental sustainability, social the implementation experience to inclusion, or improved nutrition date, five main recommendations outcomes. Given its flexibility in emerge that can strengthen and design and implementation, the PA enhance the Productive Alliance approach in future projects: approach is adaptable and can be used to address several objectives »» Productive Alliance projects should in a more holistic manner. seek to build broad alliances through a stronger involvement of local 4 1 1. Introduction > 2. Productive Alliances in Latin America and the Caribbean 5 Linking Farmers to Markets through Productive Alliances INTRODUCTION 1.  Across the developing world, may sell through intermediaries due to 2.  The World Bank Agriculture Global agriculture remains the primary source the small scale of their production, the Practice has identified “Linking farmers of income and employment for the high transaction costs they face, and to markets” as a major development majority of the rural poor. Most rural their inability to provide products of challenge in recognition of the obstacles households engage in smallholder consistent quality (Collion and Friedman, faced by smallholder producers. As laid agriculture, facing multiple challenges to 2011). This remains the case even out in the Agriculture Action Plan (2013- competing in rapidly expanding modern while numerous factors, including the 2015), the Bank is working to improve the agricultural value chains. Low productivity proliferation of supermarkets that began competitiveness and entrepreneurship stemming from lack of access to inputs, in the 1990s, have changed the market of smallholder producers in developing modern technologies, and credit, as structure with increasingly stringent countries by strengthening value chains. well as asymmetric information on prices contractual requirements on volume, One approach has been to promote and marketing opportunities undermine safety, quality, and timely delivery. The productive alliances through lending smallholder producers’ negotiation disadvantages faced by smallholder operations. The Productive Alliances power with buyers (Collion and Friedman, producers are major challenges facing (PA) approach strengthens the linkages 2011). As a result, many smallholders sell efforts to foster growth and reduce between producers, buyers, and the commodities mainly in traditional and poverty and inequality by integrating public sector within agriculture value- less profitable venues such as informal, them into these rapidly evolving markets. chains through the provision of core open-air markets. Alternatively, they 6 inputs such as productive investments, in Sub-Saharan Africa and East Asia. is to compare common factors of technical assistance, and business This widespread adoption is based on implementation across projects development. It promotes horizontal increasing evidence suggesting - subject as well as the various ways in which alliances among smallholder producers to more rigorous data collection and the PA approach has been applied to coordinate production and sell analysis - that the PA approach can throughout the region. This assessment collectively. It also provides incentives lead to increases in productivity, market is based on a desk review of project for the formation of a vertical alliance integration, production, sales volume, documents, such as Project Appraisal between producers and at least one value-addition, prices and income of Documents (PAD), Operational Manuals buyer for the provision of a good smallholder farmers, while generating (OM), Aide Memoires, Mid-Term in a specific value chain through a on-farm and non-farm employment, Reviews, project monitoring system commercial agreement, with the public improving the quality of jobs, as well databases, Implementation Completion sector playing the role of the convener as the inclusion of vulnerable groups. Reports (ICR), and other project files. who brings the parties together and Furthermore, it takes into account data 4.  The purpose of this assessment sets the ground rules/regulations. and information collected during project is to describe the Productive Alliance supervision missions, ex-post economic 3.  The Productive Alliance approach approach and its core elements,1 and financial analyses, and impact was introduced in the early 2000s as well as to present the results of evaluation results where available. Finally, in Latin America, a region in which a systematic review of World Bank- lessons learned and expert opinions on the World Bank had provided supported PA projects across Latin the PA approach have been collected approximately US$1 billion in America and the Caribbean. through workshops and seminars. financing to support 21 projects in 5.  With an in-depth assessment ten countries by the end of 2015. 6.  While the focus of this assessment of Productive Alliance projects, Projects promoting productive alliances was Latin America and the Caribbean, the methodology of this report have also been introduced in countries it provides lessons applicable to all 7 Linking Farmers to Markets through Productive Alliances developing regions. Learning from core elements. Section 3 presents 1. Project documents use different the experience of the PA approach is the experience of PA projects in Latin terminologies for a productive alliance, important, because by 2040 the number America and the Caribbean. Section such as rural alliance, commercial alliance, productive partnership, of people living in rural areas is not 4 details the variation in PA project productive subprojects and producer expected to decline from today’s rural implementation based on geography, organization productive subprojects. For consistency, this report uses only the population and a high proportion of that value chains and beneficiary targeting. term productive alliances. population will continue to be engaged Section 5 describes how PA subprojects in agriculture (World Bank, 2015). Hence, are selected, and Section 6 summarizes both the World Bank and its partners have the financial support typically provided to an interest in identifying effective and the alliance partners. Section 7 presents efficient instruments to link smallholder the monitoring activities of PA projects, farmers to markets, which can be adapted while Section 8 discusses key outcomes to idiosyncratic production, markets and and impacts of these operations. Section sector priorities. This report claims that 9 provides suggestions for developing the PA approach is a promising candidate. evaluation strategies for PA projects, and Section 10 summarizes the main 7.  The structure of this report is lessons learned. Section 11 provides as follows: Section 2 describes the conclusions and recommendations. concept of the PA approach and its 8 2 < 1. Introduction 2. Productive Alliances in Latin America and the Caribbean > 3. Targeting Strategies: Geographic, Value Chains, Beneficiaries 9 Linking Farmers to Markets through Productive Alliances THE PRODUCTIVE ALLIANCE APPROACH 8.  The Productive Alliance approach by the Ministry of Agriculture. These 10.  The design of the Productive provides integrated solutions for three agents are connected through Alliance approach encourages the addressing market imperfections a business proposition, or “business development of a horizontal alliance that constrain smallholder producers’ plan”, which describes the capital and among the producers and a vertical socio-economic progress, including: services needed by the producers and alliance between the producers and (i) limited scale of production and low proposes improvements that would the buyer(s). The difference between the productivity; (ii) inferior standing in allow them to upgrade their production two is important due to the respective market negotiations with buyers and capacities and skills to strengthen their nature of the relationship in each input providers to obtain better prices linkage with the markets, i.e. the buyer(s). case. The logic behind the horizontal and more stable market relationships; (iii) The implementation of this business alliance is quite straightforward, as poor knowledge of modern production plan within a PA project is typically the producers supported through a practices, technologies and market supported through three core inputs and/ PA project are typically grouped in requirements, and of the entrepreneurial or activities associated with producers’ a producer organization and jointly and management skills required to needs: (i) productive investments; (ii) develop and implement the business become more competitive and resilient technical assistance; and, (iii) business plan. With the support of a PA project, to economic and climatic shocks; development. Productive investments individual smallholders are encouraged (iv) inadequate (access to) financial typically include the provision of to engage in collective action to generate resources for productive investments machinery and equipment, infrastructure economies of scale (e.g. collective sales to increase efficiency and comply with (on-farm or off-farm), and production or setting-up a financial accounting market requirements; and, (v) lack of inputs (e.g. seeds, fertilizer, veterinary system that is subject to common review direct access to buyers and markets supplies) to producers. Similarly, by producer organization members) to successfully integrate into local, technical assistance entails the delivery and to invest in and share common national and/or international value of extension services technology and goods (e.g. warehouses or processing chains. The PA approach has been specialized assistance on technical equipment). Hence, their relationship implemented in diverse ways, adapting matters related to production, processing, is of a socio-economic nature, focused to the circumstances and needs of health/sanitation and environmental on identifying market opportunities individual countries, target beneficiaries, aspects, as well as market studies. within a concrete business plan while and market realities. Despite these Finally, business development focuses strengthening social cohesion, both of differences in implementation modalities on strengthening producers’ capacities which are further reinforced through a and institutional arrangements, all in management, accounting, business PA project. In contrast, the establishment PA projects share a common set administration and marketing. PA of an alliance between the producers of core elements (Figure 2.1). projects vary in their emphasis on each and the buyer(s) is a more complex of these core inputs, but all use support process due to the market imperfections 9.  The Productive Alliance approach goods and services to promote the described above. The nature and level involves three core agents: (i) a group fulfilment of a business plan. One key of the producer-buyer vertical alliance is of smallholder producers; (ii) one characteristic of a PA project is that the primarily a business relationship, based or more buyers; and, (iii) the public core inputs to support the business plan on a rational economic arrangement sector. The producers are typically united are financed through grants provided for the supply and purchase of goods in a producer organization (PO), the by the public sector, which are matched and/or services. Experience shows that buyer(s) can be active at different levels by contributions from the producers and establishing horizontal alliances is often of a value chain in either commercial in some cases also from the buyer(s). a prerequisite for successfully integrating or institutional markets, and the public smallholder producers into markets sector agent is commonly represented and vertical alliances with buyers. 10 Figure 2.1 Core Elements of a Productive Alliance Public Sector Business Plan Productive Technical Business Investments Assistance Development Production inputs Extension services Management Small infrastructure Technology transfer Accounting Producers Commercial agreement (Vertical Alliance) Buyer(s) Collective Action Product specifications (Horizontal Alliance) Quantity of product Common goods Quality requirements Shared capacity-building Price arrangements Financial accountability Delivery specifications Duration of alliance Support goods and services 11.  Stable prices and sales in addition are assured sales for producers, assured 12.  Strict implementation of the to improved product quality are product delivery for buyers and stable Productive Alliance approach requires primary advantages of the Productive prices. Both groups mentioned the that the business plan identify a Alliance approach for both producers improved product quality resulting from specific buyer upfront as a condition and buyers. The motivations driving project intervention leading to better for obtaining financing to overcome producers and buyers to enter a vertical, prices and hence revenues. In addition, market imperfections and to establish a long-term commercial relationship producers also valued the opportunity to direct, formal producer-buyer relationship. are multiple but generally relate to its obtain technical assistance, to improve This specification is typically based advantages. Based on survey evidence their negotiating power, and to receive on a commercial agreement between from PA projects, the crucial benefits prompt payment from the buyers. the group of producers and the buyer 11 Linking Farmers to Markets through Productive Alliances regarding the provision of a good in a “unconditional” PA projects in the Latin with fundamental PA project objectives specific value chain (Collion, 2012), in America and Caribbean portfolio. of improving smallholder production line with the objectives and activities and market integration. Similarly, the 14.  The basic concept of the of the business plan. The nature of role of the public sector in this enabling Productive Alliance approach is such commercial agreement can have environment varies across projects, straightforward and flexible permitting varying degrees of formality ranging with some governments going beyond adjustment to a wide range of market from a verbal commitment to a written their convening role and being actively realities and policy objectives. As contract detailing product specifications, involved in providing complementary seen in Figure 2.1, the PA approach is quantity required, product quality, and services (e.g. establishing market designed to resolve a series of constraints the respective unit price to be paid, information systems, promoting risk by providing integrated solutions that are delivery specifications and payment management tools). Section 3 details adapted to local conditions. There is also modalities. Further, the involvement the different forms of the “core model” a set of criteria that have been applied of the buyer in the formulation and displayed in Figure 2.1, implemented flexibly/differently across PA.projects implementation of the business plan across Latin America and the Caribbean. beyond the core elements displayed in differs across projects, ranging from Figure 2.1, reflecting the ability of the a rather limited engagement focused PA approach to adapt to differences in on product and delivery specifications policy priorities, market opportunities to stronger commitment where, for and countries’ economic conditions. example, the commercial agreement Each PA project determines through its might include direct buyer support to design and execution how these criteria the producers in the form of goods (e.g. are applied: for example, whether it is packaging material, production inputs) implemented at the national, state or or services (e.g. credit) (Collion, 2012). municipal level; whether business plan 13.  The up-front identification proposals by the alliance partners are of a buyer has not always been selected competitively or in a targeted implemented. A few PA projects share fashion; whether all products and/ the core elements of the PA approach or value chains are eligible or only a and foster the horizontal alliance among selected few; which type of producers producers. However, their design is not are eligible; and/or the level of their conditioned on the identification of a required co-financing contributions. specific buyer and an explicit, upfront 15.  In addition to differences in design vertical alliance with producers. This criteria, each project is subject to the has occurred in market environments enabling environment within which it where finding a specific buyer is not operates, and needs to adjust to and considered to be the major constraint utilize this environment to maximize to smallholder producers, but rather results. Where favorable conditions where identifying and accessing market exist, PA projects can benefit from opportunities more generally is the main complementary support systems. For concern. These types of PA projects example, in countries where the capacity go beyond a basic producer support and support from sector actors is strong, project by providing the three core inputs another form of alliance with the broader displayed in Figure 2.1., and requiring public and private sector agents can arise smallholder producers to adopt a market- (e.g. input providers, public/private/NGO oriented mindset and to develop a entrepreneurial service providers, local business plan which includes upfront the secretariats of agriculture, or financial targeted market, but not a specific buyer. institutions), encouraging a harmonization Section 3l distinguishes between these of public and private services in line types of “conditional” and the broader 12 3 < 2. The Productive Alliance Approach 3. Productive Alliances in Latin America and the Caribbean > 4. Targeting Strategies: Geographic, Value Chains, Beneficiaries 13 Linking Farmers to Markets through Productive Alliances PRODUCTIVE ALLIANCES IN LATIN AMERICA AND THE CARIBBEAN 16.  In Latin America and the account the broader portfolio of PA, 21 as two separate projects for this report. Caribbean agriculture remains an such projects region-wide have supported PA projects are typically designed and important sector for the rural poor, over 3,500 business plans and their negotiated at the national level, while with its contribution to GDP ranging from respective subprojects. The Bank now implementation takes place either a high of 22% (Nicaragua) to a low of has 15 years of cumulative experience nationwide or in specific priority regions 3.5% (Mexico) (WDI, 2012). Smallholder with diverse applications of the PA of the country. A distinctive feature of agriculture predominates, with about 72% approach following initiation of the first the Brazil PA projects is that they have of the regional rural population involved PA project in Colombia in 2002. With total all been designed, negotiated and in small-scale farming, compared to just funding exceeding US$1 billion, the PA implemented at the state level due to 3% engaged in large-scale commercial approach has proved both flexible and the ability of Brazil’s State Governments production.2 The remaining 25% adaptable to the individual circumstances to obtain loans from the World Bank consists of landless laborers employed and priorities of the 10 countries where (which other countries in the region primarily in agriculture (Collion, 2012). it has been implemented. Figure 3.1 cannot). Also, the Brazil PA projects shows the geographic coverage and are commonly multi-sector operations, 17.  Since the early 2000s, the chronology of PA projects in the region. where the establishment of PA World Bank has been implementing constitutes one of several components.3 the Productive Alliance approach 18.  The Productive Alliance approach Annex 1 presents a brief description throughout the Latin America and started out in the Andean countries of the 21 PA projects implemented in Caribbean Region to strengthen before being introduced in Central Latin America and the Caribbean. smallholder producers’ integration in America, the Caribbean and Brazil. local, national and international value Bolivia and Colombia have by now chains and thereby their access to better implemented two phases of their income opportunities. When taking into respective PA projects; these are assessed Figure 3.1 Geography of Productive Alliance Projects in Latin America Mexico (2012-17) Haiti (2012-16) Guatemala (2007-14) Jamaica (2010-17) Honduras (2008-16) Brazil: Para (2006-14) Panama (2007-15) Alto Solimoes (08-14) Colombia (2002-15) Sergipe (2008-12) Sta Catarina (2010-17) Sao Paolo (2009-17) Peru (2008-17) Rio (2010-18) Ceara (2012-18) Bolivia (2006-17) Pernambuco (2012-19) Parana (2012-17) Bahia (2014-21) 14 Table 3.1 Conditional versus unconditional Productive Alliance projects Conditional on identified buyer Unconditional on identified buyer Bolivia: Rural Alliances I & II (PAR I & I) Brazil: Alto Solimôes Basic Services and Sustainable Development Colombia: Rural Productive Partnerships I & II (PAAP I & II) Brazil: Bahia Sustainable Rural Development Guatemala: Support Rural Economic Development (PDER) Brazil: Para Integrated Rural Development Haiti: Strengthening Agriculture Public Services II (RESEPAG II) Brazil: Parana Multi Sector Development Honduras: Rural Competitiveness (COMRURAL) Brazil: Pernambuco Rural Economic Inclusion Jamaica: Rural Economic Development Initiative (REDI) Brazil: Rio de Janeiro Sustainable Rural Development Panama: Rural Productivity (PRORURAL) Brazil: Santa Catarina Rural Competitiveness Peru: Sierra Rural Development (ALIADOS) Brazil: Sao Paulo Sustainable Rural Development and Access to Markets Brazil Sergipe Integrated Project: Rural Poverty Mexico: Sustainable Production and Biodiversity Conservation (SPSB) 19.  It is important to distinguish to be effective in addressing key AREAS OF EMPHASIS between conditional and unconditional constraints. In Mexico, the unconditional 21.  Productive Alliance projects Productive Alliance projects among PA approach was chosen because the diverge in the degree of emphasis on the 21 operations in Latin America. SPSB project in its initial phase has been certain aspects of implementation. To reiterate: (i) those providing support testing innovative production methods Based on interviews with PA project conditional on the upfront identification (i.e. biodiversity-friendly practices) for leaders, De Salvo (2014) identified of a specified buyer willing to enter which no well-defined market existed at several areas of emphasis and divided into an alliance with the producers; the project design stage. In its planned them into the following categories: (i) and, (ii) those not strictly requiring this second phase, the project is considering Market integration: increasing smallholder condition. Table 3.1 divides the projects making the upfront identification of producers’ linkages to markets; (ii) of the region into these two categories. a specific buyer a requirement in the Competitiveness: improving production, Considering the overall portfolio, business plan to foster a strong vertical productivity and/or sales of smallholder the projects in Brazil and the Mexico alliance between producers and buyers. producers; (iii) Social Inclusion: facilitating Sustainable Production Systems and 20.  Diverse modalities of the access of disadvantaged groups to Biodiversity Project (SPSB) are those Productive Alliance approach have markets, productive infrastructure with the least stress on the vertical developed in areas of emphasis, and services; and; (iv) Environmental alliance between producers and a buyer types of beneficiary producers, end Sustainability: promoting environmentally- given their respective market structure, markets and contractual formality. friendly production practices among but otherwise share the core elements Given the wide range of projects region- smallholders. De Salvo also categorized and implementation approach of a PA wide with each adapting to the needs PA projects according to their emphasis project. In Brazil, the unconditional PA of the respective country and markets, on the three core inputs present in approach has been favored because the some PA projects place more emphasis all PA projects but implemented to a need to identify a specific buyer upfront on certain core elements than others different degree depending on project has not proved essential due to the in their implementation. As a result, objectives and sector conditions: (v) size of the local market and a generally they can be differentiated based on Access to Productive Goods: supporting strong demand from institutional buyers areas of emphasis in implementation, investment in (common) productive (e.g. school feeding programs, hospital beneficiaries, end markets and the goods, such processing equipment and/ food supply). Rather, improving market formality of commercial agreements or facilities to be shared by members of a access and integrating rural smallholder between producers and buyers. producer organization (e.g. warehouses, producers in general through the core processing machinery); (vi) Access elements of the PA approach proved to Technical Assistance: providing 15 Linking Farmers to Markets through Productive Alliances Table 3.2 Areas of emphasis Development Competitive- Environmen- Integration tal Sustain- Productive Assistance Access to Access to Access to Technical Inclusion Business Training Market Goods ability Social ness Country Project Name Rural Alliances I (PAR I) ** * *** * * ** *** Bolivia Rural Alliances II (PAR II) ** * *** ** ** ** *** Alto Solimôes Basic Services and ** ** *** *** *** * ** Sustainable Development Bahia *** *** ** * * ** *** Para Integrated Rural Development ** ** *** ** *** ** ** Parana Multi Sector Development * * ** ** *** ** ** Pernambuco Rural Economic Inclusion ** *** ** ** *** * * Brazil Rio de Janeiro Sustainable Rural ** ** ** *** ** ** ** Development Santa Catarina Rural Competitiveness * *** * ** *** ** * Sao Paulo Sustainable Rural *** ** * ** ** ** ** Development and Access to Markets Sergipe Integrated Project: Rural ** ** *** * *** * * Poverty Rural Productive Partnerships I (PAAP I) ** *** ** * * ** *** Colombia Rural Productive Partnerships II *** *** ** * * ** ** (PAAP II) Support Rural Economic Development Guatemala *** ** *** * ** ** * (PDER) Strengthening Agriculture Public Haiti *** * ** * ** *** *** Services (RESEPAG II) Honduras Rural Competitiveness (COMRURAL) *** *** ** * * ** *** Rural Economic Development Initiative Jamaica *** ** ** * ** * * (REDI) Sustainable Production and Mexico ** * * *** *** ** * Biodiversity Conservation (SPSB) Panama Panama Rural Productivity (PRORURAL) *** ** ** *** * ** * Peru Sierra Rural Development (ALIADOS) ** *** ** * * *   Note: A stronger emphasis by the larger number in stars/darker color shading, ranging from one to three stars/shadings. Source: Modified from De Salvo (2014) smallholder producers with access to differences in these areas of emphasis their production, productivity and sales improved extension services and other across PA projects are shown in Table 3.2. and increasing their market linkages. technical assistance; and, (vii) Access In addition, some projects address 22.  Generally, improving market to Business Development Training: transversal issues such as social inclusion integration and competitiveness are supporting producer groups in organizing and environmental sustainability. Social commonly the highest-rated emphases themselves to improve their position in inclusion is the second most highly rated of Productive Alliance projects. This market negotiations, reaching economies emphasis and of particular relevance aligns with the model’s overall objectives of scale in production, and enhancing in Bolivia, Guatemala and Northern of supporting smallholders’ upgrading of their administrative management. The Brazil, where indigenous peoples and 16 other disadvantaged groups represent towards Brazil where PA projects have BENEFICIARY TYPES a large share of the population and are emerged from former Community- 23.  Productive Alliance projects in generally among the poorest. Promoting Driven Development (CDD) interventions the Latin America region can also environmentally sustainable farming that supported the provision of small- be classified according to the stage practices has increasingly become the scale common goods requiring local of organization of their beneficiary focus of several Productive Alliance coordination. Similarly, Access to Business producers and corresponding needs projects in the Latin America region, such Development Training to strengthen for improvements and services. The as in Mexico and Panama. The three producer organizations is mentioned as degree of organizational structure remaining areas of emphasis on core a key emphasis in Bolivia, Brazil (State of varies greatly among beneficiary inputs can be considered as means to Bahia), Colombia and Honduras, where producers and also between business achieve the above-mentioned goals and agricultural markets are generally quite plan subprojects within the same a prerequisite for achieving overall PA competitive. Finally, Access to Technical PA project. Figure 3.2 distinguishes objectives. However, projects differ in Assistance is not rated as a high-level between three organizational stages: the importance they assign to each core focus in any of the PA projects, but the unorganized, organized, and organized input. For example, Access to Productive provision of improved technical assistance with well-established services. Goods is considered of high importance services is an important means of in seven projects, yet with a strong bias achieving the objectives of these projects. Figure 3.2 Stages of beneficiary producer organizations Need for im- Unorganized producers / Organized producer groups / provements and unregistered producer groups follow-on phase beneficiaries Organized producer groups with well-es- services in: (“Stage 1”) (“Stage 2”) tablished services (“Stages 3”) Production scale, productivity, Same as Stage 2 plus value addition, Needs form Organizational capacity, higher quality standards, product upgrading and differentiation, improvements production, minimum quality certification, competitiveness, sanitary and phytosanitary standards, and services standards, market access financial services, market share, sophisticated marketing and (export) market diversification logistics Haiti RESEPAG II Brazil Bahia Honduras COMRURAL Jamaica REDI Brazil Sao Paolo Brazil Alto Solimoes Brazil Sergipe Colombia PAAP I and II Brazil Rio Brazil Santa Catarina Bolivia PAR I and II PA projects Peru ALIADOS Mexico SPSB Guatemala PDER Brazil Pemambuco Brazil Ceará Brazil Pará Brazil Paraná 17 Linking Farmers to Markets through Productive Alliances Figure 3.3 End Market Levels + specifications for more + specifications for advanced/quite specific more advanced and process standards, yet + specifications Upgrading often quite specific implemented in the context + internal quality for selected basic process standards of highly integrated supply characteristics of standards, basic GAP, steps + quality grades and chains and where the suppli- good hygiene, and with associated varietals preferences products greater detail in er has a relatively sophisti- approaches to safe + visual character- record-keeping cated management structure + consistent quality + basic requirements pesticide use/storage istics for quality control and risk and quantities on pesticide use and record-keeping + 2nd and 3rd party management systems + visual inspecta- conformity assess- + visual inspectations 1st and 2nd party tions ment + 2nd and 3rd party confor- inspections / testing + 2nd and 3rd party mity assessment conformity assessment Level 1 Level 2 Level 3 Level 4 Level 5 Level 6 Stringency of official and buyer’s requirements Level of sophistication of the conformity assessment systems Developing Countries Industrialized Countries Traditional retail Small local super- High-end supermar- markets & small markets kets Retail markets stores High-end supermar- Local stores kets Discount supermarkets Source: Jaffee et al. (2011) Table 3.3 Formality of Commercial Agreements Written contract (with specifics)Written contract (no specifics) No contract per se / agreement is part of business plan Bolivia PAR I* & II Guatemala PDER Colombia PAAP I* & II* Haiti RESEPAG II Honduras COMRURAL* Jamaica REDI Panama PRORURAL Peru ALIADOS* *The requirement was effectively executed during project implementation 18 24.  Many Productive Alliance include buyers like wholesalers, food commercial agreement. As previously projects support subprojects that processing companies, supermarkets and mentioned, when strictly implemented work with producers at different specialized distributers. In Brazil however, the PA approach requires an established stages of organization. As a result, producer organizations rely heavily on alliance between producers and at least several projects work across the three institutional markets focused on local one buyer for eligibility to have a business categories of Figure 3.2. For example, (public) procurement of goods, where plan financed. These include most the Guatemala PDER by design adopted buyers are typically linked to school countries in the PA project portfolio in a mixed strategy to support both feeding, hospital food supply or similar Latin America. Among these conditional “Stage 1” and “Stage 2”-type producer programs that promote the procurement PA projects, the required formality of organizations through different activities. of products from smallholder producers. the commercial agreement between Moreover, while in Colombia most of the two parties varies from direct 26.  Productive alliances also differ in the beneficiary producer organizations written contracts to indirect agreements regard to product quality, depending can be considered as “Stage 2”, the embedded in the business plans (Table on the degree of formality of their project introduced a “street light” 3.3). Only the Colombia PAAP I & II end markets and respective product classification which demonstrated required a written agreement between requirements and standards. Jaffee differences in organizational structure the producer organization(s) and the et al. (2011) define six levels of market and capacity among its beneficiary buyer(s) with key product specifications requirements, which are typically producer organizations by dividing them (e.g., product volume, quality and prices) incremental on an ascending scale. Figure into a green, yellow, and red category. prior to obtaining subproject financing. 3.3 describes these levels of market In Brazil, projects in some states (e.g. To a lesser extent, the Guatemala PDER, requirements and the respective typical Amazonas, Sergipe) have worked with Honduras COMRURAL and Panama end markets, starting from traditional producers through community-level PRORURAL require a buyer’s written (informal) local markets and moving projects instead of formalized producer commitment to purchase products from up the scale of “increasingly higher- organizations, as a result of their transition the partnering producer organization, value domestic and international value from CDD-based interventions to more but without further specifications. chains. This change is accompanied by market-oriented activities. Similarly, Some conditional PA projects do not progressively stricter food safety, quality, the Peru ALIADOS project made a necessarily require a written contract and other requirements” (Jaffee et al., clear distinction between subprojects between producers and buyers. The 2011). Most productive alliance projects supporting productive alliances commercial relationship between the two serve markets between Level 1 (informal involving producers and subprojects for parties can be documented in a letter domestic markets) and Level 3 (high-end food security at the community level. of intent to purchase or in the business domestic markets). Mid-level markets Given this diversity of implementation plan to be supported by the PA project. are considered the most appropriate for methodology, the project examples listed In Bolivia, for example, while a letter smallholder producers in developing in Figure 3.2 are based on the most of intent to purchase was a minimum countries, as the costs of achieving common type of producer organizations requirement, medium-term offtake and maintaining the required scale and encountered under these projects. contracts, improvised arrangements standards are more bearable (Jaffee et or other types of agreements between al., 2011). However, several high-value END MARKETS exceptions such as specialty coffee from parties were also considered acceptable. 25.  A further distinctive feature of Honduras and Colombia have proven 28.  The requirement that business Productive Alliance projects is the their ability to compete in export markets. plan financing be contingent on variation between end markets as well a written contract has not always as their respective product quality FORMALITY OF been implemented. For example, in requirements. End markets are first, Guatemala and Panama, few contracts COMMERCIAL AGREEMENT distinguished by whether they are private between producers and their alliance 27.  Productive Alliance projects sector commercial markets or public partner buyers were formally signed. conditional on a specific buyer entering sector institutional markets. In most Latin Similarly, the commercial agreements an alliance with producers differ American PA projects links are established in Bolivia PAR I and II were/are not in the degree of formality of their to commercial markets, which typically formalized by a written contract between 19 Linking Farmers to Markets through Productive Alliances the beneficiary POs and buyers – which is 2. Based on Collion (2012), smallholder not in any case, a requirement - but for all agriculture includes households involved subprojects the terms of the agreements in livestock-raising, artisanal fishing and forestry. were specified in the business plans 3. The other project components in Brazil supporting the subproject application. are typically related to sectors other than agriculture, such as the provision of water and sanitation or small-scale road 29.  The success of a vertical alliance infrastructure to beneficiaries. Projects in is not necessarily guaranteed by a pro other countries (i.e. Guatemala, Jamaica, Peru and Mexico) also implement forma commercial agreement between components complementary to the producers and buyers but rather by the establishment of productive alliances, but usually with proportionally lower commitment of both alliance partners funding compared to the projects to implement the business plan. While in Brazil. formulating the specifics of a vertical alliance between producers and buyers potential misunderstandings and conflicts should be avoided and it is highly important that all parties comply with the agreed (written or unwritten) terms of the business plan. Thus, PA projects conditioned on the identification of a specific buyer are more straightforward as they require the involvement of this buyer in the implementation of the business plan from the beginning. In other words, while a formal upfront agreement can foster and clarify the specifications of a vertical alliance between producers and buyers, its success depends on the commitment of the alliance partners to act on this agreement during project implementation. 20 4 < 3. Productive Alliances in Latin America and the Caribbean 4. Targeting Strategies: Geographic, Value Chains, Beneficiaries > 5. Subproject Preparation and Selection 21 Linking Farmers to Markets through Productive Alliances TARGETING STRATEGIES: GEOGRAPHIC, VALUE CHAINS, BENEFICIARIES 30.  The specific objectives and result, specific eligibility criteria for of the rural population; and (iii) poverty expected outcomes of a Productive subproject financing have evolved, density. Following this first selection, Alliance project determine its targeting as described and compared below. other municipalities within a broader strategies. Given the flexibility of the PA geographic area were also included in approach, governments can use it as an GEOGRAPHIC TARGETING the project coverage to avoid spatial instrument to achieve diverse outcomes gaps and ensure territorial cohesion. 32.  Most Productive Alliance projects related to the overall objectives of As a result, implementation of Bolivia’s focus on specific priority areas in a supporting smallholders in production PAR project covers 120 municipalities in given country based on poverty and upgrades and market integration. Such five Departments nationwide. Similarly, market criteria. Despite some differences outcomes might include: (i) improved Peru’s ALIADOS project targets six of in geographic targeting and scope, all commercialization of smallholder the country’s 25 Regions characterized PA projects state explicitly that they agriculture in specific priority regions, or by higher levels of (extreme) poverty intend to reach poor rural producers in another case, nationwide; (ii) targeting and poor areas that experienced high with little access to markets and weak support to selected value chains, while in levels of violence during the civil conflicts linkages to agro-processors and/or other cases considering all value chains as of the 1980s and 1990s. In Panama’s established buyers in areas with market eligible; or, (iii) supporting certain groups PRORURAL project, Provinces with a potential. The geographic dispersion within the rural population, or targeting mean monthly income below the national of the PA projects in Latin America the rural poor in general. Hence, once average were selected first, followed is described briefly in Table 4.1. the objectives and expected outcomes by specific Districts with higher levels of a PA project are clear, appropriate of extreme poverty (i.e., double the 33.  Productive Alliance projects targeting strategies can be developed. national average). A different, innovative generally define specific socio-economic strategy for geographic targeting was and demographic criteria to ensure 31.  Productive Alliance projects applied by the Mexico SPSB project, effective geographic targeting. For across Latin America have adopted which used a three-step prioritization example, Bolivia’s PAR project selected various targeting strategies: methodology. Given its objective of municipalities on the basis of inclusion geographic, agricultural value biodiversity conservation, a municipal- criteria including: (i) a higher than chains and beneficiary-based. As a level map with priority sites for terrestrial average population growth; (ii) share Table 4.1 Geographical Coverage of Productive Alliance Projects Project National coverage Sub-national coverage Bolivia PAR I & II 5 out of 9 Departments Brazil all PA projects State-level projects (priority areas within States) Colombia PAAP I & II 31 of 32 Departments Guatemala PDER 20 of 22 Departments Haiti RESEPAG II 5 of 10 Departments Honduras COMRURAL 7 out of 24 Departments Jamaica REDI 12 of 14 Parishes Mexico SPSB 6 out of 31 States Panama PRORURAL 3 out of 10 Provinces Peru ALIADOS 6 out of 25 Regions 22 biodiversity conservation was established population, taking into account a geographic and socio-economic first, then overlaid by a map of priority project’s objectives and countries’ specific eligibility criteria, while others focus municipalities based on indices of poverty poverty and market conditions. Based on selected value chains, with project and marginalization. Areas of overlap – or on the overall goals of poverty reduction support targeting only productive “hot spots” – were identified as having and market integration, the geographic alliances from those chains declared significant potential for mainstreaming selectivity of PA projects based on socio- eligible. For example, in Honduras’ biodiversity while improving economic, vulnerability and environmental COMRURAL the eligible products wellbeing, including improving/ criteria makes sense. However, the level are horticulture, specialty coffee, increasing producers’ organizational of effectiveness might be even higher if beekeeping, aquaculture, fruit-culture, level, production area and value of more emphasis were placed during the rural tourism, dairy production, cattle production. This process resulted in the design phase on assessing the availability production, swine production, poultry/ selection of six priority target states. of existing infrastructure (e.g. warehouses, hen farming and traditional art crafts. roads, ports) and accessibility to Mexico’s SPSB has established eligibility 34.  The geographic focus of a national and international markets. criteria supporting subprojects in just Productive Alliance project can seven products: cocoa, coffee, honey, change over time due to a country’s shifting needs or, interventions which VALUE CHAIN TARGETING silvo-pastoral, forestry (timber and non-timber), wildlife, and ecotourism. scale-up the operation. Guatemala’s 36.  Productive Alliance projects PDER project was designed initially to in Latin America have financed 38.  The targeting strategy for value focus on eight Departments, based on subprojects with activities in both chains and/or agricultural products can indicators of indigenous population, level agricultural production and rural change during project implementation, of poverty, agricultural potential and non-farm economic activities. The when market circumstances or policy productive infrastructure. During project wide range of agricultural products priorities shift. Jamaica’s REDI project implementation, external factors (a severe and services supported by PA projects for example, was initially open to storm) called for the project’s expansion shows an emphasis on high-value crops all agricultural products, but due to to 20 Departments and re-focusing some including: (i) specialty coffee, (ii) cocoa, changes in sector policies the third call activities on infrastructure investments. In (iii) livestock (beef and dairy), (iv) rubber, for business plan proposals was more the case of Colombia, PAAP I was initially (v) fruits (short- and long-cycled like specific in focusing on investments focused on 27 out of the countries’ blackberry, mango and peaches), (vi) in non-traditional, high-value crops 32 Departments before expanding to quinoa, (vii) potato seeds, (vii) sesame, (e.g., pepper, ginger, strawberries). national coverage during the second (viii) native bee honey, (ix) Brazil nuts, (x) 39.  The experience of the Productive project phase due to the positive maize, (xi) beets, (xii) aquaculture, (xiii) Alliance approach has demonstrated outcomes of first-stage interventions agroforestry, as well as art crafts and the importance of identifying early and an increased national budget. eco-tourism services. The majority of on – preferably before any subproject investments has focused on agricultural 35.  In many cases, the geographic investments have been made – crops and/or livestock production. targeting of Productive Alliance those value chains and/or products projects is a well-justified, as it ensures 37.  Some Productive Alliance projects with the highest long-term market the effective provision of goods support productive alliances in any potential. Although it is difficult to and services to a specified target value chain within the established estimate the future developments of Table 4.2 Targeting of Value Chains All producers / value chains Selected products / value chains Brazil Alto Solimôes Brazil Bahía Colombia PAAP I & II Honduras COMRURAL Bolivia PAR II Jamaica REDI Mexico SPSB 23 Linking Farmers to Markets through Productive Alliances certain value chains, it is crucial that objectives, but could suffer from criticism 3% of the rural population), smallholder expected projections and the role of for favoring certain groups in the sector. producers (about 72%), and the rural certain products for the project target landless (about 25%). As with the Latin 41.  The selection of value chains to population are taken into account at America region, smallholder producers be supported through a Productive the design stage. While PA projects represent the majority of agricultural Alliance project should be based have generally led to adequate, producers in most developing countries. on technical criteria, well-defined average financial and economic Internal Even so, in any given country, there is objectives and market potential to Rates of Return (IRR), these rates vary great heterogeneity among smallholder ensure effectiveness and sustainability. greatly among different products, as producers in terms of access to land, Technical criteria should relate to described in more detail in Section 8. labor, physical and human capital social, economic and environmental and natural resources. Collion (2012) 40.  Open and selective targeting relevance. Given the flexibility of the defines three distinct categories of of value chains has advantages and PA approach, value chain targeting smallholder producers: (i) subsistence disadvantages, depending on the can be adjusted in a follow-up project producers, i.e. households engaged market structure. Whether open or or during project implementation if in agricultural production mostly selective targeting of value chains and/ several rounds of the business plan for own consumption with limited or agricultural products is a better subproject selection process take place. assets and market access due to low choice depends on the specific project productivity and poor infrastructure; objectives, the situation of the agricultural PRODUCER TARGETING (ii) small investor producers, i.e. small- sector in each country, and the project and medium-sized producers with 42.  Productive Alliance projects alignment with public sector priorities. favorable endowments of assets, skills generally target “transition smallholder While the open selection of subprojects and market access who commercialize producers” who have the potential and permits fair competition for support their products competitively; and (iii) willingness to increase their productive across all agricultural products, it might “transition smallholder producers”, i.e. and entrepreneurial capacities. Collion lead to investments that are not aligned producers in between subsistence and (2012) describes the distribution of with sector needs and where results are commercial farming who have some the rural population in Latin America, difficult to compare. On the other hand, assets and resource endowments but distinguishing between large-scale a selective approach is more focused lack well-established linkages to buyers commercial agricultural producers (about and more likely to be aligned with sector and markets. Productive Alliance projects Figure 4.1 Distribution of rural population and smallholder producers 3% Large-scale 9% Small investor producers 72% Smallholder 43% Transition smallholder producers producers 20% Subsistence smallholder producers 25% Rural landless Source: Modified from Collion (2012). 24 Table 4.3 Eligibility criteria for Beneficiary Producers Minimum Maximum Assets / Minimum Level of Experience in Formally Grouped with Project Income Age Education Activity Legal Recognition Bolivia PAR I & II * * * * Colombia PAAP I & II * * * * * Guatemala PDER * Honduras COMRURAL * * * Jamaica REDI * * * Mexico SPSB * * Panama PRORURAL * * * Peru ALIADOS * * focus on strengthening backward exceed the equivalent of 200 minimum requirements. Common requirements for linkages for these “transition smallholder wages and, that monthly income would PO include: (i) being legally-established; producers” who have the potential and not exceed four minimum wages.4 At (ii) demonstrating the administrative and willingness to increase their productive the end of the second project phase, an financial capacity to manage subprojects; and entrepreneurial capacities to engage independent impact evaluation found and, (iii) having well-established rules in modern agri-food markets. In other that some 90% of the projects’ direct and decision-making processes. Annex 3 words, the aim of the PA approach is beneficiaries were registered in the includes a table with specifics for some to attract producers with a market- public identification system for eligible PA projects with PO-level requirements. oriented mindset and likely to adopt and recipients of conditional cash transfer 45.  Some Productive Alliance projects prosper from the technical assistance programs. Similarly, in Guatemala’s require producer groups to be formally and investments offered by the projects PDER project the beneficiaries’ yearly registered at the time of application income could not exceed national, per 43.  Productive Alliance projects, to for project support, but generally do not capita GDP.5 Table 4.3 presents project ensure effective targeting, usually establish minimum requirements in terms examples with the most commonly used define eligibility criteria for beneficiary of length of existence or operational eligibility criteria. Annex 3 provides producers based on socio-economic performance prior to those applications. details on the varying definitions and and demographic factors. For In practice, many of Colombia’s PAAP specifications used across PA projects. example, in four countries (Colombia, producer groups had already been Guatemala, Jamaica and Panama), PA 44.  Membership of a legally formed before the project started projects explicitly established maximum registered producer organization implementation, which implied a strong amounts of household income, assets, is another frequent criterion for social cohesion amongst their members landholdings and education as eligibility Productive Alliance projects. This and a willingness to sell collectively. criteria for producers. Thus, while requirement for producers becomes This proved essential to build effective the PA approach is generally not an relevant at the stage of obtaining and sustainable alliances, as highlighted instrument for targeting the poorest subproject financing (once the business by better results compared to younger of the poor, these criteria ensure that plan has been selected), not necessarily producer organizations in the impact poorer segments of the rural population for being able to participate in the call evaluation of PAAP. In contrast, the REDI - not better-off investor producers - will for business plan proposals. In several project in Jamaica played a strong role in be direct beneficiaries. For example, PA projects, producer organizations the formation of producer groups. There, Colombia’s PAAP I and II projects (PO), not just the beneficiary producers social cohesion and cooperation among required that household assets not themselves, must meet certain eligibility producers have proved to be challenging 25 Linking Farmers to Markets through Productive Alliances during project implementation. Similarly, 47.  Another common requirement Brazil Sergipe, Brazil Rio). Other projects in Mexico the SPSB project provided established by Productive Alliance in Brazil however, support productive specific support enabling producer projects is a minimum number of subproject implementation through groups to be legally registered. member producers, which, in the producer organizations of family farmers design of the Latin American PA projects (e.g. Brazil Bahia, Brazil Ceará, Brazil 46.  Other Productive Alliance projects varied between 10 and 30 members. Pernambuco, Brazil Santa Catarina, help producer groups selected for In practice, none of these PA projects Brazil Sao Paolo), similar to the non- subproject financing to identify and imposed a maximum limit on members, Brazilian PA projects. The Peru ALIADOS adopt the most appropriate legal which permitted the formation of larger project also emerged from a history of structure by financing legal advice and producer groups able to benefit from CDD-projects, adopting two lines of the costs associated with formalization. project support and scale. As a result, subproject support: a productive line The types of legal associations for the average number of members per based on business plans aligned with producers and the respective pre- PO in the Bolivia PAR I was 50, in the PA approach and another line for requisites vary greatly between the Colombia PAAP I 58 and in the community-level territorial development. countries. As a result, because each Guatemala PDER, 96 producers. form of legal status involves the group 49.  Overall, a clear strategy for in complying with specific organizational 48.  Brazil is a special case for the targeting beneficiary producers is requirements, it is important that targeting of beneficiary producers crucial for defining the success of producers become familiar with these under Productive Alliance projects, a Productive Alliance project (and requirements before settling on a single since these have emerged from former of any development project). To option. For example, in Guatemala’s Community-Driven Development (CDD) ensure credibility and a fair selection PDER project, beneficiary producers projects (particularly in the Northeast process, beneficiary targeting should typically formalized as civil society region). Productive subprojects have be executed in a transparent manner organizations, which turned out to be been implemented with producer and based on technical criteria (i.e. the least expensive and most expedient associations at the community level social, economic, environmental form of organization for registration. instead of with formalized producer soundness) in alignment with sector organizations (e.g. Brazil Alto Solimôes, objectives and market opportunities. 26 BUYER TARGETING lead potentially to collusion and low beneficiary producers, buyers also prices for producers. Only in Bolivia, benefit through resulting improvements 50.  Targeting financially and Colombia, Guatemala and Panama in product quantity and quality. Several commercially strong buyers and their was the formal screening of buyers PA projects have reported that the commitment to an alliance with the required, with the specific requirements advantages for buyers include better producers, are critical, given that being described in In Practice Box 1.6 prices and sales from improved product establishing alliances between producer quality and more stable product delivery. organizations and buyer(s) is the key 52.  Requiring buyers to co-invest These and other advantages have objective of PA projects (particularly the in supported business plans has not been confirmed through larger-scale ones that make the vertical alliance a been very successful. The PA projects surveys and stakeholder workshops in conditional requirement for financing). in Colombia, Guatemala and Panama the Colombia PAAP-II project where Most PA projects in Latin America required buyers to co-invest in the more than 90% of buyers stated/agreed seek to link smallholder producers to alliance subproject in their designs. In that their motivation to enter a PA with buyers in commercial markets. Typically, practice, only in the case of Colombia smallholder producer organizations these buyers are private firms and/ was this requirement consistently applied was driven by business reasons. Such or individuals such as brokers, traders, during implementation, as private evidence should be disseminated more exporters, wholesalers, supermarkets, funds from buyers were part of the intensively and widely to promote the specialized distributors, processors and total partnership investment package. PA approach with potential buyers. restaurant chains. In contrast, the PA As a consequence, PA projects could projects in Brazil generally do not share more forcefully require criteria for active this orientation towards the private buyer involvement and monitor these 4. Colombia’s minimum wage was US$328 a month in December 2011. commercial market, although this result during implementation to ensure greater 5. Annual GDP per capita (2013 US$ PPP): is not due to their design. Despite the financial commitment from buyers. US$5,300 (2013 estimation), US$5,200 (2012 est.), US$5,200 (2011 estimation). option and even with encouragement However, enforcing such commitment 6. Throughout this report, “In to make the link to commercial markets, from early on, especially in a new PA Practice” boxes highlight practical many subprojects in Brazil (especially in project, might be challenging. As the recommendations, innovative activities or lessons learned based on examples the Northeast region) provide support to subproject grant financing primarily drawn from PA projects that the World beneficiary producers and communities goes to the producers, buyers might Bank has implemented in Latin America. enabling them to sell primarily to public not perceive the benefit of their sector institutional markets (e.g. school involvement in a productive alliance, feeding programs, hospital food supply). in addition to the general reluctance This focus on institutional buyers has, in to enter into a business relationship several cases, been encouraged by these with smallholder producers. public procurement programs paying 53.  While criteria for economic above-market prices and/or providing viability should be mandatory for tax benefits for smallholder producers. the selection of buyers in Productive 51.  Few Productive Alliance have Alliance projects, requiring their active established eligibility criteria for (financial) contributions might be more buyers. So far, PA projects have focused feasible during a second phase project on establishing criteria to ensure the or in a longer-established alliance. rigorous selection of viable producers. Alternatively, PA projects could put An equivalent effort for the selection much greater effort into explaining the of buyers has not yet occurred, which benefits to buyers of a vertical alliance increases the risk of: (i) electing a buyer with smallholder producers. Specifically, who is not sufficiently competitive to although the subproject productive perform sustainably in the market; or, investments, technical assistance and (ii) promoting elite capture in imperfect business development activities of PA markets with few buyers which can projects are directed primarily towards 27 Linking Farmers to Markets through Productive Alliances In Practice Box 1: Eligibility Criteria for Targeting Buyers Buyer targeting differs across PA projects, as illustrated by the list of eligibility criteria from four projects: Bolivia PAR I & II: (i) purchase commitment between buyer and producer organization; (ii) at least two years of experience in the national market; (iii) legally established and with proven commercial infrastructure; and, (iv) clear technical specifications (quantity, quality, terms, conditions of storage, packaging and other related to the specificity of each product marketed). Colombia PAAP I & II: One or more buyers who demonstrate: (i) competence; (ii) experience; (iii) financial strength; and, (iv) a commitment to buy all or at least 70% of the products from the alliance. Guatemala PDER: (i) a national or international company specialized in the production and sale of goods and services; (ii) prepared to co-invest with partners in Guatemala to establish a long term commercial relationship; (iii) has good and reliable distribution channels; and, (iv) economically, financially and legally solvent. Panama PRORURAL: Agro-processor or wholesale agent with: (i) experience in the specific products; (ii) willing and able to contribute to the alliance with financial resources or in kind (e.g., through technical assistance); (iii) willing to guarantee a minimum purchase volume and reference price through a purchase agreement/contract; and, (iv) having sufficient purchasing power, appropriate facilities, positive references and experience with small producers. 28 5 < 4. Targeting Strategies: Geographic, Value Chains, Beneficiaries 5. Subproject Preparation and Selection > 6. Subproject Financing 29 Linking Farmers to Markets through Productive Alliances SUBPROJECT PREPARATION AND SELECTION 54.  Productive Alliance projects 56.  The initial subproject proposals plan is between 12 and 18 months, but typically start with public information submitted are then screened and usually not exceeding 24 months. Once campaigns in the project target areas evaluated against the eligibility criteria finalized, the business plans are again (Step 1 of Figure 5.1). The objective of set by the implementing agency or submitted to the project (e.g. the PCU, these information campaigns is to raise project steering committee (Step 3). As a project steering committee, or an awareness among potential beneficiaries described in Section 4, these criteria independent evaluation agency) to be (i.e., producers groups and buyers, and typically include applying producers’ evaluated and ranked according to pre- on occasion specific disadvantaged maximum income, asset and/or land established selection criteria. In Practice groups) about the PA approach, the ownership, age, education, experience Box 3 briefly describes differences in the support offered by the project, the in agricultural production, the quality institutional arrangements of PA projects related eligibility and selection criteria, of the subproject description, as well for the subproject selection process. and the application procedure. as the producer organization’s legal 58.  Finally, subproject proposals which status.7 Most commonly, the Project 55.  Subsequently, a call for initial meet (a minimum of) the established Coordination Unit (PCU) selects viable subproject proposals is launched (Step criteria are approved for financing initial subproject proposals and invites 2). Interested producers are typically (Step 5).8 Typically, after this approval, the proponents to prepare a more invited to submit a proposal that briefly a memorandum of agreement between detailed subproject business plan. describes the key elements of their the implementing agency and the intended productive alliance and lays out 57.  Viable initial proposals are alliance partners is processed legally and the justification for project support. The developed into more detailed, administratively. Details on the support preparation of these initial subproject comprehensive subproject business financed/provided to productive alliance proposals is usually the responsibility plans (Step 4). The PA project plays an subprojects are presented in Section 6. of the producer organization, which in important role in supporting this step. 59.  Productive Alliance projects differ most cases uses a standardized proposal Specifically, the PA project often recruits in the number of calls for subproject format provided/facilitated by the project. and pays for private service providers proposals launched during project In some cases, the initial proposals (usually individual consultants and/or implementation. Some PA projects need to be presented jointly with the consultant firms pre-approved by the have established a regular schedule of potential buyer(s) (e.g. Bolivia PAR). In project and agreed with the POs) to calls, while others only call for proposals many PA projects, the calls for subproject prepare and finalize the details of the once during project implementation. proposals are open and competitive. business plans working with the alliance For example, the Bolivia PAR and However, in some cases they can also be partners. Business plans provide greater Colombia PAAP conducted multiple selective, allowing only certain regions or detail than the initial subproject proposal (virtually annual) calls for proposals in beneficiary groups to apply, or products and typically include the PA objectives, all target regions on specific dates, to qualify (as in the case of Brazil and implementation arrangements, needs while the Mexico PSPB executed just Mexico). Some of the Brazil PA did not for technical assistance, productive one call early in its implementation. The call for proposals, but the opportunity infrastructure and/or seed capital, budget Brazil Pernambuco project, in contrast, to develop subproject proposals was projections, financial feasibility including engages in continuous information disseminated by the Project Coordination cash flows, and results indicators. In dissemination and accepts proposals Unit (PCU) of the State Secretariat Practice Box 2 describes best practices on a rolling basis. Annex 4 provides of Agriculture or similar agency or, for the preparation and selection a more detailed description of the through producers’/communities’ criteria of subproject business plans. subproject preparation and selection connections with local government In general, the envisaged subproject process and specific project examples. or other entities (e.g. Brazil Pará). duration after approval of the business 30 Figure 5.1 Steps of the Subproject Selection Process Step 1 Information Campaigns & Call for Proposals Information Campaigns Step 5 Outreach to potential benefiaries (producers, commercial partners, disadvantaged groups) Selection of subproject Call for subproject profile proposals Step 2 business plans Preparation of subproject Subproject business plans are evaluated on technical, finacial, environmental, proposals social feasibility criteria Producer organizations prepare initial Approved business plans are selected subproject proposals for financing Step 4 Step 3 Preparation of subproject Evaluation of subproject business plans proposals Producer organizations prepare subproject Subproject proposals are screened and business plans with support from private evaluated on eligibility criteria service providers and commercial partners Eligible proposals are selected for business plan preparation stage 7. Some projects have supported the legal formalization of producer groups before subproject financing was provided (e.g. Guatemala PDER, Jamaica REDI, Mexico SPSB). 8. In some circumstances, not all subprojects that comply with the minimum criteria can be financed, so projects set a cap on the number of subprojects to be selected (e.g. the top 40% of subprojects, as in the case of the Colombia PAAP). 31 Linking Farmers to Markets through Productive Alliances In Practice Box 2: Preparation and Selection Criteria for Subproject Business Plans Ideally, business plans should be developed jointly between the alliance partners (producer organizations and buyers) and the service providers contracted as part of the subprojects. To minimize bias and rent-seeking, the service providers developing the business plans are generally different from the ones supporting the selected business plan subprojects during implementation. To ensure good quality, the selection criteria for subprojects across all PA projects are typically based on a mix of: i.  Technical feasibility (e.g. production capacity and quality); ii.  Financially viability (e.g. financial sustainability beyond the duration of project support); iii.  Market linkage strength (e.g. quality of market analysis, identification of technical assistance needs, participation in marketing events); iv.  Alliance partners’ capacity (e.g. quality of production planning, quality control, identification of production bottlenecks);, v.  Social aspects (e.g. potential for job creation, belonging to a disadvantaged group); and, vi.  Environmental sustainability (e.g. the subproject’s promotion of improved or climate-efficient production practices). In Practice Box 3: Institutional Arrangements of PA Subproject Selection Process The involvement of local actors in the evaluation and selection of subproject proposals is key to validate the information provided and to assess the soundness of the proposals against eligibility criteria. However, PA projects vary in the degree to which decentralized actors are involved in the selection process. As shown in the table, while several projects have regional PCUs supporting the screening and evaluation of proposals before transmitting their opinion to the centralized institutions (e.g. PCU, project steering committee, or independent evaluation agency) for final approval, only the Bolivia PAR, Colombia PAAP and Peru ALIADOS projects used regional multi-stakeholder committees to perform these tasks. Centralized Decentralized Guatemala PDER Bolivia PAR I and II Honduras COMRURAL Colombia PAAP I and II Jamaica REDI Peru ALIADOS Panama PRORURAL Mexico SPSB Brazil PA projects (State-level) 32 6 < 5. Subproject Preparation and Selection 6. Subproject Financing > 7. Results Monitoring 33 Linking Farmers to Markets through Productive Alliances SUBPROJECT FINANCING 60.  The financial support for municipalities, other development a Trust Company and were paid out into Productive Alliance projects is partners) to cover the remaining individual trust accounts controlled by provided through matching grants. costs not financed by the beneficiary each beneficiary producer organization. These grants co-finance the core producer contributions. In Practice Box The project’s operational manual inputs of Productive Investments (e.g., 4 describes the creation of a broader specified that the grant disbursements machinery and equipment, fertilizers, alliance, including agents of the enabling were conditioned on the funds being seeds), Technical Assistance (e.g., environment, around the PA project. used in accordance with the subproject consulting services, studies and surveys) business plan and under the supervision 62.  Beneficiary producer organizations and Business Development (e.g., of the Trust Company and the Ministry are responsible for implementing the training in organizational management, of Agriculture and Rural Development. approved subprojects and thus the business administration, accounting The use of one-time disbursements to key recipient of the grant funds. In the and marketing). The activities are trusts managed by the POs themselves majority of cases, the matching grants typically implemented by the producer helped to streamline project activities provided by a PA Project are disbursed organizations with technical support and eliminated restrictions to project into the subproject-specific beneficiary from private service providers, while the execution within the same fiscal year. producer (or community) organization public sector Project Coordination Unit Another example is the Guatemala bank account. Thus, they are accountable (PCU) focuses on facilitation, monitoring PDER project, which contracted and for the proper use and management of and supervision to ensure quality and directly paid Business Development grant resources in line with their approved adequate procurement. Details of Service providers who also managed subproject business plans. These co-financing arrangements for specific the implementation of other technical accounts and the use of the grant funds projects are described in Annex 5. assistance. The matching grants for managed and overseen by a subproject seed capital and infrastructure were 61.  The levels of grant support per committee to ensure compliance with paid directly in three tranches into subproject and the arrangements for procurement rules. The most common the POs’ bank accounts, once all co-financing vary considerably between procurement procedure used across requirements were met. In the Peru Productive Alliance projects. In most PA projects is “Community-based ALIADOS project the large number of cases, PA projects co-finance between Procurement”, which is explained to the subprojects - which were small in scope 70% and 90% of the total subproject POs during training and capacity-building and technically simple - led the project investment costs. Beneficiary producer activities organized by the projects. to authorize single lump-sum payments organizations (or other partners) are Typically, matching grants are disbursed to POs (instead of several tranches). typically required to contribute between in three tranches into the account of the 10% and 30%. A higher contribution PO during the life of the PA project. Prior 64.  There are established ceilings from the beneficiary producers leads to the transfer of resources to a PO bank for the grant amount supporting a to a greater sense of ownership account, in all PA projects the PCU has to subproject. These ceilings typically and commitment to the successful first verify and clear the PO’s requests for range from a minimum of US$1,800 per implementation of their PA subproject payment both technically and financially. beneficiary household in Bolivia’s PAR business plan. Colombia’s PAAP I and projects to a maximum of US$3,600 63.  Some alternatives to the II projects capped co-financing support in Honduras’ COMRURAL project. A direct disbursements of Productive from the PA project at 40% and 30% notable exception is Peru’s ALIADOS Alliance project grants to producer of subproject costs, respectively. In project which has a much lower grant organizations have developed. For addition, both phases successfully ceiling of US$600 per beneficiary example, in Colombia, financial resources leveraged additional funds from other household. Similarly, PA project designs of the PAAP project were administered by partners (e.g. local governments/ established maximum grant amounts 34 per producer organization/subproject. of access, and collateral for commercial and needed. Some experiences are In some projects the total grant ceiling finance. The provision of matching described in the next paragraph. amounts per PO were/are quite high, grants from the public sector to support 67.  In their design, almost all ranging from US$200,000 to US$300,000 smallholder agricultural production in Productive Alliance projects mention (i.e., Jamaica REDI: up to US$200,000; rural areas raises the question of whether the goal of enhancing producers’ access Panama PRORURAL: up to US$250,000; such financing could and should be to commercial financial services to Mexico SPSB: up to US$300,000). In better-provided by commercial financial complement project grant financing contrast, other PA projects allow for institutions. There are three main and beneficiary contributions, but lower ceilings of around US$30,000 to arguments for why such public-sector in practice few such linkages have US$50,000 per PO (i.e., Guatemala PDER: matching financing is warranted: (i) materialized. Activities to stimulate up to US$30,000; Bolivia PAR: up to positive externalities: the activities of a PA this access include (i) specific technical US$50,000; Peru ALIADOS: US$10,000 project can be expected to have positive assistance to producer organizations to for microenterprises, US$20,000 for externalities on producers that are not improve the “bankability” of subproject small enterprises, and US$30,000 for directly benefiting from project support. business plans, i.e., meeting the medium enterprises, as the expected In the case of the Colombia PAAP II, the standard requirements of commercial number of beneficiaries increases from final impact evaluation quantified such financial institutions; (ii) specific technical the former to the latter). In several PA positive spillovers and found a positive, assistance to financial institutions to projects in Brazil, financial support to significant income effect for producers adjust their services to the needs of subprojects resulted in over-dimensioned living near direct project beneficiaries. the beneficiary producers; (iii) a credit investments, given the much lower than Moreover, PA projects have been shown line to eligible financial institutions at expected number of subprojects and to increase the adoption of improved market interest rates; or, (iv) a diagnostic beneficiaries (e.g. Brazil Pará). The actual production practices, often leading to study of local financial institutions and average investments per alliance and positive environmental externalities; (ii) their coverage and service provision to beneficiary household are presented in lack of access to commercial finance: smallholder producers (e.g. Guatemala Section 8 for completed PA projects. the coverage of commercial finance PDER and Peru ALIADOS). Along the institutions in remote rural areas is 65.  Most Productive Alliance same line, the Mexico SPSB project still low (although in some countries projects accept both in-kind and cash establishes commercial financial this situation is changing with mobile contributions by the producers, which institutions as key technical service banking), so that access for smallholder has led to a bias towards in-kind providers which are hired to support the producers to financial services is not contributions. To encourage ownership management of financial services for each always a given; and, (iii) lack of collateral: and shared risk-taking assumed by of the project’s seven targeted products. smallholder farmers often do not have producers beyond their existing in- However, among the PA projects that (sufficient) collateral acceptable for kind contributions (e.g., machinery, have closed, the implementation of financial institutions to give them a land, productive assets and labor), these activities was often limited and loan. A fourth argument could be made some projects have made cash (or bank did not lead to systematic improvement that PA projects provide smallholder credits) contributions by producers a in producers’ access to commercial producers with longer-term finance requirement. For example, the Bolivia credit. The reasons for this are manifold, combined with capacity-building, a PAR II accepts only cash contributions as including: low coverage of financial service that commercial banks do not producers’ financial counterpart. In the services in the rural areas where the usually offer. However, these arguments Honduras COMRURAL, the co-financing PA projects operate, producers’ lack do not preclude an increased role for level is 70% to 90%, of which at least of collateral, and regulatory issues financial intuitions in PA projects, as their 30% has to be covered through a loan preventing financial institutions from selection process lowers the transaction from a financial partner. In Practice Box making loans to groups of producers (and costs of identifying quality business plans, 5 provides more details on this case. not individuals). The only project that which makes sense from a financial and has systematically involved commercial an agribusiness viewpoint. In general, 66.  The use of matching grants in financial institutions in subproject more could be done to include the Productive Alliance projects can be financing is the COMRURAL project, commercial financial sector formally justified by positive externalities, lack with its requirement that 30% of total in the PA approach, where desirable 35 Linking Farmers to Markets through Productive Alliances subproject cost be provided upfront but less formally, the Colombia PAAP through a loan from a financial service project established a mandatory revolving provider (In Practice Box 4). As such, fund for producer organizations, given Honduras COMRURAL introduces a fourth the difficulty of its beneficiary producers main agent - the financial sector - to the in accessing commercial credit. In other three core agents of producers, Practice Box 6 describes this alternative buyer and public sector. Alternatively, financial mechanism of revolving funds. In Practice Box 4: Leveraging Subproject Funding through broader Alliances (Colombia) In addition to the Project grant and co-financing from individual producers or the producer organization, Colombia’s PAAP I and II projects leveraged funds from different public and private sources, including local government agencies, municipalities and other development partners. Hence, while the matching grant provided by the Project was by design limited to a maximum of 30% of the total subproject investment costs, the remaining 70% came from various sources, including: i.  Producer/Producer organizations contributions (usually in-kind: land, equipment and materials) ii.  Private sector partner contributions (e.g., technical assistance, packaging materials, seeds) iii.  Grants from the contributions of Departmental Secretariats of Agriculture iv.  Grants from other institutions such as the National Service for Apprenticeship (SENA) and local governments (municipalities) While this approach successfully leveraged funds from public and private funding sources other than the Project, it also entailed certain risks. For example, the Colombia implementation experience showed that delays in the disbursement of these complementary contributions could delay the entire investment. Moreover, public funds for agriculture from local sources may be concentrated around a relatively limited number of producers, at the expense of producers in other locations. Finally, the producer contributions are usually provided in kind through existing assets (e.g., land, family labor) which - if valued at market prices - might inflate their value to meet co-financing requirements, whereas cash contributions are provided entirely through the Project grants. 36 In Practice Box 5: Upfront Engagement of the Commercial Financial Sector (Honduras) The Honduras COMRURAL project required that a subproject secure a loan from a financial partner, covering at least of 30% of total subproject investment costs, in addition to the required 10% in cash or in kind contribution by the PO. From the beginning of the project, a broad range of financial partners were identified as eligible to participate in co-financing the subproject: i.  Commercial banks, finance associations, private finance development institutions regulated by the Banking and National Insurance Commission; ii.  Credit and savings cooperatives affiliated with the Honduran Federation of Credit and Savings Cooperatives; and, iii.  Other micro-finance institutions, and other buyers such as input providers. During implementation, the Producer Organizations’ contribution was often higher than required, in some cases almost 50% of subproject costs. A lesson learned from the Honduras COMRURAL Project is that smallholders are able to obtain access to commercial finance despite the widely-held view that it is too risky for banks to lend to smallholders. The project has demonstrated that financial actors can lend funds to organized smallholders if an effective enabling environment exists. This includes: strong communication between counterparts and the banks; the setting up of a project guarantee fund; and, a longer-term engagement to build trust and confidence among all actors. As a result of their participation in the project, various producer organizations have received further loans. Thus, the project was able to demonstrate that smallholder producers can be attractive clients who repay their loans on time. Based on this experience, a new PDO-level indicator was introduced for the project’s Additional Financing, measuring the percentage of producer organizations without loans in arrears, and at the intermediate results level, the ratio of loans to grants in the business plans. 37 Linking Farmers to Markets through Productive Alliances In Practice Box 6: Revolving Funds as an Alternative Financing Mechanism (Colombia) In Colombia, many participating producer organizations experienced difficulty accessing commercial credit under PAAP I. The design of the second phase project PAAP II therefore included stronger measures to entice financial institutions to provide credit to beneficiary producers. As an intermediate step, a mandatory revolving fund was established which was initially financed using funds recovered through the partial repayment scheme of Project grant managed by the producer organizations. Producer organizations were encouraged to use and manage their revolving fund as a tool to finance their working and investment capital requirements. A secondary objective of the revolving fund was to allow producer organizations to build up their credit-worthiness, which was expected to enable them to obtain commercial financial credit. While access to commercial credit remained low during PAAP II, the revolving funds turned out to be a successful measure to finance (beneficiary) producer investments, technical assistance and to purchase specialized machinery. By completion of PAAP II, 399 revolving funds were in operation. Despite the popularity of the revolving funds, the Colombia PAAP project envisaged several steps to also gradually improve producer organizations’ (PO) access to commercial finance: i.  POs’ financial transactions managed by the trust companies could help them build a track record in the financial system ii.  POs should prepare audited financial statements iii.  Forward contracts could be established to obtain supplier credit iv.  Existing vertical alliances from PAAP I that need capital to expand should be assisted by a specialized person in their negotiations with the banking system v.  The PCU and OGR should approach financial institutions that offer credit to small enterprises such as banks, micro- credit institutions and NGOs 38 7 < 6. Subproject Financing 7. Results Monitoring > 8. Achievements and Outcomes 39 Linking Farmers to Markets through Productive Alliances RESULTS MONITORING 68.  Productive Alliance projects 69.  The Project Coordination Units limits the ability to monitor project engage in continuous monitoring and at the central and/or regional level are results at the outcome level and to regular reporting of project activities responsible for monitoring subproject conduct final evaluations. Consequently, and achievements in relation to activities and achievements. Such more guidance and training on data established project targets. These monitoring of subproject implementation needs and collection methods should targets are spelled out in the Project should be conducted according to the be provided to PCUs and POs by PA Results Framework, which includes both results indicators spelled out in the projects, as they often do not have the Project Development Objective and subproject business plans (which are specific skills required for these tasks. In Intermediate Results Indicators (with their typically aligned with the overall Project response to this identified need, some respective targets by year and for the end Results Framework). The beneficiary PA projects have been strengthening of project implementation).9 The Project producer organizations are required to their M&E Systems and provided training Coordination Unit (PCU) is responsible collect and report data periodically to the to consultants who accompany the for providing the necessary information PCU, which then systematically records subproject partners in recording key at the aggregate project level and for this information in a Project Monitoring data such as production costs, sales sharing it with the World Bank Task Team. and Evaluation System. The technical revenues and the use of hired and non- Depending on a project’s organizational assistance service providers hired by hired labor. Conducting periodic quality structure, this monitoring data collection the project are commonly expected checks of these data bases is crucial to can be done either centrally (if there is to validate the data provided by the enable measurement of the impact of the only a national PCU) or with the support POs. In addition, PCU staff regularly Productive Alliance approach at project of decentralized Regional Project Units visit each subproject for verification. completion. The issue of adequate (RPUs) that collect information for their monitoring data collection will be 70.  Many of the Productive Alliance respective region and share it with the addressed further in the Sections 8 and 9. projects reviewed could improve national PCU. Typically, projects develop their efforts to fulfill data needs and a Project Monitoring and Evaluation improve data collection methods for 9. Annex 7 displays the PDO indicators (M&E) System for this purpose. In the case from the Results Frameworks of the better results monitoring, as is the analyzed Productive Alliance projects. of PA projects, some have developed case also for other World Bank-financed sophisticated on-line platforms that lending operations. Regarding data include results as well as fiduciary availability at the subproject level, monitoring (e.g. Brazil Pernambuco, the productive alliance business plans Bolivia PAR, Colombia PAAP, Mexico typically contain baseline information SPSB), while others rely on Excel- on some important indicators (such as based data sets (e.g. Guatemala PDER, sales volumes, prices obtained and end Jamaica REDI, Panama PRORURAL). markets). However, experiences from In some cases, results monitoring has Mid-Term Reviews (MTR), Implementation been incomplete due to the lack of Completion and Results Reports (ICR), an adequate system, which had led to and ex-post economic analyses of PA partial, sub-optimal data collection (e.g. projects have shown that data on key Brazil Alto Solimôes, Brazil Pará). Annex indicators, such as production costs 6 provides details on monitoring and and sales incomes before and after the data collection activities at different subproject investment, have often not stages of project implementation, as been accurately collected. This lack of well as some examples of M&E online adequate and consistent data collection systems developed for PA projects. 40 8 < 7. Results Monitoring 8. Achievements and Outcomes > 9. Evaluation Strategies 41 Linking Farmers to Markets through Productive Alliances ACHIEVEMENTS AND OUTCOMES 71.  The achievements of Productive PRORURAL) to 104 (Guatemala the average amount between Colombia Alliance projects can be presented PDER). Details on project scope are PAAP I and II is assumed to be the under five dimensions: (i) scope, (ii), presented in Table A8.1 in Annex 8. result of a large increase in financial social inclusion, (iii) socio-economic contributions from other agents (e.g. 73.  Indirect beneficiaries of Productive impacts, (iv) efficiency, and (v) municipalities and other partners of Alliance projects remain largely sustainability. Of the PA projects in broader alliances). As projects did not unaccounted for. Similar to other World the Latin America region, eight have consistently report co-financing from Bank investment projects, the focus of completed their implementation. sources other than the Project, producers results monitoring and analyses is on the These are, in chronological order of and buyers, it is not possible to provide direct beneficiaries. However, it is likely their completion: Colombia PAAP I, this information. Details on the scope that a considerable number of producers Brazil Sergipe, Bolivia PAR I, Brazil Alto of the project investment support given benefit indirectly from positive spillover Solimôes, Brazil Pará, Guatemala PDER, are presented in Table A8.2 in Annex 8. effects of the activities undertaken by Panama PRORURAL and Colombia PAAP PA projects. These indirect effects (both II. The main achievements compared positive and negative) are usually not SOCIAL INCLUSION to targets and consequent outcomes measured in project monitoring systems. 75.  Productive Alliance projects of these projects are summarized for For the PA projects, only Colombia PAAP have performed well in including the five afore-mentioned dimensions, II started monitoring indirect beneficiaries women. Most of the PA projects have based on the projects’ Implementation during project implementation (since not established specific requirements Completion and Results Reports. 2010). The results show that in addition to or strategies with regard to including the 42,552 direct beneficiary households, women in their activities. Yet, many SCOPE at least 11,057 households benefited PA projects report on the number of 72.  Almost all Productive Alliance indirectly from the project. The related female project beneficiaries in their projects in Latin America exceeded spillover effects are discussed below project monitoring systems and include their appraisal targets for the number under socio-economic impacts. women’s participation in the projects’ of alliance subprojects, but not for the Results Frameworks (with specific end- 74.  The actual grant investment number of beneficiary households. of-project targets in the intermediate amounts per beneficiary household vary With respect to the number of vertical results indicators, which usually related greatly among projects, ranging from productive alliances with buyers (i.e. to a certain percentage of total project an average of US$514 (Colombia PAAP subprojects) to be established, all projects beneficiaries). A special case is Brazil II) to US$4,186 Panama (PRORURAL), and except for the Guatemala PDER and the Sergipe, which even included a results per alliance from US$30,370 (PAAP II) to projects in Brazil exceeded their targets. indicator on women’s participation at the US$189,450 (Brazil Pará). The variation However, only the Colombia PAAP I Project Development Objective (PDO) in those amounts is due to the different and II met and actually exceeded the level. According to these monitoring types of investments and upgrades targets for number of direct beneficiary systems, in most PA projects women’s promoted by each project (e.g. some producer households. This is because involvement and participation in projects promoted the procurement projects generally supported a larger productive alliances is much higher than of equipment or machinery, which than expected number of alliances, expected. For example, the Colombia resulted in higher investment costs per but with fewer beneficiary producers PAAP I successfully reached households beneficiary). These figures do not include per participating organization than headed by women (175% of the Results grants from other public sources and, anticipated at appraisal. The actual Framework indicator target). On a much hence, the actual total grant amounts average number of producer households higher level, PAAP II directly benefited received by alliances is expected to be per alliance ranges from 35 (Panama more than 9,900 female-headed producer higher. For example, the decrease in 42 households, nearly four times higher eligibility criteria and the usually rural SOCIO-ECONOMIC IMPACTS than the end-of-project target of 2,550 geographic focus. For example, in the 77.  Available evidence shows that households (23% of all beneficiaries, Bolivia PAR I project, 70% of the direct Productive Alliance projects have compared to the projected 10%). The project beneficiaries identified themselves generated significant positive impacts Brazil Sergipe project measured women’s as belonging to indigenous groups. No in terms of increases in production, leadership in productive investment specific target was set in the case of Brazil sales, income and employment. Table subprojects at the PDO-level and found Alto Solimôes, where about 31% of the A8.3 in Annex 8 summarizes the results that 31% of subprojects were led by State’s population are indigenous people. for the key outcomes and impacts of: women (86% of the end-of-project Yet, significant results were achieved (i) sales, (ii) income, (iii) employment, target) and 45% of total beneficiary by the project in terms of indigenous and, (iv) spillovers for the completed households were female-headed. Also in participation as about 60% of the 3,252 PA projects, as well as the source of Bolivia’s PAR I project, female producer direct beneficiaries were indigenous. the evaluation data. Evidence of results heads of household accounted for 32% In the case of Guatemala, indigenous for key indicators related to the Project of the membership of POs and 27% of peoples were a specific target population Development Objective stems from the representatives in PO management included in the PDO. At appraisal, the project monitoring systems, smaller-scale and monitoring committees. Although project expected that 80% of the 30,000 beneficiary surveys, and stakeholder no specific target was set, at project beneficiaries to be reached would consultations. Based on these, several completion 34% of Guatemala PDER belong to indigenous groups. At project projects have shown that the PA approach beneficiaries turned out to be women. completion, this target was surpassed, has led to increases in smallholder with 90% of the PDER beneficiaries being 76.  Vulnerable groups like indigenous producers’ production volume, indigenous. The Colombia PAAP II also or producers in post-conflict zones productivity and access to improved performed well regarding indigenous have been able to benefit from the inputs and productive equipment, as and Afro-Colombian households - even Productive Alliance approach on a well as to integration into new markets. though they were not a PDO-level target par with comparable non-vulnerable To complement these intermediate group - systematically monitoring their producers. PA projects typically have outcomes, some PA projects have participation from 2010 onwards. At the no specific requirements regarding conducted external impact evaluations completion of PAAP II, 9,250 indigenous the participation of indigenous or at project completion. The Bolivia PAR and Afro-Colombian households had other vulnerable groups. Yet, in many I and Colombia PAAP I engaged in benefited directly, exceeding by over countries across Latin America and the representative evaluations that included six-fold the end-of-project target of Caribbean, these groups constitute a control groups to rigorously measure 1,275 households. An assessment of major segment of the project target their impact.10 The results of these the performance of these groups in the population given the socio-economic evaluations have shown that beneficiary PAAP is described in In Practice Box 7. Figure 8.1 Main areas of PA project impact Sales Increases Employment 20% to 60% Quality improvements PA higher than baseline Generation of new jobs Project Income Increases Spillovers About 30% higher Improved practices than control group Income increases 43 Linking Farmers to Markets through Productive Alliances In Practice Box 7: Productive Alliances as a Tool for Effective Social Inclusion The International Center for Tropical Agriculture (CIAT) used data from the Colombia PAAP’s M&E system to assess the effectiveness of the project in reaching indigenous people, people of Afro-Caribbean descent) and internally displaced populations. The analysis showed that a much higher than expected participation of those groups in the productive alliances had occurred, sometimes exceeding the proportion of these groups in the overall population (9.9% of indigenous peoples as PAAP beneficiaries compared to 3.4% in the country’s overall population, 8.8% of displaced people compared to 7.3% in the population, and 8.5% of Afro-Colombians compared to 10.6% in the population). Furthermore, the study also compared the performance of alliances involving these disadvantaged groups to alliances involving producers from non-disadvantaged groups (CIAT, 2014). It found that the former performed as well and in some cases even better than the latter, demonstrating the suitability of the Productive Alliance approach for disadvantaged groups. producers benefit from better product jobs (on- and off-farm). For example, the 79.  Measuring results for capacity- quality and diversification and hence Colombia PAAP II is estimated to have building and institutional strengthening increased sales volume and prices and, generated around 10,000 full-time jobs has been challenging. Similar to other higher income of beneficiary households. between 2008 and 2015. However, the development projects, measurement of Specifically, increases in sales have evidence on jobs is based on relatively the strengthened institutional capacity of been reported to range between 20% small samples and varies strongly across producer organizations and implementing and 60% (measured by sales receipts, value chains. In terms of spillover effects, agencies promoted by PA projects has sales volume marketed, net sales only the Colombia PAAP has collected not been done systematically. However, revenue, or absolute sales value). information. The independent impact these results can be expected to have evaluation at completion of PAAP II a positive impact on the commercial 78.  The effect of Productive Alliance estimated that the indirect beneficiaries viability and sustainability of productive projects on beneficiary household in nearby project areas also adopted alliances. For example, in Colombia the income has generally been positive, improved production practices promoted results of the beneficiary survey showed although changes in income have been through the project or were able to that the capacity-building activities measured differently across projects benefit from collective goods paid supported by PAAP II benefited not and are based on varying sample sizes. for by the project. Another significant only individual producers and producer The strongest evidence on income economic spillover effect is the 24.4% organizations, but also buyers. One of the arises from the Colombia and Bolivia higher gross income found for these most important benefits identified during independent impact evaluations. The nearby indirect beneficiary producers the national and regional stakeholder average increase in net (agricultural) compared to a distant control group, consultation workshops was the hands- income of beneficiary households demonstrating that the impact of PAAP II on training provided to community-level compared to the control group was found on Colombia’s rural sector was amplified organizations, which allowed them to be 29% and 28%, respectively (at the beyond direct beneficiaries. Details on to assist producer organizations “in a 5% significance level). With respect to the impact dimensions for each closed holistic manner” (ICR Colombia PAAP II, employment, some PA projects were PA project are presented in Annex 8. 2015). In general, PA projects provide found to have led to improvements in capacity training to producers and employment and the generation of new 44 producer organizations, which includes data collection using small samples when conducting (ex-ante and ex- the teaching of improved production of alliances (without control groups).11 post) economic and financial analyses practices (Good Agricultural Practices Hence, as with other World Bank lending and longer time-frames should be and Good Manufacturing Practices) and operations, more evidence is needed considered. Moreover, the PA approach improved environmental management, for the rigorous assessment of results. can be seen as a public instrument which can be expected to contribute to for overcoming the initial investment positive environmental externalities (e.g. EFFICIENCY constraints related to perennial crops, prevention of soil erosion, improved justified by the positive environmental 81.  Most Productive Alliance projects soil quality). Moreover, one core input externalities of perennials and a longer- have resulted in satisfactory average of the PA approach is related to training term vision for the development of rates of return, but are characterized activities in business development, a country’s agricultural sector. by a high variability across products. covering topics like accounting All closed PA projects conducted ex- principles and methods, enterprise post economic and financial analyses SUSTAINABILITY management and marketing strategies. (EFA) at completion. In general, the 83.  The sustainability of a Productive 80.  More evidence is needed to financial internal rate of return (FIRR) Alliance project should be measured substantiate the results for socio- has been satisfactory at the commonly using different factors, such as: (i) economic impacts found so far. All Latin assumed discount rate of 12% and continuation of the initial vertical American PA projects have measured 10-year estimation period. However, alliance between producers and buyers results for social and economic outcomes all analyses found a high variability or the expansion of producers into through the Monitoring Systems and across the different products supported new or additional vertical alliances; Project Results Frameworks. In many PA by these projects (Table 8.1). This (ii) POs maintaining or increasing projects, final evaluations were part of the heterogeneity is not surprising, given their membership; (iii) POs obtaining original design, but rigorous empirical the different production and market improved access to credit or other evidence remains limited (as many of the characteristics across products and commercial financial services, or projects are still under implementation). value chains. However, this experience establishing self-managed revolving This evidence gap is not particular to underlines the importance of early funds for more sustainable funding; or, the PA portfolio, but to development assessment of the long-term market (iv) POs establishing maintenance plans interventions in agriculture in general potential of the products to be financed and collecting fees to maintain capital (Goldstein, 2016; IEG, 2011).To help under a PA project. Annex 8 presents investments. In general, assessing fill this knowledge gap, the Bolivia details on the methodologies and sustainability is challenging, given that PAR I and Colombia PAAP I conducted results of the economic and financial the majority of subprojects start their independent, large-scale impact analysis of specific projects. productive cycle (and hence sales) evaluations with control groups. In the during the last years of subproject 82.  The proportion of investments in case of Brazil Pará, the final evaluation implementation. Moreover, evidence is perennial crops influences a project’s included a post-restructuring baseline scarce, as only a few ex-post evaluations estimated rate of return. In PA projects, and an end-line data collection were are ever done more than five years many subprojects have been linked to conducted with treatment and control after overall project completion investments in perennial crop production groups. However, the results need to be (which is common for all World Bank (e.g., cacao, coffee, rubber, lime, mango, interpreted cautiously due to the small financed projects). If monitored at avocado). Due to their long gestation sample. Other completed projects (Brazil all, PA projects usually assess the periods, production start-up - and Alto Solimôes, Brazil Sergipe, Colombia sustainability of alliances one or two hence sales and revenues - is delayed PAAP I, Guatemala PDER, Panama years after project support has finished. compared to other crops. Slow sales PRORURAL) collected information An exception with a longer-term view growth lowers the projections for the for evaluating their performance, but is the Independent Evaluation Group’s value of sales and hence the estimated the lack of representative baselines Project Performance Assessment Report internal rates of return at a standard and control groups meant that they (PPAR) on the Colombia PAAP projects. project completion time of five to six relied mostly on data available in years. This should be taken into account the project M&E systems or primary 45 Linking Farmers to Markets through Productive Alliances In Practice Box 8: Dissolution versus Failure of a Productive Alliance (Colombia) It is critically important to distinguish between the dissolution of a vertical productive alliance between producers and buyers, and the failure of a PO. A vertical alliance can be dissolved for many reasons. It might be assumed that the reason is the non-viability (failure) of the PO, but this assumption is not supported by evidence from Colombia PAAP I. Other possible reasons include: the inability of the PO to produce the agreed volume of product(s) and/or meet the agreed quality criteria; the buyer’s inability to meet purchasing commitments; changing market conditions; the buyer’s evolving business priorities; or, the failure of the buyer. Data from the Colombia PAAP M&E System show that of the 51 PAAP I vertical alliances that were dissolved, 50% could be attributed to changes in market conditions and/or the business cycle. Moreover, the dissolution rate diminished steadily during the life of the project: the dissolution rate averaged 54% from 2002 to 2004 before dropping to 31% from 2005 to 2007, and to 25% by 2007. Results of the CIAT impact evaluation underline the fact that the dissolution of a vertical alliance between producers and buyers does not necessarily signal the failure of the PO. This study showed that following the dissolution of a productive alliance, 70% of the POs entered into a new vertical alliance with a new buyer. Of those, more than 80% did so within six months after the dissolution of the initial alliance, and 98% within two years. Table 8.1 Financial Rates of Return Average Average Project Sample FIRR* FIRR** Variability in FIRR* Bolivia PAR I 535 alliances 49% 35%   43% of subprojects with FIRR>=12%; Brazil Alto Solimôes 26 subprojects 14%*** N/A 57%<12% (29% with FIRR <0). 43% of mi- cro-projects with FIRR>12% and 57%<5% 40% of subprojects with FIRR>=12%; Brazil Pará 10 subprojects N/A N/A 60%<12% 35%(56%) of subprojects with FIRR>=10%; Brazil Sergipe 23 subprojects 39% N/A 65(45%),10% Colombia PAAP I 23 alliances 23% 19% 69% of alliances with FIRR>=12%; 31%<12% Colombia PAAP II 56 alliances 29% 27% 79% of alliances with FIRR>=12%; 21%<12% Guatemala PDER 39 alliances 20% 17%   Panama PRORURAL 12 alliances 11% N/A About 33% of alliances with FIRR<0 *includes only direct costs **includes direct and indirect costs (i.e., proposal preparation, technical assistance, project management and monitoring). *** EFA not clear on which costs were included in analysis, but it is assumed only direct costs 46 84.  There is evidence of long-term 85.  The prospects for sustainability 10. The methdologies of these impact vertical alliances between smallholder seem better for producer organizations evaluations are summarized in Annex producers and buyers. For example, that existed prior to the Productive 8. Details can be found in the projects’ Implementation Completion and Results in the Panama PRORURAL project it Alliance project intervention, according Reports. was found that 80% of the 130 alliances to available data from project impact 11. For example, the Guatemala PDER continued to operate one year after evaluations and anecdotal evidence project conducted a final external evaluation through an independent project support ended. In addition, from stakeholder assessments (In consultant firm based on a sample of 16 documentary evidence shows that Practice Box 9). These organizations alliances out of 174 alliances operating at project completion. new members of beneficiary producer benefit from having an established organizations increased 54%, implying track record in collective action, rather that membership of a producer than organizations formed just for the organization supported by PRORURAL purpose of accessing project support. was perceived as valuable. Continued monitoring of Colombia PAAP alliances even after project closure, showed that most of the productive alliances had continued to operate even seven years after project completion. Specifically, 62% of first phase partnerships were continuing to operate seven years after PAAP-I closure. In the case of PAAP- II vertical alliances, 80% of producer organizations continued to sell to a buyer 24 months after project support ended (In Practice Box 8). Furthermore, the CIAT study demonstrated that the vast majority of producer organizations were capable of establishing new alliances when an earlier alliance dissolved (CIAT, 2014). Regarding other dimensions of sustainability, the survey results from 83 beneficiary vertical alliances in the impact evaluation of PAAP II show that 82% of beneficiary producer organizations are collecting annual membership fees to finance their operations and to maintain capital infrastructure. Another factor influencing the sustainability of production is productive reinvestments, with PAAP II beneficiary producers being more likely to reinvest part of their net revenues in productive activities than the control group. 47 Linking Farmers to Markets through Productive Alliances In Practice Box 9: Factors of Sustainability (Bolivia and Colombia) The following findings regarding the factors influencing the sustainability of productive vertical alliances between producers and buyers were documented in Implementation Completion Reports: Bolivia PAR I: Alliance sustainability depends on: (i) production viability at the level of the producers themselves; (ii) continuation of PO services to members; and, (iii) sustained relations with market agents. Close to 94% of POs had an IRR exceeding 12.5% at closing, showing that most of the producers are currently engaged in financially viable initiatives. Anecdotal evidence indicates that this is more likely in the cases where the organization itself assumes a role in the production cycle (e.g. input purchases, technical assistance, post-production processing and sales). Results from the beneficiary survey indicate that alliances which reconcile the main interests of producers (higher income) and buyers (product quality) are likely to be sustainable. Both these conditions are dominant in the project outcomes. About 33% of the POs changed their buyer between alliance identification and completion, and preliminary results from the beneficiary survey indicate that about 25% had quit a formal arrangement with a buyer. Colombia PAAP I: The “age of the alliances matters, as profitability matures over time. However, even if an alliance fails, the impact on production, income and marketing may still be significant: producer organizations will have gained experience in dealing with buyers and should be able to identify and negotiate with other buyers” (ICR Colombia PAAP II). One major conclusion from the beneficiary survey is that the sustainability of the alliance depends largely on the operation of the POs; beneficiary producers perceived that positive results flow from well-organized operations, teamwork and financing. 48 9 > 8. Achievements and Outcomes 9. Evaluation Strategies < 10. Lessons Learned 49 Linking Farmers to Markets through Productive Alliances EVALUATION STRATEGIES 86.  Productive Alliance projects often monitoring of project indicators. latter should be measured through lack adequate baseline data. As with Lessons from Brazil Alto Solimôes, Brazil complementary, final project evaluations other development projects, missing Pará, Colombia PAAP II, and Panama conducted several years beyond project baselines are often due to the high PRORURAL highlight the need for an closing); and, (iii) product documents cost of household-level data collection M&E unit with a clear operational plan to (PAD, MTR) make more frequent in rural areas and the demand-driven be established early on within the Project references to project-specific baseline nature of most subprojects, which make Coordination Unit. Furthermore, to data. Furthermore, there is an ongoing it impossible to know ex-ante which facilitate impact evaluation data collection transformation of PA projects to monitor products will be supported during project at project completion (especially for the and evaluate not only socio-economic implementation. To reduce the likelihood ex-post economic and financial analyses), effects, but also to assess co-benefits of ending up without robust baseline PA projects should train beneficiary for the environment (Mexico SPSB and data, establishing a representative producer organizations to establish Panama PRORURAL), social inclusion baseline and control groups should be and maintain records and accounts (Brazil Sergipe, Panama PRORURAL), and a priority following effectiveness, taking on production, yields, costs, sales nutrition (Brazil Bahia, Haiti RESEPAG II). into account data collection for vulnerable volumes, prices and/or sales revenues 90.  Productive Alliance project teams groups. Where available and possible, associated with their productive activities are increasingly developing a Theory such efforts should explore whether and the measurement of key project of Change to more systematically useful data may already exist, and they indicators. A systematic monitoring of include evaluation strategies in their may want to seek out opportunities compliance with these activities would project designs. The concept of a for securing additional (external) limit inconsistencies in the analysis of Theory of Change is to demonstrate how funding for baseline data collection. subproject profitability, from which an intervention is expected to lead to several PA projects have suffered. Thus, 87.  Furthermore, outcomes for desired results (Kusek and Rist, 2004). a project monitoring and complementary Productive Alliance projects are Designs of a Theory of Change can vary, evaluation strategy should be part of often too ambitious for the project but typically include the results chain the design of each PA and be spelled engagement period. Several PA projects elements of: (i) inputs, (ii) activities, out in the Operational Manual. have defined their Results Framework (iii) outputs, (iv) outcomes, and, (v) PDO indicators and the respective 89.  In recent years, Productive impacts (Morra Imas and Rist, 2009). In targets at levels which can only be Alliance projects have strengthened addition to the five main elements, a achieved longer-term (e.g. changes in their M&E strategies and revised their Theory of Change can also include the household income). Such higher-level Results Frameworks. Recognizing the description of target groups, risk factors outcomes are unlikely to materialize by limited, rigorous evidence - beyond and mitigation activities, or responsibility the end of an initial PA project of five the impact evaluations of the Bolivia assignments. The details depend on or six years. Consequently, PA projects PAR I and Colombia PAAP II projects the needs and purposes of the project should focus on realistic and achievable - trends in the M&E of PA projects team and client. A Theory of Change medium-term outcomes. Table A9.2 have been promising: (i) (Restructured) should ideally be developed at the in Annex 9 provides a prototypical Results Frameworks contain fewer and project design stage, but the concept theory of change indicating varying more precise PDO indicators (e.g. can also be used post-effectiveness levels of outcomes for a PA project. moving from absolute values towards to assess the consistency of the logic percentage increases compared to behind the existing Results Framework, 88.  As a result, an increased focus defined benchmarks); (ii) PDO indicators which may imply restructuring. Annex is needed on developing evaluation are increasingly defined at the outcome- 9 provides a brief description of the strategies for Productive Alliance level instead of impact-level (as the elements and logic behind the workings projects which are aligned with the 50 of a Theory of Change, and presents groups at different organizational of total buyers involved in the project- a prototypical Theory of Change for stages, as Section 3 describes) sponsored alliances who meet the terms a Productive Alliance project based described in the subproject business plan. »» The PDO should be clearly related on workshops with project teams. Similarly, the Brazil Pernambuco project to the specific PA activities (to innovated by including an Intermediate 91.  Key recommendations for reflect if PDO is related to increasing Results Indicator on the maintenance of the Monitoring and Evaluation of market access, market share, project-financed productive infrastructure, Productive Alliance projects have competitiveness, productivity, or else) also linked to Recommendation 7. emerged. When working out the details »» Information on buyers should be In that case, the Results Framework of the development or revision of a collected systematically in the Results was revised at the Mid-Term Review project’s M&E design, the most commonly Framework (so far, PA projects have by adding a maintenance indicator found challenges relate to unclear focused almost entirely on data for the complementary water and definitions of indicators in project Results collection from producers only) sanitation infrastructure investments: Frameworks. These challenges are not monitoring the percentage of beneficiary unique to PA projects, but project teams »» Information on maintenance provisions producer organizations “with organized in the Latin America region have sought should be collected systematically in arrangements for the management, to address them specifically for PA-type the Results Framework (as PA projects operation and maintenance of financed projects. Based on several workshops provide productive infrastructure productive investments.” By organized and group discussions, the following which requires regular maintenance arrangements the project refers to: “(i) recommendations were identified: to ensure long-term operation) designated people responsible for it; (ii) an established reserve fund for operation »» PDO indicators should not go 92.  Several Productive Alliance and maintenance (including replacement); beyond (medium-term) outcome level projects have internalized these and (iii) [producer organization] (due to difficulties in measurement recommendations and developed members are actively paying an agreed and unrealistic expectations for innovative Results Frameworks. For and adequate sum on a regular basis the typical 5-year project) instance, in the case of Brazil Pernambuco (annually) into that fund.” In Practice Box and the Additional Financing of Honduras »» Baseline and impact evaluation 10 describes an additional, new approach COMRURAL the PDO indicators were data collection should be part of adopted by the Brazil Pernambuco revised to the outcome level and not at M&E design and budget planning project at Mid-Term to define key the impact or output level. Both projects (complementing the standard Results success factors of productive alliances. as well as the Brazil Paraíba Sustainable Framework monitoring to have a Rural Development Project appraised 93.  An increasing number of control group for adequate comparison in 2016 have all (re-)defined their PDO Productive Alliance projects have and attribution at project completion) indicators related to sales as a percentage engaged in larger-scale data collection increase in the value of gross sales of »» Standardized measurements of PDO activities to enable rigorous assessment the beneficiary POs. This will permit the indicators should be developed of their effectiveness at completion comparison of results across projects, across PA projects for purposes at the producer organization - or even taking care to consider countries’ specific of comparability (e.g. % increases household - level (e.g. Bolivia PAR I, circumstances influencing sales and instead of absolute values) Colombia PAAP, Mexico PSPB, Brazil market conditions. Furthermore, the Brazil Pernambuco). Most recent projects »» The commonly used PDO indicators Paraíba project has made an innovative have taken the above-mentioned on sales should be interpreted with addition to the Results Framework based Recommendation 2 seriously and caution due to related factors beyond on Recommendation 7 whereby the are including evaluation strategies a project’s control (e.g. business Appraisal Package Intermediate Results and data collection activities within cycle, market conditions, regulation) Indicators include “buyers who fulfill their a year of effectiveness in their M&E obligations under the commercialization design (e.g. Bolivia PAR II). This results- »» The PDO should be defined to agreement or business plan”. This oriented approach can be expected include all beneficiary groups (as first-time inclusion of a buyer-related to continue and increase in future PA PA projects have target beneficiary indicator is defined as the proportion projects. In Practice Box 11 describes 51 Linking Farmers to Markets through Productive Alliances the partnership between the World of the PA approach. In Practice Box 12 Bank Agriculture Global Practice and highlights the experiences and lessons the Development Impact Evaluation learned from using different evaluation Unit, launched in November 2009 to methodologies for PA projects. generate rigorous evidence on the impact In Practice Box 10: Measuring Successful Productive Alliances (Brazil Pernambuco) During the Mid-Term Review, the Brazil Pernambuco Task Team revised the definition of a PDO indicator measuring successful productive alliances. As the original description of what was meant by “successful” was unclear, the Team selected six key criteria that it considered indispensable for a successfully functioning productive alliance. Although developed for one specific project, these criteria could be applied to other PA project Results Frameworks, as they refer, among others, to: (i) the sustainable operation of project-supported productive investments; (ii) their maintenance; (iii) the producer organization’s accounting capacity; and, (iv) diversification of buyers. Furthermore, the revised indicator targets took into account that realistically, not all beneficiary producer organizations would comply with all six criteria at project completion. The concept of “successful” in this context was based on compliance of 60% (i.e. 4 out of 6) of the criteria listed. The revised PDO indicator for a successful productive alliance, its related targets and descriptions are displayed below. Cumulative Target Values Unit of Data PDO Indicator Measure Baseline YR3 YR6 Freq. Source Description “Successfully” is based on compliance with the Beneficiary Producer Organizations Successfully inserted into LPAs. % of beneficiary Producer Organizations complying with at least 4 out of the 6 criteria N/A 50 75 Twice: Mid term and end of project Mid term and final evaluations following criteria at least 12 months after conclusion of subproject financing: beneficiary. Producer Organization has: 1. Investments of business plan implemented 2. Financed investments operational 3. Production volume is at minimun 80% of target as in approved business plan 4. Functioning accounting system 5. Maintenance plan and reserve fund for financed investments 6. Sales to at least one non-institutional buyer 52 In Practice Box 11: Building Partnerships for Obtaining Evidence The partnership between the World Bank Agriculture Global and the Development Impact Evaluation Unit (DIME) includes several projects in Brazil and Haiti. The evaluations seek to identify the observed changes to the outcome indicators and quantify how much can be attributed to the PA project interventions. DIME works closely with the projects to ensure that a suitable control group is identified and that data is collected at baseline and end-line. Most evaluations look at the overall PA approach to see if providing a package of matching grants for productive investments, technical assistance, business development training, and the identification of business opportunities can relax the constraints on smallholder producers in terms of technology adoption and innovation, productivity growth and access to markets. In the case of Brazil Ceará, it will also be possible to assess the relative importance of the different barriers. By assigning the eligible producer organizations to receive either managerial training or financial support plus business plan development, the evaluation team will be able to disentangle the impact of the different components. In Haiti, it will look at the additive impact of a business skills program for women within the organizations. Besides generating knowledge on the impact of the PA projects, direct engagements with the government counterparts seek to strengthen their capacity to implement evidence-based policymaking using a learning-by-doing approach and training events. 53 Linking Farmers to Markets through Productive Alliances In Practice Box 12: Learning from Different Evaluation Methodologies The evaluation work of DIME assessing the impact of Productive Alliance projects can be divided into two phases. The first phase used non-experimental evaluation methods to identify the impact in Brazil São Paulo and Santa Catarina as well as Haiti. At that time, a plausible empirical strategy seemed to be to compare producer organizations whose business plans ranked just below and just above the threshold for being selected for support (using the method of Regression Discontinuity Design, or RDD). Although RDD is seen as the non-experimental that most resembles a randomized controlled trial, it has limitations. First, it requires a large number of observations around the threshold. Second, when successfully implemented, it informs the impact of the program only for the organizations close to the threshold. After the data collection, it turned out that in all projects the initial take-up was relatively low and the score distribution was very heterogeneous. This resulted in a small number of producer organizations around the threshold and lack of statistical power to evaluate the projects using the original design. As a result, the evaluation team decided to use Propensity Score Matching (PSM) to select a group of control organizations that at baseline most resembled the selected organizations. The credibility of PSM relies on the availability of lots of baseline characteristics of both beneficiaries and a large group of non-beneficiaries, which is often not available. With these experiences in mind, the second phase of evaluations have been set up with the experimental evaluation method of Randomized Controlled Trials (RCT) instead. RCT ensure credible estimates of program impact, requiring far fewer assumptions and small samples compared to non- experimental methods. The projects for which the RCT method is used are in the Brazilian Northeastern States of Ceará and Bahia (see table below). Project Method Brazil Santa Catarina Propensity Score Matching Brazil São Paulo Propensity Score Matching Brazil Ceará Randomized Controlled Trial Brazil Bahia Randomized Controlled Trial Propensity Score Matching, RCT of individual Haiti RESEPAG II women First results of these evaluations are expected in early 2017. To the extent known, they will be based on one of the few experimental evaluations of matching grant programs in agriculture like the PA approach. Hence, the findings are expected to contribute to filling a large gap of evidence in this field. 54 10 < 9. Evaluation Strategies 10. Lessons Learned > 11.Conclusions and Recommendations 55 Linking Farmers to Markets through Productive Alliances LESSONS LEARNED 94.  Several lessons emerge from Reports of closed PA projects are 96.  Beneficiary targeting and the Productive Alliance experience in presented at the project preparation, selection must consider producers’ Latin America, which are relevant not implementation, and closing phases. existing endowments and assess only for a particular project but across their ability to comply with market the PA portfolio and future operations. LESSONS FROM PROJECT requirements. Not all small producers Although the mode of implementation have the capacity or risk profile required PREPARATION of the PA approach has varied between for market-driven partnership with a 95.  More emphasis should be placed countries, the experience has shown that buyer. Buyers’ incentives to deal with on identifying and analyzing promising PA projects have succeeded in supporting small producers relate to product quality, value chains. Such analysis should be producer organizations to improve volume and timely delivery. Small done during project preparation and organizational capacity, develop more producers can reach these requirements must be based on a realistic assessment entrepreneurial mindsets, and to increase by being part of a producer organization. of the local/regional comparative the production, product quality, diversity However, this collective action cannot advantage in specific value chains. In and productivity of the organization’s overcome a producer’s absolute environments where competition is likely members. Technical assistance and constraints in access to land and water. to be limited, as in remote areas with business development support provided logistical and technical challenges, a 97.  Setting realistic outcomes aligned by PA projects on a continuous, medium- pre-investment market study should be with project activities and investments term basis have proven to be essential carried out to identify potential buyers is critical for success. Over-ambitious for strengthening POs and laying the and understand their size and limitations. project targets and unclear indicators can ground for entrepreneurial sustainability. pose significant challenges. For example, In the following, main lessons learned increased sales is a challenging indicator and success factors identified in of competitiveness in demand-driven Implementation Completion and Results 56 projects like the PAs, as the exact size LESSONS FROM PROJECT 101.  Accompanying producer of the gains to be made from project groups over an extended period while IMPLEMENTATION investments (in value terms) is difficult they grow and mature is crucial for 99.  An open, competitive subproject to know a priori. Also, the “sales value” ensuring long-term success. Instead selection process based on clearly- indicator (particularly in global commodity of providing a one-off injection of defined technical evaluation criteria sectors) is linked to exogenous factors resources, PA projects should accompany can be crucial for establishing outside the project’s control, such as beneficiary producer organizations credibility among stakeholders and changes in commodity prices. Expressing (and buyers) over the course of their avoiding political interference. PA the increase in sales as a percentage alliance subproject implementation project beneficiaries have indicated of initial sales or in volume terms (in period and into the operational phase, that a competitive process ensured percentage) would be more appropriate. with the goal of building the capacity that subprojects were approved based Overall, PDO outcome indicators needed to ensure long-term survival. on their merit and viability, leaving aligned with gains in productivity, few opportunities for financing to be 102.  An assessment of potential efficiencies, quality improvements, steered by government officials or local buyers can ensure continuity of a and reduction of postharvest losses authorities to friends and associates. vertical alliance. In some projects, seem appropriate for PA projects. producer organizations found that their 100.  Requiring cash contributions 98.  Operations involving a multi- increased productive capacity surpassed or bank loans as co-financing from sector approach with different their current buyer’s capacity. Measures producers can ensure a stronger institutions, components, and areas to mitigate this include improved buy-in. The provision of financial of concentration are challenging buyer identification and their market resources which are not in-kind to implement. Ground-breaking competitiveness and viability, brokerage creates greater ownership of the operations with new borrowers need services for organizations that outgrow success of the supported subproject to be kept simple, use piloting and their current alliance arrangements, and allows producer organizations to evaluation for future scale-up, and and an increased outreach to potential strengthen their entrepreneurial skills receive close supervision so that the buyers outside the local sphere. and commitment to the subproject. learning opportunity is optimized and the client is capable of entering a follow-on phase with confidence. 57 Linking Farmers to Markets through Productive Alliances 103.  Capacity-building activities for 106.  Building broader alliances is producer organizations need to adapt crucial for post-completion operation to specific business needs. In some PA and sustainability. Ideally, the borrower projects, capacity-building of producer country will sustain its commitment to organizations has focused on general the PA approach and ensure budget subproject managerial requirements (e.g. support. However, this is often not the accounting, procurement), which have case. Hence, mechanisms are needed been of high importance. However, it to graduate from donor dependence has also become evident that capacity- to ensure continued support for certain building needs to be tailored to the elements of the project. An important specific functions each producer role can be played by broader alliances organization assumes in the vertical with state agencies and/or municipalities, alliance established with a buyer. For which in some PA projects were key for example, organizations may be asked to some of the successful activities. Such reduce transaction costs for producers alliances within the broader institutional (e.g. input purchases) and negotiate framework can be strategic partners with current and future partners, but for further development. Furthermore, they may also need to operate collective alliances with the financial sector infrastructure (e.g. storage, processing such as commercial banks are highly- and packaging), or organize the provision relevant for financial sustainability. of technical assistance services to 107.  Productive Alliance projects members. All these needs require specific should develop a plan for the transition knowledge, for which support should to post-World Bank financing. To be included in the business plan. ensure that activities move smoothly to government agencies following the LESSONS FROM PROJECT closing of the World Bank-financed CLOSING PA project, an exit strategy should be 104.  Productive Alliances can defined early on. Ideally, transition be a cost-efficient way to boost planning should begin at the time productivity, expand production, of the Mid-term Review and be improve competitiveness, and link monitored continuously thereafter. farmers to markets. A large number of closed PA projects have shown satisfactory financial rates of return and largely achieved their objectives. 105.  The Productive Alliance approach is flexible and can be adapted to fit the needs of many different target groups, value chains, and production environments. It has been an effective tool for targeting not only well-established producer groups and value chains, but has also shown the capacity to include disadvantaged groups successfully. 58 11 < 10. Lessons Learned 11.Conclusions and Recommendations 59 Linking Farmers to Markets through Productive Alliances CONCLUSIONS AND RECOMMENDATIONS 108.  This assessment describes the (Northeastern) Brazil where projects have productivity, and access to improved Productive Alliance approach and transitioned from former Community- inputs and productive equipment, as presents the results of a systematic Driven Development interventions to well as integration into new markets. review of World Bank-financed the PA approach. As a result, some Furthermore, beneficiary producers have Productive Alliances projects across projects in Brazil have been implemented been found to benefit from (i) better Latin America and the Caribbean. with producers on the community product quality and diversification, (ii) The PA approach has been adopted in level and heavily focused on providing increased sales volume and prices, and 21 projects across ten countries with production investment support rather hence (iii) higher income. Also, a few funding of over US$1 billion. Given than ensuring coordinated sales to PA projects have assessed their effect the recent expansion of the approach buyers. Furthermore, while PA projects on employment generation and quality, to other regions, the findings of this outside Brazil seek to establish vertical which have largely been positive. Results assessment - based on a detailed alliances with private sector commercial from beneficiary and stakeholder surveys overview of different PA projects in markets, Brazilian producers rely heavily of PA projects have highlighted their terms of targeting strategies, subproject on supplying to public institutional prominent role in capacity-building selection, implementation, monitoring markets. Recently, a new emphasis and in institutional strengthening of and evaluation - provide valuable lessons on environmental sustainability and producer organizations and implementing on the different design features and climate smart agriculture has emerged agencies. Finally, ex-post Economic and implementation experiences across among PA projects, promoting “green” Financial Analyses have shown that most Latin America since the early 2000s. PA as in Mexico or Panama. Similarly, PA projects have acceptable internal rate a greater focus has been placed of returns, although performance is highly 109.  The concept of the Productive on beneficiary targeting and social variable across different value chains. Alliance approach is based on three inclusion of vulnerable groups in PA main agents (Producers, Buyers, Public 111.  The following recommendations projects, and in some cases on their Sector) and three core inputs and/ consider the needs for strengthening nutritional effects. This amplification or activities (Productive Investments, future projects to enhance the impact of objectives and related co-benefits Technical Assistance, and Business and sustainability of the Productive demonstrates the flexibility of the PA Development). As this assessment Alliance approach, in consideration of approach, while preserving its core demonstrates, PA projects can vary in the achievements and lessons learned objective of smallholder integration the emphasis given to each individual described in this assessment:12 in agricultural value chains. core element, but all aim to strengthen smallholder producers’ capacity to 110.  The effectiveness of completed »» Productive Alliance projects need improve production and competiveness Productive Alliance projects has been to prominently promote building (through horizontal alliances) and access assessed in terms of scope, social broader alliances through a stronger markets through linkages with buyers inclusion, socio-economic impacts, involvement of local actors in (through vertical alliances). The distinct efficiency and and sustainabilty. In the enabling environment (e.g. types of PA projects can be categorized general, the PA approach has been municipalities, local chambers of based on different objectives, degree successful in reaching smallholder commerce, NGOs) in the initial of organization among beneficiary producers and including vulnerable subproject selection process and in the producers, or end markets. In general, groups like women, indigenous peoples supervision of their implementation. the entry point of the PA approach is and young producers. The available This would not only enable them to the producer organization. In practice, evidence points to the PA approach as tap into local knowledge and business the degree of organizational structure having led to increases in smallholder opportunities, but also to better varies greatly, particularly for the case of producers’ production volume, integrate the alliance subprojects in 60 municipal development plans and projects have already done this (e.g. 12. Further areas of attention in future related investments and services. Mexico SPSB, Panama PRORURAL) by projects and proposed actions to adding to their typical socio-economic address them are described in Annex 10. »» Productive Alliance projects could development objectives certain develop and adopt a more systematic desirable higher-level objectives related approach linking beneficiary producers to: (i) environmental sustainability with the commercial financial (“productive landscapes”); (ii) social sector. This might be done through inclusion and to a lesser extent guarantees provided by the overall (iii) nutrition. Given its flexibility in project to stimulate the willingness design and implementation, the PA of financial intuitions to co-finance a approach is adaptable and can be significant part of the business plan used to address several objectives through a loan. A good example is in a more holistic manner. the Honduras COMRURAL project, which requires a financial institution to »» Productive Alliance projects should be part of the commercial agreement incorporate an impact evaluation from the beginning. Furthermore, strategy and respective budget early PA projects could increasingly focus in their design stage. As this report on meeting other financing needs has documented, despite widespread of producers, such as payment application of the Productive Alliance systems, insurance, or savings. approach, little large-scale, rigorous evidence exists on its effectiveness »» Productive Alliance projects need at the producer organization and to more intensively promote the household level. Because of this lack diversification of buyers and markets of evidence, key design questions are to which beneficiary producers are as yet, unanswerable. For example, seeking to connect. A more diversified a primary issue resulting from this portfolio of offtake markets can assessment is how the performance of serve as an instrument to increase conditional PA projects relates to their producers’ resilience to external unconditional counterparts. Although shocks (e.g. negative economic conditional PA seem intuitively more or climatic events) and to protect effective at integrating producers producers from exploitative behavior into markets, there is insufficient by buyers (especially in imperfect evidence to explain the basis for market environments. Diversification this intuition. Therefore, future PA is especially important for the case of projects should include relevant Brazil, where producers rely heavily evaluation questions in their results on selling most of their production monitoring design and incorporate to institutional markets (e.g. school baseline and impact evaluation data feeding programs, public hospitals, collection in their budget planning. and the like). Similarly, diversification of products (in addition to the main project-supported products) could be promoted by future PA projects as well. »» Productive Alliance projects might consider adopting a multi-sector approach, where appropriate and desired by the client. Some, recent PA 61 Linking Farmers to Markets through Productive Alliances ANNEX 1 DESCRIPTIONS PA PROJECTS IN LATIN AMERICA AND THE CARIBBEAN This Annex briefly describes in million, of which IBRD provided US$30 project cost amounted to US$79.7 million, chronological order the 21 Productive million. A distinguishing feature of PAAP II with the contribution of US$19.8 million by Alliance projects that had been was its ability to leverage funding from the the producer organizations and municipal implemented across ten countries in public and private sectors. At completion, governments being nearly 50% higher Latin America and the Caribbean. the total cost of the second phase was than estimated. In total, the project had well beyond appraisal estimates and financed 768 productive alliances and In Colombia, the first Productive amounted to US$346 million, of which counted 28,527 direct beneficiaries. Partnerships Support Project (PAAP I) was US$190 million were contributed by the implemented between May 2002 and A follow-on Bolivia Second Rural Alliance producer organizations themselves, US$90 September 2008. The Project Development Project (PAR II) started in September 2013 million by the national government, US$32 Objective (PDO) was to generate income, and the estimated closing date is November million by local governments, and US30 create employment, and promote social 2017. The PDO was slightly modified million by IBRD as planned. In total, 744 cohesion of poor rural communities in an from PAR I to improving accessibility productive alliances were financed, directly economic and environmentally sustainable to markets for small rural producers in benefiting over 45,000 producer households manner through the development and selected areas by: (i) promoting productive in 31 of the country’s 32 Departments. implementation of a demand-driven alliances between different small rural productive alliance scheme with the In Bolivia, the Rural Alliances Project producer organizations and purchasers; private sector. The main intervention (PAR I) was implemented between May (ii) empowering rural producers through mechanism consisted of matching grants 2006 and March 2014. The PDO was the establishment and strengthening of for participating smallholder producer to test a model to improve accessibility self-managed grass-root organizations; organizations to co-finance productive to markets for poor rural producers in (iii) increasing access to productive assets, investments, complemented by technical selected pilot areas by: (i) promoting technology and financial services; (iv) assistance, supervision, and training for strategic productive alliances between promoting more effective, responsive and the productive partners. The actual total different economic actors at the local level, accountable service organizations at the project costs amounted to US$30 million, (ii) empowering rural producers through local level; and, (v) enhancing environmental of which US$22 million were funded by the development of self-managed grass- sustainability of productive practices. The the International Bank for Reconstruction root organizations, (iii) increasing access total project cost is estimated at US$64.54 and Development (IBRD) and US$8.44 to productive assets and technology, and million, comprising a US$50 million IDA million by the Government of Colombia. iv) promoting more effective, responsive credit and co-financing of US$13.50 At project completion, US$19.7 million and accountable service organizations million from producer organizations and had been provided as grant investment at the local level. The estimated total US$1.04 million from municipalities. support to 136 vertical alliances, directly project cost at appraisal was US$34.88 According to the Results Framework, the benefiting 11,714 producer households. million, of which US$28.40 million were project aims to establish 645 productive planned to be financed by the International alliances that will directly benefit 25,000 A follow-on, Colombia Second Rural Development Association (IDA), US$5.54 poor rural households, most of which are Productive Partnerships Project (PAAP million by producer organizations, and indigenous peoples. Moreover, the project II) started in January 2008 and closed US$0.95 million by municipal governments. is expected to fund municipal subprojects in June 2015. The PDO was to increase In 2009, an Additional Financing (AF) of (e.g. rural road rehabilitation, small bridges) rural competitiveness and build up rural US$36.57 million was approved, comprising to benefit about 10,100 households. entrepreneurship in poor rural communities an additional US$30 million IDA credit, in a sustainable manner through demand- In Guatemala, IBRD co-financed US$5.71 from producers’ organizations, driven partnership schemes with the together with the Inter-American and US$0.86 million from municipal commercial private sector. Project cost Development Bank (IADB) the Rural governments. At completion, the total estimated at appraisal was US$122.40 Economic Development Program (PDER), 62 which was implemented between November Apurimac, Ayacucho, Huancavelica, Junin, alliances/subprojects to support revenue 2007 and December 2014. The PDO Huanuco and Pasco, and to strengthen generating activities in agriculture and was to: (i) improve the competitiveness government capacity to implement an tourism (Type 1) and to a lesser extent of rural productive supply chains with integrated Sierra development strategy. provide critical infrastructure, marketing strong indigenous participation; and, (ii) to ALIADOS consisted of two components: and management in the agriculture strengthen the institutional capacity of the (i) Rural Business, financing subprojects and tourism sectors (Type 2). The total public entities participating in the Program proposed by groups of rural producers to project costs amount to US$17.5 million, for the adoption of a territorial management build strategic productive alliances and comprising US$15 million from IBRD, model with indigenous participation. The increase market access and income; and, US$2 million from the beneficiaries, and actual project costs amounted US$45 (ii) Community Development, financing US$0.5 million from the Government of million, of which US$29 million were funded subprojects proposed by communities Jamaica. As of latest reporting, the project by IBRD and US$16 by IADB. In addition, and groups of families to increase basic has approved 93 subprojects with about the beneficiary producers contributed agricultural and livestock production to 1,500 direct beneficiary households. US$1.04 million in cash. At completion, 174 improve their socioeconomic and food In Honduras, the Rural Competitiveness productive alliance business plans were security. Total project costs amounted to Project (COMRURAL) became effective successfully implemented. In addition, US$34.93 million, with US$20 million from in July 2008 and is expected to close the project had financed 27 business IBRD, US$7.83 million from the Borrower, in December 2016. The PDO is to strengthening plans. In total, 18,115 direct and US$7.1 million from beneficiaries. At contribute to increased productivity beneficiaries were reached by PDER. the end of the five-year project period, and competitiveness among organized 876 productive alliances were financed In Panama, the Rural Productivity rural small-scale producers through their benefiting 17,303 producers. In addition, Project (PRORURAL) was implemented participation in productive alliances. 864 community development subprojects between November 2007 and January The target by the end of the project benefited 35,277 rural households. In 2015. The PDO was to contribute to is to have established 150 productive 2013, an Additional Financing (AF) of increased productivity among organized alliances implemented that reach 6,700 US$38.8 million was approved, consisting rural small-scale producers, through their rural producers. As of latest reporting, of a US$20 million IBRD loan, US$5.6 participation in productive alliances, while 122 productive alliances have been under borrower contribution, and US$7.2 ensuring the sustainable use of natural implementation benefiting 6,441 producers. million through beneficiaries’ counterpart resources and the conservation of globally In 2015 the Government of Honduras funds. The AF introduced a stronger important biodiversity. The actual total requested the closing date extension as focus on rural business investment cost of this project was US$39.25 million, well as an Additional Financing (AF) of subprojects to respond to robust comprising US$38.8 million from IBRD, US$12.6 million to consolidate the existing demand. The AF makes a clear distinction US$8.72 million in kind from producer business plans and to expand the project between rural business investments and organizations, and US$1.09 million from 13 to new geographical areas. Part of the community development subprojects , the Government of Panama. Funding IDA grant for the project was cancelled, and provides further incentives for specifically for productive alliances but in total US$26.1 million are financed the selection of larger business. An amounted to US25.1 million, from which through IDA (including the AF), US$12.1 estimated 694 productive alliances 130 subprojects were financed, reaching million from the commercial financial sector, (i.e. about 16,600 direct beneficiaries) 4,577 direct small producer households. A US$1.2 million from the Government of in addition to the original number of follow-on project (PRORURAL Incluyente) Honduras, US$3.9 million from POs, and productive alliances, are expected. is currently under preparation, with a US$4.0 million from the Swiss Agency for stronger focus on indigenous populations. In Jamaica, the Rural Economic Development and Cooperation (COSUDE). Development Initiative (REDI) started in In Peru, the Sierra Rural Development In Mexico, the Sustainable Production January 2010 and will close in July 2017. Project (ALIADOS) was implemented Systems and Biodiversity Project (SPSB) The PDO is to improve market access for between July 2008 and June 2013. became effective in February 2013 and is micro and small-scale rural agricultural The PDO was to improve the economic expected to close in August 2017. The PDO producers, and tourism operators and conditions and asset base of approximately is to conserve and protect nationally and service providers. The project is expected 53,600 rural families in the selected areas of globally significant biodiversity in Mexico to finance the establishment of productive 63 Linking Farmers to Markets through Productive Alliances through mainstreaming biodiversity-friendly In Brazil, ten projects with a productive 13. The Additional Financing PERU management practices in productive alliance component started operating in ALIADOS ensures that territorial landscapes in priority biological corridors. 2007, but the majority became effective development subprojects are not used to promote market- oriented activities The total project cost amounts to US$30.89 more recently, from 2010 onwards. As and that they maintain their focus on million, including US$11.69 million from mentioned in the main text, the projects in poverty-reduction and natural resources management. the Global Environmental Facility (GEF) Brazil are typically multi-sector operations, and US$19.2 million from the Federal in which productive alliances constitute Government. 185 producers groups only one of several project components. have been participating in biodiversity- The projects in Brazil in chronological friendly production initiatives, directly order of their starting date are: Pará benefiting nearly 16,000 producers. Integrated Rural Development Project (2007-2014), Alto Solimôes Basic Services In Haiti, the Strengthening Agriculture and Sustainable Development Project Public Services II Project (RESEPAG II) (2008-2014), Sergipe Integrated Project became effective in April 2012 and is (2009-2012), São Paulo Sustainable Rural expected to be completed in June 2018. Development and Access to Markets Project The PDO is to: (i) reinforce the capacity (2010-2017), Rio de Janeiro Sustainable of the Ministry of Agriculture, Natural Rural Development Project (2010-2016), Resources and Rural Development to Santa Catarina Rural Competitiveness provide or facilitate access to services in Project (2010-2016), Pernambuco Rural the agricultural sector; (ii) increase market Economic Inclusion Project (2012-2019), access to small producers and food security Ceará Rural Sustainable Development in selected areas, and, (iii) provide financial and Competitiveness Project (2012-2016), assistance in the case of an Agriculture Parana Multi Sector Development Project Sector Emergency. The total project cost (2013-2018 with extension/Additional is US50 million, of which US$40 million are Financing), and Bahia Sustainable Rural from IDA and US$10 million from the Global Development Project (2014-2021). Agriculture and Food Security Program Trust Fund (GAFSP). The establishment of productive alliances is linked to Component 2, which is financed with US$ 36 million. As of June 2016, 51 matching grant subprojects were implemented in the North/North Eastern Departments, directly benefiting over 10,000 households. Additional subprojects were expected to be approved as the screening of further subproject proposals was underway. 64 ANNEX 2 KEY PA PROJECT CHARACTERISTICS Table A2.1 Productive Alliance Projects in Latin America and the Caribbean Total Investment Country / Project name Start Closing (in US$ million) 79.74 Bolivia First Rural Alliances Project (PAR I) May 22, 2006 March 31, 2014 (of which 59.94 IDA) 64.54 Bolivia Second Rural Alliances Project (PAR II) May 9, 2013 November 30, 2017 (of which 50 IDA) Colombia Productive Partnership Support Project 30.44 May 15, 2002 September 30, 2008 (PAAP I) (of which 22 IBRD) Colombia Productive Partnership Support Project 122.40 January 17, 2008 June 30, 2015 (PAAP II) (of which 30 IBRD) Guatemala Rural Economic Development Pro- 45.45 November 26, 2007 December 31, 2014 gram (PDER) (of which 29.14 IBRD) 50.00 Haiti Strengthening Agriculture Public Services II April 12, 2012 June 30, 2018 (of which 40 IDA and 10 (RESEPAG II) GAFSP) Honduras Rural Competitiveness Project 51.20 June 17, 2008 December 31, 2016 (COMRURAL) (of which 30 IBRD) Jamaica Rural Economic Development Initiative 17.5 May 28, 2010 November 30, 2015 (REDI) (of which 15 IBRD) Mexico Sustainable Production Systems and Bio- 30.89 January 31, 2010 July 31, 2017 diversity Project (SPSB) (of which 11.69 GEF) 46.90 Panama Rural Productivity Project (PRORURAL) February 28, 2013 August 31, 2017 (of which 39.40 IBRD) 67.73 Peru Sierra Rural Development Project (ALIADOS) November 1, 2007 January 27, 2015 (of which 40 IBRD) Brazil Alto Solimoes Basic Services and Sustain- 35 February 26, 2008 June 25, 2014 able Development (of which 7 for PA) Brazil Bahia Sustainable Rural Development 260 June 27, 2014 March 31, 2021 Project (of which 153 for PA) Brazil Ceara Rural Sustainable Development and 150 April 5, 2012 April 30, 2018 Competitiveness (of which 70 for PA) 100 Brazil Para Integrated Rural Development December 12, 2006 December 31, 2014 (of which 63 for PA) 713.24 Brazil Parana Multi Sector Development November 6, 2012 November 30, 2017 (of which 59 for PA) 135.25 Brazil Pernambuco Rural Economic Inclusion March 6, 2012 January 31, 2019 (of which 110 for PA) Brazil Rio de Janeiro Sustainable Rural Develop- 79 September 10, 2009 November 30, 2018 ment (of which 66 for PA) 180 Brazil Santa Catarina Rural Competitiveness September 2, 2010 June 30, 2017 (of which 43 for PA) 65 Linking Farmers to Markets through Productive Alliances Total Investment Country / Project name Start Closing (in US$ million) Brazil Sao Paulo Sustainable Rural Development 130 May 25, 2010 September 30, 2017 and Access to Markets (of which 59 for PA) 27.05 Brazil Sergipe Integrated Project: Rural Poverty September 23, 2008 June 20, 2012 (of which 24 for PA) Table A2.2 PA Stand-alone versus PA Component Stand-alone / PA Components Project PDO component (US$ million) »»Institutional support (US$5.14) Test a model to improve accessibility »»Implementation of rural productive alliances Bolivia PAR I to markets for poor rural producers in Stand-alone (US$63.35) selected pilot areas of the country »»Project management (US$9.26) Same as PAR I. The project is a follow »»Institutional strengthening (US$3.6) on operation to the successful Rural »»Implementation of rural productive alliances Bolivia PAR II Alliances Project (PAR I), which tested Stand-alone (US$46.85) the model for improving market »»Project management (US$8.44) access. Generate income, create employ- ment and promote social cohesion of poor rural communities in an »»Preparation of productive alliances (US$3.1) economic and environmentally »»Implementation of productive alliances Colombia PAAP I Stand-alone sustainable manner through the (US$20.6) development and implementation of »»Project Management (US$4.6) a demand-driven productive alliance scheme with the private sector To increase rural competitiveness and build up rural entrepreneurship in »»Partnership Promotion and Preparation poor rural communities in a sustain- (US$5.10) Colombia PAAP II Stand-alone able manner through demand-driven »»Partnership Implementation (US$114.70) partnership schemes with the com- »»Project Management and M&E (US$2.6) mercial private sector To improve the competitiveness of rural productive supply chains with »»Investment in Producer Chains (including strong indigenous participation, and rehabilitation of bridges due to Tropical Storm to strengthen the institutional capac- Agatha of 2010) (US$34.25) Guatemala PDER PA Component ity of the public entities participating »»Territorial Capacity Reinforcement, (US$8.96) in the program through the adoption »»Management, Monitoring and Evaluation of a territorial management model (US$2.24) with indigenous participation Contribute to increased productivity and competitiveness among orga- »»Support for Productive Alliances (US$9.0) Honduras nized rural small scale producers Stand-alone »»Productive Investments (US$39.6) COMRURAL through their participation in produc- »»Project Coordination, M&E (US$2.6) tive alliances 66 Stand-alone / PA Components Project PDO component (US$ million) »»Agriculture & Rural Tourism Subprojects The PDO is to improve market (US$13.72) access for micro and small-scale rural »»National Technical and Capacity Building Jamaica REDI PA Component agricultural producers and tourism subprojects (US$1.25) product and service providers. »»Project Management (US$2) »»Sustainable Production Systems and Biodi- To conserve and protect nationally versity Mainstreaming (US$4.59) and globally significant biodiversity »»Producer Associations and Biodiversi- in Mexico through mainstreaming ty-Friendly Market Initiatives (US$3.63) Mexico SPSB PA Component biodiversity-friendly management »»Institutions, Labels, and South-South Coop- practices in productive landscapes in eration (US$2.30) priority biological corridors. »»Project Management and Monitoring (US$1.17) To contribute to increased productiv- »»Support for Productive Alliances (US$7.1) ity among organized rural small-scale »»Productive Alliances (US$19.8) producers, through their participation »»Environmental Investments and support to Panama PRORU- in productive alliances, while ensur- Stand-alone the NPAS (US$10.0) RAL ing the use of natural resources and »»Project Management, Monitoring and Evalu- the conservation of globally import- ation (US$2.5s) ant biodiversity Assist the Borrower in improving the assets and economic conditions »»Promotion of Rural Businesses (US$34.76) of rural families in selected areas of Peru ALIADOS PA Component »»Territorial Development (US$20.66) the Borrower’s Apurímac, Ayacucho, »»Project Management (US$12.02) Huancavelica, Junín, Huánuco and Pasco regions. Table A2.3 Target Areas and Direct Beneficiaries14 Project Target area Direct beneficiaries (actual) Productive Alliances (actual) 5 Departments: »»Uyuni Salt Lake »»Valleys of Cochabamba »»28,527 producer Bolivia PAR I »»Tropics/Norte Santa Cruz »»768 alliances households »»Lago Titicaca »»La Paz-Beni (El Chaco under certain requirements) 5 selected areas: »»“Central valleys” area »»“Southern valleys” area »»25,800 producer Bolivia PAR II »»645 alliances »»Tropic households »»Chaco area »»“North” areas Progressively, the project became active in »»11,714 producer Colombia PAAP I »»136 alliances 27 of the 32 Departments in Colombia. households »»42,552 producer Colombia PAAP II PAAP II operated in 31 of 32 Departments. »»725 alliances households 67 Linking Farmers to Markets through Productive Alliances Project Target area Direct beneficiaries (actual) Productive Alliances (actual) Initially 8 Departments. Then, project area »»20,001 producer Guatemala PDER covered 20 of the 22 Departments of Gua- »»201 alliances households temala (91% of the national territory) Priority Regions defined in National Haiti RESEPAG II Agricultural Development Plan (South, South-Eastern, North, and North-Eastern) 7 out of 24 Departments: Compayagua, »»7,085 producer Honduras COMRURAL Copán, Intibucá, La Paz, Lempira, Ocote- »»74 alliances households peque and Santa Barbara (Western) Entire country - Currently project is imple- »»1,500 producer Jamaica REDI mented in all Parishes except Kingston & »»93 alliances households St. Andrew. 6 of 31 States: »»3,141 producer Mexico SPSB Yucatan, Quintana Roo, Campeche, Oaxa- »»185 alliances households ca, Tabasco, and Chiapas 3 out of 10 Provinces: Herrera, Los Santos »»4,577 producer Panama PRORURAL »»130 alliances and Veraguas (26 districts) households 6 of 25 Departments, or 255 of 483 Dis- tricts. »»17,303 producer Peru ALIADOS I »»876 alliances 1. Apurimac, 2. Ayacucho, 3. Huancavelica, households 4. Junin, 5.Huanuco, and 6.Pasco »»16,610 producer Peru AF Same as ALIADOS I »»694 alliances households 14. The figure on beneficiaries and alliances in this table are based on the total number of alliances that were still operating at project completion/ latest reporting. 68 ANNEX 3 TARGETING STRATEGIES Table A3.1 Producer Eligibility Criteria for Subproject Proposals Formally Grouped with Occupation/Minimum Maximum Assets (income any Type of Legal Recog- Experience in the Proposed Residence / Age / Minimum Project and land) per Family nition Activity Level of Education Small producers living in centers with less than 2000 At least two years’ expe- inhabitants Being members of POs rience in the proposed At least, one member May not exceed the small formally grouped with any activity for the alliance or Bolivia PAR I of the family must have land ownership, within the type of legal recognition, alternatively, commitment & II attended the third grade limits established by law or with intent to formalize to participate in re-training of school or be able to an economic organization programs needed to per- write and read. May not form the new activity be civil servants or retired annuitants Heads of household with or without land Assets Farm’s size not exceeding two Family Agricultural Units (PAAP I) or Max land tenure & income depend- At least one family mem- ing on the geographical At least 3 years of agricul- ber with complete primary Colombia area & cropping patterns tural experience relevant schooling or participating PAAP (PAAP II) to one of the alliance in adult education pro- Not more than 200 min. activities grams and between 18-50 wages years old. 75% of the income derived from agricultural activities and wages do not exceed 4 (2 in case of PAAP II) minimum wages per month Guatemala Income below the national PDER GDP per capita Main occupation is ag- Honduran citizenship Honduras Participate in an organiza- riculture or related rural Residence in rural areas COMRURAL tion, association or firm employment Older than 18 years Formed a legally recog- Jamaica REDI nized organization Legal personality Predominantly agricultural, Be integrated into Mexico SPSB livestock, harvesting or producer associations or mining in rural areas networks. 69 Linking Farmers to Markets through Productive Alliances Formally Grouped with Occupation/Minimum Maximum Assets (income any Type of Legal Recog- Experience in the Proposed Residence / Age / Minimum Project and land) per Family nition Activity Level of Education Access to land suitable Be a member of an eligi- for agriculture (owned, ble PO. If not a member, Have as a main occupation leased, or under certified he/she may benefit from Panama the activity of the alliance possession) the project’s training PRORURAL requesting funding, Maximum of 10 Ha for component and then be- agricultural crops and 25 come a member eligible Ha for livestock activities organization Small rural producers in districts that suffered high Peru ALIA- Entrepreneurial, have a levels of violence during DOS market oriented mindset the civil conflicts of the 1980s and 1990s Table A3.2 Producer Organization Requirements Project Requirement »»Legally established with a minimum size of 20 producer households (agricultural activities) and 10 (non-agricultural activities) Bolivia PAR I & II »»Not be in a position of imminent bankruptcy or internal conflict »»Clear, explicit and agreed management rules and provisions for distributing costs and revenues Colombia PAAP I »»Legally established with a minimum of 20 members (until 2011) and of 30 members (from 2012 to 2014), & II and of 15 members for the new Departments incorporated in 2013. »»Legally established and active for over one year with at least 20 participating members Guatemala PDER »»An asset base not exceeding US$1.5 million Honduras COMRU- »»Second-tier cooperatives with audited financial statements with at least 12 participating members RAL »»At least one year of experience as producer organization »»Be legally recognized (i.e. a Cooperative, Friendly or Benevolent society) with at least 10 participating members Jamaica REDI »»Have an asset base (excluding land and buildings) not exceeding US$10,000 (for micro) and US$100,000 (for small-scale) and an annual turnover of less than US$125,000. Mexico SPSB »»Legal status and demonstrated administrative capacity. »»Legally-constituted entity/civil associations with at least 15 participating members Panama PRORURAL »»Open membership for producers who fulfill the PO’s requirements »»Formally and legally recognized as a group with at least 10 members Peru ALIADOS »»Eligibility for participation is not contingent on degree of market access. 70 ANNEX 4 SUBPROJECT SELECTION PROCESS Table A4.1 Institutional arrangements subproject selection Project Roles of local / regional institutions Role of central institutions »»The Ministry of Rural Development and Land, through a National Coordination Program (EMPODERAR), launches CfPs through oral and written press in each re- »»Regional Project Units (RPUs) manage the evaluation gion, conducts the final approval of business evaluation of the initial subproject proposals (oportunidad), and by RPUs, and supervises their implementation then conduct the technical evaluation of Business Plans »»A specialized service provider (Intermediate Financial Bolivia PAR I »»Brokers (facilitadores) hired by the Project support Institution, IFI) is hired to conduct the ex-ante financial & II15 POs in business plan’s preparation evaluation of business plan and validate the technical »»Service Providers (acompañantes) assist alliances soundness (PAR II) business plan implementation and through capacity »»The National Fund for Productive and Social Invest- building in business management. ments (Fondo Nacional de Inversión Productiva y Social y Productiva, FPS) implements complimentary Municipal subprojects (productive infrastructure) »»Regional Intersector Committees (RICs) launch CfPs in each region, approve the initial subproject proposals selected by SA and the business plans »»The Ministry of Agriculture and Rural Development »»The Departmental Secretariats of Agriculture (SA) (MADR) acts as executing agency through a PCU in inform about the CfP and rank the initial proposals charge of final decision on subproject selection, and according to technical criteria. They would also partici- a National Technical Committee in charge of general pate in the RIC and provide co-financing and follow-up oversight and policy guidance. to alliances. »»The National Inter-sectorial Committee supervises »»Services Providers (NGOs or consulting firms (called overall project implementation (including policy priorities Organización Gestora Acompañante – OGA) (i) sup- Colombia and monitoring). It also has the right to refuse alliance port the preparation of alliance proposals and facilitate PAAP I & II approvals of the RICs on the basis of technical criteria. their implementation,(ii) transfer management skills »»A Trust Company (i) transfers the funds from MADR to the PO including through “on the job training” of to the trust accounts of the alliances, ii) supervises the Manager (“Gerente aprendiz”) financed by the projects trust accounts management, iii) oversees procurement over two years and working under the guidance of the of POs in line with the procedures for community-based Partnership Management Committee. procurement including the expenditures of all funds and »»RPUs (called Regional Management Organizations, contributions of the alliances that are kept in the trust OGR) manage the alliances subproject cycle including accounts a) evaluation of initial proposals, b) supervision of sub- project preparation and implementation, e) assessment of OGAs’ performance. 71 Linking Farmers to Markets through Productive Alliances Project Roles of local / regional institutions Role of central institutions »»The Presidential Secretariat for Planning and Pro- »»Guidance Groups established at departmental level gramming (SEGEPLAN) acts as executing agency for coordination, learning and social monitoring in through a PCU. order to ensure the involvement and participation of »»The Program Board of Directors oversees and pro- indigenous people program implementation. vides strategic direction. Guatemala »»Entrepreneurial services providers (Socio de De- »»The Ministry of Economy (MINECO), through a PCU, PDER sarrollo Empresarial – SDE –) organize the technical launches CfPs, reviews initial subproject proposals and and entrepreneurial assistance including the timely business plans, and supervises implementation of pro- disbursement of seed capital to POs and the hiring ductive alliances of specialized technical services providers, if needed. »»An Investment Coordination Committee approves SDEs can be NGOs or private consultant companies the business plans and supervises the implementation of investments. »»The Project Orientation Council overlooks the sub- »»Ministry of Agriculture and Livestock (SAG) hosts the project selection process. PCU which conducts the final subproject selection. SAG »»Private service providers implement technical/legal/ also maintains a regional presence in the targeted seven commercial assistance in the preparation and imple- departments to closely supervise project activities. Honduras mentation of the business plans (including subprojects) »»The Project Administration Unit, housed in the Min- COMRURAL of the productive alliances istry of Finance is responsible for the overall financial »»Private Financial Institutions (PFIs) - such as com- administration of COMRURAL, specifically flow of funds mercial banks, credit and savings cooperatives, and and overall procurement private institutions of financial development co-finance »»A central Evaluation Committee conducts the subproj- the business plans of productive alliances. ect selection »»Inter-Ministerial Project Steering Committee ensures »»Private Service Providers (i.e. Jamaica Business De- that the project is in line with national development velopment Center, consultants) assist applicants in the priorities preparation of business plans »»The Jamaica Social Investment Fund (JSIF) hosts the »»The Rural Agricultural Development Agency‘s PCU in charge of ranking eligible proposals, assessing (RADA) local offices - in collaboration with the Project business plans, and monitoring their implementation Jamaica Enterprise Development Officers, provide field exten- »»Rural Economic Evaluation Committees evaluate & REDI sion services for the agricultural subprojects recommend the viable subprojects »»The Tourism Product Development Company (TPD- »»Management Review Committee: Responsible for Co), a specialized public entity, conducts initial product clearance the subprojects that have been approved by development assessment locally for proposed rural REEC. tourism subproject »»JSIF Board of Directors (Projects Committee): Re- sponsible for final clearance of subproject proposals approved by MRC. »»The Ministry of Agricultural Development hosts the »»POs prepare initial subproject proposals PCU, responsible for evaluating the business plans for Panama »»Consultants prepare alliances subproject business the proposed productive alliance and recommending PRORURAL plans viable proposals for approval »»The Project Steering Committee provides final ap- proval »»Regional Project Units (RPUs) select initial subproject »»The PCU hosted by the |Ministry of Agriculture assesses proposals and evaluate business plans (through exter- the business feasibility nal consultants) »»Project Steering Committee - Responsible for provid- Peru »»Service Providers / Agents (mainly individual consul- ing policy oversight and guidance for project implemen- ALIADOS tants) assist POs with business plan tation »»Regional Resource Allocation Committees are re- »»Assessment of Business Plans is done by (1) external sponsible for the final approval of Business Plans evaluators (economic and social evaluation) 72 Specific project examples of the and business plan preparation, while a (MINECO) to launch the calls for proposals, institutional arrangements of the subproject radical change occurred under the second review the submitted subproject proposals selection process are described below: phase, in which the producer organizations and business plans, and supervise the and their buyers were made responsible subprojects. In coordination with the PCU, In Bolivia’s PAR projects, the Regional for these tasks, usually with the support an Investment Coordination Committee, Project Units (RPU) are the first layer of a specialized service provider (e.g. comprised of technical representatives from responsible for verifying the eligibility NGO, agricultural consulting firm, etc.). the various co-executing agencies of the of the beneficiaries, evaluating initial project, approved the business plans and subproject proposals, and conducting In Peru’s ALIADOS project, the supervised their implementation. At the the technical evaluation of the business preparation of the initial subproject departmental level, guidance groups for plans. Business plans are prepared by proposals is the responsibility of the coordination, learning, and social monitoring local specialists hired by the project, to producer organization. The evaluation were established to ensure the involvement support the producer organizations and and selection of these proposals is done and participation of indigenous peoples in their buyer(s) in this matter. A specialized by the project’s regional offices, who the implementation of the subprojects. consultant firm validates the technical approve eligible proposals for the business feasibility of the proposal and conducts plan preparation phase. For this, the an ex-ante financial evaluation of the project provides support to the producer 15. Bolivia PAR II: Same as PAR I, but the role of the service providers business plan. The Regional Project Units organizations and their buyer(s) by hiring (faciliatadores & acompañantes) then pass their recommendations on each a consultant who assists them in writing was strengthened to support the productive alliance preparation and business plan to the National Coordination the business plan (80% of the consultant implementation, as well as of the IFI Program (EMPODERAR) of the Ministry cost is funded by the project and 20% to validate the technical choice of the proposal, a part of certifying the ex-ante of Rural Development and Land, which by the beneficiaries). The subsequent financial evaluation. conducts the final approval of the business evaluation of the prepared business plan and supervises its implementation. plans is done by: (i) external evaluators (responsible for the economic and social In Colombia’s PAAP projects, Regional evaluation); (ii) the PCU (who assesses Management Organizations (OGRs) the business feasibility); (iii) the project’s are responsible for launching annual regional offices; and, (iv) a Regional/Local regional competitive calls for the initial Resource Allocation Committee (CLAR) subproject proposals in line with sector composed of representatives of the public policy priorities determined by a National and private sectors as well as civil society Inter-sector Committee established entities. The CLAR are the final decision- by the project. After the receipt of the makers on subproject approval. During the proposals, the Departmental Secretariats Additional Financing, ALIADOS allowed of Agriculture rank these proposals and for a more active involvement of regional pass them on to the OGRs which further governments in the subproject selection evaluate the technical, financial, economic, process through their participation in environmental and social soundness of the decision-making process of Rural the proposals. After this evaluation, they Business and Territorial Development submit those that comply with the criteria Plans. The AF simplified organizational to Regional Inter-sector Committees, which arrangements and project procedures by are composed of local representatives eliminating the requirement that every of the private and public sector. These subproject had to be approved by the committees then review the evaluation by National Public Investment System (SNIP). the OGRs and either approve the proposals for further elaboration into business plans In contrast to the examples above, or reject them with a justification. In the Guatemala’s PDER project relied on a first phase of the project (PAAP I), the centralized Project Coordination Unit OGRs were responsible for the proposal (PCU) within the Ministry of Economics 73 Linking Farmers to Markets through Productive Alliances ANNEX 5 SUBPROJECT FINANCING Table A5.1 Financial Support to Producer Organizations Project Support and Co-Financing Arrangements Initially three models for subproject financing were envisaged, as described under PAR II below. However, PAR I Bolivia PAR I only implemented Model 1. Model 1: Types A and B A) Alliances with integral support »»Max. project grant: US$1,800 per household and USD 50,000 per PO »»Project co-finances up to 70% of total investment costs, PO at least 30% B) Complementary municipal infrastructure: »»Max. total investment US$ 350,000 Bolivia PAR II »»Project co-finances up to 80% of total costs, Local government at least 20% Model 2: Technical Assistance / Training support (new alliances for transformation and 2nd generation alliances to deepen innovation) »»Max. project grant: US$200 per household »»Project co-finances up to 70% and PO at least 30% Model 3: TA to access to credit (new and 2nd generation alliances) »»Conditions same as in Model 2 »»Max. project grant: US$2,600 per household (could increase to US$7,600 if the project finances the purchase Colombia of land for landless farmers) PAAP I »»Project co-finances up to 40% of the total investment costs »»Max. project grant: US$1,818 per household (could increase to US$2,726 if the partnership obtains additional Colombia US$909 per family as commercial credit) PAAP II »»Project co-finances up to 30% of total investment costs, which can be leveraged by other government grants or buyer contributions »»Max. project grant: US$2,000 per household and US$30,000 per PO (of which up to US$10,000 could be used for seed capital for new technology, working capital, and other start-up expenditures, while the remainder Guatemala could fund productive infrastructure and Business Development Services16) PDER »»Project co-finances up to 100% of small infrastructure costs, 80% of seed capital (20% funded by the POs’ counterpart in cash), and 90% of the costs for Technical Assistance (Business Development Services) (10% fund- ed by POs’ own contribution in kind). »»Max. project grant: US$3,600 per household Honduras »»Project co-finances up to 60% of total investment costs, Private Financial Partner (PFP) at least 30% of total COMRURAL investment costs, and PO at least 10% »»Max project grant ranges from US$50,000 to US$200,000 per subproject Jamaica REDI »»Subprojects receiving more than US$70,00 as grants are required to provide benefits to the wider community and not solely to direct beneficiaries »»Max project grant: US$300,000 per subproject (including capacity building, technical assistance and training), depending on the number of beneficiary producers comprising the PO Mexico SPSB »»The budget for Technical Assistance for subprojects is maximum 10% of the total investment costs and/or not more than US$40,000 »»Max projects grant: US$250,000 per subproject (initially US$500,000) Panama »»Max ceiling of US$2,500 per household (US$5,000 in exceptional cases) PRORURAL »»Project co-finances up to 80% of the total investment costs (and working capital) and PO at least 20% percent (in cash or in kind) 74 Project Support and Co-Financing Arrangements »»Max project grant ranges from US$10,000 (micro-enterprises), 20,000 (small enterprises) to US$30,000 (medium enterprises). Peru ALIADOS17 »»Project co-finances up to 70% (max US$21,000 for medium enterprises; US$14,000 for small enterprises; US$7,000 for microenterprises) »»PO co-finances at least 30% in cash (15% in kind and 15% in cash for microenterprises) Table A5.2 Financed Support and Cost Categories Project Cost categories financed by PA projects Bolivia PAR I »»(i) Goods, (ii) Works or infrastructure, (iii) Technical assistance, and (iv) Livestock purchases »»Model 1: (i) Infrastructure, (ii) incremental inputs to produce, (iii) equipment (minimum to produce ), (iv) techni- cal assistance and training/capacity building »»Model 2: (i) technical assistance, (ii) Goods and equipment (bienes demostrativos) Bolivia PAR II »»Model 3: (i) technical assistance (facilitator to prepare studies required by the Financial Entities), (ii) technical assistance for management purposes of POs, (iii) technical support (acompañamiento) of investment execution and credit scheme` »»Grant (Incentivo Modular): Machinery and equipment, vegetative materials, fertilizers, On-farm infrastructure, Colombia PAAP labor costs, studies, surveys, etc. I & II »»Technical Assistance, follow-up and training for vertical alliances Guatemala »»(i) Seed Capital, (ii) Productive infrastructure, (iii) Business Development Partner PDER Honduras »»Fixed capital (e.g. plant and equipment, minor infrastructure), (ii) working capital and, (iii) Technical Assistance COMRURAL »»Grants: Productive infrastructure, goods, equipment and operational costs, Jamaica REDI »»Specific TA and training to help the enterprises and partner organizations to implement their business or imple- mentation plans »»Goods (e.g. acquisition of storage or processing equipment), minor works, operations costs (day labor for re- Mexico SPSB habilitation of forest) and consultants and non-consulting services for the development of farm-based biodiver- sity management and technical assistance Panama »»Fixed capital (e.g., plant and equipment, minor infrastructure), working capital and technical assistance for the PRORURAL first two years. »»(i) Services (technical assistance, marketing studies, surveys, laboratory test, publicity, organic certification), (ii) Peru ALIADOS Goods (equipment machinery, vehicles and other inputs), (iii) Small infrastructure (storage facilities, processing plants, water supply, irrigation, treatment and disposal of solid waste and effluent) 16. Business Development Service providers were responsible for holistic technical assistance, which included pre-investment costs of business plan preparation as well as management, negotiation and quality assurance costs of partnerships (capacity building and knowledge transfer on productive system, good agriculture practices and good manufacturing practices, administrative and entrepreneurial management, legal services, marketing services, certification services, supervision activities). The maximum acceptable cost of Business Development Services (including the cost of preparing the business plan and management expenses, negotiation and quality assurance) was US$40,000, of which PDER financed up to 90% and PO at least 10%. 17. In the initial implementation phase of ALIADOS, financing of goods was limited to US$4,200 per subproject, irrespective of its size. As a consequence, producers requested the smallest subproject that would allow maximizing this financing, which resulted in a portfolio of smaller subprojects than anticipated at appraisal. As a result, the Additional Financing incentivizes larger business by increasing the value of the capital goods eligible for financing to 70%. The AF introduces additional support for about 15% of the most successful Business Plans financed under the original project, for both TA and co-financing (up to the financing ceiling). 75 Linking Farmers to Markets through Productive Alliances ANNEX 6 PA PROJECT MONITORING AND EVALUATION SYSTEMS Table A6.1 Subproject Monitoring Project Monitoring Activities and Data Collection During implementation: Technical Assistance (TA) providers of the alliance collect and systematize technical, economic, social, environmental and administrative data for each alliance, which is then reported and recorded in the project’s M&E System.  At subproject completion: The TA prepares an ex-post financial assessment, which is based on the Bolivia PAR I & II structure of the business plan and includes revenue and cost “with and without PAR”. For PAR I, con- solidated business information was presented to POs in participatory workshops for their validation. A final report or closing memory is recorded. Project M&E: Online Geo-referenced Management Information System During implementation: The planned M&E System was never established. As a result, only partial monitoring of subprojects was carried out by the PCU. At subproject completion: Due to the lack of sufficient or representative data and quantitative infor- Brazil Alto Solimôes mation, an ex-post assessment during the ICR mission collected its own data during field visits directly from beneficiary communities and relied on partial studies and progress reports for some subproject activities. Project M&E: Excel sheets and Word documents (State-level PCU) During implementation: The planned M&E System was integrated in the central State-level Secretari- at of Agriculture’s online Monitoring Information System (MIS). As a result, subproject-level information is captured by the PCU with data received from the regional offices. The Mid-Term Review revealed, however, that some information of the subproject selection process is not recorded in the MIS, while Brazil Pernambuco other data records are very detailed and require time-consuming data entry. At subproject completion: The PCU is responsible for an ex-post financial assessment of subproject performance. At Mid-Term Review, it was agreed that a complementary PO-level evaluation would be implemented. Project M&E: Online Monitoring Information System During implementation: The planned M&E System was not established as planned at project apprais- al. At subproject completion: At completion, only a disbursement monitoring system was in place. A Brazil Pará results monitoring system was not established until the final year of the project (2014) and much sim- plified compared to the original design to capture information on subproject results. Project M&E: Excel sheets and Word documents (Non-aggregated data was captured only partially at the State Secretariat of Environment and the Pará Land Institute) During implementation: The Monitoring Information System established in a former project was upgraded and linked to the financial management system (TOTVS/Microsiga PROTEUS) during project implementation. The MIS was fully self-financed by the State (not by the Project). However, despite it being a well-managed, modern instrument, it remained a database on the subproject cycle, physical performance and fiduciary aspects. Its envisaged capacity to measure changes in and linkages be- Brazil Sergipe tween project investments and poverty reduction and changes in behavior never materialized. At subproject completion: At completion, the PCU was responsible for an ex-post financial assess- ment of subprojects and final evaluation. However, there were insufficient data to evaluate perfor- mance as well as some indicators from the Results Framework. Project M&E: Monitoring Information System (called SIG) 76 Project Monitoring Activities and Data Collection During implementation: TA providers (OGA and OGR) are in charge of collecting physical/financial data, sales, performance rating, and contingencies for each alliance. Data is not collected on income and production costs. Colombia PAAP I & II At subproject completion: The last subproject M&E report consolidates business performance and main indicators, but does not include ex-post investment returns. Project M&E: Online Project M&E System During implementation: The PCU collect data on physical/financial implementation, and sales. Data is not collected on financial income and costs by subproject. At subproject completion: TA providers (SDE) prepare a report once the subproject execution is final- Guatemala PDER ized and the PCU prepares a comparative analysis of the PO before and after the subproject invest- ment Project M&E: Excel sheets and Word documents During implementation: Project TA providers are in charge of collecting physical/financial data, key indicators, and loan repayments. Data not collected: financial income and costs by business plan. Honduras COMRURAL Project M&E: Online M&E system, which is participatory through regional councils of value chains. Before the system, data from 2010-2013 was captured with Excel sheets and Word documents. During implementation: POs are not directly responsible for data collection. The PCU conducts data collection through the TA provides (Enterprise Development Officers). The monitoring database includes the basic physical and financial records, the details of inputs and services provided to the beneficiaries (e.g., funding and training) and data obtained from surveys and other recording mecha- Jamaica REDI nisms designed specifically to collect information from the subprojects. At subproject completion: A final evaluation of the subproject design and implementation to docu- ment lessons learned is planned. Project M&E: Excel sheets and Word documents. The project also uses the Electronic Fund Manager platform. During implementation: The PCU has the overall responsibility for the M&E of project activities and managed data inputs. At subproject completion: Rural Invest software was used for ex-post financial evaluations of a sample Panama PRORURAL of subprojects. Field visits were done to alliances to collect data on cost and incomes. Project M&E: Excel sheets and Word documents. The project also uses Pentagon, a computerized financial management system. During implementation: The RPUs are responsible for the monitoring of the subprojects and periodi- cally collecting data on general demographic, socioeconomic, and physical/financial data on subproj- ect investments. Peru ALIADOS At subproject completion: The M&E System has been strengthened during the Additional Financing, which requires all (new) business plans to include measurable key indicators. Project M&E: Computerized project management information system During implementation: Producer-level baseline data was collected prior to the implementation of Mexico SPSB subproject activity for the four main products supported, as a basis for a final evaluation on environ- mental and socio-economic impacts. At subproject completion: An end-line survey and resulting impact evaluation analysis is planned at project completion. Project M&E: Online Integrated System for Information Management and Project Evaluation (SIGIEP) 77 Linking Farmers to Markets through Productive Alliances Figure A6.1 M&E System Example 1: Colombia PAAP Link: http://alianzasproductivas.minagricultura.gov.co/inicio/index.aspx Figure A6.2 M&E System Example 2: Mexico SPSB Link: www.pspsb.info 78 ANNEX 7 RESULTS FRAMEWORKS Table A7.1 Project Development Objectives and Indicators Target PDO Indicator (latest / ICR data for Project PDO (latest version) Baseline completed projects) Actual Indicator 1: Alliance model imple- No Yes Yes mented in project areas 41,711 Indicator 2: Growth (%) in income 60,481 Bs Bolivia- of rural productive units (gross after 3 implementa- +39% nos (Bs) Test a model to improve income per household per year) tion years (+45%) Bolivia PAR I accessibility to markets Indicator 3: Number of financed for poor rural producers in 0 540 720 alliances with IRR>12.5% selected pilot areas of the country. Indicator 4: Number of new 308,506 (SIGG) or wage-earning jobs generated 0 809,590 112,946 per year (person-day) (economic analysis) Indicator 5: % increase in the vol- 45% higher after 3 40,694 60% higher than ume marketed per rural productive alliances implemen- Bs baseline unit tation years Indicator 1: Increase in the 0 average volume of sales of the 50% 0.00 Same as PAR I. The proj- product(s) involved in the alliances ect is a follow on opera- Indicator 2: Producer organiza- Bolivia PAR II tion to the successful Rural tions that register income and 0 80 0.00 Alliances Project (PAR I), costs, and are accountable to their which tested the model members for improving market Indicator 3: Producer organiza- access. tions that maintain or improve their 0 70 0.00 commercial relations (alliances) for at least two productive cycles To generate income, At the end of a create employment and partnership cycle promote social cohesion Economic and finan- at least 70% of the Colombia PAAP I of poor rural communi- cial analysis of the producers partici- ties in an economic and Indicator 1: Producers participat- sample of 23 part- pating in produc- environmentally sustain- ing in productive alliances increase nerships (out of 136) None tive alliance have able manner through their income by 20% compared to indicates that average increased their the development and the baseline incomes increased by income by 10% implementation of a 77% with important compared to the demand-driven productive variations baseline of each alliance scheme with the partnership private sector 79 Linking Farmers to Markets through Productive Alliances Target PDO Indicator (latest / ICR data for Project PDO (latest version) Baseline completed projects) Actual Economic and finan- cial analysis of the sample of 23 part- nerships (out of 136) indicates that em- At the end of the ployment increased Indicator 2: An increase of 50% in project, employ- by about 0.8 per- employment in the participating ment in at least None son-years per family, a To generate income, production units of which 40% is 70% of the part- 70% increase. 15% of create employment and employment for women nerships will have the partnerships are promote social cohesion increased by 10% with women house- Colombia PAAP I of poor rural communi- hold heads. However, ties in an economic and no data were collect- environmentally sustain- ed on overall women able manner through employment the development and implementation of a demand-driven productive Social impact analysis alliance scheme with the shows that 35% of private sector Indicator 3: At the end of the partnership index. project at least 70% of the produc- 70% of associations Partnerships will need er associations will have attained reach adequate more time and effort None an adequate social partnership social partnerships to be socially sustain- index measuring social cohesion index able. Nevertheless, and partnership principles beneficiary survey indicates important social achievements Indicator 1: Total sales volume of the producer organizations reaches 571.9 0 350 a cumulative 350 billion Colombi- an pesos in year 5 To increase rural com- Indicator 2: 75% of the partici- 75% Colombia PAAP II petitiveness and build up pating producer organizations will 0 83% rural entrepreneurship in have a manager poor rural communities Indicator 3: 75% of participating 80% in a sustainable manner producer organizations will main- 0 75% through demand-driven tain a system of accounts partnership schemes with Indicator 4: 75% of alliances have the commercial private cumulatively recuperated at least sector 70% of the competitive grant they 75% 0 50% were expected to return to the revolving fund as programmed in their annual operating plan 80 Target PDO Indicator (latest / ICR data for Project PDO (latest version) Baseline completed projects) Actual Indicator 5: 80% of the OGR have 0 80 88% their contracts yearly renewed Indicator 6: Percentage of POs still formally selling to a buyer N/A Colombia PAAP II 75% 80% 24 months after end of project support Indicator 7: Local governments provide 22 billion Colombian pe- 0 22 104.5 sos of funding Indicator 8: 75% of participat- ing POs remained linked to their 0 75% 72 partners 24 months after project support termination Indicator 1: Increase of total sales To improve the competi- of the rural productive supply 0 tiveness of rural produc- chain partnerships with US$50 US$35 million US$16.31 million tive supply chains with million during the project period Guatemala PDER strong indigenous partici- pation, and to strengthen the institutional capacity of the public entities par- Indicator 2: At least 25% of ticipating in the program municipal projects stem from the through the adoption of 0 25% 35% participatory planning process a territorial management model with indigenous participation Indicator 1: 10% increase in the value of gross sales of the rural producer organizations based on 0 10% 61% implementation of the business plan Honduras COMRURAL Indicator 2: At least 80% of the To contribute to increased producers with approved business 0 80% 72% productivity and competi- plans are satisfied with the project tiveness among organized Indicator 3: Private actors invest at rural small scale producers least 12.05 million USD as loans to 0 12.05 12.50 through their participation producer organizations in productive alliances Indicator 4: Rural producers par- ticipating in the project experience 20% increase in land and labor 0 20% 36 productivity 81 Linking Farmers to Markets through Productive Alliances Target PDO Indicator (latest / ICR data for Project PDO (latest version) Baseline completed projects) Actual Indicator 1: Number of participat- 27 ing rural enterprises that realize an N/A 44 increase in turnover (sales) Indicator 2: Number of participat- ing rural enterprises functioning 56 0 27 Jamaica REDI To improve market access as registered business entities one for micro and small-scale year after final disbursement rural agricultural produc- Indicator 3: Number of direct ers and tourism product jobs created in rural communities 0 150 291 and service providers. receiving project assistance Indicator 4: Percent of critical infrastructure subprojects that 0 78 26 achieve their expected results for improvement in the value chain To conserve and protect Indicator 1: Areas brought under nationally and globally enhanced biodiversity protection 0 34,500 8,754 significant biodiversity in (ha) Mexico SPSB Mexico through main- Indicator 2: Producers applying streaming biodiversi- biodiversity- friendly production 0 6,900 3,141 ty-friendly management practices practices in productive Indicator 3: Share of sales of landscapes in priority goods and services produced un- 0 12 1.48 biological corridors. der biodiversity- friendly practices Indicator 1: By EOP, at least a 25% increase in sales receipts of small- 0 25% 22.3% increase scale producers via PRORURAL-fi- To contribute to increased nanced productive alliances productivity among or- Indicator 2: At least 40% increase Panama PRORURAL ganized rural small-scale in net revenues for the partici- 0 40% 80% producers, through their pating RPAs via the productive participation in productive alliances (EOP) alliances, while ensuring Indicator 3: At least a 20% 1,806 the use of natural resourc- increase in membership in RPAs active 20% 54% es and the conservation for the project area, relative to mem- of globally important baseline (EOP) bers biodiversity Indicator 4: 10% in area under annual crops and cattle in project 3,781.4 10% reduction in 8% of project area or area on land appropriate for forest- ha area (min.378 ha) 302.5 ha ry uses (EOP) Indicator 1: Increase in net value To assist the Borrower in of sales of families participating in 42 20% - improving the assets and Rural Business subprojects Peru ALIADOS economic conditions of Indicator 2: By EOP, value of prin- rural families in selected cipal productive assets of 75% of 44 30% - areas of the Borrower’s beneficiaries increased by 30% Apurímac, Ayacucho, Huancavelica, Junín, Indicator 3: By the EOP, at least Huánuco and Pasco 80% of subprojects have an index 0 80% 80% regions value of objectives achievement over 70% 82 ANNEX 8 PRODUCTIVE ALLIANCE PROJECT OUTCOMES AND IMPACT EVALUATIONS Details on selected PA projects regarding scope and efficiency, as well as the methodologies used for the Bolivia PAR I and Colombia PAAP II impact evaluations are presented below Table A8.3. Table A8.1 Scope of Projects Direct beneficiary Direct beneficiary house- Alliances/ Subprojects Alliances/ Subprojects households holds Project (appraisal target) (completion)18 (appraisal target) (completion) Bolivia PAR I 675 768 33,700 28,527 Brazil Alto Solimôes 50 26 3,500 3,252 200 36,000 19 Brazil Pará 41 3,148 (restructured to 49) (restruct. to 4,067) Brazil Sergipe N/A20 247 N/A 10,800 Colombia PAAP I 100 136 10,000 11,714 Colombia PAAP II 300 725 25,500 42,552 Guatemala PDER 200 174 30,000 18,115 Panama PRORURAL 60 21 130 5,000 4,577 Table A8.2 Investment Support Average number of PO Actual investment Actual investment perActual investment per Alli- members/ households per financed by IDA/IBRD* beneficiary household ance/ Project subproject (US$ millions) (US$) Subproject (US$) Bolivia PAR I 37 60.25 2,112 78,449 Brazil Alto Solimôes 12522 2.44** 765 93,976 Brazil Pará 73 9.7** 3,080*** 189,450*** Brazil Sergipe 44 10** 920 40,486 Colombia PAAP I 86 20.60 1,759 151,507 Colombia PAAP II 59 23.84 514 30,371 Guatemala PDER 104 16.85 930 96,822 Panama PRORURAL 35 19.16 4,186 147,379 * This figure refers to IDA/IBRD financing related to productive alliance investments (as in case of Brazil, complementary investments in other sectors were done). It excludes counterpart funds. It includes the matching grant as well as the costs for technical assistance, but excludes costs for business plan preparation (except for Guatemala). ** Due to lack of data, this figure represents the total direct investment in subprojects and/or micro-projects (not only IBRD). For Brazil Pará, the figure includes US$1.44 m spent on matching grants for subprojects. *** The investment per beneficiary figures are based on provided data on household figures and assume four members per house- hold due to lack of detailed data provision. Similarly, the investments per subproject are based on the total amount of financing and total number of alliances stated, being a simple average not reflecting the heterogeneity across subprojects. 83 Linking Farmers to Markets through Productive Alliances Table A8.3 Key Outcomes and Impacts of Completed PA Projects Out- come/ Control Impact Project Key finding Data source group Additional information 60% higher sales volume market- 133% of end-of-project target (which M&E System (PDO ed per rural productive unit com-   had been revised from 54% to 45% indicator) pared to start of project baseline increase) Bolivia 2014 Impact Evaluation PAR I 29% to 39% higher sales volume based on M&E system for project beneficiary producer Yes   data (n=6000 house- organization than control group holds) Sales were analyzed as part of the Brazil No specific results on sales pre- Economic and Financial Analysis, but Alto     sented average aggregate data not present- Solimôes ed in ICR. Sales were analyzed as part of the Economic and Financial Analysis, but average aggregate data not present- Brazil No specific results on sales pre-     ed in ICR. The ICR states that 55% of Pará sented all sales remain individual despite the project’s vision of a more organized marketing framework. Sales were analyzed as part of the Brazil No specific results on sales pre- Economic and Financial Analysis, but     Sergipe sented average aggregate data not present- ed in ICR. Sales were analyzed as part of the Sales Colombia No specific results on sales pre- Economic and Financial Analysis, but     PAAP I sented average aggregate data not present- ed in ICR. Colombia 572 billion Colombian pesos in M&E System (PDO   163% of end-of-project target PAAP II sales value generated indicator) 47% of end-of-project target (revised US$16.31 million increase of total from US$50m to US$35m). Equiva- M&E System (PDO sales of rural productive supply   lent to a 64% increase in sales value indicator) chain alliances from baseline reported in subproject business plans. Guatema- 20% increase in sales for allianc- la PDER es that had received collective Collective productive investments productive investments compared Final External Evaluation No refer to storage rooms, processing to alliances that had received only (n= 16 alliances) facilities, and the like. technical assistance and other investments 133% of end-of-project target. M&E data shows that 58.4% of beneficiary Panama 22.3% increase in sales receipts M&E System (PDO producers continued to sell through PRORU-   by beneficiary producers indicator) commercial intermediaries, which was RAL not reported. Hence, results on sales are under-estimated. Panama 80% increase in net revenues of M&E System (PDO PRORU- beneficiary producer organiza-   200% of end-of-project target indicator) RAL tions 84 Out- come/ Control Impact Project Key finding Data source group Additional information 39% increase in agricultural 87% of end-of-project target (which income of beneficiary produc- M&E System (PDO   had been revised from 54% to 45% ers compared to start of project indicator) increase) baseline 2014 Impact Evaluation 28% to 37% higher agricultural Bolivia based on M&E system income for project beneficiaries Yes   PAR I data (n=6000 house- than control group producers holds) Economic and Financial US$2,382 average increase in Analysis ICR based M&E beneficiary producers’ household     system data (n= 535 income alliances) 90% of end-of-project target of 2,000 households. This estimation is based on field data collected during the ICR mission. The main increases in family income came from the most financially Brazil M&E System (PDO indi- At least 15% increase in house- successful community subprojects: Alto cator) / Economic and No hold income of 1,800 households fish management, cassava process- Solimôes Financial Analysis ICR ing and non-timber forest products Income (Brazil nuts). The Borrower Completion Report estimates that about 1,400 families have benefitted from a 20 percent raise in incomes. 51% of treatment group registered positive changes in real income as a result of project agro-livestock activi- M&E System (restruc- ties, compared to 38% of the control 30% or higher increase in house- tured PDO indicator) / group. In general terms, the treatment hold income for 43% of treatment 2014 Evaluation (n=237 Yes group had overall income growth of group, compared to 35% in the treatment households, 73% compared to 67% in the control control group n=31 control house- group. The data collection was done holds) with households with at least one year Brazil of production following subproject Pará completion. Five non-experimental, randomly Strong variations in stated income selected complementary case studies gains (from very significant of in- were analyzed in 2015 to gain a cremental income increase of one deeper understanding of the effects of 2015 Case Studies (n=5 monthly minimum salary, interme- No interventions on project beneficiaries. cases) diate results of around one-half a Quantitative information collected monthly minimum salary, to low/ during last year of project implemen- very low increases) tation and declaratory and not based on evidence/receipts. 85 Linking Farmers to Markets through Productive Alliances Out- come/ Control Impact Project Key finding Data source group Additional information Despite a PDO indicator on income, M&E System (PDO indi- Average increase of US$986 the lack of baseline data prevents cator) / Economic and per household from productive   calculation of a percentage increase. Financial Analysis ICR subprojects The increase refers to the year post-in- (n=1,175 households) vestment. US$990 average increase in M&E System (PDO indi- Brazil capital cator) / Economic and Sergipe     assets per household from pro- Financial Analysis ICR ductive subprojects (n=1,175 households) 78.6% of surveyed beneficiaries 2012 Final Evaluation state that subprojects created (n=224 beneficiaries and Interviews were held in seven of the No opportunities for productive activ- 33 Municipal Council eight State territories. ities and income (qualitative) members) Equivalent to 77% average increase US$280 increase in annual incre- Economic and Finan- in producer household net income mental beneficiary household net cial Analysis ICR (n= 23   compared to without-project situation. Colombia income alliances) Results have high degree of variability Income PAAP I across alliances. 12% to 32% increase in total 2008 Impact Evaluation Methodology not considered robust net income of direct beneficiary No (n= 17 alliances) and based on a small sample. households 29% average increase in net 2015 Impact Evaluation The results on net income hold on a income of beneficiary households (n=2380 households: significance level of 5%; the results on Colombia compared to the control group; 899 beneficiary house- Yes gross income on a 10% significance PAAP II 20% average increase in gross holds and 1,481 control level. The controls include nearby and income of beneficiary households group households) distant control groups. compared to the control group High variance in sub-sector net Net income increases were analyzed income increases (ranging from Economic and Finan- by sub-sector, but data no overall Guatema- $489 for wood to US$-368 for ba- cial Analysis ICR (n= 39   average presented in ICR. The fiscal la PDER sic grains). Positive net increases alliances) impact analysis considered a 30% for 6 out of 7 analyzed sub-sec- income tax on incremental revenues. tors. 43% of beneficiary producers had a Panama 69% average incremental increase net profit compared to 33% in the External final evaluation PRORU- in profits from US$698/producer/ No without-project scenario. The most (n=2,439 households) RAL year to US$1,180 profitable crops were milk, plantain, fish, corn, and beans. 38% of end-of-project target. Indicator Bolivia M&E System (PDO 308,506 person-days generated   did not measure family labor or up- Employment PAR I indicator) and downstream employment effects 122 full-time jobs generated Brazil in 2014, mainly in fisheries and Economic and Finan- The full-time jobs generation was Alto Brazil nuts sectors. In cassava cial Analysis ICR (n= 26 No estimated for the year 2014 only. Solimôes processing subprojects, 12 jobs subprojects) were reduced. Brazil No specific results on employ-       Pará ment presented 86 Out- come/ Control Impact Project Key finding Data source group Additional information Economic and Finan- Brazil 4 person-days reduction as incre- cial Analysis ICR (n= 23     Sergipe mental annual labor use subprojects) Equivalent to a 70% increase com- 201 person-days increase in em- Economic and Finan- Colombia pared to without-project situation. ployment per household (i.e. 0.8 cial Analysis ICR (n= 23   PAAP I One person-year means 260 per- person-years per household) alliances) son-days. Economic and Finan- Employment 10,444 new jobs generated cial Analysis ICR (n= 56     alliances) Colombia Employment refers to at least half- PAAP II Increase in employment from an time jobs. The three most common 2015 Impact Evaluation average 1.7 persons to 2.8 per- Yes types of new jobs created are admin- (n=2380 households) sons per producer organization istrative manager, accountant, and technical support. Increase in permanent employ- Permanent employment created Economic and Finan- Guatema- ment by on average 1 per- mostly in accounting. High variability cial Analysis ICR (n= 39   la PDER son-year per producer organiza- in seasonal jobs created in various alliances) tion. sub-sectors. Panama No specific analysis on employ- PRORU-       ment RAL Bolivia No specific analysis on spillovers       PAR I Brazil Alto No specific analysis on spillovers       Solimôes Brazil No specific analysis on employ-       Pará ment Brazil No specific analysis on spillovers       Spillovers Sergipe Economic and Finan- Colombia 6% increase in total net income of cial Analysis ICR (n= 23     PAAP I indirect beneficiary households alliances) 24.4% higher gross income These spillovers were estimated based Colombia found for “nearby” control group 2015 Impact Evaluation Yes on the difference between the nearby PAAP II producers compared to “distant” (n=2380 households) and distant control groups. control group producers. Guatema- No specific analysis on spillovers       la PDER Panama PRORU- No specific analysis on spillovers       RAL 87 Linking Farmers to Markets through Productive Alliances 18. These figures are the total number of vertical alliances/subprojects established during project implementation that were still operating at project completion. For example, in Colombia 744 had been established with 46,362 direct beneficiaries, but 725 were still operating at project completion. Similarly, in Guatemala, a total of 216 alliances had been established(189 Business Plans and 27 Business Strengthening Plans), but 15 BP were cancelled during project, resulting in a total 174 alliances funded 19. The PAD was inconsistent in defining the number of target beneficiary households in its Results Framework, referring to 36,000 as well as 25,000 households. The PDO indicator referred to 36,000 households. 20. The Brazil Sergipe project implemented “socio-economic infrastructure” subprojects, which included productive market- linked investments with technical assistance support, next to other types of subprojects. 1000 (restructured to 500) of these socio-economic subprojects for 20,000 households were the appraisal targets. At completion, 547 subprojects had been implemented in total, benefiting 22,006 households. For this assessment only results for productive subprojects are reported. 21. The original target of 60 alliance subprojects was based on a US$500,000 ceiling per subproject. This ceiling was reduced to US$ 250,000 in 2012, which automatically doubled the possible number of alliances, resulting in an actual financing of 130 alliances. 22. The Brazil Alto Solimôes project was implemented with producers through community associations. Hence, no information on producer organization membership can be provided, but instead the average number of beneficiary households per subproject. SCOPE: Economic and Financial Analysis of PAAP financed (40.7% of the total investments II indicate that the bulk of the co-financing in productive subprojects), followed by Regarding the individual projects resulted from the beneficiary producers, coffee (28.5%), and basic grains (6.5%). The analyzed, Colombia’s PAAP I reached mainly through in kind contributions of Panama PRORURAL project reached 4,577 11,714 producers through the establishment assets and family labor. Grants for other producers (91% of the appraisal target) of 136 alliance subprojects, surpassing sources were typically in the order of 10% to through 130 productive alliances. The the appraisal targets by 17% and 30%, 20% of total subproject investment costs. project provided the highest grant support respectively. In terms of distribution of per beneficiary household (US$4,187) products supported, 76% of alliance In Bolivia, PAR I reached 28,527 direct and the second highest grant support per subprojects were related to crops (mainly beneficiary households (84% of the end- alliance (US$147,379). Investments were perennial crops), 14% to livestock, and the of-project target) through 768 partnerships concentrated in the following productive remaining 10% to aquaculture and forestry. (114% of the target). Nine products were chains: dairy and cattle products, maize, The average IBRD grant amount was the most largely supported by the project roots and tubers, traditional fishing, US$1,759 per beneficiary household and US and accounted for 72% of investment costs beats, and others (e.g. sugarcane, $151,507 per alliance. The follow-on project (535 subprojects) and about 70% of the apiculture, fruits, small livestock, etc.). PAAP II had a much wider scope, with more project’s direct beneficiaries. These were than 42,000 direct beneficiaries in 725 livestock (49%), quinoa (16%), coffee (9%), In Brazil, the Alto Solimôes project productive alliances at project completion and a mix of others (pig breeding, cocoa, reached 3,252 households (90% of the (167% and 242% of the appraisal targets, potato seed, peach and sesame, 26%). end-of-project target) through seven respectively). At the same time, compared The average grant support given by PAR subprojects and nineteen micro-projects. Of to the first phase, the grant resources from I amounted to US$2,112 per beneficiary the Project’s direct productive investment IBRD declined significantly to US$514 per household and US$78,449 per alliance. of US$2,443,383, 90% percent was invested beneficiary household and US$30,370 per In Guatemala, the PDER project was in the seven subprojects and the remaining alliance. It is important to note, however, expected to reach 30,000 beneficiaries 10% in the nineteen micro-projects. The that these figures do not include grants from through 200 productive alliances. At the former were concentrated in inland fisheries other public sources (as no comprehensive end of the project, only 18,115 beneficiaries (4 subprojects, representing about 40% of data on these sources is available). Thus, (60%) were reached through 174 alliances. PA investments), Brazil nuts (1 subproject, the overall grant amount (IBRD and other) This lower than expected achievement in representing 28% of PA investments) as well received by beneficiaries and alliances Guatemala might be partially explained as fingerlings and indigenous bees honey under PAAP II is higher than the figures by the complex project design and processing. Micro-projects focused mainly displayed in Table 8.2. The fact that the implementation arrangements, which linked on cassava processing (9 micro-projects, grants financed by the project dropped business plan subprojects to broader local representing about 5% of PA investments). significantly under PAAP II indicates that public investments in various sectors, which According to the ex-post Economic and PAAP II resources were leveraged to a required coordinated implementation Financial Analysis, the average investment much larger degree through other funding strategies between several line ministries. per beneficiary household was about sources, allowing for a strong increase While the PDER invested in 13 products, US$765. Although on the lower side in project scope. The findings from the fruits and vegetable were by far the most compared to other PA projects, according 88 to the ICR this “could be considered as a main products and geographic location. not sufficient to cover total costs when reasonable amount given the characteristics The FIRR was estimated at 29% when only accounting for the time value of money. of the areas (isolated and logistically considering direct costs. When indirect The EFA for Brazil Alto Solimôes was challenging) and the type of beneficiaries costs were included, the FIRR decreased done individually for each component, (small farmers with low levels of market to 27%. To test the robustness of these which allowed for an analysis of the integration)” (ICR Brazil Alto Solimôes, results in the face of possible adverse efficiency of the productive alliance 2015). In the Brazilian State of Pará, the technical, climate-, or market-related events, activities. Given the lack of representative PA project reached 3,148 households (9% net incremental income was reduced data in the M&E system, the main source of the initial appraisal target of 36,000 by 20%. Under this scenario, the FIRR of information for this analysis was the which was restructured to 4,067 households decreased further to 25%, still a robust collection of data carried out by the ICR during implementation). The project result. Average FIRR range from a high mission directly from the management of financed the preparation of subprojects, of of 38% for plantains to 18% for rubber. the communities benefitting from productive which 41 were actually implemented. The The EFA for the Bolivia PAR I project was subprojects and micro-projects only for the investment per subproject and beneficiary based on data from the project M&E system year 2014. Another complementary source household is high in comparison to other for a random sample of 535 productive was the partial assessment reports prepared PA projects. The main products financed alliances. These represented the nine major by the PCU and consultants reports. As through subprojects were (i) agroforestry, (ii) products supported, equivalent to 70% of in the other projects, the financial results honey processing, (iii) cassava processing, direct beneficiaries and 72% of investment were highly heterogeneous, showing (iv) fruit processing, (v) an ice factory, (vi) in alliances (dairy, quinoa, coffee beef, pigs, FIRR above 24%percent for subprojects poultry and (vii) aquaculture, representing cocoa, potato seeds, peaches, sesame). It supporting fishing activities, an average 85% of subprojects and around 84% of is the only EFA conducted without primary FIRR of 21% for micro-projects financing the total investment in subprojects. In data collection. The average FIRR was found cassava processing units, and FIRR below the State of Sergipe, 247 productive to be 35% (including direct and indirect the reference rate of 12% for the remaining subprojects directly benefited 10,800 investments costs). The products with the subproject investments (honey from native households. The most common activities higher performance were quinoa and coffee, bees, Brazil nuts and aquaculture). The supported by the subprojects are fishing, followed by potato seed production and incremental net income margin for 2014 cattle, goat production and agricultural peaches. The EFA of Guatemala’s PDER varied across products and subproject, mechanization (particularly the procurement project was based on a random sample of with an average of about US$350 per of tractors). The average production 39 productive alliances. The aggregated household for inland fisheries subprojects, investment support given by the project IRR on direct investment costs was 20%, US$250 for Brazil nuts, US$320 for cassava amounted to US$930 per beneficiary and considering project management processing micro-projects and only US$30 household and US$40,500 per subproject. and monitoring costs proportionate for indigenous bees honey production. to the investment is 17%. Similar to EFFICIENCY: Colombia PAAP II, a 20% reduction of The EFA of Brazil Pará is based on a Regarding specific projects, the EFA for cost-benefit analysis conducted for 10 net incremental income was calculated to the Colombia PAAP I project was based on illustrative cases out of seven economic test the robustness of the results against a random sample of 23 productive alliances. activities, representing about 84% of the adverse technical, climate, or market-related Of these, 69% were found to have a FIRR total investment in subprojects. A weakness events. As a result, the FIRR decreased to higher than 12%, and 31% lower than of this is analysis is that the illustrative cases 12%. The EFA of the Panama PRORURAL 12%. Only 13% had a negative return. As are not representative of all subproject project is based on 12 randomly selected in 2008 the rainfall in the project areas was activities due to the heterogeneous level alliances, a smallest sample due to time significantly higher than average, affecting of maturity of each activity financed and in and resource limitations and hence not productivity of annual and perennial crops, their size and characteristics. Instead, the representative of the sampling universe the returns can be expected to be higher in results should be interpreted as viability with a high degree of statistical confidence. the future under normal rainfall conditions. estimates of each product. Three scenarios The average IRR was 11%. However, The EFA for the Colombia PAAP II was were analyzed: the current situation, the about one third of alliances resulted in based on a sample of 56 alliances selected minimum scenario and a best case scenario. negative financial feasibility indicators. In through a two-stage process stratified by The financial analysis shows that if current other words, the generated revenues were 89 Linking Farmers to Markets through Productive Alliances bottlenecks are not solved (i.e. current survey collected information only from The beneficiaries were then matched with scenario), only four of the 10 cases would people who were already effectively similar non-beneficiaries in accordance attain an FIRR equal or above 12% and receiving support from the project. with the Mahalanobis distance, having positive net margin (in agroforestry and taken into account a set of pre-treatment, General sampling strategy. The baseline aquaculture). The minimum scenario, in observable characteristics. In order to survey for the follow-up project (PAR II) which the existing bottlenecks are assumed evaluate the robustness of the results, a was used to estimate the counterfactual to be overcome to attain both positive Propensity Score Matching was also carried outcomes of the project (PAR). A PAR net margin and NPV, is deemed likely to out even though the treated and non- II baseline survey was carried out with be achieved by four productions systems treated groups did not participate in the producers who had applied to the first (agroforestry, honey, ice, aquaculture), same selection process. Several variables call for proposals and who had passed while uncertainty about the resolution were included in this matching process, the pre-feasibility filter (“opportunity of existing problems remains for the comprising characteristics related to the evaluation”). This survey of potential remaining cases (cassava processing, producer, his/her household and family, as PAR II beneficiaries was used to build fruit pulp and poultry) is significant. collected through the survey: sex, age, years a counterfactual control group for the of education, number of family members, The EFA of the productive investments project’s (PAR) impact evaluation, as migration status, ethnicity, department done in Brazil Sergipe is based on a simple those surveyed were considered by the of residence, access to basic services of 23 subprojects. According to the ICR, project but had yet to receive support, (electricity, water and sanitation), use of these are proportionally representative thus eliminating selection bias. A sample combustibles to cook, number of rooms of the most common subproject types: of PAR beneficiaries was surveyed using per household members and household (i) fishing equipment (8% of productive the same questionnaire employed for the construction materials. Non-significant investments), cattle and goat production PAR II baseline to allow for comparison. variables at 10% of significance within support (8%) and agricultural mechanization the binary model were excluded from the The survey was designed to show impact (46%, mainly tractors). The financial results matching processes. Although the analysis in four intervention areas. Given that it were heterogeneous across products, with uses the ‘5-nearest neighbors’ procedure, is impossible to know which producer an average FIRR of 39%. However, the EFA summarized results of ‘nearest neighbor’ organization will effectively receive PAR states that 8-13% of the 23 subprojects and ‘3-nearest neighbors’ are also reported II support, the project’s baseline survey showed an FIRR below 10%, implying that to show the robustness of the results in includes a random sample of 3,824 56-35% of subprojects generated marginal accordance with different specifications. producers pulled from all organizations or negative financial returns, depending which had passed the pre-feasibility phase. on assumptions of reduced benefit. The matching allowed for a re-weighting In accordance with the trend observed in of the non-beneficiary group to bring it PAR I, it is expected that about half of these IMPACT EVALUATION organizations will receive project support. closer to the beneficiary group in order to reduce the selection bias caused by METHODOLOGIES Simultaneously, a sample of 2,142 PAR observable characteristics. These weights I beneficiaries was surveyed in the four Bolivia PAR I regions where both PAR and PAR II have had were used to estimate the counterfactual outcomes which were compared to The Bolivia PAR I project aimed to activities (Norte, Valle, Chaco, Trópico). This the outcomes effectively attained by implement an impact evaluation since its sample was designed to ensure significant beneficiaries. In order to properly evaluate inception. Although two baseline surveys results in each region so as to evaluate PAR’s the project’s impact,, four income variables were carried out to inform project design regional impact given that the intervention were assessed, including: (i) agricultural and execution, they could not in the end areas are varied in terms of agro-ecological, sales that include the market value of be used for the impact evaluation as they cultural, marketing and socio-economic crops, cattle and derivative productions; (ii) did not allow for the adequate identification settings. The samples were extracted from total agricultural income that includes the of treated and non-treated groups. The the project’s information system (SIGG). market value of sales and the estimated 2006 survey collected information on local value of subsistence production; (iii) net producers within the project intervention Methodologies used for analyses. agricultural income that subtracts reported areas but did not incorporate sufficient PAR beneficiaries were matched with costs from total agricultural income; and, information to identify the surveyed producers who had applied to PAR II’s first (iv) total household labor income that farmers for follow-up. Similarly, the 2008 call and passed the pre-feasibility phase. 90 includes all beneficiary labor income after participation (e.g., the criteria defined mirrored the proportion in the population. taxes and cost resulting from primary in the calls for proposals, such as social, It is important to note that the impact and/or secondary activities as salaried or demographic, and economic characteristics, evaluation’s sampling strategy is based self-employed persons of all household but also in terms of agricultural production on project data as of December 31, 2014. members. The agricultural income and size). This similarity between beneficiary As a result, some figures changed slightly variables were estimated using product-by- and control groups is crucial to obtain by the end of the project (e.g., total product information requested during the reliable results which can be attributed to number of partnerships and beneficiaries, survey; total labor income variables were the project, as it controls for self-selection investments made per productive system). calculated using a different set of questions in the impact analysis and enables the However, the overall results of the impact regarding income by labor activity. estimation of spillover effects generated evaluation can be expected to be valid. by the project. The sampling strategy The sampling strategy resulted in a Colombia PAAP II was designed to be able to estimate final sample of 899 beneficiary producers statistically significant differences in An impact evaluation of the PAAP II from 83 PAAP-II partnerships located in 86 average results between treatment and project was conducted in the first half of municipalities. The distribution of those control at a 95% confidence level. 2015. To estimate the project’s intermediary beneficiary producers across the product results and impacts, an independent firm Sampling of beneficiary partnerships strata is displayed in the table below. implemented structured household surveys and producers (treatment group). To select with 899 beneficiary producers and 1,481 Beneficiary the beneficiary producers to be interviewed comparable control group producers. The producers for the impact evaluation, in a first step household survey questionnaire modules partnerships of the project were selected Product interviewed captured main socio-economic household Cocoa 134 according to the following criteria: (i) had and production characteristics to assess Specialty coffee 148 to have finalized all investments financed the project’s effect on intermediate results, through the government financial incentive Dairy cattle 153 PDO outcomes, as well as longer-term grant provided through PAAP-II; (ii) had Blackberry 139 impacts. The sampling strategies to select to have passed at least one production Other (short-cycled beneficiary producers and control group 325 cycle; and, (iii) had to have passed one fruits) producers for the evaluation, as well as cycle of sales. At the time of the impact Total 899 the quasi-experimental methodologies evaluation design, these criteria resulted used for the intermediate results and in a population of 202 partnerships for the Sampling of control group producers. impact analyses are described below. analysis. Furthermore, it is important to As mentioned before, the sampling design More details on the impact evaluation are note that the impact evaluation sample is for the impact evaluation included two presented in the project’s Implementation stratified on the main productive systems control groups, “distant” and “nearby”. The Completion and Results Report. supported through the government financial distant control group can be considered incentive grant (i.e., specialty coffee, cacao, the control group commonly used in impact General sampling strategy. Three types dairy cattle livestock, blackberry, other analyses. As the producers sampled in of producers were sampled for the impact short-cycle fruits, such as plantain and the distant control group are expected to evaluation: (i) beneficiary producers of others). Within each of these productive have not been influenced at all by project the project, i.e. the “treatment group”; (ii) system strata, simple random sampling activities (given that they are far outside its control group producers who lived in or near was applied to select the partnerships for radius of action), the results found for these the localities of the beneficiary producers, the impact evaluation sample. Of those groups can be considered as the “pure” i.e. the “nearby control group”; and, (iii) selected partnerships, beneficiary producers counterfactual to the beneficiary household control group producers who lived far from were divided in strata based on the size treatment group. In addition, the impact the localities of the beneficiary producers, and amount of the government financial evaluation included a nearby control group i.e. the “distant control group”. To be incentive grant they had received. For the to assess whether the project generated selected into the (nearby and distant) control final impact evaluation sample, beneficiary externalities to non-beneficiaries and group sample, non-beneficiary producers producers within each stratum were thereby created potential spillover effects had to have the same or very similar selected randomly, so that the proportion beyond the project’s direct beneficiaries. characteristics as the beneficiary producers of strata in the impact evaluation sample The measurement of potential positive in terms of the eligibility criteria for project 91 Linking Farmers to Markets through Productive Alliances spillover effects is important, because far from the localities of the beneficiary very similar characteristics as the beneficiary ignoring their existence can lead to an producers (distant control group). The final producers. As a result, propensity score under-estimation of the overall impact control group sample was composed of matching was used to pair beneficiary of a project (e.g., when producers who 582 distant control group producers and producers and control group producers with benefited from spillovers are selected into 899 nearby control group producers. the same or very similar characteristics to the control group). The nearby control reduce confounding biases in the impact group for the impact evaluation of PAAP- Methodologies used for analyses. Due analysis. For the analysis of intermediate II was composed of producers who lived to the lack of representative baseline data results, differences in means in the variables in or near the localities of the beneficiary for the project, the analyses on intermediary of interest between beneficiary producers producers, but did not directly benefit results and impacts used quasi-experimental (i.e., treatment group) and control group from the project. Producers in both control methodologies. Regarding the comparison producers are presented. For estimating the groups were randomly-selected from non- of results for the beneficiary producers to project’s impact, difference-in-differences beneficiary producers who lived in or near the control group producers, it is important analysis was used. Finally, expansion factors the localities of the beneficiary producers to recall that both nearby and distant control were used to make the results representative (nearby control group) and/or who lived group producers had to have the same or of the population of partnerships evaluated. 92 ANNEX 9 THEORY OF CHANGE FOR PRODUCTIVE ALLIANCE PROJECTS Figure A9. 1 Theory of Change Logic Description Long-term, widespread changes in society that result from an accumulation of Impacts outcomes. They typically take place only over the long term beyond the implementation period of an individual project. Behavioral changes that result from the project outputs. They can be both, Outcomes intended and unintended, positive and negative. An outcome cannot be measured directly, but must be translated into a set of measureable indicators. Proucts and services produced / delivered as a result of the activities. Outputs They are generally tangible and can be counted. Activities Tasks undertaken to transform inputs into outputs. Inputs Financial, human, and material resources that go into a project or program. This figure is based on Morra Imas and Rist (2009), Kusek and Rist (2004) and Binnendijk (2000). 93 Medi- Long-term Project De- um-term outcomes velopment outcomes Short-term Outputs Activities Inputs Overall Goall Objective Precondi- outcomes / Impact PDO tions to achieve PDO Increased pro- Poverty Reduc- ductivity and Matching grants tion competitiveness Producers work Trust relation- for productive (e.g. increase in among rural in alliances Implementation ship established Subproject selec- investments, income / assets small-scale pro- (organizational, of subproject among alliance tion finalized technical assis- / consumption ducers through commercial, selection process partners tance, business / access to ser- their participa- technical) development vices) tion in produc- tive alliances Employment Generation Producers adopt Alliance partners Agreement Networking with (e.g. new and / new technolo- comply / imple- Service produc- among im- potential private or more prof- gies and prac- ment business ers active plementation and / or public itable employ- tices plan partners sector buyers ment) Table A9. 2 Prototypical Theory of Change for Productive Alliance Projects Inequality Reduc- tion Producers make Monitoring sys- (e.g. decrease productive in- tems in place of rural-urban vestments disparities) Producers are integrated in markets Buyers fulfill their obligations un- der the business plan 94 Linking Farmers to Markets through Productive Alliances ANNEX 10 SELECTED ISSUES AND POSSIBLE SOLUTIONS Issue Possible Solutions BENEFICIARY ELIGIBILITY AND SELECTION Involve local/municipal institutions in oversight of Beneficiaries selected are not within eli- Beneficiary eligibility and beneficiary eligibility, selection and approval of gibility criteria and/or are not sufficiently selection subprojects, and field-verification where possible committed Beneficiaries form groups without prior relations or social cohesion solely for Establish requirement of pre-existing groups with a Beneficiary group formation purpose of project compliance, which minimum time of operation complicates group decision-making and investment choice Considerable time is spent on subproject Streamline procedures (e.g. through standard for- design, selection and contracting mats, online registration) Strengthen participatory approach in process of subproject proposal preparation. Place greater em- Poor quality of business plans (e.g. long phasis on PO and buyer participation (not relying and detailed documents which say little too heavily on external consultants only) Preparation of subproject about the proposed business, suffer from Consider beneficiary financial contributions to proposals/ business plans lack of data or inconsistency) proposal preparation to enhance ownership and accountability of service provider Design the technological package according to the Beneficiaries receive the same technolog- assets and needs of individual beneficiaries ical package without taking into consider- Oversight of service providers and satisfaction ation differences in assets and capabilities surveys with beneficiaries Include preliminary financial viability during initial subproject proposal preparation and a meaningful financial analysis in the business plans Evaluation of subproject pro- Subprojects are approved without demon- Train service providers preparing business plans posals/ business plans strated financial viability of investments and PCUs staff in assessing financial feasibility of investments Involve financial institutions in preparation, apprais- al and financing of business plans BUYER ELIGIBILITY AND THE COMMERCIAL AGREEMENT Some productive alliances do not estab- Reinforce the buyer’s commitments to the alliance Buyer eligibility and selection lish long-term relations between POs and by spelling them out at appraisal and monitoring buyers them consistently The relationship between POs and buyers are not always reflected in a firm commer- Assess possibility to require a formal commercial Commercial agreement cial agreement (“unconditional” Productive agreement as condition for all subproject financing Alliances) Some producers sell a proportion of their Formally allow more than one buyer as buyer Marketing produce outside of the commercial agree- Allow flexibility in the alliance agreement to re- ments (“side-selling”) spond to changing market demands 95 Linking Farmers to Markets through Productive Alliances Issue Possible Solutions IMPLEMENTATION OF THE PRODUCTIVE ALLIANCE SUBPROJECT Use business plans as flexible management tools Changes in productive alliances can occur and amend them if major adjustments have oc- Business plan implementa- between proposal and implementation curred tions due to lengthy subproject selection and Evaluate subprojects results according to measur- effectiveness procedures able (standardized) results indicators identified in the business plans SUBPROJECT INVESTMENTS Transfer subproject resources to POs only after all Beneficiary’s cash contribu- Not all PO members contribute with cash PO members have made their cash contribution tion to the investment despite former agreement into PO accounts PA project design (PAD) and Operational Manuals generally do not include mech- Buyer’s contributions to the Consider mandatory buyer contribution, i.e. re- anism to stimulate financial contribution partnership quirement to co-invest in productive alliance of buyers to an alliance (some only specify “willingness to contribute”) Some PA projects include establishment Intensify training and awareness-building of POs on Implementation of Revolving of a PO-managed Revolving Fund despite the establishment and management of Revolving Fund POs not having enough experience Funds Strengthen activities to increase access to finan- cial services (e.g. where logical from a business Financial services are not systematically in- perspective, require a percentage of the total Access to (private sector) volved in co-financing alliances subprojects subproject cost to be provided through a loan from financial services a financial service provider using its own funds) Engage with financial service providers from early on in the project preparation MONITORING AND EVALUATION OF PRODUCTIVE ALLIANCES Strengthen capacities of project officers and consul- tants who accompany productive alliances agents in recording key data (e.g. production/marketing costs and incomes, hired and non-paid labor). Also, training of producer groups in record-keeping of data appropriate to the particular economic activity Projects do not adequately collect the data financed (e.g. costs, expenditures, yields). Ensure Measuring the effects of the need for Economic and Financial Analyses data collection at key points in time: (1) ex-ante, subproject investment on the at project completion (e.g. to calculate the (2) ex-post at the end of the subproject and ideally productive system and on returns on investment and effects on the (3) one or two years after completion (for a random beneficiary income supported products) sub-sample of subprojects and products) Carry out periodic quality controls of the subproj- ect-level data base Ensure a common use of parameters, which would allow aggregation and comparison of key per- formance indicators across subprojects (and PA projects) 96 Issue Possible Solutions SUSTAINABILITY OF PRODUCTIVE ALLIANCES In most PA projects, a large proportion of Launch more frequent but smaller calls for propos- productive alliances start effective execu- als to reduce the PCU’s workload and ensure con- tion during the last year of overall project tinuous implementation and effective monitoring implementation. As a result, many of Consider increasing overall project implementation alliances are immature by the time the final duration (e.g. from the typical 5 to 7 years) project evaluation The participation of decentralized local institutions has been limited in most PA PA projects should become process facilitators projects. As a result, after project support instead of solely executors, by networking from ends, there is a lack of support by local early on with institutional actors (especially local institutions to (partly) continue with support governments) and main stakeholders to encourage (e.g. technical assistance provision, moni- stronger involvements of these actors toring) 97 Linking Farmers to Markets through Productive Alliances REFERENCES GENERAL »» Berdégué, J. A. and Fuentalba, R. (2011). Latin America: The State of Smallholders in Agriculture. IFAD Conference on New Directions for Smallholder Agriculture. »» Binnendijk, A. (2000). Results-Based Management in the Development Cooperation Agencies: A Review of Experience. Paper prepared for the OECD/DAC Working Party on Aid Evaluation, Paris, France. »» Collion, M.H. (2012). Rural Productive alliances. An Inclusive Agribusiness Model for Overcoming Small-holder Market Barriers. Draft paper (unpublished) »» Collion, M.H. and Friedman, M. (2011). Productive Alliances: A Model for Overcoming Market Barriers. World Bank Innovative Activity Profile 6. »» De Salvo, P. (2014). Productive Alliances in Latin America and the Caribbean. A Review. FAO Agro-noticias. »» Goldstein, M. (2016). What works for improving welfare in agriculture: version 0.001. World Bank Development Impact Blog 03/09/2016: http://blogs.worldbank.org/impactevaluations/what-works-improving-welfare-agriculture-version-0001 »» IEG. (2011). Impact Evaluations in Agriculture: An Assessment of the Evidence. Independent Evaluation Group. Washington, DC, USA. »» Jaffee, S., Henson, S. and Diaz Rios, L. (2011). Making the Grade: Smallholder Farmers, Emerging Standards, and Development Assistance Programs in Africa. A Research Program Synthesis. World Bank Report No. 2324-AFR. Washington DC, USA. »» Kusek, J.Z. and Rist, R.C (2004). Ten Steps to a Results-Based Monitoring and Evaluation System. World Bank. Washington, DC. »» Morra Imas, L. and Rist, R.C. (2009). The Road to Results: Designing and Conducting Effective Development Evaluations. World Bank. Washington DC, USA. »» Reardon, T. and Gulati, A. (2008). The Supermarket Revolution in Developing Countries. Policies for “Competitiveness with Inclusiveness”. IFPRI Policy Brief 2. »» World Bank. (2015). Ending Poverty and Hunger by 2030. Washington DC, USA. BOLIVIA RURAL ALLIANCES PROJECT I & II (PAR I & PAR II) »» EMPODERAR Program, PAR, Ministry of Rural Development and Land (MDRyT) (2013). Operational Manual. PAR II. »» EMPODERAR Program, PAR, Ministry of Rural Development and Land (MDRyT). Operational Manual (2009) PAR I. »» World Bank. (2015). Implementation Status & Results PAR II (Mid-Term Review). Washington, DC »» World Bank. (2014). Implementation Completion and Results Report (ICR): PAR I. Report No: ICR00003388. Washington, DC »» World Bank. (2013). Implementation Status & Results Report: PAR I. Report No: ISR12872. Washington, DC »» World Bank. (2012). Project Appraisal Document (PAD): PAR II. Report No: 71702-BO. Washington, DC »» World Bank. (2009). Proposed Additional Financing Credit: PAR I. Report No: 47478 – BO. Washington, DC »» World Bank. (2005). Project Appraisal Document (PAD): PAR I. Report No: 32080 – BO. Washington, DC 98 BRAZIL ALTO SOLIMÔES BASIC SERVICES AND SUSTAINABLE DEVELOPMENT PROJECT »» World Bank. (2007). Project Appraisal Document. »» World Bank. (2015). Implementation Status Reports. »» World Bank. (2015). Implementation Completion and Results Report. BRAZIL BAHIA STATE INTEGRATED PROJECT RURAL POVERTY AND SUSTAINABLE RURAL DEVELOPMENT PROJECT »» World Bank. (2005). Project Appraisal Document. »» World Bank. (2014). Implementation Completion and Results Report. »» World Bank. (2014). Project Appraisal Document. »» World Bank. (2016). Implementation Status Reports. BRAZIL CEARÁ RURAL SUSTAINABLE DEVELOPMENT AND COMPETITIVENESS PROJECT »» World Bank. (2012). Project Appraisal Document. »» World Bank. (2014). Mid-Term Review Report. »» World Bank. (2016). Implementation Status Reports. »» Brazil Pará Integrated Rural Development Project »» World Bank. (2006). Project Appraisal Document. »» World Bank. (2015). Implementation Status Reports. »» World Bank. (2015). Implementation Completion and Results Report. BRAZIL PARAÍBA SUSTAINABLE RURAL DEVELOPMENT PROJECT »» World Bank. (2016). Project Appraisal Package. »» World Bank. (2016). Implementation Status Reports. BRAZIL PARANA MULTI SECTOR DEVELOPMENT PROJECT »» World Bank. (2009). Project Appraisal Document. »» World Bank. (2016). Implementation Status Reports. BRAZIL PERNAMBUCO RURAL ECONOMIC INCLUSION PROJECT »» World Bank. (2012). Project Appraisal Document. »» World Bank. (2015). Mid-Term Review Report. »» World Bank. (2016). Implementation Status Reports. BRAZIL SÃO PAULO SUSTAINABLE RURAL DEVELOPMENT AND ACCESS TO MARKETS PROJECT »» World Bank. (2010). Project Appraisal Document. »» World Bank. (2016). Implementation Status Reports. 99 Linking Farmers to Markets through Productive Alliances BRAZIL SANTA CATARINA RURAL COMPETITIVENESS PROJECT »» World Bank. (2010). Project Appraisal Document. »» World Bank. (2016). Implementation Status Reports. BRAZIL SERGIPE INTEGRATED PROJECT »» World Bank. (20X). Project Appraisal Document. »» World Bank. (2016). Implementation Status Reports. BRAZIL RIO DE JANEIRO SUSTAINABLE RURAL DEVELOPMENT PROJECT »» World Bank. (2008). Project Appraisal Document. »» World Bank. (2012). Implementation Completion and Results Report. »» World Bank. (2016). Implementation Status Reports. COLOMBIA PRODUCTIVE PARTNERSHIPS PROJECT I & II (PAAP I & PAAP II) »» CIAT. (2014). Productive Partnerships: An effective instrument for the inclusion of vulnerable population groups in diverse contexts of Colombia. International Center for Tropical Agriculture. Cali, Colombia. »» Escobar, G., María Teresa Palacios, M.T., Cruz Doren, M.E., Lizarazo Beltrán, M.O. (2010). Mid Term Review: PAAP II. »» Ministry of Agriculture and Rural Development (2013). Operational Manual. PAAP I »» Ministry of Agriculture and Rural Development (2008). Final Report. PAAP I »» Ministry of Agriculture and Rural Development (2003). Operational Manual. PAAP I »» World Bank. (2015). Implementation Status & Results Report: PAAP II. Washington, DC »» World Bank (2013). Paper on a proposed project modification: PAAP II. Report No: 74863-CO. Washington, DC »» World Bank. (2009). Implementation Completion and Results Report (ICR): PAAP I. Report No: ICR0000860. Washington, DC »» World Bank. (2015). Implementation Completion and Results Report (ICR): PAAP I. Report No: ICR0003038. Washington, DC »» World Bank. (2007). Project Appraisal Document (PAD): PAAP II. Report No: 39472-CO »» World Bank. (2001). Project Appraisal Document (PAD): PAAP I. Report No: 23246-CO GUATEMALA PROJECT TO SUPPORT THE RURAL ECONOMIC DEVELOPMENT PROGRAM (PDER) »» Ministry of Economy (2014). Final Completion Report. »» SEGEPLAN, Ministry of Economy, FONAPAZ (2013). Operational Manual. »» World Bank. (2015). Implementation Completion Report: Washington, DC »» World Bank. (2014). Implementation Status & Results. Report No: ISR16955. Washington, DC »» World Bank. (2013).Third Amendment to Loan Agreement: Guatemala Project to Support the Rural Economic Development Program. Washington, DC »» World Bank. (2006). Project Appraisal Document: Report No: 34978-GT. Washington, DC 100 HONDURAS RURAL COMPETITIVENESS PROJECT (COMRURAL) »» Mid Term Review (2013). COMRURAL. Honduras Rural Competitiveness Project. »» República de Honduras Secretaría de Agricultura y Ganadería- COMRURAL (2009). Operational Manual. »» World Bank. (2008). Project Appraisal Document. Report No. 43539 – HN. Washington, DC »» World Bank. (2015). Implementation Status & Results Report. Washington, DC JAMAICA RURAL ECONOMIC DEVELOPMENT INITIATIVE PROJECT (REDI) »» Implementation Support Missions to REDI, Aide-Memoires of Mid-Term and January 2015 »» Jamaica Social Investment Fund- JSIF (2009). Operational Manual Supplement. »» World Bank. (2009). Project Appraisal Document (PAD). Report No: 49392-JM. Washington, DC »» World Bank. (2014). Implementation Status & Results Report. Washington, DC MEXICO SUSTAINABLE PRODUCTION SYSTEMS AND BIODIVERSITY PROJECT (SPSB) »» Corredor Biológico Mesoamérica, SPSB, GEF, World Bank (2013). Operational Manual. »» World Bank. (2015). Implementation Status & Results Report. Washington, DC »» World Bank. (2012). Project Appraisal Document (PAD). Report No: 66664-MX. Washington, DC PANAMA RURAL PRODUCTIVITY PROJECT (PRORURAL) »» Ministry of Agricultural Development. (2015). Final Evaluation Report. »» Ministry of Agricultural Development. (2011). Final Report: Mid-term Review. »» Ministry of Agricultural Development. (2007, revised 2009). Operational Manual. »» World Bank. (2015). Implementation Completion and Results Report (ICR). Report No: ICR03095. Washington, DC. »» World Bank. (2012). Restructuring: Panama Rural Productivity Project. Washington, DC »» World Bank. (2007). Project Appraisal Document (PAD): Panama Rural Productivity Project. Report No: 38056 – PA. Washington, DC PERU SIERRA RURAL DEVELOPMENT PROJECT (ALIADOS) »» ALIADOS (2011). Final Draft Evaluation Report. »» Ministry of Agriculture of Peru - ALIADOS (2008). Operational Manual. »» World Bank. (2015). Implementation Status & Results Report. Washington, DC »» World Bank. (2013). Project Paper on Proposed Additional Loan. Report No: 69275-PE. Washington, DC »» World Bank. (2012). Restructuring paper on a Proposed Project Restructuring of ALIADOS. Report No: 73254-PE. Washington, DC »» World Bank. (2007). Project Appraisal Document (PAD). Report No: 37733-PE. Washington, DC 101 Linking Farmers to Markets through Productive Alliances An Assessment of the World Bank Experience in Latin America November 2016